A novel fracture liaison service using digital health: impact on mortality in hospitalized elderly osteoporotic fracture patients.

ABSTRACT

We examined the performance of an intelligent fracture liaison service (FLS) assisted by digital health (DH) to reduce all-cause mortality (ACM) risk. According to our findings, the new FLS reduced ACM by 36%. INTRODUCTION: A well-designed secondary prevention program known as FLS enhances the bone densitometry-based assessment rate as well as osteoporosis (OP) medication usage following a fracture. However, there are only a few reports on FLS incorporating DH, and it remains unclear whether this integration has influenced patient ACM, which refers to the overall death rate from any cause during the study period. METHODS: This retrospective observational study was conducted on data from the Fragility Fracture Registration System database linked to the Regional Health Registration Platform of Kunshan City and the Population Death Registration System of Jiangsu Province for one tertiary-level A hospital in China. Patients aged ≥ 50 years, who experienced an OP fracture between January 1, 2017, and July 27, 2022, requiring hospitalization, were selected for analysis. We compared the outcomes of patients who received routine fragility fracture management (the no-FLS group) or FLS (the FLS group). We employed multivariable Cox regression with inverse probability weighting based on the propensity score (PS). RESULTS: Of 2317 patients, 756 (32.6%) received FLS and 1561 (67.4%) did not. Using PS matching, we minimized the baseline characteristic differences between the two groups in the propensity score-matched samples, relative to the unmatched samples. Based on our analysis, the new FLS reduced ACM by 36% (hazard ratio [HR], 0.64; 95% confidence interval [CI], 0.47 to 0.87; P-value = 0.004). Moreover, FLS patients experienced further reductions in fall-related mortality, refracture rate, and total refracture-related hospital costs, and had increased dual-energy X-ray absorptiometry (DXA) testing and treatment initiation rates, relative to the no-FLS patients. CONCLUSIONS: A new FLS model implementation assisted by DH can effectively reduce ACM among elderly patients with OP fractures requiring surgery. In future investigations, we recommend examining the scalability of this model.