A multifaceted hospitalist quality improvement intervention: Decreased frequency of common labs

PURPOSE Common labs such as a daily complete blood count or a daily basic metabolic panel represent possible waste and have been targeted by professional societies and the Choosing Wisely campaign for critical evaluation. We undertook a multifaceted quality-improvement (QI) intervention in a large community hospitalist group to decrease unnecessary common labs. METHODS The QI intervention was composed of academic detailing, audit and feedback, and transparent reporting of the frequency with which common labs were ordered as daily within the hospitalist group. We performed a pre-post analysis, comparing a cohort of patients during the 10-month baseline period before the QI intervention and the 7-month intervention period. Demographic and clinical data were collected from the electronic medical record. The primary endpoint was number of common labs ordered per patient-day as estimated by a clustered multivariable linear regression model clustering by ordering hospitalist. Secondary endpoints included length of stay, hospital mortality, 30-day readmission, blood transfusion, amount of blood transfused, and laboratory cost per patient. RESULTS The baseline (n = 7824) and intervention (n = 5759) cohorts were similar in their demographics, though the distribution of primary discharge diagnosis-related groups differed. At baseline, a mean of 2.06 (standard deviation 1.40) common labs were ordered per patient-day. Adjusting for age, sex, and principle discharge diagnosis, the number of common labs ordered per patient-day decreased by 0.22 (10.7%) during the intervention period compared to baseline (95% confidence interval [CI], 0.34 to 0.11; P