Development and usage of a health recommendation web tool (HeaRT) designed to inform women of personalized preventive health recommendations

Internet Interv. 2022 Dec 24:31:100599. doi: 10.1016/j.invent.2022.100599. eCollection 2023 Mar.

Abstract

Background: Implementation of guidelines for evidence-based screening and disease prevention remains a core challenge in health care. The lack of access to accurate and personalized health recommendations may contribute to sub-optimal performance of medical screening, and ultimately increased risk for communicable and non-communicable disease. Many women do not monitor their cardiovascular disease (CVD) risk or receive regular medical screenings. A health recommendation tool (HeaRT) that provides women with profiled, individually tailored information about recommended tests and screening was designed to improve women's engagement in preventive health. This study characterized utilization of the tool in a real world setting.

Objective: To describe the development and usage patterns of HeaRT, a novel health web-tool that provides personalized health recommendations for women.

Methods: Extracted web-tool data including user input (age, BMI, smoking status and family history of CVD) and time spent in the results screen were analysed. Engagement was assessed by time spent in each results category, number of clicks and whether the user emailed/printed the recommendations. Usage patterns were analysed using multivariate analyses, logistic regression and cluster analyses.

Results: HeaRT was used 13,749 times in the years between its launch and data extraction three years later. Web-tool analysis found that 68.6 % of users accessed results and approximately 15 % printed or emailed the list of recommendations. Further analysis found that almost all the users entered the nutrition category (78 %), followed by the risk-factor category (69.5 %) and Physical activity category (61.9 %). Three usage patterns were identified by cluster analysis, including a nutrition/physical activity cluster, a risk-factor cluster and an all-categories cluster. Cluster affiliation analysis found BMI and smoking status were not predictors of cluster affiliation, whereas users over the age of 65 were more likely to solely enter the risk-factor tab (P < .001) and users with family history of CVD were more likely to either enter only the risk-factor tab or to enter all tabs (P < .01).

Conclusions: HeaRT users looked at health recommendations on a variety of health topics, and 15 % printed or emailed the recommendations. A tailored health recommendation web-tool may empower women to seek preventive-care and health maintenance, and help them interact with health care providers from a position of shared responsibility. This tool and similar programs may enable health care consumers to actively participate in directing their own health maintenance by providing consumers with personalized health recommendations. Additionally, user characteristics may inform future web-tool designers on target population profile and usage patterns.

Keywords: Decision aid; Ehealth; Ehealth literacy; Gender-medicine; Screening; Web-tool.