Physician practice variation in the pediatric emergency department and its impact on resource use and quality of care

Pediatr Emerg Care. 2010 Dec;26(12):902-8. doi: 10.1097/PEC.0b013e3181fe9108.


Objective: To evaluate variation in case-mix adjusted resource use among pediatric emergency department (ED) physicians and its correlation with ED length of stay (LOS) and return rates.

Methods: Resource use patterns at 2 EDs for 36 academic physicians (163,669 patients at ED1) and 45 private physicians (289,199 patients at ED2) from 2003 to 2006 were abstracted for common laboratory tests, imaging studies, intravenous therapy (fluids/antibiotics), LOS and 72-hour return rate for discharged patients, and hospital admissions for all patients. Case-mix adjustment was based on triage acuity, diagnostic category, demographics, and temporal measures.

Outcome measures: (1) adjusted overall resource use for ED1 and ED2 physicians and (2) observed-to-expected ratios for ED1 physicians.

Results: Case-mix adjusted hospital admission rates among physicians varied nearly 3-fold (6.3%-18%) for ED1 and 8-fold (2.5%-19.4%) for ED2. Intravenous therapy use varied 2-fold (4.9%-10.4%) at ED1 and 3-fold (3.6%-11.4%) at ED2. Emergency department 2 physicians had an almost 2-fold (10.9%-20.6%) variation in imaging use. Variation in head computed tomography use was 2-fold (1.1%-2.5%) at ED1 and 5-fold (0.9%-4.8%) at ED2. Physicians had longer than expected LOS if they had higher than expected use of laboratory tests (r, 0.41; 95% confidence interval [CI], 0.09-0.65; P < 0.05) and imaging (r, 0.48; 95% CI, 0.17-0.69; P < 0.01). Return rate was not significantly correlated with resource use in any category. Physicians with higher than expected use of laboratory tests had higher than expected use of imaging (r, 0.62; 95% CI, 0.36-0.78; P < 0.001), head computed tomography (r, 0.49; 95% CI, 0.19-0.70; P < 0.01), and intravenous therapy (r, 0.51; 95% CI, 0.20-0.71; P < 0.01).

Conclusions: Significant variation exists in physician use of common ED resources. Higher resource use was associated with increased LOS but did not reduce return to ED. Practice variation such as this may represent an opportunity to improve health care quality and decrease costs.

Publication types

  • Multicenter Study

MeSH terms

  • Child
  • Clinical Laboratory Techniques / statistics & numerical data*
  • Diagnosis-Related Groups
  • Diagnostic Imaging / statistics & numerical data*
  • Electronic Health Records
  • Emergency Service, Hospital / statistics & numerical data*
  • Georgia
  • Health Resources / statistics & numerical data*
  • Hospitals, Pediatric / statistics & numerical data*
  • Hospitals, Urban / statistics & numerical data
  • Humans
  • Infusions, Intravenous / statistics & numerical data
  • Length of Stay
  • Patient Admission
  • Practice Patterns, Physicians'* / statistics & numerical data
  • Quality of Health Care*
  • Tomography, X-Ray Computed / statistics & numerical data
  • Triage