Outcomes Associated With High- Versus Low-Frequency Laboratory Testing Among Hospitalized Children.

Document Type


Publication Date



DOI: 10.1542/hpeds.2020-005561


BACKGROUND AND OBJECTIVES: Previous pediatric studies have revealed substantial variation in laboratory testing for specific conditions, but clinical outcomes associated with high- versus low-frequency testing are unclear. We hypothesized that hospitals with high- versus low-testing frequency would have worse clinical outcomes.

METHODS: We conducted a multicenter retrospective cohort study of patients 0 to 18 years old with low-acuity hospitalizations in the years 2018-2019 for 1 of 10 common All Patient Refined Diagnosis Related Groups. We identified hospitals with high-, moderate-, and low-frequency testing for 3 common groups of laboratory tests: complete blood cell count, basic chemistry studies, and inflammatory markers. Outcomes included length of stay, 7- and 30-day emergency department revisit and readmission rates, and hospital costs, comparing hospitals with high- versus low-frequency testing.

RESULTS: We identified 132 391 study encounters across 44 hospitals. Laboratory testing frequency varied by hospital and condition. We identified hospitals with high- (13), moderate- (20), and low-frequency (11) laboratory testing. When we compared hospitals with high- versus low-frequency testing, there were no differences in adjusted hospital costs (rate ratio 0.89; 95% confidence interval 0.71-1.12), length of stay (rate ratio 0.98; 95% confidence interval 0.91-1.06), 7-day (odds ratio 0.99; 95% confidence interval 0.81-1.21) or 30-day (odds ratio 1.01; 95% confidence interval 0.82-1.25) emergency department revisit rates, or 7-day (odds ratio 0.84; 95% confidence interval 0.65-1.25) or 30-day (odds ratio 0.91; 95% confidence interval 0.76-1.09) readmission rates.

CONCLUSIONS: In a multicenter study of children hospitalized for common low-acuity conditions, laboratory testing frequency varied widely across hospitals, without substantial differences in outcomes. Our results suggest opportunities to reduce laboratory overuse across conditions and children's hospitals.

Journal Title

Hosp Pediatr





Library Record