Jesse Cambon1, Tristan Cordier2, Elizabeth L Munnich3, Andrew Renda4, Bobby Kapur5, Shkelzen Hoxhaj6, Meredith Williams7. 1. Analytics Consultant with Humana, Louisville, KY, during this study. 2. Principal Data Scientist, Humana. 3. Assistant Professor, Department of Economics, University of Louisville, KY. 4. Director, Bold Goal, Humana. 5. Chief of Emergency Medicine, Jackson Memorial Hospital, Miami, FL. 6. Chief Medical Officer, Jackson Memorial Hospital. 7. Lead Medical Director, North Division, Humana.
Abstract
BACKGROUND: The impact of messaging campaigns on influencing urgent care- and emergent care-seeking behaviors, including the use of in-network providers, is not well-understood. Although out-of-network healthcare utilization can have negative financial consequences for patients in narrow network Affordable Care Act plans, individuals with time-sensitive medical conditions, and especially patients visiting the emergency department, may not think about out-of-network issues. Inappropriate or avoidable emergency department visits can also create unnecessary costs for patients. OBJECTIVE: To evaluate the impact of 5 messaging strategies to educate individuals about the use of in-network providers and when care should be sought in the emergency department, urgent care center, or other sites of care. METHODS: Using a retrospective analysis, individuals aged ≥18 years who were enrolled in an individually purchased Affordable Care Act-compliant Humana plan as of July 1, 2015, were randomized to 1 of 5 messaging arms (e-mail, magnet mailer with or without e-mail, and key-tag mailer with or without e-mail) or to a control group. The outreach was implemented and evaluated in 2 distinct, geographically defined populations of Orlando, Palm Beach, and Tampa, Florida (Population 1); and Atlanta, Georgia, and San Antonio and Austin, Texas (Population 2). The relative number of each emergency department, urgent care, and out-of-network visits during follow-up was modeled using negative binomial regression. Cox proportional hazard models were used to calculate the risk for ≥1 of each visit type (assessed separately) and high emergency department utilization (defined as ≥3 visits during follow-up) relative to the control, while accounting for variable follow-up time. RESULTS: The relative numbers of each visit type assessed were not significantly different for any message group compared with the control in either population. The risk for an emergency department visit was 4% lower in the e-mail arm of Population 2 (hazard ratio [HR], 0.96; 95% confidence interval [CI], 0.94-0.99; P = .005) and 7% lower in the e-mail/key-tag arm of Population 1 (HR, 0.93; 95% CI, 0.89-0.97; P = .001). The risk for high emergency department utilization was significantly reduced by the key-tag, magnet, and e-mail/key-tag strategies in Population 1, but no impact was found in Population 2. CONCLUSION: Despite the mixed results, the study provides new insights into how different messaging strategies could be used to educate patients and influence healthcare utilization decisions by people with health insurance.
BACKGROUND: The impact of messaging campaigns on influencing urgent care- and emergent care-seeking behaviors, including the use of in-network providers, is not well-understood. Although out-of-network healthcare utilization can have negative financial consequences for patients in narrow network Affordable Care Act plans, individuals with time-sensitive medical conditions, and especially patients visiting the emergency department, may not think about out-of-network issues. Inappropriate or avoidable emergency department visits can also create unnecessary costs for patients. OBJECTIVE: To evaluate the impact of 5 messaging strategies to educate individuals about the use of in-network providers and when care should be sought in the emergency department, urgent care center, or other sites of care. METHODS: Using a retrospective analysis, individuals aged ≥18 years who were enrolled in an individually purchased Affordable Care Act-compliant Humana plan as of July 1, 2015, were randomized to 1 of 5 messaging arms (e-mail, magnet mailer with or without e-mail, and key-tag mailer with or without e-mail) or to a control group. The outreach was implemented and evaluated in 2 distinct, geographically defined populations of Orlando, Palm Beach, and Tampa, Florida (Population 1); and Atlanta, Georgia, and San Antonio and Austin, Texas (Population 2). The relative number of each emergency department, urgent care, and out-of-network visits during follow-up was modeled using negative binomial regression. Cox proportional hazard models were used to calculate the risk for ≥1 of each visit type (assessed separately) and high emergency department utilization (defined as ≥3 visits during follow-up) relative to the control, while accounting for variable follow-up time. RESULTS: The relative numbers of each visit type assessed were not significantly different for any message group compared with the control in either population. The risk for an emergency department visit was 4% lower in the e-mail arm of Population 2 (hazard ratio [HR], 0.96; 95% confidence interval [CI], 0.94-0.99; P = .005) and 7% lower in the e-mail/key-tag arm of Population 1 (HR, 0.93; 95% CI, 0.89-0.97; P = .001). The risk for high emergency department utilization was significantly reduced by the key-tag, magnet, and e-mail/key-tag strategies in Population 1, but no impact was found in Population 2. CONCLUSION: Despite the mixed results, the study provides new insights into how different messaging strategies could be used to educate patients and influence healthcare utilization decisions by people with health insurance.