Expert Commentary

This interesting study from the Ohio Perinatal Quality Collaborative (OPQC) tackles an important topic—reducing the number of scheduled deliveries that take place before 39 weeks’ gestation. Infants born before 39 weeks have a higher rate of NICU admission in addition to measurable morbidity. The March of Dimes and other organizations have recognized the reduction of these births as an important goal, and ACOG recently refined and clarified the medical indications for them.¹

Details of the study

Twenty maternity hospitals in Ohio collected baseline data on the rate of scheduled delivery between 36.0 and 38.6 weeks’ gestation over a 60-day period. Members of the OPQC (which included perinatologists) then visited each participating hospital to introduce the project and explain how it would be sustained using techniques from the Institute for Healthcare Improvement Breakthrough Series.² Each hospital developed a three-person team—a physician, a nurse, and a data manager—to lead the effort. OPQC faculty developed a list of interventions, including the following broad techniques:

• promote optimal determination of gestational age using ultrasonography (US)
• use ACOG criteria for the indication and timing of scheduled births
• educate physicians, nurses, and pregnant women of the risks and benefits of birth between 36.0 and 38.6 weeks’ gestation
• improve communication between obstetricians and pediatricians through chart documentation of clear patient hand-offs and monthly reporting of statistics to physicians, nurses, and administrators
• include scheduled births in an overall “culture of safety,” with regular discussion of the matter at department and quality meetings.

Staff at participating hospitals had much latitude in selecting and refining interventions for their institution. After implementation of specific initiatives, each hospital tracked and reported relevant outcomes, especially the rate of scheduled deliveries between 36.0 and 38.6 weeks’ gestation. Each hospital received a monthly report, and staff members were encouraged to share the results internally.

Dramatic result was validated

The result of this initiative is impressive: The rate of inappropriate scheduled births was reduced from 25% to less than 5% over 14 months. This finding was validated in two ways:

• by checking Ohio birth certificate data
• by determining whether there was a concomitant rise in the rate of deliveries between 39 and 41 weeks’ gestation in the same hospitals.

Both validations confirmed that the OPQC interventions were significantly effective.

Project was not truly “statewide”

No study is perfect. A potential limitation of this study is the fact that it did not include all hospitals in the state of Ohio—only a sampling of larger hospitals. This makes it unclear whether the approach would be as effective in smaller institutions. Moreover, it is possible that physicians simply got better at documenting a reason for scheduled delivery as the study progressed—although I doubt that this reason alone would explain the striking findings.

This interesting study from the Ohio Perinatal Quality Collaborative (OPQC) tackles an important topic—reducing the number of scheduled deliveries that take place before 39 weeks’ gestation. Infants born before 39 weeks have a higher rate of NICU admission in addition to measurable morbidity. The March of Dimes and other organizations have recognized the reduction of these births as an important goal, and ACOG recently refined and clarified the medical indications for them.¹

Details of the study

Twenty maternity hospitals in Ohio collected baseline data on the rate of scheduled delivery between 36.0 and 38.6 weeks’ gestation over a 60-day period. Members of the OPQC (which included perinatologists) then visited each participating hospital to introduce the project and explain how it would be sustained using techniques from the Institute for Healthcare Improvement Breakthrough Series.² Each hospital developed a three-person team—a physician, a nurse, and a data manager—to lead the effort. OPQC faculty developed a list of interventions, including the following broad techniques:

• promote optimal determination of gestational age using ultrasonography (US)
• use ACOG criteria for the indication and timing of scheduled births
• educate physicians, nurses, and pregnant women of the risks and benefits of birth between 36.0 and 38.6 weeks’ gestation
• improve communication between obstetricians and pediatricians through chart documentation of clear patient hand-offs and monthly reporting of statistics to physicians, nurses, and administrators
• include scheduled births in an overall “culture of safety,” with regular discussion of the matter at department and quality meetings.

Staff at participating hospitals had much latitude in selecting and refining interventions for their institution. After implementation of specific initiatives, each hospital tracked and reported relevant outcomes, especially the rate of scheduled deliveries between 36.0 and 38.6 weeks’ gestation. Each hospital received a monthly report, and staff members were encouraged to share the results internally.

Dramatic result was validated

The result of this initiative is impressive: The rate of inappropriate scheduled births was reduced from 25% to less than 5% over 14 months. This finding was validated in two ways:

• by checking Ohio birth certificate data
• by determining whether there was a concomitant rise in the rate of deliveries between 39 and 41 weeks’ gestation in the same hospitals.

Both validations confirmed that the OPQC interventions were significantly effective.

Project was not truly “statewide”

No study is perfect. A potential limitation of this study is the fact that it did not include all hospitals in the state of Ohio—only a sampling of larger hospitals. This makes it unclear whether the approach would be as effective in smaller institutions. Moreover, it is possible that physicians simply got better at documenting a reason for scheduled delivery as the study progressed—although I doubt that this reason alone would explain the striking findings.

This interesting study from the Ohio Perinatal Quality Collaborative (OPQC) tackles an important topic—reducing the number of scheduled deliveries that take place before 39 weeks’ gestation. Infants born before 39 weeks have a higher rate of NICU admission in addition to measurable morbidity. The March of Dimes and other organizations have recognized the reduction of these births as an important goal, and ACOG recently refined and clarified the medical indications for them.¹

Details of the study

Twenty maternity hospitals in Ohio collected baseline data on the rate of scheduled delivery between 36.0 and 38.6 weeks’ gestation over a 60-day period. Members of the OPQC (which included perinatologists) then visited each participating hospital to introduce the project and explain how it would be sustained using techniques from the Institute for Healthcare Improvement Breakthrough Series.² Each hospital developed a three-person team—a physician, a nurse, and a data manager—to lead the effort. OPQC faculty developed a list of interventions, including the following broad techniques:

• promote optimal determination of gestational age using ultrasonography (US)
• use ACOG criteria for the indication and timing of scheduled births
• educate physicians, nurses, and pregnant women of the risks and benefits of birth between 36.0 and 38.6 weeks’ gestation
• improve communication between obstetricians and pediatricians through chart documentation of clear patient hand-offs and monthly reporting of statistics to physicians, nurses, and administrators
• include scheduled births in an overall “culture of safety,” with regular discussion of the matter at department and quality meetings.

Staff at participating hospitals had much latitude in selecting and refining interventions for their institution. After implementation of specific initiatives, each hospital tracked and reported relevant outcomes, especially the rate of scheduled deliveries between 36.0 and 38.6 weeks’ gestation. Each hospital received a monthly report, and staff members were encouraged to share the results internally.

Dramatic result was validated

The result of this initiative is impressive: The rate of inappropriate scheduled births was reduced from 25% to less than 5% over 14 months. This finding was validated in two ways:

• by checking Ohio birth certificate data
• by determining whether there was a concomitant rise in the rate of deliveries between 39 and 41 weeks’ gestation in the same hospitals.

Both validations confirmed that the OPQC interventions were significantly effective.

Project was not truly “statewide”

No study is perfect. A potential limitation of this study is the fact that it did not include all hospitals in the state of Ohio—only a sampling of larger hospitals. This makes it unclear whether the approach would be as effective in smaller institutions. Moreover, it is possible that physicians simply got better at documenting a reason for scheduled delivery as the study progressed—although I doubt that this reason alone would explain the striking findings.

CASE: Heavy bleeding and apprehension over hormones

Your patient, R.L., 34 years old, has been bothered by gradually worsening menorrhagia for 2 years when she schedules an appointment to discuss treatment options. She prefers to avoid surgery, if at all possible, and is reluctant to use hormonal medications because she’s concerned about side effects. Nonsteroidal anti-inflammatory drugs provide inadequate relief.

R.L. asks if there is any other agent that can reduce her menstrual blood loss.

What can you offer to her?

Heavy menstrual bleeding (menorrhagia) impairs quality of life and prompts many women such as R.L. to seek a gynecologic consultation—many of them hoping to avoid surgery.

In the United States, medical-management options for the treatment of heavy menstrual bleeding include nonsteroidal anti-inflammatory drugs (NSAIDs) and off-label use of oral contraceptives, injectable depot medroxyprogesterone acetate, and high-dose oral progestin (e.g., norethindrone acetate, 5-mg tablets). The levonorgestrel intrauterine system (LNG-IUS; Mirena) is also widely used to treat heavy menstrual bleeding in women and was approved for this indication in October 2009. FDA approval of the LNG-IUS for heavy menstrual bleeding was based on a pivotal randomized trial demonstrating high efficacy.^1,2

Very soon an additional option will be available—the antifibrinolytic agent tranexamic acid (to be sold by Ferring Pharmaceuticals under the brand name Lysteda). Although the LNG-IUS remains first-line treatment for heavy menstrual bleeding, tranexamic acid is appropriate for women who need to, or prefer to, avoid hormones.

How tranexamic acid works

Women who experience heavy menstrual bleeding have an elevated level of plasminogen activators in the endometrium, enzymes that facilitate the dissolution of clots. These enzymes can be inhibited by antifibrinolytic agents such as tranexamic acid, which competitively blocks the conversion of plasminogen to plasmin, thereby reducing fibrinolysis (FIGURE).³

FIGURE How tranexamic acid stanches heavy menstrual bleeding

Tranexamic acid (Lysteda) competitively blocks the conversion of plasminogen to plasmin, thereby reducing fibrinolysis.Clinical trials of tranexamic acid to treat menorrhagia were first performed in the 1960s. The agent proved to be superior to placebo, NSAIDs, and cyclic oral progestin.⁴ However, the LNG-IUS has been found to be more effective than tranexamic acid in reducing menstrual blood loss.⁵

In the United States, until very recently, tranexamic acid was FDA-indicated only as an injectable agent for use at the time of dental extraction in hemophilic patients. However, an oral formulation has been widely used abroad, including in Canada, to treat heavy menstrual bleeding. In Sweden, the agent is available over the counter.

In November 2009, the FDA approved 650-mg tablets of tranexamic acid as a treatment for heavy menstrual bleeding.⁶

Recommended dosage

In women who have normal renal function, the recommended dosage of tranexamic acid is 1,300 mg (two 650-mg tablets) three times daily (a total of 3,900 mg) for as long as 5 days during menstruation. In women who have impaired renal function, the dosage should be adjusted according to the package insert.

In randomized clinical trials, tranexamic acid did not produce more side effects than placebo did. However, because of concerns that tranexamic acid may increase risk of venous thromboembolism (VTE), the agent is contraindicated in women who have a history of VTE.

Although women who use hormonal contraception have not been included in clinical trials of tranexamic acid, the FDA recommends that they take tranexamic acid only if there is “a strong medical need and the benefit of treatment will outweigh the potential increased risk of a thrombotic event.”⁶

Tranexamic acid will likely be of particular utility for women who desire treatment for heavy menstrual bleeding but who should, or would prefer to, avoid hormonal agents. British guidelines recommend that, when medical management is appropriate in the treatment of heavy menstrual bleeding, the LNG-IUS be considered first-line treatment, with tranexamic acid representing a second-line strategy.⁷

CASE: Resolved

When she hears about tranexamic acid, R.L. is eager to try it. She is pleased that it does not contain hormones and needs to be taken only 5 days or fewer each month. You give her a prescription and schedule a follow-up visit 3 months later, at which time she reports being satisfied with the drug.

CASE: Heavy bleeding and apprehension over hormones

Your patient, R.L., 34 years old, has been bothered by gradually worsening menorrhagia for 2 years when she schedules an appointment to discuss treatment options. She prefers to avoid surgery, if at all possible, and is reluctant to use hormonal medications because she’s concerned about side effects. Nonsteroidal anti-inflammatory drugs provide inadequate relief.

R.L. asks if there is any other agent that can reduce her menstrual blood loss.

What can you offer to her?

Heavy menstrual bleeding (menorrhagia) impairs quality of life and prompts many women such as R.L. to seek a gynecologic consultation—many of them hoping to avoid surgery.

In the United States, medical-management options for the treatment of heavy menstrual bleeding include nonsteroidal anti-inflammatory drugs (NSAIDs) and off-label use of oral contraceptives, injectable depot medroxyprogesterone acetate, and high-dose oral progestin (e.g., norethindrone acetate, 5-mg tablets). The levonorgestrel intrauterine system (LNG-IUS; Mirena) is also widely used to treat heavy menstrual bleeding in women and was approved for this indication in October 2009. FDA approval of the LNG-IUS for heavy menstrual bleeding was based on a pivotal randomized trial demonstrating high efficacy.^1,2

Very soon an additional option will be available—the antifibrinolytic agent tranexamic acid (to be sold by Ferring Pharmaceuticals under the brand name Lysteda). Although the LNG-IUS remains first-line treatment for heavy menstrual bleeding, tranexamic acid is appropriate for women who need to, or prefer to, avoid hormones.

How tranexamic acid works

Women who experience heavy menstrual bleeding have an elevated level of plasminogen activators in the endometrium, enzymes that facilitate the dissolution of clots. These enzymes can be inhibited by antifibrinolytic agents such as tranexamic acid, which competitively blocks the conversion of plasminogen to plasmin, thereby reducing fibrinolysis (FIGURE).³

FIGURE How tranexamic acid stanches heavy menstrual bleeding

Tranexamic acid (Lysteda) competitively blocks the conversion of plasminogen to plasmin, thereby reducing fibrinolysis.Clinical trials of tranexamic acid to treat menorrhagia were first performed in the 1960s. The agent proved to be superior to placebo, NSAIDs, and cyclic oral progestin.⁴ However, the LNG-IUS has been found to be more effective than tranexamic acid in reducing menstrual blood loss.⁵

In the United States, until very recently, tranexamic acid was FDA-indicated only as an injectable agent for use at the time of dental extraction in hemophilic patients. However, an oral formulation has been widely used abroad, including in Canada, to treat heavy menstrual bleeding. In Sweden, the agent is available over the counter.

In November 2009, the FDA approved 650-mg tablets of tranexamic acid as a treatment for heavy menstrual bleeding.⁶

Recommended dosage

In women who have normal renal function, the recommended dosage of tranexamic acid is 1,300 mg (two 650-mg tablets) three times daily (a total of 3,900 mg) for as long as 5 days during menstruation. In women who have impaired renal function, the dosage should be adjusted according to the package insert.

In randomized clinical trials, tranexamic acid did not produce more side effects than placebo did. However, because of concerns that tranexamic acid may increase risk of venous thromboembolism (VTE), the agent is contraindicated in women who have a history of VTE.

Although women who use hormonal contraception have not been included in clinical trials of tranexamic acid, the FDA recommends that they take tranexamic acid only if there is “a strong medical need and the benefit of treatment will outweigh the potential increased risk of a thrombotic event.”⁶

Tranexamic acid will likely be of particular utility for women who desire treatment for heavy menstrual bleeding but who should, or would prefer to, avoid hormonal agents. British guidelines recommend that, when medical management is appropriate in the treatment of heavy menstrual bleeding, the LNG-IUS be considered first-line treatment, with tranexamic acid representing a second-line strategy.⁷

CASE: Resolved

When she hears about tranexamic acid, R.L. is eager to try it. She is pleased that it does not contain hormones and needs to be taken only 5 days or fewer each month. You give her a prescription and schedule a follow-up visit 3 months later, at which time she reports being satisfied with the drug.

CASE: Heavy bleeding and apprehension over hormones

Your patient, R.L., 34 years old, has been bothered by gradually worsening menorrhagia for 2 years when she schedules an appointment to discuss treatment options. She prefers to avoid surgery, if at all possible, and is reluctant to use hormonal medications because she’s concerned about side effects. Nonsteroidal anti-inflammatory drugs provide inadequate relief.

R.L. asks if there is any other agent that can reduce her menstrual blood loss.

What can you offer to her?

Heavy menstrual bleeding (menorrhagia) impairs quality of life and prompts many women such as R.L. to seek a gynecologic consultation—many of them hoping to avoid surgery.

In the United States, medical-management options for the treatment of heavy menstrual bleeding include nonsteroidal anti-inflammatory drugs (NSAIDs) and off-label use of oral contraceptives, injectable depot medroxyprogesterone acetate, and high-dose oral progestin (e.g., norethindrone acetate, 5-mg tablets). The levonorgestrel intrauterine system (LNG-IUS; Mirena) is also widely used to treat heavy menstrual bleeding in women and was approved for this indication in October 2009. FDA approval of the LNG-IUS for heavy menstrual bleeding was based on a pivotal randomized trial demonstrating high efficacy.^1,2

Very soon an additional option will be available—the antifibrinolytic agent tranexamic acid (to be sold by Ferring Pharmaceuticals under the brand name Lysteda). Although the LNG-IUS remains first-line treatment for heavy menstrual bleeding, tranexamic acid is appropriate for women who need to, or prefer to, avoid hormones.

How tranexamic acid works

Women who experience heavy menstrual bleeding have an elevated level of plasminogen activators in the endometrium, enzymes that facilitate the dissolution of clots. These enzymes can be inhibited by antifibrinolytic agents such as tranexamic acid, which competitively blocks the conversion of plasminogen to plasmin, thereby reducing fibrinolysis (FIGURE).³

FIGURE How tranexamic acid stanches heavy menstrual bleeding

Tranexamic acid (Lysteda) competitively blocks the conversion of plasminogen to plasmin, thereby reducing fibrinolysis.Clinical trials of tranexamic acid to treat menorrhagia were first performed in the 1960s. The agent proved to be superior to placebo, NSAIDs, and cyclic oral progestin.⁴ However, the LNG-IUS has been found to be more effective than tranexamic acid in reducing menstrual blood loss.⁵

In the United States, until very recently, tranexamic acid was FDA-indicated only as an injectable agent for use at the time of dental extraction in hemophilic patients. However, an oral formulation has been widely used abroad, including in Canada, to treat heavy menstrual bleeding. In Sweden, the agent is available over the counter.

In November 2009, the FDA approved 650-mg tablets of tranexamic acid as a treatment for heavy menstrual bleeding.⁶

Recommended dosage

In women who have normal renal function, the recommended dosage of tranexamic acid is 1,300 mg (two 650-mg tablets) three times daily (a total of 3,900 mg) for as long as 5 days during menstruation. In women who have impaired renal function, the dosage should be adjusted according to the package insert.

In randomized clinical trials, tranexamic acid did not produce more side effects than placebo did. However, because of concerns that tranexamic acid may increase risk of venous thromboembolism (VTE), the agent is contraindicated in women who have a history of VTE.

Although women who use hormonal contraception have not been included in clinical trials of tranexamic acid, the FDA recommends that they take tranexamic acid only if there is “a strong medical need and the benefit of treatment will outweigh the potential increased risk of a thrombotic event.”⁶

Tranexamic acid will likely be of particular utility for women who desire treatment for heavy menstrual bleeding but who should, or would prefer to, avoid hormonal agents. British guidelines recommend that, when medical management is appropriate in the treatment of heavy menstrual bleeding, the LNG-IUS be considered first-line treatment, with tranexamic acid representing a second-line strategy.⁷

CASE: Resolved

When she hears about tranexamic acid, R.L. is eager to try it. She is pleased that it does not contain hormones and needs to be taken only 5 days or fewer each month. You give her a prescription and schedule a follow-up visit 3 months later, at which time she reports being satisfied with the drug.

The dermatoses of pregnancy are a poorly understood group of conditions. Their only common feature is a tendency to appear during pregnancy.

Only three of these conditions are considered unique to pregnancy, however; the others are probably exacerbations of preexisting conditions triggered by pregnancy. There isn’t even complete agreement on what to call them. To make management even more complex, two patients—mother and fetus—need to be considered in decisions about care.

Who manages these patients is another matter. These conditions fall into overlapping areas of health care, where family physicians, obstetricians, and dermatologists all might have some share in responsibility for diagnosis and treatment. You need to be sufficiently familiar with these conditions so that you can differentiate those that can be treated symptomatically and those that require referral to a specialist. This review and the handy TABLE, will help you toward that end.

TABLE

Skin disorders of pregnancy: What you’ll see, how to treat

DERMATOSES UNIQUE TO PREGNANCY

1. Pemphigoid gestationis

Years ago, this disorder was referred to as herpes gestationis, because the lesions are herpetiform. Pemphigoid gestationis (PG) has an incidence of approximately 1 in 10,000 pregnancies.^1,2 Time of onset is usually about the 21st week of gestation, although, in about 20% of cases, the eruption appears immediately postpartum.³

Presentation. The disease usually begins with urticarial papules and plaques around the umbilicus and extremities. Bullous lesions tend to develop as the disease progresses, and are often not present on first presentation (FIGURE 1). Lesions of PG tend to spare the face, palms, and soles. Mucosal surfaces are involved in fewer than 20% of cases. In about 75% of cases, PG flares around the time of delivery, regressing spontaneously after the baby is born.⁴

FIGURE 1 Pemphigoid gestationis

As the disease progresses, bullous lesions tend to develop.

Pathophysiology. The pathophysiology of PG is nearly identical to that of bullous pemphigoid, a blistering skin disorder seen more often in elderly patients.⁵ Pemphigoid disorders are immune processes, involving an immunoglobulin G (IgG) immune response directed at a 180-kDa hemidesmosome transmembrane glycoprotein. This protein is the common target in several subepidermal blistering diseases.

Differential diagnosis. Disorders that may have some of the same features as PG include pruritic urticarial papules and plaques of pregnancy (PUPPP), erythema multiforme, intrahepatic cholestasis of pregnancy (ICP), contact dermatitis, and drug reactions.

Diagnosis. A biopsy is necessary for definitive diagnosis. Direct immunofluorescence (DIF) microscopy of a sample of perilesional skin can show tissue-bound immunoreactants. Linear deposition of the complement component protein C3 along the basement membrane zone is diagnostic for PG. IgG is also deposited about 40% of the time.³

Serum enzyme-linked immunosorbent assay (ELISA) studies are also helpful in diagnosis. They have excellent sensitivity and specificity, as well as the capacity to monitor levels of antibody, which correlate with the severity of disease.¹

Treatment. Oral corticosteroids are the first-line treatment for PG, typically 20 to 60 mg/d of prednisone. Oral corticosteroids are generally most effective at ameliorating symptoms. Prednisone at a dosage of 40 to 80 mg/d for a short time has not been associated with congenital abnormalities.⁶ PG patients can also be treated successfully with intravenous immunoglobulin (IVIG) and cyclosporine in refractory cases.⁷

Pruritus associated with this condition can interfere with day-to-day activities and with the patient’s ability to sleep. Patients may also complain that the rash is painful, particularly if bullae rupture, leading to superficial ulcerations. Fortunately, the patient’s quality of life can be dramatically improved with systemic corticosteroids—with no significant risk to the fetus.

Sequelae. PG uniformly resolves within a few weeks, but the mother’s autoantibodies can be passively transferred to the fetus, causing vesicles and bullae in the newborn.⁸ An increased incidence of small-for-gestational age (SGA) infants has also been noted in PG, although no lasting morbidity or mortality in the offspring has been noted.⁵ The disease tends to recur in future pregnancies.

2. Pruritic urticarial papules and plaques of pregnancy

This condition is known by many names besides its acronym PUPPP: polymorphic eruption of pregnancy, toxemic erythema of pregnancy, and late prurigo of pregnancy.¹ It is a pruritic, inflammatory skin disorder that has been variously estimated to occur in anywhere from 1 in 120 to 1 in 240 pregnancies.⁸ PUPPP is second only to eczema as the most common dermatosis of pregnancy.

Presentation. As the name suggests, the lesions of PUPPP are itchy, red papules that often coalesce into plaques (FIGURE 2). Lesions usually occur in primigravidas after the 34th week of gestation, although they may be seen at any time from the first trimester through the postpartum period.⁹

Lesions are classically found on the abdomen, sparing the umbilical area, and are found primarily in the striae. This distribution helps you to differentiate PUPPP from PG, in which lesions typically cluster around the umbilicus. Most PUPPP lesions (80% in one study) are dispersed on the abdomen, legs, arms, buttocks, chest, and back. Another 17% appear only on the abdomen and proximal thighs, and the remaining 3% on the limbs.¹⁰ Nearly 50% of the time, lesions also include discrete vesicles.¹¹ There are no reported cases of mucosal involvement.

Patients with this condition are often very uncomfortable. The associated pruritus is severe enough to interfere with sleep. Despite the itching, however, lesions are seldom excoriated.

FIGURE 2 Pruritic urticarial papules and plaques of pregnancy

The itchy, red papules of PUPP often coalesce into plaques.

Pathophysiology. The disorder has been strongly associated with maternal weight gain and multiple gestations. One working hypothesis is that rapid abdominal distention observed in the third trimester leads to damage of the connective tissue, which then releases antigenic molecules, causing an inflammatory reaction.¹² Another hypothesis is that increased levels of fetal DNA that have been detected in the skin of PUPPP patients may contribute to the pathology. One study detected male DNA in six of 10 PUPPP sufferers, but found none in any of 26 controls—pregnant women without PUPPP pathology.⁵ There is some evidence that patients with atopy may be predisposed to PUPPP, as well as patients who are hypertensive or obese.^10,13

Differential diagnosis. Initially, PUPPP lesions can be difficult to differentiate from urticarial PG lesions. The distribution of the lesions is the best clue: PG lesions cluster around the umbilicus, whereas PUPPP lesions uniformly spare the umbilical area. Additional disorders in the PUPPP differential are atopic dermatitis, superficial urticarial allergic eruption, viral exanthema, and contact or irritant dermatitis.

Diagnosis. PUPPP can be diagnosed only by clinical observation. None of the available laboratory tests—immunofluorescence, histology, serology—yield findings specific for PUPPP, although histology and immunofluorescence can readily differentiate between this condition and PG.

Treatment. Because the disease holds no real danger for mother or fetus, treatment can be aimed solely at symptomatic relief. Mild-to-potent topical corticosteroids (consider triamcinolone or fluocinonide) should relieve pruritus within 48 to 72 hours.⁸ Antihistamines and, occasionally, low-dose systemic corticosteroids may also be used. Consider hydroxyzine, although diphenhydramine has the more proven safety profile in pregnancy.

Nonpharmaceutical treatments such as oil baths and emollients should also be considered. If the condition appears classic for PUPPP, it can be managed symptomatically. If there is any question about the diagnosis, however, referral to a dermatologist is prudent.

Sequelae. No increase in maternal or fetal morbidity or mortality is associated with PUPPP. Recurrence is fairly uncommon, as the disease primarily affects women during their first pregnancy.

3. Intrahepatic cholestasis of pregnancy

This condition is also called recurrent or idiopathic jaundice of pregnancy, obstetric cholestasis, and pruritus gravidarum. Intrahepatic cholestasis of pregnancy (ICP) is caused by disruption of hepatic bile flow during pregnancy. It has been recorded at a rate of approximately 10 to 150 of every 10,000 pregnancies in Europe and 70 of every 10,000 in the United States.¹² In 80% of patients, time of onset is after the 30th week.¹⁴

Although this disorder is not primarily a dermatosis of pregnancy, it is a pruritic condition that often presents with excoriations in pregnant women and is associated with fetal morbidity and mortality. It’s important to be able to identify this disease early to minimize sequelae.

Presentation. There are no primary lesions with ICP. The primary presenting symptom is a generalized pruritus affecting the palms and soles, and sometimes extending to the legs and abdomen (FIGURE 3). This itching is often so severe that it leads to chronic insomnia. You may see secondary skin lesions, such as erythema and excoriations. Observable jaundice occurs in 10% to 20% of patients.³ These patients do not develop the encephalopathy that is associated with cholestasis in the nonpregnant state, however.¹⁴

FIGURE 3 Intrahepatic cholestasis of pregnancy

ICP lacks primary lesions. Shown here are the secondary erythema and excoriations that results from scratching the intense pruritis.

Pathophysiology. The genesis of this condition is thought to be a combination of genetic and environmental factors. A family history of the disorder is present in one half of cases; cases with a familial component tend to be more severe.¹⁵ ICP may be an exaggerated response to increased estrogen levels in pregnancy, but the mechanism of this response is unknown.¹⁶

Differential diagnosis. Other conditions that must be considered in making the diagnosis are viral hepatitis, gallbladder disease, PG, PUPPP, drug hepatotoxicity, primary biliary cirrhosis, and uremia.

Diagnosis. Laboratory values are the definitive diagnostic tool in this condition. Increased levels of serum bile acids are the single most sensitive test. Average levels of serum bile acids in pregnancy are 6.6 µmol/L, with an upper limit of 11 µmol/L. The average value in women who have ICP is 47 µmol/L.¹⁷

Although serum bile acids remain the gold standard, a recent study showed that elevated urine bile acids have 100% sensitivity and 83% specificity for ICP.¹⁸ In 55% to 60% of cases, the liver enzymes aspartate aminotransferase (AST) and alanine aminotransferase (ALT) are mildly increased. Steatorrhea is often noted by the patient, and is followed by vitamin K deficiency.¹⁷

Treatment. The current standard of care for ICP is treatment with ursodeoxycholic acid (UDCA). In four controlled trials, UDCA caused a sustained decrease in serum bile acids.^19-22 The dosage used in these trials ranged from 450 to 1,200 mg/d.

Before UDCA treatment was available, ICP was treated with cholestyramine, which could bring about a 70% rate of response. The drawback to cholestyramine treatment is that it precipitates vitamin K, which is already compromised by the disease process. Further, the onset of action of cholestyramine is slow.³

Elective delivery is indicated for ICP, particularly in patients who have a significant clinical presentation.¹² Delivery for ICP should be performed around weeks 37 to 38, as stillbirths tend to cluster around weeks 37 to 39.¹⁴ Given the significant fetal mortality associated with ICP (see the next paragraph), the condition should be managed by a clinician experienced with the disease—likely, a gastroenterologist.

Sequelae. The impact of this maternal disorder on the fetus can be disastrous: a 10% to 15% rate of perinatal death, and a 30% to 40% rate of premature labor.¹⁴ Fortunately, the rate of preterm labor correlates strongly with the level of bile acids, so that as bile acid levels are reduced with UDCA treatment, the rate of preterm labor also falls. Management of ICP has reduced the rate of perinatal death to 3.5%. No evidence of fetal growth retardation has been noted.¹⁴

DERMATOSES TRIGGERED BY PREGNANCY

4. Eczema of pregnancy/prurigo of pregnancy

Eczema of pregnancy/prurigo of pregnancy (EP/PP) may not actually be correlated with the pregnant state. Both conditions manifest as eczematous lesions in an atopic distribution. Although they have been described in the literature as separate entities, the lack of clinical distinction between them led Ambros-Rudolph and colleagues to combine them under the umbrella term, atopic eruption of pregnancy.²³

In some patients, at least, EP/PP may be preexisting conditions that are exacerbated by pregnancy. One study of 255 patients with the condition found that 20% had had the lesions before they became pregnant.²³ The tendency of the condition to be made markedly worse by pregnancy, however, leads us to include it here.

PP has an estimated incidence of 1 in 450 pregnancies.¹¹ Although many authorities consider EP to be the most common dermatosis of pregnancy, no clear estimation of its prevalence has been established.^23,24 Taken together, the two conditions have the highest prevalence of all pregnancy-induced dermatoses.

PP is also known as popular dermatitis of Spangler, Nurse’s early prurigo of pregnancy, and linear IgM disease of pregnancy.^3,4,23

Presentation. The typical presentation is grouped, crusted, erythematous papules, patches, and plaques—frequently with excoriations. Lesions typically present on the extensor surfaces of the arms and legs or on the abdomen (FIGURE 4).⁴ Recurrence in later pregnancies is common.

FIGURE 4 Eczema of pregnancy

EP/PP typically manifests as grouped, crusted, erythematous papules, patches, and plaques, often with excoriations.

Pathophysiology. The pathophysiology of EP/PP is not understood. Many patients who develop EP/PP have a history of atopy.¹⁰

Differential diagnosis. Conditions that need to be considered in making the diagnosis include Tinea infection, scabies, contact dermatitis, ICP, pruritic folliculitis of pregnancy (PFP), and PG.

Diagnosis. The history and the physical examination determine the diagnosis. Serology, histopathology, and immunofluorescence are nonspecific. Correlation of EP/PP with increased IgE is marginal, at best.^24,25

Treatment. These conditions are treated symptomatically with topical corticosteroids or systemic antihistamines.

Sequelae. No increase in maternal or fetal morbidity or mortality is associated with EP/PP.

5. Acute pustular psoriasis of pregnancy

Whether or not acute pustular psoriasis of pregnancy (APPP) is actually a pregnancy-induced dermatosis is subject to debate.

There is evidence that APPP is not unique to pregnancy but simply a manifestation of ordinary psoriasis. Clinically and histologically, APPP is indistinguishable from pustular psoriasis. Unlike most cases of acute psoriasis, however, APPP often appears in pregnancy without any personal or family history of psoriasis and usually ceases when the pregnancy ends. This fact, combined with reports of increased fetal and maternal morbidity and mortality associated with APPP, leads us to include it here.²⁶

Presentation. APPP is a rare condition that may have an onset at any point in pregnancy. Characteristic lesions begin as erythematous plaques with pustules on the inner thighs, flexural areas, and groin and spread to the trunk and extremities. As plaques enlarge, their center becomes eroded and crusted.

The nails may become onycholytic. Hands, feet, and face are usually spared. Oral and esophageal erosions can occur. Pruritus is typically mild, although the lesions are often painful. Flu-like symptoms are often present.²⁷

Pathophysiology. The pathophysiology of APPP is unknown.

Differential diagnosis. Conditions with similar presentations include an adverse drug reaction, pityriasis rosea, lichen simplex chronicus, eczema, lupus, and pityriasis rubra pilaris.

Diagnosis. The clinical history and an association with systemic illness are the basis for a diagnosis of APPP. Cultures of pustules are negative for any infective pathology, although, as the disease progresses, pustules may become superinfected. Laboratory testing may show an increased erythrocyte sedimentation rate, hypocalcemia, and a low level of vitamin D.

Treatment. Prednisone, 15 to 60 mg/d, is often sufficient to control the disease.²⁷ Cyclosporine, 100 mg twice daily, has also been shown to be useful.²⁸ Cyclosporine in pregnancy is a Category C drug. Data on fetal malformation associated with cyclosporine therapy are limited, but risk appears minimal.⁶

Maternal hypocalcemia should be monitored and treated appropriately. If disease progression is judged serious enough, early induction of labor is indicated, because delivery will almost always lead to swift resolution.

Sequelae. A number of case reports link APPP to serious sequelae, including fetal growth retardation, hypocalcemia, and stillbirth.^27,29,30 The condition is too rare, however, for good data on specific sequelae. Although APPP does give significant cause for concern, it appears that some of the traditional apprehension comes from older publications reporting a rate of maternal mortality of 70% to 90%.³¹ This statistic has not been borne out in practice. It does appear that the mother will frequently suffer systemic symptoms, including fever and malaise.

6. Pruritic folliculitis of pregnancy

Accounts of the prevalence of pruritic folliculitis of pregnancy (PFP) vary widely. Some sources report fewer than 30 cases in all of the literature; others indicate that the prevalence is equivalent to that of PG—one in every 10,000 pregnancies.^3,11 PFP most often presents in the third trimester. It often resolves before delivery, but uniformly clears within 2 weeks of delivery.

Presentation. PFP presents as papules and pustules concentrated around hair follicles (FIGURE 5). Often, lesions begin on the abdomen and spread to the extremities.^24,28 The condition is often, but not always, pruritic. Patients are more likely to be concerned about what the condition means for their health than distressed by the symptoms.

FIGURE 5 Pruritic folliculitis of pregnancy

The papules and pustules of PFP are concentrated around hair follicles.

Pathophysiology. Like many other dermatoses of pregnancy, the pathophysiology of PFP is unknown. There is little evidence that the condition is immunologically or hormonally mediated, and there is no evidence of an infectious component.^24,28

Differential diagnosis. PFP must be distinguished from infectious folliculitis, acneiform disorders, HIV-associated eosinophilic folliculitis, and a drug reaction.

Diagnosis. The clinical diagnosis is based on presenting symptoms and third-trimester onset. No specific laboratory or histologic analysis can be used to make a definitive diagnosis.

Treatment. As the condition is, by definition, a nonmicrobial folliculitis, the most effective therapy tends to be with a low- or midpotency topical corticosteroid, such as triamcinolone or desonide. A benzoyl peroxide wash can also be effective.

Sequelae. One study reports an increased incidence of low birth weight, but no associated morbidity or mortality has been reported in recent studies.²⁴

The dermatoses of pregnancy are a poorly understood group of conditions. Their only common feature is a tendency to appear during pregnancy.

Only three of these conditions are considered unique to pregnancy, however; the others are probably exacerbations of preexisting conditions triggered by pregnancy. There isn’t even complete agreement on what to call them. To make management even more complex, two patients—mother and fetus—need to be considered in decisions about care.

Who manages these patients is another matter. These conditions fall into overlapping areas of health care, where family physicians, obstetricians, and dermatologists all might have some share in responsibility for diagnosis and treatment. You need to be sufficiently familiar with these conditions so that you can differentiate those that can be treated symptomatically and those that require referral to a specialist. This review and the handy TABLE, will help you toward that end.

TABLE

Skin disorders of pregnancy: What you’ll see, how to treat

DERMATOSES UNIQUE TO PREGNANCY

1. Pemphigoid gestationis

Years ago, this disorder was referred to as herpes gestationis, because the lesions are herpetiform. Pemphigoid gestationis (PG) has an incidence of approximately 1 in 10,000 pregnancies.^1,2 Time of onset is usually about the 21st week of gestation, although, in about 20% of cases, the eruption appears immediately postpartum.³

Presentation. The disease usually begins with urticarial papules and plaques around the umbilicus and extremities. Bullous lesions tend to develop as the disease progresses, and are often not present on first presentation (FIGURE 1). Lesions of PG tend to spare the face, palms, and soles. Mucosal surfaces are involved in fewer than 20% of cases. In about 75% of cases, PG flares around the time of delivery, regressing spontaneously after the baby is born.⁴

FIGURE 1 Pemphigoid gestationis

As the disease progresses, bullous lesions tend to develop.

Pathophysiology. The pathophysiology of PG is nearly identical to that of bullous pemphigoid, a blistering skin disorder seen more often in elderly patients.⁵ Pemphigoid disorders are immune processes, involving an immunoglobulin G (IgG) immune response directed at a 180-kDa hemidesmosome transmembrane glycoprotein. This protein is the common target in several subepidermal blistering diseases.

Differential diagnosis. Disorders that may have some of the same features as PG include pruritic urticarial papules and plaques of pregnancy (PUPPP), erythema multiforme, intrahepatic cholestasis of pregnancy (ICP), contact dermatitis, and drug reactions.

Diagnosis. A biopsy is necessary for definitive diagnosis. Direct immunofluorescence (DIF) microscopy of a sample of perilesional skin can show tissue-bound immunoreactants. Linear deposition of the complement component protein C3 along the basement membrane zone is diagnostic for PG. IgG is also deposited about 40% of the time.³

Serum enzyme-linked immunosorbent assay (ELISA) studies are also helpful in diagnosis. They have excellent sensitivity and specificity, as well as the capacity to monitor levels of antibody, which correlate with the severity of disease.¹

Treatment. Oral corticosteroids are the first-line treatment for PG, typically 20 to 60 mg/d of prednisone. Oral corticosteroids are generally most effective at ameliorating symptoms. Prednisone at a dosage of 40 to 80 mg/d for a short time has not been associated with congenital abnormalities.⁶ PG patients can also be treated successfully with intravenous immunoglobulin (IVIG) and cyclosporine in refractory cases.⁷

Pruritus associated with this condition can interfere with day-to-day activities and with the patient’s ability to sleep. Patients may also complain that the rash is painful, particularly if bullae rupture, leading to superficial ulcerations. Fortunately, the patient’s quality of life can be dramatically improved with systemic corticosteroids—with no significant risk to the fetus.

Sequelae. PG uniformly resolves within a few weeks, but the mother’s autoantibodies can be passively transferred to the fetus, causing vesicles and bullae in the newborn.⁸ An increased incidence of small-for-gestational age (SGA) infants has also been noted in PG, although no lasting morbidity or mortality in the offspring has been noted.⁵ The disease tends to recur in future pregnancies.

2. Pruritic urticarial papules and plaques of pregnancy

This condition is known by many names besides its acronym PUPPP: polymorphic eruption of pregnancy, toxemic erythema of pregnancy, and late prurigo of pregnancy.¹ It is a pruritic, inflammatory skin disorder that has been variously estimated to occur in anywhere from 1 in 120 to 1 in 240 pregnancies.⁸ PUPPP is second only to eczema as the most common dermatosis of pregnancy.

Presentation. As the name suggests, the lesions of PUPPP are itchy, red papules that often coalesce into plaques (FIGURE 2). Lesions usually occur in primigravidas after the 34th week of gestation, although they may be seen at any time from the first trimester through the postpartum period.⁹

Lesions are classically found on the abdomen, sparing the umbilical area, and are found primarily in the striae. This distribution helps you to differentiate PUPPP from PG, in which lesions typically cluster around the umbilicus. Most PUPPP lesions (80% in one study) are dispersed on the abdomen, legs, arms, buttocks, chest, and back. Another 17% appear only on the abdomen and proximal thighs, and the remaining 3% on the limbs.¹⁰ Nearly 50% of the time, lesions also include discrete vesicles.¹¹ There are no reported cases of mucosal involvement.

Patients with this condition are often very uncomfortable. The associated pruritus is severe enough to interfere with sleep. Despite the itching, however, lesions are seldom excoriated.

FIGURE 2 Pruritic urticarial papules and plaques of pregnancy

The itchy, red papules of PUPP often coalesce into plaques.

Pathophysiology. The disorder has been strongly associated with maternal weight gain and multiple gestations. One working hypothesis is that rapid abdominal distention observed in the third trimester leads to damage of the connective tissue, which then releases antigenic molecules, causing an inflammatory reaction.¹² Another hypothesis is that increased levels of fetal DNA that have been detected in the skin of PUPPP patients may contribute to the pathology. One study detected male DNA in six of 10 PUPPP sufferers, but found none in any of 26 controls—pregnant women without PUPPP pathology.⁵ There is some evidence that patients with atopy may be predisposed to PUPPP, as well as patients who are hypertensive or obese.^10,13

Differential diagnosis. Initially, PUPPP lesions can be difficult to differentiate from urticarial PG lesions. The distribution of the lesions is the best clue: PG lesions cluster around the umbilicus, whereas PUPPP lesions uniformly spare the umbilical area. Additional disorders in the PUPPP differential are atopic dermatitis, superficial urticarial allergic eruption, viral exanthema, and contact or irritant dermatitis.

Diagnosis. PUPPP can be diagnosed only by clinical observation. None of the available laboratory tests—immunofluorescence, histology, serology—yield findings specific for PUPPP, although histology and immunofluorescence can readily differentiate between this condition and PG.

Treatment. Because the disease holds no real danger for mother or fetus, treatment can be aimed solely at symptomatic relief. Mild-to-potent topical corticosteroids (consider triamcinolone or fluocinonide) should relieve pruritus within 48 to 72 hours.⁸ Antihistamines and, occasionally, low-dose systemic corticosteroids may also be used. Consider hydroxyzine, although diphenhydramine has the more proven safety profile in pregnancy.

Nonpharmaceutical treatments such as oil baths and emollients should also be considered. If the condition appears classic for PUPPP, it can be managed symptomatically. If there is any question about the diagnosis, however, referral to a dermatologist is prudent.

Sequelae. No increase in maternal or fetal morbidity or mortality is associated with PUPPP. Recurrence is fairly uncommon, as the disease primarily affects women during their first pregnancy.

3. Intrahepatic cholestasis of pregnancy

This condition is also called recurrent or idiopathic jaundice of pregnancy, obstetric cholestasis, and pruritus gravidarum. Intrahepatic cholestasis of pregnancy (ICP) is caused by disruption of hepatic bile flow during pregnancy. It has been recorded at a rate of approximately 10 to 150 of every 10,000 pregnancies in Europe and 70 of every 10,000 in the United States.¹² In 80% of patients, time of onset is after the 30th week.¹⁴

Although this disorder is not primarily a dermatosis of pregnancy, it is a pruritic condition that often presents with excoriations in pregnant women and is associated with fetal morbidity and mortality. It’s important to be able to identify this disease early to minimize sequelae.

Presentation. There are no primary lesions with ICP. The primary presenting symptom is a generalized pruritus affecting the palms and soles, and sometimes extending to the legs and abdomen (FIGURE 3). This itching is often so severe that it leads to chronic insomnia. You may see secondary skin lesions, such as erythema and excoriations. Observable jaundice occurs in 10% to 20% of patients.³ These patients do not develop the encephalopathy that is associated with cholestasis in the nonpregnant state, however.¹⁴

FIGURE 3 Intrahepatic cholestasis of pregnancy

ICP lacks primary lesions. Shown here are the secondary erythema and excoriations that results from scratching the intense pruritis.

Pathophysiology. The genesis of this condition is thought to be a combination of genetic and environmental factors. A family history of the disorder is present in one half of cases; cases with a familial component tend to be more severe.¹⁵ ICP may be an exaggerated response to increased estrogen levels in pregnancy, but the mechanism of this response is unknown.¹⁶

Differential diagnosis. Other conditions that must be considered in making the diagnosis are viral hepatitis, gallbladder disease, PG, PUPPP, drug hepatotoxicity, primary biliary cirrhosis, and uremia.

Diagnosis. Laboratory values are the definitive diagnostic tool in this condition. Increased levels of serum bile acids are the single most sensitive test. Average levels of serum bile acids in pregnancy are 6.6 µmol/L, with an upper limit of 11 µmol/L. The average value in women who have ICP is 47 µmol/L.¹⁷

Although serum bile acids remain the gold standard, a recent study showed that elevated urine bile acids have 100% sensitivity and 83% specificity for ICP.¹⁸ In 55% to 60% of cases, the liver enzymes aspartate aminotransferase (AST) and alanine aminotransferase (ALT) are mildly increased. Steatorrhea is often noted by the patient, and is followed by vitamin K deficiency.¹⁷

Treatment. The current standard of care for ICP is treatment with ursodeoxycholic acid (UDCA). In four controlled trials, UDCA caused a sustained decrease in serum bile acids.^19-22 The dosage used in these trials ranged from 450 to 1,200 mg/d.

Before UDCA treatment was available, ICP was treated with cholestyramine, which could bring about a 70% rate of response. The drawback to cholestyramine treatment is that it precipitates vitamin K, which is already compromised by the disease process. Further, the onset of action of cholestyramine is slow.³

Elective delivery is indicated for ICP, particularly in patients who have a significant clinical presentation.¹² Delivery for ICP should be performed around weeks 37 to 38, as stillbirths tend to cluster around weeks 37 to 39.¹⁴ Given the significant fetal mortality associated with ICP (see the next paragraph), the condition should be managed by a clinician experienced with the disease—likely, a gastroenterologist.

Sequelae. The impact of this maternal disorder on the fetus can be disastrous: a 10% to 15% rate of perinatal death, and a 30% to 40% rate of premature labor.¹⁴ Fortunately, the rate of preterm labor correlates strongly with the level of bile acids, so that as bile acid levels are reduced with UDCA treatment, the rate of preterm labor also falls. Management of ICP has reduced the rate of perinatal death to 3.5%. No evidence of fetal growth retardation has been noted.¹⁴

DERMATOSES TRIGGERED BY PREGNANCY

4. Eczema of pregnancy/prurigo of pregnancy

Eczema of pregnancy/prurigo of pregnancy (EP/PP) may not actually be correlated with the pregnant state. Both conditions manifest as eczematous lesions in an atopic distribution. Although they have been described in the literature as separate entities, the lack of clinical distinction between them led Ambros-Rudolph and colleagues to combine them under the umbrella term, atopic eruption of pregnancy.²³

In some patients, at least, EP/PP may be preexisting conditions that are exacerbated by pregnancy. One study of 255 patients with the condition found that 20% had had the lesions before they became pregnant.²³ The tendency of the condition to be made markedly worse by pregnancy, however, leads us to include it here.

PP has an estimated incidence of 1 in 450 pregnancies.¹¹ Although many authorities consider EP to be the most common dermatosis of pregnancy, no clear estimation of its prevalence has been established.^23,24 Taken together, the two conditions have the highest prevalence of all pregnancy-induced dermatoses.

PP is also known as popular dermatitis of Spangler, Nurse’s early prurigo of pregnancy, and linear IgM disease of pregnancy.^3,4,23

Presentation. The typical presentation is grouped, crusted, erythematous papules, patches, and plaques—frequently with excoriations. Lesions typically present on the extensor surfaces of the arms and legs or on the abdomen (FIGURE 4).⁴ Recurrence in later pregnancies is common.

FIGURE 4 Eczema of pregnancy

EP/PP typically manifests as grouped, crusted, erythematous papules, patches, and plaques, often with excoriations.

Pathophysiology. The pathophysiology of EP/PP is not understood. Many patients who develop EP/PP have a history of atopy.¹⁰

Differential diagnosis. Conditions that need to be considered in making the diagnosis include Tinea infection, scabies, contact dermatitis, ICP, pruritic folliculitis of pregnancy (PFP), and PG.

Diagnosis. The history and the physical examination determine the diagnosis. Serology, histopathology, and immunofluorescence are nonspecific. Correlation of EP/PP with increased IgE is marginal, at best.^24,25

Treatment. These conditions are treated symptomatically with topical corticosteroids or systemic antihistamines.

Sequelae. No increase in maternal or fetal morbidity or mortality is associated with EP/PP.

5. Acute pustular psoriasis of pregnancy

Whether or not acute pustular psoriasis of pregnancy (APPP) is actually a pregnancy-induced dermatosis is subject to debate.

There is evidence that APPP is not unique to pregnancy but simply a manifestation of ordinary psoriasis. Clinically and histologically, APPP is indistinguishable from pustular psoriasis. Unlike most cases of acute psoriasis, however, APPP often appears in pregnancy without any personal or family history of psoriasis and usually ceases when the pregnancy ends. This fact, combined with reports of increased fetal and maternal morbidity and mortality associated with APPP, leads us to include it here.²⁶

Presentation. APPP is a rare condition that may have an onset at any point in pregnancy. Characteristic lesions begin as erythematous plaques with pustules on the inner thighs, flexural areas, and groin and spread to the trunk and extremities. As plaques enlarge, their center becomes eroded and crusted.

The nails may become onycholytic. Hands, feet, and face are usually spared. Oral and esophageal erosions can occur. Pruritus is typically mild, although the lesions are often painful. Flu-like symptoms are often present.²⁷

Pathophysiology. The pathophysiology of APPP is unknown.

Differential diagnosis. Conditions with similar presentations include an adverse drug reaction, pityriasis rosea, lichen simplex chronicus, eczema, lupus, and pityriasis rubra pilaris.

Diagnosis. The clinical history and an association with systemic illness are the basis for a diagnosis of APPP. Cultures of pustules are negative for any infective pathology, although, as the disease progresses, pustules may become superinfected. Laboratory testing may show an increased erythrocyte sedimentation rate, hypocalcemia, and a low level of vitamin D.

Treatment. Prednisone, 15 to 60 mg/d, is often sufficient to control the disease.²⁷ Cyclosporine, 100 mg twice daily, has also been shown to be useful.²⁸ Cyclosporine in pregnancy is a Category C drug. Data on fetal malformation associated with cyclosporine therapy are limited, but risk appears minimal.⁶

Maternal hypocalcemia should be monitored and treated appropriately. If disease progression is judged serious enough, early induction of labor is indicated, because delivery will almost always lead to swift resolution.

Sequelae. A number of case reports link APPP to serious sequelae, including fetal growth retardation, hypocalcemia, and stillbirth.^27,29,30 The condition is too rare, however, for good data on specific sequelae. Although APPP does give significant cause for concern, it appears that some of the traditional apprehension comes from older publications reporting a rate of maternal mortality of 70% to 90%.³¹ This statistic has not been borne out in practice. It does appear that the mother will frequently suffer systemic symptoms, including fever and malaise.

6. Pruritic folliculitis of pregnancy

Accounts of the prevalence of pruritic folliculitis of pregnancy (PFP) vary widely. Some sources report fewer than 30 cases in all of the literature; others indicate that the prevalence is equivalent to that of PG—one in every 10,000 pregnancies.^3,11 PFP most often presents in the third trimester. It often resolves before delivery, but uniformly clears within 2 weeks of delivery.

Presentation. PFP presents as papules and pustules concentrated around hair follicles (FIGURE 5). Often, lesions begin on the abdomen and spread to the extremities.^24,28 The condition is often, but not always, pruritic. Patients are more likely to be concerned about what the condition means for their health than distressed by the symptoms.

FIGURE 5 Pruritic folliculitis of pregnancy

The papules and pustules of PFP are concentrated around hair follicles.

Pathophysiology. Like many other dermatoses of pregnancy, the pathophysiology of PFP is unknown. There is little evidence that the condition is immunologically or hormonally mediated, and there is no evidence of an infectious component.^24,28

Differential diagnosis. PFP must be distinguished from infectious folliculitis, acneiform disorders, HIV-associated eosinophilic folliculitis, and a drug reaction.

Diagnosis. The clinical diagnosis is based on presenting symptoms and third-trimester onset. No specific laboratory or histologic analysis can be used to make a definitive diagnosis.

Treatment. As the condition is, by definition, a nonmicrobial folliculitis, the most effective therapy tends to be with a low- or midpotency topical corticosteroid, such as triamcinolone or desonide. A benzoyl peroxide wash can also be effective.

Sequelae. One study reports an increased incidence of low birth weight, but no associated morbidity or mortality has been reported in recent studies.²⁴

The dermatoses of pregnancy are a poorly understood group of conditions. Their only common feature is a tendency to appear during pregnancy.

Only three of these conditions are considered unique to pregnancy, however; the others are probably exacerbations of preexisting conditions triggered by pregnancy. There isn’t even complete agreement on what to call them. To make management even more complex, two patients—mother and fetus—need to be considered in decisions about care.

Who manages these patients is another matter. These conditions fall into overlapping areas of health care, where family physicians, obstetricians, and dermatologists all might have some share in responsibility for diagnosis and treatment. You need to be sufficiently familiar with these conditions so that you can differentiate those that can be treated symptomatically and those that require referral to a specialist. This review and the handy TABLE, will help you toward that end.

TABLE

Skin disorders of pregnancy: What you’ll see, how to treat

DERMATOSES UNIQUE TO PREGNANCY

1. Pemphigoid gestationis

Years ago, this disorder was referred to as herpes gestationis, because the lesions are herpetiform. Pemphigoid gestationis (PG) has an incidence of approximately 1 in 10,000 pregnancies.^1,2 Time of onset is usually about the 21st week of gestation, although, in about 20% of cases, the eruption appears immediately postpartum.³

Presentation. The disease usually begins with urticarial papules and plaques around the umbilicus and extremities. Bullous lesions tend to develop as the disease progresses, and are often not present on first presentation (FIGURE 1). Lesions of PG tend to spare the face, palms, and soles. Mucosal surfaces are involved in fewer than 20% of cases. In about 75% of cases, PG flares around the time of delivery, regressing spontaneously after the baby is born.⁴

FIGURE 1 Pemphigoid gestationis

As the disease progresses, bullous lesions tend to develop.

Pathophysiology. The pathophysiology of PG is nearly identical to that of bullous pemphigoid, a blistering skin disorder seen more often in elderly patients.⁵ Pemphigoid disorders are immune processes, involving an immunoglobulin G (IgG) immune response directed at a 180-kDa hemidesmosome transmembrane glycoprotein. This protein is the common target in several subepidermal blistering diseases.

Differential diagnosis. Disorders that may have some of the same features as PG include pruritic urticarial papules and plaques of pregnancy (PUPPP), erythema multiforme, intrahepatic cholestasis of pregnancy (ICP), contact dermatitis, and drug reactions.

Diagnosis. A biopsy is necessary for definitive diagnosis. Direct immunofluorescence (DIF) microscopy of a sample of perilesional skin can show tissue-bound immunoreactants. Linear deposition of the complement component protein C3 along the basement membrane zone is diagnostic for PG. IgG is also deposited about 40% of the time.³

Serum enzyme-linked immunosorbent assay (ELISA) studies are also helpful in diagnosis. They have excellent sensitivity and specificity, as well as the capacity to monitor levels of antibody, which correlate with the severity of disease.¹

Treatment. Oral corticosteroids are the first-line treatment for PG, typically 20 to 60 mg/d of prednisone. Oral corticosteroids are generally most effective at ameliorating symptoms. Prednisone at a dosage of 40 to 80 mg/d for a short time has not been associated with congenital abnormalities.⁶ PG patients can also be treated successfully with intravenous immunoglobulin (IVIG) and cyclosporine in refractory cases.⁷

Pruritus associated with this condition can interfere with day-to-day activities and with the patient’s ability to sleep. Patients may also complain that the rash is painful, particularly if bullae rupture, leading to superficial ulcerations. Fortunately, the patient’s quality of life can be dramatically improved with systemic corticosteroids—with no significant risk to the fetus.

Sequelae. PG uniformly resolves within a few weeks, but the mother’s autoantibodies can be passively transferred to the fetus, causing vesicles and bullae in the newborn.⁸ An increased incidence of small-for-gestational age (SGA) infants has also been noted in PG, although no lasting morbidity or mortality in the offspring has been noted.⁵ The disease tends to recur in future pregnancies.

2. Pruritic urticarial papules and plaques of pregnancy

This condition is known by many names besides its acronym PUPPP: polymorphic eruption of pregnancy, toxemic erythema of pregnancy, and late prurigo of pregnancy.¹ It is a pruritic, inflammatory skin disorder that has been variously estimated to occur in anywhere from 1 in 120 to 1 in 240 pregnancies.⁸ PUPPP is second only to eczema as the most common dermatosis of pregnancy.

Presentation. As the name suggests, the lesions of PUPPP are itchy, red papules that often coalesce into plaques (FIGURE 2). Lesions usually occur in primigravidas after the 34th week of gestation, although they may be seen at any time from the first trimester through the postpartum period.⁹

Lesions are classically found on the abdomen, sparing the umbilical area, and are found primarily in the striae. This distribution helps you to differentiate PUPPP from PG, in which lesions typically cluster around the umbilicus. Most PUPPP lesions (80% in one study) are dispersed on the abdomen, legs, arms, buttocks, chest, and back. Another 17% appear only on the abdomen and proximal thighs, and the remaining 3% on the limbs.¹⁰ Nearly 50% of the time, lesions also include discrete vesicles.¹¹ There are no reported cases of mucosal involvement.

Patients with this condition are often very uncomfortable. The associated pruritus is severe enough to interfere with sleep. Despite the itching, however, lesions are seldom excoriated.

FIGURE 2 Pruritic urticarial papules and plaques of pregnancy

The itchy, red papules of PUPP often coalesce into plaques.

Pathophysiology. The disorder has been strongly associated with maternal weight gain and multiple gestations. One working hypothesis is that rapid abdominal distention observed in the third trimester leads to damage of the connective tissue, which then releases antigenic molecules, causing an inflammatory reaction.¹² Another hypothesis is that increased levels of fetal DNA that have been detected in the skin of PUPPP patients may contribute to the pathology. One study detected male DNA in six of 10 PUPPP sufferers, but found none in any of 26 controls—pregnant women without PUPPP pathology.⁵ There is some evidence that patients with atopy may be predisposed to PUPPP, as well as patients who are hypertensive or obese.^10,13

Differential diagnosis. Initially, PUPPP lesions can be difficult to differentiate from urticarial PG lesions. The distribution of the lesions is the best clue: PG lesions cluster around the umbilicus, whereas PUPPP lesions uniformly spare the umbilical area. Additional disorders in the PUPPP differential are atopic dermatitis, superficial urticarial allergic eruption, viral exanthema, and contact or irritant dermatitis.

Diagnosis. PUPPP can be diagnosed only by clinical observation. None of the available laboratory tests—immunofluorescence, histology, serology—yield findings specific for PUPPP, although histology and immunofluorescence can readily differentiate between this condition and PG.

Treatment. Because the disease holds no real danger for mother or fetus, treatment can be aimed solely at symptomatic relief. Mild-to-potent topical corticosteroids (consider triamcinolone or fluocinonide) should relieve pruritus within 48 to 72 hours.⁸ Antihistamines and, occasionally, low-dose systemic corticosteroids may also be used. Consider hydroxyzine, although diphenhydramine has the more proven safety profile in pregnancy.

Nonpharmaceutical treatments such as oil baths and emollients should also be considered. If the condition appears classic for PUPPP, it can be managed symptomatically. If there is any question about the diagnosis, however, referral to a dermatologist is prudent.

Sequelae. No increase in maternal or fetal morbidity or mortality is associated with PUPPP. Recurrence is fairly uncommon, as the disease primarily affects women during their first pregnancy.

3. Intrahepatic cholestasis of pregnancy

This condition is also called recurrent or idiopathic jaundice of pregnancy, obstetric cholestasis, and pruritus gravidarum. Intrahepatic cholestasis of pregnancy (ICP) is caused by disruption of hepatic bile flow during pregnancy. It has been recorded at a rate of approximately 10 to 150 of every 10,000 pregnancies in Europe and 70 of every 10,000 in the United States.¹² In 80% of patients, time of onset is after the 30th week.¹⁴

Although this disorder is not primarily a dermatosis of pregnancy, it is a pruritic condition that often presents with excoriations in pregnant women and is associated with fetal morbidity and mortality. It’s important to be able to identify this disease early to minimize sequelae.

Presentation. There are no primary lesions with ICP. The primary presenting symptom is a generalized pruritus affecting the palms and soles, and sometimes extending to the legs and abdomen (FIGURE 3). This itching is often so severe that it leads to chronic insomnia. You may see secondary skin lesions, such as erythema and excoriations. Observable jaundice occurs in 10% to 20% of patients.³ These patients do not develop the encephalopathy that is associated with cholestasis in the nonpregnant state, however.¹⁴

FIGURE 3 Intrahepatic cholestasis of pregnancy

ICP lacks primary lesions. Shown here are the secondary erythema and excoriations that results from scratching the intense pruritis.

Pathophysiology. The genesis of this condition is thought to be a combination of genetic and environmental factors. A family history of the disorder is present in one half of cases; cases with a familial component tend to be more severe.¹⁵ ICP may be an exaggerated response to increased estrogen levels in pregnancy, but the mechanism of this response is unknown.¹⁶

Differential diagnosis. Other conditions that must be considered in making the diagnosis are viral hepatitis, gallbladder disease, PG, PUPPP, drug hepatotoxicity, primary biliary cirrhosis, and uremia.

Diagnosis. Laboratory values are the definitive diagnostic tool in this condition. Increased levels of serum bile acids are the single most sensitive test. Average levels of serum bile acids in pregnancy are 6.6 µmol/L, with an upper limit of 11 µmol/L. The average value in women who have ICP is 47 µmol/L.¹⁷

Although serum bile acids remain the gold standard, a recent study showed that elevated urine bile acids have 100% sensitivity and 83% specificity for ICP.¹⁸ In 55% to 60% of cases, the liver enzymes aspartate aminotransferase (AST) and alanine aminotransferase (ALT) are mildly increased. Steatorrhea is often noted by the patient, and is followed by vitamin K deficiency.¹⁷

Treatment. The current standard of care for ICP is treatment with ursodeoxycholic acid (UDCA). In four controlled trials, UDCA caused a sustained decrease in serum bile acids.^19-22 The dosage used in these trials ranged from 450 to 1,200 mg/d.

Before UDCA treatment was available, ICP was treated with cholestyramine, which could bring about a 70% rate of response. The drawback to cholestyramine treatment is that it precipitates vitamin K, which is already compromised by the disease process. Further, the onset of action of cholestyramine is slow.³

Elective delivery is indicated for ICP, particularly in patients who have a significant clinical presentation.¹² Delivery for ICP should be performed around weeks 37 to 38, as stillbirths tend to cluster around weeks 37 to 39.¹⁴ Given the significant fetal mortality associated with ICP (see the next paragraph), the condition should be managed by a clinician experienced with the disease—likely, a gastroenterologist.

Sequelae. The impact of this maternal disorder on the fetus can be disastrous: a 10% to 15% rate of perinatal death, and a 30% to 40% rate of premature labor.¹⁴ Fortunately, the rate of preterm labor correlates strongly with the level of bile acids, so that as bile acid levels are reduced with UDCA treatment, the rate of preterm labor also falls. Management of ICP has reduced the rate of perinatal death to 3.5%. No evidence of fetal growth retardation has been noted.¹⁴

DERMATOSES TRIGGERED BY PREGNANCY

4. Eczema of pregnancy/prurigo of pregnancy

Eczema of pregnancy/prurigo of pregnancy (EP/PP) may not actually be correlated with the pregnant state. Both conditions manifest as eczematous lesions in an atopic distribution. Although they have been described in the literature as separate entities, the lack of clinical distinction between them led Ambros-Rudolph and colleagues to combine them under the umbrella term, atopic eruption of pregnancy.²³

In some patients, at least, EP/PP may be preexisting conditions that are exacerbated by pregnancy. One study of 255 patients with the condition found that 20% had had the lesions before they became pregnant.²³ The tendency of the condition to be made markedly worse by pregnancy, however, leads us to include it here.

PP has an estimated incidence of 1 in 450 pregnancies.¹¹ Although many authorities consider EP to be the most common dermatosis of pregnancy, no clear estimation of its prevalence has been established.^23,24 Taken together, the two conditions have the highest prevalence of all pregnancy-induced dermatoses.

PP is also known as popular dermatitis of Spangler, Nurse’s early prurigo of pregnancy, and linear IgM disease of pregnancy.^3,4,23

Presentation. The typical presentation is grouped, crusted, erythematous papules, patches, and plaques—frequently with excoriations. Lesions typically present on the extensor surfaces of the arms and legs or on the abdomen (FIGURE 4).⁴ Recurrence in later pregnancies is common.

FIGURE 4 Eczema of pregnancy

EP/PP typically manifests as grouped, crusted, erythematous papules, patches, and plaques, often with excoriations.

Pathophysiology. The pathophysiology of EP/PP is not understood. Many patients who develop EP/PP have a history of atopy.¹⁰

Differential diagnosis. Conditions that need to be considered in making the diagnosis include Tinea infection, scabies, contact dermatitis, ICP, pruritic folliculitis of pregnancy (PFP), and PG.

Diagnosis. The history and the physical examination determine the diagnosis. Serology, histopathology, and immunofluorescence are nonspecific. Correlation of EP/PP with increased IgE is marginal, at best.^24,25

Treatment. These conditions are treated symptomatically with topical corticosteroids or systemic antihistamines.

Sequelae. No increase in maternal or fetal morbidity or mortality is associated with EP/PP.

5. Acute pustular psoriasis of pregnancy

Whether or not acute pustular psoriasis of pregnancy (APPP) is actually a pregnancy-induced dermatosis is subject to debate.

There is evidence that APPP is not unique to pregnancy but simply a manifestation of ordinary psoriasis. Clinically and histologically, APPP is indistinguishable from pustular psoriasis. Unlike most cases of acute psoriasis, however, APPP often appears in pregnancy without any personal or family history of psoriasis and usually ceases when the pregnancy ends. This fact, combined with reports of increased fetal and maternal morbidity and mortality associated with APPP, leads us to include it here.²⁶

Presentation. APPP is a rare condition that may have an onset at any point in pregnancy. Characteristic lesions begin as erythematous plaques with pustules on the inner thighs, flexural areas, and groin and spread to the trunk and extremities. As plaques enlarge, their center becomes eroded and crusted.

The nails may become onycholytic. Hands, feet, and face are usually spared. Oral and esophageal erosions can occur. Pruritus is typically mild, although the lesions are often painful. Flu-like symptoms are often present.²⁷

Pathophysiology. The pathophysiology of APPP is unknown.

Differential diagnosis. Conditions with similar presentations include an adverse drug reaction, pityriasis rosea, lichen simplex chronicus, eczema, lupus, and pityriasis rubra pilaris.

Diagnosis. The clinical history and an association with systemic illness are the basis for a diagnosis of APPP. Cultures of pustules are negative for any infective pathology, although, as the disease progresses, pustules may become superinfected. Laboratory testing may show an increased erythrocyte sedimentation rate, hypocalcemia, and a low level of vitamin D.

Treatment. Prednisone, 15 to 60 mg/d, is often sufficient to control the disease.²⁷ Cyclosporine, 100 mg twice daily, has also been shown to be useful.²⁸ Cyclosporine in pregnancy is a Category C drug. Data on fetal malformation associated with cyclosporine therapy are limited, but risk appears minimal.⁶

Maternal hypocalcemia should be monitored and treated appropriately. If disease progression is judged serious enough, early induction of labor is indicated, because delivery will almost always lead to swift resolution.

Sequelae. A number of case reports link APPP to serious sequelae, including fetal growth retardation, hypocalcemia, and stillbirth.^27,29,30 The condition is too rare, however, for good data on specific sequelae. Although APPP does give significant cause for concern, it appears that some of the traditional apprehension comes from older publications reporting a rate of maternal mortality of 70% to 90%.³¹ This statistic has not been borne out in practice. It does appear that the mother will frequently suffer systemic symptoms, including fever and malaise.

6. Pruritic folliculitis of pregnancy

Accounts of the prevalence of pruritic folliculitis of pregnancy (PFP) vary widely. Some sources report fewer than 30 cases in all of the literature; others indicate that the prevalence is equivalent to that of PG—one in every 10,000 pregnancies.^3,11 PFP most often presents in the third trimester. It often resolves before delivery, but uniformly clears within 2 weeks of delivery.

Presentation. PFP presents as papules and pustules concentrated around hair follicles (FIGURE 5). Often, lesions begin on the abdomen and spread to the extremities.^24,28 The condition is often, but not always, pruritic. Patients are more likely to be concerned about what the condition means for their health than distressed by the symptoms.

FIGURE 5 Pruritic folliculitis of pregnancy

The papules and pustules of PFP are concentrated around hair follicles.

Pathophysiology. Like many other dermatoses of pregnancy, the pathophysiology of PFP is unknown. There is little evidence that the condition is immunologically or hormonally mediated, and there is no evidence of an infectious component.^24,28

Differential diagnosis. PFP must be distinguished from infectious folliculitis, acneiform disorders, HIV-associated eosinophilic folliculitis, and a drug reaction.

Diagnosis. The clinical diagnosis is based on presenting symptoms and third-trimester onset. No specific laboratory or histologic analysis can be used to make a definitive diagnosis.

Treatment. As the condition is, by definition, a nonmicrobial folliculitis, the most effective therapy tends to be with a low- or midpotency topical corticosteroid, such as triamcinolone or desonide. A benzoyl peroxide wash can also be effective.

Sequelae. One study reports an increased incidence of low birth weight, but no associated morbidity or mortality has been reported in recent studies.²⁴

A recent report¹ by the National Highway Traffic Safety Administration revealed that deaths from automobile accidents have declined strikingly since 1954, reaching the lowest level ever recorded today: 1.16 fatalities for every 100 million vehicle-miles traveled. Why this improvement? Reasons include an increase in seat belt use and a lower rate of drunk driving, both secondary effects of better law enforcement, safer roads, and people just driving less.

There are lessons to be learned from this success that can be applied to obstetrical care and, potentially, to lower the unacceptably high rate of maternal mortality in this country.

Things aren’t working as we need them to

Truly, it is a reflection of a broken health care system that we, as health care providers, cannot say that we have seen a dramatic decrease in maternal mortality. In 2006, the latest year for which statistics are available, maternal mortality in the United States was at the highest level recorded since 1991: 13.3 deaths for every 100,000 live births (in California, the rate is 16.9 for every 100,000!).

Remember the Healthy People 2010 benchmark of 4.3 maternal deaths for every 100,000 live births? Clearly, we will not see that number any time in the near future; achieving it is as likely to happen as getting the national cesarean delivery rate to fall below 15% from the current level above 30%, or the return of vaginal breech deliveries!

Maternal mortality numbers are even more distressing when you appreciate that 1) African American women have three to four times the death rate from pregnancy complications that white women do and 2) minorities (African Americans, Latinos) are having the majority of pregnancies.

Why the rise in maternal deaths?

Multiple reasons are cited for the rise in maternal mortality. They include:

• difficult access to care
• racial discrimination
• increase in substance abuse
• barriers to communication with non–English-speaking patients
• lack of health insurance
• inflexible appointment hours
• lack of transportation to visits
• serious comorbidities (e.g., diabetes).

Each of these has some validity. But my opinion (and many of you will disagree with me) is that our society has not addressed the real issues driving the problem—issues that are difficult to face:

First, some women have too many pregnancies.

Second, certain women who have serious medical conditions, including morbid obesity and substance abuse, are having pregnancies when they should not have any at all.

It is an interesting observation that a person needs a license to drive a car or hunt deer but only a simple act of intercourse with no contraception to become pregnant.

What I’ve learned from my work

I have spent more than 35 years caring for pregnant women of low socioeconomic status. I am proud to have played that role. The two constant themes in my work with these patients are that 1) they feel a lack of control over most aspects of their life and 2) they lack self-esteem.

Pregnancy, however, is the one thing that they can control. It is the one time when others become interested in their welfare, and their best opportunity to have a something (a baby) that is all theirs.

But their adherence to appointments and laboratory testing is often poor. And their use of contraception between pregnancies is sporadic.

The 18-year-old woman who is in her second pregnancy often has a mother who is in her early 30s and a grandmother around 50. For these women, and for their children and grandchildren, the cycle of poverty is repeated endlessly. No matter how much money or effort is put into providing resources to care for them, little has changed in the past 30 years, and there is, I believe, little hope for change in the future. The only solutions that I can discern that will decisively break the cycle of poverty, lower the rate of maternal mortality, and improve the well being of women and their newborns are education and contraception.

My three-pronged proposal

Several actions can improve the outcome of a pregnancy for mothers of low socioeconomic status and their babies. I understand that many people will find these actions coercive and prejudicial, which truly they are not.

• All forms of contraception should be free for women, and access to contraception should be easy. Such a policy would result in fewer unintended pregnancies and abortions and better pregnancy spacing. The expense of free and universal contraception offers a great return on investment compared to the cost of care for a premature baby who spends any length of time in the neonatal intensive care unit.

Young women should be compensated for using whatever contraceptive they choose by being given vouchers that can be redeemed for material goods at select venues or free minutes for their cellular phones.

• All women who qualify for Medicaid coverage of pregnancy should be aggressively encouraged to have a preconception visit at least 3 to 6 months before they plan to become pregnant to assess any risks to themselves or their fetus and have a program put in place to maximize outcomes. This encouragement can take place at any visit in which contraception is discussed.

At this visit, an attempt can be made to optimize their clinical condition by:

• counseling them about drug and alcohol use and smoking
• assessing their risk of genetic disorders
• initiating folic acid and dietary modifications
• performing appropriate screening tests (i.e., blood glucose).

Appropriate consideration should be given, and discussion held, during the visit about whether the patient should even consider becoming pregnant. She should be given a realistic assessment of the risk of pregnancy and childbirth to her and the baby—including the potential for death.

The preconception assessment should be conducted by a physician, a midwife, or a nurse practitioner who has not been involved in the care of the patient. Doing so will minimize the introduction of any bias into the conversation by a treating physician.

When it is in a woman’s medical best interest not to conceive, she should become eligible for expedited adoption and be compensated for each reproductive year in which she does not conceive.

Last, a strong financial incentive should be offered to women who complete this preconception evaluation.

Many will say that such a program is unfair and prejudicial to women of lower socioeconomic status. But a precedent exists: Medicaid prohibits the performance of a sterilization procedure unless a signed permit has been in place 30 days or longer.

• Pregnant women who adhere to a prenatal care plan should be compensated with vouchers that can be redeemed for baby items, maternity clothes, or food for the family at select venues. They should also be compensated for keeping prenatal appointments; obtaining timely laboratory tests; attending prenatal classes; avoiding drugs, alcohol and smoking; returning for postpartum assessment; and using reliable contraception.

Education + contraception = fewer deaths?

Would such a plan work? I am convinced that it is worth trying. What do you think? Send your comments to me at [email protected]!

A recent report¹ by the National Highway Traffic Safety Administration revealed that deaths from automobile accidents have declined strikingly since 1954, reaching the lowest level ever recorded today: 1.16 fatalities for every 100 million vehicle-miles traveled. Why this improvement? Reasons include an increase in seat belt use and a lower rate of drunk driving, both secondary effects of better law enforcement, safer roads, and people just driving less.

There are lessons to be learned from this success that can be applied to obstetrical care and, potentially, to lower the unacceptably high rate of maternal mortality in this country.

Things aren’t working as we need them to

Truly, it is a reflection of a broken health care system that we, as health care providers, cannot say that we have seen a dramatic decrease in maternal mortality. In 2006, the latest year for which statistics are available, maternal mortality in the United States was at the highest level recorded since 1991: 13.3 deaths for every 100,000 live births (in California, the rate is 16.9 for every 100,000!).

Remember the Healthy People 2010 benchmark of 4.3 maternal deaths for every 100,000 live births? Clearly, we will not see that number any time in the near future; achieving it is as likely to happen as getting the national cesarean delivery rate to fall below 15% from the current level above 30%, or the return of vaginal breech deliveries!

Maternal mortality numbers are even more distressing when you appreciate that 1) African American women have three to four times the death rate from pregnancy complications that white women do and 2) minorities (African Americans, Latinos) are having the majority of pregnancies.

Why the rise in maternal deaths?

Multiple reasons are cited for the rise in maternal mortality. They include:

• difficult access to care
• racial discrimination
• increase in substance abuse
• barriers to communication with non–English-speaking patients
• lack of health insurance
• inflexible appointment hours
• lack of transportation to visits
• serious comorbidities (e.g., diabetes).

Each of these has some validity. But my opinion (and many of you will disagree with me) is that our society has not addressed the real issues driving the problem—issues that are difficult to face:

First, some women have too many pregnancies.

Second, certain women who have serious medical conditions, including morbid obesity and substance abuse, are having pregnancies when they should not have any at all.

It is an interesting observation that a person needs a license to drive a car or hunt deer but only a simple act of intercourse with no contraception to become pregnant.

What I’ve learned from my work

I have spent more than 35 years caring for pregnant women of low socioeconomic status. I am proud to have played that role. The two constant themes in my work with these patients are that 1) they feel a lack of control over most aspects of their life and 2) they lack self-esteem.

Pregnancy, however, is the one thing that they can control. It is the one time when others become interested in their welfare, and their best opportunity to have a something (a baby) that is all theirs.

But their adherence to appointments and laboratory testing is often poor. And their use of contraception between pregnancies is sporadic.

The 18-year-old woman who is in her second pregnancy often has a mother who is in her early 30s and a grandmother around 50. For these women, and for their children and grandchildren, the cycle of poverty is repeated endlessly. No matter how much money or effort is put into providing resources to care for them, little has changed in the past 30 years, and there is, I believe, little hope for change in the future. The only solutions that I can discern that will decisively break the cycle of poverty, lower the rate of maternal mortality, and improve the well being of women and their newborns are education and contraception.

My three-pronged proposal

Several actions can improve the outcome of a pregnancy for mothers of low socioeconomic status and their babies. I understand that many people will find these actions coercive and prejudicial, which truly they are not.

• All forms of contraception should be free for women, and access to contraception should be easy. Such a policy would result in fewer unintended pregnancies and abortions and better pregnancy spacing. The expense of free and universal contraception offers a great return on investment compared to the cost of care for a premature baby who spends any length of time in the neonatal intensive care unit.

Young women should be compensated for using whatever contraceptive they choose by being given vouchers that can be redeemed for material goods at select venues or free minutes for their cellular phones.

• All women who qualify for Medicaid coverage of pregnancy should be aggressively encouraged to have a preconception visit at least 3 to 6 months before they plan to become pregnant to assess any risks to themselves or their fetus and have a program put in place to maximize outcomes. This encouragement can take place at any visit in which contraception is discussed.

At this visit, an attempt can be made to optimize their clinical condition by:

• counseling them about drug and alcohol use and smoking
• assessing their risk of genetic disorders
• initiating folic acid and dietary modifications
• performing appropriate screening tests (i.e., blood glucose).

Appropriate consideration should be given, and discussion held, during the visit about whether the patient should even consider becoming pregnant. She should be given a realistic assessment of the risk of pregnancy and childbirth to her and the baby—including the potential for death.

The preconception assessment should be conducted by a physician, a midwife, or a nurse practitioner who has not been involved in the care of the patient. Doing so will minimize the introduction of any bias into the conversation by a treating physician.

When it is in a woman’s medical best interest not to conceive, she should become eligible for expedited adoption and be compensated for each reproductive year in which she does not conceive.

Last, a strong financial incentive should be offered to women who complete this preconception evaluation.

Many will say that such a program is unfair and prejudicial to women of lower socioeconomic status. But a precedent exists: Medicaid prohibits the performance of a sterilization procedure unless a signed permit has been in place 30 days or longer.

• Pregnant women who adhere to a prenatal care plan should be compensated with vouchers that can be redeemed for baby items, maternity clothes, or food for the family at select venues. They should also be compensated for keeping prenatal appointments; obtaining timely laboratory tests; attending prenatal classes; avoiding drugs, alcohol and smoking; returning for postpartum assessment; and using reliable contraception.

Education + contraception = fewer deaths?

Would such a plan work? I am convinced that it is worth trying. What do you think? Send your comments to me at [email protected]!

A recent report¹ by the National Highway Traffic Safety Administration revealed that deaths from automobile accidents have declined strikingly since 1954, reaching the lowest level ever recorded today: 1.16 fatalities for every 100 million vehicle-miles traveled. Why this improvement? Reasons include an increase in seat belt use and a lower rate of drunk driving, both secondary effects of better law enforcement, safer roads, and people just driving less.

There are lessons to be learned from this success that can be applied to obstetrical care and, potentially, to lower the unacceptably high rate of maternal mortality in this country.

Things aren’t working as we need them to

Truly, it is a reflection of a broken health care system that we, as health care providers, cannot say that we have seen a dramatic decrease in maternal mortality. In 2006, the latest year for which statistics are available, maternal mortality in the United States was at the highest level recorded since 1991: 13.3 deaths for every 100,000 live births (in California, the rate is 16.9 for every 100,000!).

Remember the Healthy People 2010 benchmark of 4.3 maternal deaths for every 100,000 live births? Clearly, we will not see that number any time in the near future; achieving it is as likely to happen as getting the national cesarean delivery rate to fall below 15% from the current level above 30%, or the return of vaginal breech deliveries!

Maternal mortality numbers are even more distressing when you appreciate that 1) African American women have three to four times the death rate from pregnancy complications that white women do and 2) minorities (African Americans, Latinos) are having the majority of pregnancies.

Why the rise in maternal deaths?

Multiple reasons are cited for the rise in maternal mortality. They include:

• difficult access to care
• racial discrimination
• increase in substance abuse
• barriers to communication with non–English-speaking patients
• lack of health insurance
• inflexible appointment hours
• lack of transportation to visits
• serious comorbidities (e.g., diabetes).

Each of these has some validity. But my opinion (and many of you will disagree with me) is that our society has not addressed the real issues driving the problem—issues that are difficult to face:

First, some women have too many pregnancies.

Second, certain women who have serious medical conditions, including morbid obesity and substance abuse, are having pregnancies when they should not have any at all.

It is an interesting observation that a person needs a license to drive a car or hunt deer but only a simple act of intercourse with no contraception to become pregnant.

What I’ve learned from my work

I have spent more than 35 years caring for pregnant women of low socioeconomic status. I am proud to have played that role. The two constant themes in my work with these patients are that 1) they feel a lack of control over most aspects of their life and 2) they lack self-esteem.

Pregnancy, however, is the one thing that they can control. It is the one time when others become interested in their welfare, and their best opportunity to have a something (a baby) that is all theirs.

But their adherence to appointments and laboratory testing is often poor. And their use of contraception between pregnancies is sporadic.

The 18-year-old woman who is in her second pregnancy often has a mother who is in her early 30s and a grandmother around 50. For these women, and for their children and grandchildren, the cycle of poverty is repeated endlessly. No matter how much money or effort is put into providing resources to care for them, little has changed in the past 30 years, and there is, I believe, little hope for change in the future. The only solutions that I can discern that will decisively break the cycle of poverty, lower the rate of maternal mortality, and improve the well being of women and their newborns are education and contraception.

My three-pronged proposal

Several actions can improve the outcome of a pregnancy for mothers of low socioeconomic status and their babies. I understand that many people will find these actions coercive and prejudicial, which truly they are not.

• All forms of contraception should be free for women, and access to contraception should be easy. Such a policy would result in fewer unintended pregnancies and abortions and better pregnancy spacing. The expense of free and universal contraception offers a great return on investment compared to the cost of care for a premature baby who spends any length of time in the neonatal intensive care unit.

Young women should be compensated for using whatever contraceptive they choose by being given vouchers that can be redeemed for material goods at select venues or free minutes for their cellular phones.

• All women who qualify for Medicaid coverage of pregnancy should be aggressively encouraged to have a preconception visit at least 3 to 6 months before they plan to become pregnant to assess any risks to themselves or their fetus and have a program put in place to maximize outcomes. This encouragement can take place at any visit in which contraception is discussed.

At this visit, an attempt can be made to optimize their clinical condition by:

• counseling them about drug and alcohol use and smoking
• assessing their risk of genetic disorders
• initiating folic acid and dietary modifications
• performing appropriate screening tests (i.e., blood glucose).

Appropriate consideration should be given, and discussion held, during the visit about whether the patient should even consider becoming pregnant. She should be given a realistic assessment of the risk of pregnancy and childbirth to her and the baby—including the potential for death.

The preconception assessment should be conducted by a physician, a midwife, or a nurse practitioner who has not been involved in the care of the patient. Doing so will minimize the introduction of any bias into the conversation by a treating physician.

When it is in a woman’s medical best interest not to conceive, she should become eligible for expedited adoption and be compensated for each reproductive year in which she does not conceive.

Last, a strong financial incentive should be offered to women who complete this preconception evaluation.

Many will say that such a program is unfair and prejudicial to women of lower socioeconomic status. But a precedent exists: Medicaid prohibits the performance of a sterilization procedure unless a signed permit has been in place 30 days or longer.

• Pregnant women who adhere to a prenatal care plan should be compensated with vouchers that can be redeemed for baby items, maternity clothes, or food for the family at select venues. They should also be compensated for keeping prenatal appointments; obtaining timely laboratory tests; attending prenatal classes; avoiding drugs, alcohol and smoking; returning for postpartum assessment; and using reliable contraception.

Education + contraception = fewer deaths?

Would such a plan work? I am convinced that it is worth trying. What do you think? Send your comments to me at [email protected]!

Tamoxifen is widely used by women who have receptor-positive breast cancer to lower their risk of recurrence. Tamoxifen is a prodrug converted by the hepatic cytochrome P450 enzyme system (particularly isoenzyme CYP2D6) to active metabolites, which have much higher affinity for the estrogen receptor than the parent prodrug has. Selective serotonin reuptake inhibitors (SSRIs) inhibit CYP2D6 to variable degrees, with paroxetine thought to be the most potent inhibitor.

The fact that SSRIs inhibit CYP2D6 is important because as many as one in every four women who have breast cancer experiences depression as well.¹ SSRIs are widely prescribed for these patients—not only as treatment for their depressive disorder but also because the drugs have been shown to offer some relief from hot flashes.

In this cohort study from Ontario, where health care is delivered in a centralized fashion, Kelly and colleagues assessed the survival of breast cancer survivors 66 years and older who were treated with both tamoxifen and an SSRI, or with both tamoxifen and venlafaxine (which inhibits reuptake of both serotonin and norepinephrine).

The median age of the 2,430 eligible and evaluable women at the time of tamoxifen initiation was 74 years. These women began using tamoxifen between 1993 and 2005, with a median duration of use of 4 years. Paroxetine was the most commonly used SSRI in this study (25.9%), followed by sertraline (Zoloft, 22.3%), citalopram (Celexa, 19.2%), venlafaxine (Effexor, 15.0%), fluoxetine (Prozac, Sarafem, 10.4%), and fluvoxamine (Luvox, 7.2%).

Neither duloxetine (Cymbalta) nor escitalopram (Lexapro) were included in this analysis.

By the end of follow-up, 1,074 women (44.2%) in the cohort had died, 374 of them from breast cancer.

When taken with tamoxifen, paroxetine increased breast cancer mortality

Concomitant use of paroxetine and tamoxifen was associated with a significantly increased risk of death from breast cancer. In addition, as the proportion of concomitant use of the drugs increased to 25%, 50%, and 75% of total tamoxifen use, the adjusted risk of death increased 24%, 54%, and 91%, respectively. Among women taking tamoxifen, concomitant use of other SSRIs or venlafaxine was not associated with an increased risk of death.

Based on their data, the investigators determined that use of paroxetine during 41% of tamoxifen treatment (the median overlap observed in this study) would result in one additional breast cancer death at 5 years for every 19.7 women treated in this fashion. If paroxetine were used for the entire duration of tamoxifen treatment—an overlap of 100%—an additional death would occur for every 6.9 women treated.

The authors also noted that fluoxetine is a “strong inhibitor” of tamoxifen, although it was not associated with an increased risk of breast cancer death in this trial. One reason may be the small number of women who took tamoxifen and fluoxetine concomitantly in this study.

“Our results should not be viewed as evidence that fluoxetine can be safely used in combination with tamoxifen,” Kelly and coworkers write. “Similarly, we cannot exclude the possibility that insufficient sample size explains the nonsignificant mortality results with other SSRIs.”

Tamoxifen is widely used by women who have receptor-positive breast cancer to lower their risk of recurrence. Tamoxifen is a prodrug converted by the hepatic cytochrome P450 enzyme system (particularly isoenzyme CYP2D6) to active metabolites, which have much higher affinity for the estrogen receptor than the parent prodrug has. Selective serotonin reuptake inhibitors (SSRIs) inhibit CYP2D6 to variable degrees, with paroxetine thought to be the most potent inhibitor.

The fact that SSRIs inhibit CYP2D6 is important because as many as one in every four women who have breast cancer experiences depression as well.¹ SSRIs are widely prescribed for these patients—not only as treatment for their depressive disorder but also because the drugs have been shown to offer some relief from hot flashes.

In this cohort study from Ontario, where health care is delivered in a centralized fashion, Kelly and colleagues assessed the survival of breast cancer survivors 66 years and older who were treated with both tamoxifen and an SSRI, or with both tamoxifen and venlafaxine (which inhibits reuptake of both serotonin and norepinephrine).

The median age of the 2,430 eligible and evaluable women at the time of tamoxifen initiation was 74 years. These women began using tamoxifen between 1993 and 2005, with a median duration of use of 4 years. Paroxetine was the most commonly used SSRI in this study (25.9%), followed by sertraline (Zoloft, 22.3%), citalopram (Celexa, 19.2%), venlafaxine (Effexor, 15.0%), fluoxetine (Prozac, Sarafem, 10.4%), and fluvoxamine (Luvox, 7.2%).

Neither duloxetine (Cymbalta) nor escitalopram (Lexapro) were included in this analysis.

By the end of follow-up, 1,074 women (44.2%) in the cohort had died, 374 of them from breast cancer.

When taken with tamoxifen, paroxetine increased breast cancer mortality

Concomitant use of paroxetine and tamoxifen was associated with a significantly increased risk of death from breast cancer. In addition, as the proportion of concomitant use of the drugs increased to 25%, 50%, and 75% of total tamoxifen use, the adjusted risk of death increased 24%, 54%, and 91%, respectively. Among women taking tamoxifen, concomitant use of other SSRIs or venlafaxine was not associated with an increased risk of death.

Based on their data, the investigators determined that use of paroxetine during 41% of tamoxifen treatment (the median overlap observed in this study) would result in one additional breast cancer death at 5 years for every 19.7 women treated in this fashion. If paroxetine were used for the entire duration of tamoxifen treatment—an overlap of 100%—an additional death would occur for every 6.9 women treated.

The authors also noted that fluoxetine is a “strong inhibitor” of tamoxifen, although it was not associated with an increased risk of breast cancer death in this trial. One reason may be the small number of women who took tamoxifen and fluoxetine concomitantly in this study.

“Our results should not be viewed as evidence that fluoxetine can be safely used in combination with tamoxifen,” Kelly and coworkers write. “Similarly, we cannot exclude the possibility that insufficient sample size explains the nonsignificant mortality results with other SSRIs.”

Tamoxifen is widely used by women who have receptor-positive breast cancer to lower their risk of recurrence. Tamoxifen is a prodrug converted by the hepatic cytochrome P450 enzyme system (particularly isoenzyme CYP2D6) to active metabolites, which have much higher affinity for the estrogen receptor than the parent prodrug has. Selective serotonin reuptake inhibitors (SSRIs) inhibit CYP2D6 to variable degrees, with paroxetine thought to be the most potent inhibitor.

The fact that SSRIs inhibit CYP2D6 is important because as many as one in every four women who have breast cancer experiences depression as well.¹ SSRIs are widely prescribed for these patients—not only as treatment for their depressive disorder but also because the drugs have been shown to offer some relief from hot flashes.

In this cohort study from Ontario, where health care is delivered in a centralized fashion, Kelly and colleagues assessed the survival of breast cancer survivors 66 years and older who were treated with both tamoxifen and an SSRI, or with both tamoxifen and venlafaxine (which inhibits reuptake of both serotonin and norepinephrine).

The median age of the 2,430 eligible and evaluable women at the time of tamoxifen initiation was 74 years. These women began using tamoxifen between 1993 and 2005, with a median duration of use of 4 years. Paroxetine was the most commonly used SSRI in this study (25.9%), followed by sertraline (Zoloft, 22.3%), citalopram (Celexa, 19.2%), venlafaxine (Effexor, 15.0%), fluoxetine (Prozac, Sarafem, 10.4%), and fluvoxamine (Luvox, 7.2%).

Neither duloxetine (Cymbalta) nor escitalopram (Lexapro) were included in this analysis.

By the end of follow-up, 1,074 women (44.2%) in the cohort had died, 374 of them from breast cancer.

When taken with tamoxifen, paroxetine increased breast cancer mortality

Concomitant use of paroxetine and tamoxifen was associated with a significantly increased risk of death from breast cancer. In addition, as the proportion of concomitant use of the drugs increased to 25%, 50%, and 75% of total tamoxifen use, the adjusted risk of death increased 24%, 54%, and 91%, respectively. Among women taking tamoxifen, concomitant use of other SSRIs or venlafaxine was not associated with an increased risk of death.

Based on their data, the investigators determined that use of paroxetine during 41% of tamoxifen treatment (the median overlap observed in this study) would result in one additional breast cancer death at 5 years for every 19.7 women treated in this fashion. If paroxetine were used for the entire duration of tamoxifen treatment—an overlap of 100%—an additional death would occur for every 6.9 women treated.

The authors also noted that fluoxetine is a “strong inhibitor” of tamoxifen, although it was not associated with an increased risk of breast cancer death in this trial. One reason may be the small number of women who took tamoxifen and fluoxetine concomitantly in this study.

“Our results should not be viewed as evidence that fluoxetine can be safely used in combination with tamoxifen,” Kelly and coworkers write. “Similarly, we cannot exclude the possibility that insufficient sample size explains the nonsignificant mortality results with other SSRIs.”

A new study reveals a trend that most of us on the front line of medical education see daily: The number of hours that any given physician works in a week has decreased dramatically over the past decade, along with the fees that he or she collects.¹

The retrospective analysis used data from the US Census Bureau Current Population Survey from 1976 through 2008 (n = 116,733). Trends were estimated among all US physicians and by residency status, sex, age, and work setting. As the authors point out, the decrease in work hours was “broad-based,” occurring in different demographic groups and in different settings.¹

Among the findings:

• The mean weekly workload remained steady through the early 1990s, but declined 7.2% between 1996 and 2008 among all physicians, from 54.9 hours in 1996–1998 to 51.0 hours in 2006–2008 (95% confidence interval [CI], 5.3%–9.0%; P < .001).
• The hours worked by nonresident physicians decreased by 5.7% (95% CI, 3.8%–7.7%; P < .001), versus 9.8% among residents (95% CI, 5.8%–13.7%; P < .001).
• Overall, among nonresident physicians, those younger than 45 years experienced the biggest decrease in mean work hours (7.4%; 95% CI, 4.7%–10.2%; P < .001). Non-resident physicians who worked outside a hospital setting also saw a significant decrease (6.4%; 95% CI, 4.1%–8.7%; P < .001).
• Overall, nonresident physicians older than 45 years had the smallest decrease in mean work hours (3.7%; 95% CI, 1.0%–6.5%; P < .008). Nonresident physicians who worked in a hospital setting also experienced a slim reduction (4.0%; 95% CI, 0.4%–7.6%; P =.03).
• Mean physician fees declined 25% (after adjustment for inflation) between 1995 and 2006, coincident with diminishing work hours.
• The largest reduction in mean work hours—9.8% among resident physicians—came after the imposition of limits on resident work hours by the American Council of Graduate Medical Education in 2003.¹

Which came first—lower reimbursement or fewer hours?

“The observed decrease in physician fees is a potential economic factor behind the decrease in physician hours,” the authors write.¹ For example, hours diminished first in metropolitan statistical areas where fees were lowest in 2001.

The General Accounting Office found that these areas also tended to feature greater competition among physicians, as well as greater penetration by managed care, which may have contributed to lower fees. However, these conditions also could have “weakened physician autonomy and reduced job satisfaction, which in turn could have led to fewer hours.”¹

Study sheds light on gender shift, generational differences

In this study, Staiger and colleagues documented the reality of the “gender shift” in medicine: Women constituted 20.6% of respondents and worked a mean of 44.4 hours a week in 2006–2008, compared with 51.7 hours for men.¹ However, the authors point out, “[a]lthough female physicians work fewer mean hours and represent an increasing proportion of physicians, decreases in hours were observed for both male and female physicians.”¹

Although physicians of all ages worked fewer hours than before, the reduction was greatest among physicians younger than 45 years.¹

The study also demonstrated that physicians work significantly longer hours than lawyers, engineers, and nurses. The work hours of these other professions have remained relatively stable since 1976.¹

How do fewer hours affect the supply of physicians?

Over the past 10 years, much has been written about a projected shortage in the physician workforce and what should be done about it. In response, the Association of American Medical Colleges requested an expansion in medical school capacity, and many medical schools increased class size. In most cases, however, the reason for the increase in class size was financial: a desire for more tuition dollars, one of the few sources of income under the control of a medical school.

In addition, medical school tuition has continued to rise, and the increase in tuition has markedly outpaced the rate of inflation. The result: Medical students graduate with a large financial debt that influences their choice of specialty. Specialties that have high remuneration such as radiology, dermatology, anesthesiology, and orthopedics have become prime choices for graduating medical students.

The system needs to catch up to reality

As Staiger and his coauthors point out, a 5.7% decrease in the mean number of hours worked among approximately 630,000 physicians is equivalent to a loss of 36,000 physicians from the workforce.¹ If the current system is sincere about increasing the physician workforce, why haven’t there been 1) a decrease in the length of medical school, 2) control of tuition escalation, and 3) an increase in the number of residency positions? And why has nothing been done to bring some of the lost physicians back in to the workplace by offering part time or flexible positions?

We won’t solve the physician shortage simply by increasing enrollment and building new medical schools. It’s time to realize that physicians want a work-lifestyle balance similar to that of other professionals. If we don’t make the practice of medicine more enjoyable and align it with the needs and desires of Generation X and Generation Y physicians, we’ll fall short of the number of physicians that we need to care for a growing, and an aging, population.

A new study reveals a trend that most of us on the front line of medical education see daily: The number of hours that any given physician works in a week has decreased dramatically over the past decade, along with the fees that he or she collects.¹

The retrospective analysis used data from the US Census Bureau Current Population Survey from 1976 through 2008 (n = 116,733). Trends were estimated among all US physicians and by residency status, sex, age, and work setting. As the authors point out, the decrease in work hours was “broad-based,” occurring in different demographic groups and in different settings.¹

Among the findings:

• The mean weekly workload remained steady through the early 1990s, but declined 7.2% between 1996 and 2008 among all physicians, from 54.9 hours in 1996–1998 to 51.0 hours in 2006–2008 (95% confidence interval [CI], 5.3%–9.0%; P < .001).
• The hours worked by nonresident physicians decreased by 5.7% (95% CI, 3.8%–7.7%; P < .001), versus 9.8% among residents (95% CI, 5.8%–13.7%; P < .001).
• Overall, among nonresident physicians, those younger than 45 years experienced the biggest decrease in mean work hours (7.4%; 95% CI, 4.7%–10.2%; P < .001). Non-resident physicians who worked outside a hospital setting also saw a significant decrease (6.4%; 95% CI, 4.1%–8.7%; P < .001).
• Overall, nonresident physicians older than 45 years had the smallest decrease in mean work hours (3.7%; 95% CI, 1.0%–6.5%; P < .008). Nonresident physicians who worked in a hospital setting also experienced a slim reduction (4.0%; 95% CI, 0.4%–7.6%; P =.03).
• Mean physician fees declined 25% (after adjustment for inflation) between 1995 and 2006, coincident with diminishing work hours.
• The largest reduction in mean work hours—9.8% among resident physicians—came after the imposition of limits on resident work hours by the American Council of Graduate Medical Education in 2003.¹

Which came first—lower reimbursement or fewer hours?

“The observed decrease in physician fees is a potential economic factor behind the decrease in physician hours,” the authors write.¹ For example, hours diminished first in metropolitan statistical areas where fees were lowest in 2001.

The General Accounting Office found that these areas also tended to feature greater competition among physicians, as well as greater penetration by managed care, which may have contributed to lower fees. However, these conditions also could have “weakened physician autonomy and reduced job satisfaction, which in turn could have led to fewer hours.”¹

Study sheds light on gender shift, generational differences

In this study, Staiger and colleagues documented the reality of the “gender shift” in medicine: Women constituted 20.6% of respondents and worked a mean of 44.4 hours a week in 2006–2008, compared with 51.7 hours for men.¹ However, the authors point out, “[a]lthough female physicians work fewer mean hours and represent an increasing proportion of physicians, decreases in hours were observed for both male and female physicians.”¹

Although physicians of all ages worked fewer hours than before, the reduction was greatest among physicians younger than 45 years.¹

The study also demonstrated that physicians work significantly longer hours than lawyers, engineers, and nurses. The work hours of these other professions have remained relatively stable since 1976.¹

How do fewer hours affect the supply of physicians?

Over the past 10 years, much has been written about a projected shortage in the physician workforce and what should be done about it. In response, the Association of American Medical Colleges requested an expansion in medical school capacity, and many medical schools increased class size. In most cases, however, the reason for the increase in class size was financial: a desire for more tuition dollars, one of the few sources of income under the control of a medical school.

In addition, medical school tuition has continued to rise, and the increase in tuition has markedly outpaced the rate of inflation. The result: Medical students graduate with a large financial debt that influences their choice of specialty. Specialties that have high remuneration such as radiology, dermatology, anesthesiology, and orthopedics have become prime choices for graduating medical students.

The system needs to catch up to reality

As Staiger and his coauthors point out, a 5.7% decrease in the mean number of hours worked among approximately 630,000 physicians is equivalent to a loss of 36,000 physicians from the workforce.¹ If the current system is sincere about increasing the physician workforce, why haven’t there been 1) a decrease in the length of medical school, 2) control of tuition escalation, and 3) an increase in the number of residency positions? And why has nothing been done to bring some of the lost physicians back in to the workplace by offering part time or flexible positions?

We won’t solve the physician shortage simply by increasing enrollment and building new medical schools. It’s time to realize that physicians want a work-lifestyle balance similar to that of other professionals. If we don’t make the practice of medicine more enjoyable and align it with the needs and desires of Generation X and Generation Y physicians, we’ll fall short of the number of physicians that we need to care for a growing, and an aging, population.

A new study reveals a trend that most of us on the front line of medical education see daily: The number of hours that any given physician works in a week has decreased dramatically over the past decade, along with the fees that he or she collects.¹

The retrospective analysis used data from the US Census Bureau Current Population Survey from 1976 through 2008 (n = 116,733). Trends were estimated among all US physicians and by residency status, sex, age, and work setting. As the authors point out, the decrease in work hours was “broad-based,” occurring in different demographic groups and in different settings.¹

Among the findings:

• The mean weekly workload remained steady through the early 1990s, but declined 7.2% between 1996 and 2008 among all physicians, from 54.9 hours in 1996–1998 to 51.0 hours in 2006–2008 (95% confidence interval [CI], 5.3%–9.0%; P < .001).
• The hours worked by nonresident physicians decreased by 5.7% (95% CI, 3.8%–7.7%; P < .001), versus 9.8% among residents (95% CI, 5.8%–13.7%; P < .001).
• Overall, among nonresident physicians, those younger than 45 years experienced the biggest decrease in mean work hours (7.4%; 95% CI, 4.7%–10.2%; P < .001). Non-resident physicians who worked outside a hospital setting also saw a significant decrease (6.4%; 95% CI, 4.1%–8.7%; P < .001).
• Overall, nonresident physicians older than 45 years had the smallest decrease in mean work hours (3.7%; 95% CI, 1.0%–6.5%; P < .008). Nonresident physicians who worked in a hospital setting also experienced a slim reduction (4.0%; 95% CI, 0.4%–7.6%; P =.03).
• Mean physician fees declined 25% (after adjustment for inflation) between 1995 and 2006, coincident with diminishing work hours.
• The largest reduction in mean work hours—9.8% among resident physicians—came after the imposition of limits on resident work hours by the American Council of Graduate Medical Education in 2003.¹

Which came first—lower reimbursement or fewer hours?

“The observed decrease in physician fees is a potential economic factor behind the decrease in physician hours,” the authors write.¹ For example, hours diminished first in metropolitan statistical areas where fees were lowest in 2001.

The General Accounting Office found that these areas also tended to feature greater competition among physicians, as well as greater penetration by managed care, which may have contributed to lower fees. However, these conditions also could have “weakened physician autonomy and reduced job satisfaction, which in turn could have led to fewer hours.”¹

Study sheds light on gender shift, generational differences

In this study, Staiger and colleagues documented the reality of the “gender shift” in medicine: Women constituted 20.6% of respondents and worked a mean of 44.4 hours a week in 2006–2008, compared with 51.7 hours for men.¹ However, the authors point out, “[a]lthough female physicians work fewer mean hours and represent an increasing proportion of physicians, decreases in hours were observed for both male and female physicians.”¹

Although physicians of all ages worked fewer hours than before, the reduction was greatest among physicians younger than 45 years.¹

The study also demonstrated that physicians work significantly longer hours than lawyers, engineers, and nurses. The work hours of these other professions have remained relatively stable since 1976.¹

How do fewer hours affect the supply of physicians?

Over the past 10 years, much has been written about a projected shortage in the physician workforce and what should be done about it. In response, the Association of American Medical Colleges requested an expansion in medical school capacity, and many medical schools increased class size. In most cases, however, the reason for the increase in class size was financial: a desire for more tuition dollars, one of the few sources of income under the control of a medical school.

In addition, medical school tuition has continued to rise, and the increase in tuition has markedly outpaced the rate of inflation. The result: Medical students graduate with a large financial debt that influences their choice of specialty. Specialties that have high remuneration such as radiology, dermatology, anesthesiology, and orthopedics have become prime choices for graduating medical students.

The system needs to catch up to reality

As Staiger and his coauthors point out, a 5.7% decrease in the mean number of hours worked among approximately 630,000 physicians is equivalent to a loss of 36,000 physicians from the workforce.¹ If the current system is sincere about increasing the physician workforce, why haven’t there been 1) a decrease in the length of medical school, 2) control of tuition escalation, and 3) an increase in the number of residency positions? And why has nothing been done to bring some of the lost physicians back in to the workplace by offering part time or flexible positions?

We won’t solve the physician shortage simply by increasing enrollment and building new medical schools. It’s time to realize that physicians want a work-lifestyle balance similar to that of other professionals. If we don’t make the practice of medicine more enjoyable and align it with the needs and desires of Generation X and Generation Y physicians, we’ll fall short of the number of physicians that we need to care for a growing, and an aging, population.