Participation in Fitness-Related Activities of an Incentive-Based Health Promotion Program and Hospital Costs: A Retrospective Longitudinal Study

• Published in: American journal of health promotion 2011-05, Vol.25 (5), p.341-348
• Main Authors: Patel, Deepak, Lambert, Estelle V., da Silva, Roseanne, Greyling, Mike, Kolbe-Alexander, Tracy, Noach, Adam, Conradie, Jaco, Nossel, Craig, Borresen, Jill, Gaziano, Thomas
• Format: Article
• Language: English
• Subjects: Adolescent; Adult; Aged; Female; Health Promotion - economics; Health Promotion - methods; Health technology assessment; Hospital Costs; Hospitalization - economics; Hospitalization - statistics & numerical data; Humans; Insurance Claim Review - economics; Insurance Claim Review - statistics & numerical data; Insurance, Health - organization & administration; Logistic Models; Longitudinal Studies; Male; Middle Aged; Motivation; Physical Fitness; Retrospective Studies; South Africa; Young Adult
• ISSN: 0890-1171; 2168-6602
• DOI: 10.4278/ajhp.100603-QUAN-172

Purpose. A retrospective, longitudinal study examined changes in participation in fitness-related activities and hospital claims over 5 years amongst members of an incentivized health promotion program offered by a private health insurer. Design. A 3-year retrospective observational analysis measuring gym visits and participation in documented fitness-related activities, probability of hospital admission, and associated costs of admission. Setting. A South African private health plan, Discovery Health and the Vitality health promotion program. Participants. 304,054 adult members of the Discovery medical plan, 192,467 of whom registered for the health promotion program and 111,587 members who were not on the program. Intervention. Members were incentivised for fitness-related activities on the basis of the frequency of gym visits. Measures. Changes in electronically documented gym visits and registered participation in fitness-related activities over 3 years and measures of association between changes in participation (years 1—3) and subsequent probability and costs of hospital admission (years 4—5). Hospital admissions and associated costs are based on claims extracted from the health insurer database. Analysis. The probability of a claim modeled by using linear logistic regression and costs of claims examined by using general linear models. Propensity scores were estimated and included age, gender, registration for chronic disease benefits, plan type, and the presence of a claim during the transition period, and these were used as covariates in the final model. Results. There was a significant decrease in the prevalence of inactive members (76% to 68%) over 5 years. Members who remained highly active (years 1—3) had a lower probability (p < .05) of hospital admission in years 4 to 5 (20.7%) compared with those who remained inactive (22.2%). The odds of admission were 13% lower for two additional gym visits per week (odds ratio, .87; 95% confidence interval [CI], .801—.949). Conclusion. We observed an increase in fitness-related activities over time amongst members of this incentive-based health promotion program, which was associated with a lower probability of hospital admission and lower hospital costs in the subsequent 2 years. (Am J Health Promot 2011;25[5]:341-348.)