Prescribing Cascades among Nursing Home Residents with ADRD

PROJECT SUMMARY Polypharmacy, usually defined as the concurrent use of five or more medications, occurs in over 90% of nursing home (NH) residents with Alzheimer’s Disease and Related Dementias (ADRD). Prescribing cascades are a key driver of inappropriate polypharmacy. A prescribing cascade occurs when adverse events from one medication

(the “precipitant medication”) are misinterpreted as manifestations of a new medical condition, resulting in the addition of another unnecessary medication to manage symptoms (the “resultant medication”). Because the underlying cause of the adverse event is never resolved and a new medication with additional risks is started,

prescribing cascades may result in adverse outcomes such as hospitalizations, injurious falls, decreased physical and cognitive function, and death. Our research team’s preliminary data suggest over 50% of NH residents may experience prescribing cascades. There is a severe lack of evidence on the causal effects of

prescribing cascades on health outcomes in NH residents, especially those with ADRD, who are at high risk of polypharmacy and its adverse effects. To optimize prescribing and deprescribing, it is essential to generate evidence on prescribing cascades and their effects among NH residents with and without ADRD. Without

additional evidence, providers will be unable to identify clinically significant prescribing cascades that could be intervened on to improve NH resident health outcomes. Thus, the overall objective of this proposal is to understand how prescribing cascades arise in NH residents and to quantify their effects on outcomes. Our central

hypothesis is that prescribing cascades increase the risks of adverse health outcomes (e.g., hospitalizations, decreased physical function, fall-related injuries, mortality) and the risks will be greatest among residents with severe cognitive impairment. This hypothesis will be tested through three specific aims: Aim 1, Identify risk

factors for clinically relevant prescribing cascades among NH residents with and without ADRD; Aim 2, Estimate the effect of precipitant medication use on subsequent resultant medication use among NH residents with and without ADRD; and Aim 3, Quantify the effects of the prescribing cascades on adverse outcomes among all

older adults in NHs and those in high-risk subgroups. To accomplish the proposed aims, we will leverage a large innovative database of NH electronic health record information linked to national Minimum Data Set clinical assessment records, Medicare health insurance claims, and other relevant datasets. We will also innovate by

developing the use of novel causal inference approaches, including sequential target trial emulation and doubly robust estimation methods, for studying prescribing cascades and other related exposures. The proposed research is highly significant because it will provide clinically actionable empirical evidence to help guide the

prescribing, deprescribing, and management of medications to minimize the risk of important prescribing cascades and adverse outcomes among NH residents with ADRD. This proposal is directly responsive to PAR- 22-093 and advances the National Institute on Aging’s Strategic Goal C to improve the safe use of medications.