Psychosocial distress screening in the health care safety net.

Abstract

91 Background: Recognizing that psychosocial distress (PSD) is underestimated in patients with cancer, the Commission on Cancer mandated screening using a validated tool. Studies of PSD screening exist, but none to date in a diverse, multicultural safety net setting where patients face challenges such as homelessness, mental illness, and substance abuse, which may augment PSD. Methods: We performed a retrospective cohort study of patients with cancer offered PSD screening during 2015. Overall distress scores and problems in each domain were analyzed. Chart review identified potential predictors of distress including age, gender, race, language, housing, psychiatric illness, substance abuse, and cancer stage. Results: Of 177 eligible patients, 113 (64%) completed screening. The most common reasons patients were not screened were refusal, too symptomatic (physically or emotionally), or language barriers. Of screened patients, 40.7% were female, 57.5% male, and 1.7% transgender. 31% were Caucasian, 27% Asian/Pacific Islander, 25% Hispanic, and 17% African American. 35% were non-English speaking. 29% had history of mental illness and 34% of substance abuse. 23% were marginally housed or homeless. 63% reported moderate to severe levels of PSD as defined by the NCCN as ≥ 4. Patients with mental illness were nearly twice as likely to report PSD ≥ 4 (p = 0.012) and had higher mean PSD scores (5.78 vs. 4.03, p = 0.002). English speaking patients had a mean PSD score of 5.01 compared to 3.6 and 3.2 for Spanish and Chinese speaking patients, respectively (p = 0.02 for English v. Chinese) and more domains causing PSD (p = 0.028 for English v. Chinese). Lack of stable housing also correlated with more domains causing PSD (p = 0.05). Conclusions: This proved to be an ethnically diverse cohort with high rates of mental illness, substance abuse, and homelessness, with the majority reporting moderate to severe distress. Even with a small cohort, English language and mental illness were significant predictors of PSD, and housing status correlated with more domains contributing to PSD. Several other variables trended toward significance, suggesting a larger cohort may be needed to determine if additional characteristics predict higher levels of PSD.