Effect of population-based interventions on laboratory utilization: a time-series analysis.

CONTEXT Previous studies have identified methods of decreasing laboratory utilization. However, most were hospital-based, relatively small, single-centered, or of limited duration. OBJECTIVE To determine the effect of 3 population-based interventions (physician guidelines, laboratory requisition form modification, and changes to funding policy) on laboratory utilization in Ontario. DESIGN Interventional time-series analysis in which data analysis was based on all claims made to the Ontario Health Insurance Program between July 1, 1991, and April 1997 for laboratory tests affected by the interventions. SETTING All clinical laboratories (not based in hospitals) in Ontario. INTERVENTIONS Physician guidelines, modification of laboratory requisition form, and changes in funding policy for the use of the erythrocyte sedimentation rate test (ESR), microscopic urinalysis, tests for renal function, iron stores, serum urea, and serum iron determinations, and tests for thyroid dysfunction (total thyroxine and thyroid-stimulating hormone [TSH]). MAIN OUTCOME MEASURES Change from 1991 to 1997 in utilization rates of ESR, microscopic urinalysis, serum urea and iron determinations, and tests for total thyroxine and TSH. RESULTS Age- and sex-standardized rates for laboratory tests unaffected by the interventions were stable during the study period. Utilization of ESR and urea determination decreased by 58% (P<.001) and 57% (P<.001), respectively, after they were removed from the requisition form and guidelines discouraging their use were disseminated. Rates for urinalyses without microscopy increased by 1700% (P<.001), while microscopic urinalysis decreased by 14% (P<.001), after a policy change eliminated microscopic urinalysis from routine urinalysis. Rates of iron determination declined by 80% (P<.001) and ferritin rates increased by 34% (P= .05) when policy changes eliminated iron testing when ordered with ferritin and guidelines advocating ferritin alone for investigating iron deficiency were disseminated. Utilization of total thyroxine testing declined by 96% (P = .02) when the provincial health plan stopped its funding. When TSH was removed from the laboratory requisition form, a 12% decline (P= .03) in its use was observed. Through April 1997, these interventions saved more than 625000 tests or $210400. CONCLUSIONS The combination of guideline dissemination, laboratory requisition form modification, and changes to funding policy was associated with significant reductions in laboratory utilization.