Susmita Pati, MD, MPH

Patient flow refers to the management and movement of patients in health care settings and is linked to quality, safety, and cost.16 The intensive care unit (ICU) is crucial in patient flow.7, 8 The limited number of beds and the resource‐intensive services and staffing associated with them require that hospitals optimize their utilization, as is increasingly true of all hospital resources. To maximize delivery of services to patients who need them and minimize real and opportunity losses (eg, postponed surgery, diverted transports, or inability to accept patients), patients in ICU beds should receive critical care medicine/nursing services while there and be transferred or discharged when appropriate.

The time between arrival and departure from any area of the hospital, including the ICU, is considered the time when a patient is receiving needed clinical carethe value‐added portion of health care operationsand time waiting to move on to the next step.911 This period includes both necessary logistics (eg, signing out a patient or waiting a reasonable amount of time for room cleaning) and nonvalue‐added time (eg, an excessively long amount of time for room cleaning). Operations management labels nonvalue‐added time as waste, and its reduction is vital for high‐quality health care.9, 12, 13 As in other industries, one important way to understand value versus waste is through direct observation.11, 14 Although operating rooms have been the subject of several published process improvement projects to improve efficiency,1518 inpatient beds have not been the subject of such scrutiny. The objectives of this study were to generate a direct observation method and use it to describe pediatric ICU (PICU) bed utilization from a value‐added perspective.

METHODS

An interdisciplinary work group of physicians, nurses, quality improvement specialists, and 1 operations management expert developed an Excel spreadsheet to categorize hour‐by‐hour status of PICU beds. The clinicians generated a list of 27 activities. A critical care nurse trained in quality improvement piloted the list for 3 separate 4‐hour blocks over 2 weeks adding 18 activities; 2 additional activities were added during the 5 weeks of observation (Table 1). (The recording tool is provided in the Supporting Information Appendix.) Three observers with knowledge of medical terminology (2 third‐year medical students and 1 premedical student with years of experience as an emergency medical technician) were trained over 12 hours to conduct the observations. Prior to the observations, the 3 observers also spent time in the PICU, and terminology used for recordings was reviewed. Interobserver reliability was checked during 3 sets of observation circuits by all 3 observers and the principal investigator, as well as by spot checks during the study.

The targeted area included 24 single‐patient rooms. The activity of each bed was recorded hourly. Real‐time recording in to the Excel spreadsheet on a dedicated laptop occurred from 8:00 _AM until 11:00 _PM. The most visible or critical event was recorded. Although some activities were not mutually exclusive (eg, a patient could be ventilated and on a continuous infusion simultaneously), the objective was to identify when a room was being used for any critical care service, not enumerate all of them. The observers noted overnight events that occurred from 11:00 _PM to 8:00 _AM in the morning by reviewing the bedside record and talking to the staff to complete each day's 24‐hour recording. The observers also recorded the hospital‐wide census and the census for the other half of the PICU every 4 hours. The observations occurred over 5 noncontiguous weeks between January 2009 and April 2009.

After all observations were complete, activities were classified as critical care services (CCS) or noncritical‐care services (NCCS). NCCSs were further divided into necessary logistics (defined for analysis purposes as the first hour of any NCCS activity) or nonvalue‐added (the second or greater hour of NCCS). A time limit of 1 hour was chosen to define necessary logistics based on a consensus that nonclinical activities optimally would not take more than 1 hour each. We also analyzed results with 2 hours as the cutoff for necessary logistics. Admission, discharge, and transfer records were reviewed to check for returns to the PICU or hospital within 48 hours of transfer or discharge from the PICU.

Analyses were conducted using Microsoft Excel (Microsoft, Redmond, WA) and Stata 10.0 (StataCorp, College Station, TX). The study was approved by the Children's Hospital of Philadelphia Institutional Review Board with waiver of consent.

RESULTS

A total of 824 hours of recordings included 19,887 bed‐hours with 219 unique patients; among them, 2 remained from the first day of recording in January to the last day in April (sample recording in Figure 1). A total of 50 patients (range, 812 per week) stayed for the entirety of each 1‐week observation period. Of the 47 possible activities, 45 of them were recorded for at least 1 hour in the 5 weeks. Overall, 14 activities accounted for 95% of the observed bed‐hours and 31 activities accounted for the remaining 5%. CCS accounted for 82% of observed bed‐hours, NCCS accounted for 10.4%, and empty unassigned accounted for 8% (Figure 2). Using the 1‐hour cutoff for necessary services, 77% of NCCS time was nonvalue‐added, whereas 23% of it was necessary logistics; using the 2‐hour cutoff, 54% was nonvalue‐added, and 46% was necessary logistics.

Figure 1

Figure 2

During the observation period, <1% of bed‐hours were used for CCS for overflow patients from the neonatal ICU (NICU), cardiac care unit (CCU), cardiac ICU (CICU), or progressive care unit (PCU; tracheostomy/ventilator unit). Although only 4 patients required transport to a rehabilitation facility, their wait time comprised 99 hours (<1%) of total recordings. Eight patients waited a mean of 2.6 hours for transportation home (maximum, 10 hours).

To demonstrate the cycle of room use, activities were divided into 4 categories: room preparation, critical care services, disposition pending, and postcritical care services (Figure 3). As an example of detailed data revealed by direct observation, we identified 102 instances totaling 919 hours when a patient was waiting for a bed assignment on another floor (5% of all bed‐hours). The mean wait time was 9 hours (range, 188 hours) and the median time was 5.5 hours. There were only 15 instances when floor bed assignment took 1 hour or less, and only 9 instances when it took 12 hours. Similarly, considerable time was spent on cleaning rooms between patients: only 66 of 146 instances of cleaning took 1 hour or less. The mean time for cleaning was 2.2 hours (range, 115), and the median was 2 hours. (There were 136 recorded instances of room cleaning and 10 additional episodes that were not recorded but had to be completed for the room to turnover from one patient to the next, yielding a total of 146 instances of cleaning.)

Figure 3

From the 824 hours of recording, we identified 200 hours (25% of time) when there were zero empty unassigned beds available in the section of the PICU being observed. Episodes of full occupancy occurred mostly on weekdays, with 23% of hours of full capacity on Thursdays, 21% on Mondays, and 21% on Wednesdays; only 8% were on Saturdays and <1% on Sundays. These 200 hours fell into 36 separate episodes of complete occupancy, each lasting 122 hours. Each patient, on average, received 3.1 hours of NCCS during each episode of full occupancy (range, 111 hours). Within these 200 hours at capacity, we identified only 15 hours (8%) when all 24 beds were used for CCS. For 72% of the time, there was at least 1 bed with NCCS, and for 37% at least 2 beds. A small portion of the time (7%), the lack of beds was affected by occupancy by patients who should have been in the NICU, CICU, CCU, or PCU.

Data collected through direct observation can be used to understand aggregated and averaged experiences, but also more specific time periods. For example, we identified 1 week with the highest consistent level of occupancy and turnover: March 915 had empty unassigned beds for only 4% of the week. Of the 168 hours in the week, 68 (40%) had full capacity. However, for 90% of the time, at least 1 bed was used for a NCCS. Other analytic options included varying the assumptions around time needed for logistics. Overall, NCCS time on necessary logistics changes from 23% to 46% using 1 hour versus 2 hours as the cutoff. For floor bed assignments, assuming that the first hour of this activity is necessary logistics and any hour thereafter is not, 817 hours were wasted. Even after assuming 2 hours of necessary logistical time (which may also include steps such as nursing and physician sign‐out to the receiving team, often not recorded in the observations), this left 715 hours of NCCS time in which patients waited to be placed elsewhere in the hospital. For room cleaning, because recordings were hourly, but room cleaning could take less time, we performed a sensitivity analysis, converting all 1‐hour recordings to half‐hour recordings to half‐hour recordings (an exaggerated shortening since industry‐standard cleaning may take longer).

Of the 219 patients directly observed, 15 were noted to be waiting for a transfer out of the PICU but experienced a change in disposition before the transfer. On average, these patients waited 8 hours for a floor bed assignment (range, 221) before reverting to a CCS, which then lasted an average of 16.5 hours (range, 149). (Included in this group are 2 patients who experienced this change in disposition twice.) In post hoc review across the 5 weeks, no patients were transferred back to the PICU within 48 hours after being transferred out. During the study period, 19 patients were discharged directly from the PICU (8 to home, 7 by transport to another facility, and 4 to rehabilitation). One patient returned to the hospital (but not the PICU) within 48 hours of being discharged home from the PICU.

During the study period, using the highest census value for recorded for each 24‐hour period and the number of beds available that day, median hospital‐wide occupancy was 93% (interquartile range, 90%96%). During the 35 days of observation, 71% of the days had occupancy >90%, 29% of days had occupancy >95%, and 3% of days had occupancy >100%.

DISCUSSION

In this direct observation of a PICU, we found high usage of beds for delivery of CCS. We identified many episodes in which the half of the PICU we observed was fully occupied (200 of 824 hours), but not necessarily delivering PICU‐level care to all patients. In fact, 75% of the full‐capacity hours had at least 1 patient receiving NCCS and 37% had at least 2. Patients waiting for a floor bed assignment represented nearly 5% of bed‐hours observed (mean 9 hours per patient). That full occupancy was not random, but rather clustered on weekdays, is consistent with other work showing that hospitals are at greater risk for midweek crowding due to the way in which scheduled admissions enter and leave.1925

Our methods provide the basis for operational analysis and improvement to patient flow, such as value stream mapping.9, 26 Process improvement work could be directed to areas of delay uncovered through this analysis and inform clinical and nonclinical management. For example, one of the key problems faced by the PICU was finding floor bed assignments for patients leaving the unit. Simply building more beds in the PICU will not solve this problemand at an estimated cost of $2 million to add a bed, it is likely not an efficient means of responding to poor flow. In these cases, the problem seems to lie downstream, and could suggest shortage of regular floor beds or inefficient bed assignment procedures within the hospital. The output also suggests that variation in nonclinical processes should be a target for improvement, such as time to clean rooms, because variation is known to be a source of nonvalue‐added time in many operations.9, 26 High occupancy on weekdays but low occupancy on weekends also emphasizes the potential for smoothing occupancy to reduce the risk of midweek crowding and to better manage bed utilization and staffing.24, 25

When seeking to reduce nonvalue‐added time, one must weigh the risks of increased efficiency against clinical outcomes. For example, if patients could be transferred out of the PICU faster, would the risk of returns to the PICU be higher? In this study, 15 patients (7%) had a change in disposition from awaiting transfer back to a CCS. The fact that transfers did not happen instantaneously may serve as a safety check to reduce rapid returns, but it is not possible for us to evaluate the reasons why patients did not actually complete the pending transfers. Specifically, we cannot determine whether the patient's clinical status objectively deteriorated, the ICU team made a judgment call to hold the patient, or the floor team refused to accept the transfer. Given this fact, although it appears in this study (and in the health care system more broadly) that there are opportunities to increase efficiency and reduce nonvalue‐added time, it is not realistic (nor advisable) that such time be reduced to zero. Along this line, one must consider separately purely nonclinical functions such as room cleaning and those that include some clinical element, such as time waiting for a patient to be transferred.

Beyond the direct findings of this study, the method should be replicable in other settings and can reveal important information about health care efficiency, capacity, and flexibility. The bottlenecks identified would have been difficult to identify through administrative record review. The exact amount of time to spend on observation may vary from place to place and would depend on the expected variation over time and the level of detail sought. In general, the more common the event and the less variation, the less time needed to observe it.

This study has several limitations that should be considered in terms of interpreting the results and in seeking to reproduce the approach. First, hourly recordings may not be discrete enough for events that took less than 1 hour. To assess the degree to which this would affect our results, we reanalyzed all NCCS by subtracting 30 minutes (0.5 hour) from all recordings, which increased total CCS from 82% to 87% and decreased NCCS by the same 5 percentage points. In a related fashion, our recordings were truncated at the start and end of each 1‐week period, so we could only observe a maximum of 168 hours for any given activity and did not record how long an activity was happening before or after the recordings started or stopped, respectively. Second, each recording could only be for 1 activity per hour. Separate from the level of granularity already noted, this also limits interpretation of critical care activities that may have been simultaneous. However, because the goal of the study was not to describe the provision of critical care services, but rather the times when they were not being delivered, this does not influence our conclusions. For movement of patients, however, we missed instances of physician and nursing calling sign‐out on patients to receiving units, as these events last less than 1 hour (and in the case of surgical patients, generally do not occur as the team provides continuous coverage). The time for such events is then included in other activities. To the extent that this may influence the results, it would increase the perceived time for nonvalue‐added services, but to a limited degree, and never by more than 59 minutes. Third, the overnight hours (11:00 _PM to 8:00 _AM) were not directly observed, but retrospectively recorded each morning by reviewing the records and discussing the overnight events with the clinical staff. For example, if a patient was intubated at 11:00 _PM and at 8:00 _AM, the observer would confirm this and record that status for the intervening hours. This is unlikely to result in a substantial impact on the findings, because the overnight hours have a relative degree of stability even for unstable patients in terms of their status of needing or not needing a CCS. Fourth, we did not evaluate the appropriateness of CCS delivered (eg, how long a patient was ventilated). Our definitions for CCS and NCSS were based on Children's Hospital of Philadelphia practices, which may not be the same as those of other facilities. The categorization of CCS was objective for activities such as ventilation or continuous infusion, but was less clear for the not otherwise specified recordings, which represented patients with a complex illness or projected organ, respiratory, cardiac, or neurological failure. These patients were not receiving a specific critical care intervention, but were deemed to need to be in the PICU as opposed to a regular floor (eg, for frequent monitoring of potential respiratory failure). It would also include patients receiving combinations of therapies more efficiently delivered in the PICU. For that, the observers relied on the judgment of clinicians (primarily nurses) to determine whether the patient needed to be in the PICU or not; if no specific reason could be provided, not otherwise specified was applied. These 192 instances accounted for 2982 aggregate bed‐hours (15% of total). It is difficult to judge the direction of bias, because overestimation of need to be in the PICU may be as likely to occur as underestimation. Fifth, the very presence of the observers may have changed behavior. Knowing that they were being observed staff may have acted with greater efficiency than otherwise. We expect that such a finding would lead to less time appearing as necessary logistics or NCCS. Finally, results may not be generalizable to other hospitals or hospital settings. There are clearly important contextual factors, not only for the location but also for the duration. For example, staffing was never an issue during the 5 weeks of observation, but there are locations where an empty bed is not necessarily usable due to lack of staffing. Nonetheless, we believe the results provide a generalizable approach and methodology for other settings (and staffing could be a reason for an empty bed).

In terms of the setting, as noted, we observed one discrete 24‐bed unit, which comprises half of the total PICU. Thus, statements that the PICU was at full capacity must be interpreted in the context that additional rooms may have been available on the other side. Patients are generally admitted alternately to each unit, so the occupancies should parallel each other. We recorded the census every 4 hours for both sides from the electronic system (Sunrise Clinical Manager [SCM]). However, this only accounts for patients physically in beds, not beds held for patients in other locations. Thus, we would expect a discrepancy between direct observation and the SCM value. Through analysis of the entire pediatric intensive care unit,* that part which observed directly, and that which we did not observe directly using census data, we think it reasonable to assert that both units of the total PICU had constrained capacity during the times we directly observed and recorded such constraint on one side.

This study demonstrates the use of direct observation for inpatient settings to learn about resource utilization and identification of value‐added services. PubMed searches for the terms efficiency, flow, process redesign, and time management bring up many more references for operating rooms than for ICUs or inpatient beds. Some examples of ICU‐directed work include videography of an ICU in Australia27 and human factor analysis in ICU nursing.5 Time‐motion studies have also been conducted on clinical staff, such as physicians.28, 29

In conclusion, we found that direct observation provided important insights into the utilization of patient rooms in an important inpatient setting. Data such as these are valuable for clinical and process improvement work, as well as understanding how best to match capacity to patient need. Finally, the methodology is reproducible for other settings and would be an additional tool to measuring and improving the efficiency and value of the health system. When appropriate, this approach can also evaluate the effectiveness of process improvement, help identify and reduce waste,13 and contribute to the growing field that merges operations management with hospital administration and clinical care: in other words, evidence‐based management.30

Patient flow refers to the management and movement of patients in health care settings and is linked to quality, safety, and cost.16 The intensive care unit (ICU) is crucial in patient flow.7, 8 The limited number of beds and the resource‐intensive services and staffing associated with them require that hospitals optimize their utilization, as is increasingly true of all hospital resources. To maximize delivery of services to patients who need them and minimize real and opportunity losses (eg, postponed surgery, diverted transports, or inability to accept patients), patients in ICU beds should receive critical care medicine/nursing services while there and be transferred or discharged when appropriate.

The time between arrival and departure from any area of the hospital, including the ICU, is considered the time when a patient is receiving needed clinical carethe value‐added portion of health care operationsand time waiting to move on to the next step.911 This period includes both necessary logistics (eg, signing out a patient or waiting a reasonable amount of time for room cleaning) and nonvalue‐added time (eg, an excessively long amount of time for room cleaning). Operations management labels nonvalue‐added time as waste, and its reduction is vital for high‐quality health care.9, 12, 13 As in other industries, one important way to understand value versus waste is through direct observation.11, 14 Although operating rooms have been the subject of several published process improvement projects to improve efficiency,1518 inpatient beds have not been the subject of such scrutiny. The objectives of this study were to generate a direct observation method and use it to describe pediatric ICU (PICU) bed utilization from a value‐added perspective.

METHODS

An interdisciplinary work group of physicians, nurses, quality improvement specialists, and 1 operations management expert developed an Excel spreadsheet to categorize hour‐by‐hour status of PICU beds. The clinicians generated a list of 27 activities. A critical care nurse trained in quality improvement piloted the list for 3 separate 4‐hour blocks over 2 weeks adding 18 activities; 2 additional activities were added during the 5 weeks of observation (Table 1). (The recording tool is provided in the Supporting Information Appendix.) Three observers with knowledge of medical terminology (2 third‐year medical students and 1 premedical student with years of experience as an emergency medical technician) were trained over 12 hours to conduct the observations. Prior to the observations, the 3 observers also spent time in the PICU, and terminology used for recordings was reviewed. Interobserver reliability was checked during 3 sets of observation circuits by all 3 observers and the principal investigator, as well as by spot checks during the study.

The targeted area included 24 single‐patient rooms. The activity of each bed was recorded hourly. Real‐time recording in to the Excel spreadsheet on a dedicated laptop occurred from 8:00 _AM until 11:00 _PM. The most visible or critical event was recorded. Although some activities were not mutually exclusive (eg, a patient could be ventilated and on a continuous infusion simultaneously), the objective was to identify when a room was being used for any critical care service, not enumerate all of them. The observers noted overnight events that occurred from 11:00 _PM to 8:00 _AM in the morning by reviewing the bedside record and talking to the staff to complete each day's 24‐hour recording. The observers also recorded the hospital‐wide census and the census for the other half of the PICU every 4 hours. The observations occurred over 5 noncontiguous weeks between January 2009 and April 2009.

After all observations were complete, activities were classified as critical care services (CCS) or noncritical‐care services (NCCS). NCCSs were further divided into necessary logistics (defined for analysis purposes as the first hour of any NCCS activity) or nonvalue‐added (the second or greater hour of NCCS). A time limit of 1 hour was chosen to define necessary logistics based on a consensus that nonclinical activities optimally would not take more than 1 hour each. We also analyzed results with 2 hours as the cutoff for necessary logistics. Admission, discharge, and transfer records were reviewed to check for returns to the PICU or hospital within 48 hours of transfer or discharge from the PICU.

Analyses were conducted using Microsoft Excel (Microsoft, Redmond, WA) and Stata 10.0 (StataCorp, College Station, TX). The study was approved by the Children's Hospital of Philadelphia Institutional Review Board with waiver of consent.

RESULTS

A total of 824 hours of recordings included 19,887 bed‐hours with 219 unique patients; among them, 2 remained from the first day of recording in January to the last day in April (sample recording in Figure 1). A total of 50 patients (range, 812 per week) stayed for the entirety of each 1‐week observation period. Of the 47 possible activities, 45 of them were recorded for at least 1 hour in the 5 weeks. Overall, 14 activities accounted for 95% of the observed bed‐hours and 31 activities accounted for the remaining 5%. CCS accounted for 82% of observed bed‐hours, NCCS accounted for 10.4%, and empty unassigned accounted for 8% (Figure 2). Using the 1‐hour cutoff for necessary services, 77% of NCCS time was nonvalue‐added, whereas 23% of it was necessary logistics; using the 2‐hour cutoff, 54% was nonvalue‐added, and 46% was necessary logistics.

Figure 1

Figure 2

During the observation period, <1% of bed‐hours were used for CCS for overflow patients from the neonatal ICU (NICU), cardiac care unit (CCU), cardiac ICU (CICU), or progressive care unit (PCU; tracheostomy/ventilator unit). Although only 4 patients required transport to a rehabilitation facility, their wait time comprised 99 hours (<1%) of total recordings. Eight patients waited a mean of 2.6 hours for transportation home (maximum, 10 hours).

To demonstrate the cycle of room use, activities were divided into 4 categories: room preparation, critical care services, disposition pending, and postcritical care services (Figure 3). As an example of detailed data revealed by direct observation, we identified 102 instances totaling 919 hours when a patient was waiting for a bed assignment on another floor (5% of all bed‐hours). The mean wait time was 9 hours (range, 188 hours) and the median time was 5.5 hours. There were only 15 instances when floor bed assignment took 1 hour or less, and only 9 instances when it took 12 hours. Similarly, considerable time was spent on cleaning rooms between patients: only 66 of 146 instances of cleaning took 1 hour or less. The mean time for cleaning was 2.2 hours (range, 115), and the median was 2 hours. (There were 136 recorded instances of room cleaning and 10 additional episodes that were not recorded but had to be completed for the room to turnover from one patient to the next, yielding a total of 146 instances of cleaning.)

Figure 3

From the 824 hours of recording, we identified 200 hours (25% of time) when there were zero empty unassigned beds available in the section of the PICU being observed. Episodes of full occupancy occurred mostly on weekdays, with 23% of hours of full capacity on Thursdays, 21% on Mondays, and 21% on Wednesdays; only 8% were on Saturdays and <1% on Sundays. These 200 hours fell into 36 separate episodes of complete occupancy, each lasting 122 hours. Each patient, on average, received 3.1 hours of NCCS during each episode of full occupancy (range, 111 hours). Within these 200 hours at capacity, we identified only 15 hours (8%) when all 24 beds were used for CCS. For 72% of the time, there was at least 1 bed with NCCS, and for 37% at least 2 beds. A small portion of the time (7%), the lack of beds was affected by occupancy by patients who should have been in the NICU, CICU, CCU, or PCU.

Data collected through direct observation can be used to understand aggregated and averaged experiences, but also more specific time periods. For example, we identified 1 week with the highest consistent level of occupancy and turnover: March 915 had empty unassigned beds for only 4% of the week. Of the 168 hours in the week, 68 (40%) had full capacity. However, for 90% of the time, at least 1 bed was used for a NCCS. Other analytic options included varying the assumptions around time needed for logistics. Overall, NCCS time on necessary logistics changes from 23% to 46% using 1 hour versus 2 hours as the cutoff. For floor bed assignments, assuming that the first hour of this activity is necessary logistics and any hour thereafter is not, 817 hours were wasted. Even after assuming 2 hours of necessary logistical time (which may also include steps such as nursing and physician sign‐out to the receiving team, often not recorded in the observations), this left 715 hours of NCCS time in which patients waited to be placed elsewhere in the hospital. For room cleaning, because recordings were hourly, but room cleaning could take less time, we performed a sensitivity analysis, converting all 1‐hour recordings to half‐hour recordings to half‐hour recordings (an exaggerated shortening since industry‐standard cleaning may take longer).

Of the 219 patients directly observed, 15 were noted to be waiting for a transfer out of the PICU but experienced a change in disposition before the transfer. On average, these patients waited 8 hours for a floor bed assignment (range, 221) before reverting to a CCS, which then lasted an average of 16.5 hours (range, 149). (Included in this group are 2 patients who experienced this change in disposition twice.) In post hoc review across the 5 weeks, no patients were transferred back to the PICU within 48 hours after being transferred out. During the study period, 19 patients were discharged directly from the PICU (8 to home, 7 by transport to another facility, and 4 to rehabilitation). One patient returned to the hospital (but not the PICU) within 48 hours of being discharged home from the PICU.

During the study period, using the highest census value for recorded for each 24‐hour period and the number of beds available that day, median hospital‐wide occupancy was 93% (interquartile range, 90%96%). During the 35 days of observation, 71% of the days had occupancy >90%, 29% of days had occupancy >95%, and 3% of days had occupancy >100%.

DISCUSSION

In this direct observation of a PICU, we found high usage of beds for delivery of CCS. We identified many episodes in which the half of the PICU we observed was fully occupied (200 of 824 hours), but not necessarily delivering PICU‐level care to all patients. In fact, 75% of the full‐capacity hours had at least 1 patient receiving NCCS and 37% had at least 2. Patients waiting for a floor bed assignment represented nearly 5% of bed‐hours observed (mean 9 hours per patient). That full occupancy was not random, but rather clustered on weekdays, is consistent with other work showing that hospitals are at greater risk for midweek crowding due to the way in which scheduled admissions enter and leave.1925

Our methods provide the basis for operational analysis and improvement to patient flow, such as value stream mapping.9, 26 Process improvement work could be directed to areas of delay uncovered through this analysis and inform clinical and nonclinical management. For example, one of the key problems faced by the PICU was finding floor bed assignments for patients leaving the unit. Simply building more beds in the PICU will not solve this problemand at an estimated cost of $2 million to add a bed, it is likely not an efficient means of responding to poor flow. In these cases, the problem seems to lie downstream, and could suggest shortage of regular floor beds or inefficient bed assignment procedures within the hospital. The output also suggests that variation in nonclinical processes should be a target for improvement, such as time to clean rooms, because variation is known to be a source of nonvalue‐added time in many operations.9, 26 High occupancy on weekdays but low occupancy on weekends also emphasizes the potential for smoothing occupancy to reduce the risk of midweek crowding and to better manage bed utilization and staffing.24, 25

When seeking to reduce nonvalue‐added time, one must weigh the risks of increased efficiency against clinical outcomes. For example, if patients could be transferred out of the PICU faster, would the risk of returns to the PICU be higher? In this study, 15 patients (7%) had a change in disposition from awaiting transfer back to a CCS. The fact that transfers did not happen instantaneously may serve as a safety check to reduce rapid returns, but it is not possible for us to evaluate the reasons why patients did not actually complete the pending transfers. Specifically, we cannot determine whether the patient's clinical status objectively deteriorated, the ICU team made a judgment call to hold the patient, or the floor team refused to accept the transfer. Given this fact, although it appears in this study (and in the health care system more broadly) that there are opportunities to increase efficiency and reduce nonvalue‐added time, it is not realistic (nor advisable) that such time be reduced to zero. Along this line, one must consider separately purely nonclinical functions such as room cleaning and those that include some clinical element, such as time waiting for a patient to be transferred.

Beyond the direct findings of this study, the method should be replicable in other settings and can reveal important information about health care efficiency, capacity, and flexibility. The bottlenecks identified would have been difficult to identify through administrative record review. The exact amount of time to spend on observation may vary from place to place and would depend on the expected variation over time and the level of detail sought. In general, the more common the event and the less variation, the less time needed to observe it.

This study has several limitations that should be considered in terms of interpreting the results and in seeking to reproduce the approach. First, hourly recordings may not be discrete enough for events that took less than 1 hour. To assess the degree to which this would affect our results, we reanalyzed all NCCS by subtracting 30 minutes (0.5 hour) from all recordings, which increased total CCS from 82% to 87% and decreased NCCS by the same 5 percentage points. In a related fashion, our recordings were truncated at the start and end of each 1‐week period, so we could only observe a maximum of 168 hours for any given activity and did not record how long an activity was happening before or after the recordings started or stopped, respectively. Second, each recording could only be for 1 activity per hour. Separate from the level of granularity already noted, this also limits interpretation of critical care activities that may have been simultaneous. However, because the goal of the study was not to describe the provision of critical care services, but rather the times when they were not being delivered, this does not influence our conclusions. For movement of patients, however, we missed instances of physician and nursing calling sign‐out on patients to receiving units, as these events last less than 1 hour (and in the case of surgical patients, generally do not occur as the team provides continuous coverage). The time for such events is then included in other activities. To the extent that this may influence the results, it would increase the perceived time for nonvalue‐added services, but to a limited degree, and never by more than 59 minutes. Third, the overnight hours (11:00 _PM to 8:00 _AM) were not directly observed, but retrospectively recorded each morning by reviewing the records and discussing the overnight events with the clinical staff. For example, if a patient was intubated at 11:00 _PM and at 8:00 _AM, the observer would confirm this and record that status for the intervening hours. This is unlikely to result in a substantial impact on the findings, because the overnight hours have a relative degree of stability even for unstable patients in terms of their status of needing or not needing a CCS. Fourth, we did not evaluate the appropriateness of CCS delivered (eg, how long a patient was ventilated). Our definitions for CCS and NCSS were based on Children's Hospital of Philadelphia practices, which may not be the same as those of other facilities. The categorization of CCS was objective for activities such as ventilation or continuous infusion, but was less clear for the not otherwise specified recordings, which represented patients with a complex illness or projected organ, respiratory, cardiac, or neurological failure. These patients were not receiving a specific critical care intervention, but were deemed to need to be in the PICU as opposed to a regular floor (eg, for frequent monitoring of potential respiratory failure). It would also include patients receiving combinations of therapies more efficiently delivered in the PICU. For that, the observers relied on the judgment of clinicians (primarily nurses) to determine whether the patient needed to be in the PICU or not; if no specific reason could be provided, not otherwise specified was applied. These 192 instances accounted for 2982 aggregate bed‐hours (15% of total). It is difficult to judge the direction of bias, because overestimation of need to be in the PICU may be as likely to occur as underestimation. Fifth, the very presence of the observers may have changed behavior. Knowing that they were being observed staff may have acted with greater efficiency than otherwise. We expect that such a finding would lead to less time appearing as necessary logistics or NCCS. Finally, results may not be generalizable to other hospitals or hospital settings. There are clearly important contextual factors, not only for the location but also for the duration. For example, staffing was never an issue during the 5 weeks of observation, but there are locations where an empty bed is not necessarily usable due to lack of staffing. Nonetheless, we believe the results provide a generalizable approach and methodology for other settings (and staffing could be a reason for an empty bed).

In terms of the setting, as noted, we observed one discrete 24‐bed unit, which comprises half of the total PICU. Thus, statements that the PICU was at full capacity must be interpreted in the context that additional rooms may have been available on the other side. Patients are generally admitted alternately to each unit, so the occupancies should parallel each other. We recorded the census every 4 hours for both sides from the electronic system (Sunrise Clinical Manager [SCM]). However, this only accounts for patients physically in beds, not beds held for patients in other locations. Thus, we would expect a discrepancy between direct observation and the SCM value. Through analysis of the entire pediatric intensive care unit,* that part which observed directly, and that which we did not observe directly using census data, we think it reasonable to assert that both units of the total PICU had constrained capacity during the times we directly observed and recorded such constraint on one side.

This study demonstrates the use of direct observation for inpatient settings to learn about resource utilization and identification of value‐added services. PubMed searches for the terms efficiency, flow, process redesign, and time management bring up many more references for operating rooms than for ICUs or inpatient beds. Some examples of ICU‐directed work include videography of an ICU in Australia27 and human factor analysis in ICU nursing.5 Time‐motion studies have also been conducted on clinical staff, such as physicians.28, 29

In conclusion, we found that direct observation provided important insights into the utilization of patient rooms in an important inpatient setting. Data such as these are valuable for clinical and process improvement work, as well as understanding how best to match capacity to patient need. Finally, the methodology is reproducible for other settings and would be an additional tool to measuring and improving the efficiency and value of the health system. When appropriate, this approach can also evaluate the effectiveness of process improvement, help identify and reduce waste,13 and contribute to the growing field that merges operations management with hospital administration and clinical care: in other words, evidence‐based management.30

Patient flow refers to the management and movement of patients in health care settings and is linked to quality, safety, and cost.16 The intensive care unit (ICU) is crucial in patient flow.7, 8 The limited number of beds and the resource‐intensive services and staffing associated with them require that hospitals optimize their utilization, as is increasingly true of all hospital resources. To maximize delivery of services to patients who need them and minimize real and opportunity losses (eg, postponed surgery, diverted transports, or inability to accept patients), patients in ICU beds should receive critical care medicine/nursing services while there and be transferred or discharged when appropriate.

The time between arrival and departure from any area of the hospital, including the ICU, is considered the time when a patient is receiving needed clinical carethe value‐added portion of health care operationsand time waiting to move on to the next step.911 This period includes both necessary logistics (eg, signing out a patient or waiting a reasonable amount of time for room cleaning) and nonvalue‐added time (eg, an excessively long amount of time for room cleaning). Operations management labels nonvalue‐added time as waste, and its reduction is vital for high‐quality health care.9, 12, 13 As in other industries, one important way to understand value versus waste is through direct observation.11, 14 Although operating rooms have been the subject of several published process improvement projects to improve efficiency,1518 inpatient beds have not been the subject of such scrutiny. The objectives of this study were to generate a direct observation method and use it to describe pediatric ICU (PICU) bed utilization from a value‐added perspective.

METHODS

An interdisciplinary work group of physicians, nurses, quality improvement specialists, and 1 operations management expert developed an Excel spreadsheet to categorize hour‐by‐hour status of PICU beds. The clinicians generated a list of 27 activities. A critical care nurse trained in quality improvement piloted the list for 3 separate 4‐hour blocks over 2 weeks adding 18 activities; 2 additional activities were added during the 5 weeks of observation (Table 1). (The recording tool is provided in the Supporting Information Appendix.) Three observers with knowledge of medical terminology (2 third‐year medical students and 1 premedical student with years of experience as an emergency medical technician) were trained over 12 hours to conduct the observations. Prior to the observations, the 3 observers also spent time in the PICU, and terminology used for recordings was reviewed. Interobserver reliability was checked during 3 sets of observation circuits by all 3 observers and the principal investigator, as well as by spot checks during the study.

The targeted area included 24 single‐patient rooms. The activity of each bed was recorded hourly. Real‐time recording in to the Excel spreadsheet on a dedicated laptop occurred from 8:00 _AM until 11:00 _PM. The most visible or critical event was recorded. Although some activities were not mutually exclusive (eg, a patient could be ventilated and on a continuous infusion simultaneously), the objective was to identify when a room was being used for any critical care service, not enumerate all of them. The observers noted overnight events that occurred from 11:00 _PM to 8:00 _AM in the morning by reviewing the bedside record and talking to the staff to complete each day's 24‐hour recording. The observers also recorded the hospital‐wide census and the census for the other half of the PICU every 4 hours. The observations occurred over 5 noncontiguous weeks between January 2009 and April 2009.

After all observations were complete, activities were classified as critical care services (CCS) or noncritical‐care services (NCCS). NCCSs were further divided into necessary logistics (defined for analysis purposes as the first hour of any NCCS activity) or nonvalue‐added (the second or greater hour of NCCS). A time limit of 1 hour was chosen to define necessary logistics based on a consensus that nonclinical activities optimally would not take more than 1 hour each. We also analyzed results with 2 hours as the cutoff for necessary logistics. Admission, discharge, and transfer records were reviewed to check for returns to the PICU or hospital within 48 hours of transfer or discharge from the PICU.

Analyses were conducted using Microsoft Excel (Microsoft, Redmond, WA) and Stata 10.0 (StataCorp, College Station, TX). The study was approved by the Children's Hospital of Philadelphia Institutional Review Board with waiver of consent.

RESULTS

A total of 824 hours of recordings included 19,887 bed‐hours with 219 unique patients; among them, 2 remained from the first day of recording in January to the last day in April (sample recording in Figure 1). A total of 50 patients (range, 812 per week) stayed for the entirety of each 1‐week observation period. Of the 47 possible activities, 45 of them were recorded for at least 1 hour in the 5 weeks. Overall, 14 activities accounted for 95% of the observed bed‐hours and 31 activities accounted for the remaining 5%. CCS accounted for 82% of observed bed‐hours, NCCS accounted for 10.4%, and empty unassigned accounted for 8% (Figure 2). Using the 1‐hour cutoff for necessary services, 77% of NCCS time was nonvalue‐added, whereas 23% of it was necessary logistics; using the 2‐hour cutoff, 54% was nonvalue‐added, and 46% was necessary logistics.

Figure 1

Figure 2

During the observation period, <1% of bed‐hours were used for CCS for overflow patients from the neonatal ICU (NICU), cardiac care unit (CCU), cardiac ICU (CICU), or progressive care unit (PCU; tracheostomy/ventilator unit). Although only 4 patients required transport to a rehabilitation facility, their wait time comprised 99 hours (<1%) of total recordings. Eight patients waited a mean of 2.6 hours for transportation home (maximum, 10 hours).

To demonstrate the cycle of room use, activities were divided into 4 categories: room preparation, critical care services, disposition pending, and postcritical care services (Figure 3). As an example of detailed data revealed by direct observation, we identified 102 instances totaling 919 hours when a patient was waiting for a bed assignment on another floor (5% of all bed‐hours). The mean wait time was 9 hours (range, 188 hours) and the median time was 5.5 hours. There were only 15 instances when floor bed assignment took 1 hour or less, and only 9 instances when it took 12 hours. Similarly, considerable time was spent on cleaning rooms between patients: only 66 of 146 instances of cleaning took 1 hour or less. The mean time for cleaning was 2.2 hours (range, 115), and the median was 2 hours. (There were 136 recorded instances of room cleaning and 10 additional episodes that were not recorded but had to be completed for the room to turnover from one patient to the next, yielding a total of 146 instances of cleaning.)

Figure 3

From the 824 hours of recording, we identified 200 hours (25% of time) when there were zero empty unassigned beds available in the section of the PICU being observed. Episodes of full occupancy occurred mostly on weekdays, with 23% of hours of full capacity on Thursdays, 21% on Mondays, and 21% on Wednesdays; only 8% were on Saturdays and <1% on Sundays. These 200 hours fell into 36 separate episodes of complete occupancy, each lasting 122 hours. Each patient, on average, received 3.1 hours of NCCS during each episode of full occupancy (range, 111 hours). Within these 200 hours at capacity, we identified only 15 hours (8%) when all 24 beds were used for CCS. For 72% of the time, there was at least 1 bed with NCCS, and for 37% at least 2 beds. A small portion of the time (7%), the lack of beds was affected by occupancy by patients who should have been in the NICU, CICU, CCU, or PCU.

Data collected through direct observation can be used to understand aggregated and averaged experiences, but also more specific time periods. For example, we identified 1 week with the highest consistent level of occupancy and turnover: March 915 had empty unassigned beds for only 4% of the week. Of the 168 hours in the week, 68 (40%) had full capacity. However, for 90% of the time, at least 1 bed was used for a NCCS. Other analytic options included varying the assumptions around time needed for logistics. Overall, NCCS time on necessary logistics changes from 23% to 46% using 1 hour versus 2 hours as the cutoff. For floor bed assignments, assuming that the first hour of this activity is necessary logistics and any hour thereafter is not, 817 hours were wasted. Even after assuming 2 hours of necessary logistical time (which may also include steps such as nursing and physician sign‐out to the receiving team, often not recorded in the observations), this left 715 hours of NCCS time in which patients waited to be placed elsewhere in the hospital. For room cleaning, because recordings were hourly, but room cleaning could take less time, we performed a sensitivity analysis, converting all 1‐hour recordings to half‐hour recordings to half‐hour recordings (an exaggerated shortening since industry‐standard cleaning may take longer).

Of the 219 patients directly observed, 15 were noted to be waiting for a transfer out of the PICU but experienced a change in disposition before the transfer. On average, these patients waited 8 hours for a floor bed assignment (range, 221) before reverting to a CCS, which then lasted an average of 16.5 hours (range, 149). (Included in this group are 2 patients who experienced this change in disposition twice.) In post hoc review across the 5 weeks, no patients were transferred back to the PICU within 48 hours after being transferred out. During the study period, 19 patients were discharged directly from the PICU (8 to home, 7 by transport to another facility, and 4 to rehabilitation). One patient returned to the hospital (but not the PICU) within 48 hours of being discharged home from the PICU.

During the study period, using the highest census value for recorded for each 24‐hour period and the number of beds available that day, median hospital‐wide occupancy was 93% (interquartile range, 90%96%). During the 35 days of observation, 71% of the days had occupancy >90%, 29% of days had occupancy >95%, and 3% of days had occupancy >100%.

DISCUSSION

In this direct observation of a PICU, we found high usage of beds for delivery of CCS. We identified many episodes in which the half of the PICU we observed was fully occupied (200 of 824 hours), but not necessarily delivering PICU‐level care to all patients. In fact, 75% of the full‐capacity hours had at least 1 patient receiving NCCS and 37% had at least 2. Patients waiting for a floor bed assignment represented nearly 5% of bed‐hours observed (mean 9 hours per patient). That full occupancy was not random, but rather clustered on weekdays, is consistent with other work showing that hospitals are at greater risk for midweek crowding due to the way in which scheduled admissions enter and leave.1925

Our methods provide the basis for operational analysis and improvement to patient flow, such as value stream mapping.9, 26 Process improvement work could be directed to areas of delay uncovered through this analysis and inform clinical and nonclinical management. For example, one of the key problems faced by the PICU was finding floor bed assignments for patients leaving the unit. Simply building more beds in the PICU will not solve this problemand at an estimated cost of $2 million to add a bed, it is likely not an efficient means of responding to poor flow. In these cases, the problem seems to lie downstream, and could suggest shortage of regular floor beds or inefficient bed assignment procedures within the hospital. The output also suggests that variation in nonclinical processes should be a target for improvement, such as time to clean rooms, because variation is known to be a source of nonvalue‐added time in many operations.9, 26 High occupancy on weekdays but low occupancy on weekends also emphasizes the potential for smoothing occupancy to reduce the risk of midweek crowding and to better manage bed utilization and staffing.24, 25

When seeking to reduce nonvalue‐added time, one must weigh the risks of increased efficiency against clinical outcomes. For example, if patients could be transferred out of the PICU faster, would the risk of returns to the PICU be higher? In this study, 15 patients (7%) had a change in disposition from awaiting transfer back to a CCS. The fact that transfers did not happen instantaneously may serve as a safety check to reduce rapid returns, but it is not possible for us to evaluate the reasons why patients did not actually complete the pending transfers. Specifically, we cannot determine whether the patient's clinical status objectively deteriorated, the ICU team made a judgment call to hold the patient, or the floor team refused to accept the transfer. Given this fact, although it appears in this study (and in the health care system more broadly) that there are opportunities to increase efficiency and reduce nonvalue‐added time, it is not realistic (nor advisable) that such time be reduced to zero. Along this line, one must consider separately purely nonclinical functions such as room cleaning and those that include some clinical element, such as time waiting for a patient to be transferred.

Beyond the direct findings of this study, the method should be replicable in other settings and can reveal important information about health care efficiency, capacity, and flexibility. The bottlenecks identified would have been difficult to identify through administrative record review. The exact amount of time to spend on observation may vary from place to place and would depend on the expected variation over time and the level of detail sought. In general, the more common the event and the less variation, the less time needed to observe it.

This study has several limitations that should be considered in terms of interpreting the results and in seeking to reproduce the approach. First, hourly recordings may not be discrete enough for events that took less than 1 hour. To assess the degree to which this would affect our results, we reanalyzed all NCCS by subtracting 30 minutes (0.5 hour) from all recordings, which increased total CCS from 82% to 87% and decreased NCCS by the same 5 percentage points. In a related fashion, our recordings were truncated at the start and end of each 1‐week period, so we could only observe a maximum of 168 hours for any given activity and did not record how long an activity was happening before or after the recordings started or stopped, respectively. Second, each recording could only be for 1 activity per hour. Separate from the level of granularity already noted, this also limits interpretation of critical care activities that may have been simultaneous. However, because the goal of the study was not to describe the provision of critical care services, but rather the times when they were not being delivered, this does not influence our conclusions. For movement of patients, however, we missed instances of physician and nursing calling sign‐out on patients to receiving units, as these events last less than 1 hour (and in the case of surgical patients, generally do not occur as the team provides continuous coverage). The time for such events is then included in other activities. To the extent that this may influence the results, it would increase the perceived time for nonvalue‐added services, but to a limited degree, and never by more than 59 minutes. Third, the overnight hours (11:00 _PM to 8:00 _AM) were not directly observed, but retrospectively recorded each morning by reviewing the records and discussing the overnight events with the clinical staff. For example, if a patient was intubated at 11:00 _PM and at 8:00 _AM, the observer would confirm this and record that status for the intervening hours. This is unlikely to result in a substantial impact on the findings, because the overnight hours have a relative degree of stability even for unstable patients in terms of their status of needing or not needing a CCS. Fourth, we did not evaluate the appropriateness of CCS delivered (eg, how long a patient was ventilated). Our definitions for CCS and NCSS were based on Children's Hospital of Philadelphia practices, which may not be the same as those of other facilities. The categorization of CCS was objective for activities such as ventilation or continuous infusion, but was less clear for the not otherwise specified recordings, which represented patients with a complex illness or projected organ, respiratory, cardiac, or neurological failure. These patients were not receiving a specific critical care intervention, but were deemed to need to be in the PICU as opposed to a regular floor (eg, for frequent monitoring of potential respiratory failure). It would also include patients receiving combinations of therapies more efficiently delivered in the PICU. For that, the observers relied on the judgment of clinicians (primarily nurses) to determine whether the patient needed to be in the PICU or not; if no specific reason could be provided, not otherwise specified was applied. These 192 instances accounted for 2982 aggregate bed‐hours (15% of total). It is difficult to judge the direction of bias, because overestimation of need to be in the PICU may be as likely to occur as underestimation. Fifth, the very presence of the observers may have changed behavior. Knowing that they were being observed staff may have acted with greater efficiency than otherwise. We expect that such a finding would lead to less time appearing as necessary logistics or NCCS. Finally, results may not be generalizable to other hospitals or hospital settings. There are clearly important contextual factors, not only for the location but also for the duration. For example, staffing was never an issue during the 5 weeks of observation, but there are locations where an empty bed is not necessarily usable due to lack of staffing. Nonetheless, we believe the results provide a generalizable approach and methodology for other settings (and staffing could be a reason for an empty bed).

In terms of the setting, as noted, we observed one discrete 24‐bed unit, which comprises half of the total PICU. Thus, statements that the PICU was at full capacity must be interpreted in the context that additional rooms may have been available on the other side. Patients are generally admitted alternately to each unit, so the occupancies should parallel each other. We recorded the census every 4 hours for both sides from the electronic system (Sunrise Clinical Manager [SCM]). However, this only accounts for patients physically in beds, not beds held for patients in other locations. Thus, we would expect a discrepancy between direct observation and the SCM value. Through analysis of the entire pediatric intensive care unit,* that part which observed directly, and that which we did not observe directly using census data, we think it reasonable to assert that both units of the total PICU had constrained capacity during the times we directly observed and recorded such constraint on one side.

This study demonstrates the use of direct observation for inpatient settings to learn about resource utilization and identification of value‐added services. PubMed searches for the terms efficiency, flow, process redesign, and time management bring up many more references for operating rooms than for ICUs or inpatient beds. Some examples of ICU‐directed work include videography of an ICU in Australia27 and human factor analysis in ICU nursing.5 Time‐motion studies have also been conducted on clinical staff, such as physicians.28, 29

In conclusion, we found that direct observation provided important insights into the utilization of patient rooms in an important inpatient setting. Data such as these are valuable for clinical and process improvement work, as well as understanding how best to match capacity to patient need. Finally, the methodology is reproducible for other settings and would be an additional tool to measuring and improving the efficiency and value of the health system. When appropriate, this approach can also evaluate the effectiveness of process improvement, help identify and reduce waste,13 and contribute to the growing field that merges operations management with hospital administration and clinical care: in other words, evidence‐based management.30

Disparities in patterns of care and outcomes for ambulatory‐care sensitive conditions remain a persistent problem for children.19 Many studies have focused on disparities in hospitalization rates and length of stay (LOS) related to asthma, however, few studies have focused on community‐acquired pneumonia (CAP) despite the fact that pneumonia is the most common, preventable, and potentially serious infection in childhood.10 Providers, payers, and families have a common interest in minimizing hospital LOS for different reasons (eg, minimizing costs, lost wages, exposure to antibiotic‐resistant bacteria), however, this interest is balanced against the potentially greater risk of readmission and adverse outcomes if LOS is inappropriately short. To date, the relationship between insurance status and LOS for CAP remains unexplored.

As in other conditions, substantial variation exists with respect to patterns of care and outcomes for children hospitalized with CAP.11 For example, children hospitalized in rural settings have a shorter LOS for pneumonia than those hospitalized in large urban settings.12 Children from racial/ethnic minorities tend to have higher rates of CAP‐associated complications, including death.11 Decades of prior studies have documented that uninsured children are less likely than insured children to make preventive care visits and obtain prescription medications, but differences in LOS or hospitalization rates between insured and uninsured children with CAP have not been studied.6, 8, 13, 14 Though imperfect, insurance status is 1 proxy for healthcare access, and current healthcare reform efforts aim to improve healthcare access and decrease socioeconomic gradients in health by increasing the number of insured American children. Nonetheless, quantifying the relationship between insurance status on LOS for children hospitalized with CAP is a first step towards understanding the influence of ambulatory care access on hospitalization for ambulatory‐care sensitive conditions.

The purpose of this study was to investigate the influence of insurance status and type on LOS for children hospitalized with CAP. In addition, we sought to determine if there were consistent trends over time in the association between insurance status and type with LOS for children hospitalized with CAP.

METHODS

RESULTS

The more than 10.5 million children sampled (unweighted) in KID during these 4 time periods (1997, 2000, 2003, and 2006) are representative of the more than 28.9 million children hospitalized in the United States. In each of these sample years, there were approximately 150,000 children hospitalized with pneumonia across the United States (Table 1). Of those hospitalized, 23% to 28% had a concomitant diagnosis of asthma; 6% to 8% had a pneumonia‐associated complication; and mortality was <0.01% in each sample year for patients hospitalized with pneumonia. In all years, among those with racial/ethnic data, the sample population was predominantly white boys less than 6 years old. The greatest proportion of children were hospitalized in urban non‐teaching settings, and also those children living in the southern regions of the United States.

There was little variation in the insurance status of children hospitalized with CAP between 1997 and 2006. In each of the sampled years, at least 40% of sampled children were privately insured, at least 40% were publicly insured, and approximately 5% were uninsured (Table 1). In all years, there were significant racial/ethnic disparities in insurance coverage such that whites were 4 to 6 times more likely to have private insurance than blacks, however, the large amount of missing race/ethnicity data warrant caution in interpreting this finding (Table 2; also see Supporting Information Appendix B in the online version of this article). We also found that children less than 1 year old were the most likely to be publicly insured in all years (see Supporting Appendix C in the online version of this article). There were also regional differences related to insurance coverage such that a greater proportion of children hospitalized in facilities located in the southern part of the United States were publicly insured. Notably, there were no significant differences in CAP‐associated mortality or asthma related to insurance coverage (Table 2). In 2006, CAP‐associated complications occurred in 8.5% of children with private insurance, 6.5% of children with public insurance, and 7.7% of uninsured children; the relative distribution of complications by insurance type were similar in previous years of the KID survey.

After examining the general and demographic characteristics, we then examined mean LOS for all children with CAP in each sample year (Table 3). The mean LOS for children with CAP was 3.44 days in 1997, with marginal decreases in subsequent years to a mean LOS of 3.18 days in 2006. The distribution of LOS for children with CAP revealed that nearly 70% of children were hospitalized for fewer than 3 days, another 22% to 28% were hospitalized for less than 1 week, and only 3% were hospitalized for more than 1 week. This distribution did not change substantially between 1997 and 2006. Next, we compared mean LOS by insurance type and race/ethnicity in unadjusted analyses. In each sample year, publicly insured children hospitalized with CAP had significantly longer LOS than privately insured children (P < 0.001). Similarly, in all years excepting 1997, uninsured children hospitalized with CAP had significantly shorter LOS than privately insured children. There were also significant racial differences in LOS for children with CAP, such that black children had longer LOS than white children with CAP. However, the large amount of missing data for race/ethnicity limited the robustness of this finding, and subsequent sensitivity analyses demonstrated that there were no consistent racial/ethnic disparities in LOS (see Supporting Appendix B in the online version of this article). These sensitivity analyses for missing race data did not alter our primary finding of shorter LOS for uninsured versus publicly or privately insured children.

After controlling for child age, race/ethnicity, gender, hospital type, transfer status, and presence of asthma or pneumonia‐associated complications, our multivariable analyses examining the relationship between insurance coverage and hospital LOS yielded the following results (Table 4). First, publicly insured children had significantly longer hospital stays than privately insured children, and uninsured children had significantly shorter hospital stays than privately insured children in all years except 1997. Second, children admitted with CAP at urban teaching children's hospitals had significantly longer LOS than those admitted to urban non‐teaching hospitals, and, in 2003, children admitted with CAP to rural hospitals had significantly shorter LOS than those admitted to urban non‐teaching hospitals. Third, children older than 1 year consistently had shorter hospital stays than infants less than 1 year old. Finally, though concomitant diagnosis of asthma did not consistently influence LOS, children who developed any complications had significantly longer LOS than those who did not. The cumulative impact of seemingly small differences in LOS is great. For example, in 2006, our model suggests that, for every 1000 children hospitalized with CAP in a given year, after adjusting for differences in sex, age, race, hospital‐type, region, transfer status, and diagnosis of asthma or complications, publicly insured children spend 90 to 130 more days in the hospital than privately insured children, whereas uninsured children spend between 40 to 90 fewer days in the hospital than privately insured children.

DISCUSSION

In this nationally representative sample selected over the past 10 years, we found that publicly insured children hospitalized with CAP have significantly longer LOS than those who are privately insured, and that, since 2000, uninsured children hospitalized with CAP have significantly shorter LOS than those who are privately insured. Though these observed differences are small, they are consistent across all 4 sampled years and, because CAP is one of the most common pediatric inpatient diagnoses, the cumulative impact of the observed differences on hospital LOS is great. Insurance status is often considered a proxy for access to preventive and ambulatory healthcare services or socioeconomic status. However, the underlying mechanisms relating insurance status to healthcare access, utilization, and ultimately, health outcomes are highly complex and difficult to elucidate.17 The observed variation in this study raises questions about the potential influence of insurance status on hospital discharge practices. Additional research is necessary to understand whether there are differences in processes of care (eg, performance of blood cultures or chest radiographs), quality of care, or other outcomes, such as readmissions, related to CAP inpatient management for children with different insurance coverage.

Apart from differences in hospital discharge practices, another possible explanation for uninsured children with CAP having shorter LOS is that these children have less severe disease than privately insured. This may occur if uninsured children with CAP are evaluated in the emergency department rather than the office setting, because emergency department providers may be more likely to admit children with CAP who lack a consistent access to ambulatory primary care services. Countering this alternative, prior studies have shown that uninsured groups are more likely to have greater disease severity than privately insured groups at the time of hospital admission.18, 19 In this study, we attempted to identify children with greater severity of disease using ICD‐9 codes for CAP‐associated complications. Though this is a relatively crude method that might lead to an underestimate of the total number of children with complications, we found that there were no significant differences in the prevalence of CAP‐associated complications between uninsured and insured groups in all sampled years.

On the other hand, uninsured patients may be released earlier by providers in order to reduce the amount of uncompensated care provided, or possibly because parents may urge providers to discharge their children, given their inability to pay forthcoming hospital bills and/or avoid further lost wages due to work absence.20, 21 In California, Bindman et al. demonstrated that decreasing the frequency of Medicaid recertification, and consequently increasing the likelihood of continuous insurance coverage, was associated with a decreased risk of hospitalization for ambulatory‐care sensitive conditions.5

We also found that children admitted to urban teaching children's hospitals with CAP had significantly longer LOS than those admitted to urban non‐teaching hospitals, whereas children in rural hospitals had significantly shorter LOS than those in urban non‐teaching hospitals in 2003. These findings are consistent with prior data from 1996 to1998 demonstrating that children admitted to rural hospitals in New York and Pennsylvania had significantly shorter LOS than large urban hospitals for 19 medical and 9 surgical conditions, including pneumonia.12 These findings may reflect underlying differences in between rural and urban hospital transfer practices, whereby rural hospitals may be more likely than urban hospitals to transfer children with relatively more severe illness to urban referral centers and retain children with less severe illness, leading to shorter LOS.12 Though our empiric understanding of differences in LOS between teaching and non‐teaching hospitals is currently limited, clinical experience supports the notion that there may be decreases in efficiency that occur in teaching hospitals, and are a result of the supervision required for care provided by trainees. It is also possible that, despite our exclusion of comorbid conditions, some children with complex or chronic medical conditions were included in this study. These children are often cared for at teaching hospitals, regardless of the primary cause for admission, and are more likely to have public insurance than other children, thus confounding the relationship between hospital type, insurance type and status, and LOS for children with CAP. The limitations of this dataset preclude further examination of this issue.

There are some limitations to this study. First, the KID data are cross‐sectional and causal inferences are limited. However, our results demonstrating that uninsured children hospitalized with CAP had shorter LOS than privately insured children were quite consistent in each sample year, suggesting that our results are a true association. Additionally, insurance status in KID is typically collected at admission, however, it is not possible to determine whether specific changes to insurance status that occurred during the hospitalization were applied to the data. The impact of this limitation would depend on the type of insurance obtained by the patient. If uninsured patients obtained public insurance, our study would underestimate the increased LOS for publicly insured patients, compared with privately insured patients, but have no effect on the difference in LOS between uninsured and privately insured patients. In the unlikely event that uninsured patients obtained private insurance, then our study would underestimate the difference for uninsured patients, compared with privately insured patients, biasing our current study results towards the null. Second, a substantial proportion of sampled children had missing data for race/ethnicity. To assess the impact of the missing race/ethnicity data on our results, we conducted sensitivity analyses and found that, though difficult to make any definitive conclusions about the relationship between race/ethnicity and LOS for children with CAP, there were no changes to our primary findings regarding differences in LOS between children with different insurance status and type. Third, KID does not include data about other unmeasured confounders (eg, parent income, parent education, regular source of care) that might be related to LOS, as well as a broad spectrum of pediatric outcomes. Serious consideration of expanding KID to include these variables is warranted. Fourth, the other category of insurance is not uniformly coded across states in the KID database. While some states use this category to classify public insurance options other than Medicare and Medicaid, other states include private insurance options in this group. Thus, it is possible that some patients with public insurance are misclassified as having other insurance. We would expect such misclassification to bias our findings towards the null hypothesis. Finally, we focused on the relationship between child health insurance status and CAP, only 1 ambulatory care‐sensitive condition. Additional research examining the relationship between insurance type and other ambulatory care‐sensitive conditions is warranted.

In summary, we found that, after multivariable adjustment, uninsured children hospitalized with community‐acquired pneumonia had significantly shorter LOS than privately insured children, and publicly insured children had a significantly longer hospital stay than privately insured children in these 4 nationally representative samples from 1997 to 2006. Current federal and state efforts to increase enrollment of children into insurance programs are a first step in reducing healthcare disparities. However, insurance coverage alone does not guarantee access to healthcare, thus, these efforts in isolation will likely be insufficient to achieve optimal health for the children of our country. As healthcare reform legislation is implemented, these findings provide hospitals and policy makers additional impetus to develop ways to achieve the ideal length of stay for every child; this ideal state will be achieved when clinical status and course, rather than nonclinical factors such as insurance type or provider's unease with ambulatory follow‐up, determine the duration of hospitalization for every child.

Disparities in patterns of care and outcomes for ambulatory‐care sensitive conditions remain a persistent problem for children.19 Many studies have focused on disparities in hospitalization rates and length of stay (LOS) related to asthma, however, few studies have focused on community‐acquired pneumonia (CAP) despite the fact that pneumonia is the most common, preventable, and potentially serious infection in childhood.10 Providers, payers, and families have a common interest in minimizing hospital LOS for different reasons (eg, minimizing costs, lost wages, exposure to antibiotic‐resistant bacteria), however, this interest is balanced against the potentially greater risk of readmission and adverse outcomes if LOS is inappropriately short. To date, the relationship between insurance status and LOS for CAP remains unexplored.

As in other conditions, substantial variation exists with respect to patterns of care and outcomes for children hospitalized with CAP.11 For example, children hospitalized in rural settings have a shorter LOS for pneumonia than those hospitalized in large urban settings.12 Children from racial/ethnic minorities tend to have higher rates of CAP‐associated complications, including death.11 Decades of prior studies have documented that uninsured children are less likely than insured children to make preventive care visits and obtain prescription medications, but differences in LOS or hospitalization rates between insured and uninsured children with CAP have not been studied.6, 8, 13, 14 Though imperfect, insurance status is 1 proxy for healthcare access, and current healthcare reform efforts aim to improve healthcare access and decrease socioeconomic gradients in health by increasing the number of insured American children. Nonetheless, quantifying the relationship between insurance status on LOS for children hospitalized with CAP is a first step towards understanding the influence of ambulatory care access on hospitalization for ambulatory‐care sensitive conditions.

The purpose of this study was to investigate the influence of insurance status and type on LOS for children hospitalized with CAP. In addition, we sought to determine if there were consistent trends over time in the association between insurance status and type with LOS for children hospitalized with CAP.

METHODS

RESULTS

The more than 10.5 million children sampled (unweighted) in KID during these 4 time periods (1997, 2000, 2003, and 2006) are representative of the more than 28.9 million children hospitalized in the United States. In each of these sample years, there were approximately 150,000 children hospitalized with pneumonia across the United States (Table 1). Of those hospitalized, 23% to 28% had a concomitant diagnosis of asthma; 6% to 8% had a pneumonia‐associated complication; and mortality was <0.01% in each sample year for patients hospitalized with pneumonia. In all years, among those with racial/ethnic data, the sample population was predominantly white boys less than 6 years old. The greatest proportion of children were hospitalized in urban non‐teaching settings, and also those children living in the southern regions of the United States.

There was little variation in the insurance status of children hospitalized with CAP between 1997 and 2006. In each of the sampled years, at least 40% of sampled children were privately insured, at least 40% were publicly insured, and approximately 5% were uninsured (Table 1). In all years, there were significant racial/ethnic disparities in insurance coverage such that whites were 4 to 6 times more likely to have private insurance than blacks, however, the large amount of missing race/ethnicity data warrant caution in interpreting this finding (Table 2; also see Supporting Information Appendix B in the online version of this article). We also found that children less than 1 year old were the most likely to be publicly insured in all years (see Supporting Appendix C in the online version of this article). There were also regional differences related to insurance coverage such that a greater proportion of children hospitalized in facilities located in the southern part of the United States were publicly insured. Notably, there were no significant differences in CAP‐associated mortality or asthma related to insurance coverage (Table 2). In 2006, CAP‐associated complications occurred in 8.5% of children with private insurance, 6.5% of children with public insurance, and 7.7% of uninsured children; the relative distribution of complications by insurance type were similar in previous years of the KID survey.

After examining the general and demographic characteristics, we then examined mean LOS for all children with CAP in each sample year (Table 3). The mean LOS for children with CAP was 3.44 days in 1997, with marginal decreases in subsequent years to a mean LOS of 3.18 days in 2006. The distribution of LOS for children with CAP revealed that nearly 70% of children were hospitalized for fewer than 3 days, another 22% to 28% were hospitalized for less than 1 week, and only 3% were hospitalized for more than 1 week. This distribution did not change substantially between 1997 and 2006. Next, we compared mean LOS by insurance type and race/ethnicity in unadjusted analyses. In each sample year, publicly insured children hospitalized with CAP had significantly longer LOS than privately insured children (P < 0.001). Similarly, in all years excepting 1997, uninsured children hospitalized with CAP had significantly shorter LOS than privately insured children. There were also significant racial differences in LOS for children with CAP, such that black children had longer LOS than white children with CAP. However, the large amount of missing data for race/ethnicity limited the robustness of this finding, and subsequent sensitivity analyses demonstrated that there were no consistent racial/ethnic disparities in LOS (see Supporting Appendix B in the online version of this article). These sensitivity analyses for missing race data did not alter our primary finding of shorter LOS for uninsured versus publicly or privately insured children.

After controlling for child age, race/ethnicity, gender, hospital type, transfer status, and presence of asthma or pneumonia‐associated complications, our multivariable analyses examining the relationship between insurance coverage and hospital LOS yielded the following results (Table 4). First, publicly insured children had significantly longer hospital stays than privately insured children, and uninsured children had significantly shorter hospital stays than privately insured children in all years except 1997. Second, children admitted with CAP at urban teaching children's hospitals had significantly longer LOS than those admitted to urban non‐teaching hospitals, and, in 2003, children admitted with CAP to rural hospitals had significantly shorter LOS than those admitted to urban non‐teaching hospitals. Third, children older than 1 year consistently had shorter hospital stays than infants less than 1 year old. Finally, though concomitant diagnosis of asthma did not consistently influence LOS, children who developed any complications had significantly longer LOS than those who did not. The cumulative impact of seemingly small differences in LOS is great. For example, in 2006, our model suggests that, for every 1000 children hospitalized with CAP in a given year, after adjusting for differences in sex, age, race, hospital‐type, region, transfer status, and diagnosis of asthma or complications, publicly insured children spend 90 to 130 more days in the hospital than privately insured children, whereas uninsured children spend between 40 to 90 fewer days in the hospital than privately insured children.

DISCUSSION

In this nationally representative sample selected over the past 10 years, we found that publicly insured children hospitalized with CAP have significantly longer LOS than those who are privately insured, and that, since 2000, uninsured children hospitalized with CAP have significantly shorter LOS than those who are privately insured. Though these observed differences are small, they are consistent across all 4 sampled years and, because CAP is one of the most common pediatric inpatient diagnoses, the cumulative impact of the observed differences on hospital LOS is great. Insurance status is often considered a proxy for access to preventive and ambulatory healthcare services or socioeconomic status. However, the underlying mechanisms relating insurance status to healthcare access, utilization, and ultimately, health outcomes are highly complex and difficult to elucidate.17 The observed variation in this study raises questions about the potential influence of insurance status on hospital discharge practices. Additional research is necessary to understand whether there are differences in processes of care (eg, performance of blood cultures or chest radiographs), quality of care, or other outcomes, such as readmissions, related to CAP inpatient management for children with different insurance coverage.

Apart from differences in hospital discharge practices, another possible explanation for uninsured children with CAP having shorter LOS is that these children have less severe disease than privately insured. This may occur if uninsured children with CAP are evaluated in the emergency department rather than the office setting, because emergency department providers may be more likely to admit children with CAP who lack a consistent access to ambulatory primary care services. Countering this alternative, prior studies have shown that uninsured groups are more likely to have greater disease severity than privately insured groups at the time of hospital admission.18, 19 In this study, we attempted to identify children with greater severity of disease using ICD‐9 codes for CAP‐associated complications. Though this is a relatively crude method that might lead to an underestimate of the total number of children with complications, we found that there were no significant differences in the prevalence of CAP‐associated complications between uninsured and insured groups in all sampled years.

On the other hand, uninsured patients may be released earlier by providers in order to reduce the amount of uncompensated care provided, or possibly because parents may urge providers to discharge their children, given their inability to pay forthcoming hospital bills and/or avoid further lost wages due to work absence.20, 21 In California, Bindman et al. demonstrated that decreasing the frequency of Medicaid recertification, and consequently increasing the likelihood of continuous insurance coverage, was associated with a decreased risk of hospitalization for ambulatory‐care sensitive conditions.5

We also found that children admitted to urban teaching children's hospitals with CAP had significantly longer LOS than those admitted to urban non‐teaching hospitals, whereas children in rural hospitals had significantly shorter LOS than those in urban non‐teaching hospitals in 2003. These findings are consistent with prior data from 1996 to1998 demonstrating that children admitted to rural hospitals in New York and Pennsylvania had significantly shorter LOS than large urban hospitals for 19 medical and 9 surgical conditions, including pneumonia.12 These findings may reflect underlying differences in between rural and urban hospital transfer practices, whereby rural hospitals may be more likely than urban hospitals to transfer children with relatively more severe illness to urban referral centers and retain children with less severe illness, leading to shorter LOS.12 Though our empiric understanding of differences in LOS between teaching and non‐teaching hospitals is currently limited, clinical experience supports the notion that there may be decreases in efficiency that occur in teaching hospitals, and are a result of the supervision required for care provided by trainees. It is also possible that, despite our exclusion of comorbid conditions, some children with complex or chronic medical conditions were included in this study. These children are often cared for at teaching hospitals, regardless of the primary cause for admission, and are more likely to have public insurance than other children, thus confounding the relationship between hospital type, insurance type and status, and LOS for children with CAP. The limitations of this dataset preclude further examination of this issue.

There are some limitations to this study. First, the KID data are cross‐sectional and causal inferences are limited. However, our results demonstrating that uninsured children hospitalized with CAP had shorter LOS than privately insured children were quite consistent in each sample year, suggesting that our results are a true association. Additionally, insurance status in KID is typically collected at admission, however, it is not possible to determine whether specific changes to insurance status that occurred during the hospitalization were applied to the data. The impact of this limitation would depend on the type of insurance obtained by the patient. If uninsured patients obtained public insurance, our study would underestimate the increased LOS for publicly insured patients, compared with privately insured patients, but have no effect on the difference in LOS between uninsured and privately insured patients. In the unlikely event that uninsured patients obtained private insurance, then our study would underestimate the difference for uninsured patients, compared with privately insured patients, biasing our current study results towards the null. Second, a substantial proportion of sampled children had missing data for race/ethnicity. To assess the impact of the missing race/ethnicity data on our results, we conducted sensitivity analyses and found that, though difficult to make any definitive conclusions about the relationship between race/ethnicity and LOS for children with CAP, there were no changes to our primary findings regarding differences in LOS between children with different insurance status and type. Third, KID does not include data about other unmeasured confounders (eg, parent income, parent education, regular source of care) that might be related to LOS, as well as a broad spectrum of pediatric outcomes. Serious consideration of expanding KID to include these variables is warranted. Fourth, the other category of insurance is not uniformly coded across states in the KID database. While some states use this category to classify public insurance options other than Medicare and Medicaid, other states include private insurance options in this group. Thus, it is possible that some patients with public insurance are misclassified as having other insurance. We would expect such misclassification to bias our findings towards the null hypothesis. Finally, we focused on the relationship between child health insurance status and CAP, only 1 ambulatory care‐sensitive condition. Additional research examining the relationship between insurance type and other ambulatory care‐sensitive conditions is warranted.

In summary, we found that, after multivariable adjustment, uninsured children hospitalized with community‐acquired pneumonia had significantly shorter LOS than privately insured children, and publicly insured children had a significantly longer hospital stay than privately insured children in these 4 nationally representative samples from 1997 to 2006. Current federal and state efforts to increase enrollment of children into insurance programs are a first step in reducing healthcare disparities. However, insurance coverage alone does not guarantee access to healthcare, thus, these efforts in isolation will likely be insufficient to achieve optimal health for the children of our country. As healthcare reform legislation is implemented, these findings provide hospitals and policy makers additional impetus to develop ways to achieve the ideal length of stay for every child; this ideal state will be achieved when clinical status and course, rather than nonclinical factors such as insurance type or provider's unease with ambulatory follow‐up, determine the duration of hospitalization for every child.

Disparities in patterns of care and outcomes for ambulatory‐care sensitive conditions remain a persistent problem for children.19 Many studies have focused on disparities in hospitalization rates and length of stay (LOS) related to asthma, however, few studies have focused on community‐acquired pneumonia (CAP) despite the fact that pneumonia is the most common, preventable, and potentially serious infection in childhood.10 Providers, payers, and families have a common interest in minimizing hospital LOS for different reasons (eg, minimizing costs, lost wages, exposure to antibiotic‐resistant bacteria), however, this interest is balanced against the potentially greater risk of readmission and adverse outcomes if LOS is inappropriately short. To date, the relationship between insurance status and LOS for CAP remains unexplored.

As in other conditions, substantial variation exists with respect to patterns of care and outcomes for children hospitalized with CAP.11 For example, children hospitalized in rural settings have a shorter LOS for pneumonia than those hospitalized in large urban settings.12 Children from racial/ethnic minorities tend to have higher rates of CAP‐associated complications, including death.11 Decades of prior studies have documented that uninsured children are less likely than insured children to make preventive care visits and obtain prescription medications, but differences in LOS or hospitalization rates between insured and uninsured children with CAP have not been studied.6, 8, 13, 14 Though imperfect, insurance status is 1 proxy for healthcare access, and current healthcare reform efforts aim to improve healthcare access and decrease socioeconomic gradients in health by increasing the number of insured American children. Nonetheless, quantifying the relationship between insurance status on LOS for children hospitalized with CAP is a first step towards understanding the influence of ambulatory care access on hospitalization for ambulatory‐care sensitive conditions.

The purpose of this study was to investigate the influence of insurance status and type on LOS for children hospitalized with CAP. In addition, we sought to determine if there were consistent trends over time in the association between insurance status and type with LOS for children hospitalized with CAP.

METHODS

RESULTS

The more than 10.5 million children sampled (unweighted) in KID during these 4 time periods (1997, 2000, 2003, and 2006) are representative of the more than 28.9 million children hospitalized in the United States. In each of these sample years, there were approximately 150,000 children hospitalized with pneumonia across the United States (Table 1). Of those hospitalized, 23% to 28% had a concomitant diagnosis of asthma; 6% to 8% had a pneumonia‐associated complication; and mortality was <0.01% in each sample year for patients hospitalized with pneumonia. In all years, among those with racial/ethnic data, the sample population was predominantly white boys less than 6 years old. The greatest proportion of children were hospitalized in urban non‐teaching settings, and also those children living in the southern regions of the United States.

There was little variation in the insurance status of children hospitalized with CAP between 1997 and 2006. In each of the sampled years, at least 40% of sampled children were privately insured, at least 40% were publicly insured, and approximately 5% were uninsured (Table 1). In all years, there were significant racial/ethnic disparities in insurance coverage such that whites were 4 to 6 times more likely to have private insurance than blacks, however, the large amount of missing race/ethnicity data warrant caution in interpreting this finding (Table 2; also see Supporting Information Appendix B in the online version of this article). We also found that children less than 1 year old were the most likely to be publicly insured in all years (see Supporting Appendix C in the online version of this article). There were also regional differences related to insurance coverage such that a greater proportion of children hospitalized in facilities located in the southern part of the United States were publicly insured. Notably, there were no significant differences in CAP‐associated mortality or asthma related to insurance coverage (Table 2). In 2006, CAP‐associated complications occurred in 8.5% of children with private insurance, 6.5% of children with public insurance, and 7.7% of uninsured children; the relative distribution of complications by insurance type were similar in previous years of the KID survey.

After examining the general and demographic characteristics, we then examined mean LOS for all children with CAP in each sample year (Table 3). The mean LOS for children with CAP was 3.44 days in 1997, with marginal decreases in subsequent years to a mean LOS of 3.18 days in 2006. The distribution of LOS for children with CAP revealed that nearly 70% of children were hospitalized for fewer than 3 days, another 22% to 28% were hospitalized for less than 1 week, and only 3% were hospitalized for more than 1 week. This distribution did not change substantially between 1997 and 2006. Next, we compared mean LOS by insurance type and race/ethnicity in unadjusted analyses. In each sample year, publicly insured children hospitalized with CAP had significantly longer LOS than privately insured children (P < 0.001). Similarly, in all years excepting 1997, uninsured children hospitalized with CAP had significantly shorter LOS than privately insured children. There were also significant racial differences in LOS for children with CAP, such that black children had longer LOS than white children with CAP. However, the large amount of missing data for race/ethnicity limited the robustness of this finding, and subsequent sensitivity analyses demonstrated that there were no consistent racial/ethnic disparities in LOS (see Supporting Appendix B in the online version of this article). These sensitivity analyses for missing race data did not alter our primary finding of shorter LOS for uninsured versus publicly or privately insured children.

After controlling for child age, race/ethnicity, gender, hospital type, transfer status, and presence of asthma or pneumonia‐associated complications, our multivariable analyses examining the relationship between insurance coverage and hospital LOS yielded the following results (Table 4). First, publicly insured children had significantly longer hospital stays than privately insured children, and uninsured children had significantly shorter hospital stays than privately insured children in all years except 1997. Second, children admitted with CAP at urban teaching children's hospitals had significantly longer LOS than those admitted to urban non‐teaching hospitals, and, in 2003, children admitted with CAP to rural hospitals had significantly shorter LOS than those admitted to urban non‐teaching hospitals. Third, children older than 1 year consistently had shorter hospital stays than infants less than 1 year old. Finally, though concomitant diagnosis of asthma did not consistently influence LOS, children who developed any complications had significantly longer LOS than those who did not. The cumulative impact of seemingly small differences in LOS is great. For example, in 2006, our model suggests that, for every 1000 children hospitalized with CAP in a given year, after adjusting for differences in sex, age, race, hospital‐type, region, transfer status, and diagnosis of asthma or complications, publicly insured children spend 90 to 130 more days in the hospital than privately insured children, whereas uninsured children spend between 40 to 90 fewer days in the hospital than privately insured children.

DISCUSSION

In this nationally representative sample selected over the past 10 years, we found that publicly insured children hospitalized with CAP have significantly longer LOS than those who are privately insured, and that, since 2000, uninsured children hospitalized with CAP have significantly shorter LOS than those who are privately insured. Though these observed differences are small, they are consistent across all 4 sampled years and, because CAP is one of the most common pediatric inpatient diagnoses, the cumulative impact of the observed differences on hospital LOS is great. Insurance status is often considered a proxy for access to preventive and ambulatory healthcare services or socioeconomic status. However, the underlying mechanisms relating insurance status to healthcare access, utilization, and ultimately, health outcomes are highly complex and difficult to elucidate.17 The observed variation in this study raises questions about the potential influence of insurance status on hospital discharge practices. Additional research is necessary to understand whether there are differences in processes of care (eg, performance of blood cultures or chest radiographs), quality of care, or other outcomes, such as readmissions, related to CAP inpatient management for children with different insurance coverage.

Apart from differences in hospital discharge practices, another possible explanation for uninsured children with CAP having shorter LOS is that these children have less severe disease than privately insured. This may occur if uninsured children with CAP are evaluated in the emergency department rather than the office setting, because emergency department providers may be more likely to admit children with CAP who lack a consistent access to ambulatory primary care services. Countering this alternative, prior studies have shown that uninsured groups are more likely to have greater disease severity than privately insured groups at the time of hospital admission.18, 19 In this study, we attempted to identify children with greater severity of disease using ICD‐9 codes for CAP‐associated complications. Though this is a relatively crude method that might lead to an underestimate of the total number of children with complications, we found that there were no significant differences in the prevalence of CAP‐associated complications between uninsured and insured groups in all sampled years.

On the other hand, uninsured patients may be released earlier by providers in order to reduce the amount of uncompensated care provided, or possibly because parents may urge providers to discharge their children, given their inability to pay forthcoming hospital bills and/or avoid further lost wages due to work absence.20, 21 In California, Bindman et al. demonstrated that decreasing the frequency of Medicaid recertification, and consequently increasing the likelihood of continuous insurance coverage, was associated with a decreased risk of hospitalization for ambulatory‐care sensitive conditions.5

We also found that children admitted to urban teaching children's hospitals with CAP had significantly longer LOS than those admitted to urban non‐teaching hospitals, whereas children in rural hospitals had significantly shorter LOS than those in urban non‐teaching hospitals in 2003. These findings are consistent with prior data from 1996 to1998 demonstrating that children admitted to rural hospitals in New York and Pennsylvania had significantly shorter LOS than large urban hospitals for 19 medical and 9 surgical conditions, including pneumonia.12 These findings may reflect underlying differences in between rural and urban hospital transfer practices, whereby rural hospitals may be more likely than urban hospitals to transfer children with relatively more severe illness to urban referral centers and retain children with less severe illness, leading to shorter LOS.12 Though our empiric understanding of differences in LOS between teaching and non‐teaching hospitals is currently limited, clinical experience supports the notion that there may be decreases in efficiency that occur in teaching hospitals, and are a result of the supervision required for care provided by trainees. It is also possible that, despite our exclusion of comorbid conditions, some children with complex or chronic medical conditions were included in this study. These children are often cared for at teaching hospitals, regardless of the primary cause for admission, and are more likely to have public insurance than other children, thus confounding the relationship between hospital type, insurance type and status, and LOS for children with CAP. The limitations of this dataset preclude further examination of this issue.

There are some limitations to this study. First, the KID data are cross‐sectional and causal inferences are limited. However, our results demonstrating that uninsured children hospitalized with CAP had shorter LOS than privately insured children were quite consistent in each sample year, suggesting that our results are a true association. Additionally, insurance status in KID is typically collected at admission, however, it is not possible to determine whether specific changes to insurance status that occurred during the hospitalization were applied to the data. The impact of this limitation would depend on the type of insurance obtained by the patient. If uninsured patients obtained public insurance, our study would underestimate the increased LOS for publicly insured patients, compared with privately insured patients, but have no effect on the difference in LOS between uninsured and privately insured patients. In the unlikely event that uninsured patients obtained private insurance, then our study would underestimate the difference for uninsured patients, compared with privately insured patients, biasing our current study results towards the null. Second, a substantial proportion of sampled children had missing data for race/ethnicity. To assess the impact of the missing race/ethnicity data on our results, we conducted sensitivity analyses and found that, though difficult to make any definitive conclusions about the relationship between race/ethnicity and LOS for children with CAP, there were no changes to our primary findings regarding differences in LOS between children with different insurance status and type. Third, KID does not include data about other unmeasured confounders (eg, parent income, parent education, regular source of care) that might be related to LOS, as well as a broad spectrum of pediatric outcomes. Serious consideration of expanding KID to include these variables is warranted. Fourth, the other category of insurance is not uniformly coded across states in the KID database. While some states use this category to classify public insurance options other than Medicare and Medicaid, other states include private insurance options in this group. Thus, it is possible that some patients with public insurance are misclassified as having other insurance. We would expect such misclassification to bias our findings towards the null hypothesis. Finally, we focused on the relationship between child health insurance status and CAP, only 1 ambulatory care‐sensitive condition. Additional research examining the relationship between insurance type and other ambulatory care‐sensitive conditions is warranted.

In summary, we found that, after multivariable adjustment, uninsured children hospitalized with community‐acquired pneumonia had significantly shorter LOS than privately insured children, and publicly insured children had a significantly longer hospital stay than privately insured children in these 4 nationally representative samples from 1997 to 2006. Current federal and state efforts to increase enrollment of children into insurance programs are a first step in reducing healthcare disparities. However, insurance coverage alone does not guarantee access to healthcare, thus, these efforts in isolation will likely be insufficient to achieve optimal health for the children of our country. As healthcare reform legislation is implemented, these findings provide hospitals and policy makers additional impetus to develop ways to achieve the ideal length of stay for every child; this ideal state will be achieved when clinical status and course, rather than nonclinical factors such as insurance type or provider's unease with ambulatory follow‐up, determine the duration of hospitalization for every child.

High levels of hospital occupancy are associated with compromises to quality of care and access (often referred to as crowding), 18 while low occupancy may be inefficient and also impact quality. 9, 10 Despite this, hospitals typically have uneven occupancy. Although some demand for services is driven by factors beyond the control of a hospital (eg, seasonal variation in viral illness), approximately 15%30% of admissions to children's hospitals are scheduled from days to months in advance, with usual arrivals on weekdays. 1114 For example, of the 3.4 million elective admissions in the 2006 Healthcare Cost and Utilization Project Kids Inpatient Database (HCUP KID), only 13% were admitted on weekends. 14 Combined with short length of stay (LOS) for such patients, this leads to higher midweek and lower weekend occupancy. 12

Hospitals respond to crowding in a number of ways, but often focus on reducing LOS to make room for new patients. 11, 15, 16 For hospitals that are relatively efficient in terms of LOS, efforts to reduce it may not increase functional capacity adequately. In children's hospitals, median lengths of stay are 2 to 3 days, and one‐third of hospitalizations are 1 day or less. 17 Thus, even 10%20% reductions in LOS trims hours, not days, from typical stays. Practical barriers (eg, reluctance to discharge in the middle of the night, or family preferences and work schedules) and undesired outcomes (eg, increased hospital re‐visits) are additional pitfalls encountered by relying on throughput enhancement alone.

Managing scheduled admissions through smoothing is an alternative strategy to reduce variability and high occupancy. 6, 12, 1820 The concept is to proactively control the entry of patients, when possible, to achieve more even levels of occupancy, instead of the peaks and troughs commonly encountered. Nonetheless, it is not a widely used approach. 18, 20, 21 We hypothesized that children's hospitals had substantial unused capacity that could be used to smooth occupancy, which would reduce weekday crowding. While it is obvious that smoothing will reduce peaks to average levels (and also raise troughs), we sought to quantify just how large this difference wasand thereby quantify the potential of smoothing to reduce inpatient crowding (or, conversely, expose more patients to high levels of occupancy). Is there enough variation to justify smoothing, and, if a hospital does smooth, what is the expected result? If the number of patients removed from exposure to high occupancy is not substantial, other means to address inpatient crowding might be of more value. Our aims were to quantify the difference in weekday versus weekend occupancy, report on mathematical feasibility of such an approach, and determine the difference in number of patients exposed to various levels of high occupancy.

Methods

Results

The characteristics of the 39 hospitals are provided in Table 1. Based on standardization with KID, 23.6% of PHIS admissions were scheduled (range: 18.1%35.8%) or a median of 81.5 scheduled admissions per week per hospital; 26.6% of weekday admissions were scheduled versus 16.1% for weekends. Overall, 12.4% of scheduled admissions entered on weekends. For all patients, median LOS was three days (interquartile range [IQR]: twofive days), but median LOS for scheduled admissions was two days (IQR: onefour days). The median LOS and IQR were the same by day of admission for all days of the week. Most hospitals had an overall SLOSR close to one (median: 0.9, IQR: 0.91.1). Overall, hospital mean midnight occupancy ranged from 70.9% to 108.1% on weekdays and 65.7% to 94.9% on weekends. Uniformly, weekday occupancy exceeded weekend occupancy, with a median difference of 8.2% points (IQR: 7.2%9.5% points). There was a wide range of median hospital weekdayweekend occupancy differences across hospitals (Figure 1). The overall difference was less in winter (median difference: 7.7% points; IQR: 6.3%8.8% points) than in summer (median difference: 8.6% points; IQR: 7.4%9.8% points (Wilcoxon Sign Rank test, P < 0.001). Thirty‐five hospitals (89.7%) exceeded the 85% occupancy threshold and 29 (74.4%) exceeded the 95% occupancy threshold on at least 20% of weekdays (Table 2). Across all the hospitals, the median difference in weekly maximum and weekly mean occupancy was 6.6% points (IQR: 6.2%7.4% points) (Figure 2).

Figure 1

Figure 2

Smoothing reduced the number of hospitals at each occupancy threshold, except 85% (Table 2). As a linear relationship, the reduction in weekday peak occupancy (y) based on a hospital's median difference in weekly maximum and weekly mean occupancy (x) was y = 2.69 + 0.48x. Thus, a hospital with a 10% point difference between weekday and weekend occupancy could reduce weekday peak by 7.5% points.

Smoothing increased the number of patients exposed to the lower thresholds (85% and 90%), but decreased the number of patients exposed to >95% occupancy (Table 2). For example, smoothing at the 95% threshold resulted in 630 fewer patients per hospital exposed to that threshold. If all 39 hospitals had within‐week smoothing, a net of 39,607 patients would have been protected from exposure to >95% occupancy and a net of 50,079 patients from 100% occupancy.

To demonstrate the varied effects of smoothing, Table 3 and Figure 3 present representative categories of response to smoothing depending on pre‐smoothing patterns. While not all hospitals decreased occupancy to below thresholds after smoothing (Types B and D), the overall occupancy was reduced and fewer patients were exposed to extreme levels of high occupancy (eg, >100%).

Figure 3

To achieve within‐week smoothing, a median of 7.4 patient‐admissions per week (range: 2.314.4) would have to be scheduled on a different day of the week. This equates to a median of 2.6% (IQR: 2.25%, 2.99%; range: 0.02%9.2%) of all admissionsor 9% of a typical hospital‐week's scheduled admissions.

Discussion

This analysis of 39 children's hospitals found high levels of occupancy and weekend occupancy lower than weekday occupancy (median difference: 8.2% points). Only 12.4% of scheduled admissions entered on weekends. Thus, weekend capacity is available to offset high weekday occupancy. Hospitals at the higher end of the occupancy thresholds (95%, 100%) would reduce the number of days operating at very high occupancy and the number of patients exposed to such levels by smoothing. This change is mathematically feasible, as a median of 7.4 patients would have to be proactively scheduled differently each week, just under one‐tenth of scheduled admissions. Since LOS by day of admission was the same (median: two days), the opportunity to affect occupancy by shifting patients should be relatively similar for all days of the week. In addition, these admissions were short, conferring greater flexibility. Implementing smoothing over the course of the week does not necessarily require admitting patients on weekends. For example, Monday admissions with an anticipated three‐day LOS could enter on Friday with anticipated discharge on Monday to alleviate midweek crowding and take advantage of unoccupied weekend beds. 26

At the highest levels of occupancy, smoothing reduces the frequency of reaching these maximum levels, but can have the effect of actually exposing more patient‐days to a higher occupancy. For example, for nine hospitals in our analysis with >20% of days over 100%, smoothing decreased days over 100%, but exposed weekend patients to higher levels of occupancy (Figure 3). Since most admissions are short and most scheduled admissions currently occur on weekdays, the number of individual patients (not patient‐days) newly exposed to such high occupancy may not increase much after smoothing at these facilities. Regardless, hospitals with such a pattern may not be able to rely solely on smoothing to avoid weekday crowding, and, if they are operating efficiently in terms of SLOSR, might be justified in building more capacity.

Consistent with our findings, the Institute for Healthcare Improvement, the Institute for Healthcare Optimization, and the American Hospital Association Quality Center stress that addressing artificial variability of scheduled admissions is a critical first step to improving patient flow and quality of care while reducing costs. 18, 21, 27 Our study suggests that small numbers of patients need to be proactively scheduled differently to decrease midweek peak occupancy, so only a small proportion of families would need to find this desirable to make it attractive for hospitals and patients. This type of proactive smoothing decreases peak occupancy on weekdays, reducing the safety risks associated with high occupancy, improving acute access for emergent patients, shortening wait‐times and loss of scheduled patients to another facility, and increasing procedure volume (3%74% in one study). 28 Smoothing may also increase quality and safety on weekends, as emergent patients admitted on weekends experience more delays in necessary treatment and have worse outcomes. 2932 In addition, increasing scheduled admissions to span weekends may appeal to some families wishing to avoid absence from work to be with their hospitalized child, to parents concerned about school performanceand may also appeal to staff members seeking flexible schedules. Increasing weekend hospital capacity is safe, feasible, and economical, even when considering the increased wages for weekend work. 33, 34 Finally, smoothing over the whole week allows fixed costs (eg, surgical suites, imaging equipment) to be allocated over 7 days rather than 5, and allows for better matching of revenue to the fixed expenses.

Rather than a prescriptive approach, our work suggests hospitals need to identify only a small number of patients to proactively shift, providing them opportunities to adapt the approach to local circumstances. The particular patients to move around may also depend on the costs and benefits of services (eg, radiologic, laboratory, operative) and the hospital's existing patterns of staffing. A number of hospitals that have engaged in similar work have achieved sustainable results, such as Seattle Children's Hospital, Boston Medical Center, St. John's Regional Health Center, and New York University Langone Medical Center. 19, 26, 3537 In these cases, proactive smoothing took advantage of unused capacity and decreased crowding on days that had been traditionally very full. Hospitals that rarely or never have high‐occupancy days, and that do not expect growth in volume, may not need to employ smoothing, whereas others that have crowding issues primarily in the winter may wish to implement smoothing techniques seasonally.

Aside from attempting to reduce high‐occupancy through modification of admission patterns, other proactive approaches include optimizing staffing and processes around care, improving efficiency of care, and building additional beds. 16, 25, 38, 39 However, the expense of construction and the scarcity of capital often preclude this last option. Among children's hospitals, with SLOSR close to one, implementing strategies to reduce the LOS during periods of high occupancy may not result in meaningful reductions in LOS, as such approaches would only decrease the typical child's hospitalization by hours, not days. In addition to proactive strategies, hospitals also rely on reactive approaches, such as ED boarding, placing patients in hallways on units, diverting ambulances or transfers, or canceling scheduled admissions at the last moment, to decrease crowding. 16, 39, 40

This study has several limitations. First, use of administrative data precluded modeling all responses. For example, some hospitals may be better able to accommodate fluctuations in census or high occupancy without compromising quality or access. Second, we only considered intra‐week smoothing, but hospitals may benefit from smoothing over longer periods of time, especially since children's hospitals are busier in winter months, but incoming scheduled volume is often not reduced. 11 Hospitals with large occupancy variations across months may want to consider broadening the time horizon for smoothing, and weigh the costs and benefits over that period of time, including parental and clinician concerns and preferences for not delaying treatment. At the individual hospital level, discrete‐event simulation would likely be useful to consider the trade‐offs of smoothing to different levels and over different periods of time. Third, we assumed a fixed number of beds for the year, an approach that may not accurately reflect actual available beds on specific days. This limitation was minimized by counting all beds for each hospital as available for all the days of the year, so that hospitals with a high census when all available beds are included would have an even higher percent occupancy if some of those beds were not actually open. In a related way, then, we also do not consider how staffing may need to be altered or augmented to care for additional patients on certain days. Fourth, midnight census, the only universally available measure, was used to determine occupancy rather than peak census. Midnight census provides a standard snapshot, but is lower than mid‐day peak census. 41 In order to account for these limitations, we considered several different thresholds of high occupancy. Fifth, we smoothed at the hospital level, but differential effects may exist at the unit level. Sixth, to determine proportion of scheduled admissions, we used HCUP KID proportions on PHIS admissions. Overall, this approach likely overestimated scheduled medical admissions on weekends, thus biasing our result towards the null hypothesis. Finally, only freestanding children's hospitals were included in this study. While this may limit generalizability, the general concept of smoothing occupancy should apply in any setting with substantial and consistent variation.

In summary, our study revealed that children's hospitals often face high midweek occupancy, but also have substantial unused weekend capacity. Hospitals facing challenges with high weekday occupancy could proactively use a smoothing approach to decrease the frequency and severity of high occupancy. Further qualitative evaluation is also warranted around child, family, and staff preferences concerning scheduled admissions, school, and work.

High levels of hospital occupancy are associated with compromises to quality of care and access (often referred to as crowding), 18 while low occupancy may be inefficient and also impact quality. 9, 10 Despite this, hospitals typically have uneven occupancy. Although some demand for services is driven by factors beyond the control of a hospital (eg, seasonal variation in viral illness), approximately 15%30% of admissions to children's hospitals are scheduled from days to months in advance, with usual arrivals on weekdays. 1114 For example, of the 3.4 million elective admissions in the 2006 Healthcare Cost and Utilization Project Kids Inpatient Database (HCUP KID), only 13% were admitted on weekends. 14 Combined with short length of stay (LOS) for such patients, this leads to higher midweek and lower weekend occupancy. 12

Hospitals respond to crowding in a number of ways, but often focus on reducing LOS to make room for new patients. 11, 15, 16 For hospitals that are relatively efficient in terms of LOS, efforts to reduce it may not increase functional capacity adequately. In children's hospitals, median lengths of stay are 2 to 3 days, and one‐third of hospitalizations are 1 day or less. 17 Thus, even 10%20% reductions in LOS trims hours, not days, from typical stays. Practical barriers (eg, reluctance to discharge in the middle of the night, or family preferences and work schedules) and undesired outcomes (eg, increased hospital re‐visits) are additional pitfalls encountered by relying on throughput enhancement alone.

Managing scheduled admissions through smoothing is an alternative strategy to reduce variability and high occupancy. 6, 12, 1820 The concept is to proactively control the entry of patients, when possible, to achieve more even levels of occupancy, instead of the peaks and troughs commonly encountered. Nonetheless, it is not a widely used approach. 18, 20, 21 We hypothesized that children's hospitals had substantial unused capacity that could be used to smooth occupancy, which would reduce weekday crowding. While it is obvious that smoothing will reduce peaks to average levels (and also raise troughs), we sought to quantify just how large this difference wasand thereby quantify the potential of smoothing to reduce inpatient crowding (or, conversely, expose more patients to high levels of occupancy). Is there enough variation to justify smoothing, and, if a hospital does smooth, what is the expected result? If the number of patients removed from exposure to high occupancy is not substantial, other means to address inpatient crowding might be of more value. Our aims were to quantify the difference in weekday versus weekend occupancy, report on mathematical feasibility of such an approach, and determine the difference in number of patients exposed to various levels of high occupancy.

Methods

Results

The characteristics of the 39 hospitals are provided in Table 1. Based on standardization with KID, 23.6% of PHIS admissions were scheduled (range: 18.1%35.8%) or a median of 81.5 scheduled admissions per week per hospital; 26.6% of weekday admissions were scheduled versus 16.1% for weekends. Overall, 12.4% of scheduled admissions entered on weekends. For all patients, median LOS was three days (interquartile range [IQR]: twofive days), but median LOS for scheduled admissions was two days (IQR: onefour days). The median LOS and IQR were the same by day of admission for all days of the week. Most hospitals had an overall SLOSR close to one (median: 0.9, IQR: 0.91.1). Overall, hospital mean midnight occupancy ranged from 70.9% to 108.1% on weekdays and 65.7% to 94.9% on weekends. Uniformly, weekday occupancy exceeded weekend occupancy, with a median difference of 8.2% points (IQR: 7.2%9.5% points). There was a wide range of median hospital weekdayweekend occupancy differences across hospitals (Figure 1). The overall difference was less in winter (median difference: 7.7% points; IQR: 6.3%8.8% points) than in summer (median difference: 8.6% points; IQR: 7.4%9.8% points (Wilcoxon Sign Rank test, P < 0.001). Thirty‐five hospitals (89.7%) exceeded the 85% occupancy threshold and 29 (74.4%) exceeded the 95% occupancy threshold on at least 20% of weekdays (Table 2). Across all the hospitals, the median difference in weekly maximum and weekly mean occupancy was 6.6% points (IQR: 6.2%7.4% points) (Figure 2).

Figure 1

Figure 2

Smoothing reduced the number of hospitals at each occupancy threshold, except 85% (Table 2). As a linear relationship, the reduction in weekday peak occupancy (y) based on a hospital's median difference in weekly maximum and weekly mean occupancy (x) was y = 2.69 + 0.48x. Thus, a hospital with a 10% point difference between weekday and weekend occupancy could reduce weekday peak by 7.5% points.

Smoothing increased the number of patients exposed to the lower thresholds (85% and 90%), but decreased the number of patients exposed to >95% occupancy (Table 2). For example, smoothing at the 95% threshold resulted in 630 fewer patients per hospital exposed to that threshold. If all 39 hospitals had within‐week smoothing, a net of 39,607 patients would have been protected from exposure to >95% occupancy and a net of 50,079 patients from 100% occupancy.

To demonstrate the varied effects of smoothing, Table 3 and Figure 3 present representative categories of response to smoothing depending on pre‐smoothing patterns. While not all hospitals decreased occupancy to below thresholds after smoothing (Types B and D), the overall occupancy was reduced and fewer patients were exposed to extreme levels of high occupancy (eg, >100%).

Figure 3

To achieve within‐week smoothing, a median of 7.4 patient‐admissions per week (range: 2.314.4) would have to be scheduled on a different day of the week. This equates to a median of 2.6% (IQR: 2.25%, 2.99%; range: 0.02%9.2%) of all admissionsor 9% of a typical hospital‐week's scheduled admissions.

Discussion

This analysis of 39 children's hospitals found high levels of occupancy and weekend occupancy lower than weekday occupancy (median difference: 8.2% points). Only 12.4% of scheduled admissions entered on weekends. Thus, weekend capacity is available to offset high weekday occupancy. Hospitals at the higher end of the occupancy thresholds (95%, 100%) would reduce the number of days operating at very high occupancy and the number of patients exposed to such levels by smoothing. This change is mathematically feasible, as a median of 7.4 patients would have to be proactively scheduled differently each week, just under one‐tenth of scheduled admissions. Since LOS by day of admission was the same (median: two days), the opportunity to affect occupancy by shifting patients should be relatively similar for all days of the week. In addition, these admissions were short, conferring greater flexibility. Implementing smoothing over the course of the week does not necessarily require admitting patients on weekends. For example, Monday admissions with an anticipated three‐day LOS could enter on Friday with anticipated discharge on Monday to alleviate midweek crowding and take advantage of unoccupied weekend beds. 26

At the highest levels of occupancy, smoothing reduces the frequency of reaching these maximum levels, but can have the effect of actually exposing more patient‐days to a higher occupancy. For example, for nine hospitals in our analysis with >20% of days over 100%, smoothing decreased days over 100%, but exposed weekend patients to higher levels of occupancy (Figure 3). Since most admissions are short and most scheduled admissions currently occur on weekdays, the number of individual patients (not patient‐days) newly exposed to such high occupancy may not increase much after smoothing at these facilities. Regardless, hospitals with such a pattern may not be able to rely solely on smoothing to avoid weekday crowding, and, if they are operating efficiently in terms of SLOSR, might be justified in building more capacity.

Consistent with our findings, the Institute for Healthcare Improvement, the Institute for Healthcare Optimization, and the American Hospital Association Quality Center stress that addressing artificial variability of scheduled admissions is a critical first step to improving patient flow and quality of care while reducing costs. 18, 21, 27 Our study suggests that small numbers of patients need to be proactively scheduled differently to decrease midweek peak occupancy, so only a small proportion of families would need to find this desirable to make it attractive for hospitals and patients. This type of proactive smoothing decreases peak occupancy on weekdays, reducing the safety risks associated with high occupancy, improving acute access for emergent patients, shortening wait‐times and loss of scheduled patients to another facility, and increasing procedure volume (3%74% in one study). 28 Smoothing may also increase quality and safety on weekends, as emergent patients admitted on weekends experience more delays in necessary treatment and have worse outcomes. 2932 In addition, increasing scheduled admissions to span weekends may appeal to some families wishing to avoid absence from work to be with their hospitalized child, to parents concerned about school performanceand may also appeal to staff members seeking flexible schedules. Increasing weekend hospital capacity is safe, feasible, and economical, even when considering the increased wages for weekend work. 33, 34 Finally, smoothing over the whole week allows fixed costs (eg, surgical suites, imaging equipment) to be allocated over 7 days rather than 5, and allows for better matching of revenue to the fixed expenses.

Rather than a prescriptive approach, our work suggests hospitals need to identify only a small number of patients to proactively shift, providing them opportunities to adapt the approach to local circumstances. The particular patients to move around may also depend on the costs and benefits of services (eg, radiologic, laboratory, operative) and the hospital's existing patterns of staffing. A number of hospitals that have engaged in similar work have achieved sustainable results, such as Seattle Children's Hospital, Boston Medical Center, St. John's Regional Health Center, and New York University Langone Medical Center. 19, 26, 3537 In these cases, proactive smoothing took advantage of unused capacity and decreased crowding on days that had been traditionally very full. Hospitals that rarely or never have high‐occupancy days, and that do not expect growth in volume, may not need to employ smoothing, whereas others that have crowding issues primarily in the winter may wish to implement smoothing techniques seasonally.

Aside from attempting to reduce high‐occupancy through modification of admission patterns, other proactive approaches include optimizing staffing and processes around care, improving efficiency of care, and building additional beds. 16, 25, 38, 39 However, the expense of construction and the scarcity of capital often preclude this last option. Among children's hospitals, with SLOSR close to one, implementing strategies to reduce the LOS during periods of high occupancy may not result in meaningful reductions in LOS, as such approaches would only decrease the typical child's hospitalization by hours, not days. In addition to proactive strategies, hospitals also rely on reactive approaches, such as ED boarding, placing patients in hallways on units, diverting ambulances or transfers, or canceling scheduled admissions at the last moment, to decrease crowding. 16, 39, 40

This study has several limitations. First, use of administrative data precluded modeling all responses. For example, some hospitals may be better able to accommodate fluctuations in census or high occupancy without compromising quality or access. Second, we only considered intra‐week smoothing, but hospitals may benefit from smoothing over longer periods of time, especially since children's hospitals are busier in winter months, but incoming scheduled volume is often not reduced. 11 Hospitals with large occupancy variations across months may want to consider broadening the time horizon for smoothing, and weigh the costs and benefits over that period of time, including parental and clinician concerns and preferences for not delaying treatment. At the individual hospital level, discrete‐event simulation would likely be useful to consider the trade‐offs of smoothing to different levels and over different periods of time. Third, we assumed a fixed number of beds for the year, an approach that may not accurately reflect actual available beds on specific days. This limitation was minimized by counting all beds for each hospital as available for all the days of the year, so that hospitals with a high census when all available beds are included would have an even higher percent occupancy if some of those beds were not actually open. In a related way, then, we also do not consider how staffing may need to be altered or augmented to care for additional patients on certain days. Fourth, midnight census, the only universally available measure, was used to determine occupancy rather than peak census. Midnight census provides a standard snapshot, but is lower than mid‐day peak census. 41 In order to account for these limitations, we considered several different thresholds of high occupancy. Fifth, we smoothed at the hospital level, but differential effects may exist at the unit level. Sixth, to determine proportion of scheduled admissions, we used HCUP KID proportions on PHIS admissions. Overall, this approach likely overestimated scheduled medical admissions on weekends, thus biasing our result towards the null hypothesis. Finally, only freestanding children's hospitals were included in this study. While this may limit generalizability, the general concept of smoothing occupancy should apply in any setting with substantial and consistent variation.

In summary, our study revealed that children's hospitals often face high midweek occupancy, but also have substantial unused weekend capacity. Hospitals facing challenges with high weekday occupancy could proactively use a smoothing approach to decrease the frequency and severity of high occupancy. Further qualitative evaluation is also warranted around child, family, and staff preferences concerning scheduled admissions, school, and work.

High levels of hospital occupancy are associated with compromises to quality of care and access (often referred to as crowding), 18 while low occupancy may be inefficient and also impact quality. 9, 10 Despite this, hospitals typically have uneven occupancy. Although some demand for services is driven by factors beyond the control of a hospital (eg, seasonal variation in viral illness), approximately 15%30% of admissions to children's hospitals are scheduled from days to months in advance, with usual arrivals on weekdays. 1114 For example, of the 3.4 million elective admissions in the 2006 Healthcare Cost and Utilization Project Kids Inpatient Database (HCUP KID), only 13% were admitted on weekends. 14 Combined with short length of stay (LOS) for such patients, this leads to higher midweek and lower weekend occupancy. 12

Hospitals respond to crowding in a number of ways, but often focus on reducing LOS to make room for new patients. 11, 15, 16 For hospitals that are relatively efficient in terms of LOS, efforts to reduce it may not increase functional capacity adequately. In children's hospitals, median lengths of stay are 2 to 3 days, and one‐third of hospitalizations are 1 day or less. 17 Thus, even 10%20% reductions in LOS trims hours, not days, from typical stays. Practical barriers (eg, reluctance to discharge in the middle of the night, or family preferences and work schedules) and undesired outcomes (eg, increased hospital re‐visits) are additional pitfalls encountered by relying on throughput enhancement alone.

Managing scheduled admissions through smoothing is an alternative strategy to reduce variability and high occupancy. 6, 12, 1820 The concept is to proactively control the entry of patients, when possible, to achieve more even levels of occupancy, instead of the peaks and troughs commonly encountered. Nonetheless, it is not a widely used approach. 18, 20, 21 We hypothesized that children's hospitals had substantial unused capacity that could be used to smooth occupancy, which would reduce weekday crowding. While it is obvious that smoothing will reduce peaks to average levels (and also raise troughs), we sought to quantify just how large this difference wasand thereby quantify the potential of smoothing to reduce inpatient crowding (or, conversely, expose more patients to high levels of occupancy). Is there enough variation to justify smoothing, and, if a hospital does smooth, what is the expected result? If the number of patients removed from exposure to high occupancy is not substantial, other means to address inpatient crowding might be of more value. Our aims were to quantify the difference in weekday versus weekend occupancy, report on mathematical feasibility of such an approach, and determine the difference in number of patients exposed to various levels of high occupancy.

Methods

Results

The characteristics of the 39 hospitals are provided in Table 1. Based on standardization with KID, 23.6% of PHIS admissions were scheduled (range: 18.1%35.8%) or a median of 81.5 scheduled admissions per week per hospital; 26.6% of weekday admissions were scheduled versus 16.1% for weekends. Overall, 12.4% of scheduled admissions entered on weekends. For all patients, median LOS was three days (interquartile range [IQR]: twofive days), but median LOS for scheduled admissions was two days (IQR: onefour days). The median LOS and IQR were the same by day of admission for all days of the week. Most hospitals had an overall SLOSR close to one (median: 0.9, IQR: 0.91.1). Overall, hospital mean midnight occupancy ranged from 70.9% to 108.1% on weekdays and 65.7% to 94.9% on weekends. Uniformly, weekday occupancy exceeded weekend occupancy, with a median difference of 8.2% points (IQR: 7.2%9.5% points). There was a wide range of median hospital weekdayweekend occupancy differences across hospitals (Figure 1). The overall difference was less in winter (median difference: 7.7% points; IQR: 6.3%8.8% points) than in summer (median difference: 8.6% points; IQR: 7.4%9.8% points (Wilcoxon Sign Rank test, P < 0.001). Thirty‐five hospitals (89.7%) exceeded the 85% occupancy threshold and 29 (74.4%) exceeded the 95% occupancy threshold on at least 20% of weekdays (Table 2). Across all the hospitals, the median difference in weekly maximum and weekly mean occupancy was 6.6% points (IQR: 6.2%7.4% points) (Figure 2).

Figure 1

Figure 2

Smoothing reduced the number of hospitals at each occupancy threshold, except 85% (Table 2). As a linear relationship, the reduction in weekday peak occupancy (y) based on a hospital's median difference in weekly maximum and weekly mean occupancy (x) was y = 2.69 + 0.48x. Thus, a hospital with a 10% point difference between weekday and weekend occupancy could reduce weekday peak by 7.5% points.

Smoothing increased the number of patients exposed to the lower thresholds (85% and 90%), but decreased the number of patients exposed to >95% occupancy (Table 2). For example, smoothing at the 95% threshold resulted in 630 fewer patients per hospital exposed to that threshold. If all 39 hospitals had within‐week smoothing, a net of 39,607 patients would have been protected from exposure to >95% occupancy and a net of 50,079 patients from 100% occupancy.

To demonstrate the varied effects of smoothing, Table 3 and Figure 3 present representative categories of response to smoothing depending on pre‐smoothing patterns. While not all hospitals decreased occupancy to below thresholds after smoothing (Types B and D), the overall occupancy was reduced and fewer patients were exposed to extreme levels of high occupancy (eg, >100%).

Figure 3

To achieve within‐week smoothing, a median of 7.4 patient‐admissions per week (range: 2.314.4) would have to be scheduled on a different day of the week. This equates to a median of 2.6% (IQR: 2.25%, 2.99%; range: 0.02%9.2%) of all admissionsor 9% of a typical hospital‐week's scheduled admissions.

Discussion

This analysis of 39 children's hospitals found high levels of occupancy and weekend occupancy lower than weekday occupancy (median difference: 8.2% points). Only 12.4% of scheduled admissions entered on weekends. Thus, weekend capacity is available to offset high weekday occupancy. Hospitals at the higher end of the occupancy thresholds (95%, 100%) would reduce the number of days operating at very high occupancy and the number of patients exposed to such levels by smoothing. This change is mathematically feasible, as a median of 7.4 patients would have to be proactively scheduled differently each week, just under one‐tenth of scheduled admissions. Since LOS by day of admission was the same (median: two days), the opportunity to affect occupancy by shifting patients should be relatively similar for all days of the week. In addition, these admissions were short, conferring greater flexibility. Implementing smoothing over the course of the week does not necessarily require admitting patients on weekends. For example, Monday admissions with an anticipated three‐day LOS could enter on Friday with anticipated discharge on Monday to alleviate midweek crowding and take advantage of unoccupied weekend beds. 26

At the highest levels of occupancy, smoothing reduces the frequency of reaching these maximum levels, but can have the effect of actually exposing more patient‐days to a higher occupancy. For example, for nine hospitals in our analysis with >20% of days over 100%, smoothing decreased days over 100%, but exposed weekend patients to higher levels of occupancy (Figure 3). Since most admissions are short and most scheduled admissions currently occur on weekdays, the number of individual patients (not patient‐days) newly exposed to such high occupancy may not increase much after smoothing at these facilities. Regardless, hospitals with such a pattern may not be able to rely solely on smoothing to avoid weekday crowding, and, if they are operating efficiently in terms of SLOSR, might be justified in building more capacity.

Consistent with our findings, the Institute for Healthcare Improvement, the Institute for Healthcare Optimization, and the American Hospital Association Quality Center stress that addressing artificial variability of scheduled admissions is a critical first step to improving patient flow and quality of care while reducing costs. 18, 21, 27 Our study suggests that small numbers of patients need to be proactively scheduled differently to decrease midweek peak occupancy, so only a small proportion of families would need to find this desirable to make it attractive for hospitals and patients. This type of proactive smoothing decreases peak occupancy on weekdays, reducing the safety risks associated with high occupancy, improving acute access for emergent patients, shortening wait‐times and loss of scheduled patients to another facility, and increasing procedure volume (3%74% in one study). 28 Smoothing may also increase quality and safety on weekends, as emergent patients admitted on weekends experience more delays in necessary treatment and have worse outcomes. 2932 In addition, increasing scheduled admissions to span weekends may appeal to some families wishing to avoid absence from work to be with their hospitalized child, to parents concerned about school performanceand may also appeal to staff members seeking flexible schedules. Increasing weekend hospital capacity is safe, feasible, and economical, even when considering the increased wages for weekend work. 33, 34 Finally, smoothing over the whole week allows fixed costs (eg, surgical suites, imaging equipment) to be allocated over 7 days rather than 5, and allows for better matching of revenue to the fixed expenses.

Rather than a prescriptive approach, our work suggests hospitals need to identify only a small number of patients to proactively shift, providing them opportunities to adapt the approach to local circumstances. The particular patients to move around may also depend on the costs and benefits of services (eg, radiologic, laboratory, operative) and the hospital's existing patterns of staffing. A number of hospitals that have engaged in similar work have achieved sustainable results, such as Seattle Children's Hospital, Boston Medical Center, St. John's Regional Health Center, and New York University Langone Medical Center. 19, 26, 3537 In these cases, proactive smoothing took advantage of unused capacity and decreased crowding on days that had been traditionally very full. Hospitals that rarely or never have high‐occupancy days, and that do not expect growth in volume, may not need to employ smoothing, whereas others that have crowding issues primarily in the winter may wish to implement smoothing techniques seasonally.

Aside from attempting to reduce high‐occupancy through modification of admission patterns, other proactive approaches include optimizing staffing and processes around care, improving efficiency of care, and building additional beds. 16, 25, 38, 39 However, the expense of construction and the scarcity of capital often preclude this last option. Among children's hospitals, with SLOSR close to one, implementing strategies to reduce the LOS during periods of high occupancy may not result in meaningful reductions in LOS, as such approaches would only decrease the typical child's hospitalization by hours, not days. In addition to proactive strategies, hospitals also rely on reactive approaches, such as ED boarding, placing patients in hallways on units, diverting ambulances or transfers, or canceling scheduled admissions at the last moment, to decrease crowding. 16, 39, 40

This study has several limitations. First, use of administrative data precluded modeling all responses. For example, some hospitals may be better able to accommodate fluctuations in census or high occupancy without compromising quality or access. Second, we only considered intra‐week smoothing, but hospitals may benefit from smoothing over longer periods of time, especially since children's hospitals are busier in winter months, but incoming scheduled volume is often not reduced. 11 Hospitals with large occupancy variations across months may want to consider broadening the time horizon for smoothing, and weigh the costs and benefits over that period of time, including parental and clinician concerns and preferences for not delaying treatment. At the individual hospital level, discrete‐event simulation would likely be useful to consider the trade‐offs of smoothing to different levels and over different periods of time. Third, we assumed a fixed number of beds for the year, an approach that may not accurately reflect actual available beds on specific days. This limitation was minimized by counting all beds for each hospital as available for all the days of the year, so that hospitals with a high census when all available beds are included would have an even higher percent occupancy if some of those beds were not actually open. In a related way, then, we also do not consider how staffing may need to be altered or augmented to care for additional patients on certain days. Fourth, midnight census, the only universally available measure, was used to determine occupancy rather than peak census. Midnight census provides a standard snapshot, but is lower than mid‐day peak census. 41 In order to account for these limitations, we considered several different thresholds of high occupancy. Fifth, we smoothed at the hospital level, but differential effects may exist at the unit level. Sixth, to determine proportion of scheduled admissions, we used HCUP KID proportions on PHIS admissions. Overall, this approach likely overestimated scheduled medical admissions on weekends, thus biasing our result towards the null hypothesis. Finally, only freestanding children's hospitals were included in this study. While this may limit generalizability, the general concept of smoothing occupancy should apply in any setting with substantial and consistent variation.

In summary, our study revealed that children's hospitals often face high midweek occupancy, but also have substantial unused weekend capacity. Hospitals facing challenges with high weekday occupancy could proactively use a smoothing approach to decrease the frequency and severity of high occupancy. Further qualitative evaluation is also warranted around child, family, and staff preferences concerning scheduled admissions, school, and work.