Home-Based Care and Mental Health during a Disaster: A Qualitative Analysis

Study Sample and Recruitment

The study sample included home health care agencies located in areas affected by Hurricanes Irma and Harvey. Participant inclusion criteria included English speaking, registered providers that provided care in a home-based setting during either of these hurricanes. The definition of home-based setting included community-based residences such as houses and apartment complexes but not institutional settings such as skilled nursing facilities or assisted living facilities.

The criteria for county inclusion were derived from counties with Federal Emergency Management Agency (FEMA) disaster declarations for these two hurricanes and further narrowed to those that also received Individual Assistance funding. A priority list of home health care agencies was generated by using open-source FEMA data, as well as race/ethnicity and socioeconomic characteristics by county. Counties with higher statistics of residents who describe themselves as Black, Hispanic, Asian, and “other” were also considered priority. This was done in order to learn from the experiences of HBCP working in counties with historically underserved patient populations. Priority was also given to counties with a median household income closest to the US Department of Health and Human Services Federal Poverty Guidelines, which were $61,372 in 2017.

Enrollment

Home health care agencies meeting study criteria discussed above were contacted by telephone, using a priority calling list of agencies based on criteria stated above. Two trained research assistants (LK, SD), both of whom have a background in research and health care, contacted each agency. After a fifth unanswered call or request for a call back, the agency was removed from the active calling list. Out of 164 agencies called, focus groups were held with a convenience sample of participants from five agencies.

Data Collection

Five in-person focus groups with a total of 25 participants were held in the two disaster-affected locations: Southern Florida (Hurricane Irma, September 2017) and the greater Houston area (Hurricane Harvey, August 2017). The focus groups took place in January, October, and November of 2019. The groups were moderated by the principal investigator who has doctoral level training in qualitative research methods (SB) and a trained research assistant (LK) as the secondary moderator. The secondary moderator was responsible for the audio recording, consent and incentive documents, and note taking. The focus groups lasted approximately 60 minutes, starting with an introduction to promote a comfortable environment. Written informed consent was obtained. The groups were then conducted in a semi-structured fashion with open discussion from questions and prompts. Participants received a $50 gift card as a compensation for their time and expertise. Study rigor was supported by use of the same semi-structured interview guide across all focus groups, use of a moderator with a background in nursing who was familiar with the research problem and topics of discussion, and selection of moderator who was not in a position of authority over focus group participants (Loeb et al., 2006; Morrison & Peoples, 1999; Prince & Davies, 2001; Smithson, 2000; Stewart et al., 2007).

Data Analysis

Once the focus groups were completed, the process of coding and analysis began. The focus group conversations were digitally recorded. These audio recordings were sent to a professional transcription service through a secured platform. Identifying information, such as facility, participant, clinician, or family member names, was removed from each transcript.

Data were analyzed using an abductive analytic approach (Timmermans & Tavory, 2012). This method decreases the likelihood of bias by combining inductive and deductive approaches to allow for deliberate examination of a range of explanations. The Social Ecological Model (Bronfenbrenner, 1981) and the Disaster Management Cycle (Neal, 1997) provided theoretical guidance for data analysis. The Social Ecological Model is a framework that examines influences on the health of individuals at interpersonal, community, societal, and policy levels. The Disaster Life Cycle provides a framework for effective interventions to address disasters’ effects through four phases: preparedness, mitigation, response, and recovery. Two trained research assistants (LK, SD), each of whom has a background in research and health care, individually read the transcriptions and developed codes informed by both frameworks. Forty-nine codes were initially developed. The codes were refined into themes that represent the over-arching concepts extracted from the data. Study rigor was supported by the use of a code book, a detailed audit trail, and weekly meetings between members of the coding team (LK, SD, and SAB) to review and arbitrate coding differences.

Demographics

A total of five focus groups were conducted, each with a total of between two and six participants. A total of 25 participants completed focus groups. Participants reported having worked in their profession for between 3.5 and 39 years (mean 16.5 years). Table 1 summarizes participant characteristics.

OPEN IN VIEWER

Table 1.

Participant Characteristics.

Characteristic	n (%)
Male gender	3 (12)
Age
18–30	3 (12)
31–45	8 (32)
46–60	10 (40)
61–75	1 (4)
No response	3 (12)
Race
Black	1 (4)
White/Caucasian	21 (84)
Hispanic/Latinx	3 (12)
Asian	0 (0)
Role on care team a
Physical therapist	1 (4)
Registered nurse	13 (52)
Licensed vocational nurse	4 (16)
Administrator	5 (20)
No response	5 (20)
Length of time in occupation (years)
0–5	1 (4)
6–10	8 (32)
11–20	5 (20)
21+	7 (28)
No response	4 (16)
Highest level of education
Diploma/certificate (i.e., LVN)	4 (16)
Associates’ degree	5 (20)
Bachelor’s degree	10 (40)
Master’s degree	1 (4)
Doctorate	1 (4)
No response	4 (16)

Study Themes

Four major themes emerged. The themes were (a) providers’ disaster-related mental health needs; (b) HBCP resilience; (c) tension between caring for self and family and caring for patients; (d) adapting to meet patients’ holistic needs during and after disaster. Each theme is described below.

Implications

We suggest further quantitative study of the effects of disasters on HBCPs’ mental health to guide future policy decisions regarding the allocation and accessibility of mental health care resources for health care workers providing care in the home setting during disasters. Given the documented risks that disasters, such as hurricanes, pose to mental health, we propose further study of potential mental health screening and treatment interventions that may benefit HBCP in the context of disaster (Shultz & Galea, 2017). The presence of resilience, or processes of adapting well to adversity, trauma, tragedy, threats, or significant sources of stress (American Psychological Association, 2022), in providers is a protective factor from negative mental health impacts, and resilience can be increased by training, preparedness, managerial practices, and post-disaster counseling (Turner, 2015). An opportunity exists for home care agencies to consider ways to increase provider resilience, which may help to support HBCPs’ mental health during the post-disaster recovery process. One example might be personal wellness check-in calls to HBCP in the aftermath of a disaster.

We also suggest further study of potential strategies to reduce mental health stigma and burnout among HBCP providing care during disasters. The COVID-19 pandemic has highlighted burnout as a critical issue in the nursing profession (National Academies of Sciences, 2019; National Academy of Medicine, 2021). HBCP, in our study, described stressful situations they encountered while providing care in the home during and after major hurricanes that may lead to health care provider burnout. However, challenges still exist surrounding health care providers’ willingness to speak out about mental health care needs. Although we have suggested initiatives that may improve mental health wellness among HBCP, reducing occupational stigma around receiving mental health care is imperative.

In addition to identifying a potential need for better mental health supports for HBCP working during and after disasters, our study highlighted the need for mental health care strategies HBCP could use in the care of their patients. Although we described examples of providers who displayed adaptability in order to meet patient needs in the context of major hurricanes, HBCP still need additional support in the form of formal training on mental health care. Trainings using case identification and simple intervention skills, such as brief interventions, may be beneficial to HBCP attempting to provide mental health support to patients during and after disasters (Sundram et al., 2008). Other examples include psychologist first aid containing the “five essential elements: safety, calming, connectedness, self-efficacy, and hope” (North & Pfefferbaum, 2013; Shultz & Forbes, 2014) and critical incident stress management (CISM) (Everly, JrFlannery, & JrEyler, 2002 2003). To facilitate broader uptake and understanding of CISM among HBCP, it may be beneficial to require a one-time CISM course as part of HBCP continuing education requirements.

Our initial research indicates that home-based care agencies may benefit from the inclusion of a mental health component in their emergency preparedness plan (Emergency Nurses Association, n.d.). These mental health-specific resources may assist HBCP in the early detection of patients with mental health concerns although future research is needed to determine the optimal approach. Standardized screening tools can be easily administered, provide valuable insight into an individual’s risk for psychological problems, and facilitate referral to a mental health provider (McFarlane, 1986; North & Pfefferbaum, 2013). Other resources that may be beneficial to include in an emergency preparedness plan include existing resources such as the Substance Abuse and Mental Health Service’s Administration Disaster Mobile App (SAMHSA, 2022).

Limitations and Future Research

This study included a total of 25 HBCP participants drawn from two disaster-affected locations. Participants shared insight into their experiences providing care after hurricanes; however, the limited sample size makes it difficult to generalize our findings beyond communities with similar demographics and disaster types. Additionally, most of our participants were white women, excluding valuable insights from HBCP of color and those identifying as men. Future studies would benefit from a focus on the perspectives of HBCP from historically marginalized groups including Black, Indigenous, and other people of color to provide diverse views on health care delivery in disaster settings. Because spirituality can impact individuals’ decision-making processes and coping mechanisms during major stressors, we were surprised that themes related to spirituality were not identified in our analysis and believe that this represents an area worthy of future study in disaster settings. Finally, there is a risk for recall and recency bias in our data, as data collection occurred 2 years after the hurricanes of interest. However, recruitment of health care providers during active disasters remains an important methodological challenge. As our data show, health care personnel are often focused on the tension between providing high quality patient care and maintaining the safety of their homes and families (Bell et al., 2020), which limits time available for participation in research. For this reason, we accept the limitations associated with our study’s time frame and instead focus on the insights gained from in-depth focus groups with HBCP, who may not have been able give so much of their time in the immediate aftermath of both hurricanes.