Predictive Models for Spine and Lower Extremity Injury After Discharge From Rehab

Subjects will be recruited across 4 medical centers after having completed a regimen of physical therapy for a spine or lower extremity injury. Upon discharge back to full duty, they will be given the opportunity to enroll in the study and undergo a battery of physical performance tests and associated surveys. The subjects will then be followed for a year to identify the occurrence of any injuries. Prediction algorithms will be derived using regression analysis to predict injury based on performance on the physical evaluation tests. The overall hypothesis is that Service Member performance on a battery of physical performance tests performed upon discharge from care and return to duty, will be able to predict 1) the risk of sustaining any injury as well as 2) reoccurrence of the same injury that they were seeking care for during the year following discharge from rehabilitation. The current assumption is that when a Service Member is discharged from medical care, it has been done based on the expectation that it is appropriate and safe for them to return to function in their operational environment. Because history of prior injury is a well-established risk factor, every single Service Member that is returned to duty after medical care for a musculoskeletal (MSK) injury is already at a higher risk for future injury than his or her non-injured counterpart. The investigators hypothesize that decreased performance on the proposed testing protocol will be related to increase in the risk of 1 year-injury and recurrence of injury. Successfully identifying those at increased risk of recurrence provides the ability for secondary and tertiary prevention programs to optimize return to duty rates. Injury will be defined as any new musculoskeletal injury or the re-occurrence of the same injury during the 1-year surveillance period. The battery of physical performance tests will include: Selective Functional Movement Assessment (SFMA), Functional Movement Screen (FMS), Upper Quarter Y-balance Test (YBT-UQ), Lower Quarter Y-balance Test (YBT-LQ), Closed Kinetic Chain Dorsiflexion (CKC DF), a Single Hop Test, Triple Hop Test, Triple Crossover Hop Test, Carry Test, and a un-weighted and weighted 300 yard Shuttle Run Test. Each subject will then also be contacted monthly via a SMS (Short Message Service, e.g. text message) survey for the following year to identify information about additional injury or profile that they may have sustained during the prior period of time. Information about injury will also be calculated from patient chart reviews and Department of Defense healthcare utilization database (claims data). This will provide a robust method in which to capture data injury data regardless of subject availability for follow-up. Subjects will be dichotomized as injured or non-injured based on the injury surveillance data. Key demographic, physical performance (FMS, YBT, SFMA, Hop Test, Carry test, \& Shuttle Run), and self-report measures will be examined for group differences. Potential predictor variables will be entered into a backward stepwise logistic regression model to determine the most accurate set of variables predictive of musculoskeletal injury status. Risk stratification (low, moderate, or high) will be based on likelihood ratios (LR) associated with the clinical prediction rule for injury outlined above. A positive LR \> 10 will place the individual as high risk, a LR between 2 and 10 would place the individual as moderate risk. Those with a positive LR less than 2 will be listed as low risk.