Objective. Pediatric teaching hospitals provide particular expertise in caring for children with complex or severe illnesses, yet most patients within teaching hospitals have common pediatric conditions. No study has determined whether children with common conditions remain hospitalized at teaching institutions longer than at nonteaching institutions. The objective of this study was to compare length of stay (LOS) for common pediatric conditions between teaching and nonteaching hospitals.
Methods. This study uses Massachusetts’s hospital data for all discharged children ages 0 to 17 years for 1995 and 1996. Discharges were included when the principal diagnosis indicated asthma, bacterial pneumonia, convulsions, dehydration, failure to thrive, gastroenteritis, or urinary tract infections. Hospitals were classified as either teaching or nonteaching using the 1995–1996 American Hospital Association Guide. Children were identified as having a chronic condition when any discharge diagnosis was 1 of those on a previously published catalog of chronic childhood illnesses. The analysis tested the association of hospital type with LOS, controlling for chronic conditions, insurance type, age, race, diagnosis, mortality, and disposition using multivariate linear regression.
Results. Of 17 890 discharges for a common pediatric condition during the study period, 52.3% were from teaching hospitals. Twelve percent of common condition discharges also had a chronic disease diagnosis; 75.1% of these were discharged from a teaching hospital. LOS from nonteaching hospitals was shorter than from teaching hospitals (2.42 days vs 3.20 days). Although LOS for stays with a chronic diagnosis were longer than those without (4.75 days vs 2.56 days), controlling for chronic illness and other covariates did not eliminate the difference between LOS for nonteaching hospitals versus teaching hospitals (1.65 days vs 2.23 days).
Conclusion. Pediatric patients with common conditions have a shorter LOS in nonteaching hospitals than those admitted to teaching hospitals by a little more than half a day. These results are unchanged when accounting for chronic conditions despite the expected results of preferential admissions to teaching hospitals for this group of patients. Additional studies should better characterize differences in patient populations, describe differences in processes, and identify differences in patient experience and outcomes to understand better the potential benefits of treating children with specific conditions at particular types of hospitals.
Pediatric teaching hospitals provide particular expertise in caring for children with complex or severe illnesses,1 yet most patients within teaching hospitals have common pediatric conditions. Whether teaching hospitals compared with nonteaching hospitals are more or less efficient in the care of these children is unknown.
Costs are higher in teaching hospitals that care for adult patients compared with costs for adults in community hospitals.2 Some of the higher costs are associated with differences in case mix, but not all.2 These costs remain higher for similar patients in teaching and nonteaching hospitals.3
Legitimate reasons exist for costs of care to be higher in teaching hospitals. Teaching programs typically experience a greater concentration of children with chronic conditions, who are appropriately more expensive to care for in general.4 Costs in academic settings may also be higher because of the necessary expense of providing supervision to students and residents. Less easy to justify, however, is inefficiency in the routine processes of care. Such inefficiencies might result in extending the length of stay (LOS) of children with comparable levels of illness in academic versus nonacademic centers.
In pediatrics, both teaching and nonteaching hospitals see a high volume of common conditions. The effect of the presence or absence of a chronic condition on resource utilization for a group of children hospitalized with common pediatric conditions has been studied; these reports demonstrate that costs increase—not surprisingly—when caring for a child who has a chronic condition, even if hospitalized for what seems to be a common problem (eg, pneumonia).5 Another study examined how LOS for children with asthma varies by hospital type. The study found no difference in LOS for nonteaching hospitals versus a single tertiary care children’s hospital in 1 county in the state of Washington.6
In our study, we included every hospital in the state of Massachusetts that cared for children, and we also looked at a group of common conditions to determine whether there was any variation in LOS between hospital types, holding constant the presence or absence of a chronic condition. We wanted to compare LOS of hospitalization, as a proxy for efficiency, for common pediatric conditions between teaching and nonteaching hospitals.
This study is a cross-sectional analysis using Massachusetts’ hospital discharge data for all discharged children for 1995 and 1996. The study population consisted of all children who were between 0 and 17 years of age and discharged from a Massachusetts hospital with 1 of the following 7 principal diagnoses: asthma, bacterial pneumonia, convulsions, dehydration, failure to thrive, gastroenteritis, or urinary tract infections. Discharges of children with these diagnoses were identified through a list of International Classification of Diseases, Ninth Revision codes defined by the Massachusetts Division of Health Care, Finance and Policy as Ambulatory Sensitive Conditions7 (Table 1 ).
Hospitals were classified as either teaching or nonteaching using the American Hospital Association Guide for 1995–1996.8 Teaching hospitals had to meet criteria as defined by the American Association of Medical Colleges. The remaining hospitals were classified as nonteaching.
Children who were hospitalized for an acute condition were considered also to have a chronic condition when any discharge diagnosis other than the principal was 1 of those on a previously published catalog of chronic childhood illnesses.9–11 Children who were transferred from other hospitals were excluded from the analyses as they are preferentially admitted to teaching hospitals. Age was classified into children younger than 2 years, 2 to 5 years, 5 to 10 years, 10 to 15 years, and 15 to 18 years.12 Insurance status was divided into those with public, private, or uninsured patients. Race was trichotomized into white, black, and other.
χ2 tests were used to test the relationship of the nominal predictor variables (eg, insurance status) to type of hospital (teaching or nonteaching) for univariate analyses. Wilcoxon rank sum tests were used to test the relationship for continuous variables to type of hospital.
Variables that were significantly associated with mean LOS at P < .10 were included in a multivariate linear regression model. The continuous predictor variables were analyzed with Spearman rank correlations to the outcome variable (mean LOS), as a result of the nonnormal distribution of the outcome. The dichotomous and nominal predictor variables (with >2 categories) were analyzed to the nonnormal outcome variable with the Wilcoxon rank sum and Kruskal-Wallis tests, respectively. The analysis tested the association of the main independent variable of interest—hospital type—with the dependent variable—mean LOS—controlling for chronic conditions, insurance type, age, race, principal diagnosis, mortality, source of admission, and type of admission using multivariate linear regression.
The analyses were repeated for LOS truncated by 3 standard deviations, LOS log transformed, and with the removal of all patients with chronic conditions. When the log of LOS was examined, parametric tests were used, because the distribution of the outcome variable was now normal. The data were also analyzed treating the hospitals as a random effect (instead of a fixed effect by standard linear regression) using mixed model regression to ensure the validity of the first model. All statistical analyses were performed using Statistical Analysis Software (version 8.2; SAS Institute, Cary, NC).
A total of 81 of 126 hospitals in the state of Massachusetts discharged at least 1 child during the years 1995 and 1996. Fifteen (18.5%) hospitals were identified as teaching.
There were 17 890 discharges for the identified hospitals in Massachusetts with a principal diagnosis of 1 of the 7 pediatric conditions during the years 1995 and 1996. A total of 9349 (52.3%) were from teaching hospitals. A total of 2146 (12%) common condition discharges also had a chronic disease diagnosis; as anticipated, a high proportion of these (75.1%) were discharged from a teaching hospital. Asthma accounted for the most number of patients discharged from a hospital in Massachusetts for children, 6012 patients (34%; Table 1).
Teaching hospitals cared for patients who were more likely to have a chronic condition (17% vs 6%), be younger in age (4.5 years vs 5.0 years), be nonwhite (47% vs 31%), have managed care insurance (58% vs 45%), be uninsured (16% vs 14%), and be admitted from the emergency department (76% vs 50%) than nonteaching hospitals (P < .001; Table 2 ).
LOS from nonteaching hospitals was shorter than from teaching hospitals (2.42 days vs 3.20 days, unadjusted). Although LOS for stays with a chronic diagnosis were longer than those without (4.75 days vs 2.56 days, unadjusted), controlling for chronic illness and other covariates narrowed but did not eliminate the difference between LOS for nonteaching hospitals versus teaching hospitals (1.65 days vs 2.23; P < .001; Table 3 ). The results were unchanged when repeating analyses with truncated LOS to within 3 standard deviations, log transforming LOS, or removing all patients with chronic conditions (although the last analyses reduced the explanatory power of the model by 50%), and using mixed model regression (teaching hospital effect of 0.73 days; P = 0.0008).
We found that of those pediatric patients with 1 of 7 acute conditions discharged from a nonteaching hospital in Massachusetts in 1995 and 1996, there was a significantly shorter LOS of 0.59 days compared with a teaching hospital, when adjusting for patients with chronic conditions (P < .001). The major limitation using this database is the difficulty accounting for severity of illness. The literature suggests various methods to try to account for severity of illness using databases, although there is no standard approach.13–18 We attempted to account for severity of illness by making the populations similar by adjusting for patients with chronic conditions.6,19 Although this addresses the issue of preferential admissions to teaching hospitals for these patients, the issue of severity of illness for the same disease in similar patients remains. This database did not have sufficient information to allow additional severity adjustments.
There may be nonrandom selection bias in how physicians who care for pediatric patients with 1 of these acute conditions decide where to admit these patients, aside from severity of illness. Social factors and convenience to both the family and the physician may affect which type of hospital to which patients are admitted. In our classification of hospitals based on the American Hospital Association classification guide, there is a degree of heterogeneity among teaching hospitals that may be diluting the true effect of differences in LOS.
To define our population, we chose those conditions that were designated ambulatory sensitive conditions defined by Massachusetts Division of Health Care, Finance and Policy. We chose these conditions on the basis of the recent publication on International Classification of Diseases, Ninth Revision codes by Massachusetts, the relation of our research to potential policy implications using similar methods, and the recognition of the conditions being common and having relatively equal admissions across the hospitals in the state. Although recognizing that ambulatory sensitive conditions have been used as a marker of adequate primary care, with a potential for reducing the need for hospital admission, we did not draw conclusions about access to primary care for the geographic locations of the teaching and nonteaching hospitals, as this was not the primary focus of our study.
Small area variation in health care has been documented for admissions to a hospital for both adults and pediatrics.20,21 Hospitalization rates also vary for “discretionary” conditions (ie, conditions uniformly requiring hospitalization, eg, appendicitis, bacterial meningitis). Many pediatric conditions are discretionary, and the rates of hospitalization can vary 3-fold.22 Because of the nature of the analytic approach used in this study, we were unable to account for underlying hospitalization rates for children in different regions of the state. However, the relationship between hospitalization rates and severity of illness has been inconsistent across studies.21–23
Despite these limitations, our findings are suggestive of a difference in LOS for pediatric patients who have common conditions and are cared for in either a nonteaching or a teaching hospital. These data do not allow any judgments as to whether the longer LOS for patients with similar conditions (and increased cost) in teaching hospitals may be attributable to increased intensity of services (eg, child life specialists, specialized laboratory technicians, equipment for dealing with children with rare conditions) and produces added value—better outcomes, reduced burden on families—or merely indicates inefficiency and waste.
Pediatric patients with common conditions have a shorter LOS in nonteaching hospitals than those who are admitted to teaching hospitals by a little more than half a day. The difference remains significant when accounting for chronic conditions despite the expected results of preferential admissions to teaching hospitals for this group of patients. Additional studies should use more robust case-mix approaches and analyze patients at a level more in depth (including surveys and chart reviews) to include differences in processes of care and differences in outcomes and to identify differences in patient experience and outcomes to understand better the benefits of treating particular conditions at particular types of hospitals.
This study was supported by grant T32 PE10018 from the Health Resources and Services Administration (Dr Srivastava) to the Harvard Pediatric Health Services Research Fellowship Program.
- Received March 15, 2002.
- Accepted December 3, 2002.
- Reprint requests to (R.S.) University of Utah Health Sciences Center, Department of Pediatrics, 100 North Medical Dr, MAPS, Salt Lake City, UT 84113. E-mail:
This work was presented in part at the Ambulatory Pediatric Association Meeting; May 2000; Boston, MA.
- ↵Byrke CR, Tunnessen WW Jr, Scully TJ, Oski FA. Pediatric residencies: differences between 1959/1960 and 1984/1985. Pediatrics.1988;82 :752– 755
- ↵Dosa NP, Boeing NM, Ms N, Kanter RK. Excess risk of severe acute illness in children with chronic health conditions. Pediatrics.2001;107 :499– 504
- ↵Samuels BN, Novack AH, Martin DP, Connell FA. Comparison of length of stay for asthma by hospital type. Pediatrics. 1999;101(4). Available at: http://www.pediatrics.org/cgi/content/full/101/4/e13
- ↵Division of Health Care Finance and Policy. Preventable Hospitalization in Massachusetts. Update for Fiscal Years 1995 and 1996. Boston, MA: The Division; 1998
- ↵American Hospital Association. American Hospital Association Guide for 1995–1996. Chicago, IL: American Hospital Association; 1995
- ↵Perrin JM, Kuhlthau K, Ettner SL, McLaughlin TJ, Gortmaker SL. Previous Medicaid status of children newly enrolled in Supplemental Security Income. Health Care Financ Rev.1998;19 :117– 127
- ↵Wennberg J, Gittelsohn A. Small area variations in health care delivery: a population based health information system can guide planning and regulatory decision making. Science.1973;182 :1102– 1107
- ↵McConnachie KM, Roghmann KJ, Liptak GS. Socioeconomic variation in discretionary and mandatory hospitalization of infants: an ecologic analysis. Pediatrics.1997;99 :774– 784
- Copyright © 2003 by the American Academy of Pediatrics