Background: High value healthcare has been increasingly emphasized during residency training. The Choosing Wisely campaign, an initiative of the American Board of Internal Medicine foundation and the Society of Hospital Medicine recommend against performing repetitive laboratory tests in the setting of clinical and laboratory stability. Unnecessary phlebotomy draws increase healthcare cost, decrease patient satisfaction, and place patients at risk for phlebotomy-associated anemia. An observational study at an academic center identified unnecessary phlebotomy draws as a potential area for improvement. We aimed to decrease repetitive laboratory testing in an inpatient setting by providing education to the interprofessional team regarding high value care principles and applying those principals to diagnostic laboratory ordering practices.
Methods: Baseline data was collected for one week prior to the intervention. Nurses and residents on a 28-bed general medicine unit participated in an educational session during a ward orientation session and were provided with materials for high value care. Data was collected by daily chart audits and included the number of daily basic metabolic panels (BMP), complete blood counts (CBC), patients with more than one blood draw per day, and patients with lab-free hospital days. The estimated amount of blood saved was also calculated and communicated to the team by multiplying the lab-free patient days by 8 ml of blood saved (5 ml for a BMP and 3 ml for a CBC). Three tests of change were performed, and the results were tracked using a run chart on the unit. The tests of change included the educational session, use of a tracking board to communicate the patients without morning labs, and repeat interprofessional training when a new team arrived on the unit.
Results: A total of 365 patient-days were tracked over a 6-week period, including a one week baseline data collection. Results were compared using t-tests. A ratio of BMPs/patient/day decreased from 0.78 at baseline to 0.69, p value of 0.015. The mean ratio of CBCs/patient/day decreased from 0.72 to 0.60, p value of 0.02. The mean patients with lab-free days increased from 5.83 to 6.28, p value 0.54. The mean patients with more than one blood draw per day decreased from 5.86 to 3.48, p value 0.88. The estimated amount of blood saved was 1900mL over the course of our pilot study.
Conclusions: While two of the metrics tracked did not meet statistical significance, there was improvement overall. We plan to expand to other units, which will increase our sample size. Our study demonstrated that an interprofessional educational session and team-led tests of change were successful in reducing the number of daily morning labs on a general medicine ward.