Home  |   Subscribe  |   Resources  |   Reprints  |   Writers' Guidelines

Summer 2024 Issue

Electronic Health Records: Identifying Patients With Dementia
By Sue Coyle, MSW
For The Record
Vol. 36 No. 3 P. 28

Investigators at Cedars-Sinai aim to use EHRs to identify patients with dementia upon hospital admission and improve outcomes.

Individuals with dementia experience higher rates of hospitalization. The Canadian Institute for Health Information (CIHI) found that one in five older adults with dementia were hospitalized over the course of 2015 and 2016, with nearly 40% of that cohort being hospitalized more than once. CIHI reported that “Hospitalization rates were 65% higher for seniors with dementia: 33 hospitalizations per 100 seniors with dementia compared with 20 hospitalizations per 100 seniors without.”1

On top of higher rates of hospitalization, individuals with dementia are more likely to have poorer outcomes once hospitalized. “People with dementia are more vulnerable to poor outcomes and suboptimal discharge because people with dementia are less likely to be able to advocate for themselves. They’re more likely to get confused. They’re more likely to get delirium. The other thing is that they have prolonged hospital stays because they’re more prone to falls and delirium and that sort of complication, and when it comes to getting discharged are thus harder to discharge,” says Zaldy Tan, MD, MPH, director of the Memory and Healthy Aging Program and the C.A.R.E.S. Program at Cedars-Sinai in Los Angeles.

Reducing the vulnerabilities for individuals with dementia in the hospital is multifaceted but starts first and foremost with the providers knowing of either the diagnosis or the potential for a dementia diagnosis—a first step that’s harder than one might assume.

Barriers to Identification
Most commonly when individuals with dementia are admitted to the hospital, they’re not admitted because of a dementia diagnosis. Tan explains. They’re “typically admitted for pneumonia, heart attack, stroke, sepsis, and things like that or for a fall,” which can result in an orthopedic fracture, for example.

“If they come alone and they’re admitted to the hospital, unless someone has documented dementia in the medical records or alerts the hospital team that this person they’re treating has dementia, the diagnosis will be easily missed,” he says.

It’s possible, of course, that the individual or, if they’re not alone, their family member or caregiver, may disclose the diagnosis, but that’s not a given. For one, they might not know they’ve been diagnosed with dementia. A survey released in 2015 by the Alzheimer’s Association found that only 45% of patients and caregivers had been informed of their Alzheimer’s diagnosis. Excluding caregivers, only 33% of patients had been told.2

There’s also the high likelihood that an individual is admitted to the hospital with cognitive impairment or signs of cognitive impairment but without a diagnosis. Dementia is an underdiagnosed condition. Alzheimer’s Disease International reports that “75% of 55 million people with dementia are not diagnosed worldwide. This figure is as high as 90% in lower-to-middle income countries.”3

Utilizing EHRs
To better identify patients who are admitted to the hospital and have dementia, investigators at Cedars-Sinai, including Tan, developed a method to flag such individuals using EHRs. They created an algorithm that will search patients’ EHRs for specific indications that the individual may have dementia or a cognitive impairment, such as a diagnosis or prescribed medication.

“There are really two types of patients that we were identifying—patients with known dementia and then patients who may have dementia or may have a cognitive decline but have not necessarily been diagnosed with dementia,” according to Cameron Escovedo, MD, MS, physician leader of enterprise information services at Cedars-Sinai. “Amongst the former group—those who have known dementia— it’s not always documented as such. So for example, somebody might write a note somewhere that this patient has dementia, but they might not have added it to the patient’s diagnoses in the chart.”

It also may not be as clear as a written note about dementia. When creating the algorithm, investigators had to look for other means of identifying the patient, including medications the patient has been prescribed and cognitive elements of the physician’s physical exam.

“There is not an explicit statement that they have dementia, but there are these clues—these breadcrumbs—and so we have to try to go find [the clues] and figure out if there enough of these breadcrumbs that it’s worth writing these rules to pull those patients into our cohort to say ‘Yeah, they may have dementia and somebody on their team needs to do a more in-depth investigation to see if they’ve ever been diagnosed with dementia or not and if they need further evaluation for that,” Escovedo says. “And then there are patients who maybe nobody has ever identified as having dementia before, but they have indications of cognitive decline. We don’t want those patients to be lost in the study.”

When the algorithm, which was developed specifically for the Epic EHR system, identifies a patient as potentially having dementia or a cognitive decline, a yellow banner appears on the patient’s chart, signaling the identification to the care team.

The algorithm was initially tested from October 2022 to May 2023. In that time, 344 individuals were flagged as having or potentially having dementia. Of those identified, 280 had a diagnosis of dementia or had been prescribed an FDA-approved dementia medication. Only 9% of those flagged by the algorithm had no indication of cognitive decline.

Providing Care
At this time, the use of the algorithm at Cedars-Sinai is in the completion phase. The investigators are focused on completion rates and have not yet started measuring how better identifying individuals with dementia influences hospitalization outcomes, though they plan to soon.

The goal, however, is to be able to provide care that’s best suited to the individual. Thus, in practice when the banner appears, the care team at Cedars-Sinai can treat the admitting diagnosis, as well as the dementia, more appropriately, as they would if a dementia diagnosis had been known from the beginning.

For instance, decisions made pertaining to the presenting concern may be approached differently. “When you admit someone for a heart attack or a stroke, you’re not typically looking for a particular diagnosis that’s unrelated to the reason they’re admitted, and yet a diagnosis like dementia, even if it’s buried in a visit six months ago, may inform which hospital unit the patient should be admitted to, what services the patient should have, and what types of behaviors or challenges the hospital team should anticipate,” Tan says.

“A good example is asking for consent for a procedure,” he adds. “Let’s say a person came in for a heart attack and they need to get an intervention [such as] open heart surgery or an angioplasty, for example, you’ll need to get consent for that. Even if this person has dementia, especially in the hospital setting, he probably shouldn’t be making that decision on his own.

The care team will also look beyond the reason for hospitalization and address, as needed, the dementia diagnosis. For instance, Tan explains that a Cedars-Sinai patient will be plugged into the CEDARS-6 care path, a Cedars-Sinai tool that details the six domains deemed most relevant to people with dementia and offers care teams a holistic view of the individual. The six domains are cognition and behavior, environment, diseases and medications, advanced directive, resources, and social determinants of health.

“Each of these six domains are filled out by a different health professional. Then the hospital EHR Epic system will put them all together as one unified note. There appears a tab in Epic that any of the care teams who have access to [it can view],” Tan says.

This is just one example of how care may change when a patient is identified as potentially having dementia.

Moving Forward
It’s hoped that this algorithm will help improve the outcomes for patients with dementia when admitted to Cedars-Sinai, and Tan would like to see it tested in a multisite study to determine its efficacy outside of Cedars-Sinai as well.

Additionally, the focus right now is on inpatient care and addressing the vulnerabilities and outcomes of individuals while they are in the hospital. Both Tan and Escovedo would like that focus to expand.

“Right now, the work to date has been mostly on the inpatient side, capturing patients who have potential cognitive decline in the hospital. A good amount of work can happen in the hospital, but that’s a snapshot in time,” Escovedo says.

“Now what we’re trying to do,” he continues, “is patients who are identified in the hospital—how do we then identify their primary care providers who are seeing them during hospital follow up to make sure that their primary care providers can maybe pick up the outstanding tasks that weren’t completed in the hospital and help the patient get plugged in to the services they need after they have left the hospital.”

— Sue Coyle, MSW, is a freelance writer in the Philadelphia suburbs.

 

References
1. Dementia in hospitals. Canadian Institute for Health Information website. https://www.cihi.ca/en/dementia-in-canada/dementia-care-across-the-health-system/dementia-in-hospitals#:~:text=Back%20to%20top-,Higher%20hospitalization%20rates,reversed%20in%20older%20age%20groups

2. Leonard K. Most Alzheimer’s patients not told about their diagnosis. U.S. News & World Report. March 24, 2015. https://www.usnews.com/news/articles/2015/03/24/most-alzheimers-patients-not-told-about-their-diagnosis#:~:text=Many%20doctors%20do%20not
%20inform,other%20long%2Dterm%20emotional%20problems

3. Over 41 million cases of dementia go undiagnosed across the globe – World Alzheimer Report reveals. Alzheimer’s Disease International website. https://www.alzint.org/news-events/news/over-41-million-cases-of-dementia-go-undiagnosed-across-the-globe-world-alzheimer-report-reveals/. Published September 21, 2021.