The burden of poor health management: Evidence from an incentive-based health promotion program in South Africa

Research Objective: To quantify the relationship between maintaining a healthy lifestyle, assessed by in-range biometrics and wellness engagement, and health care utilization and costs. We hypothesize that individuals who maintain healthy biometrics and engage in health wellness activities, utilize healthcare less frequently for secondary and tertiary care and incur less costs to the system, as compared to individuals who do not maintain healthy biometrics and do not engage in health wellness. We further hypothesize that the healthcare savings are larger for individuals with pre-existing chronic conditions.

Research design: A retrospective longitudinal analysis of data obtained from the Discovery Health medical plan, which is the largest health insurer in South Africa, the Discovery Health Promotion Vitality program, and medical claims.  Predictor variables include age, gender, plan type (comprehensive, saver, and core), and pre-existing chronic conditions. Levels of engagement in Vitality include non-engaged, low engagement, and high engagement. Biometrics include body mass index (BMI), systolic blood pressure, diastolic blood pressure, total cholesterol level, random glucose level, smoking status, and weekly alcohol consumption. Outcome variables include all hospital inpatient and outpatient/ambulatory care utilization and claims submitted and paid by Discovery Health. Claims were also categorized by type and illness categories (chronic vs. non-chronic; cancers, asthma, diabetes, etc.) Multivariate analyses were conducted to measure the associations between biometrics, levels of wellness engagement, and healthcare utilization patterns and costs, controlling for all predictor variables, year, and members’ intra-class correlation. All analyses controlled for time variant effects and length of participation in the program, the longest observed period being for three years. The two-part model was used to model health care costs.

Population studied: Adult members of the Discovery Health medical plan in South Africa ages 21 and older whose health benefits has been effective for the full 12 months during the study period 2010-2012, and who were registered for the incentivized health promotion program Vitality (N=370,892).

Principal Findings: About 49.4% of the Discovery Health members were enrolled in Vitality between 2010 and 2012. About 27.3% of Vitality members have completed at least one biometrics measurement each year. After controlling for member level effects, levels of engagement in Vitality increased each year. Preliminary results controlling only for member level effects and age, suggested that higher wellness engagement was associated with increased likelihood of in-range biometrics including BMI, systolic blood pressure, diastolic blood pressure, total cholesterol levels, and random glucose level; non-smoking status and less alcohol consumption. Membership enrollment and biometrics data were matched by enrollee and year with insurance medical claims, about half of the medical claims were submitted for hospital inpatient services and the remaining half were submitted for hospital outpatient and ambulatory care.

Conclusions: Results will describe healthcare utilization patterns and estimate marginal costs that are associated with poorly maintained health, in healthy individuals as well as individuals with chronic illnesses.

Implications for Policy, Delivery or Practice: Results will estimate the burden of poor health management and may generate evidence for wellness programs to curb the growth in health care expenditures.