Impact of Admission-Day Crowding on the Length of Stay of Pediatric Hospitalizations
【摘要】 OBJECTIVE. Increased crowding may affect the care that is delivered to hospitalized patients, particularly around the time of admission. There is little information about the impact of admission-day crowding on the outcome of children who are hospitalized with common pediatric conditions.
翘楚医学论文网:专业提供医学论文的写作、修改、翻译、发表
官方网站：https://www.qclunwen.com 服务QQ：55755004
邮箱：55755004@qq.com 咨询电话：18986130979
服务宗旨:原创、专业、诚信、安全可靠
METHODS. A population cohort was constructed of children who were aged 1 to 17 years and were hospitalized in Pennsylvania and New York between April 1, 1996, and June 30, 1998, with 1 of 19 common pediatric conditions (N = 116235). Condition-specific Cox regression and logit models were developed to estimate the effect of admission-day occupancy on 4 outcome measures after controlling for illness severity and site of care: length of stay; 21-day readmission; prolonged stay or a stay longer than the typical, uncomplicated stay for that condition as a measure of care delivered to patients with uncomplicated courses; and conditional length of stay as a measure of care delivered to patients whose stays are prolonged.
RESULTS. For children who were admitted with respiratory disease, increasing admission-day occupancy from 60% to 100% was associated with a 0.25-day increase in the average length of stay. Increased admission-day occupancy above 60% was also associated with higher odds of a prolonged stay but not with a change in 21-day readmission rates or conditional length of stay. For children who were admitted with nonrespiratory conditions, increased admission-day occupancy was not associated with changes to any length-of-stay outcome.
CONCLUSIONS. Increased admission-day occupancy was associated with longer lengths of stay for less complicated respiratory admissions but not for children who were admitted with the most serious conditions. These results suggest that medical professionals, during times of increased workload, first focus their attention on more acutely ill children with a complicated course and thus delay treatment of children who have less complicated courses but require time-consuming management and treatment.
【关键词】 hospital crowding length of stay pediatric hospitalizations
Increased workload or occupancy could adversely affect the care that is delivered to hospitalized patients and patients' ultimate outcome. This threat to efficient care is of particular interest immediately after admission; not only is the illness severity of the patient likely to be highest, but also medical care that is delivered at this time begins the diagnostic and treatment course for the patient.1,2
The literature on this subject, though, offers contradictory information. Several studies have shown an adverse relationship between periods of increased workload and mortality, particularly for newborn infants3,4 and adults with myocardial infarction5 and other emergent surgical conditions.6 Other studies have either failed to replicate these results7,8 or found an association between higher occupancy and improved processes of care.9–11 These contradictory results may occur because of how workload, or occupancy, was measured. Most studies did not measure the actual environment experienced by a patient; instead, these studies used proxy measurements of this environment, such as average monthly occupancy rates, or compared weekend with weekday admissions as a surrogate for high and low workloads per provider, respectively.
These studies also offer little information on pediatric patients. Most studies of crowding focus on adult patients in intensive care settings12–16 or with acute conditions that require emergent interventions.5,6 The primary goal of these studies was to assess the association between crowding and mortality. Most pediatric admissions, though, do not require such intensive care and rarely result in the death of the child. The goal of this study was to determine the association between hospital occupancy on admission, as a measure of increased workload, and the length of stay (LOS) for 19 common pediatric diagnoses. LOS is 1 measure of the care that is provided during a hospitalization, particularly when death is a rare outcome. This study used 4 measures of LOS that we developed in previous work.17–19 We hypothesized that increased occupancy on admission would be associated with longer lengths of stay, particularly for (1) conditions that require complex, ongoing management because of the number of decisions or caregivers involved in care, such as respiratory conditions, and (2) conditions with a high severity of illness, such as diabetic ketoacidosis, sickle cell crises, and bacterial meningitis. Children who are admitted for respiratory illness require ongoing assessment of the respiratory status to wean from therapies such as supplemental oxygen and nebulizer treatments and involve multiple health care providers during the day of the admission, such as nurses, respiratory therapists, and physicians. For this reason, many hospitals and pediatric practices have instituted guidelines to standardize the treatment of conditions such as bronchiolitis.20–22 Children who are admitted for conditions with high severity of illness, such as diabetic ketoacidosis and sickle cell crises, may also be affected by increased occupancy on admission because they may not receive treatments that are needed to stabilize their medical course in a timely manner. To answer this question, we used data on all admissions to Pennsylvania and New York for 19 common pediatric medical conditions.
METHODS
Data Sources and Patient Population
We obtained claims data on all hospital admissions to acute care hospitals for patients who were aged 1 to 17 years in Pennsylvania and New York between April 1, 1996, and June 30, 1998. These data are checked for accuracy and validity by the respective state agencies that oversee data collection.23,24 In both data sets, we linked patient charts to allow for calculation of previous admission and rehospitalization rates. We identified 19 common, unscheduled medical conditions from the principal diagnosis codes (Table 1). Because these conditions were unplanned, a hospital could not have counted on these specific admissions when determining their staffing levels for the following day. Four conditions (viral and bacterial pneumonia, asthma, and croup/bronchitis) were classified as respiratory conditions, with the remaining 15 conditions classified as nonrespiratory conditions. Patients from 323 hospitals were identified in this study. Using data from the American Hospital Association,25 the hospitals during the time of the study had a median total bed size of 253 beds (interquartile range [IQR]: 166–427) and a total nurse-to-bed ratio of 1.1 (IQR: 0.9–1.3). No data were available on specific staffing plans or the number of beds or nurses specifically allocated to pediatric patients at each individual hospital.
TABLE 1 Principal Diagnoses Included in the Analysis (N = 116 235)
Construction of Occupancy Rate Variable
We constructed a hospital's daily occupancy rate, standardized by the hospital's average occupancy in that year, by the following method. We defined the occupancy count, Chd, as the number of pediatric patients residing in hospital h at 12 AM on each day d of the study, excluding newborn nursery and NICU patients. For each hospital h, we sorted the 365 Chd values and found the 18th largest Chd. This value, Ch,5%, was the upper fifth percentile point of occupancy for hospital h. We then defined the occupancy rate for hospital h on day d as ORhd = Chd/Ch,5%, where OR is odds ratio. We used the top fifth percentile occupancy count of that year as the denominator for our occupancy measure because there may also be days of extremely high occupancy that represent an outlier for that hospital; with this occupancy measure, we allowed these outliers to occur without influencing the occupancy rates on the remaining 95% of the days.
Definition of Confounding Variables
To control for expected increases in LOS associated with patient-level characteristics, our models included demographic variables and comorbid conditions listed in Table 2, the median income based on patient zip code,26 distance from the hospital as a proxy for severity of illness,27 and a state variable to account for regional differences in overall LOS.17 There were no significant differences in the results of the study when we included the year of hospitalization in our analyses. A viral illness variable was added to the models because concurrent viral illness may increase the severity of many chronic illnesses, such as asthma.28,29 This variable was coded as "yes" when the patient had 1 of the following secondary diagnoses: 079.6 (respiratory syncytial virus), 464.4 (croup), 466.x (bronchiolitis), 480.x (viral pneumonia), or 487.x (influenza). We also analyzed the occupancy data for children's hospitals separately then for nonchildren's hospitals as in previous work17; children's hospitals consisted of institutions where either (1) at least 90% of patients who were admitted to the hospital were younger than 17 years or (2) the hospital was in the top fifth percentile for pediatric admissions and the hospital had either a pediatric residency program or a primary medical school affiliation.
TABLE 2 Demographic and Comorbidity Data
Outcome Definitions and Modeling
This study analyzed 4 measures of LOS as defined in previous work17–19: (1) LOS in days, (2) readmission within 21 days of discharge, (3) odds of a prolonged stay, and (4) conditional LOS (CLOS) in days. LOS was calculated as the number of days between hospital admission and discharge. It provides insight into the use of resources by a provider and a hospital. Twenty-one-day readmission rates were used to account for potentially inadequate care during the hospitalization. Higher readmission rates could also represent inadequate outpatient care.
Earlier work suggested that a hospital LOS could be divided into 2 periods, on the basis of the daily rate of discharge for that condition.17,19 If we take all patients from the study period with a specific condition and code the first day of each patient's hospitalization as day 1, etc, then the daily rate of discharge is the percentage of hospitalized patients at the start of a given hospitalization day who were discharged on that day. The daily rate of discharge increases during the first few days of a hospitalization, then decreases rapidly.17–19 For each condition, we defined the day when the rate of discharge began to decline as the prolongation point. We determined the prolongation point by applying the "new-worse-than-used" test of Hollander and Proschan to the hospital discharge data for each condition.17,19,30,31 This test defines the technical concept of an "extended discharge" if a patient who has already stayed b days is more likely to stay t more days than a newly admitted patient is to stay t days. We used diagnosis-specific prolongation points reported in Table 1 in the 2 secondary outcome measures. First, we calculated the odds of a prolonged stay, defined as an admission lasting beyond the prolongation point. This outcome measure describes the ability of providers to treat and discharge quickly the patient with a less complicated course,17,19 where higher ORs suggest slower care. CLOS was defined as the number of days a patient remained hospitalized beyond the prolongation point. CLOS describes the provider's ability to treat patients who had complicated courses and stayed beyond the prolongation point.17–19
Model Development and Presentation of Data
We developed Cox regression models to estimate the effects of occupancy rate on LOS and CLOS, reporting a hazard ratio (HR) for discharge. A discharge HR of <1 means that patients stayed longer when occupancy was higher. We developed logit models to estimate the effects of occupancy rate on prolonged stay and readmissions. Stata 9.2 (Stata Corp, College Station, TX) was used for all analyses. We assigned the outcome of a given hospitalized child to the initial admission hospital regardless of transfers after the admission. Deaths were modeled as the longest stay in the data set in the Cox regressions and as prolonged stays in the logit models regardless of the day on which the death occurred. Modeling the few deaths on the basis of the day when the child died changed all results by <1%. A set of indicator variables for individual hospitals were included in the analyses to control for fixed differences between hospitals in admission and discharge practices.17–19 By including these fixed effects, we compared each hospital with itself as occupancy fluctuates, rather than comparing crowded hospitals with uncrowded ones.
Previous work suggested that the effect of hospital occupancy changes once a specific threshold is reached within a given hospital.32 For each outcome, we determined the occupancy threshold associated with the best prediction of the outcome variable and compared this model with a model without a threshold variable. Our final analyses included the threshold variable when the log-likelihood test between these 2 models was significant at the P < .05 level after correction for multiple testing with the Bonferroni method.
RESULTS
We present the results of these analyses in 2 ways. First, we present either HRs or ORs in tabular form. These ratios show the impact of a 10% change in hospital occupancy on 1 of the 4 outcome measures; however, we typically think about changes to a patient's LOS as a change in the number of hospitalized days. We also typically think about dichotomous variables as probabilities rather than odds. Thus, we also present the study results in graphic form for the average child either with or without respiratory disease in our data set. Thus, for LOS and CLOS, we constructed a predicted hazard curve for each average patient and report the LOS or CLOS as a predicted number of days at a given admission-day occupancy. For 21-day readmission rates and prolonged stay, we predicted the probability of each outcome at a given admission day occupancy.
We identified 67029 respiratory admissions and 49206 nonrespiratory medical admissions in Pennsylvania and New York during our study period. Table 1 shows the number of patients who were admitted with each of the 19 medical conditions and the condition-specific prolongation point used in the prolonged and CLOS outcome measures. Relevant demographic and comorbidity data are shown in Table 2.
Hospital Occupancy During Study Period
Average daily occupancy rates varied within a given week, because children were more likely to be admitted for routine procedures during weekdays (Fig 1). Many smaller hospitals had days on which their occupancy rate was 0. Throughout the year, hospital occupancy was generally increased during the winter and early spring months and dropped around Christmas day and between June and September. These fluctuations mirrored changes in the percentage of patients who were admitted with a secondary diagnosis of viral illness.
FIGURE 1 Average admission-day occupancy and percentage of hospitalized children with a secondary diagnosis of viral illness during the time span of the study.
For this study, the median occupancy rate that children experienced on admission was 75.0% with an IQR of 58.8% to 87.5%. Admission-day occupancy rates for children who were admitted with respiratory conditions were slightly higher than for children who were admitted with nonrespiratory conditions (median: 75.0% vs 73.8%; P < .001 by rank-sum test). Overall, 21.2% of the children identified in this study were admitted when the occupancy rate was 90%, and 11.4% of the children were admitted when the occupancy rate was 100%. The percentage of respiratory and nonrespiratory admissions that occurred when the occupancy rates were at least 90% or 100% was similar.
Admission-Day Occupancy and LOS
We first examined the impact of increased admission-day occupancy on the LOS of children who were admitted with 1 of 4 respiratory conditions (Table 3). After adjustment for case severity and demographic characteristics, increased admission-day occupancy above 60% was associated with longer LOS (HR per 10% increase in occupancy: 0.978; 95% confidence interval: 0.973–0.983) without changing the odds of a readmission within 21 days of discharge. The effect of admission-day occupancy was not uniform; as shown in Fig 2, the increase in LOS was greatest above a threshold of 60% occupancy. When the admission-day occupancy increased from 60% to 100%, our models predicted an extra 25 hospital days per 100 typical children admitted with respiratory conditions. Similarly, when the admission-day occupancy increased from 59% to 88.5%, which was the IQR of admission-day occupancies for children who were admitted with respiratory conditions, our models predicted an extra 16 hospital days per 100 typical respiratory admissions.
TABLE 3 Association of Admission-Day Occupancy and LOS
FIGURE 2 Predicted LOS (A), CLOS (B), and probability of a prolonged stay (C) or 21-day readmission (D) for a specific admission-day occupancy. Children who were admitted with a respiratory disease (solid line) experienced longer LOS and higher probabilities of a prolonged stay as admission-day occupancy increased above 60%. The outcome of a nonrespiratory admission (dashed line) did not change substantially as admission-day occupancy increased. The influence of admission-day occupancy on the outcome of all 19 conditions included in the study is shown with the dotted line.
We next examined 2 additional LOS outcome measures: prolonged stay and CLOS. Increased admission-day occupancy above 60% was associated with increased odds of a prolonged stay (OR: 1.045 per 10% increase in admission-day occupancy, 95% confidence interval: 1.033–1.057) but did not predict a change in CLOS. As with LOS, the effect of admission-day occupancy on the prolonged stay outcome was not uniform, because there was a significant threshold effect above an occupancy rate of 60%. On the basis of these results, our models predicted a 16% increase in the probability of a prolonged stay for an average child who was admitted with respiratory disease when the admission-day occupancy increased from 60% to 100% (Fig 2). These findings suggest that the longer LOS associated with increased admission-day occupancy often occurred for children with uncomplicated courses, who otherwise may be discharged quickly from the hospital.
For nonrespiratory conditions, increases in admission-day occupancy were not associated with changes in any of the 4 LOS measures used in this study (Table 3). No threshold effects were detected. Fig 2 shows the relatively flat predicted change in the 4 outcome measures for nonrespiratory hospitalizations as admission-day occupancy increased.
Subgroup Analyses: Increased Occupancy and Individual Medical Conditions
Compared with other hospitals in Pennsylvania and New York, children's hospitals displayed the same associations between increased admission-day occupancy and measures of LOS. Table 4 shows the impact of admission-day occupancy on the outcomes of each medical condition separately. Increased admission-day occupancy was associated with a longer LOS and higher odds of a prolonged stay for 3 of the 4 respiratory conditions, with croup/bronchitis admissions showing the greatest increases. In contrast to the respiratory condition, the LOS of most nonrespiratory conditions was not significantly changed by increased admission-day outcome on the basis of our study outcomes. Our analyses also failed to detect an association between increased admission-day occupancy and longer LOS outcomes for nonrespiratory conditions that typically are associated with increased severity of illness, such as bacterial meningitis or sickle cell crises. For diabetic ketoacidosis, increased admission-day occupancy was associated with shorter LOS and lower odds of a prolonged stay.
TABLE 4 Influence of Increasing Admission-Day Occupancy on the Outcome of Individual Conditions
DISCUSSION
Improving the efficiency of care that is delivered to hospitalized patients remains an important goal of hospital administrators and policy makers, who strive to eliminate unnecessary health care costs, and families, who desire adequate care without unnecessary disruptions to their lives. Increased hospital occupancy, as a measure of the overall workload experienced by the hospital, is a potential threat to these goals. This study found different effects of increased occupancy for 2 groups of at-risk children. For children who were admitted with common respiratory conditions, increased occupancy on admission was associated with longer LOS. When the occupancy increased from 60% to 100%, the average increase in LOS translated to 25 extra hospital days per 100 typical respiratory patients even after controlling for other factors that may increase the LOS. This increase in LOS seems to be focused on the patients with a less complicated course, because increasing hospital occupancy was associated with a 16% increase in the predicted probability of a prolonged stay—a measure of the care delivered to children without significant complications during their hospitalization—whereas no change was observed in the CLOS statistic, a measure based only on the children who had a more complicated course and whose course was already prolonged. The outcome of children who were admitted with nonrespiratory conditions was not significantly affected by admission-day occupancy, even for conditions that typically are associated with the highest severity of illness, such as diabetic ketoacidosis and bacterial meningitis.
Our study suggests that the increase in LOS preferentially affects children who have a less complicated course and require complex management and treatment, not children who are admitted with high-severity conditions. One potential explanation is that medical professionals, during times of increased workload, first focus their attention on the more acutely ill children with a complicated course and thus delay treatment of the children with less complex courses, who make up the majority of pediatric hospitalizations, particularly respiratory admissions. This relationship may not have been noted by other studies that focused on sicker patients who were admitted to intensive care settings.5–7,13–16 Beyond being the most common reason for admitting children to the hospital, children with respiratory conditions require specific management once admitted. These children require ongoing assessment of both their respiratory status and their ability to wean from therapies such as supplemental oxygen and nebulizer treatments. Their care also involves multiple health care providers, such as nurses, respiratory therapists, and physicians, throughout admission. Conditions that require such complex coordination of multiple treatment plans may be more vulnerable to admission-day occupancy than other conditions for which the initial assessment of the child allows for the implementation of a treatment plan that can carry on with less oversight and adjustment. For example, the increase in LOS identified in this study could arise from delayed administration and weaning of bronchodilators and corticosteroids for children with asthma or delayed weaning of supplemental oxygen. For this reason, many institutions have implemented guidelines to standardize the treatment that is received by children with common respiratory conditions, such as bronchiolitis.20–22 Additional work is needed to understand both the coping methods used by hospitals during times of increased occupancy and processes such as treatment pathways and guidelines that may facilitate the care of less severely ill but equally time-consuming patients.
This study used actual occupancy data on the day of admission for each child, which was standardized by the observed occupancy at each hospital for that year. This value provided a more accurate measure of hospital occupancy and workload than using monthly or yearly occupancy,9–11 which may reflect some other factor, such as hospital volume or the day of the week of the admis-sion.4–6,8,13,15,16,33 This method of calculating the actual hospital occupancy on the day of a patient's admission worked well for this study question; however, this method cannot determine how changing hospital occupancy during a hospitalization influences the decision to discharge a patient, because we cannot determine which factor—the change to hospital occupancy or the decision to discharge a patient on that day—came first. Answering this interesting but different question would require other techniques to eliminate the endogeneity bias between daily hospital occupancy and the decision to discharge a patient.
The detailed daily data that were needed to construct hospital occupancy were available from Pennsylvania and New York. Because these 2 states have many different types of hospitals, rural and urban communities, and large numbers of pediatric admissions, we believe that these results reflect the way hospitals operate in the United States. We were unable to tease out the effect of specific policies that some hospitals may use during periods of high occupancy, such as the use of float nurses, because these data and their actual implementation on a specific day were not available; however, this inability to control for specific hospital staffing patterns on a given day biases our results toward the null hypothesis; our results demonstrate a persistent impact of admission-day occupancy after these hospital-level adjustments have been made. Finally, this study used International Classification of Diseases, Ninth Revision, Clinical Modification codes to identify groups of hospitalized children and define medical conditions to control for illness severity on admission, as in previous studies.17–19 Whereas studies in adult medicine use comorbidity indices for this purpose,34–36 most children are hospitalized with acute, self-limited illnesses that do not lend themselves to these indices. Our results were seen in hospitals of all types and geographic locations, suggesting that the types of patients seen by a hospital did not influence the association between admission-day occupancy and LOS. Also, this study detected a significant effect at an occupancy level of 60%. It is unlikely that admission patterns in a community or at a specific hospital would change in response to this relatively low occupancy rate or that increased occupancy rates would affect the admission of children only with a few medical conditions. Finally, our data suggest that the children who are most affected by increased admission-day occupancy are the less ill children with an uncomplicated course. Thus, although these results do not rule out the possibility that children are sicker in general during periods of increased occupancy, they do suggest that other factors besides illness severity are contributing to these observations.
The results of this study, then, suggest that increased occupancy at the time of admission is associated with longer lengths of stay for less complicated pediatric admissions and for children who are admitted with respiratory disease. Increased occupancy was not associated with changes to the LOS for children who were admitted with the most serious conditions, such as bacterial meningitis and diabetic ketoacidosis. These results suggest that medical professionals, during times of increased workload, first focus their attention on the more acutely ill children with a complicated course and thus delay treatment of the children with less complex courses, who make up the majority of pediatric hospitalizations, particularly respiratory admissions. Policies that improve the care that is received by children who are admitted during times of high occupancy with less severe conditions that require complex management plans and primary prevention of the causes for increased occupancy, such as the use of influenza vaccines for young children37 and immunoprophylaxis against respiratory syncytial virus in high-risk infants,38 could improve the efficiency of delivered care within a specific hospital and parental satisfaction with the health care system.
APPENDIX Complete Regression Models for Table 3
【参考文献】
Rivers E, Nguyen B, Havstad S, et al. Early goal-directed therapy in the treatment of severe sepsis and septic shock.N Engl J Med. 2001;345(19):1368–1377
Ioachimescu OC, Ioachimescu AG, Iannini PB. Severity scoring in community-acquired pneumonia caused by Streptococcus pneumoniae: a 5-year experience.Int J Antimicrob Agents. 2004;24(5):485–490
Stephansson O, Dickman PW, Johansson AL, Kieler H, Cnattingius S. Time of birth and risk of intrapartum and early neonatal death.Epidemiology. 2003;14(2):218–222
Hamilton P, Restrepo E. Weekend birth and higher neonatal mortality: a problem of patient acuity or quality of care?J Obstet Gynecol Neonatal Nurs. 2003;32(6):724–733
Kostis WJ, Demissie K, Marcella SW, Shao YH, Wilson AC, Moreyra AE. Weekend versus weekday admission and mortality from myocardial infarction.N Engl J Med. 2007;356(11):1099–1109
Bell CM, Redelmeier DA. Mortality among patients admitted to hospitals on weekends as compared with weekdays.N Engl J Med. 2001;345(9):663–668
Abdel-Latif ME, Bajuk B, Oei J, Lui K. Mortality and morbidities among very premature infants admitted after hours in an Australian neonatal intensive care unit network.Pediatrics. 2006;117(5):1632–1639
翘楚医学论文网:专业提供医学论文的写作、修改、翻译、发表
官方网站：https://www.qclunwen.com 服务QQ：55755004
邮箱：55755004@qq.com 咨询电话：18986130979
服务宗旨:原创、专业、诚信、安全可靠
Gould JB, Qin C, Marks AR, Chavez G. Neonatal mortality in weekend vs weekday births.JAMA. 2003;289(22):2958–2962
Hartz AJ, Krakauer H, Kuhn EM, et al. Hospital characteristics and mortality rates.N Engl J Med. 1989;321(25):1720–1725
Kuhn EM, Hartz AJ, Gottlieb MS, Rimm AA. The relationship of hospital characteristics and the results of peer review in six large states.Med Care. 1991;29(10):1028–1038
Boscarino JA. Patients' perception of quality hospital care and hospital occupancy: are there biases associated with assessing quality care based on patients' perceptions?Int J Qual Health Care. 1996;8(5):467–477
Goldfrad C, Rowan K. Consequences of discharges from intensive care at night.Lancet. 2000;355(9210):1138–1142
Ensminger SA, Morales IJ, Peters SG, et al. The hospital mortality of patients admitted to the ICU on weekends.Chest. 2004;126(4):1292–1298
Luyt CE, Combes A, Aegerter P, et al. Mortality among patients admitted to intensive care units during weekday day shifts compared with "off" hours.Crit Care Med. 2007;35(1):3–11
Arabi Y, Alshimemeri A, Taher S. Weekend and weeknight admissions have the same outcome of weekday admissions to an intensive care unit with onsite intensivist coverage.Crit Care Med. 2006;34(3):605–611
Hixson ED, Davis S, Morris S, Harrison AM. Do weekends or evenings matter in a pediatric intensive care unit?Pediatr Crit Care Med. 2005;6(5):523–530
Silber JH, Rosenbaum PR, Even-Shoshan O, et al. Length of stay, conditional length of stay, and prolonged stay in pediatric asthma.Health Serv Res. 2003;38(3):867–886
Lorch SA, Zhang X, Rosenbaum PR, Evan-Shoshan O, Silber JH. Equivalent length of stay in pediatric patients hospitalized in rural and non-rural hospitals.Pediatrics. 2004;114(4). Available at: www.pediatrics.org/cgi/content/full/114/4/e400
Silber JH, Rosenbaum PR, Koziol LF, Sutaria N, Marsh RR, Even-Shoshan O. Conditional length of stay.Health Serv Res. 1999;34(1 pt 2):349–363
Kwan-Gett TS, Lozano P, Mullin K, Marcuse EK. One-year experience with an inpatient asthma clinical pathway.Arch Pediatr Adolesc Med. 1997;151(7):684–689
Todd J, Bertoch D, Dolan S. Use of a large national database for comparative evaluation of the effect of a bronchiolitis/viral pneumonia clinical care guideline on patient outcome and resource utilization.Arch Pediatr Adolesc Med. 2002;156(11):1086–1090
Harrison AM, Boeing NM, Domachowske JB, Piedmonte MR, Kanter RK. Effect of RSV bronchiolitis practice guidelines on resource utilization.Clin Pediatr (Phila). 2001;40(9):489–495
Pennsylvania Health Care Cost Containment Council. Hospital Performance Report: 49 Common Medical Procedures and Treatments; 2006. Available at: www.phc4.org/reports/hpr/05/docs/hpr2005technotes.pdf. Accessed July 19, 2007
New York State Department of Health. 2002 Annual Report. The SPARCS Data System. Available at: www.health.state.ny.us/statistics/sparcs/sysdoc/operguid.htm. Accessed February 4, 2008
American Hospital Association. Annual Survey of Hospitals Database, 1996. Available at: www.associationstores.org/OA_HTML/ibeCCtpItmDspRte.jsp?a=b&item=3086. Accessed February 4, 2008
Goodman DC, Stukel TA, Chang CH. Trends in pediatric asthma hospitalization rates: regional and socioeconomic differences.Pediatrics. 1998;101(2):208–213
Luft HS, Garnick DW, Mark DH, et al. Does quality influence choice of hospital?JAMA. 1990;263(21):2899–2906
Johnston SL, Pattemore PK, Sanderson G, et al. Community study of role of viral infections in exacerbations of asthma in 9–11 year old children.BMJ. 1995;310(6989):1225–1229
Gergen PJ, Mitchell H, Lynn H. Understanding the seasonal pattern of childhood asthma: results from the National Cooperative Inner-City Asthma Study (NCICAS).J Pediatr. 2002;141(5):631–636
Hollander M, Proschan F. Testing whether new is better than used.Ann Mathematical Stat. 1972;43(4):1136–1146
Hollander M, Wolfe DA.Nonparametric Statistical Methods. 2nd ed. New York, NY: John Wiley & Sons; 1999:504–513
Bagust A, Place M, Posnett JW. Dynamics of bed use in accommodating emergency admissions: stochastic simulation model.BMJ. 1999;319(7203):155–158
Tarnow-Mordi WO, Hau C, Warden A, Shearer AJ. Hospital mortality in relation to staff workload: a 4-year study in an adult intensive-care unit.Lancet. 2000;356(9225):185–189
Elixhauser A, Steiner C, Harris DR, Coffey RM. Comorbidity measures for use with administrative data.Med Care. 1998;36(1):8–27
Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation.J Chron Dis. 1987;40(5):373–383
Deyo RA, Cherkin DC, Ciol MA. Adapting a clinical comorbidity index for use with ICD-9-CM administrative databases.J Clin Epidemiol. 1992;45(6):613–619
American Academy of Pediatrics Committee on Infectious Diseases. Prevention of influenza: recommendations for influenza immunization of children, 2006–2007.Pediatrics. 2007;119(4):846–851
American Academy of Pediatrics Committee on Infectious Diseases and Committee on Fetus and Newborn. Revised indications for the use of palivizumab and respiratory syncytial virus immune globulin intravenous for the prevention of respiratory syncytial virus infections.Pediatrics. 2003;112(6 pt 1):1442–1446