Choosing Wisely: A Quality Improvement Initiative to Decrease Unnecessary Preoperative Testing

BMJ Qual Improv Rep. 2017 May 23;6(1):bmjqir.u216281.w6691. doi: 10.1136/bmjquality.u216281.w6691. eCollection 2017.

Abstract

Dartmouth-Hitchcock Medical Center is a rural, academic medical center in the northeastern United States; its General Internal Medicine (GIM) division performs about 900 low and intermediate surgical risk preoperative evaluations annually. Routine preoperative testing in these evaluations is widely considered a low-value service. Our baseline data sample showed unnecessary testing rates of approximately 36%. A multi-disciplinary team used a micro-systems approach to analyze the existing process and formulate a rapid cycle improvement strategy. Our improvement efforts focused on implementation of a Nurse Practitioner and Physician Assistant (Associate Provider) clinic to incorporate standardized protocols for preoperative assessment. Plan-Do-Study-Act (PDSA) cycles included creation of a dedicated Associate Provider run preoperative clinic, modifying and operationalizing a scheduling scheme, and creating and implementing Electronic Health Record (EHR) tools. We used Statistical Process Control (SPC) methods to analyze time ordered data for the usual care process and to compare performance with the novel preoperative clinic. The Associate Provider preoperative clinic showed unnecessary testing rates of 4% compared with 23% in the usual care cohort (p<.001) within 3 months of implementation. When testing rates across the entire division were analyzed, there was no significant change. In our GIM division this preoperative clinic was effectively staffed with Associate Providers. Dedicated leadership support, incorporating input from a diverse improvement team, and balancing innovation with other clinical needs are important elements for success. We hypothesize that protecting clinical time to focus on preoperative care, monitoring and modifying scheduling processes, and improving support for electronic health record tool implementation would have yielded further performance improvements. Our experience provides valuable learning for other primary care practices with similar challenges. Identifying appropriate patients for inclusion in these clinic visits while optimizing primary care provider collaboration are important future challenges.