Association between Hospital Volume and Care for Gastroenteritis

Acute gastroenteritis is a common condition with established guidelines for treatment that have been shown to improve outcomes and reduce costs, length of hospital stay, and duration of illness. According to researchers, >30% of children admitted to hospitals in the United States receive care inconsistent with the current guidelines; reasons for the low adherence rate are not known. The number of patients admitted to the hospital (patient volume) is a determinant of variation in quality of care: for various common adult conditions, higher patient volumes have been association with lower mortality. In some cases, the volume effect can be explained, in part, by increased adherence to evidence-based guidelines for treatment. Little is known, however, about the association of volume with quality in the case of children hospitalized for common conditions. Researchers recently conducted a retrospective cohort study to examine the relationship between adherence to recommended treatment guidelines and the volume of children admitted to the hospital with acute gastroenteritis. They reported study results in Archives of Pediatrics & Adolescent Medicine [2011;165(9):857-863]. The study cohort comprised 12,604 children ages 3 months to 10 years who were admitted to 280 hospitals in the United States between January 2, 2007, and December 31, 2009. The cohort population was identified using International Classification of Diseases, Ninth Revision diagnosis codes for acute gastroenteritis. The primary outcome measures were indicators for overuse and misuse of care in managing acute gastroenteritis based on nationally published guidelines. The indicators included blood testing, stool studies, use of antibiotics, and use of nonrecommended medications (antiemetic or antidiarrheal medications). At the 280 hospitals included in the study, the overall number of children admitted with acute gastroenteritis ranged from 1 to 174 per year; median patient volumes within small, medium, and large categories of hospitals were 7, 13, and 30 patients per year, respectively. Approximately 46% of the hospitals were in Southern urban locations. In addition to admitting patients with the highest severity of illness, large hospitals were more likely to be Southern, urban, and teaching hospitals; payer mix, including percentage of patients covered by government insurance plans, did not vary by hospital size. The majority of children (80%) in the study cohort were between the ages of 3 months and 7 years. Overall, 58% were admitted from the emergency department, 79% were cared for in the hospital by a pediatrician, 92% stayed in the hospital between 1 and 3 days, and 78% had a rating of moderate severity of illness at discharge. Patient characteristics were well balanced across volume categories. Analyses of misuse of care revealed that 6% of patients received nonrecommended medications, a percentage that varied by location (1.5% in Northeastern urban hospitals to 16% in Southern rural hospitals). Antibiotics were administered to 26% of the patients; use was higher if the diagnosis was bacterial gastroenteritis rather than viral or nonspecific gastroenteritis (63% vs 23% of patients). There was a variation in antibiotic use by location: 14% of children in Northeastern urban hospitals were given antibiotics compared with 37% of patients in Southern rural hospitals. Analysis following adjustment for patient- and hospital-level variables found that higher volumes of patients admitted with acute gastroenteritis were associated with less frequent use of nonrecommended medications (odds ratio [OR], 0.84; 95% confidence interval [CI], 0.76-0.93; P<.001) and antibiotics (OR, 0.93; 95% CI, 0.86-0.99; P=.04). There were similar correlations between higher total patient volumes and frequency of use of nonrecommended medications and antibiotics. Analyses on overuse of care revealed that blood tests were performed for 80 of the cohort patients; the percentage ranged from 76% of children in urban hospitals in the Northeast to 94% in rural facilities in the South. Higher volumes of hospital admissions for acute gastroenteritis were associated with less use of blood tests (OR, 0.67; 95% CI, 0.50-0.89). Stool testing was conducted in 46% of children; rotavirus testing in 56%. There was little variation in testing rates across regions. Compared with children admitted to facilities in the 75th percentile of patient volume, those admitted to hospitals in the 25th percentile were 10% more likely to have blood tests, 30% more likely to receive nonrecommended medications, and 10% more likely to receive antibiotics.