Examining the Impact of COVID-19 on People With Dementia From the Perspective of Family and Friends: Thematic Analysis of Tweets

Background The COVID-19 pandemic is taking a serious toll on people with dementia. Given the rapidly evolving COVID-19 context, policymakers and practitioners require timely, evidence-informed research to address the changing needs and challenges of people with dementia and their family care partners. Objective Using Twitter data, the objective of this study was to examine the COVID-19 impact on people with dementia from the perspective of their family members and friends. Methods Using the Twint application in Python, we collected 6243 relevant tweets over a 15-month time frame. The tweets were divided among 11 coders and analyzed using a 6-step thematic analysis process. Results Based on our analysis, 3 main themes were identified: (1) frustration and structural inequities (eg, denied dignity and inadequate supports), (2) despair due to loss (eg, isolation, decline, and death), and (3) resiliency, survival, and hope for the future. Conclusions As the COVID-19 pandemic persists and new variants emerge, people with dementia and their family care partners are facing complex challenges that require timely interventions. More specifically, tackling COVID-19 challenges requires revisiting pandemic policies and protocols to ensure equitable access to health and support services, recognizing the essential role of family care partners, and providing financial assistance and resources to help support people with dementia in the pandemic. Revaluating COVID-19 policies is critical to mitigating the pandemic’s impact on people with dementia and their family care partners.


Introduction
The COVID-19 pandemic is taking a serious toll on people living with dementia. Beyond having 1 of the highest COVID-19 mortality rates [1], people with dementia have experienced challenges to their mental, physical, and cognitive health [2]. In attempts to mitigate the spread of the COVID-19 virus, governments have implemented various infection control measures, such as public curfews, social distancing protocols, regional lockdowns, required masking, quarantines, visitation bans, and travel restrictions [3].
Studies show that people with dementia and their family care partners (eg, a spouse, common-law partner, or other family member who provides care to a person with dementia) [4] have experienced major barriers to accessing health services and community-based supports during the first wave of the pandemic. More specifically, challenges to service use included reduced or terminated adult day programs, respite services, cognitive rehabilitation therapy, friendly visitor services, senior center programs, and limited knowledge of or access to technology [5][6][7]. In addition, people with dementia and their family care partners experienced a range of unintended consequences from COVID-19 lockdowns and infection control measures, such as social isolation, loss of informal supports, and care partner fatigue and burnout [8][9][10][11].
As the pandemic continues to unfold, new coronavirus variants, such as Omicron, continue to create challenges for health and support services [12], and future variants are expected [13]. With this rapidly changing situation, policymakers and practitioners require timely and comprehensive data to address the changing needs of at-risk population groups, including people living with dementia. More specifically, research-based evidence is needed to inform COVID-19 policies to help mitigate the impact of the pandemic on people with dementia and their family care partners. However, new publications on dementia continue to report findings primarily based on the early stages of the pandemic with little knowledge of the COVID-19 barriers over time [5,9,14,15].
Given the rapidly evolving COVID-19 context, conducting in-person research remains a critical challenge with frequent changes to public health orders ranging from regional lockdowns to travel bans [16]. With a daily average of 500 million tweets, Twitter provides a novel and timely means for people to share their COVID-19 experiences [17]. Specifically, health care providers, policymakers, family members, and friends of people with dementia have been using Twitter to share their opinions, experiences, and concerns about the pandemic [18,19]. Accordingly, the purpose of this study was to use Twitter to examine the COVID-19 impact on people with dementia from the perspective of their family members and friends.

Tweet Extraction
Tweets were collected using Twint, an advanced scraping tool that enables users to scrape tweets without the use of Twitter's application programming interface. This feature enables Twint scrapers to avoid certain restrictions, such as the number of tweets scraped, the frequency and period of scrapes, and the requirement of a Twitter account [20]. Thus, Twint was used to asynchronously scrape tweets for the period of September 8, 2020, to December 8, 2021.
Given that existing Twitter studies already explore the early stages of the pandemic (ie, February 15-September 7, 2020, and March 17-24, 2020) on people with dementia [19,21], our study's time frame focused further into the pandemic (ie, September 8, 2020, to December 8, 2021). Our full set of search terms included "dementia" OR "Alzheimer" used in combination with "COVID-19" OR "COVID" OR "Corona," resulting in 110,528 tweets. We searched for tweets worldwide rather than focusing on tweets only from a certain country. Filters were applied to exclude non-English language tweets, reply tweets, advertising or spam tweets, and unrelated tweets, such as political tweets about "Trump" or "Biden" (refer to Figure 1). Only original tweets were cited, and replies to tweets were removed as they often were missing information and only contained half of the conversation. These filters are consistent with previous research analyzing Twitter data [18,22]. To improve the likelihood of scraping tweets that described personal COVID-19 experiences of dementia, we excluded tweets that did not include synonyms for familial relationships (eg, "mother," "father," "grandparent") or friends or acquaintances (eg, "neighbor" and "pal"). The remaining 6243 tweets were put into a Microsoft Excel spreadsheet for data analysis.

Data Analysis
Tweets were analyzed using a 6-step thematic analysis process [23]. First, 2 researchers (authors JB and MEO) read and reread 200 tweets to support familiarization with the data. Memos were used to document any new codes or ideas about the data to help supplement and revise an a priori code book [19]. This coding scheme comes from our previous Twitter research on the COVID-19 impact on dementia that was conducted at the early stages of the pandemic (ie, February 15-September 7, 2020) [19]. Second, the 2 researchers refined an existing coding scheme by deleting 2 codes (eg, questioning the cause of death and positive stories of survival) that were not identified within the current data. The final version of the revised codebook consisted of 9 codes: (1) death, (2) fear for the person with dementia's health/well-being, (3) challenges and unmet needs, team meetings using group consensus. Sixth, once the coding was completed, 2 team meetings were held to discuss the patterns and relationships (eg, similarities, differences, overlap) among the codes, and theme piles were used to identify the overarching themes. The theme titles were reviewed and refined through group discussion to ensure clarity of the theme titles and properties.

Rigor
We used 4 measures to support rigor in our study. First, a comprehensive audit trail was kept to document notes about the context of the research, methodological decisions (eg, code book revisions), and the 6-step thematic analysis process [24]. Second, each researcher used reflexive memoing to document interesting findings, similarities, differences, emerging patterns, and relationships throughout the coding process [23]. Third, our research team consisted of multidisciplinary researchers (eg, psychology, community health and epidemiology, computer science) with diverse skills and theoretical perspectives that provided a more insightful and nuanced approach to interpreting our study's findings. Fourth, the first author managed intercoder reliability by randomly reviewing 25% of the codes to cross-check the data analysis, resulting in an intercoder reliability average of 83.4%.

Ethical Considerations
Our research used publicly available tweets posted on Twitter.
No interaction was made with the users who posted the tweets. We removed the Twitter handles and usernames to help protect the anonymity of the tweeters. Consequently, our Twitter research did not require review by an institutional ethics review board.

Results
Drawing on our thematic analysis, 3 main themes were identified: (1) frustration and structural inequities (eg, denied dignity and inadequate supports), (2) despair due to loss (eg, isolation, decline, and death), and (3) resiliency, survival, and hope for the future. Many tweets, however, conveyed complicated reactions to their loss of a loved one who was living with dementia to COVID-19. These reactions, which span anger, helplessness, and a sense of futility, seem to be factors that could interrupt the process of making meaning from loss and thereby disrupt the grieving process and create psychological vulnerability to future mental health challenges (eg, from a depressed mood).

Frustration and Structural Inequities: Denied Dignity and Inadequate Supports
Hes had dementia, a few strokes, and heart surgery, but what finally did him in was a combo of COVID and the selfishness of some I used to call family. This week felt like a long panic attack but with anger. We knew wed lose him, but him being taken feels so much worse.
Some lamented the challenges of concomitant dementia and active COVID-19 infection requiring hospitalization or medical interventions that made the loss even more challenging to process, accept, and make meaning from. Many expressed anger when describing their experiences with losing a loved one with dementia due to COVID-19. Although anger was reported earlier in the pandemic related to mask usage and lack of social distancing [19], as the medical science has evolved to illuminate ways to prevent or mitigate COVID-19 infection, the nature of the tweeters' anger has similarly evolved. For some, the anger was directed at those who were unvaccinated, those who interacted without wearing masks, and government-led mitigation strategies. These tweets convey a sense of helplessness and struggles to find meaning from deaths that these tweeters felt could have been avoided.

This covid shit is no joke at all. My mom mom passed this morning. After a total of 20+ years living with a heart AND kidney transplant + recently being diagnosed w/ dementia, my gmom fought long and hard just to be taken out by something that could have been prevented…
The experiences of losing someone with dementia who resided in long-term care were described. Many of these conveyed the sense of helplessness and sense of tragedy at the numerous cases of COVID-19 in long-term care, further compounding the sense of loss and likely to interfere with the process of grieving.

My paternal grandfather passed away in October, 1 month after being admitted to a nursing home rampant with COVID b/c Drs decided my 88 year old grandmother could not longer care for him and his
worsening dementia at home. It was the only facility their insurance would cover...

I left mum 13/3 in her dementia care home -she died 1/4 one of many who died after hospital discharged covid positive patients into homes. The voice of the voiceless #neverforgotten
Disrupted grieving rituals were also commonly described. Adherence to grieving rituals helps with reminiscence through a shared experience with others [25]. Sharing the loss with others creates a sense of universality in the experience and is also psychologically healing [26]. The pandemic has disrupted the shared grieving process of gathering at funerals, potentially making the grieving process unduly prolonged and creating psychological vulnerability to future mental health concerns. People described hope by sharing and envisioning their future plans related to the easing of COVID-19 restrictions. For example, many looked forward to reconnecting in person with people with dementia in care facilities. These tweets often described long-awaited anticipation to engage in face-to-face visits, partake in activities, and provide hugs and a human touch. This sense of hope and anticipation to reconnect is highlighted in tweets such as:

Principal Findings
Our study aimed to examine the COVID-19 impact on people with dementia from the perspective of their family members and friends. Understanding COVID-19 challenges is critical to informing targeted policies, programs, and resources to support people living with dementia. Drawing on thematic analysis, our study identified 3 main themes: (1) frustration and structural inequities (eg, denied dignity and inadequate supports), (2) despair due to loss (eg, isolation, decline, and death), and (3) [28], increased confinement and social isolation [29][30][31], reduced physician visits, decreased access to specialist services, and reduced hospital transfers for illness [32]. Consequently, improving COVID-19 outcomes for people with dementia requires addressing structural inequities in health and support services.
Policymakers need to recognize the essential role of family care partners for people with dementia during the pandemic. Many tweets highlighted care partner challenges related to health care navigation, barriers in health care communication, and COVID-19 visitation restrictions that inhibited their support for people with dementia in care facilities (eg, hospitals, emergency departments, medical clinics, and long-term care homes). Moreover, visitation restrictions were often blamed for psychological loss and accelerating cognitive, psychological, and physical decline in people with dementia. Consequently, COVID-19 policies must be loosened to allow family care partners to accompany, access, and provide necessary support to people with dementia in care facilities. In the United States alone, over 16 million people provide unpaid care to people with dementia at an economic value of US $244 billion [33]. Furthermore, numerous tweets identified financial challenges, issues of economic uncertainty, and inadequate resources among people with dementia and their family care partners. These issues of financial hardship among people with dementia and their care partners are reflected in the existing COVID-19 literature [34,35]. Accordingly, policies and programs are needed to provide financial assistance and resources to help support and sustain people with dementia and their family care partners.
Although many tweets highlighted COVID-19 challenges, some tweets also discussed resiliency and hope for the future. Numerous tweets described people with dementia beating the virus and surviving against multiple odds. People also shared a strong sense of optimism and hope related to COVID-19 vaccines and the easing of restrictions. Many of these tweets discussed future plans to engage in face-to-face visits, partake in activities, and provide hugs and a human touch. These findings suggest that resiliency and hope may play an important role for people with dementia and their care partners during times of adversity, such as a pandemic. However, further research is needed to examine this area.

Limitations
Although a rigorous process was undertaken to perform this study, our research includes limitations. For example, our study did not collect any sociodemographic information about the tweeters. Accordingly, it is difficult to make specific inferences or generalizations about our findings in relation to age, ethnicity/culture, and gender or sex. Moreover, our study did not collect data based on the tweeter's country or geographic location.  [36]. Consequently, it is interesting to note that our themes were consistent across our data.
Another limitation of our Twitter study is that it is based on cross-sectional research rather than longitudinal data. For example, longitudinal data over different years would provide greater insight into COVID-19 policy responses, changes, and implications over time. Accordingly, further research with longitudinal data would provide useful information on the temporal aspects of different policies and their related outcomes (eg, social, emotional, cognitive, and physical health).
In addition, our research is limited by our scraping strategy that focused on tweets with familial relationships, friends, and acquaintances. Although we aimed to have an inclusive scraping strategy, it is possible that we may have missed relevant data on the COVID-19 impact on people with dementia. For example, future research could use a broader search filter to ensure no relevant findings were overlooked. Moreover, further Twitter research is needed to focus on the lived experiences of people with dementia during the pandemic.
Lastly, each tweet is limited to Twitter's 280-character limit. Consequently, it is possible that we might be missing important background details or contextual factors needed to fully understand the tweet's meaning. To help mitigate this issue, we held research team meetings to discuss and clarify any coding uncertainties by using group consensus.

Conclusion
Given the rapidly evolving COVID-19 context, policymakers and practitioners require timely, evidence-informed research to address the changing needs of marginalized population groups, including people with dementia. Using Twitter data, the objective of our study was to examine the COVID-19 impact on people with dementia from the perspective of their family members and friends. Our study identified 3 main themes: frustration and structural inequities (eg, denied dignity and inadequate supports), despair due to loss (eg, isolation, decline, and death), and resiliency, survival, and hope for the future.
As the pandemic persists and new variants emerge, people with dementia and their family care partners are facing complex challenges that require timely interventions. However, addressing these challenges is neither a straightforward nor a simple task. Rather, tackling these challenges requires interventions informed by lived experiences and evidence-based research. More specifically, tackling COVID-19 challenges requires revisiting infection control policies to ensure equitable access to health and support services, recognizing the essential role of family care partners, and providing financial assistance and resources to help support people with dementia throughout the pandemic. Re-evaluating COVID-19 policies is critical to mitigating the pandemic's impact on people with dementia and their family care partners.