American Journal of Medicine, internal medicine, medicine, health, healthy lifestyles, cancer, heart disease, drugs

Predicting Falls in People Aged 65 Years and Older from Insurance Claims

Comparison of predictive model variations by number of members identified as high-risk during validation (>20%). The number of people who actually fell within each model's estimated high-risk group are provided (in blue) and the number within each bar denotes the fraction of the high-risk members who fell, ie, positive predictive value.

Comparison of predictive model variations by number of members identified as high-risk during validation (>20%). The number of people who actually fell within each model’s estimated high-risk group are provided (in blue) and the number within each bar denotes the fraction of the high-risk members who fell, ie, positive predictive value.

Accidental falls among people aged 65 years and older caused approximately 2,700,000 injuries, 27,000 deaths, and cost more than 34 billion dollars in the US annually in recent years. Here, we derive and validate a predictive model for falls based on a retrospective cohort of those 65 years and older.

Methods

Insurance claims from a 1-year observational period were used to predict a fall-related claim in the following 2 years. The predictive model takes into account a person’s age, sex, prescriptions, and diagnoses. Through random assignment, half of the people had their claims used to derive the model, while the remaining people had their claims used to validate the model.

Results

Of 120,881 individuals with Aetna health insurance coverage, 12,431 (10.3%) members fell. During validation, people were risk stratified across 20 levels, where those in the highest risk stratum had 10.5 times the risk as those in the lowest stratum (33.1% vs 3.1%).

Conclusions

Using only insurance claims, individuals in this large cohort at high risk of falls could be readily identified up to 2 years in advance. Although external validation is needed, the findings support the use of the model to better target interventions.

Accidental falls are the leading cause of injury-related death among people aged 65 years and older.1 The Centers for Disease Control and Prevention report that more than 2.7 million people 65 years and older are injured in falls annually in the US.2 Surveys reveal that women and men aged 65 to 74 years had a 12-month fall incident rate of 42.6 and 41.3 per 100, respectively; once over 74 years, the incident rate climbed to 50.6 and 62.0, respectively.3 Nationally, the direct medical cost attributable to falls is 34 billion dollars.4

Focusing interventions on individuals at high risk is most likely to reduce falls and realize a net cost savings. Current guidelines for risk assessment focus on several factors.56 The initial screening typically takes into account self-reported history of falls, balance problems, and an unsteady gait. Additionally, the guidelines recommend checking for orthostatic hypotension, visual impairment, and cognitive impairment, as well as review of psychoactive medications. However, use of the guidelines relies on in-person clinical evaluation and often takes into account only a small portion of the risk factors among medications,78910 diagnoses,811 and social issues.12

With the widespread digitization of medical records, predictive analytics can be applied to combine the contribution of multiple risk factors into a single probabilistic estimate to guide care.13141516171819 Models for risk of future falls have been developed for residents of nursing homes,20 inpatients,21 and members of a cohort study on aging and mobility,22 but not yet for the general elderly population. We sought to develop a predictive model that could run universally for those aged 65 years and older based solely on insurance claims.

To read this article in its entirety please visit our website.

-Mark L. Homer, PhD, MMSc, Nathan P. Palmer, PhD, Kathe P. Fox, PhD, Joanne Armstrong, MD, MPH, Kenneth D. Mandl, MD, MPH

This article originally appeared in the June 2017 issue of The American Journal of Medicine.

Comments are closed.

UA-42320404-1