Factors associated with recovery from homelessness among veterans in permanent supportive housing

Aims We sought to identify person- and program-level factors distinguishing permanent supportive housing (PSH) residents with higher versus lower social integration; and higher versus lower instrumental functioning. Methods Among 60 PSH residents at Los Angeles' VA, surveys and medical records captured person-level factors. Using a median split, we dichotomized participants with higher versus lower social integration; and higher versus lower instrumental functioning. Recursive partitioning (RP) identified variables that best-differentiated these subgroups. Interviews with 26 participants captured their perceptions on social integration and instrumental functioning. Results Using RP, health-related quality of life, psychiatric symptoms and case management frequency best-differentiated the social integration subgroups. Few perceived that PSH affected social integration. RP did not yield a stable model to differentiate the instrumental functioning subgroups; participants perceived that PSH addressed most functional deficits. Conclusions Services that enhance social integration may benefit PSH residents with poor health; existing services may adequately address instrumental functioning.