Association of a Multifaceted Intervention With Ordering of Unnecessary Laboratory Tests Among Caregivers in Internal Medicine Departments

IMPORTANCE: Inappropriate use of laboratory testing is a challenging problem. Estimated overuse rates of approximately 20% have been reported. Effective, sustainable solutions to stimulate optimal use are needed. OBJECTIVE: To determine the association of a multifaceted intervention with laboratory test volume. DESIGN, SETTING, AND PARTICIPANTS: A before-after quality improvement study was performed between August 1, 2016, and April 30, 2018, in the internal medicine departments of 4 teaching hospitals in the Netherlands. Data on laboratory order volumes from 19 comparable hospitals were used as controls. The participants were clinicians ordering laboratory tests. INTERVENTIONS: The intervention included creating awareness through education and feedback, intensified supervision of residents, and changes in order entry systems. Interventions were performed by local project teams and guided by a central project team during a 6-month period. Sustainability was investigated during an 8-month follow-up period. MAIN OUTCOMES AND MEASURES: The primary outcome was the change in slope for laboratory test volume. Secondary outcomes were change in slope for laboratory expenditure, order volumes and expenditure for other diagnostic procedures, and clinical outcomes. Data were collected on duration of hospital stay, rate of repeated outpatient visits, 30-day readmission rate, and rate of unexpected prolonged duration of hospital stay for patients admitted for pneumonia. RESULTS: The numbers of internists and residents ordering tests in hospitals 1 to 4 were 16 and 30, 18 and 20, 13 and 17, and 21 and 60, respectively. Statistically significant changes in slope for laboratory test volume per patient contact were found at hospital 1 (change in slope, −1.55; 95% CI, −1.98 to −1.11; P < .001), hospital 3 (change in slope, −0.74; 95% CI, −1.42 to −0.07; P = .03), and hospital 4 (change in slope, −2.18; 95% CI, −3.27 to −1.08; P < .001). At hospital 2, the change in slope was not statistically significant (−0.34; 95% CI, −2.27 to 1.58; P = .73). Laboratory test volume per patient contact decreased by 11.4%, whereas the volume increased by 2.4% in 19 comparable hospitals. Statistically significant changes in slopes for laboratory costs and volumes and costs for other diagnostic procedures were also observed. Clinical outcomes were not associated with negative changes. Important facilitators were education, continuous attention for overuse, feedback, and residents’ involvement. Important barriers were difficulties in data retrieval, difficulty in incorporation of principles in daily practice, and high resident turnover. CONCLUSIONS AND RELEVANCE: A set of interventions aimed at changing caregivers’ mindset was associated with a reduction in the laboratory test volume in all departments, whereas the volume increased in comparable hospitals in the Netherlands. This study provides a framework for nationwide implementation of interventions to reduce unnecessary laboratory testing.