Predicting nonrecovery in adults with incident traffic injuries including post-traumatic headache.

Importance: The management of traffic injuries is challenging for clinicians. Knowledge about predictors of nonrecovery from traffic injuries may help to improve patient care.

Objective: To develop a prediction model for self-reported overall nonrecovery from traffic injuries six months post-collision in adults with incident traffic injuries including post-traumatic headache (PTH).

Design: Inception cohort studies of adults with incident traffic injuries (including PTH) injured in traffic collisions between November 1997 and December 1999 in Saskatchewan, Canada; and between January 2004 and January 2005 in Sweden.

Methods: Prediction model development and geographical external validation.

Setting: The Saskatchewan cohort (development) was population-based (N = 4,162). The Swedish cohort (validation) (N = 379) were claimants from two insurance companies covering 20% of cars driven in Sweden in 2004.

Participants: All adults injured in traffic collisions who completed a baseline questionnaire within 30 days of collision. Excluded were those hospitalized > 2 days, lost consciousness > 30 min, or reported headache < 3/10 on the numerical rating scale. Follow-up rates for both cohorts were approximately 80%.

Predictors: Baseline sociodemographic, pre-injury, and injury factors.

Outcome: Self-reported nonrecovery from all injuries (not "all better (cured)" on the self-perceived recovery scale) six months after traffic collision.

Results: Both cohorts were predominantly female (69.8% in Saskatchewan, 65.2% in Sweden), with median ages 35.9 years (Saskatchewan), and 38.0 years (Sweden). Predictors were age, low back pain, symptoms in arms or hands, hearing problems, sleeping problems, pre-existing headache, and lower recovery expectations. With a positive score (i.e., ≥0.85 probability), the model can rule in the presence of self-reported nonrecovery from all injuries at six months (development: specificity = 91.3%, 95% CI 89.2%-93.0%; sensitivity = 27.8%, 95% CI 26.0%-29.7%; positive likelihood ratio (LR + ) = 3.2, 95% CI 2.5-4.0; negative likelihood ratio (LR-) = 0.79, 95% CI 0.76-0.82; validation: specificity = 72.6%, 95% CI 61.4%-81.5%; sensitivity = 60.5%, 95% CI 53.9%-66.7%); LR+ = 2.2, 95% CI 1.5-3.3; LR- = 0.5, 95% CI 0.4-0.7).

Conclusions And Relevance: In adults with incident traffic injuries including PTH, predictors other than those related to baseline head and neck pain drive overall nonrecovery. Developing and testing interventions targeted at the modifiable predictors may help to improve outcomes for adults after traffic collision.