It was dusk, and I stared glumly at cold rain falling onto steaming rooftop vents outside the clouded window of my husband's hospital room. I was feeling more than a little sorry for myself. Out there, it was a Friday night full of the promise of weekend diversions. In here, it was much like the night before, and the night before thata waiting game.

Waiting to see if Doug would live or die. Waiting to see if he could overcome the terrible malaise that gripped his body and come home. Waiting to see if he would ever be able to move his arms and legs again.

Turning from the window, I found Doug awake. He had only been off the ventilator for a short time and wasn't able to talk. I had just come from work, having been away from him since late morning. It seemed as though there should be plenty of things to tell him, but patter about the office and traffic did not belong in this room, and he'd already heard endlessly that his family and friends were pulling hard for him.

So I held his hand and leaned on the side rail of his bed, getting my face as close to his as I could. We locked eyes and smiled, and words flowed silently between us, just as they had so many times over our 25 years of marriage. God, how I'd missed that!

The nurses had dimmed the lights in the ICU for the night, and though it was far from dark, the room had a nicer ambience than usual. Straightening up, I searched the channels on Doug's TV for something more suitable than CNN. Suddenly, there were Jake and Elwood sauntering into Aretha Franklin's eatery on their mission from God. Hey, Doug. I said, It's The Blues Brothers. Moments later, Aretha was belting out R‐E‐S‐P‐E‐C‐T, and I was gyrating. Doug was doing the only thing he could, swinging his head from side to side in time to the music.

It was just a tiny moment, a vignette unnoticed by anyone but us two in the life of that ICU. But it is the sweetest memory I have of that time. Just days before his death at age 55, the spark that was us had flamed briefly to life.

It was dusk, and I stared glumly at cold rain falling onto steaming rooftop vents outside the clouded window of my husband's hospital room. I was feeling more than a little sorry for myself. Out there, it was a Friday night full of the promise of weekend diversions. In here, it was much like the night before, and the night before thata waiting game.

Waiting to see if Doug would live or die. Waiting to see if he could overcome the terrible malaise that gripped his body and come home. Waiting to see if he would ever be able to move his arms and legs again.

Turning from the window, I found Doug awake. He had only been off the ventilator for a short time and wasn't able to talk. I had just come from work, having been away from him since late morning. It seemed as though there should be plenty of things to tell him, but patter about the office and traffic did not belong in this room, and he'd already heard endlessly that his family and friends were pulling hard for him.

So I held his hand and leaned on the side rail of his bed, getting my face as close to his as I could. We locked eyes and smiled, and words flowed silently between us, just as they had so many times over our 25 years of marriage. God, how I'd missed that!

The nurses had dimmed the lights in the ICU for the night, and though it was far from dark, the room had a nicer ambience than usual. Straightening up, I searched the channels on Doug's TV for something more suitable than CNN. Suddenly, there were Jake and Elwood sauntering into Aretha Franklin's eatery on their mission from God. Hey, Doug. I said, It's The Blues Brothers. Moments later, Aretha was belting out R‐E‐S‐P‐E‐C‐T, and I was gyrating. Doug was doing the only thing he could, swinging his head from side to side in time to the music.

It was just a tiny moment, a vignette unnoticed by anyone but us two in the life of that ICU. But it is the sweetest memory I have of that time. Just days before his death at age 55, the spark that was us had flamed briefly to life.

It was dusk, and I stared glumly at cold rain falling onto steaming rooftop vents outside the clouded window of my husband's hospital room. I was feeling more than a little sorry for myself. Out there, it was a Friday night full of the promise of weekend diversions. In here, it was much like the night before, and the night before thata waiting game.

Waiting to see if Doug would live or die. Waiting to see if he could overcome the terrible malaise that gripped his body and come home. Waiting to see if he would ever be able to move his arms and legs again.

Turning from the window, I found Doug awake. He had only been off the ventilator for a short time and wasn't able to talk. I had just come from work, having been away from him since late morning. It seemed as though there should be plenty of things to tell him, but patter about the office and traffic did not belong in this room, and he'd already heard endlessly that his family and friends were pulling hard for him.

So I held his hand and leaned on the side rail of his bed, getting my face as close to his as I could. We locked eyes and smiled, and words flowed silently between us, just as they had so many times over our 25 years of marriage. God, how I'd missed that!

The nurses had dimmed the lights in the ICU for the night, and though it was far from dark, the room had a nicer ambience than usual. Straightening up, I searched the channels on Doug's TV for something more suitable than CNN. Suddenly, there were Jake and Elwood sauntering into Aretha Franklin's eatery on their mission from God. Hey, Doug. I said, It's The Blues Brothers. Moments later, Aretha was belting out R‐E‐S‐P‐E‐C‐T, and I was gyrating. Doug was doing the only thing he could, swinging his head from side to side in time to the music.

It was just a tiny moment, a vignette unnoticed by anyone but us two in the life of that ICU. But it is the sweetest memory I have of that time. Just days before his death at age 55, the spark that was us had flamed briefly to life.

My dad was a hero. I suppose it's natural that I feel that way, growing up as I did in a small rural town where my father was the only doctor. Once, he was called to attend to a farmer who had climbed down from his combine, stating flatly that he was going to die. After this pronouncement, the farmer sat on the ground, where over the next hour he proceeded to do just that.

With the benefit of my own 30‐year career in medicine, it's easy to opine that this farmer probably had a pulmonary embolism from a lower extremity deep vein thrombosis (hours spent on the seat of a tractor are the rural equivalent of a long plane ride). I'm sure he was experiencing the feeling of impending doom that can signal such an event. What's harder to conjecture is how my father was able to quickly assess the situation, know his limitations, and still have the guts to carry on. There was no dialing 911, no starting of IVs, and no CPR once the chain of events started. Even if the doomed patient had been taken to the hospital, he would likely have died en route. If not, the personnel, equipment, and therapeutics there would not have been much more than those available in the wheat field. My father had all the tools he needed right there at the scene to bring comfort to that poor individual in the final moments of his life. He also had the courage to use them. What a hero!

Other events and memories of my father flash through my mind as I remember those times. Dozens of antibiotics rendered for cold virusesbut what a placebo effect from that big shot in the butt! Tonsillectomies performed right in the office, with a nurse‐anesthetist coming in once a week to render sedation. (That is until one of a set of twins transiently quit breathing, after which all surgeries were moved to the hospital!) A shot of adrenaline, given unsuccessfully, via the intracardiac route to a high school football player who suddenly collapsed on the 30‐yard line while the usually boisterous Friday‐night crowd watched in stunned horror. Countless hours waiting in the car for my dad to make house calls or finish rounds at the local hospital.

There, in that place and at that time, my dad was it. He embodied medical science, such as it was, in our little community. His black bag and bow tie helped complete the image. He did what he could, limited as it may have been, and he loved every minute of it. Sure, he sometimes complained. It was tough when a patient showed up at the back door and interrupted dinner. He didn't much like it when then‐president LBJ tried to socialize medicine with the Medicare Act. Most of all, he hated it when my mother insisted he take a job with regular hours at a VA hospital after he had two heart attacks and a bout with colon cancer. Mostly though, I remember a happy, self‐actualized guy, especially when he was at work.

My dad died when I was a freshman in medical school, so exactly what he thought about those times I really don't know. I do know, however, how much I loved him and how much I wanted to be just like him.

In medical school and later during my residency, I had other heroes. Appropriately, for someone who ended up an internist and hospitalist, most were master diagnosticians. There was J. Willis Hurst, the renowned cardiologist, who I once saw diagnose cardiac sarcoid solely on the basis of a 12‐lead electrocardiogram. And there was Jay Sanford, author of that little book on antibiotics known as the bugs and drugs book, tucked to this day into my lab coat pocket, who I once heard tell of going to war‐torn Vietnam to collect water samples in order to make the diagnosis of babesiosis. Finally, there was Walter McDonald, then chief of medicine at the Portland VA hospital, later executive vice‐president of the American College of Physicians, who, to a third‐year medical student like I was then, seemed omniscient about each case relayed to him at morning report. They all seemed so confident and clever; so dedicated, diligent, and proud. I wanted to be just like them.

Now, as a vice‐president for medical affairs, given the task of improving quality at a large teaching hospital, I herd cats. I recite from memory the embarrassing statistics on medical errors revealed by the Institute of Medicine in 1999. I plead for standardization and strive to eliminate variability in hospital practice. From the evidence, I extract guidelines and implement them via standard order sets. But frequently I look back and wonder.

Would the practice of medicine, where the goal is standardization and lack of variability, appeal to my older heroes? How would the practices of exceeding benchmarks, following pathways, and complying with indicators play to the icons of my past? Would they be satisfied to practice in today's health care environment? Or would they perish the thought if asked to standardize their orders and comply with best‐practice norms? After all, there was nothing normal about these guys! Sure they knew the literature and would be the first to insist that practice be evidenced based, but for them, that was never enough. What made them so attractive was their ability to go beyond what any of us able to read the journals could achieve. These men (and, sadly, most, but not all, were men) treasured autonomy, yearned for diagnostic brilliance, and doggedly pursued therapeutic breakthroughs. They set the standards that mere mortals like me aspired to achieve. They were heroes.

So we must be careful not to stifle genius while promoting compliance. We must not push the standardization of health care to a point where an individual's ability to rise above the pack is limited. We should remember that with decreased variability comes the risk of denying innovation. For keep in mind that improved methods like those of hospital medicine exist because those before us sought a better way. They were able to try, and sometimes fail, to use their intuition and individual street smarts and to take risks for the greater good. To use a tired phrase, they were able to think outside the box. In the name of quality, we must not further limit the confines of that box. We must assure that however much we strive to elevate the norm, we do not restrict those few who set the curve. We must allow for heroes.

I believe our profession can produce those who will carry the banner forward. I already have some new heroesDon Berwick, Peter Pronovost, and Bob Wachter. These individuals have demonstrated the ability to combine the patient‐centered care practiced by my father with the evidenced‐based knowledge and intuitive genius of my academic mentors. They are then able to apply this admixture of competencies to the problems facing health care today such as the deficiencies in patient safety and the inefficiencies of delivery.

New heroes will attract another generation of the best and the brightest, and the cycle will repeat. With careful foresight we can assure that this will happen. To do otherwise is unthinkable. We must have heroes.

My dad was a hero. I suppose it's natural that I feel that way, growing up as I did in a small rural town where my father was the only doctor. Once, he was called to attend to a farmer who had climbed down from his combine, stating flatly that he was going to die. After this pronouncement, the farmer sat on the ground, where over the next hour he proceeded to do just that.

With the benefit of my own 30‐year career in medicine, it's easy to opine that this farmer probably had a pulmonary embolism from a lower extremity deep vein thrombosis (hours spent on the seat of a tractor are the rural equivalent of a long plane ride). I'm sure he was experiencing the feeling of impending doom that can signal such an event. What's harder to conjecture is how my father was able to quickly assess the situation, know his limitations, and still have the guts to carry on. There was no dialing 911, no starting of IVs, and no CPR once the chain of events started. Even if the doomed patient had been taken to the hospital, he would likely have died en route. If not, the personnel, equipment, and therapeutics there would not have been much more than those available in the wheat field. My father had all the tools he needed right there at the scene to bring comfort to that poor individual in the final moments of his life. He also had the courage to use them. What a hero!

Other events and memories of my father flash through my mind as I remember those times. Dozens of antibiotics rendered for cold virusesbut what a placebo effect from that big shot in the butt! Tonsillectomies performed right in the office, with a nurse‐anesthetist coming in once a week to render sedation. (That is until one of a set of twins transiently quit breathing, after which all surgeries were moved to the hospital!) A shot of adrenaline, given unsuccessfully, via the intracardiac route to a high school football player who suddenly collapsed on the 30‐yard line while the usually boisterous Friday‐night crowd watched in stunned horror. Countless hours waiting in the car for my dad to make house calls or finish rounds at the local hospital.

There, in that place and at that time, my dad was it. He embodied medical science, such as it was, in our little community. His black bag and bow tie helped complete the image. He did what he could, limited as it may have been, and he loved every minute of it. Sure, he sometimes complained. It was tough when a patient showed up at the back door and interrupted dinner. He didn't much like it when then‐president LBJ tried to socialize medicine with the Medicare Act. Most of all, he hated it when my mother insisted he take a job with regular hours at a VA hospital after he had two heart attacks and a bout with colon cancer. Mostly though, I remember a happy, self‐actualized guy, especially when he was at work.

My dad died when I was a freshman in medical school, so exactly what he thought about those times I really don't know. I do know, however, how much I loved him and how much I wanted to be just like him.

In medical school and later during my residency, I had other heroes. Appropriately, for someone who ended up an internist and hospitalist, most were master diagnosticians. There was J. Willis Hurst, the renowned cardiologist, who I once saw diagnose cardiac sarcoid solely on the basis of a 12‐lead electrocardiogram. And there was Jay Sanford, author of that little book on antibiotics known as the bugs and drugs book, tucked to this day into my lab coat pocket, who I once heard tell of going to war‐torn Vietnam to collect water samples in order to make the diagnosis of babesiosis. Finally, there was Walter McDonald, then chief of medicine at the Portland VA hospital, later executive vice‐president of the American College of Physicians, who, to a third‐year medical student like I was then, seemed omniscient about each case relayed to him at morning report. They all seemed so confident and clever; so dedicated, diligent, and proud. I wanted to be just like them.

Now, as a vice‐president for medical affairs, given the task of improving quality at a large teaching hospital, I herd cats. I recite from memory the embarrassing statistics on medical errors revealed by the Institute of Medicine in 1999. I plead for standardization and strive to eliminate variability in hospital practice. From the evidence, I extract guidelines and implement them via standard order sets. But frequently I look back and wonder.

Would the practice of medicine, where the goal is standardization and lack of variability, appeal to my older heroes? How would the practices of exceeding benchmarks, following pathways, and complying with indicators play to the icons of my past? Would they be satisfied to practice in today's health care environment? Or would they perish the thought if asked to standardize their orders and comply with best‐practice norms? After all, there was nothing normal about these guys! Sure they knew the literature and would be the first to insist that practice be evidenced based, but for them, that was never enough. What made them so attractive was their ability to go beyond what any of us able to read the journals could achieve. These men (and, sadly, most, but not all, were men) treasured autonomy, yearned for diagnostic brilliance, and doggedly pursued therapeutic breakthroughs. They set the standards that mere mortals like me aspired to achieve. They were heroes.

So we must be careful not to stifle genius while promoting compliance. We must not push the standardization of health care to a point where an individual's ability to rise above the pack is limited. We should remember that with decreased variability comes the risk of denying innovation. For keep in mind that improved methods like those of hospital medicine exist because those before us sought a better way. They were able to try, and sometimes fail, to use their intuition and individual street smarts and to take risks for the greater good. To use a tired phrase, they were able to think outside the box. In the name of quality, we must not further limit the confines of that box. We must assure that however much we strive to elevate the norm, we do not restrict those few who set the curve. We must allow for heroes.

I believe our profession can produce those who will carry the banner forward. I already have some new heroesDon Berwick, Peter Pronovost, and Bob Wachter. These individuals have demonstrated the ability to combine the patient‐centered care practiced by my father with the evidenced‐based knowledge and intuitive genius of my academic mentors. They are then able to apply this admixture of competencies to the problems facing health care today such as the deficiencies in patient safety and the inefficiencies of delivery.

New heroes will attract another generation of the best and the brightest, and the cycle will repeat. With careful foresight we can assure that this will happen. To do otherwise is unthinkable. We must have heroes.

My dad was a hero. I suppose it's natural that I feel that way, growing up as I did in a small rural town where my father was the only doctor. Once, he was called to attend to a farmer who had climbed down from his combine, stating flatly that he was going to die. After this pronouncement, the farmer sat on the ground, where over the next hour he proceeded to do just that.

With the benefit of my own 30‐year career in medicine, it's easy to opine that this farmer probably had a pulmonary embolism from a lower extremity deep vein thrombosis (hours spent on the seat of a tractor are the rural equivalent of a long plane ride). I'm sure he was experiencing the feeling of impending doom that can signal such an event. What's harder to conjecture is how my father was able to quickly assess the situation, know his limitations, and still have the guts to carry on. There was no dialing 911, no starting of IVs, and no CPR once the chain of events started. Even if the doomed patient had been taken to the hospital, he would likely have died en route. If not, the personnel, equipment, and therapeutics there would not have been much more than those available in the wheat field. My father had all the tools he needed right there at the scene to bring comfort to that poor individual in the final moments of his life. He also had the courage to use them. What a hero!

Other events and memories of my father flash through my mind as I remember those times. Dozens of antibiotics rendered for cold virusesbut what a placebo effect from that big shot in the butt! Tonsillectomies performed right in the office, with a nurse‐anesthetist coming in once a week to render sedation. (That is until one of a set of twins transiently quit breathing, after which all surgeries were moved to the hospital!) A shot of adrenaline, given unsuccessfully, via the intracardiac route to a high school football player who suddenly collapsed on the 30‐yard line while the usually boisterous Friday‐night crowd watched in stunned horror. Countless hours waiting in the car for my dad to make house calls or finish rounds at the local hospital.

There, in that place and at that time, my dad was it. He embodied medical science, such as it was, in our little community. His black bag and bow tie helped complete the image. He did what he could, limited as it may have been, and he loved every minute of it. Sure, he sometimes complained. It was tough when a patient showed up at the back door and interrupted dinner. He didn't much like it when then‐president LBJ tried to socialize medicine with the Medicare Act. Most of all, he hated it when my mother insisted he take a job with regular hours at a VA hospital after he had two heart attacks and a bout with colon cancer. Mostly though, I remember a happy, self‐actualized guy, especially when he was at work.

My dad died when I was a freshman in medical school, so exactly what he thought about those times I really don't know. I do know, however, how much I loved him and how much I wanted to be just like him.

In medical school and later during my residency, I had other heroes. Appropriately, for someone who ended up an internist and hospitalist, most were master diagnosticians. There was J. Willis Hurst, the renowned cardiologist, who I once saw diagnose cardiac sarcoid solely on the basis of a 12‐lead electrocardiogram. And there was Jay Sanford, author of that little book on antibiotics known as the bugs and drugs book, tucked to this day into my lab coat pocket, who I once heard tell of going to war‐torn Vietnam to collect water samples in order to make the diagnosis of babesiosis. Finally, there was Walter McDonald, then chief of medicine at the Portland VA hospital, later executive vice‐president of the American College of Physicians, who, to a third‐year medical student like I was then, seemed omniscient about each case relayed to him at morning report. They all seemed so confident and clever; so dedicated, diligent, and proud. I wanted to be just like them.

Now, as a vice‐president for medical affairs, given the task of improving quality at a large teaching hospital, I herd cats. I recite from memory the embarrassing statistics on medical errors revealed by the Institute of Medicine in 1999. I plead for standardization and strive to eliminate variability in hospital practice. From the evidence, I extract guidelines and implement them via standard order sets. But frequently I look back and wonder.

Would the practice of medicine, where the goal is standardization and lack of variability, appeal to my older heroes? How would the practices of exceeding benchmarks, following pathways, and complying with indicators play to the icons of my past? Would they be satisfied to practice in today's health care environment? Or would they perish the thought if asked to standardize their orders and comply with best‐practice norms? After all, there was nothing normal about these guys! Sure they knew the literature and would be the first to insist that practice be evidenced based, but for them, that was never enough. What made them so attractive was their ability to go beyond what any of us able to read the journals could achieve. These men (and, sadly, most, but not all, were men) treasured autonomy, yearned for diagnostic brilliance, and doggedly pursued therapeutic breakthroughs. They set the standards that mere mortals like me aspired to achieve. They were heroes.

So we must be careful not to stifle genius while promoting compliance. We must not push the standardization of health care to a point where an individual's ability to rise above the pack is limited. We should remember that with decreased variability comes the risk of denying innovation. For keep in mind that improved methods like those of hospital medicine exist because those before us sought a better way. They were able to try, and sometimes fail, to use their intuition and individual street smarts and to take risks for the greater good. To use a tired phrase, they were able to think outside the box. In the name of quality, we must not further limit the confines of that box. We must assure that however much we strive to elevate the norm, we do not restrict those few who set the curve. We must allow for heroes.

I believe our profession can produce those who will carry the banner forward. I already have some new heroesDon Berwick, Peter Pronovost, and Bob Wachter. These individuals have demonstrated the ability to combine the patient‐centered care practiced by my father with the evidenced‐based knowledge and intuitive genius of my academic mentors. They are then able to apply this admixture of competencies to the problems facing health care today such as the deficiencies in patient safety and the inefficiencies of delivery.

New heroes will attract another generation of the best and the brightest, and the cycle will repeat. With careful foresight we can assure that this will happen. To do otherwise is unthinkable. We must have heroes.

A 47‐year‐old woman was brought to the emergency department by her family because of 1 week of abdominal pain. The pain had begun in the epigastrium but had spread across the abdomen. She described it as constant and 10 of 10 in intensity but could not identify aggravating or alleviating factors. She also complained of nausea and vomiting, beginning 4 days prior to presentation, occurring 25 times per day. She noted poor oral intake and mild diarrhea. She denied melena or hematochezia. She reported no recent fever, dysuria, chills, or night sweats; however, she reported upper respiratory symptoms 2 weeks prior to presentation. On the day of presentation, her family felt she was becoming increasingly lethargic.

Epigastric pain in a middle‐aged woman suggests several possible diagnoses. Conditions such as acute cholecystitis begin abruptly, whereas small bowel obstruction, appendicitis, and diverticulitis start gradually. Nausea and vomiting are common concomitants of abdominal pain and are nonspecific. The absence of fever and chills is reassuring. Of greatest concern is the mental status. Initially, I think of enterohemorrhagic E. coli syndromes with associated glomerulonephritis. With a more systemic metabolic abnormality such as this, the rapid development of the disease tends to exaggerate symptoms.

The patient had a history of nephrolithiasis and underwent total abdominal hysterectomy and bilateral salpingo‐oopherectomy secondary to uterine fibroids in the past. She took occasional acetaminophen, smoked two cigarettes per day, and rarely consumed alcohol. Temperature was 38.5C, heart rate was 160 beats/minute, respiratory rate was 28/minute, and blood pressure was 92/52 mm Hg; oxygen saturation was 100% breathing 2 L of oxygen by nasal cannula. She was a moderately obese African American woman in moderate distress, lying in bed moaning. Mucous membranes were dry. There was no lymphadenopathy or thyromegaly. Heart rate was regular without appreciable murmur, rub, or gallop. Lungs were clear. Abdomen was soft and nondistended, with diffuse tenderness to palpation; bowel sounds were present; there was no rebound or guarding. She had normal rectal tone with brown, guaiac‐negative stool. There was no costovertebral angle tenderness. She was oriented to person, place, and time but lethargic; deep tendon reflexes were 3+ bilaterally, and no focal signs were elicited.

Renal stones certainly produce abdominal pain, and the rare patient undergoes laparotomy for this reason. The hysterectomy tells us that small bowel obstruction could be a reason for her symptoms, although abnormal mental status would not be expected without additional problems such as infection. The tachycardia seems out of proportion to her temperature. Hyperpnea and absent respiratory symptoms, along with hypotension and tachycardia, suggest a sepsis syndrome. Her physical exam confirms dehydration. Examination of the abdomen makes me speculate about whether she has a nonsurgical cause of acute abdomen. The lethargy remains unexplained. Sepsis syndrome, possibly from a perinephric abscess, is my leading diagnosis.

White blood cell count was 15.9/mm³ with 78% neutrophils, a hemoglobin of 14.3 g/dL with a MCV of 76 and a platelet count of 320/mm³. Sodium was 159 mmol/L, chloride 128 mmol/L, bicarbonate 19 mmol/L, blood urea nitrogen 120 mmol/L, creatinine 3.1 mg/dL, calcium 11.7 mg/dL, albumin 3.3 g/dL, serum aspartate aminotransferase 65 U/L, serum alanine aminotransferase 72 U/L, total bilirubin 0.7 mg/dL, amylase 137 U/L (normal 30100), and lipase 92 IU/dL (normal 424). Urine obtained from a Foley catheter revealed negative nitrite and leukocyte esterase, 5075 red blood cells, and 1025 white blood cells per high‐powered field.

The elevated serum sodium is likely contributing to her abnormal mental status. It is unusual for a previously healthy and conscious woman to become this hypernatremic because persons with a normal mental status will defend their sodium balance strenuously, assuming regulatory mechanisms are intact. Generally, this level of hypernatremia indicates 2 things. One, a patient was not allowed, or did not seek access to, free water. The other is the presence of diabetes insipidus. It is unlikely she became this dehydrated from the initial gastrointestinal episode as described. The low MCV suggests she may be a thalassemia carrier, as microcytosis with iron deficiency typically does not occur until the patient is anemic, although she may be when rehydrated. Serum calcium, while elevated, also will likely return to the normal range with hydration. The metabolic abnormalities strongly suggest a problem in the central nervous system. The hematuria in the urinalysis continues to raise the possibility of nephrolithiasis as a cause of abdominal pain, though it does not fit well with the rest of the patient's clinical picture. The hematuria and pyuria both could still indicate a urinary tract infection such as pyelonephritis or perinephric abscess causing a sepsis syndrome.

An acute abdominal series and chest radiograph revealed a paucity of gas in the abdomen but no free air under the diaphragm or active cardiopulmonary disease. Abdominal ultrasound showed cholelithiasis without biliary dilation. There was no evidence of hydronephrosis, hydroureter, or perinephric abscess. A noncontrast abdominal‐pelvic computed tomography (CT) scan demonstrated no peripancreatic stranding or fluid collection and no nephrolithiasis or fluid collection suggestive of abscess. The admission electrocardiogram, read as sinus tachycardia with a rate of 160, is displayed in Figure 1.

Figure 1

I have long believed that unexplained sinus tachycardia is one of the most ominous rhythms in clinical medicine; it is expected after vigorous exercise, among other situations, but not in the condition in which this woman finds herself. The nature of the tracing does not indicate the likelihood of a supraventricular arrhythmia, particularly atrial flutter, which should be considered given the rate. The absence of free air under the diaphragm on chest radiography is reassuring. Though the pancreatic enzymes are mildly elevated, they are usually far more striking in gallstone pancreatitis. Hypercalcemia may result in abdominal pain by several mechanisms. I remain concerned about her central nervous system.

The patient was admitted to the intensive care unit (ICU), where she received intravenous antibiotics and aggressive rehydration. The following morning, she continued to complain of abdominal pain. Her systolic blood pressure was 115 mmHg, and her heart rate ranged between 140 and 150 beats/minute. The remainder of her physical exam was unchanged. Repeat laboratory tests revealed a white blood cell count of 14.7/mm³, a blood urea nitrogen of 66 mg/dL, a creatinine of 1.3 mg.dL, amylase of 67 IU/L, and lipase of 70 IU/dL. A contrast‐enhanced abdominal‐pelvic CT scan did not reveal intra‐abdominal pathology. Blood and urine cultures obtained at admission were negative for any growth.

The patient was appropriately admitted to the ICU. When caring for a critically ill patient, establishing a diagnosis is less important initially than addressing treatable conditions with dispatch. The negative CT scans rule out previously entertained diagnoses like nephrolithiasis and perinephric abscess. It is possible that the initially positive urinalysis was a result of urinary catheter placement trauma. Given the course to date, I believe this patient likely has a nonsurgical cause of abdominal pain. I am considering entities such as lead intoxication, hypercalcemia, a tear of the rectus abdominus caused by vomiting, systemic vasculitis, or a hypercoagulable state leading to intra‐abdominal venous thrombosis.

By hospital day 3 her sodium decreased to 149 mmol/L and her creatinine was 1.0 mg/dL. Abdominal pain persisted, unchanged from admission. Her systolic blood pressure had stabilized at 120 mmHg, but the heart rate remained near 150 beats/minute. Her abdomen remained soft and nondistended on exam but diffusely tender to palpation. Her amylase and lipase continued to decrease, and her repeat electrocardiogram demonstrated tachycardia with a rate of 144.

We are gratified to see that her serum sodium has waned but not with the persistence of the tachycardia. It must be assumed that this patient has an infectious disease that we are not clever enough to diagnose at this time. I am also considering an autoimmune process, such as systemic lupus erythematosus. It is difficult to envision a neoplastic disorder causing these problems. The differential remains broad, however, because we have not ruled out metabolic or endocrine causes. It is difficult to imagine she could have Addison's diseasea common cause of severe abdominal pain, tachycardia, and hypotensiongiven her serum sodium level. Hyperthyroidism has been known to produce mild hypercalcemia and abdominal complaints and is an intriguing possibility. The striking elevation of her serum sodium makes me consider the possibility of a problem in the posterior pituitary gland such as sarcoidosis. I cannot explain how sarcoidosis would cause her abdominal pain, unless the hypercalcemia were related. The tachycardia remains of concern, especially if she is otherwise improving. Thus, I would likely administer a small dose of adenosine to ascertain that this is not a different supraventricular tachycardia. In sinus tachycardia, the rate is usually attendant to the clinical picture and thus begs explanation given her clinical improvement.

After receiving 6 mg of intravenous adenosine, the patient's heart rate declined; atrial flutter waves were observed.

This case nicely demonstrates a key teaching point: a fast regular heart rate of about 150, irrespective of the electrocardiogram, suggests atrial flutter. Who gets atrial flutter? Patients with chronic lung disease, myocardial ischemia (albeit rarely), alcohol‐induced cardiomyopathy, and infiltrative cardiac disorders do. Additionally, we also have to consider thyroid dysfunction.

If forced to come up with a single unifying diagnosis at this point, I would have to say this patient most likely has sarcoidosis because this entity would account for modest hypercalcemia, the myocardial conduction disturbance, and hypernatremia because of diabetes insipidus; furthermore, it would fit the patient's demographic profile. However, I am also concerned about hyperthyroidism and would not proceed until thyroid function studies were obtained.

Thyroid studies revealed thyroid stimulating hormone of less than 0.01 mU/L (normal range, 0.305.50), free thyroxine (T4) of 5.81 ng.dL (normal range, 0.731.79), free triiodothyronine (T3) of 15.7 pg/mL (normal range, 2.85.3), and total triiodothyronine (T3) of 218 ng/dL (normal range, 95170). The patient was diagnosed with thyroid crisis and was started on propranolol, propylthiouracil, hydrocortisone, and a saturated solution of potassium iodine. Thyroid stimulating immunoglobulins were obtained and found to be markedly elevated (3.4 TSI index; normal 1.3), suggestive of Grave's disease. Over the next several days, the patient's abdominal pain and tachycardia resolved. Her mental status returned to normal. A workup for her microcytic anemia revealed beta thalassemia trait. The patient was discharged home on hospital day 9 and has done well as an outpatient.

COMMENTARY

As Sir Zachary Cope stated in his classic text Cope's Early Diagnosis of the Acute Abdomen, [I]t is only by thorough history taking and physical examination that one can propound a diagnosis.1 When first presented with a patient whose chief complaint is abdominal pain, physicians tend to focus on the disorders of both the hollow and solid organs of the abdomen as potential sources of the pain. The differential diagnosis traditionally includes disorders such as cholecystitis, peptic ulcer disease, pancreatitis, small bowel obstruction, bowel ischemia or perforation, splenic abscess and infarct, nephrolithiasis, diverticulitis, and appendicitis, all of which were initially considered by the clinicians involved in this case. But as our discussant pointed out, in this case the differential needed to be broadened to include less common disorders, particularly given the patient's altered mental status, numerous electrolyte abnormalities, and lethargy and the lack of explanation provided by the physical examination and sophisticated imaging studies.

Specifically, a myriad of systemic diseases and metabolic derangements can cause abdominal complaints and mimic surgical abdominal disease, including hypercalcemia, acute intermittent porphyria, diabetic ketoacidosis, lead intoxication, familial Mediterranean fever, vasculopathies, adrenal insufficiency, and hyperthyroidism. Unfortunately, the frequency with which abdominal pain occurs in many of these less common disease processes and the pathophysiology that underlies its occurrence are not well defined. For example, abdominal pain is well described as a typical manifestation of both diabetic ketoacidosis and lead poisoning, but the pathophysiology behind its occurrence is poorly understood in both. Further, as a manifestation of thyrotoxicosis and as one of the diagnostic criteria for thyroid storm, the reported prevalence of abdominal pain in this condition is variable, ranging from rare to 20%47%.24 Also, although other gastrointestinal manifestations of hyperthyroidism (such as nausea, vomiting, and hyperdefecation) are thought to be the result of the effect of excess thyroid hormone on gastrointestinal motility, it is unclear whether this similar mechanism is responsible for the perception of abdominal pain.4

An important clue to the underlying diagnosis in this case was the patient's marked tachycardia. Classically, a persistent heart rate of 150 should raise suspicion of atrial flutter with a 2:1 conduction block, as was eventually discovered in this case. Adenosine, in addition to other vagal maneuvers such as carotid massage or Valsalva that also block atrioventricular (AV) node conduction, has been recognized as a safe and effective means of establishing a diagnosis in tachyarrhythmias.5 In AV nodal‐dependent tachycardias, such as AV node reentrant tachycardia or AV reentrant tachycardia, adenosine will often terminate the tachyarrhythmia by blocking the anterograde limb of the reentrant circuit. In AV nodeindependent tachyarrhythmias, such as atrial flutter or atrial fibrillation, adenosine will not terminate the rhythm. However, in the case of flutter, blocking the AV node will usually transiently unmask the underlying P waves, thereby facilitating the diagnosis.5, 6

In this patient, the discovery of atrial flutter was the main clue that thyrotoxicosis may provide the unifying diagnosis. Thyroid hormone has a direct positive cardiac chronotropic effect, resulting in the increased resting heart characteristic of thyrotoxicosis. Specifically, this hormone increases sinoatrial‐node firing, shortens the refractory period of conduction tissue within the heart, and decreases the electrical threshold for atrial excitation. In addition to predisposing to sinus tachycardia (the most common rhythm associated with this disorder), thyrotoxicosis is also associated with atrial tachycardias such as atrial flutter and, more classically, atrial fibrillation.7, 8 Though no studies have specifically evaluated the incidence of atrial flutter in thyrotoxicosis, atrial fibrillation has been found in 9%22% of these patients.7

Finally, several of the patient's electrolyte derangements could explain some of her clinical findings and are clues to the underlying diagnosis. She initially presented with a mild hypercalcemia that persisted even after hydration. Potential explanations include her severe dehydration or her underlying thyrotoxicosis because hypercalcemia is present in up to 20% of patients with hyperthyroidism.9, 10 However, the presence of significant hypercalcemia in the setting of thyrotoxicosis may actually make the diagnosis of thyrotoxicosis more difficult, masking the hypermetabolic signs and symptoms of the hyperthyroid state.11 Interestingly, coexistent primary hyperparathyroidism does occur in a few of these patients, but it likely was not an underlying cause in our patient given that her calcium normalized after receipt of propylthiouracil therapy.12

The patient's marked hypernatremia is more difficult to explain. She may have developed nephrogenic diabetes insipidus secondary to hypercalcemia, explained by a renal concentrating defect that can become evident once the calcium is persistently above 11 mg/dL.13 Combined with her altered mental status, which likely limited her ability to access free water, this may be enough to explain her marked hypernatremia. Her rapid improvement with rehydration is also consistent with this explanation, mediated through the improvement of her serum free calcium.

This case highlights the importance of using all the clinical clues provided by the history, physical exam, and laboratory and imaging studies when generating an initial differential diagnosis, as well as the importance of being willing to appropriately broaden and narrow the list of possibilities as a case evolves. When this patient was initially evaluated by physicians in the emergency department, they believed her symptoms were most consistent with generalized peritonitis that was likely secondary to an infectious or inflammatory intra‐abdominal process such as pancreatitis (especially in light of her mildly elevated lipase and amylase), appendicitis, or diverticulitis. When the medical team in the intensive care unit assumed care of this patient, members of the team failed to recognize several of the early clues, including the patient's markedly abnormal mental status, electrolyte derangements, and persistent tachycardia despite aggressive rehydration, which suggested the possibility of alternative, and less common, etiologies of her abdominal pain. Instead, they continued to aggressively pursue the possibility of the initial differential diagnosis, even repeating some of the previously negative studies from the emergency department. This case illustrates the importance of constantly reevaluating the available information from physical examination and laboratory and imaging studies and not falling victim to intellectual blind spots created by suggested diagnoses by other care providers. Fortunately for this patient, her thyroid crisis was diagnosed, albeit with some delay, before any long‐term complications occurred.

A 47‐year‐old woman was brought to the emergency department by her family because of 1 week of abdominal pain. The pain had begun in the epigastrium but had spread across the abdomen. She described it as constant and 10 of 10 in intensity but could not identify aggravating or alleviating factors. She also complained of nausea and vomiting, beginning 4 days prior to presentation, occurring 25 times per day. She noted poor oral intake and mild diarrhea. She denied melena or hematochezia. She reported no recent fever, dysuria, chills, or night sweats; however, she reported upper respiratory symptoms 2 weeks prior to presentation. On the day of presentation, her family felt she was becoming increasingly lethargic.

Epigastric pain in a middle‐aged woman suggests several possible diagnoses. Conditions such as acute cholecystitis begin abruptly, whereas small bowel obstruction, appendicitis, and diverticulitis start gradually. Nausea and vomiting are common concomitants of abdominal pain and are nonspecific. The absence of fever and chills is reassuring. Of greatest concern is the mental status. Initially, I think of enterohemorrhagic E. coli syndromes with associated glomerulonephritis. With a more systemic metabolic abnormality such as this, the rapid development of the disease tends to exaggerate symptoms.

The patient had a history of nephrolithiasis and underwent total abdominal hysterectomy and bilateral salpingo‐oopherectomy secondary to uterine fibroids in the past. She took occasional acetaminophen, smoked two cigarettes per day, and rarely consumed alcohol. Temperature was 38.5C, heart rate was 160 beats/minute, respiratory rate was 28/minute, and blood pressure was 92/52 mm Hg; oxygen saturation was 100% breathing 2 L of oxygen by nasal cannula. She was a moderately obese African American woman in moderate distress, lying in bed moaning. Mucous membranes were dry. There was no lymphadenopathy or thyromegaly. Heart rate was regular without appreciable murmur, rub, or gallop. Lungs were clear. Abdomen was soft and nondistended, with diffuse tenderness to palpation; bowel sounds were present; there was no rebound or guarding. She had normal rectal tone with brown, guaiac‐negative stool. There was no costovertebral angle tenderness. She was oriented to person, place, and time but lethargic; deep tendon reflexes were 3+ bilaterally, and no focal signs were elicited.

Renal stones certainly produce abdominal pain, and the rare patient undergoes laparotomy for this reason. The hysterectomy tells us that small bowel obstruction could be a reason for her symptoms, although abnormal mental status would not be expected without additional problems such as infection. The tachycardia seems out of proportion to her temperature. Hyperpnea and absent respiratory symptoms, along with hypotension and tachycardia, suggest a sepsis syndrome. Her physical exam confirms dehydration. Examination of the abdomen makes me speculate about whether she has a nonsurgical cause of acute abdomen. The lethargy remains unexplained. Sepsis syndrome, possibly from a perinephric abscess, is my leading diagnosis.

White blood cell count was 15.9/mm³ with 78% neutrophils, a hemoglobin of 14.3 g/dL with a MCV of 76 and a platelet count of 320/mm³. Sodium was 159 mmol/L, chloride 128 mmol/L, bicarbonate 19 mmol/L, blood urea nitrogen 120 mmol/L, creatinine 3.1 mg/dL, calcium 11.7 mg/dL, albumin 3.3 g/dL, serum aspartate aminotransferase 65 U/L, serum alanine aminotransferase 72 U/L, total bilirubin 0.7 mg/dL, amylase 137 U/L (normal 30100), and lipase 92 IU/dL (normal 424). Urine obtained from a Foley catheter revealed negative nitrite and leukocyte esterase, 5075 red blood cells, and 1025 white blood cells per high‐powered field.

The elevated serum sodium is likely contributing to her abnormal mental status. It is unusual for a previously healthy and conscious woman to become this hypernatremic because persons with a normal mental status will defend their sodium balance strenuously, assuming regulatory mechanisms are intact. Generally, this level of hypernatremia indicates 2 things. One, a patient was not allowed, or did not seek access to, free water. The other is the presence of diabetes insipidus. It is unlikely she became this dehydrated from the initial gastrointestinal episode as described. The low MCV suggests she may be a thalassemia carrier, as microcytosis with iron deficiency typically does not occur until the patient is anemic, although she may be when rehydrated. Serum calcium, while elevated, also will likely return to the normal range with hydration. The metabolic abnormalities strongly suggest a problem in the central nervous system. The hematuria in the urinalysis continues to raise the possibility of nephrolithiasis as a cause of abdominal pain, though it does not fit well with the rest of the patient's clinical picture. The hematuria and pyuria both could still indicate a urinary tract infection such as pyelonephritis or perinephric abscess causing a sepsis syndrome.

An acute abdominal series and chest radiograph revealed a paucity of gas in the abdomen but no free air under the diaphragm or active cardiopulmonary disease. Abdominal ultrasound showed cholelithiasis without biliary dilation. There was no evidence of hydronephrosis, hydroureter, or perinephric abscess. A noncontrast abdominal‐pelvic computed tomography (CT) scan demonstrated no peripancreatic stranding or fluid collection and no nephrolithiasis or fluid collection suggestive of abscess. The admission electrocardiogram, read as sinus tachycardia with a rate of 160, is displayed in Figure 1.

Figure 1

I have long believed that unexplained sinus tachycardia is one of the most ominous rhythms in clinical medicine; it is expected after vigorous exercise, among other situations, but not in the condition in which this woman finds herself. The nature of the tracing does not indicate the likelihood of a supraventricular arrhythmia, particularly atrial flutter, which should be considered given the rate. The absence of free air under the diaphragm on chest radiography is reassuring. Though the pancreatic enzymes are mildly elevated, they are usually far more striking in gallstone pancreatitis. Hypercalcemia may result in abdominal pain by several mechanisms. I remain concerned about her central nervous system.

The patient was admitted to the intensive care unit (ICU), where she received intravenous antibiotics and aggressive rehydration. The following morning, she continued to complain of abdominal pain. Her systolic blood pressure was 115 mmHg, and her heart rate ranged between 140 and 150 beats/minute. The remainder of her physical exam was unchanged. Repeat laboratory tests revealed a white blood cell count of 14.7/mm³, a blood urea nitrogen of 66 mg/dL, a creatinine of 1.3 mg.dL, amylase of 67 IU/L, and lipase of 70 IU/dL. A contrast‐enhanced abdominal‐pelvic CT scan did not reveal intra‐abdominal pathology. Blood and urine cultures obtained at admission were negative for any growth.

The patient was appropriately admitted to the ICU. When caring for a critically ill patient, establishing a diagnosis is less important initially than addressing treatable conditions with dispatch. The negative CT scans rule out previously entertained diagnoses like nephrolithiasis and perinephric abscess. It is possible that the initially positive urinalysis was a result of urinary catheter placement trauma. Given the course to date, I believe this patient likely has a nonsurgical cause of abdominal pain. I am considering entities such as lead intoxication, hypercalcemia, a tear of the rectus abdominus caused by vomiting, systemic vasculitis, or a hypercoagulable state leading to intra‐abdominal venous thrombosis.

By hospital day 3 her sodium decreased to 149 mmol/L and her creatinine was 1.0 mg/dL. Abdominal pain persisted, unchanged from admission. Her systolic blood pressure had stabilized at 120 mmHg, but the heart rate remained near 150 beats/minute. Her abdomen remained soft and nondistended on exam but diffusely tender to palpation. Her amylase and lipase continued to decrease, and her repeat electrocardiogram demonstrated tachycardia with a rate of 144.

We are gratified to see that her serum sodium has waned but not with the persistence of the tachycardia. It must be assumed that this patient has an infectious disease that we are not clever enough to diagnose at this time. I am also considering an autoimmune process, such as systemic lupus erythematosus. It is difficult to envision a neoplastic disorder causing these problems. The differential remains broad, however, because we have not ruled out metabolic or endocrine causes. It is difficult to imagine she could have Addison's diseasea common cause of severe abdominal pain, tachycardia, and hypotensiongiven her serum sodium level. Hyperthyroidism has been known to produce mild hypercalcemia and abdominal complaints and is an intriguing possibility. The striking elevation of her serum sodium makes me consider the possibility of a problem in the posterior pituitary gland such as sarcoidosis. I cannot explain how sarcoidosis would cause her abdominal pain, unless the hypercalcemia were related. The tachycardia remains of concern, especially if she is otherwise improving. Thus, I would likely administer a small dose of adenosine to ascertain that this is not a different supraventricular tachycardia. In sinus tachycardia, the rate is usually attendant to the clinical picture and thus begs explanation given her clinical improvement.

After receiving 6 mg of intravenous adenosine, the patient's heart rate declined; atrial flutter waves were observed.

This case nicely demonstrates a key teaching point: a fast regular heart rate of about 150, irrespective of the electrocardiogram, suggests atrial flutter. Who gets atrial flutter? Patients with chronic lung disease, myocardial ischemia (albeit rarely), alcohol‐induced cardiomyopathy, and infiltrative cardiac disorders do. Additionally, we also have to consider thyroid dysfunction.

If forced to come up with a single unifying diagnosis at this point, I would have to say this patient most likely has sarcoidosis because this entity would account for modest hypercalcemia, the myocardial conduction disturbance, and hypernatremia because of diabetes insipidus; furthermore, it would fit the patient's demographic profile. However, I am also concerned about hyperthyroidism and would not proceed until thyroid function studies were obtained.

Thyroid studies revealed thyroid stimulating hormone of less than 0.01 mU/L (normal range, 0.305.50), free thyroxine (T4) of 5.81 ng.dL (normal range, 0.731.79), free triiodothyronine (T3) of 15.7 pg/mL (normal range, 2.85.3), and total triiodothyronine (T3) of 218 ng/dL (normal range, 95170). The patient was diagnosed with thyroid crisis and was started on propranolol, propylthiouracil, hydrocortisone, and a saturated solution of potassium iodine. Thyroid stimulating immunoglobulins were obtained and found to be markedly elevated (3.4 TSI index; normal 1.3), suggestive of Grave's disease. Over the next several days, the patient's abdominal pain and tachycardia resolved. Her mental status returned to normal. A workup for her microcytic anemia revealed beta thalassemia trait. The patient was discharged home on hospital day 9 and has done well as an outpatient.

COMMENTARY

As Sir Zachary Cope stated in his classic text Cope's Early Diagnosis of the Acute Abdomen, [I]t is only by thorough history taking and physical examination that one can propound a diagnosis.1 When first presented with a patient whose chief complaint is abdominal pain, physicians tend to focus on the disorders of both the hollow and solid organs of the abdomen as potential sources of the pain. The differential diagnosis traditionally includes disorders such as cholecystitis, peptic ulcer disease, pancreatitis, small bowel obstruction, bowel ischemia or perforation, splenic abscess and infarct, nephrolithiasis, diverticulitis, and appendicitis, all of which were initially considered by the clinicians involved in this case. But as our discussant pointed out, in this case the differential needed to be broadened to include less common disorders, particularly given the patient's altered mental status, numerous electrolyte abnormalities, and lethargy and the lack of explanation provided by the physical examination and sophisticated imaging studies.

Specifically, a myriad of systemic diseases and metabolic derangements can cause abdominal complaints and mimic surgical abdominal disease, including hypercalcemia, acute intermittent porphyria, diabetic ketoacidosis, lead intoxication, familial Mediterranean fever, vasculopathies, adrenal insufficiency, and hyperthyroidism. Unfortunately, the frequency with which abdominal pain occurs in many of these less common disease processes and the pathophysiology that underlies its occurrence are not well defined. For example, abdominal pain is well described as a typical manifestation of both diabetic ketoacidosis and lead poisoning, but the pathophysiology behind its occurrence is poorly understood in both. Further, as a manifestation of thyrotoxicosis and as one of the diagnostic criteria for thyroid storm, the reported prevalence of abdominal pain in this condition is variable, ranging from rare to 20%47%.24 Also, although other gastrointestinal manifestations of hyperthyroidism (such as nausea, vomiting, and hyperdefecation) are thought to be the result of the effect of excess thyroid hormone on gastrointestinal motility, it is unclear whether this similar mechanism is responsible for the perception of abdominal pain.4

An important clue to the underlying diagnosis in this case was the patient's marked tachycardia. Classically, a persistent heart rate of 150 should raise suspicion of atrial flutter with a 2:1 conduction block, as was eventually discovered in this case. Adenosine, in addition to other vagal maneuvers such as carotid massage or Valsalva that also block atrioventricular (AV) node conduction, has been recognized as a safe and effective means of establishing a diagnosis in tachyarrhythmias.5 In AV nodal‐dependent tachycardias, such as AV node reentrant tachycardia or AV reentrant tachycardia, adenosine will often terminate the tachyarrhythmia by blocking the anterograde limb of the reentrant circuit. In AV nodeindependent tachyarrhythmias, such as atrial flutter or atrial fibrillation, adenosine will not terminate the rhythm. However, in the case of flutter, blocking the AV node will usually transiently unmask the underlying P waves, thereby facilitating the diagnosis.5, 6

In this patient, the discovery of atrial flutter was the main clue that thyrotoxicosis may provide the unifying diagnosis. Thyroid hormone has a direct positive cardiac chronotropic effect, resulting in the increased resting heart characteristic of thyrotoxicosis. Specifically, this hormone increases sinoatrial‐node firing, shortens the refractory period of conduction tissue within the heart, and decreases the electrical threshold for atrial excitation. In addition to predisposing to sinus tachycardia (the most common rhythm associated with this disorder), thyrotoxicosis is also associated with atrial tachycardias such as atrial flutter and, more classically, atrial fibrillation.7, 8 Though no studies have specifically evaluated the incidence of atrial flutter in thyrotoxicosis, atrial fibrillation has been found in 9%22% of these patients.7

Finally, several of the patient's electrolyte derangements could explain some of her clinical findings and are clues to the underlying diagnosis. She initially presented with a mild hypercalcemia that persisted even after hydration. Potential explanations include her severe dehydration or her underlying thyrotoxicosis because hypercalcemia is present in up to 20% of patients with hyperthyroidism.9, 10 However, the presence of significant hypercalcemia in the setting of thyrotoxicosis may actually make the diagnosis of thyrotoxicosis more difficult, masking the hypermetabolic signs and symptoms of the hyperthyroid state.11 Interestingly, coexistent primary hyperparathyroidism does occur in a few of these patients, but it likely was not an underlying cause in our patient given that her calcium normalized after receipt of propylthiouracil therapy.12

The patient's marked hypernatremia is more difficult to explain. She may have developed nephrogenic diabetes insipidus secondary to hypercalcemia, explained by a renal concentrating defect that can become evident once the calcium is persistently above 11 mg/dL.13 Combined with her altered mental status, which likely limited her ability to access free water, this may be enough to explain her marked hypernatremia. Her rapid improvement with rehydration is also consistent with this explanation, mediated through the improvement of her serum free calcium.

This case highlights the importance of using all the clinical clues provided by the history, physical exam, and laboratory and imaging studies when generating an initial differential diagnosis, as well as the importance of being willing to appropriately broaden and narrow the list of possibilities as a case evolves. When this patient was initially evaluated by physicians in the emergency department, they believed her symptoms were most consistent with generalized peritonitis that was likely secondary to an infectious or inflammatory intra‐abdominal process such as pancreatitis (especially in light of her mildly elevated lipase and amylase), appendicitis, or diverticulitis. When the medical team in the intensive care unit assumed care of this patient, members of the team failed to recognize several of the early clues, including the patient's markedly abnormal mental status, electrolyte derangements, and persistent tachycardia despite aggressive rehydration, which suggested the possibility of alternative, and less common, etiologies of her abdominal pain. Instead, they continued to aggressively pursue the possibility of the initial differential diagnosis, even repeating some of the previously negative studies from the emergency department. This case illustrates the importance of constantly reevaluating the available information from physical examination and laboratory and imaging studies and not falling victim to intellectual blind spots created by suggested diagnoses by other care providers. Fortunately for this patient, her thyroid crisis was diagnosed, albeit with some delay, before any long‐term complications occurred.

A 47‐year‐old woman was brought to the emergency department by her family because of 1 week of abdominal pain. The pain had begun in the epigastrium but had spread across the abdomen. She described it as constant and 10 of 10 in intensity but could not identify aggravating or alleviating factors. She also complained of nausea and vomiting, beginning 4 days prior to presentation, occurring 25 times per day. She noted poor oral intake and mild diarrhea. She denied melena or hematochezia. She reported no recent fever, dysuria, chills, or night sweats; however, she reported upper respiratory symptoms 2 weeks prior to presentation. On the day of presentation, her family felt she was becoming increasingly lethargic.

Epigastric pain in a middle‐aged woman suggests several possible diagnoses. Conditions such as acute cholecystitis begin abruptly, whereas small bowel obstruction, appendicitis, and diverticulitis start gradually. Nausea and vomiting are common concomitants of abdominal pain and are nonspecific. The absence of fever and chills is reassuring. Of greatest concern is the mental status. Initially, I think of enterohemorrhagic E. coli syndromes with associated glomerulonephritis. With a more systemic metabolic abnormality such as this, the rapid development of the disease tends to exaggerate symptoms.

The patient had a history of nephrolithiasis and underwent total abdominal hysterectomy and bilateral salpingo‐oopherectomy secondary to uterine fibroids in the past. She took occasional acetaminophen, smoked two cigarettes per day, and rarely consumed alcohol. Temperature was 38.5C, heart rate was 160 beats/minute, respiratory rate was 28/minute, and blood pressure was 92/52 mm Hg; oxygen saturation was 100% breathing 2 L of oxygen by nasal cannula. She was a moderately obese African American woman in moderate distress, lying in bed moaning. Mucous membranes were dry. There was no lymphadenopathy or thyromegaly. Heart rate was regular without appreciable murmur, rub, or gallop. Lungs were clear. Abdomen was soft and nondistended, with diffuse tenderness to palpation; bowel sounds were present; there was no rebound or guarding. She had normal rectal tone with brown, guaiac‐negative stool. There was no costovertebral angle tenderness. She was oriented to person, place, and time but lethargic; deep tendon reflexes were 3+ bilaterally, and no focal signs were elicited.

Renal stones certainly produce abdominal pain, and the rare patient undergoes laparotomy for this reason. The hysterectomy tells us that small bowel obstruction could be a reason for her symptoms, although abnormal mental status would not be expected without additional problems such as infection. The tachycardia seems out of proportion to her temperature. Hyperpnea and absent respiratory symptoms, along with hypotension and tachycardia, suggest a sepsis syndrome. Her physical exam confirms dehydration. Examination of the abdomen makes me speculate about whether she has a nonsurgical cause of acute abdomen. The lethargy remains unexplained. Sepsis syndrome, possibly from a perinephric abscess, is my leading diagnosis.

White blood cell count was 15.9/mm³ with 78% neutrophils, a hemoglobin of 14.3 g/dL with a MCV of 76 and a platelet count of 320/mm³. Sodium was 159 mmol/L, chloride 128 mmol/L, bicarbonate 19 mmol/L, blood urea nitrogen 120 mmol/L, creatinine 3.1 mg/dL, calcium 11.7 mg/dL, albumin 3.3 g/dL, serum aspartate aminotransferase 65 U/L, serum alanine aminotransferase 72 U/L, total bilirubin 0.7 mg/dL, amylase 137 U/L (normal 30100), and lipase 92 IU/dL (normal 424). Urine obtained from a Foley catheter revealed negative nitrite and leukocyte esterase, 5075 red blood cells, and 1025 white blood cells per high‐powered field.

The elevated serum sodium is likely contributing to her abnormal mental status. It is unusual for a previously healthy and conscious woman to become this hypernatremic because persons with a normal mental status will defend their sodium balance strenuously, assuming regulatory mechanisms are intact. Generally, this level of hypernatremia indicates 2 things. One, a patient was not allowed, or did not seek access to, free water. The other is the presence of diabetes insipidus. It is unlikely she became this dehydrated from the initial gastrointestinal episode as described. The low MCV suggests she may be a thalassemia carrier, as microcytosis with iron deficiency typically does not occur until the patient is anemic, although she may be when rehydrated. Serum calcium, while elevated, also will likely return to the normal range with hydration. The metabolic abnormalities strongly suggest a problem in the central nervous system. The hematuria in the urinalysis continues to raise the possibility of nephrolithiasis as a cause of abdominal pain, though it does not fit well with the rest of the patient's clinical picture. The hematuria and pyuria both could still indicate a urinary tract infection such as pyelonephritis or perinephric abscess causing a sepsis syndrome.

An acute abdominal series and chest radiograph revealed a paucity of gas in the abdomen but no free air under the diaphragm or active cardiopulmonary disease. Abdominal ultrasound showed cholelithiasis without biliary dilation. There was no evidence of hydronephrosis, hydroureter, or perinephric abscess. A noncontrast abdominal‐pelvic computed tomography (CT) scan demonstrated no peripancreatic stranding or fluid collection and no nephrolithiasis or fluid collection suggestive of abscess. The admission electrocardiogram, read as sinus tachycardia with a rate of 160, is displayed in Figure 1.

Figure 1

I have long believed that unexplained sinus tachycardia is one of the most ominous rhythms in clinical medicine; it is expected after vigorous exercise, among other situations, but not in the condition in which this woman finds herself. The nature of the tracing does not indicate the likelihood of a supraventricular arrhythmia, particularly atrial flutter, which should be considered given the rate. The absence of free air under the diaphragm on chest radiography is reassuring. Though the pancreatic enzymes are mildly elevated, they are usually far more striking in gallstone pancreatitis. Hypercalcemia may result in abdominal pain by several mechanisms. I remain concerned about her central nervous system.

The patient was admitted to the intensive care unit (ICU), where she received intravenous antibiotics and aggressive rehydration. The following morning, she continued to complain of abdominal pain. Her systolic blood pressure was 115 mmHg, and her heart rate ranged between 140 and 150 beats/minute. The remainder of her physical exam was unchanged. Repeat laboratory tests revealed a white blood cell count of 14.7/mm³, a blood urea nitrogen of 66 mg/dL, a creatinine of 1.3 mg.dL, amylase of 67 IU/L, and lipase of 70 IU/dL. A contrast‐enhanced abdominal‐pelvic CT scan did not reveal intra‐abdominal pathology. Blood and urine cultures obtained at admission were negative for any growth.

The patient was appropriately admitted to the ICU. When caring for a critically ill patient, establishing a diagnosis is less important initially than addressing treatable conditions with dispatch. The negative CT scans rule out previously entertained diagnoses like nephrolithiasis and perinephric abscess. It is possible that the initially positive urinalysis was a result of urinary catheter placement trauma. Given the course to date, I believe this patient likely has a nonsurgical cause of abdominal pain. I am considering entities such as lead intoxication, hypercalcemia, a tear of the rectus abdominus caused by vomiting, systemic vasculitis, or a hypercoagulable state leading to intra‐abdominal venous thrombosis.

By hospital day 3 her sodium decreased to 149 mmol/L and her creatinine was 1.0 mg/dL. Abdominal pain persisted, unchanged from admission. Her systolic blood pressure had stabilized at 120 mmHg, but the heart rate remained near 150 beats/minute. Her abdomen remained soft and nondistended on exam but diffusely tender to palpation. Her amylase and lipase continued to decrease, and her repeat electrocardiogram demonstrated tachycardia with a rate of 144.

We are gratified to see that her serum sodium has waned but not with the persistence of the tachycardia. It must be assumed that this patient has an infectious disease that we are not clever enough to diagnose at this time. I am also considering an autoimmune process, such as systemic lupus erythematosus. It is difficult to envision a neoplastic disorder causing these problems. The differential remains broad, however, because we have not ruled out metabolic or endocrine causes. It is difficult to imagine she could have Addison's diseasea common cause of severe abdominal pain, tachycardia, and hypotensiongiven her serum sodium level. Hyperthyroidism has been known to produce mild hypercalcemia and abdominal complaints and is an intriguing possibility. The striking elevation of her serum sodium makes me consider the possibility of a problem in the posterior pituitary gland such as sarcoidosis. I cannot explain how sarcoidosis would cause her abdominal pain, unless the hypercalcemia were related. The tachycardia remains of concern, especially if she is otherwise improving. Thus, I would likely administer a small dose of adenosine to ascertain that this is not a different supraventricular tachycardia. In sinus tachycardia, the rate is usually attendant to the clinical picture and thus begs explanation given her clinical improvement.

After receiving 6 mg of intravenous adenosine, the patient's heart rate declined; atrial flutter waves were observed.

This case nicely demonstrates a key teaching point: a fast regular heart rate of about 150, irrespective of the electrocardiogram, suggests atrial flutter. Who gets atrial flutter? Patients with chronic lung disease, myocardial ischemia (albeit rarely), alcohol‐induced cardiomyopathy, and infiltrative cardiac disorders do. Additionally, we also have to consider thyroid dysfunction.

If forced to come up with a single unifying diagnosis at this point, I would have to say this patient most likely has sarcoidosis because this entity would account for modest hypercalcemia, the myocardial conduction disturbance, and hypernatremia because of diabetes insipidus; furthermore, it would fit the patient's demographic profile. However, I am also concerned about hyperthyroidism and would not proceed until thyroid function studies were obtained.

Thyroid studies revealed thyroid stimulating hormone of less than 0.01 mU/L (normal range, 0.305.50), free thyroxine (T4) of 5.81 ng.dL (normal range, 0.731.79), free triiodothyronine (T3) of 15.7 pg/mL (normal range, 2.85.3), and total triiodothyronine (T3) of 218 ng/dL (normal range, 95170). The patient was diagnosed with thyroid crisis and was started on propranolol, propylthiouracil, hydrocortisone, and a saturated solution of potassium iodine. Thyroid stimulating immunoglobulins were obtained and found to be markedly elevated (3.4 TSI index; normal 1.3), suggestive of Grave's disease. Over the next several days, the patient's abdominal pain and tachycardia resolved. Her mental status returned to normal. A workup for her microcytic anemia revealed beta thalassemia trait. The patient was discharged home on hospital day 9 and has done well as an outpatient.

COMMENTARY

As Sir Zachary Cope stated in his classic text Cope's Early Diagnosis of the Acute Abdomen, [I]t is only by thorough history taking and physical examination that one can propound a diagnosis.1 When first presented with a patient whose chief complaint is abdominal pain, physicians tend to focus on the disorders of both the hollow and solid organs of the abdomen as potential sources of the pain. The differential diagnosis traditionally includes disorders such as cholecystitis, peptic ulcer disease, pancreatitis, small bowel obstruction, bowel ischemia or perforation, splenic abscess and infarct, nephrolithiasis, diverticulitis, and appendicitis, all of which were initially considered by the clinicians involved in this case. But as our discussant pointed out, in this case the differential needed to be broadened to include less common disorders, particularly given the patient's altered mental status, numerous electrolyte abnormalities, and lethargy and the lack of explanation provided by the physical examination and sophisticated imaging studies.

Specifically, a myriad of systemic diseases and metabolic derangements can cause abdominal complaints and mimic surgical abdominal disease, including hypercalcemia, acute intermittent porphyria, diabetic ketoacidosis, lead intoxication, familial Mediterranean fever, vasculopathies, adrenal insufficiency, and hyperthyroidism. Unfortunately, the frequency with which abdominal pain occurs in many of these less common disease processes and the pathophysiology that underlies its occurrence are not well defined. For example, abdominal pain is well described as a typical manifestation of both diabetic ketoacidosis and lead poisoning, but the pathophysiology behind its occurrence is poorly understood in both. Further, as a manifestation of thyrotoxicosis and as one of the diagnostic criteria for thyroid storm, the reported prevalence of abdominal pain in this condition is variable, ranging from rare to 20%47%.24 Also, although other gastrointestinal manifestations of hyperthyroidism (such as nausea, vomiting, and hyperdefecation) are thought to be the result of the effect of excess thyroid hormone on gastrointestinal motility, it is unclear whether this similar mechanism is responsible for the perception of abdominal pain.4

An important clue to the underlying diagnosis in this case was the patient's marked tachycardia. Classically, a persistent heart rate of 150 should raise suspicion of atrial flutter with a 2:1 conduction block, as was eventually discovered in this case. Adenosine, in addition to other vagal maneuvers such as carotid massage or Valsalva that also block atrioventricular (AV) node conduction, has been recognized as a safe and effective means of establishing a diagnosis in tachyarrhythmias.5 In AV nodal‐dependent tachycardias, such as AV node reentrant tachycardia or AV reentrant tachycardia, adenosine will often terminate the tachyarrhythmia by blocking the anterograde limb of the reentrant circuit. In AV nodeindependent tachyarrhythmias, such as atrial flutter or atrial fibrillation, adenosine will not terminate the rhythm. However, in the case of flutter, blocking the AV node will usually transiently unmask the underlying P waves, thereby facilitating the diagnosis.5, 6

In this patient, the discovery of atrial flutter was the main clue that thyrotoxicosis may provide the unifying diagnosis. Thyroid hormone has a direct positive cardiac chronotropic effect, resulting in the increased resting heart characteristic of thyrotoxicosis. Specifically, this hormone increases sinoatrial‐node firing, shortens the refractory period of conduction tissue within the heart, and decreases the electrical threshold for atrial excitation. In addition to predisposing to sinus tachycardia (the most common rhythm associated with this disorder), thyrotoxicosis is also associated with atrial tachycardias such as atrial flutter and, more classically, atrial fibrillation.7, 8 Though no studies have specifically evaluated the incidence of atrial flutter in thyrotoxicosis, atrial fibrillation has been found in 9%22% of these patients.7

Finally, several of the patient's electrolyte derangements could explain some of her clinical findings and are clues to the underlying diagnosis. She initially presented with a mild hypercalcemia that persisted even after hydration. Potential explanations include her severe dehydration or her underlying thyrotoxicosis because hypercalcemia is present in up to 20% of patients with hyperthyroidism.9, 10 However, the presence of significant hypercalcemia in the setting of thyrotoxicosis may actually make the diagnosis of thyrotoxicosis more difficult, masking the hypermetabolic signs and symptoms of the hyperthyroid state.11 Interestingly, coexistent primary hyperparathyroidism does occur in a few of these patients, but it likely was not an underlying cause in our patient given that her calcium normalized after receipt of propylthiouracil therapy.12

The patient's marked hypernatremia is more difficult to explain. She may have developed nephrogenic diabetes insipidus secondary to hypercalcemia, explained by a renal concentrating defect that can become evident once the calcium is persistently above 11 mg/dL.13 Combined with her altered mental status, which likely limited her ability to access free water, this may be enough to explain her marked hypernatremia. Her rapid improvement with rehydration is also consistent with this explanation, mediated through the improvement of her serum free calcium.

This case highlights the importance of using all the clinical clues provided by the history, physical exam, and laboratory and imaging studies when generating an initial differential diagnosis, as well as the importance of being willing to appropriately broaden and narrow the list of possibilities as a case evolves. When this patient was initially evaluated by physicians in the emergency department, they believed her symptoms were most consistent with generalized peritonitis that was likely secondary to an infectious or inflammatory intra‐abdominal process such as pancreatitis (especially in light of her mildly elevated lipase and amylase), appendicitis, or diverticulitis. When the medical team in the intensive care unit assumed care of this patient, members of the team failed to recognize several of the early clues, including the patient's markedly abnormal mental status, electrolyte derangements, and persistent tachycardia despite aggressive rehydration, which suggested the possibility of alternative, and less common, etiologies of her abdominal pain. Instead, they continued to aggressively pursue the possibility of the initial differential diagnosis, even repeating some of the previously negative studies from the emergency department. This case illustrates the importance of constantly reevaluating the available information from physical examination and laboratory and imaging studies and not falling victim to intellectual blind spots created by suggested diagnoses by other care providers. Fortunately for this patient, her thyroid crisis was diagnosed, albeit with some delay, before any long‐term complications occurred.

Community‐acquired pneumonia (CAP) is commonly defined as an infection of the pulmonary parenchyma that is associated with at least some symptoms and signs of acute infection, accompanied by the presence of an acute infiltrate on chest radiograph, in a patient not hospitalized or residing in a long‐term‐care facility for 14 days prior to the onset of symptoms.1 CAP continues to be a common and serious illness, causing substantial morbidity and mortality in the adult population. There are an estimated 56 million cases a year in the United States, with greater than 1 million hospitalizations. Community‐acquired pneumonia is one of the most common admitting diagnoses among adults, and with a 30‐day mortality between 10% and 14% for patients admitted to the hospital, it is the leading cause of infectious death in the United States.2 In elderly patients, hospitalization for CAP portends a poor long‐term prognosis. In a Medicare database, the 1‐year mortality for patients with CAP was nearly 40%, compared to 29% in patients with other diagnoses.3 Community‐acquired pneumonia is a model illness in hospital medicineit is a common disease that allows for evidence‐based and cost‐effective management. In addition, many national organizations have proposed multiple quality indicators for community‐acquired pneumonia, thus providing an opportunity for institutional quality improvement. This review article outlines the assessment and management of patients admitted to the hospital with community‐acquired pneumonia.

Etiology

Although many pathogens can cause community‐acquired pneumonia, the clinical syndromes and microbiology of CAP have traditionally been characterized as either typical or atypical. The typical organisms include Streptococcus pneumoniae, Haemophilus influenzae, and Moraxella catarrhalis, and the atypical organisms include Chlamydia spp., Mycoplasma pneumoniae, Legionella spp., and viruses. This historical distinction has recently come into question. It is now clear that the presenting symptoms, signs, and basic laboratory findings (including the chest radiograph) cannot be reliably used to predict the etiologic pathogen or to distinguish typical from atypical organisms.4 Rather, the specific causative agent of CAP depends more on the degree of patient illness. Table 1 shows what prospective studies with comprehensive diagnostic strategies determined to be the most common pathogens in patients hospitalized for CAP in ICU and non‐ICU settings.5 Streptococcus pneumoniae remains the most common cause of CAP in hospitalized patients and is the most common cause of fatal pneumonia, whereas Legionella spp. is a common cause of severe CAP, more often found in patients requiring admission to the intensive care unit. Gram‐negative bacilli can cause CAP in elderly patients and those recently treated with broad‐spectrum antibiotics or with underlying lung disease. Notably, though, despite improved diagnostic testing, only one quarter of all admitted patients with CAP have the etiologic agent defined, and therefore empiric therapy should be directed broadly at the most likely organisms.6

Clinical Presentation

Patients admitted to the hospital with CAP typically present with a brief history of respiratory complaints, including cough (greater than 90%), dyspnea (66%), sputum production (66%), and pleuritic chest pain (50%); see Table 2.7, 8 In 10%30% of patients, nonrespiratory complaints predominate, including headache, myalgias, fatigue, and gastrointestinal symptoms.6 Elderly patients, an increasing percentage of hospitalized patients, are less likely to present with typical CAP symptoms (such as cough) and more likely to have altered mental status as a presenting symptom.9

On physical examination, patients with CAP usually have signs of fever (80%), tachypnea (70%), and tachycardia (50%); see Table 2. Most will have a focal lung exam (>90%) with findings ranging from crackles to bronchial breath sounds.10 No exam finding is specific for the diagnosis of pneumonia, but the absence of fever, tachycardia, and tachypnea significantly reduces the probability of CAP in patients with suspected pneumonia.10 Furthermore, similar to the clinical history, the physical examination of elderly patients with community‐acquired pneumonia is not specific or sensitive for the diagnosis of CAP. For example, up to 40% of elderly patients subsequently determined to have CAP may not have fever.11

Leukocytosis is common in patients with CAP; however, its absence does not rule out disease.12 A number of guidelines recommend laboratory evaluation of electrolytes, urea nitrogen, creatinine, liver enzymes, and bilirubin, although these are used primarily for prognostication and are not specifically useful in the diagnosis of CAP.

DIAGNOSIS

Diagnostic Studies

The diagnosis of community‐acquired pneumonia requires that a patient have both signs and symptoms consistent with pulmonary infection and evidence of a new radiographic infiltrate. Therefore, most guidelines recommend that all patients with a possible diagnosis of CAP be evaluated with chest radiography.1, 14, 15

The specific radiographic findings in community‐acquired pneumonia range from lobar consolidation to hazy focal infiltrate to diffuse bilateral interstitial opacities (see Figure 1). Although chest radiography has traditionally been considered the gold standard for the diagnosis of CAP, its exact performance characteristics are unknown, and it is clearly not 100% sensitive or 100% specific. The utility of the chest radiograph can be limited by patient body habitus, underlying lung disease, or dehydration. Computed tomography (CT) scanning, although not recommended for routine use, can identify pulmonary consolidation in up to 30% of patients with a normal or equivocal chest radiograph in whom pneumonia is suspected and can also identify complications of pneumonia including an empyema or pulmonary abscess.16

Figure 1

Limitations in the performance of the chest radiograph have resulted in an interest in the diagnostic performance of serologic markers of infection such as C‐reactive protein (CRP), procalcitonin, and soluble triggering receptor expressed on myeloid cells (s‐TREM).1719 Preliminary evidence suggests these inflammatory markers may ultimately prove useful in differentiating infectious from noninfectious pulmonary processes, but regular use of these new tests cannot currently be recommended.

Most expert guidelines state that 2 sets of blood cultures should be taken and analyzed prior to antibiotic administration in all patients admitted to the hospital with suspected community‐acquired pneumonia.1, 14, 15 Isolation of bacteria from blood cultures in CAP is a very specific way to identify a causative organism in order to subsequently narrow therapy and also identifies a high‐risk group of patients because bacteremia is associated with increased mortality. Obtaining blood cultures within 24 hours of admission has been associated with 10% lower odds of 30‐day mortality in patients with CAP,20 and as a result, drawing blood cultures prior to antibiotic administration is a national quality indicator for CAP.

There are, however, a number of problems with the routine acquisition of blood cultures in all patients admitted with CAP. Practically, the cultures can be difficult to obtain, can potentially delay the initiation of antibiotics, and are often contaminated, which has been shown to increase both cost and length of stay.21, 22 The yield is generally low: the true‐positive bacteremia rate for admitted patients with CAP ranges from 6% to 9%, and the culture results rarely change management or outcomes.23, 24 Given these limitations, many have argued that blood cultures should be obtained with a more targeted approach. A recent study used a Medicare database to create a decision‐support tool to help maximize the value of blood cultures in CAP.25 The predictors of a positive blood culture are shown in Table 3. Not obtaining cultures on patients who had received prior antibiotics or had no risk factors resulted in about 40% fewer overall cultures while identifying approximately 90% of bacteremias. In their guidelines, the British Thoracic Society (BTS) advocates a similar strategy, recommending blood cultures be omitted in nonsevere pneumonia and in patients without comorbidities.15, 26 Although recommendations vary for non‐severe CAP in hospitalized patients, all professional society guidelines agree that blood cultures should be obtained in critically ill patients, and if cultures are obtained, they should be drawn prior to antibiotics.1, 14, 15, 26

Table 3.Independent Predictors of Bacteremia in Patients with Community‐Acquired Pneumonia24

Comorbidities
Liver disease
Vital signs
Systolic blood pressure 90 mm Hg
Temperature 35C or 40C
Pulse 125 beats/min
Laboratory and radiographic data
Blood urea nitrogen (BUN) 30 mg/dL
Sodium 130 mmol/L
White blood cells 5000/mm3 or > 20,000/mm3
Prior use of antibiotics (negative predictor)

Substantial controversy surrounds the utility of routine sputum gram stains and cultures for patients admitted to the hospital with CAP. The Infectious Disease Society of America (IDSA) and the British Thoracic Society (BTS) both recommend that all patients admitted to the hospital with community‐acquired pneumonia should have a gram stain and culture of expectorated sputum.1, 15, 26 Both organizations argue sputum collection is a simple and inexpensive procedure that can potentially identify pathogenic organisms and can affect both initial and long‐term antibiotic therapy. Most notably, they highlight gram stain specificity of greater than 80% for pneumococcal pneumonia. Conversely, the American Thoracic Society (ATS) argues that sputum gram stains and cultures generally have low sensitivity, specificity, and positive predictive value.14 Furthermore, they argue the utility of sputum testing is also limited practically; in one study 30% of patients could not produce an adequate sputum specimen and up to 30% had received prior antibiotic therapy, substantially reducing the yield.27 In another study, good‐quality sputum with a predominant morphotype could be obtained in only 14% of patients admitted with CAP.28 However, targeting sputum analysis to patients who have not received prior antibiotics and are able to produce an adequate sample improved the yield significantly.29 In addition, with increasing rates of antibiotic resistance among common community isolates (ie, S. Pneumoniae) and the increasing prevalence of infecting organisms not targeted by routine empiric therapy (methicillin‐resistant Staphylococcus Aureus [MRSA]), isolation of potential causative pathogens is increasingly important. We believe that severely ill patients with CAP (such as patients admitted to the ICU), as well as patients with identifiable risk factors for uncommon or drug‐resistant pathogens (eg, Pseudomonas aeruginosa, enteric gram‐negative rods, MRSA, etc.) should have sputum sent for gram stain and culture. Ideally, sputum obtained for gram stain and culture should be:

• 1

  Prior to antibiotic therapy,

• 2

  A deep‐cough, expectorated specimen,

• 3

  A purulent specimen (>25 polymorphonucleacytes and less than 10 squamous cells per high‐powered field), and

• 4

  Rapidly transported to the laboratory.

Subsequent gram stain and culture results should be interpreted in the specific clinical context and antibiotic choices targeted appropriately.

ADMISSION DECISION

Once the diagnosis of CAP has been made, the initial site where treatment will occur, whether the hospital or the home, must be determined. The decision to hospitalize should be based on 3 factors: 1) evaluation of the safety of home treatment, 2) calculation of the Pneumonia Severity Index (PSI), and 3) clinical judgment of the physician.27 The PSI, or PORT (Pneumonia Outcomes Research Team) score, is a validated prediction rule that quantifies mortality and allows for risk stratification of patients with community‐acquired pneumonia.2 The PSI combines clinical history, physical examination, and laboratory data at the time of admission to divide patients into 5 risk classes and to estimate 30‐day mortality (Figure 2), which ranges from 0.1% of patients in risk class I to 27.0% in risk class V.2

Figure 2

On the basis of the estimated prognosis and in the absence of concerns about home safety or comorbidities, patients in risk classes I, II, and III should be managed at home. Many prospective trials have shown that implementation of PSI significantly increases the number of low‐risk patients managed outside the hospital, with no differences in quality of life, complications, readmissions, or short‐term mortality.30, 31 Most recently, a trial randomizing patients in risk classes II and III to treatment in the hospital or at home found no significant differences in clinical outcomes but did find that patients were more satisfied with care at home.32 Because the number of patients with CAP being treated at home is increasing, the American College of Chest Physicians recently published a consensus statement on the management of community‐acquired pneumonia in the home.33 All national guidelines for the management of community‐acquired pneumonia recommend using the PSI to help determine the initial location of treatment, with the caveat that using the prediction rule should never supersede clinical judgment in the decision about whether to admit.1, 14, 15, 26, 27 A practical decision tree for the use of the PSI is shown in Figure 3.

Figure 3

There are no reliable prediction rules for deciding on whether admission to the intensive care unit is necessary. Hemodynamic instability requiring resuscitation and monitoring or respiratory failure requiring ventilatory support are clear indications for ICU admission. Additional variables such as tachypnea (respiratory rate 30), altered mental status, multilobar disease, and azotemia are associated with severe CAP and should prompt consideration of ICU admission, especially when 2 or more variables coexist.14

TREATMENT

DISCHARGE/FOLLOW‐UP PLANS

Patients hospitalized for community‐acquired pneumonia can be safely discharged when they have reached clinical stability, are able to tolerate oral medications, have no other active comorbid conditions, and have safe, close, appropriate outpatient follow‐up (see Table 7). Clinical pathways employing these discharge criteria have been found to be safe and effective in reducing the length of stay for CAP. Most important, patients should have met most if not all of the vital sign and clinical criteria noted in Table 5 in the criteria for switching to oral therapy. Patients with 2 or more abnormal vital signs (instabilities) within 24 hours prior to discharge are at high risk of readmission and mortality, but those with one or no abnormal vital signs generally have good outcomes.65 Absent other clinical factors or extenuating circumstances (persistent hypoxia, poor functional status, etc.), most patients with CAP should reach clinical stability by day 3 or 4, be considered for a switch to oral therapy, and, if stable, be discharged shortly thereafter.

When patients with CAP are discharged from the hospital, they should be counseled about the expected course of recovery. Most important, patients and families must be informed that many symptoms of CAP may persist well after hospitalization. In one study, up to 80% of patients reported persistent cough and fatigue 1 week after discharge, and up to 50% still had dyspnea and sputum production. In some, the cough can last for 46 weeks.8

All patients discharged after treatment of community‐acquired pneumonia should have follow‐up with their outpatient provider. The physician responsible for their inpatient care should communicate directly with the provider and outline the hospital course, the discharge medications, and the duration of antibiotic therapy. There is no specific time frame within which patients must be seen, but follow‐up should be dictated by patient age, comorbidities, clinical stability at discharge, and degree of illness. The American Thoracic Society guidelines do recommend patients with a substantial smoking history who are hospitalized with CAP have a follow‐up chest radiograph 46 weeks after discharge to establish a radiographic baseline and exclude the possibility of underlying malignancy.14 However, several studies have suggested that radiographic resolution may take 3 or more months in some patients, especially the elderly and those with multilobar disease.66

PREVENTION

Prevention of community‐acquired pneumonia and pneumonic syndromes has traditionally relied on vaccination with the polysaccharide pneumococcal pneumonia vaccine and the seasonal influenza vaccine. The vaccine for S. pneumoniae used in adults is composed of the 23 serotypes that cause 85%90% of the invasive pneumococcal infections in the United States. Although in randomized trials the vaccine has not consistently prevented community‐acquired pneumonia or death in elderly patients or those with comorbidities, it likely prevents invasive pneumococcal infection.67 National guidelines and the CDC recommend the pneumococcal vaccine be given to all patients older than 65 years and those with chronic medical conditions.1, 14, 15

The seasonal influenza vaccine has clearly been shown to decrease influenza‐related illness in elderly and high‐risk patient populations. As well, in a meta‐analysis and a large observational study of patients older than 65 years, vaccination against influenza prevented pneumonia, hospitalization, and death.68, 69 Vaccination of health care workers may also confer a benefit to elderly patients of reduced mortality. The CDC recommends the influenza vaccine for all patients more than 50 years old, those with comorbidities, those at high risk for influenza, and health care workers in both inpatient and outpatient settings.

Pneumococcal and influenza vaccination have traditionally been relegated to the outpatient setting. National guidelines and the CDC recommend vaccination of all eligible hospitalized patients. Vaccination is safe and effective with almost any medical illness, and both vaccines can be given simultaneously at discharge.69 Both JCAHO and CMS have defined administration of the pneumococcal and influenza vaccines to patients hospitalized with CAP as a quality measure. Using standing orders is the most effective means of ensuring vaccination.

Some evidence suggests that tobacco smokers are at increased risk of invasive pneumococcal disease or pneumonia.70 Patients hospitalized (for all illnesses, but for CAP in particular) should be counseled about smoking cessation and offered pharmacotherapy and outpatient follow‐up. And, finally, recent observational data suggests that use of acid suppressive therapy, including proton pump inhibitors and H‐2 receptor antagonists, may be associated with an increased risk of developing CAP.71 Patients using these agents who are admitted with CAP should have their indications for treatment reviewed, especially when the pneumonia has been recurrent and there is no clear indication for continued use of acid suppressive therapy, in which case they should be discontinued in the hospital.

CONCLUSIONS

Community‐acquired pneumonia remains a common cause for hospitalization of adult patients, with significant associated morbidity and mortality. Although there are multiple expert guidelines for the management of community‐acquired pneumonia, further research is urgently needed. Clinicians need improved diagnostic tests that enable an earlier and more accurate diagnosis of CAP. In addition, the etiologic agent causing CAP is rarely discovered; improved microbiologic studies might enable antibiotic therapy to be targeted to the organisms responsible. High‐quality randomized, controlled trials examining empiric antibiotic therapy in CAP are needed, especially related to the addition of agents covering atypical organisms. Last, the general management of patients hospitalized with CAP is marked by significant heterogeneity, and research and initiatives focusing on improving the quality and process of care of patients with CAP are needed.

Community‐acquired pneumonia (CAP) is commonly defined as an infection of the pulmonary parenchyma that is associated with at least some symptoms and signs of acute infection, accompanied by the presence of an acute infiltrate on chest radiograph, in a patient not hospitalized or residing in a long‐term‐care facility for 14 days prior to the onset of symptoms.1 CAP continues to be a common and serious illness, causing substantial morbidity and mortality in the adult population. There are an estimated 56 million cases a year in the United States, with greater than 1 million hospitalizations. Community‐acquired pneumonia is one of the most common admitting diagnoses among adults, and with a 30‐day mortality between 10% and 14% for patients admitted to the hospital, it is the leading cause of infectious death in the United States.2 In elderly patients, hospitalization for CAP portends a poor long‐term prognosis. In a Medicare database, the 1‐year mortality for patients with CAP was nearly 40%, compared to 29% in patients with other diagnoses.3 Community‐acquired pneumonia is a model illness in hospital medicineit is a common disease that allows for evidence‐based and cost‐effective management. In addition, many national organizations have proposed multiple quality indicators for community‐acquired pneumonia, thus providing an opportunity for institutional quality improvement. This review article outlines the assessment and management of patients admitted to the hospital with community‐acquired pneumonia.

Etiology

Although many pathogens can cause community‐acquired pneumonia, the clinical syndromes and microbiology of CAP have traditionally been characterized as either typical or atypical. The typical organisms include Streptococcus pneumoniae, Haemophilus influenzae, and Moraxella catarrhalis, and the atypical organisms include Chlamydia spp., Mycoplasma pneumoniae, Legionella spp., and viruses. This historical distinction has recently come into question. It is now clear that the presenting symptoms, signs, and basic laboratory findings (including the chest radiograph) cannot be reliably used to predict the etiologic pathogen or to distinguish typical from atypical organisms.4 Rather, the specific causative agent of CAP depends more on the degree of patient illness. Table 1 shows what prospective studies with comprehensive diagnostic strategies determined to be the most common pathogens in patients hospitalized for CAP in ICU and non‐ICU settings.5 Streptococcus pneumoniae remains the most common cause of CAP in hospitalized patients and is the most common cause of fatal pneumonia, whereas Legionella spp. is a common cause of severe CAP, more often found in patients requiring admission to the intensive care unit. Gram‐negative bacilli can cause CAP in elderly patients and those recently treated with broad‐spectrum antibiotics or with underlying lung disease. Notably, though, despite improved diagnostic testing, only one quarter of all admitted patients with CAP have the etiologic agent defined, and therefore empiric therapy should be directed broadly at the most likely organisms.6

Clinical Presentation

Patients admitted to the hospital with CAP typically present with a brief history of respiratory complaints, including cough (greater than 90%), dyspnea (66%), sputum production (66%), and pleuritic chest pain (50%); see Table 2.7, 8 In 10%30% of patients, nonrespiratory complaints predominate, including headache, myalgias, fatigue, and gastrointestinal symptoms.6 Elderly patients, an increasing percentage of hospitalized patients, are less likely to present with typical CAP symptoms (such as cough) and more likely to have altered mental status as a presenting symptom.9

On physical examination, patients with CAP usually have signs of fever (80%), tachypnea (70%), and tachycardia (50%); see Table 2. Most will have a focal lung exam (>90%) with findings ranging from crackles to bronchial breath sounds.10 No exam finding is specific for the diagnosis of pneumonia, but the absence of fever, tachycardia, and tachypnea significantly reduces the probability of CAP in patients with suspected pneumonia.10 Furthermore, similar to the clinical history, the physical examination of elderly patients with community‐acquired pneumonia is not specific or sensitive for the diagnosis of CAP. For example, up to 40% of elderly patients subsequently determined to have CAP may not have fever.11

Leukocytosis is common in patients with CAP; however, its absence does not rule out disease.12 A number of guidelines recommend laboratory evaluation of electrolytes, urea nitrogen, creatinine, liver enzymes, and bilirubin, although these are used primarily for prognostication and are not specifically useful in the diagnosis of CAP.

DIAGNOSIS

Diagnostic Studies

The diagnosis of community‐acquired pneumonia requires that a patient have both signs and symptoms consistent with pulmonary infection and evidence of a new radiographic infiltrate. Therefore, most guidelines recommend that all patients with a possible diagnosis of CAP be evaluated with chest radiography.1, 14, 15

The specific radiographic findings in community‐acquired pneumonia range from lobar consolidation to hazy focal infiltrate to diffuse bilateral interstitial opacities (see Figure 1). Although chest radiography has traditionally been considered the gold standard for the diagnosis of CAP, its exact performance characteristics are unknown, and it is clearly not 100% sensitive or 100% specific. The utility of the chest radiograph can be limited by patient body habitus, underlying lung disease, or dehydration. Computed tomography (CT) scanning, although not recommended for routine use, can identify pulmonary consolidation in up to 30% of patients with a normal or equivocal chest radiograph in whom pneumonia is suspected and can also identify complications of pneumonia including an empyema or pulmonary abscess.16

Figure 1

Limitations in the performance of the chest radiograph have resulted in an interest in the diagnostic performance of serologic markers of infection such as C‐reactive protein (CRP), procalcitonin, and soluble triggering receptor expressed on myeloid cells (s‐TREM).1719 Preliminary evidence suggests these inflammatory markers may ultimately prove useful in differentiating infectious from noninfectious pulmonary processes, but regular use of these new tests cannot currently be recommended.

Most expert guidelines state that 2 sets of blood cultures should be taken and analyzed prior to antibiotic administration in all patients admitted to the hospital with suspected community‐acquired pneumonia.1, 14, 15 Isolation of bacteria from blood cultures in CAP is a very specific way to identify a causative organism in order to subsequently narrow therapy and also identifies a high‐risk group of patients because bacteremia is associated with increased mortality. Obtaining blood cultures within 24 hours of admission has been associated with 10% lower odds of 30‐day mortality in patients with CAP,20 and as a result, drawing blood cultures prior to antibiotic administration is a national quality indicator for CAP.

There are, however, a number of problems with the routine acquisition of blood cultures in all patients admitted with CAP. Practically, the cultures can be difficult to obtain, can potentially delay the initiation of antibiotics, and are often contaminated, which has been shown to increase both cost and length of stay.21, 22 The yield is generally low: the true‐positive bacteremia rate for admitted patients with CAP ranges from 6% to 9%, and the culture results rarely change management or outcomes.23, 24 Given these limitations, many have argued that blood cultures should be obtained with a more targeted approach. A recent study used a Medicare database to create a decision‐support tool to help maximize the value of blood cultures in CAP.25 The predictors of a positive blood culture are shown in Table 3. Not obtaining cultures on patients who had received prior antibiotics or had no risk factors resulted in about 40% fewer overall cultures while identifying approximately 90% of bacteremias. In their guidelines, the British Thoracic Society (BTS) advocates a similar strategy, recommending blood cultures be omitted in nonsevere pneumonia and in patients without comorbidities.15, 26 Although recommendations vary for non‐severe CAP in hospitalized patients, all professional society guidelines agree that blood cultures should be obtained in critically ill patients, and if cultures are obtained, they should be drawn prior to antibiotics.1, 14, 15, 26

Table 3.Independent Predictors of Bacteremia in Patients with Community‐Acquired Pneumonia24

Comorbidities
Liver disease
Vital signs
Systolic blood pressure 90 mm Hg
Temperature 35C or 40C
Pulse 125 beats/min
Laboratory and radiographic data
Blood urea nitrogen (BUN) 30 mg/dL
Sodium 130 mmol/L
White blood cells 5000/mm3 or > 20,000/mm3
Prior use of antibiotics (negative predictor)

Substantial controversy surrounds the utility of routine sputum gram stains and cultures for patients admitted to the hospital with CAP. The Infectious Disease Society of America (IDSA) and the British Thoracic Society (BTS) both recommend that all patients admitted to the hospital with community‐acquired pneumonia should have a gram stain and culture of expectorated sputum.1, 15, 26 Both organizations argue sputum collection is a simple and inexpensive procedure that can potentially identify pathogenic organisms and can affect both initial and long‐term antibiotic therapy. Most notably, they highlight gram stain specificity of greater than 80% for pneumococcal pneumonia. Conversely, the American Thoracic Society (ATS) argues that sputum gram stains and cultures generally have low sensitivity, specificity, and positive predictive value.14 Furthermore, they argue the utility of sputum testing is also limited practically; in one study 30% of patients could not produce an adequate sputum specimen and up to 30% had received prior antibiotic therapy, substantially reducing the yield.27 In another study, good‐quality sputum with a predominant morphotype could be obtained in only 14% of patients admitted with CAP.28 However, targeting sputum analysis to patients who have not received prior antibiotics and are able to produce an adequate sample improved the yield significantly.29 In addition, with increasing rates of antibiotic resistance among common community isolates (ie, S. Pneumoniae) and the increasing prevalence of infecting organisms not targeted by routine empiric therapy (methicillin‐resistant Staphylococcus Aureus [MRSA]), isolation of potential causative pathogens is increasingly important. We believe that severely ill patients with CAP (such as patients admitted to the ICU), as well as patients with identifiable risk factors for uncommon or drug‐resistant pathogens (eg, Pseudomonas aeruginosa, enteric gram‐negative rods, MRSA, etc.) should have sputum sent for gram stain and culture. Ideally, sputum obtained for gram stain and culture should be:

• 1

  Prior to antibiotic therapy,

• 2

  A deep‐cough, expectorated specimen,

• 3

  A purulent specimen (>25 polymorphonucleacytes and less than 10 squamous cells per high‐powered field), and

• 4

  Rapidly transported to the laboratory.

Subsequent gram stain and culture results should be interpreted in the specific clinical context and antibiotic choices targeted appropriately.

ADMISSION DECISION

Once the diagnosis of CAP has been made, the initial site where treatment will occur, whether the hospital or the home, must be determined. The decision to hospitalize should be based on 3 factors: 1) evaluation of the safety of home treatment, 2) calculation of the Pneumonia Severity Index (PSI), and 3) clinical judgment of the physician.27 The PSI, or PORT (Pneumonia Outcomes Research Team) score, is a validated prediction rule that quantifies mortality and allows for risk stratification of patients with community‐acquired pneumonia.2 The PSI combines clinical history, physical examination, and laboratory data at the time of admission to divide patients into 5 risk classes and to estimate 30‐day mortality (Figure 2), which ranges from 0.1% of patients in risk class I to 27.0% in risk class V.2

Figure 2

On the basis of the estimated prognosis and in the absence of concerns about home safety or comorbidities, patients in risk classes I, II, and III should be managed at home. Many prospective trials have shown that implementation of PSI significantly increases the number of low‐risk patients managed outside the hospital, with no differences in quality of life, complications, readmissions, or short‐term mortality.30, 31 Most recently, a trial randomizing patients in risk classes II and III to treatment in the hospital or at home found no significant differences in clinical outcomes but did find that patients were more satisfied with care at home.32 Because the number of patients with CAP being treated at home is increasing, the American College of Chest Physicians recently published a consensus statement on the management of community‐acquired pneumonia in the home.33 All national guidelines for the management of community‐acquired pneumonia recommend using the PSI to help determine the initial location of treatment, with the caveat that using the prediction rule should never supersede clinical judgment in the decision about whether to admit.1, 14, 15, 26, 27 A practical decision tree for the use of the PSI is shown in Figure 3.

Figure 3

There are no reliable prediction rules for deciding on whether admission to the intensive care unit is necessary. Hemodynamic instability requiring resuscitation and monitoring or respiratory failure requiring ventilatory support are clear indications for ICU admission. Additional variables such as tachypnea (respiratory rate 30), altered mental status, multilobar disease, and azotemia are associated with severe CAP and should prompt consideration of ICU admission, especially when 2 or more variables coexist.14

TREATMENT

DISCHARGE/FOLLOW‐UP PLANS

Patients hospitalized for community‐acquired pneumonia can be safely discharged when they have reached clinical stability, are able to tolerate oral medications, have no other active comorbid conditions, and have safe, close, appropriate outpatient follow‐up (see Table 7). Clinical pathways employing these discharge criteria have been found to be safe and effective in reducing the length of stay for CAP. Most important, patients should have met most if not all of the vital sign and clinical criteria noted in Table 5 in the criteria for switching to oral therapy. Patients with 2 or more abnormal vital signs (instabilities) within 24 hours prior to discharge are at high risk of readmission and mortality, but those with one or no abnormal vital signs generally have good outcomes.65 Absent other clinical factors or extenuating circumstances (persistent hypoxia, poor functional status, etc.), most patients with CAP should reach clinical stability by day 3 or 4, be considered for a switch to oral therapy, and, if stable, be discharged shortly thereafter.

When patients with CAP are discharged from the hospital, they should be counseled about the expected course of recovery. Most important, patients and families must be informed that many symptoms of CAP may persist well after hospitalization. In one study, up to 80% of patients reported persistent cough and fatigue 1 week after discharge, and up to 50% still had dyspnea and sputum production. In some, the cough can last for 46 weeks.8

All patients discharged after treatment of community‐acquired pneumonia should have follow‐up with their outpatient provider. The physician responsible for their inpatient care should communicate directly with the provider and outline the hospital course, the discharge medications, and the duration of antibiotic therapy. There is no specific time frame within which patients must be seen, but follow‐up should be dictated by patient age, comorbidities, clinical stability at discharge, and degree of illness. The American Thoracic Society guidelines do recommend patients with a substantial smoking history who are hospitalized with CAP have a follow‐up chest radiograph 46 weeks after discharge to establish a radiographic baseline and exclude the possibility of underlying malignancy.14 However, several studies have suggested that radiographic resolution may take 3 or more months in some patients, especially the elderly and those with multilobar disease.66

PREVENTION

Prevention of community‐acquired pneumonia and pneumonic syndromes has traditionally relied on vaccination with the polysaccharide pneumococcal pneumonia vaccine and the seasonal influenza vaccine. The vaccine for S. pneumoniae used in adults is composed of the 23 serotypes that cause 85%90% of the invasive pneumococcal infections in the United States. Although in randomized trials the vaccine has not consistently prevented community‐acquired pneumonia or death in elderly patients or those with comorbidities, it likely prevents invasive pneumococcal infection.67 National guidelines and the CDC recommend the pneumococcal vaccine be given to all patients older than 65 years and those with chronic medical conditions.1, 14, 15

The seasonal influenza vaccine has clearly been shown to decrease influenza‐related illness in elderly and high‐risk patient populations. As well, in a meta‐analysis and a large observational study of patients older than 65 years, vaccination against influenza prevented pneumonia, hospitalization, and death.68, 69 Vaccination of health care workers may also confer a benefit to elderly patients of reduced mortality. The CDC recommends the influenza vaccine for all patients more than 50 years old, those with comorbidities, those at high risk for influenza, and health care workers in both inpatient and outpatient settings.

Pneumococcal and influenza vaccination have traditionally been relegated to the outpatient setting. National guidelines and the CDC recommend vaccination of all eligible hospitalized patients. Vaccination is safe and effective with almost any medical illness, and both vaccines can be given simultaneously at discharge.69 Both JCAHO and CMS have defined administration of the pneumococcal and influenza vaccines to patients hospitalized with CAP as a quality measure. Using standing orders is the most effective means of ensuring vaccination.

Some evidence suggests that tobacco smokers are at increased risk of invasive pneumococcal disease or pneumonia.70 Patients hospitalized (for all illnesses, but for CAP in particular) should be counseled about smoking cessation and offered pharmacotherapy and outpatient follow‐up. And, finally, recent observational data suggests that use of acid suppressive therapy, including proton pump inhibitors and H‐2 receptor antagonists, may be associated with an increased risk of developing CAP.71 Patients using these agents who are admitted with CAP should have their indications for treatment reviewed, especially when the pneumonia has been recurrent and there is no clear indication for continued use of acid suppressive therapy, in which case they should be discontinued in the hospital.

CONCLUSIONS

Community‐acquired pneumonia remains a common cause for hospitalization of adult patients, with significant associated morbidity and mortality. Although there are multiple expert guidelines for the management of community‐acquired pneumonia, further research is urgently needed. Clinicians need improved diagnostic tests that enable an earlier and more accurate diagnosis of CAP. In addition, the etiologic agent causing CAP is rarely discovered; improved microbiologic studies might enable antibiotic therapy to be targeted to the organisms responsible. High‐quality randomized, controlled trials examining empiric antibiotic therapy in CAP are needed, especially related to the addition of agents covering atypical organisms. Last, the general management of patients hospitalized with CAP is marked by significant heterogeneity, and research and initiatives focusing on improving the quality and process of care of patients with CAP are needed.

Community‐acquired pneumonia (CAP) is commonly defined as an infection of the pulmonary parenchyma that is associated with at least some symptoms and signs of acute infection, accompanied by the presence of an acute infiltrate on chest radiograph, in a patient not hospitalized or residing in a long‐term‐care facility for 14 days prior to the onset of symptoms.1 CAP continues to be a common and serious illness, causing substantial morbidity and mortality in the adult population. There are an estimated 56 million cases a year in the United States, with greater than 1 million hospitalizations. Community‐acquired pneumonia is one of the most common admitting diagnoses among adults, and with a 30‐day mortality between 10% and 14% for patients admitted to the hospital, it is the leading cause of infectious death in the United States.2 In elderly patients, hospitalization for CAP portends a poor long‐term prognosis. In a Medicare database, the 1‐year mortality for patients with CAP was nearly 40%, compared to 29% in patients with other diagnoses.3 Community‐acquired pneumonia is a model illness in hospital medicineit is a common disease that allows for evidence‐based and cost‐effective management. In addition, many national organizations have proposed multiple quality indicators for community‐acquired pneumonia, thus providing an opportunity for institutional quality improvement. This review article outlines the assessment and management of patients admitted to the hospital with community‐acquired pneumonia.

Etiology

Although many pathogens can cause community‐acquired pneumonia, the clinical syndromes and microbiology of CAP have traditionally been characterized as either typical or atypical. The typical organisms include Streptococcus pneumoniae, Haemophilus influenzae, and Moraxella catarrhalis, and the atypical organisms include Chlamydia spp., Mycoplasma pneumoniae, Legionella spp., and viruses. This historical distinction has recently come into question. It is now clear that the presenting symptoms, signs, and basic laboratory findings (including the chest radiograph) cannot be reliably used to predict the etiologic pathogen or to distinguish typical from atypical organisms.4 Rather, the specific causative agent of CAP depends more on the degree of patient illness. Table 1 shows what prospective studies with comprehensive diagnostic strategies determined to be the most common pathogens in patients hospitalized for CAP in ICU and non‐ICU settings.5 Streptococcus pneumoniae remains the most common cause of CAP in hospitalized patients and is the most common cause of fatal pneumonia, whereas Legionella spp. is a common cause of severe CAP, more often found in patients requiring admission to the intensive care unit. Gram‐negative bacilli can cause CAP in elderly patients and those recently treated with broad‐spectrum antibiotics or with underlying lung disease. Notably, though, despite improved diagnostic testing, only one quarter of all admitted patients with CAP have the etiologic agent defined, and therefore empiric therapy should be directed broadly at the most likely organisms.6

Clinical Presentation

Patients admitted to the hospital with CAP typically present with a brief history of respiratory complaints, including cough (greater than 90%), dyspnea (66%), sputum production (66%), and pleuritic chest pain (50%); see Table 2.7, 8 In 10%30% of patients, nonrespiratory complaints predominate, including headache, myalgias, fatigue, and gastrointestinal symptoms.6 Elderly patients, an increasing percentage of hospitalized patients, are less likely to present with typical CAP symptoms (such as cough) and more likely to have altered mental status as a presenting symptom.9

On physical examination, patients with CAP usually have signs of fever (80%), tachypnea (70%), and tachycardia (50%); see Table 2. Most will have a focal lung exam (>90%) with findings ranging from crackles to bronchial breath sounds.10 No exam finding is specific for the diagnosis of pneumonia, but the absence of fever, tachycardia, and tachypnea significantly reduces the probability of CAP in patients with suspected pneumonia.10 Furthermore, similar to the clinical history, the physical examination of elderly patients with community‐acquired pneumonia is not specific or sensitive for the diagnosis of CAP. For example, up to 40% of elderly patients subsequently determined to have CAP may not have fever.11

Leukocytosis is common in patients with CAP; however, its absence does not rule out disease.12 A number of guidelines recommend laboratory evaluation of electrolytes, urea nitrogen, creatinine, liver enzymes, and bilirubin, although these are used primarily for prognostication and are not specifically useful in the diagnosis of CAP.

DIAGNOSIS

Diagnostic Studies

The diagnosis of community‐acquired pneumonia requires that a patient have both signs and symptoms consistent with pulmonary infection and evidence of a new radiographic infiltrate. Therefore, most guidelines recommend that all patients with a possible diagnosis of CAP be evaluated with chest radiography.1, 14, 15

The specific radiographic findings in community‐acquired pneumonia range from lobar consolidation to hazy focal infiltrate to diffuse bilateral interstitial opacities (see Figure 1). Although chest radiography has traditionally been considered the gold standard for the diagnosis of CAP, its exact performance characteristics are unknown, and it is clearly not 100% sensitive or 100% specific. The utility of the chest radiograph can be limited by patient body habitus, underlying lung disease, or dehydration. Computed tomography (CT) scanning, although not recommended for routine use, can identify pulmonary consolidation in up to 30% of patients with a normal or equivocal chest radiograph in whom pneumonia is suspected and can also identify complications of pneumonia including an empyema or pulmonary abscess.16

Figure 1

Limitations in the performance of the chest radiograph have resulted in an interest in the diagnostic performance of serologic markers of infection such as C‐reactive protein (CRP), procalcitonin, and soluble triggering receptor expressed on myeloid cells (s‐TREM).1719 Preliminary evidence suggests these inflammatory markers may ultimately prove useful in differentiating infectious from noninfectious pulmonary processes, but regular use of these new tests cannot currently be recommended.

Most expert guidelines state that 2 sets of blood cultures should be taken and analyzed prior to antibiotic administration in all patients admitted to the hospital with suspected community‐acquired pneumonia.1, 14, 15 Isolation of bacteria from blood cultures in CAP is a very specific way to identify a causative organism in order to subsequently narrow therapy and also identifies a high‐risk group of patients because bacteremia is associated with increased mortality. Obtaining blood cultures within 24 hours of admission has been associated with 10% lower odds of 30‐day mortality in patients with CAP,20 and as a result, drawing blood cultures prior to antibiotic administration is a national quality indicator for CAP.

There are, however, a number of problems with the routine acquisition of blood cultures in all patients admitted with CAP. Practically, the cultures can be difficult to obtain, can potentially delay the initiation of antibiotics, and are often contaminated, which has been shown to increase both cost and length of stay.21, 22 The yield is generally low: the true‐positive bacteremia rate for admitted patients with CAP ranges from 6% to 9%, and the culture results rarely change management or outcomes.23, 24 Given these limitations, many have argued that blood cultures should be obtained with a more targeted approach. A recent study used a Medicare database to create a decision‐support tool to help maximize the value of blood cultures in CAP.25 The predictors of a positive blood culture are shown in Table 3. Not obtaining cultures on patients who had received prior antibiotics or had no risk factors resulted in about 40% fewer overall cultures while identifying approximately 90% of bacteremias. In their guidelines, the British Thoracic Society (BTS) advocates a similar strategy, recommending blood cultures be omitted in nonsevere pneumonia and in patients without comorbidities.15, 26 Although recommendations vary for non‐severe CAP in hospitalized patients, all professional society guidelines agree that blood cultures should be obtained in critically ill patients, and if cultures are obtained, they should be drawn prior to antibiotics.1, 14, 15, 26

Table 3.Independent Predictors of Bacteremia in Patients with Community‐Acquired Pneumonia24

Comorbidities
Liver disease
Vital signs
Systolic blood pressure 90 mm Hg
Temperature 35C or 40C
Pulse 125 beats/min
Laboratory and radiographic data
Blood urea nitrogen (BUN) 30 mg/dL
Sodium 130 mmol/L
White blood cells 5000/mm3 or > 20,000/mm3
Prior use of antibiotics (negative predictor)

Substantial controversy surrounds the utility of routine sputum gram stains and cultures for patients admitted to the hospital with CAP. The Infectious Disease Society of America (IDSA) and the British Thoracic Society (BTS) both recommend that all patients admitted to the hospital with community‐acquired pneumonia should have a gram stain and culture of expectorated sputum.1, 15, 26 Both organizations argue sputum collection is a simple and inexpensive procedure that can potentially identify pathogenic organisms and can affect both initial and long‐term antibiotic therapy. Most notably, they highlight gram stain specificity of greater than 80% for pneumococcal pneumonia. Conversely, the American Thoracic Society (ATS) argues that sputum gram stains and cultures generally have low sensitivity, specificity, and positive predictive value.14 Furthermore, they argue the utility of sputum testing is also limited practically; in one study 30% of patients could not produce an adequate sputum specimen and up to 30% had received prior antibiotic therapy, substantially reducing the yield.27 In another study, good‐quality sputum with a predominant morphotype could be obtained in only 14% of patients admitted with CAP.28 However, targeting sputum analysis to patients who have not received prior antibiotics and are able to produce an adequate sample improved the yield significantly.29 In addition, with increasing rates of antibiotic resistance among common community isolates (ie, S. Pneumoniae) and the increasing prevalence of infecting organisms not targeted by routine empiric therapy (methicillin‐resistant Staphylococcus Aureus [MRSA]), isolation of potential causative pathogens is increasingly important. We believe that severely ill patients with CAP (such as patients admitted to the ICU), as well as patients with identifiable risk factors for uncommon or drug‐resistant pathogens (eg, Pseudomonas aeruginosa, enteric gram‐negative rods, MRSA, etc.) should have sputum sent for gram stain and culture. Ideally, sputum obtained for gram stain and culture should be:

• 1

  Prior to antibiotic therapy,

• 2

  A deep‐cough, expectorated specimen,

• 3

  A purulent specimen (>25 polymorphonucleacytes and less than 10 squamous cells per high‐powered field), and

• 4

  Rapidly transported to the laboratory.

Subsequent gram stain and culture results should be interpreted in the specific clinical context and antibiotic choices targeted appropriately.

ADMISSION DECISION

Once the diagnosis of CAP has been made, the initial site where treatment will occur, whether the hospital or the home, must be determined. The decision to hospitalize should be based on 3 factors: 1) evaluation of the safety of home treatment, 2) calculation of the Pneumonia Severity Index (PSI), and 3) clinical judgment of the physician.27 The PSI, or PORT (Pneumonia Outcomes Research Team) score, is a validated prediction rule that quantifies mortality and allows for risk stratification of patients with community‐acquired pneumonia.2 The PSI combines clinical history, physical examination, and laboratory data at the time of admission to divide patients into 5 risk classes and to estimate 30‐day mortality (Figure 2), which ranges from 0.1% of patients in risk class I to 27.0% in risk class V.2

Figure 2

On the basis of the estimated prognosis and in the absence of concerns about home safety or comorbidities, patients in risk classes I, II, and III should be managed at home. Many prospective trials have shown that implementation of PSI significantly increases the number of low‐risk patients managed outside the hospital, with no differences in quality of life, complications, readmissions, or short‐term mortality.30, 31 Most recently, a trial randomizing patients in risk classes II and III to treatment in the hospital or at home found no significant differences in clinical outcomes but did find that patients were more satisfied with care at home.32 Because the number of patients with CAP being treated at home is increasing, the American College of Chest Physicians recently published a consensus statement on the management of community‐acquired pneumonia in the home.33 All national guidelines for the management of community‐acquired pneumonia recommend using the PSI to help determine the initial location of treatment, with the caveat that using the prediction rule should never supersede clinical judgment in the decision about whether to admit.1, 14, 15, 26, 27 A practical decision tree for the use of the PSI is shown in Figure 3.

Figure 3

There are no reliable prediction rules for deciding on whether admission to the intensive care unit is necessary. Hemodynamic instability requiring resuscitation and monitoring or respiratory failure requiring ventilatory support are clear indications for ICU admission. Additional variables such as tachypnea (respiratory rate 30), altered mental status, multilobar disease, and azotemia are associated with severe CAP and should prompt consideration of ICU admission, especially when 2 or more variables coexist.14

TREATMENT

DISCHARGE/FOLLOW‐UP PLANS

Patients hospitalized for community‐acquired pneumonia can be safely discharged when they have reached clinical stability, are able to tolerate oral medications, have no other active comorbid conditions, and have safe, close, appropriate outpatient follow‐up (see Table 7). Clinical pathways employing these discharge criteria have been found to be safe and effective in reducing the length of stay for CAP. Most important, patients should have met most if not all of the vital sign and clinical criteria noted in Table 5 in the criteria for switching to oral therapy. Patients with 2 or more abnormal vital signs (instabilities) within 24 hours prior to discharge are at high risk of readmission and mortality, but those with one or no abnormal vital signs generally have good outcomes.65 Absent other clinical factors or extenuating circumstances (persistent hypoxia, poor functional status, etc.), most patients with CAP should reach clinical stability by day 3 or 4, be considered for a switch to oral therapy, and, if stable, be discharged shortly thereafter.

When patients with CAP are discharged from the hospital, they should be counseled about the expected course of recovery. Most important, patients and families must be informed that many symptoms of CAP may persist well after hospitalization. In one study, up to 80% of patients reported persistent cough and fatigue 1 week after discharge, and up to 50% still had dyspnea and sputum production. In some, the cough can last for 46 weeks.8

All patients discharged after treatment of community‐acquired pneumonia should have follow‐up with their outpatient provider. The physician responsible for their inpatient care should communicate directly with the provider and outline the hospital course, the discharge medications, and the duration of antibiotic therapy. There is no specific time frame within which patients must be seen, but follow‐up should be dictated by patient age, comorbidities, clinical stability at discharge, and degree of illness. The American Thoracic Society guidelines do recommend patients with a substantial smoking history who are hospitalized with CAP have a follow‐up chest radiograph 46 weeks after discharge to establish a radiographic baseline and exclude the possibility of underlying malignancy.14 However, several studies have suggested that radiographic resolution may take 3 or more months in some patients, especially the elderly and those with multilobar disease.66

PREVENTION

Prevention of community‐acquired pneumonia and pneumonic syndromes has traditionally relied on vaccination with the polysaccharide pneumococcal pneumonia vaccine and the seasonal influenza vaccine. The vaccine for S. pneumoniae used in adults is composed of the 23 serotypes that cause 85%90% of the invasive pneumococcal infections in the United States. Although in randomized trials the vaccine has not consistently prevented community‐acquired pneumonia or death in elderly patients or those with comorbidities, it likely prevents invasive pneumococcal infection.67 National guidelines and the CDC recommend the pneumococcal vaccine be given to all patients older than 65 years and those with chronic medical conditions.1, 14, 15

The seasonal influenza vaccine has clearly been shown to decrease influenza‐related illness in elderly and high‐risk patient populations. As well, in a meta‐analysis and a large observational study of patients older than 65 years, vaccination against influenza prevented pneumonia, hospitalization, and death.68, 69 Vaccination of health care workers may also confer a benefit to elderly patients of reduced mortality. The CDC recommends the influenza vaccine for all patients more than 50 years old, those with comorbidities, those at high risk for influenza, and health care workers in both inpatient and outpatient settings.

Pneumococcal and influenza vaccination have traditionally been relegated to the outpatient setting. National guidelines and the CDC recommend vaccination of all eligible hospitalized patients. Vaccination is safe and effective with almost any medical illness, and both vaccines can be given simultaneously at discharge.69 Both JCAHO and CMS have defined administration of the pneumococcal and influenza vaccines to patients hospitalized with CAP as a quality measure. Using standing orders is the most effective means of ensuring vaccination.

Some evidence suggests that tobacco smokers are at increased risk of invasive pneumococcal disease or pneumonia.70 Patients hospitalized (for all illnesses, but for CAP in particular) should be counseled about smoking cessation and offered pharmacotherapy and outpatient follow‐up. And, finally, recent observational data suggests that use of acid suppressive therapy, including proton pump inhibitors and H‐2 receptor antagonists, may be associated with an increased risk of developing CAP.71 Patients using these agents who are admitted with CAP should have their indications for treatment reviewed, especially when the pneumonia has been recurrent and there is no clear indication for continued use of acid suppressive therapy, in which case they should be discontinued in the hospital.

CONCLUSIONS

Community‐acquired pneumonia remains a common cause for hospitalization of adult patients, with significant associated morbidity and mortality. Although there are multiple expert guidelines for the management of community‐acquired pneumonia, further research is urgently needed. Clinicians need improved diagnostic tests that enable an earlier and more accurate diagnosis of CAP. In addition, the etiologic agent causing CAP is rarely discovered; improved microbiologic studies might enable antibiotic therapy to be targeted to the organisms responsible. High‐quality randomized, controlled trials examining empiric antibiotic therapy in CAP are needed, especially related to the addition of agents covering atypical organisms. Last, the general management of patients hospitalized with CAP is marked by significant heterogeneity, and research and initiatives focusing on improving the quality and process of care of patients with CAP are needed.

Toyota is widely recognized as one of the most successful companies in the world. Its automobiles have consistently placed at or near the top of the quality and customer satisfaction rankings published by J.D. Power and Associates and Consumer Reports. Toyota constantly focuses on the safety and well‐being of its employees and the quality of its cars through its relentless dedication to continuous improvement in everything it does. Toyota has recently become the world's number two auto manufacturer, and the company's net profit margin was more than 8 times that of the industry average. 1

How has Toyota been able to achieve such remarkable results in product quality, market share, and profit margins? Jeffrey Liker, in his book The Toyota Way, described the world‐renowned Toyota production system as supported by 2 pillars: continuous improvement and respect for people. The end result is a learning organization that values employee contributions and continuously strives to produce products of higher quality at lower cost. 1, 2 Lean production is the generic term used to describe the principles and methods of the Toyota Production System. Lean production has been implemented to improve performance in a broad array of industries, from aerospace and aluminum refining to financial services and insurance. The philosophy of lean thinking, which is derived from the Toyota Production System, is rapidly gaining a following among health care leaders, with a number of hospitals and medical groups around the country adopting a version of lean production as their systematic approach to improving quality and efficiency. In the coming years, the application of lean principles and methods could have a transformational effect on how health care is delivered, with the potential for dramatic gains in quality, safety, efficiency, and appropriateness.

LEAN CONCEPTS

To understand how lean production can be applied to improve the delivery of health care, some of the fundamental concepts and practice of lean must first be explained. 3, 4 The first step in a lean improvement initiative is to understand value as defined by our customers. 5 In clinical care delivery, external customers include patients, families, payers, and regulators. Internal customers include physicians, nurses, clerks, and others involved in the care process. What customers value usually includes care that is of high quality, safe, efficient and appropriate. The second step typically is to go to the workplace and observe firsthand how the process now operates. 6 As the flow of the process from beginning to end is seen, the observer learns to see and to understand the multiple areas of delay, inefficiency, and waste that may exist. 7, 8 A representational flowchart called a current‐state value stream map (CS VSM) is created to make the work visible and to depict graphically all the individual steps necessary to complete the process from beginning to end. It is important that the CS VSM be a factual depiction of how an entire process flows created by those who actually work in that process. The CS VSM does not state any exceptions to or provide any explanations for why certain steps are taken. It does include key measures such as process time (the actual time it takes to complete a particular step of the process), lead time (the total time it takes to complete the entire process, including waiting time), and first‐time quality (the percentage of time in which that step of the process is completed without defect); (see Figure 1).

Figure 1

In the hands of an improvement team, the current state map becomes a powerful tool that allows participants to systematically recognize and categorize waste. The CS VSM also allows workers to visualize how much opportunity there is for improving the existing process. Working from the CS VSM, the team members can identify specific areas of waste, delay, causes of error, and inefficiency. The team then brainstorms ideas for improvement, proposing how steps of the process might be combined, eliminated, error‐proofed, or otherwise improved to transform waste into value from the customer's perspective. In the third step, the team seeks to achieve the flow state in which the steps of the process follow one another without stopping. All ideas are welcomed at this stage and are placed on the current state map itself or arrayed elsewhere for consideration. Using the ideas generated by the team, a new and better process is designed and depicted on a flow map called the future‐state value stream map (FS VSM). 9, 10 The FS VSM represents an improved and streamlined or ideal way in which the process could be accomplished, as best the team was able to envision at this point (see Figure 2). Ideally, the process described in the FS VSM also allows customers to pull value when they need goods or services provided by the organization, rather than having to do the usual requesting and waiting seen in health care and other service industries. Creating processes from which customers pull what they need is the fourth step in lean design.

Figure 2

Once a future state map is devised and approved, the critical work of rapid deployment of an implementation plan for reaching the future state begins. An implementation plan explicitly identifies who is responsible for what aspect of implementation. Usually a senior leader or leadership group sponsoring the project is responsible for encouraging team members to think beyond their historical (and often political) limits and to support the team in overcoming barriers outside its control. As the individuals return to work and attempt to implement the new solutions, however, they will likely encounter areas of resistance and ambiguity that require creative solutions. The implementation phase focuses on and encourages the individual worker to experiment and work toward a solution that can be broadly adopted and disseminated for use as a standardized solution by all workers facing similar situations. 11, 12 Through this experimental development and dissemination of solutions, the agreed‐to future‐state map is revised. In this way the old future‐state map plays the role of the new current‐state map. There is an ongoing, continuous loop between the current‐ and future‐state maps through implementation and testing to develop the ideal way in which the process should flow toward the final product or service (see Figure 3). The fifth step, pursuing perfection, requires this continuous loop of all workers improving everything they do, every day. The hardest of all the steps, pursuing perfection requires an organization to commit to process improvement and the elimination of defects and waste on a daily and permanent basis. 5

Figure 3

The steps outlined above provide only the basic foundations of lean concepts, of course. A detailed description of learning about and applying lean within one's own organization requires further study and help. Readers interested in learning more about value stream mapping are referred to a workbook by Mike Rother and John Shook called Learning to See, Value Stream Mapping to Create Value and Eliminate Muda. 4 Further, there are consultants with lean expertise who are available to help hospitals get started on the lean journey.

The management philosophy of lean production methods has ties to other operational and quality‐improvement models such as total quality management (TQM)/continuous quality improvement (CQI), developed by W. E. Deming, and Six Sigma, developed by Motorola and General Electric. Although there are several overlapping points of philosophy and techniques, a feature distinguishing lean from these other models is in its value stream approach to driving change and eliminating waste within the process of providing a product for the customer. Lean is unique in its focus on the specification of value from the customer's perspective and on the identification and categorization of waste and its transformation to value using specific tools. The lean approach encourages individuals within the organization (from top to bottom) to learn to see the flow of their product's process and thus to help to identify areas of waste, with the ultimate goal of creating a product with built‐in quality with the least amount of waste. Both Six Sigma and TQM/CQI focus on the delivery of a high‐quality product. Once a system has been studied and standardized, Six Sigma utilizes rigorous statistical and data measurements to drive quality improvements in product delivery. 13 Total quality management/continuous quality improvement engages the entire organization in delivering a high‐quality product from the customer's standpoint by getting everyone in the organization involved in the continuous improvement effort. 14 Lean thinking builds directly on the plandocheckact cycle of CQI but adds tools to identify and transform waste, supplies metrics of timing and resources, and, most important, focuses intently on the creation of value as defined by the customer. A more detailed comparison of lean philosophy with these other quality‐improvement management philosophies is beyond the scope of this article on the introduction of lean methods for hospitals.

DOES HEALTH CARE NEED LEAN THINKERS?

So, should health care try to emulate the successes of an automobile manufacturing company? Before answering this question, consider the following about the health care system as we know it and the significant challenges it faces.

• A key take‐home message of the Institute of Medicine's 1999 report, To Err Is Human: Building a Safer Health System, was that errors are caused by poorly designed systems. 15

• The Centers for Medicare and Medicaid Services (CMS) reported that the cost of health care is rising rapidly and that the rate of growth is not sustainable. 16

• Studies by the RAND Corporation demonstrated considerable variability in the practice of medicine, raising important questions about the appropriateness and necessity of some medical care and procedures. 17, 18

• In Crossing the Quality Chasm: a New Health System for the 21st Century, the Institute of Medicine concluded that today's health care system functions at far lower levels than it can and should and recommended 6 aims for improving the health care system: health care should be safe, effective, patient centered, timely, efficient, and equitable. 19 This report is rich in detail about what the ideal health care system should look like and how application of lean philosophy and tools can help hospitals and physicians achieve that vision.

Although these challenges reflect a view of the health care system at a very macro level, the mechanism and process of driving change will need to be initiated at the more local and organizational level. Lean production focuses on the goal of continuously transforming waste into value from the customer's perspective. It provides a rigorous and systematic approach to process improvement, error proofing, and waste reduction. Manufacturing companies such as Toyota and Alcoa and financial service organizations such as Vanguard have enjoyed tremendous success in implementing lean production, reporting gains in both quality and efficiency. 11 It is time for health care leaders and practitioners to evaluate how lean techniques can be adapted and applied to addressing the pressing challenges of safety, quality, efficiency, and appropriateness in order to improve system reliability and timeliness. As hospitalists, we are at the forefront of these challenges and therefore are in a prime position to lead the change in how health care delivery is improved continuously using a rigorous system such as the Toyota production system.

LEAN IN HEALTH CARE: EXAMPLES OF SOME EARLY RESULTS

Lean health care is still a very novel concept to most health care institutions; however, there have been some early adopters of lean health care, and some of their experiences are described here:

• At Virginia Mason Medical Center (VMMC) in Seattle, Washington, changes implemented using lean production methods have resulted in decreased incidence of ventilator‐associated pneumoniafrom 34 cases with 5 deaths in 2002 to 4 cases with 1 death in 2004. This led to a cost reduction of nearly a half‐million dollars. VMMC has also reported increased profit margins and improvement in space utilization at its cancer center, enabling 57% more patients to be seen in the same allotted space; and it is now taking measures to decrease the number of medication errors by standardizing and mistake‐proofing the process of ordering, delivering, and administering medications, all using lean techniques. 12, 20

• At Park Nicollet Health Services (PNHS) in Minneapolis, Minnesota, implementation of lean production has enabled improved patient access through flow improvements. Results include increasing the number of CT and MRI scans performed per day by 2 and 1, respectively; creating a capacity for 10 additional chemotherapy and antibiotic infusion patients per day in the cancer center; reducing the waiting time of patients from 122 to 52 minutes at the urgent care clinic; standardizing surgical instrument use by the general surgery group, which resulted in processing more than 40,000 fewer instruments each month. These improvements achieved through applying lean concepts have resulted in Park Nicollet being recognized by the American Medical Group Association (AMGA) with its top‐rated Acclaim Award. 21 In addition, PNHS has been able to achieve a record 3.9% operating margin, which equates to a $7.5 million profit in 2004.

• In Pittsburgh, Pennsylvania, a group of hospitals participating in the Pittsburgh Regional Healthcare Initiative (PRHI) have implemented lean concepts to minimize the risk of developing central catheterrelated bloodstream infections. Several hospitals have been able to cut the incidence of central line infections by 50%‐90% through implementation of lean production methods. 12

• At Community Medical Center in Missoula, Montana, a series of pilot projects have been initiated to test lean methods. Some of the early results have demonstrated a reduction in turnaround time for pathology reports from the anatomical pathology lab from 5 to 2 days, a reduction in the number of steps and therefore the time from medication order to treatment initiation from 4 hours to 12 minutes, and a reduction in time for unit clerks to process new physician orders from an average of 43 minutes to 10 minutes during the hospital's busiest hours. 22

THE MICHIGAN LEAN EXPERIENCE

In the past year, the University of Michigan has begun to use lean production methods to improve the care of patients across various venues of hospitalization and flow toward discharge. Delays in placement of peripherally inserted central catheters (PICC) were associated with delays in appropriate and timely administration of intravenous medication, as well as in delays in discharges home or to extended‐care facilities (ECF) for continuation of medical care post‐hospitalization. Since the initiation of the lean PICC initiative, when adjusted for increased volume of demand, for 3 consecutive months 90%95% of the PICC lines have been inserted within 24 hours of request. This is a remarkable achievement, given that in the previous 12 months only 50%70% of PICC lines were placed within 24 hours of request. Even without adjusting for volume of demand, the lean PICC initiative has resulted in a 36% decrease in the average time to line placement and in a 50% decrease in the number of PICC referrals to interventional radiology (IR), thus decreasing the workload of a constrained resource.

As with all lean improvement projects, the entire value stream map was assessed in order to identify areas of intervention that would enhance the final product of the process for the patient (in this case, placing a PICC line as timely as possible). As we evaluated this value stream, one step in the process, occurring prior to placement of the line, appeared to be significantly wasteful: when the PICC nurse needed to search for data (such as locating a patient's chart for the order and reviewing labs and medication records) and to ensure that the patient was in his/her room and prepared for line placement. This step appeared to be inefficient use of the time of technically skilled individuals. The future state map of this process implemented the addition of an assisting individual who would ensure that these prework issues were prepared and completed in advance for the PICC nurses, thus making maximum use of the time of these skilled individuals in placing PICC lines, not in performing unskilled work. Another area where an intervention was believed beneficial was streamlining the process of chest x‐ray (CXR) ordering and reading in order to obtain PICC line confirmation. Performance of the previous process was not standardized and led to delays. The future state proposed a standard method of writing an order for a CXR, a standard method for getting that order to the radiology department, and a standard method for reading the films for dictation. This prevented the confusion and rework that had previously occurred. A last example of an intervention is that the PICC nurses began to internally defer to more experienced nurses if a less experienced nurse could not successfully place a PICC line in a patient. Previously, an unsuccessful attempt at bedside PICC placement warranted an IR referral, thus increasing the demand on an already constrained resource. This intervention by the PICC nurses drove down referrals to the IR suite by 50%. Although this may have led to a small increase in rework early in the process, it has led to a significant reduction in work downstream in the process. Thus, we believe that the overall work flow process has been served well by this intervention. As depicted in Figure 3, the lean process improvement method seeks to have continuous improvement, with the old future‐state map taking on the role of the new current‐state map. Since the initial development of these value stream maps, we have been working toward developing and implementing new areas of intervention, which will lead to new future‐state maps and to further improvement of this process, as the demand for PICC lines continues to rise.

Another critical segment in the care of hospitalized patients is the discharge process, including coordination of care to an outpatient or extended care facility (ECF) setting, which has several potential areas of disconnect that could result inpatients having untoward complications requiring rehospitalization, higher morbidity, or prolongation of suffering from their illnesses. The University of Michigan sees a tremendous opportunity to make a significant impact on patient care in this realm and has just initiated a lean project on the coordination of care. Team members on this project will be relevant process stakeholders, including those representing hospitalists, discharge planning, nursing, social work, a related home nursing company, a home infusion service, an ECF, ambulatory care, pharmacy, case management, nutrition, utilization review, patients and their families, and clinic physicians. The overall goal of the project is to optimize patient care from hospitalization to discharge and transfer of care to the outpatient setting.

CONCLUSION

The health care industry should learn about and consider adoption of lean techniques in order to improve its processes. More specifically, hospitals are prime locales for reaping the benefits of implementation of lean production, which can significantly affect how health care is delivered to patients. Toyota and other lean exemplars in the manufacturing industry have achieved a high level of success by utilizing the practice of lean. Early results from health care organizations suggest that utilizing lean production methods can lead to substantial improvements in the quality and efficiency of health care. To determine if the magnitude of success experienced by Toyota and other lean exemplars can also be achieved in the health care sector, it will be necessary to continuously test and evaluate the impact lean health care can have. In the hospital setting, where hospitalists are at the forefront of delivering care, it is incumbent on the hospitalist community to evaluate whether these techniques can make a difference in the quality, efficiency, and safety of the care provided to patients.

Lean thinking is still a novel idea to those in the health care sector, and as early adopters of this promising management model, we are very optimistic about the benefits of applying lean concepts in our hospital. Some of the first published reports and results presented on the benefits of lean in individual organizations are encouraging; however, as health care is a scientific community, we believe that future work should undergo rigorous evaluation on the benefits of lean and that such future works should be shared among the health care community through peer‐reviewed and published works.

Toyota is widely recognized as one of the most successful companies in the world. Its automobiles have consistently placed at or near the top of the quality and customer satisfaction rankings published by J.D. Power and Associates and Consumer Reports. Toyota constantly focuses on the safety and well‐being of its employees and the quality of its cars through its relentless dedication to continuous improvement in everything it does. Toyota has recently become the world's number two auto manufacturer, and the company's net profit margin was more than 8 times that of the industry average. 1

How has Toyota been able to achieve such remarkable results in product quality, market share, and profit margins? Jeffrey Liker, in his book The Toyota Way, described the world‐renowned Toyota production system as supported by 2 pillars: continuous improvement and respect for people. The end result is a learning organization that values employee contributions and continuously strives to produce products of higher quality at lower cost. 1, 2 Lean production is the generic term used to describe the principles and methods of the Toyota Production System. Lean production has been implemented to improve performance in a broad array of industries, from aerospace and aluminum refining to financial services and insurance. The philosophy of lean thinking, which is derived from the Toyota Production System, is rapidly gaining a following among health care leaders, with a number of hospitals and medical groups around the country adopting a version of lean production as their systematic approach to improving quality and efficiency. In the coming years, the application of lean principles and methods could have a transformational effect on how health care is delivered, with the potential for dramatic gains in quality, safety, efficiency, and appropriateness.

LEAN CONCEPTS

To understand how lean production can be applied to improve the delivery of health care, some of the fundamental concepts and practice of lean must first be explained. 3, 4 The first step in a lean improvement initiative is to understand value as defined by our customers. 5 In clinical care delivery, external customers include patients, families, payers, and regulators. Internal customers include physicians, nurses, clerks, and others involved in the care process. What customers value usually includes care that is of high quality, safe, efficient and appropriate. The second step typically is to go to the workplace and observe firsthand how the process now operates. 6 As the flow of the process from beginning to end is seen, the observer learns to see and to understand the multiple areas of delay, inefficiency, and waste that may exist. 7, 8 A representational flowchart called a current‐state value stream map (CS VSM) is created to make the work visible and to depict graphically all the individual steps necessary to complete the process from beginning to end. It is important that the CS VSM be a factual depiction of how an entire process flows created by those who actually work in that process. The CS VSM does not state any exceptions to or provide any explanations for why certain steps are taken. It does include key measures such as process time (the actual time it takes to complete a particular step of the process), lead time (the total time it takes to complete the entire process, including waiting time), and first‐time quality (the percentage of time in which that step of the process is completed without defect); (see Figure 1).

Figure 1

In the hands of an improvement team, the current state map becomes a powerful tool that allows participants to systematically recognize and categorize waste. The CS VSM also allows workers to visualize how much opportunity there is for improving the existing process. Working from the CS VSM, the team members can identify specific areas of waste, delay, causes of error, and inefficiency. The team then brainstorms ideas for improvement, proposing how steps of the process might be combined, eliminated, error‐proofed, or otherwise improved to transform waste into value from the customer's perspective. In the third step, the team seeks to achieve the flow state in which the steps of the process follow one another without stopping. All ideas are welcomed at this stage and are placed on the current state map itself or arrayed elsewhere for consideration. Using the ideas generated by the team, a new and better process is designed and depicted on a flow map called the future‐state value stream map (FS VSM). 9, 10 The FS VSM represents an improved and streamlined or ideal way in which the process could be accomplished, as best the team was able to envision at this point (see Figure 2). Ideally, the process described in the FS VSM also allows customers to pull value when they need goods or services provided by the organization, rather than having to do the usual requesting and waiting seen in health care and other service industries. Creating processes from which customers pull what they need is the fourth step in lean design.

Figure 2

Once a future state map is devised and approved, the critical work of rapid deployment of an implementation plan for reaching the future state begins. An implementation plan explicitly identifies who is responsible for what aspect of implementation. Usually a senior leader or leadership group sponsoring the project is responsible for encouraging team members to think beyond their historical (and often political) limits and to support the team in overcoming barriers outside its control. As the individuals return to work and attempt to implement the new solutions, however, they will likely encounter areas of resistance and ambiguity that require creative solutions. The implementation phase focuses on and encourages the individual worker to experiment and work toward a solution that can be broadly adopted and disseminated for use as a standardized solution by all workers facing similar situations. 11, 12 Through this experimental development and dissemination of solutions, the agreed‐to future‐state map is revised. In this way the old future‐state map plays the role of the new current‐state map. There is an ongoing, continuous loop between the current‐ and future‐state maps through implementation and testing to develop the ideal way in which the process should flow toward the final product or service (see Figure 3). The fifth step, pursuing perfection, requires this continuous loop of all workers improving everything they do, every day. The hardest of all the steps, pursuing perfection requires an organization to commit to process improvement and the elimination of defects and waste on a daily and permanent basis. 5

Figure 3

The steps outlined above provide only the basic foundations of lean concepts, of course. A detailed description of learning about and applying lean within one's own organization requires further study and help. Readers interested in learning more about value stream mapping are referred to a workbook by Mike Rother and John Shook called Learning to See, Value Stream Mapping to Create Value and Eliminate Muda. 4 Further, there are consultants with lean expertise who are available to help hospitals get started on the lean journey.

The management philosophy of lean production methods has ties to other operational and quality‐improvement models such as total quality management (TQM)/continuous quality improvement (CQI), developed by W. E. Deming, and Six Sigma, developed by Motorola and General Electric. Although there are several overlapping points of philosophy and techniques, a feature distinguishing lean from these other models is in its value stream approach to driving change and eliminating waste within the process of providing a product for the customer. Lean is unique in its focus on the specification of value from the customer's perspective and on the identification and categorization of waste and its transformation to value using specific tools. The lean approach encourages individuals within the organization (from top to bottom) to learn to see the flow of their product's process and thus to help to identify areas of waste, with the ultimate goal of creating a product with built‐in quality with the least amount of waste. Both Six Sigma and TQM/CQI focus on the delivery of a high‐quality product. Once a system has been studied and standardized, Six Sigma utilizes rigorous statistical and data measurements to drive quality improvements in product delivery. 13 Total quality management/continuous quality improvement engages the entire organization in delivering a high‐quality product from the customer's standpoint by getting everyone in the organization involved in the continuous improvement effort. 14 Lean thinking builds directly on the plandocheckact cycle of CQI but adds tools to identify and transform waste, supplies metrics of timing and resources, and, most important, focuses intently on the creation of value as defined by the customer. A more detailed comparison of lean philosophy with these other quality‐improvement management philosophies is beyond the scope of this article on the introduction of lean methods for hospitals.

DOES HEALTH CARE NEED LEAN THINKERS?

So, should health care try to emulate the successes of an automobile manufacturing company? Before answering this question, consider the following about the health care system as we know it and the significant challenges it faces.

• A key take‐home message of the Institute of Medicine's 1999 report, To Err Is Human: Building a Safer Health System, was that errors are caused by poorly designed systems. 15

• The Centers for Medicare and Medicaid Services (CMS) reported that the cost of health care is rising rapidly and that the rate of growth is not sustainable. 16

• Studies by the RAND Corporation demonstrated considerable variability in the practice of medicine, raising important questions about the appropriateness and necessity of some medical care and procedures. 17, 18

• In Crossing the Quality Chasm: a New Health System for the 21st Century, the Institute of Medicine concluded that today's health care system functions at far lower levels than it can and should and recommended 6 aims for improving the health care system: health care should be safe, effective, patient centered, timely, efficient, and equitable. 19 This report is rich in detail about what the ideal health care system should look like and how application of lean philosophy and tools can help hospitals and physicians achieve that vision.

Although these challenges reflect a view of the health care system at a very macro level, the mechanism and process of driving change will need to be initiated at the more local and organizational level. Lean production focuses on the goal of continuously transforming waste into value from the customer's perspective. It provides a rigorous and systematic approach to process improvement, error proofing, and waste reduction. Manufacturing companies such as Toyota and Alcoa and financial service organizations such as Vanguard have enjoyed tremendous success in implementing lean production, reporting gains in both quality and efficiency. 11 It is time for health care leaders and practitioners to evaluate how lean techniques can be adapted and applied to addressing the pressing challenges of safety, quality, efficiency, and appropriateness in order to improve system reliability and timeliness. As hospitalists, we are at the forefront of these challenges and therefore are in a prime position to lead the change in how health care delivery is improved continuously using a rigorous system such as the Toyota production system.

LEAN IN HEALTH CARE: EXAMPLES OF SOME EARLY RESULTS

Lean health care is still a very novel concept to most health care institutions; however, there have been some early adopters of lean health care, and some of their experiences are described here:

• At Virginia Mason Medical Center (VMMC) in Seattle, Washington, changes implemented using lean production methods have resulted in decreased incidence of ventilator‐associated pneumoniafrom 34 cases with 5 deaths in 2002 to 4 cases with 1 death in 2004. This led to a cost reduction of nearly a half‐million dollars. VMMC has also reported increased profit margins and improvement in space utilization at its cancer center, enabling 57% more patients to be seen in the same allotted space; and it is now taking measures to decrease the number of medication errors by standardizing and mistake‐proofing the process of ordering, delivering, and administering medications, all using lean techniques. 12, 20

• At Park Nicollet Health Services (PNHS) in Minneapolis, Minnesota, implementation of lean production has enabled improved patient access through flow improvements. Results include increasing the number of CT and MRI scans performed per day by 2 and 1, respectively; creating a capacity for 10 additional chemotherapy and antibiotic infusion patients per day in the cancer center; reducing the waiting time of patients from 122 to 52 minutes at the urgent care clinic; standardizing surgical instrument use by the general surgery group, which resulted in processing more than 40,000 fewer instruments each month. These improvements achieved through applying lean concepts have resulted in Park Nicollet being recognized by the American Medical Group Association (AMGA) with its top‐rated Acclaim Award. 21 In addition, PNHS has been able to achieve a record 3.9% operating margin, which equates to a $7.5 million profit in 2004.

• In Pittsburgh, Pennsylvania, a group of hospitals participating in the Pittsburgh Regional Healthcare Initiative (PRHI) have implemented lean concepts to minimize the risk of developing central catheterrelated bloodstream infections. Several hospitals have been able to cut the incidence of central line infections by 50%‐90% through implementation of lean production methods. 12

• At Community Medical Center in Missoula, Montana, a series of pilot projects have been initiated to test lean methods. Some of the early results have demonstrated a reduction in turnaround time for pathology reports from the anatomical pathology lab from 5 to 2 days, a reduction in the number of steps and therefore the time from medication order to treatment initiation from 4 hours to 12 minutes, and a reduction in time for unit clerks to process new physician orders from an average of 43 minutes to 10 minutes during the hospital's busiest hours. 22

THE MICHIGAN LEAN EXPERIENCE

In the past year, the University of Michigan has begun to use lean production methods to improve the care of patients across various venues of hospitalization and flow toward discharge. Delays in placement of peripherally inserted central catheters (PICC) were associated with delays in appropriate and timely administration of intravenous medication, as well as in delays in discharges home or to extended‐care facilities (ECF) for continuation of medical care post‐hospitalization. Since the initiation of the lean PICC initiative, when adjusted for increased volume of demand, for 3 consecutive months 90%95% of the PICC lines have been inserted within 24 hours of request. This is a remarkable achievement, given that in the previous 12 months only 50%70% of PICC lines were placed within 24 hours of request. Even without adjusting for volume of demand, the lean PICC initiative has resulted in a 36% decrease in the average time to line placement and in a 50% decrease in the number of PICC referrals to interventional radiology (IR), thus decreasing the workload of a constrained resource.

As with all lean improvement projects, the entire value stream map was assessed in order to identify areas of intervention that would enhance the final product of the process for the patient (in this case, placing a PICC line as timely as possible). As we evaluated this value stream, one step in the process, occurring prior to placement of the line, appeared to be significantly wasteful: when the PICC nurse needed to search for data (such as locating a patient's chart for the order and reviewing labs and medication records) and to ensure that the patient was in his/her room and prepared for line placement. This step appeared to be inefficient use of the time of technically skilled individuals. The future state map of this process implemented the addition of an assisting individual who would ensure that these prework issues were prepared and completed in advance for the PICC nurses, thus making maximum use of the time of these skilled individuals in placing PICC lines, not in performing unskilled work. Another area where an intervention was believed beneficial was streamlining the process of chest x‐ray (CXR) ordering and reading in order to obtain PICC line confirmation. Performance of the previous process was not standardized and led to delays. The future state proposed a standard method of writing an order for a CXR, a standard method for getting that order to the radiology department, and a standard method for reading the films for dictation. This prevented the confusion and rework that had previously occurred. A last example of an intervention is that the PICC nurses began to internally defer to more experienced nurses if a less experienced nurse could not successfully place a PICC line in a patient. Previously, an unsuccessful attempt at bedside PICC placement warranted an IR referral, thus increasing the demand on an already constrained resource. This intervention by the PICC nurses drove down referrals to the IR suite by 50%. Although this may have led to a small increase in rework early in the process, it has led to a significant reduction in work downstream in the process. Thus, we believe that the overall work flow process has been served well by this intervention. As depicted in Figure 3, the lean process improvement method seeks to have continuous improvement, with the old future‐state map taking on the role of the new current‐state map. Since the initial development of these value stream maps, we have been working toward developing and implementing new areas of intervention, which will lead to new future‐state maps and to further improvement of this process, as the demand for PICC lines continues to rise.

Another critical segment in the care of hospitalized patients is the discharge process, including coordination of care to an outpatient or extended care facility (ECF) setting, which has several potential areas of disconnect that could result inpatients having untoward complications requiring rehospitalization, higher morbidity, or prolongation of suffering from their illnesses. The University of Michigan sees a tremendous opportunity to make a significant impact on patient care in this realm and has just initiated a lean project on the coordination of care. Team members on this project will be relevant process stakeholders, including those representing hospitalists, discharge planning, nursing, social work, a related home nursing company, a home infusion service, an ECF, ambulatory care, pharmacy, case management, nutrition, utilization review, patients and their families, and clinic physicians. The overall goal of the project is to optimize patient care from hospitalization to discharge and transfer of care to the outpatient setting.

CONCLUSION

The health care industry should learn about and consider adoption of lean techniques in order to improve its processes. More specifically, hospitals are prime locales for reaping the benefits of implementation of lean production, which can significantly affect how health care is delivered to patients. Toyota and other lean exemplars in the manufacturing industry have achieved a high level of success by utilizing the practice of lean. Early results from health care organizations suggest that utilizing lean production methods can lead to substantial improvements in the quality and efficiency of health care. To determine if the magnitude of success experienced by Toyota and other lean exemplars can also be achieved in the health care sector, it will be necessary to continuously test and evaluate the impact lean health care can have. In the hospital setting, where hospitalists are at the forefront of delivering care, it is incumbent on the hospitalist community to evaluate whether these techniques can make a difference in the quality, efficiency, and safety of the care provided to patients.

Lean thinking is still a novel idea to those in the health care sector, and as early adopters of this promising management model, we are very optimistic about the benefits of applying lean concepts in our hospital. Some of the first published reports and results presented on the benefits of lean in individual organizations are encouraging; however, as health care is a scientific community, we believe that future work should undergo rigorous evaluation on the benefits of lean and that such future works should be shared among the health care community through peer‐reviewed and published works.

Toyota is widely recognized as one of the most successful companies in the world. Its automobiles have consistently placed at or near the top of the quality and customer satisfaction rankings published by J.D. Power and Associates and Consumer Reports. Toyota constantly focuses on the safety and well‐being of its employees and the quality of its cars through its relentless dedication to continuous improvement in everything it does. Toyota has recently become the world's number two auto manufacturer, and the company's net profit margin was more than 8 times that of the industry average. 1

How has Toyota been able to achieve such remarkable results in product quality, market share, and profit margins? Jeffrey Liker, in his book The Toyota Way, described the world‐renowned Toyota production system as supported by 2 pillars: continuous improvement and respect for people. The end result is a learning organization that values employee contributions and continuously strives to produce products of higher quality at lower cost. 1, 2 Lean production is the generic term used to describe the principles and methods of the Toyota Production System. Lean production has been implemented to improve performance in a broad array of industries, from aerospace and aluminum refining to financial services and insurance. The philosophy of lean thinking, which is derived from the Toyota Production System, is rapidly gaining a following among health care leaders, with a number of hospitals and medical groups around the country adopting a version of lean production as their systematic approach to improving quality and efficiency. In the coming years, the application of lean principles and methods could have a transformational effect on how health care is delivered, with the potential for dramatic gains in quality, safety, efficiency, and appropriateness.

LEAN CONCEPTS

To understand how lean production can be applied to improve the delivery of health care, some of the fundamental concepts and practice of lean must first be explained. 3, 4 The first step in a lean improvement initiative is to understand value as defined by our customers. 5 In clinical care delivery, external customers include patients, families, payers, and regulators. Internal customers include physicians, nurses, clerks, and others involved in the care process. What customers value usually includes care that is of high quality, safe, efficient and appropriate. The second step typically is to go to the workplace and observe firsthand how the process now operates. 6 As the flow of the process from beginning to end is seen, the observer learns to see and to understand the multiple areas of delay, inefficiency, and waste that may exist. 7, 8 A representational flowchart called a current‐state value stream map (CS VSM) is created to make the work visible and to depict graphically all the individual steps necessary to complete the process from beginning to end. It is important that the CS VSM be a factual depiction of how an entire process flows created by those who actually work in that process. The CS VSM does not state any exceptions to or provide any explanations for why certain steps are taken. It does include key measures such as process time (the actual time it takes to complete a particular step of the process), lead time (the total time it takes to complete the entire process, including waiting time), and first‐time quality (the percentage of time in which that step of the process is completed without defect); (see Figure 1).

Figure 1

In the hands of an improvement team, the current state map becomes a powerful tool that allows participants to systematically recognize and categorize waste. The CS VSM also allows workers to visualize how much opportunity there is for improving the existing process. Working from the CS VSM, the team members can identify specific areas of waste, delay, causes of error, and inefficiency. The team then brainstorms ideas for improvement, proposing how steps of the process might be combined, eliminated, error‐proofed, or otherwise improved to transform waste into value from the customer's perspective. In the third step, the team seeks to achieve the flow state in which the steps of the process follow one another without stopping. All ideas are welcomed at this stage and are placed on the current state map itself or arrayed elsewhere for consideration. Using the ideas generated by the team, a new and better process is designed and depicted on a flow map called the future‐state value stream map (FS VSM). 9, 10 The FS VSM represents an improved and streamlined or ideal way in which the process could be accomplished, as best the team was able to envision at this point (see Figure 2). Ideally, the process described in the FS VSM also allows customers to pull value when they need goods or services provided by the organization, rather than having to do the usual requesting and waiting seen in health care and other service industries. Creating processes from which customers pull what they need is the fourth step in lean design.

Figure 2

Once a future state map is devised and approved, the critical work of rapid deployment of an implementation plan for reaching the future state begins. An implementation plan explicitly identifies who is responsible for what aspect of implementation. Usually a senior leader or leadership group sponsoring the project is responsible for encouraging team members to think beyond their historical (and often political) limits and to support the team in overcoming barriers outside its control. As the individuals return to work and attempt to implement the new solutions, however, they will likely encounter areas of resistance and ambiguity that require creative solutions. The implementation phase focuses on and encourages the individual worker to experiment and work toward a solution that can be broadly adopted and disseminated for use as a standardized solution by all workers facing similar situations. 11, 12 Through this experimental development and dissemination of solutions, the agreed‐to future‐state map is revised. In this way the old future‐state map plays the role of the new current‐state map. There is an ongoing, continuous loop between the current‐ and future‐state maps through implementation and testing to develop the ideal way in which the process should flow toward the final product or service (see Figure 3). The fifth step, pursuing perfection, requires this continuous loop of all workers improving everything they do, every day. The hardest of all the steps, pursuing perfection requires an organization to commit to process improvement and the elimination of defects and waste on a daily and permanent basis. 5

Figure 3

The steps outlined above provide only the basic foundations of lean concepts, of course. A detailed description of learning about and applying lean within one's own organization requires further study and help. Readers interested in learning more about value stream mapping are referred to a workbook by Mike Rother and John Shook called Learning to See, Value Stream Mapping to Create Value and Eliminate Muda. 4 Further, there are consultants with lean expertise who are available to help hospitals get started on the lean journey.

The management philosophy of lean production methods has ties to other operational and quality‐improvement models such as total quality management (TQM)/continuous quality improvement (CQI), developed by W. E. Deming, and Six Sigma, developed by Motorola and General Electric. Although there are several overlapping points of philosophy and techniques, a feature distinguishing lean from these other models is in its value stream approach to driving change and eliminating waste within the process of providing a product for the customer. Lean is unique in its focus on the specification of value from the customer's perspective and on the identification and categorization of waste and its transformation to value using specific tools. The lean approach encourages individuals within the organization (from top to bottom) to learn to see the flow of their product's process and thus to help to identify areas of waste, with the ultimate goal of creating a product with built‐in quality with the least amount of waste. Both Six Sigma and TQM/CQI focus on the delivery of a high‐quality product. Once a system has been studied and standardized, Six Sigma utilizes rigorous statistical and data measurements to drive quality improvements in product delivery. 13 Total quality management/continuous quality improvement engages the entire organization in delivering a high‐quality product from the customer's standpoint by getting everyone in the organization involved in the continuous improvement effort. 14 Lean thinking builds directly on the plandocheckact cycle of CQI but adds tools to identify and transform waste, supplies metrics of timing and resources, and, most important, focuses intently on the creation of value as defined by the customer. A more detailed comparison of lean philosophy with these other quality‐improvement management philosophies is beyond the scope of this article on the introduction of lean methods for hospitals.

DOES HEALTH CARE NEED LEAN THINKERS?

So, should health care try to emulate the successes of an automobile manufacturing company? Before answering this question, consider the following about the health care system as we know it and the significant challenges it faces.

• A key take‐home message of the Institute of Medicine's 1999 report, To Err Is Human: Building a Safer Health System, was that errors are caused by poorly designed systems. 15

• The Centers for Medicare and Medicaid Services (CMS) reported that the cost of health care is rising rapidly and that the rate of growth is not sustainable. 16

• Studies by the RAND Corporation demonstrated considerable variability in the practice of medicine, raising important questions about the appropriateness and necessity of some medical care and procedures. 17, 18

• In Crossing the Quality Chasm: a New Health System for the 21st Century, the Institute of Medicine concluded that today's health care system functions at far lower levels than it can and should and recommended 6 aims for improving the health care system: health care should be safe, effective, patient centered, timely, efficient, and equitable. 19 This report is rich in detail about what the ideal health care system should look like and how application of lean philosophy and tools can help hospitals and physicians achieve that vision.

Although these challenges reflect a view of the health care system at a very macro level, the mechanism and process of driving change will need to be initiated at the more local and organizational level. Lean production focuses on the goal of continuously transforming waste into value from the customer's perspective. It provides a rigorous and systematic approach to process improvement, error proofing, and waste reduction. Manufacturing companies such as Toyota and Alcoa and financial service organizations such as Vanguard have enjoyed tremendous success in implementing lean production, reporting gains in both quality and efficiency. 11 It is time for health care leaders and practitioners to evaluate how lean techniques can be adapted and applied to addressing the pressing challenges of safety, quality, efficiency, and appropriateness in order to improve system reliability and timeliness. As hospitalists, we are at the forefront of these challenges and therefore are in a prime position to lead the change in how health care delivery is improved continuously using a rigorous system such as the Toyota production system.

LEAN IN HEALTH CARE: EXAMPLES OF SOME EARLY RESULTS

Lean health care is still a very novel concept to most health care institutions; however, there have been some early adopters of lean health care, and some of their experiences are described here:

• At Virginia Mason Medical Center (VMMC) in Seattle, Washington, changes implemented using lean production methods have resulted in decreased incidence of ventilator‐associated pneumoniafrom 34 cases with 5 deaths in 2002 to 4 cases with 1 death in 2004. This led to a cost reduction of nearly a half‐million dollars. VMMC has also reported increased profit margins and improvement in space utilization at its cancer center, enabling 57% more patients to be seen in the same allotted space; and it is now taking measures to decrease the number of medication errors by standardizing and mistake‐proofing the process of ordering, delivering, and administering medications, all using lean techniques. 12, 20

• At Park Nicollet Health Services (PNHS) in Minneapolis, Minnesota, implementation of lean production has enabled improved patient access through flow improvements. Results include increasing the number of CT and MRI scans performed per day by 2 and 1, respectively; creating a capacity for 10 additional chemotherapy and antibiotic infusion patients per day in the cancer center; reducing the waiting time of patients from 122 to 52 minutes at the urgent care clinic; standardizing surgical instrument use by the general surgery group, which resulted in processing more than 40,000 fewer instruments each month. These improvements achieved through applying lean concepts have resulted in Park Nicollet being recognized by the American Medical Group Association (AMGA) with its top‐rated Acclaim Award. 21 In addition, PNHS has been able to achieve a record 3.9% operating margin, which equates to a $7.5 million profit in 2004.

• In Pittsburgh, Pennsylvania, a group of hospitals participating in the Pittsburgh Regional Healthcare Initiative (PRHI) have implemented lean concepts to minimize the risk of developing central catheterrelated bloodstream infections. Several hospitals have been able to cut the incidence of central line infections by 50%‐90% through implementation of lean production methods. 12

• At Community Medical Center in Missoula, Montana, a series of pilot projects have been initiated to test lean methods. Some of the early results have demonstrated a reduction in turnaround time for pathology reports from the anatomical pathology lab from 5 to 2 days, a reduction in the number of steps and therefore the time from medication order to treatment initiation from 4 hours to 12 minutes, and a reduction in time for unit clerks to process new physician orders from an average of 43 minutes to 10 minutes during the hospital's busiest hours. 22

THE MICHIGAN LEAN EXPERIENCE

In the past year, the University of Michigan has begun to use lean production methods to improve the care of patients across various venues of hospitalization and flow toward discharge. Delays in placement of peripherally inserted central catheters (PICC) were associated with delays in appropriate and timely administration of intravenous medication, as well as in delays in discharges home or to extended‐care facilities (ECF) for continuation of medical care post‐hospitalization. Since the initiation of the lean PICC initiative, when adjusted for increased volume of demand, for 3 consecutive months 90%95% of the PICC lines have been inserted within 24 hours of request. This is a remarkable achievement, given that in the previous 12 months only 50%70% of PICC lines were placed within 24 hours of request. Even without adjusting for volume of demand, the lean PICC initiative has resulted in a 36% decrease in the average time to line placement and in a 50% decrease in the number of PICC referrals to interventional radiology (IR), thus decreasing the workload of a constrained resource.

As with all lean improvement projects, the entire value stream map was assessed in order to identify areas of intervention that would enhance the final product of the process for the patient (in this case, placing a PICC line as timely as possible). As we evaluated this value stream, one step in the process, occurring prior to placement of the line, appeared to be significantly wasteful: when the PICC nurse needed to search for data (such as locating a patient's chart for the order and reviewing labs and medication records) and to ensure that the patient was in his/her room and prepared for line placement. This step appeared to be inefficient use of the time of technically skilled individuals. The future state map of this process implemented the addition of an assisting individual who would ensure that these prework issues were prepared and completed in advance for the PICC nurses, thus making maximum use of the time of these skilled individuals in placing PICC lines, not in performing unskilled work. Another area where an intervention was believed beneficial was streamlining the process of chest x‐ray (CXR) ordering and reading in order to obtain PICC line confirmation. Performance of the previous process was not standardized and led to delays. The future state proposed a standard method of writing an order for a CXR, a standard method for getting that order to the radiology department, and a standard method for reading the films for dictation. This prevented the confusion and rework that had previously occurred. A last example of an intervention is that the PICC nurses began to internally defer to more experienced nurses if a less experienced nurse could not successfully place a PICC line in a patient. Previously, an unsuccessful attempt at bedside PICC placement warranted an IR referral, thus increasing the demand on an already constrained resource. This intervention by the PICC nurses drove down referrals to the IR suite by 50%. Although this may have led to a small increase in rework early in the process, it has led to a significant reduction in work downstream in the process. Thus, we believe that the overall work flow process has been served well by this intervention. As depicted in Figure 3, the lean process improvement method seeks to have continuous improvement, with the old future‐state map taking on the role of the new current‐state map. Since the initial development of these value stream maps, we have been working toward developing and implementing new areas of intervention, which will lead to new future‐state maps and to further improvement of this process, as the demand for PICC lines continues to rise.

Another critical segment in the care of hospitalized patients is the discharge process, including coordination of care to an outpatient or extended care facility (ECF) setting, which has several potential areas of disconnect that could result inpatients having untoward complications requiring rehospitalization, higher morbidity, or prolongation of suffering from their illnesses. The University of Michigan sees a tremendous opportunity to make a significant impact on patient care in this realm and has just initiated a lean project on the coordination of care. Team members on this project will be relevant process stakeholders, including those representing hospitalists, discharge planning, nursing, social work, a related home nursing company, a home infusion service, an ECF, ambulatory care, pharmacy, case management, nutrition, utilization review, patients and their families, and clinic physicians. The overall goal of the project is to optimize patient care from hospitalization to discharge and transfer of care to the outpatient setting.

CONCLUSION

The health care industry should learn about and consider adoption of lean techniques in order to improve its processes. More specifically, hospitals are prime locales for reaping the benefits of implementation of lean production, which can significantly affect how health care is delivered to patients. Toyota and other lean exemplars in the manufacturing industry have achieved a high level of success by utilizing the practice of lean. Early results from health care organizations suggest that utilizing lean production methods can lead to substantial improvements in the quality and efficiency of health care. To determine if the magnitude of success experienced by Toyota and other lean exemplars can also be achieved in the health care sector, it will be necessary to continuously test and evaluate the impact lean health care can have. In the hospital setting, where hospitalists are at the forefront of delivering care, it is incumbent on the hospitalist community to evaluate whether these techniques can make a difference in the quality, efficiency, and safety of the care provided to patients.

Lean thinking is still a novel idea to those in the health care sector, and as early adopters of this promising management model, we are very optimistic about the benefits of applying lean concepts in our hospital. Some of the first published reports and results presented on the benefits of lean in individual organizations are encouraging; however, as health care is a scientific community, we believe that future work should undergo rigorous evaluation on the benefits of lean and that such future works should be shared among the health care community through peer‐reviewed and published works.