Collaborative Project Expands Prognostic Tools for Hospitalized Older Adults With Dementia

Does having dementia herald a worse prognosis for an older patient with prostate cancer? Is it worth putting a 92-year-old through bypass surgery after a heart attack? How much will a patient’s risk of dying increase if they stop a medication that’s causing unwanted side effects? These are the kinds of questions frequently asked by patients and caregivers when expected to make informed medical decisions.

Seeking answers to these questions and others, Cleveland Clinic has joined a collaborative initiative designed to develop better clinical prediction models for geriatric patients. The project, which will further expand on a widely used ePrognosis platform developed at the University of California San Fransisco (UCSF), will mine an extensive new data set on the management of hospitalized older adults.

“Despite the fact that dementia is the fourth leading killer of older adults in the U.S., our current prediction models fall short,” explains Cleveland Clinic geriatrician Kenneth Koncilja, MD, a lead investigator for the project. “Our ability to improve and extend the lives of patients with dementia depends, to a large degree, on how well we understand the disease. Because so many questions linger, the medical community has an enormous appetite for the kind of lifesaving information this project will provide.”

In with the new

Dr. Koncilja notes that current prognostic tools for geriatric patients are based on decades-old Medicare data that have failed to keep up with new treatments, diagnostic methods and medications. Many screening formulas, including the criteria for hospice for patients with dementia, were derived from data that hasn’t been updated since the mid-90s.

The access to additional hospital data made possible by this project boosts the number of conditions and treatments addressed by diagnostic tools, he says, and increases the number of valuable data points that can be studied. Dr. Koncilja credits this welcome expansion to the extensive volume of information collected by modern electronic health records.

“It’s enormously difficult for caregivers to make decisions based on outdated constructs, and it means that patients who might benefit from hospice may not get it because our clinical definitions are so confusing and archaic,” he says. “With stronger, more accurate prognostic tools, we can help people live longer, fuller lives – and, I hope, give patients and their families the guidance they deserve.”

The new project will analyze five years’ of retrospective data from hundreds of thousands of patients. Researchers will focus on every health condition affecting hospitalized adults over age 50, with and without dementia. Algorithms will be used compare hundreds of patient data points, including demographics, vital signs, diagnoses, medications, lab values and interventions. Investigators will also review data on patients’ independence and functioning, as assessed by physical therapists, occupational therapists and other caregivers.

The goal, says Dr. Koncilja, is to tease out trends and identify which factors are most important in predicting patient outcomes.

Deriving meaning from data

The project is made possible by a five-year, $5 million grant from the National Institutes of Health/National Institute of Aging. Of the six hospitals U.S. hospitals participating in the study, Cleveland Clinic has contributed the largest volume of patient data. Researchers plan to begin releasing publications and making prediction tools publicly available within three years, Koncilja says.

The new project also enables researchers to improve the diversity of patients represented in the model by tapping into hospital systems in a variety of locations across the country, he adds.

Dr. Koncilja hopes the expanded platform will eventually help clinicians, patients and their loved ones make more informed, confident decisions about medications, treatments and end-of-life care. He notes that no effective treatment exists for dementia, despite decades of research into the diagnosis and prevention of the disease.

“That’s why early interventions aimed at preventing or delaying dementia progression are so critical,” he adds. “There is a lot riding on a diagnosis like this and a great many decisions that must be made about the patient’s current and future care. Our hope is that the data revealed through this study will help reshape the way we evaluate, manage and support geriatric patients worldwide.”