Social determinants of outcomes in nonoperatively treated proximal humerus fractures.

Background: Proximal humerus fractures (PHFs) are common fractures especially in the elderly, with most fractures being managed nonoperatively. Traditional biomedical factors such as radiological alignment have not been able to meaningfully predict comfort and capability after PHFs. Conversely, recent literature has increasingly recognized the role of psychological factors in determining comfort and capability after PHFs. Nonetheless, less is known about the impact of social factors. Additional study of these potentially modifiable social factors as targets for enhancing recovery from injury is merited. Among people recovering from a nonoperatively- treated proximal humerus fracture (PHF) we studied the social factors associated with patient-reported outcomes at 6 months and 1 year.

Methods: One hundred seventy-one patients who received nonoperative management of a PHF completed baseline measures of sociodemographic characteristics (age, gender, race, employment status, household income, educational level, presence of domestic workers, housing type, and smoking status). Six and 12 months after fracture, participants completed the Oxford Shoulder Score (OSS), Quick Disabilities of the Arm, Shoulder and Hand (QuickDASH) and EuroQol-5-Dimensions (EQ5D) measures of comfort and capability. The relationship between capability and social factors was assessed using linear regression modelling, accounting for potential confounding from age, fracture severity assessed using Neer classification, premorbid comorbidities measured by Charlson Comorbidity Index, and premorbid functional status measured by Parker Mobility Index and Barthel Index.

Results: Lower capability (higher QuickDASH scores) 6 months and 1 year after fracture were associated with being unemployed (coef: -5.02 [95% CI: -9.96 to -0.07];  = .047) and having domestic workers at home (coef: 8.63 [95% CI: 1.39 to 15.86];  = .020), but not with Neer classification. Both greater shoulder discomfort and magnitude of incapability (lower OSS scores) and worse general quality of life (lower EQ5D scores) were associated with having domestic workers (coef: -4.07 [95% CI: -6.62 to -1.53];  = .002 and coef: -0.18 [95% CI: -0.29 to -0.07];  = .001 respectively) or living in an assisted care facility (coef: -14.82 [95% CI: -22.24 to -7.39];  < .001 and coef: -0.59 [95% CI: -0.90 to -0.29]  < .001).

Conclusions: The finding that people recovering from PHF experience less incapability in proportion to their social independence (employment, absence of a caregiver such as domestic workers at home and living outside care facilities) emphasizes the important associations of social factors to musculoskeletal health, and the utility of accounting for social factors in the development and assessment of care strategies.