Project Title

Patient Choice to Opt-In or Opt-Out of Telephonic Health-Related Social Need Navigation Program

Authors' Affiliations

Sam Bailey and Dr. Nathan Hale, Department of Health Services Management and Policy, College of Public Health, East Tennessee State University, Johnson City, TN

Location

AUDITORIUM ROOM 137A

Start Date

4-12-2019 9:00 AM

End Date

4-12-2019 9:15 AM

Faculty Sponsor’s Department

Health Services Management & Policy

Name of Project's Faculty Sponsor

Dr. Nathan Hale

Type

Oral Presentation

Classification of First Author

Graduate Student-Doctoral

Project's Category

Health of Underserved Populations, Health Services Delivery, Program Evaluation in Health Sciences

Abstract Text

Background: Ballad Health participates in the Centers for Medicare and Medicaid Services’ (CMS) Accountable Health Communities (AHC) model. The AHC model is evaluating if universal screening, referral, and navigation services for health-related social needs (HRSN) can improve outcomes and reduce unnecessary utilization and costs of health care services. To ensure the evaluation of the model has sufficient statistical power, navigation services are expected to be provided to a minimum number of individuals. The purpose of this study is to analyze the characteristics of Ballad Health’s AHC navigation services that could be modified to improve opt-in rates.

Methods: The primary outcome measure was identified as whether a beneficiary contacted via telephone opted-in or –out of the navigation program. Andersen’s Behavioral Model for Health Service Use was used as the conceptual framework for selecting covariates of interest. Enabling factors were of primary interest because alternate interventions may be designed around them. Data was pulled for the time period of November 17, 2018 through February 14, 2019. Where possible, covariates were associated with data from CMS’ AHC Data Template v3.1 to accommodate replication for all AHC bridge organizations, though additional internally-collected data, which may not be available for all bridge organizations, were needed for some variables. Chi-squared tests were performed for each covariate.

Results: No statistical differences were found for the primary covariates of interest. Opt-in rates by Navigator were lowest for Navigator 5 and highest for Navigator 4 (67.53% and 88.24%). Opt-in rates by weekday of decision were lowest on Thursdays and highest on Wednesdays (64.91% and 77.42%). Opt-in rates based on time of day were lowest between 8:00am and 9:59am, and highest between 12:00pm and 1:59pm (62.50% and 100%). Opt-in rates were lowest when the decision was made six days after the screening and highest when made the same day (53.57% and 83.33%). Opt-in rates were lowest when there were five weekdays between screening and navigation decision, and highest when there were three weekdays between the screening and decision (60% and 90%). Other non-process covariates of interest that were statistically significant for opt-in rates were the presence of either food, safety, or utility needs.

Conclusions: Several groups had higher opt-in rates that were not statistically significant; small sample sizes may have impacted the significance of these differences. For example, opt-in rates were higher when made the same day as the screening than when made one day after (83.33% and 74.79%). However, only 18 beneficiary decisions were made on the same day, while 119 were made one day after. Increasing the number of same-day phone call attempts may be a method to improve opt-in rates. Importantly, date and time data for contact attempts before a beneficiary decides to opt-in or opt-out were unavailable as of the time of the analysis. These data are captured and will be added to the analysis when available, which could provide more insight into whether a beneficiary is more likely to opt-in or opt-out.

This document is currently not available here.

Share

COinS
 
Apr 12th, 9:00 AM Apr 12th, 9:15 AM

Patient Choice to Opt-In or Opt-Out of Telephonic Health-Related Social Need Navigation Program

AUDITORIUM ROOM 137A

Background: Ballad Health participates in the Centers for Medicare and Medicaid Services’ (CMS) Accountable Health Communities (AHC) model. The AHC model is evaluating if universal screening, referral, and navigation services for health-related social needs (HRSN) can improve outcomes and reduce unnecessary utilization and costs of health care services. To ensure the evaluation of the model has sufficient statistical power, navigation services are expected to be provided to a minimum number of individuals. The purpose of this study is to analyze the characteristics of Ballad Health’s AHC navigation services that could be modified to improve opt-in rates.

Methods: The primary outcome measure was identified as whether a beneficiary contacted via telephone opted-in or –out of the navigation program. Andersen’s Behavioral Model for Health Service Use was used as the conceptual framework for selecting covariates of interest. Enabling factors were of primary interest because alternate interventions may be designed around them. Data was pulled for the time period of November 17, 2018 through February 14, 2019. Where possible, covariates were associated with data from CMS’ AHC Data Template v3.1 to accommodate replication for all AHC bridge organizations, though additional internally-collected data, which may not be available for all bridge organizations, were needed for some variables. Chi-squared tests were performed for each covariate.

Results: No statistical differences were found for the primary covariates of interest. Opt-in rates by Navigator were lowest for Navigator 5 and highest for Navigator 4 (67.53% and 88.24%). Opt-in rates by weekday of decision were lowest on Thursdays and highest on Wednesdays (64.91% and 77.42%). Opt-in rates based on time of day were lowest between 8:00am and 9:59am, and highest between 12:00pm and 1:59pm (62.50% and 100%). Opt-in rates were lowest when the decision was made six days after the screening and highest when made the same day (53.57% and 83.33%). Opt-in rates were lowest when there were five weekdays between screening and navigation decision, and highest when there were three weekdays between the screening and decision (60% and 90%). Other non-process covariates of interest that were statistically significant for opt-in rates were the presence of either food, safety, or utility needs.

Conclusions: Several groups had higher opt-in rates that were not statistically significant; small sample sizes may have impacted the significance of these differences. For example, opt-in rates were higher when made the same day as the screening than when made one day after (83.33% and 74.79%). However, only 18 beneficiary decisions were made on the same day, while 119 were made one day after. Increasing the number of same-day phone call attempts may be a method to improve opt-in rates. Importantly, date and time data for contact attempts before a beneficiary decides to opt-in or opt-out were unavailable as of the time of the analysis. These data are captured and will be added to the analysis when available, which could provide more insight into whether a beneficiary is more likely to opt-in or opt-out.