Drug Reduction in Older Patients: The DROP Trial

Overview and Specific Aims Polypharmacy, defined as five or more medications, is common in older patients discharged to skilled nursing facilities (SNFs) and is associated with adverse health outcomes. As part of a recent VA project, the investigators prospectively collected medication data for 134 hospitalized, older Veterans. These data showed that, on average, 14.2 medications were ordered for Veterans discharged to SNF, which included 2.5 medications newly prescribed in the hospital that met criteria for being potentially inappropriate medications (PIMs). Moreover, 75% of these Veterans had two or more geriatric syndromes. Geriatric syndromes are clinical conditions common in older adults such as cognitive impairment, unintentional weight loss and falls. The co-occurrence of multiple syndromes is associated with loss of independence and higher healthcare utilization. All Veterans with polypharmacy also had one or more medications associated with geriatric syndromes. The prevalence of polypharmacy and multiple geriatric syndromes may partially explain why these patients often have poor health outcomes after SNF discharge. Recent data suggests that only 28% of patients discharged from SNF are living at home 100 days later, and 35% experience an adverse drug event.

There is a dearth of evidence related to the management of multiple co-existing geriatric syndromes, and few interventions have been implemented to reduce medications while also monitoring health outcomes. The relationship between polypharmacy, adverse drug events and geriatric syndromes in the VA population supports the rationale for an intervention focused on deprescribing medications before hospital discharge. The investigators have pilot-tested a multifaceted intervention (Drug Reduction in Older Patients, DROP) to engage patients and providers to reduce the number and/or dose of medications prior to hospital discharge. The proposed randomized, controlled trial is powered to evaluate the effect of this intervention on a reduction in medications as defined by the total number of prescribed medications, the number of PIMs, anticholinergic and sedative drug burden and the number of medications associated with geriatric syndromes. In addition, the investigators will collect relevant data on the prevalence and severity of geriatric syndromes and other clinical outcomes. The investigators also will use a hybrid research design to evaluate both effectiveness and implementation issues to better inform future adoption and sustainability. Our overarching hypothesis is that a hospital-based intervention to safely reduce the total number of medications represents the most feasible way to impact multiple health-related outcomes among older Veterans. Our Specific Aims reflect the primary outcomes that are the focus of the analyses, although the investigators also will measure secondary outcomes related to VA healthcare utilization and patient safety:

Specific Aim 1: Implement a patient-centered deprescribing intervention (DROP) in the hospital to reduce the total number of medications Veterans are prescribed at hospital discharge.

Specific Aim 2: Document the effects of a Veteran-centered deprescribing intervention (DROP) on medication adherence, health status, and geriatric syndromes.

Specific Aim 3: Evaluate intervention implementation to inform future adoption and sustainability.

Potential for Impact: The DROP intervention has significant potential to impact the health of older hospitalized Veterans vulnerable to poor health outcomes. During hospitalization and SNF care, older Veterans often acquire new geriatric syndromes and medications, exactly when deprescribing actions should be initiated by care providers in the context of clinical oversight. Lastly, the application of an innovative hybrid research design will allow us to collect critical data related to both intervention effectiveness and implementation to better inform future adoption and sustainability across the VA.

Innovation: Our study is advancing science in multiple ways:

1. The proposed study reflects one of the first controlled intervention trials designed to improve post discharge outcomes for the high-risk population of older Veterans transitioning from hospital to SNF to home. Healthcare costs are increasing for this patient population, and they are at high risk for medication-associated complications and newly acquired geriatric syndromes.
2. Although the number of medications is strongly associated with increases in geriatric syndromes, it is unknown whether reducing medications will improve geriatric syndromes. This trial will generate data to evaluate the impact of medication reduction on Veteran health outcomes.
3. To date, there have been no controlled intervention trials focused on reducing the number of medications prescribed for hospitalized patients discharged to SNF, and the proposed intervention is one of the first to explicitly incorporate Veteran preferences into deprescribing efforts, which is consistent with Veteran-centered care initiatives. Moreover, our preliminary data show that incorporating patient preferences in the intervention protocol significantly changes these clinical decisions.
4. The proposed trial will evaluate both clinical effectiveness and implementation to inform the potential for DROP to be adopted into VA clinical practice.

Rationale for VA Study: Dr. Simmons was recently awarded an National Institute on Aging (NIA) R01 focused on polypharmacy in hospitalized non-Veterans. The investigators summarize below the primary differences between the two studies and our rationale for the importance of the proposed VA study:

The Need for More Evidence: Our NIA study focuses on non-Veterans hospitalized at Vanderbilt University Medical Center (VUMC) and discharged to a select group of SNFs that have partnered with VUMC for post-acute care services. Unfortunately, the scope of the NIA study does not allow us to evaluate the intervention in two separate healthcare systems with two unique patient populations and a larger number of SNFs, to include VA Community Living Centers and VA contract community nursing homes.

Uniqueness of the Veteran population: Our preliminary data demonstrates important differences in older hospitalized Veterans and non-Veterans discharged to SNF. Beyond being predominately male, older Veterans in the investigators' prior study had a higher overall prevalence of geriatrics syndromes and were prescribed more medications at hospital discharge than non-Veterans, particularly antidepressants and antipsychotics. Data also suggests that Veterans may be more willing to reduce their medicines; thus, the magnitude of the intervention effects may differ in the VA. Finally, hospitalized Veterans are discharged to a diverse group of SNFs, many of which are located in rural communities. Unlike the investigators' NIA study, which only includes a select group of partner SNFs predominately located in an urban area close to Vanderbilt, the proposed study will follow Veterans from hospital to SNF to home, irrespective of which SNF they utilize. The VA study, therefore, will more closely mimic real-world practice. The innovative application of an effectiveness-implementation hybrid design will further inform unique aspects of the VA healthcare system critical to adoption.

Uniqueness of VA healthcare system: The VA healthcare system differs significantly from other hospital systems, to include medication prescribing practices and how patients access and pay for medications. Thus, the investigators do not expect the results of the NIA study at VUMC to be directly translatable into the VA healthcare system. The proposed study also will yield an understanding of Veteran and VA provider attitudes toward deprescribing and other system factors unique to the VA that may influence intervention implementation.

Note: The investigators resumed face to face enrollment for non-COVID patients as of 8/17/20. COVID+ patients were being approached by phone.

Note: Investigators focus the reporting on primary medication outcomes for two reasons: (1) to align with reporting of an NIH/NIA funded deprescribing trial that utilized a near identical intervention (NCT02979353); and (2) medication outcomes had the strongest statistical power. Variable response rates across multiple timepoints for Veteran-reported secondary outcome measures limited the statistical power to detect intervention effects on secondary outcomes. Additionally, results for the 7-day post-SNF follow up timepoint are not presented due to a protocol change mid-study that resulted in incomplete data for that timepoint. The protocol change allowed the study to retain Veterans who were referred to a SNF but ultimately discharged directly home from the hospital (instead of SNF); thus, a significant proportion of enrolled Veterans have no data for the 7-day post-SNF follow up.