"Return to duty" as an outcome metric in military concussion research: Problems, pitfalls, and potential solutions.

Objective: "Return to duty" (RTD) is often used as an outcome metric in military concussion research, but is inconsistently defined across studies and presents several key problems to researchers. Using results from the Defense and Veterans Brain Injury Center's (DVBIC) Progressive Return to Activity (PRA) study, we highlight problems with RTD, and suggest solutions to inform future efforts.

Method: 116 service members (SMs) were enrolled in one of two groups (pre-implementation and post-implementation of the PRA Clinical Recommendation [CR]). Data, including the Neurobehavioral Symptom Inventory (NSI-22), was collected within 72-hours of injury (baseline), and at 1-week, 1-month, 3-months, and 6-months post-injury. Our analyses focused on three time points: baseline, approximate RTD date, and post-RTD follow-up, with RTD data captured via self-report and electronic medical record (EMR). Secondary analyses included comparisons across PRA-CR implementation groups.

Results: Of those SMs (<50% of the sample) with both self-reported and EMR RTD dates, dates largely did not match (range 1 to 36 days). RTD (either date) also did not indicate symptom recovery, with >50% of SMs reporting "abnormally high" symptom levels (i.e., NSI-22 total ≥75 percentile) at RTD, and over 50% of SMs reporting at least one significant symptom (i.e., any NSI-22 item ≥ 2) after RTD.

Conclusions: Our data demonstrate challenges encountered with a RTD outcome metric. Military concussion researchers should strive to use a well-defined RTD outcome metric. We propose defining RTD as a . Further, researchers should utilize Department of Defense definitions of Individual Medical Readiness and Deployment Limiting conditions to increase specificity of a RTD outcome metric. Improving the way RTD is captured will improve confidence that tools used after a SM sustains concussion are adequately informing RTD decisions.