This study aimed to identify factors that predict patients at risk of unchanged or worsened functioning after hand surgery. A cohort of 421 patients with various hand-related disorders was followed for one year. Functioning was measured using the Disabilities of the arm, shoulder and hand short-form questionnaire (Q-DASH). Demographic information, sense of coherence scale (SOC-13), and pain catastrophizing scale (PCS) were collected at baseline. We applied penalized lasso regression for model selection. The most paramount model was able to distinguish between patients with improved Q-DASH scores and those with unchanged or worsened scores, with 77% accuracy at three months. Male sex and the presence of cancer or other diseases were associated with an increased risk of non-improvement, whereas student status, long-term third level education and ambidexterity were associated with a lower risk. In addition, items 3 and 13 from the SOC-13 and item 5 from the PCS were also included in the model. The model were not able to predict Q-DASH not having improved from baseline to 12 months any better than the trivial model that predicted only the overall prevalence of not improving in the cohort. While the model’s current performance provides valuable risk stratification at three months, further refinement and validation in independent cohorts are needed for broader applicability.