A team of researchers led by Feinberg Prof. Margaret Danilovich and McCormick Profs. Ankit Agrawal and Alok Choudhary used data mining and qualitative analysis to explore how users talk about Alzheimer’s disease and dementia. The study, which analyzed thousands of tweets from a one-month period referencing keywords related to Alzheimer’s and dementia, was recently published in the journal Topics in Geriatric Rehabilitation.
Agrawal said the long-term goal of the project was to reduce “caregiver burnout.” People giving care to others also need to take care of themselves, he added.
“The adult population who need care is rising,” he said. “If we are to ensure that they get proper care … we have to ensure that their caregiver’s health is good.”
Danilovich said that she initially was interested in caregivers because of her work as a physical therapist. Much of her research revolves around older adults and their caregivers, she said.
Agrawal said Danilovich approached him for the study because of his team’s previous work using data mining techniques to analyze tweets about cancer and the flu.
“Many times we assume that older people don’t use social media,” Danilovich said. “But these people who are providing care are using (Twitter) to document their experiences caregiving.”
Agrawal said the study focused on “informal caregivers,” like people who look after their own parents, because they freely share their life experiences on Twitter more often. Many times, these caregivers aren’t aware of resources that could help them and so aren’t actively looking for them, he added.
Danilovich said they focused on social media rather than surveys or interviews because social media is a quicker way to reach a larger number of people across the country. It is also a way for caregivers to share things they would be less inclined to discuss in a one-on-one interview, she added.
“If I’m a caregiver and I’m really struggling and someone is asking me about my caregiving experience, I might be less inclined to tell them just how rough it is and how much trouble I’m going through,” she said.
Danilovich added that the authors also found tweets using dementia as an insult, which she said stigmatizes the condition and prevents people from getting early care. The concern is that people may not actually know the symptoms of dementia, she said.
Reda Al-Bahrani, a McCormick graduate student who worked on the project, said the team collected tweets and categorized them based on certain keywords and phrases related to Alzheimer’s, dementia and caregiving. According to the study’s abstract, the tweets referenced the hashtags #alzheimer’s, #Alz, #dementia and #dimentia.
Agrawal stressed that while the data was categorized manually for this study, the ability to identify caregivers from tweets “paved the way” for the authors to automate the process in the future.
“If there is an automated way to identify these people at risk, then hopefully we can reduce the burnout significantly,” he said. “It’s more important than ever right now.”