Social needs screening and resource referral in the Gynecologic Oncology Clinic: Results from an innovative quality improvement intervention one title (110/111)

Abstract

Objectives: In a previously published study, we used a paper-based assessment to screen for and address social needs, the downstream actionable modifiers of the social determinants of health that contribute to racial/ethnic and socioeconomic disparities in gynecologic cancer care. Here, we examine completion rates and acceptability of social needs screening after integrating into the electronic medical record (EMR) and routine clinic workflow. Methods: We initiated paper-based social needs screening at our urban, academic gynecologic oncology clinic in 2017. In 2021, we integrated these validated assessments and referral orders into the EMR as part of routine clinic workflow (Figure 1). From July to September 2021, the screening was distributed via an online patient portal prior to the visit, or if not completed, verbally administered and entered into the EMR by clinic staff at the visit. We compared the current screening completion rate to our historical paper-based screening. To assess patient acceptance, we used multivariate Poisson regression to identify factors associated with patients desiring referral to community-based resources and opting out of future screenings. Results: Of 405 eligible patients, 92% (n=373) completed the assessment, significantly increased from the 54% historical completion rate (p<0.001). Of those who completed screening, the median age was 60 years (range: 49-69), 58% were non-Hispanic White (NHW), and 29% were non-Hispanic Black. Half (51%) had private insurance. A total of 20% (n=74) of those screened had at least one social need (range: 0-7). The most common needs were financial (14%), transportation (5%), and needing help reading hospital materials (5%). Among those with needs (n=74), 57% desired a referral for resources. Ninety percent of automatically-generated referral orders were signed. On univariate analysis, having food (n=10), housing (n=12), utility (n=5), transportation (n=19), and household items (n=19) needs was associated with an over 50% higher likelihood of desiring referral compared to patients without those needs (p<0.05). Controlling for marital status and screening modality, only Hispanic/other race compared to NHW (RR: 2.1, 95% CI: 0.8-6.2) and having two or more needs compared to one (RR: 1.5, 95% CI: 1.03-2.2) were associated with desiring referral. Of the 299 individuals without needs, 64% (n=192) opted out of future screenings. Only modality (in-person compared to online) was independently associated with opting out (RR: 1.3, 95% CI: 1.03-1.7). Insurance type was neither associated with desiring referral nor opting out. Objectives: In a previously published study, we used a paper-based assessment to screen for and address social needs, the downstream actionable modifiers of the social determinants of health that contribute to racial/ethnic and socioeconomic disparities in gynecologic cancer care. Here, we examine completion rates and acceptability of social needs screening after integrating into the electronic medical record (EMR) and routine clinic workflow. Methods: We initiated paper-based social needs screening at our urban, academic gynecologic oncology clinic in 2017. In 2021, we integrated these validated assessments and referral orders into the EMR as part of routine clinic workflow (Figure 1). From July to September 2021, the screening was distributed via an online patient portal prior to the visit, or if not completed, verbally administered and entered into the EMR by clinic staff at the visit. We compared the current screening completion rate to our historical paper-based screening. To assess patient acceptance, we used multivariate Poisson regression to identify factors associated with patients desiring referral to community-based resources and opting out of future screenings. Results: Of 405 eligible patients, 92% (n=373) completed the assessment, significantly increased from the 54% historical completion rate (p<0.001). Of those who completed screening, the median age was 60 years (range: 49-69), 58% were non-Hispanic White (NHW), and 29% were non-Hispanic Black. Half (51%) had private insurance. A total of 20% (n=74) of those screened had at least one social need (range: 0-7). The most common needs were financial (14%), transportation (5%), and needing help reading hospital materials (5%). Among those with needs (n=74), 57% desired a referral for resources. Ninety percent of automatically-generated referral orders were signed. On univariate analysis, having food (n=10), housing (n=12), utility (n=5), transportation (n=19), and household items (n=19) needs was associated with an over 50% higher likelihood of desiring referral compared to patients without those needs (p<0.05). Controlling for marital status and screening modality, only Hispanic/other race compared to NHW (RR: 2.1, 95% CI: 0.8-6.2) and having two or more needs compared to one (RR: 1.5, 95% CI: 1.03-2.2) were associated with desiring referral. Of the 299 individuals without needs, 64% (n=192) opted out of future screenings. Only modality (in-person compared to online) was independently associated with opting out (RR: 1.3, 95% CI: 1.03-1.7). Insurance type was neither associated with desiring referral nor opting out.