Multidimensional screening for predicting pain problems in adults: a systematic review of screening tools and validation studies

Abstract
Screening tools allowing to predict poor pain outcomes are widely used. Often these screening tools contain psychosocial risk
factors. This review (1) identifies multidimensional screening tools that include psychosocial risk factors for the development or
maintenance of pain, pain-related distress, and pain-related disability across pain problems in adults, (2) evaluates the quality of the
validation studies using Prediction model Risk Of Bias ASsessment Tool (PROBAST), and (3) synthesizes methodological concerns.
We identified 32 articles, across 42 study samples, validating 7 screening tools. All tools were developed in the context of
musculoskeletal pain, most often back pain, and aimed to predict the maintenance of pain or pain-related disability, not pain-related
distress. Although more recent studies design, conduct, analyze, and report according to best practices in prognosis research, risk
of bias was most often moderate. Common methodological concerns were identified, related to participant selection (eg, mixed
populations), predictors (eg, predictors were administered differently to predictors in the development study), outcomes (eg, overlap
between predictors and outcomes), sample size and participant flow (eg, unknown or inappropriate handlin