As stated by the US Federal Emergency Management Agency, efforts to build community resilience can be complicated because communities must tailor outcomes to the individual needs of their residents. 1 Older adults (ie, individuals ≥ 65 years of age) make up one of the fastest-growing segments of many communities in the United States. 2 Therefore, their unique needs and vulnerabilities, including physiological, psychological, and socioeconomic factors, have been the topic of discussion in several studies of disasters.Reference Aldrich and Benson 3 - Reference Wang 9
The growth of the older adult population, combined with an increased awareness of the economic viability of providing care in patients’ homes, has led to an increased growth in home health care delivery. The US Centers for Disease Control and Prevention (CDC) estimated the presence of 12 400 home health agencies (HHAs) in the United States in 2014, 10 with over 1 million individuals receiving home health care each day. 11 HHAs have an opportunity to play a unique role in community mitigation and preparedness efforts, as they work directly with some of the most vulnerable members of the population directly in their homes.Reference Claver, Wyte-Lake and Dobalian 12 - Reference Daugherty, Eiring and Blake 16
Veterans served by the US Veterans Health Administration (VHA) have poorer health status than that of older people enrolled in Medicare managed care.Reference Selim, Berlowitz and Fincke 17 The VHA’s Home-Based Primary Care (HBPC) program, which serves 53 000 older veterans (mean age 76.5) throughout the United States,Reference Karuza, Gillespie and Olsan 18 is a subset of home health care programs. In providing care to older veterans who present with complex chronic conditions, the VHA HBPC program serves a subset of the most vulnerable members of the US population, and can expect to do so during disaster situations. However, little research has been conducted to better characterize the needs of and resources available to HBPC veterans during times of emergency or disaster. Although HBPC programs, like all HHAs, are tasked with supporting the preparedness efforts of their patients, there is a wide range of practices in the fieldReference Wyte-Lake, Claver and Dalton 19 and practitioners often desire guidance on how to best approach these efforts.Reference Claver, Wyte-Lake and Dobalian 12
A recent Medicare ruling, in its expansion of HHA emergency preparedness requirements, explicitly states that HHA activities and resources are necessary for effective community emergency preparedness planning. 20 Accordingly, home health agencies could be tapped as a resource to maximize patient resiliency and be key players in local community resilience. This ruling discusses a general expectation that HHA practitioners assist their patients in preparedness planning. Although this may be true for a certain percentage of practitioners, research shows that the degree to which a practitioner covers preparedness with his or her patient depends on how comfortable that practitioner is with his or her own knowledge about the topic.Reference Claver, Wyte-Lake and Dobalian 12 We are unaware of any literature evaluating the extent to which disaster preparedness is covered with patients in home health programs.
Through the testing of an HBPC Patient Assessment Tool for Disaster Preparedness, baseline characteristics of veterans served by the HBPC program were collected. Additionally, this study aimed to conduct an in-depth examination of the types of disaster preparedness education provided to veterans in HBPC and how patients’ characteristics might be associated with whether they receive information.
DESIGN AND METHODS
The HBPC Patient Assessment Tool was created by the authors to address the lack of a systematic assessment tool for disaster preparedness among HBPC patients.Reference Wyte-Lake, Claver and Dobalian 21 The assessment tool is based on best practices identified through a review of current assessment items and standard operation procedure documents from HBPC sites. It contains items regarding veteran characteristics (eg, bed bound, cognitive impairment, communication difficulties), oxygen use and safety, use of medical equipment, use of safety equipment, and patient preparedness education provided during the home visit. Practitioners were instructed to derive patient characteristics per the patient’s medical records. The provision of patient preparedness education was the primary outcome measure for this study and includes 9 education-related items. These items fall into 2 categories: (1) primary or “life essentials” resources, including ability to evacuate, how to activate 911 services, and education on the importance of working carbon monoxide and smoke detectors; and (2) secondary or “good to have” resources, including discussion of emergency plan, emergency kit present, advance directive/Physician Orders for Life-Sustaining Treatment (POLST) available, copies of important documents in one place, and information on emergency shelter registration and special transport.
VHA HBPC program managers from 10 sites in 8 states designated specific practitioners (eg, nurse practitioner, social worker, physical therapist) to field the Patient Assessment Tool with all patients seen over the course of a 3-week period (April to October, 2017). At the conclusion of the 3-week data collection period, the completed Patient Assessment Tools were deidentified and sent back to the project team. The project team inputted the collected data from the Patient Assessment Tool and generated site-specific infographic reports that were presented to the sites via a brief follow-up interview approximately 1 month post–data collection. In addition to site-specific data analysis, aggregate data from all sites were analyzed.
From the 10 participating HBPC program sites, 757 patient assessments were returned. The number of patients from individual sites ranged from 29 to 155. Three patients were dropped because of unclear responses. The analytic sample size for this study is 754, except for the risk classification data, which is 567 because 1 site failed to complete the relevant section of the Patient Assessment Tool.
Descriptive statistics, including mobility level, cognitive issues, and behavioral issues, were used to understand patient profiles. Bivariate analysis using chi-square tests was conducted to test the significance of the association between patient preparedness education provided and patient risk category. To test the association between patient characteristics and preparedness education items, for each education item, 9 separate logistic regression analyses were conducted where all 8 patient characteristics were entered into each regression model as independent variables. The results of all 9 logistic regression analyses are presented in this paper. The significance level was set at P<0.05. All quantitative statistical analysis was performed using STATA v.12 (Stata Corp, College Station, TX).
The VA Greater Los Angeles Healthcare System Institutional Review Board approved this study as a quality improvement initiative, and the study was therefore exempt from human subject approval.
RESULTS
Patient Characteristics and Health Conditions
Table 1 shows the characteristics of the patient population included in the analysis. Note that identifiers such as age, sex, and race were not collected. Sites were directed to use their own risk classification system to rate each patient “high,” medium,” or “low” based on how much support he/she would require in case of an emergency or a major disaster. About 40% were rated low, 44% medium, and 16% high risk. Practitioners were then instructed to check off a series of yes/no statements about patient characteristics derived from the day’s evaluation as well as the patient’s medical records. As noted in Table 1, these questions ranged from “Patient is chair/bed bound” to “Social support readily available.” Eighty-five percent of the patients had mobility issues: 30% were chair bound (unable to ambulate), while 55% required an assistive device (to ambulate). Results also showed that 25% of the patients were oxygen dependent and 40% did not have social support readily available. In terms of cognitive functioning, the most commonly noted issue was general cognitive impairment (33%), followed by communication limitations (20%), mental illness (11%), and evidence of behavioral disturbance (7%).
Table 1 Patient Characteristics, US Veterans Health Administration Home-Based Primary Care Patient Assessment Tool for Disaster Preparedness Study, United States, 2017 (n=754)
Disaster Preparedness Education
Various types of disaster preparedness education were discussed with patients and their caregivers. Overall, the most commonly discussed items included how to activate 911 services (87%), the ability to evacuate from home by identifying the number of exits (86%), having an emergency plan (83%), and received instructions that emergency preparedness recommendations were included in the patient’s HBPC handbook (79%). Items such as education about an advance directive/POLST (66%), the importance of working CO and smoke detectors (66%), and having important documents in one place (66%) were discussed with patients with decreased frequency. The items least likely to be discussed were the importance of having an emergency kit (56%) and the provision of information on emergency shelter registration and emergency specialty transportation (44%) (data not shown).
Patient Risk Category and Disaster Preparedness Education
The 9 preparedness education items discussed with HBPC patients were compared by patient risk category where patients in the high or medium risk category were grouped together to distinguish from HBPC patients who needed less support during disasters (ie, rated as low risk). Table 2 displays the results of these comparisons in the order of the preparedness education most likely received by the high/medium risk patients. The percentages for the 2 groups (high/medium vs low) for all 9 items are listed. When compared to the low risk group, HBPC patients in the high/medium risk group were more likely to receive preparedness education materials for 6 items: how to activate 911 (90% high/medium vs 80% low, P<.001); emergency plans (87% high/medium vs 76% low, P<.001); importance of working CO and smoke detectors (68% high/medium vs 56% low, P<.01); having copies of important documents in a safe place (60% high/medium vs 53% low, P<.05); emergency kit (56% high/medium vs 46% low, P<.05); and information on emergency shelter registration and emergency specialty transportation shelter (51% high/medium vs 33% low, P<.001).
Table 2 Patient Risk Category and Preparedness Education, US Veterans Health Administration Home-Based Primary Care Patient Assessment Tool for Disaster Preparedness Study, United States, 2017 (n=567)
Abbreviations: CO, carbon monoxide; HBPC, home-based primary care; POLST, Physician Orders For Life-Sustaining Treatment.
Patient Characteristics and Disaster Preparedness Education
For the next set of comparisons, the 9 preparedness education items were grouped into 2 categories: primary (“life essentials”) and secondary (“good to have”) resources. The primary resources include 3 items: ability to evacuate, how to activate 911, and education on the importance of working CO and smoke detectors (see Table 3), and the secondary resources include the remaining 6 items (see Table 4). Table 3 presents results from 3 logistic regression analyses for the primary resources. For the first education item, only 1 patient characteristic showed significant association. After we controlled for all 8 patient characteristics, the results indicated that patients with mental illness were more likely to receive education on how to evacuate (OR: 2.7239, P<.05). For the second education item, 3 items showed significant association: patients dependent on oxygen (OR: 2.3218, P<.05) and patients with mental illness (OR: 2.8210, P<.05) were more likely to learn how to activate 911, whereas patients with cognitive impairment were less likely to receive education on how to activate 911 (OR: 0.5678, P<.05). For the last primary resource item, 4 patient characteristics showed significant associations. Chair/bed bound patients (OR: 1.5593, P<.05) and oxygen dependent patients (OR: 1.9356, P<.01) were more likely to receive educational information on having working CO and smoke detectors, whereas patients with communication limitations (OR: 0.6379, P<.05) and patients with social support (OR: 0.6247, P<.01) were less likely to receive educational information on CO and smoke detectors.
Table 3 Patient Demographics and Preparedness Education, Primary Resources (ie, Life Essentials). Odds Ratio of Those Patients With the Specific Patient Characteristic Receiving the Education. US Veterans Health Administration Home-Based Primary Care Patient Assessment Tool for Disaster Preparedness Study, United States, 2017 (n=754)
*P<.05; **P<.01
Note: Bold represents significant associations. Bold and italics represent significant associations where presence of the characteristic makes it less likely that the patient received preparedness education. All logistic regression models were significant (P<.01), except for Ability to Evacuate (P=.0514).
Table 4 Patient Demographics and Preparedness Education, Secondary Resources (ie, Good to Have). Odds Ratio of Those Patients With the Specific Patient Characteristic Receiving the Education. US Veterans Health Administration Home-Based Primary Care Patient Assessment Tool for Disaster Preparedness Study, United States, 2017 (n=754)
Abbreviations: HBPC, home-based primary care; POLST, Physician Orders for Life-Sustaining Treatment.
*P<.05; **P<.01; ***P<.001
Note: Bold represents significant associations. Bold and italics represent significant associations where presence of the characteristic makes it less likely that the patient received preparedness education. All 6 logistic regression models were significant (P<.01).
Table 4 presents results from 6 logistic regression analyses—1 for each of the 6 secondary preparedness items. For the most part, the results show a similar pattern, in which chair/bed bound patients and oxygen dependent patients were more likely, whereas patients with social support were less likely to receive secondary preparedness educational resources. For example, after controlling for the effects of all 8 patient characteristics, chair/bed bound (OR: 1.5585, P<.05) and oxygen dependent patients (OR: 1.5502, P<.05) were more likely, whereas patients with social support were less likely (OR: 0.6534, P<.01), to receive information on emergency kits. For the advance directive/POLST education item, however, there is a reverse pattern, where patients with social support, which was the only significant patient characteristic, were more likely (OR: 1.6154, P<.01) to receive the advance directive/POLST education. For patients with mental illness and patients with cognitive impairment, the findings are mixed. In the case of having copies of important documents in 1 place, patients with mental illness were less likely (OR: 0.5843, P<.05) to receive this education item. On the other hand, for the last secondary resource item, patients with mental illness (OR: 2.4403, P<.05) and patients with cognitive impairment (OR: 1.5728, P<.05) were more likely to receive instructions that emergency preparedness recommendations are located in the HBPC handbook.
DISCUSSION
In this study, we sought to examine the association between patient characteristics and whether disaster preparedness education was provided by practitioners in the VHA HBPC program. As noted by Gamble et al,Reference Gamble, Hurley and Schultz 4 the increase in managed care has led to more elderly individuals remaining in their homes and a corresponding increase in the need for appropriate, accurate, and trusted information about disaster preparedness. The new Medicare and Joint Commission policy guidelines recognize the need for home health agencies to increase their involvement in improving mitigation and preparedness amongst their patients, 20 yet research has found that program staff and leadership desire more direction in how to make their preparedness protocols more robust.Reference Wyte-Lake, Claver and Der-Martirosian 15 , Reference Wyte-Lake, Claver and Dalton 19 Few articles have evaluated the robustness of tools developed or adopted to support preparedness in home health programs,Reference Al-Rousan, Rubenstein and Wallace 5 and none that we are aware of have evaluated the extent to which disaster preparedness is actually covered with patients. Through evaluation of the HBPC Patient Assessment Tool, this survey provides an opportunity to begin to identify strengths and challenges in supporting the preparedness activities of patients served by home health programs such as the VHA HBPC program.
There is strong evidence that older adults, particularly those served by home health agencies, are more vulnerable in times of disaster because of a series of chronic comorbidities as well as poor social support.Reference Gamble, Hurley and Schultz 4 - Reference Dostal 6 , Reference Khorram-Manesh, Yttermyr and Sörensson 14 Veterans receiving care at the VHA HBPC program are a subset of the aging American population. Our results highlight the need for understanding how to best support these individuals, as well remembering the role of the home health program as a preparatory entity rather than a responder.Reference Claver, Wyte-Lake and Dobalian 12 Nonetheless, HBPC programs, like other health programs that engage with patients in their homes, have a potential unique advantage in preparing patients for disasters. Having the opportunity to provide direct education and the ability to tailor the message to the patient’s home environment, home health practitioners can act as trusted advisors, a role that has been shown to provide more valued messaging to vulnerable populations such as homebound patients.Reference Eisenman, Cordasco and Asch 22 , 23
There are marked inconsistencies in the comprehensiveness and type of preparedness education disseminated by home health programs.Reference Wyte-Lake, Claver and Der-Martirosian 15 , Reference Wyte-Lake, Claver and Dalton 19 Practitioners in the field can often feel they have an enormous amount of information to cover with patients in a very short amount of time. One of the proposed benefits of the Patient Assessment Tool is to provide a user-friendly but quick and comprehensive checklist of preparedness items to be reviewed by the practitioner with the patient. Our results show that certain items on this list are more frequently discussed. This finding might be due to the fact that certain education topics are easier to discuss or more easily remembered by patients. For example, most adults are aware of 911, and therefore a simple review of how to activate 911 services is not much more than a straightforward reminder. Similarly, reviewing how to evacuate a home or the importance of having an emergency plan requires few additional resources from either the patient or the practitioner. These are not complicated action items on the part of the patient. The study findings indicate that some preparedness topics, such as putting together a home emergency kit, which for this population is often an economic challenge, and information on emergency shelter registration and emergency specialty transport, which may not be readily available to the practitioner in the field, were less likely to be discussed. Previous research has shown that not knowing where to evacuate to and challenges with transportation were primary barriers to evacuation in large scale natural disasters.Reference Eisenman, Cordasco and Asch 22 Improved provision of this information to home health programs, which could then disseminate the information to patients via practitioners, could support mitigation efforts for this vulnerable segment of the population.
Three associations between patient characteristics and dissemination of preparedness items stand out in our study as noteworthy to help create a better understanding of how HBPC practitioners can effectively provide preparedness education to some of their most vulnerable patients. Previous studies have shown risk categorization systems play an important role in how programs triage their patients in the case of an emergency.Reference Zane and Biddinger 24 Our results show that a patient’s risk category was associated with the likelihood of receiving disaster preparedness education and could possibly influence a practitioner’s decision-making process when it comes to providing key pieces of preparedness information. As previously mentioned, during an average medical visit, practitioners often find themselves pressed for time to cover emergent medical concerns. For patients presenting with multiple complex problems, such as is found in the HBPC population, covering items outside of the immediate medical issue becomes even more challenging.Reference Fiscella and Epstein 25 And yet, for 6 of the 9 educational topics, preparedness education was actually more likely to be disseminated to the higher risk patients. Examining vulnerability from another perspective, the results also show that when a patient presents with a physical limitation, such as severe ambulation limitation or oxygen dependence, there is an increased association with receiving preparedness education. Once again, we see the practitioner highlighting the importance of sharing preparedness education with their most high-risk patients. Both sets of associations highlight the unique role of the home health practitioner in supporting the care for individuals identified to be some of the most vulnerable members of our society. It is possible that, compared to less complex patients, the relatively greater vulnerability of these patients brings preparedness to the forefront of practitioners’ concerns and makes them more likely to discuss these actions with these patients. Lack of social support, sometimes seen as social isolation, is an additional vulnerability common among the elderly.Reference Eisenman, Cordasco and Asch 22 Our findings, which suggest that patients with no social support were more likely to receive preparedness education, underscore 1 of the unique values of home health programs: connecting with individuals who have limited community networks and providing them with resources and support not otherwise available.
On the other hand, 3 items were not associated with risk category. Two of these, reminding patients of the HBPC handbook and reviewing the need for an advance directive/POLST, are educational items covered irrespective of the fact that they are also related to preparedness education. The ability to evacuate the home, the third item not associated with risk category, could perhaps be considered one of the most basic and fundamental pieces of preparedness when considering the homebound population, as it was 1 of the topics most frequently covered with all patients and has no association with risk category or patient characteristics. One of the main tenants of this population is having difficulty leaving the home, so addressing this underlying risk factor is critical in providing appropriate preparedness education.
One of the most important preparedness challenges for the homebound elderly population is the number of comorbidities with which they present, including the growing presence of cognitive impairment.Reference Toner 8 , Reference Oliva, Wexler and Gullickson 26 There is a growing body of literature addressing the unique needs of individuals presenting with cognitive impairments such as dementia, which can carry with it symptoms such as communication limitations and behavioral disturbances. 27 Our findings are mixed for these patients. In some cases, education is more likely to be disseminated to patients with cognitive impairments, while in others it is less likely to be shared. Some of these items, such as reminding patients that the preparedness recommendations are in the VHA HBPC handbook, may seem straightforward for much of the population but are possibly more of a challenge for patients with cognitive impairment. For those items where practitioners are less likely to disseminate preparedness education (eg, education on importance of CO and smoke detectors and copies of important documents in 1 place), our findings underscore the challenges of caring for this population and providing information effectively to either them or their caregivers. There are a limited number of studies that discuss the creation of specific resources for individuals (and the families caring for them) who present with dementia,Reference Toner 8 , Reference Oliva, Wexler and Gullickson 26 but organizations such as the Alzheimer’s Association, the Red Cross, and the US Administration on Aging have developed tools and resources to specifically support disaster preparedness activities for people with dementia and their caregivers, as well as training modules for first responders to be better aware of how to work with individuals with dementia at the time of an emergency or a disaster. Many items overlap with standard preparedness practices such as putting together emergency kits and having family emergency plans in place. Recommendations unique to this population include ensuring that persons with dementia are enrolled in programs to assist with identification and reunification if they become lost or separated from their family and making sure to pack calming objects in the emergency kit to help a disoriented person with dementia feel calmer in unfamiliar circumstances. 28 Training home health practitioners about these resources, as well as including the specific resources, tips, and tools in the disaster preparedness materials that home health programs distribute, would be a crucial step in supporting the preparedness of a growing segment of our population.
Limitations
Pilot sites were asked to assess every patient seen during a 3-week period. This may be a relatively random sample, because the data collection took place over a 3-week period with patients that happened to be scheduled during that period. However, it is possible that the sample may be slightly skewed toward inclusion of high-need patients that are seen more frequently (such as those with Foley catheters or needing wound care). Although practitioners were instructed to use patient medical records to derive patient designation, variability exists regarding how practitioners assessed these characteristics in particular with cognitive functioning items. Moreover, the risk categorization item within the Patient Assessment Tool instructed practitioners to defer to the sites’ own categorization system. Prior research has shown that risk categorization systems are not systematic between HBPC programs, nor arguably within HBPC programs.Reference Wyte-Lake, Claver and Dobalian 21 Therefore, there is no consistent definition for how patients were identified as high, medium, or low. Our study was conducted with VHA HBPC providers, and the extent to which our findings generalize outside the VHA is unknown. Although VHA facilities are not covered under Medicare, all VHA facilities must be certified by The Joint Commission. Thus, it is likely that our findings would not markedly differ for non-VHA home health organizations.
CONCLUSIONS
Arguments have been made that home health programs should be increasingly engaged in community resilience efforts, and the results of this study support a need for more extensive education to homebound patients concerning disaster preparedness. Our findings, which demonstrate that practitioners are providing basic preparedness education to some of their most vulnerable patients, underscore the unique opportunity for home health agencies to support populations cared for in the home regarding disaster preparedness. It is unlikely that homebound patients are receiving this information elsewhere. Home health agencies, with the aid of a tool such as the HBPC Patient Assessment Tool, can play an important role in their patients’ preparedness.
Acknowledgements and Disclosures
The authors would like to thank Roger Wasserman and Alexis Gequillo for their extensive support with data cleaning. This material is based upon work supported by the by the US Department of Veterans Affairs (VA), Veterans Health Administration, Office of Patient Care Services. The views expressed in this article are those of the authors and do not necessarily reflect the position or policy of the VA or the United States government.
Author Contributions
T. Wyte-Lake drafted the article and led the writing. Both T. Wyte-Lake and M. Claver led the conceptualization of the research, development of the Patient Assessment Tool, collection of the data, and the interpretation of the data. C. Der-Martirosian supported the statistical analyses and interpretation of the data. D. Davis supported the survey design and data collection. A. Dobalian had ongoing involvement in the design of the study. All authors read, edited, and approved the final article.
