Natural Disasters in the Americas, Dialysis Patients, and Implications for Emergency Planning: A Systematic Review

Introduction Natural hazards are elements of the physical environment caused by forces extraneous to human intervention and may be harmful to human beings. Natural hazards, such as weather events, can lead to natural disasters, which are serious societal disruptions that can disrupt dialysis provision, a life-threatening event for dialysis-dependent people. The adverse outcomes associated with missed dialysis sessions are likely exacerbated in island settings, where health care resources and emergency procedures are limited. The effect of natural disasters on dialysis patients living in geographically vulnerable areas such as the Cayman Islands is largely understudied. To inform predisaster interventions, we systematically reviewed studies examining the effects of disasters on dialysis patients and discussed the implications for emergency preparedness in the Cayman Islands. Methods Two reviewers independently screened 434 titles and abstracts from PubMed, Scopus, CINAHL, and Cochrane Library. We included studies if they were original research articles published in English from 2009 to 2019 and conducted in the Americas. Results Our search yielded 15 relevant articles, which we included in the final analysis. Results showed that disasters have both direct and indirect effects on dialysis patients. Lack of electricity, clean water, and transportation, and closure of dialysis centers can disrupt dialysis care, lead to missed dialysis sessions, and increase the number of hospitalizations and use of the emergency department. Additionally, disasters can exacerbate depression and lead to posttraumatic stress disorder among dialysis patients. Conclusion To our knowledge, this systematic review is the first study that presents a synthesis of the scientific literature on the effects of disasters on dialysis populations. The indirect and direct effects of disasters on dialysis patients highlight the need for predisaster interventions at the patient and health care system levels. Particularly, educating patients about an emergency renal diet and offering early dialysis can help to mitigate the negative effects of disasters.


Introduction
Natural hazards are elements of the physical environment caused by forces extraneous to human intervention and may be harmful to human beings. Natural hazards, such as weather events, can lead to natural disasters, which are serious societal disruptions that can disrupt dialysis provision, a life-threatening event for dialysisdependent people. The adverse outcomes associated with missed dialysis sessions are likely exacerbated in island settings, where health care resources and emergency procedures are limited. The effect of natural disasters on dialysis patients living in geographically vulnerable areas such as the Cayman Islands is largely under-studied. To inform predisaster interventions, we systematically reviewed studies examining the effects of disasters on dialysis patients and discussed the implications for emergency preparedness in the Cayman Islands.

Methods
Two reviewers independently screened 434 titles and abstracts from PubMed, Scopus, CINAHL, and Cochrane Library. We included studies if they were original research articles published in English from 2009 to 2019 and conducted in the Americas.

Results
Our search yielded 15 relevant articles, which we included in the final analysis. Results showed that disasters have both direct and indirect effects on dialysis patients. Lack of electricity, clean water, and transportation, and closure of dialysis centers can disrupt dialysis care, lead to missed dialysis sessions, and increase the number of hospitalizations and use of the emergency department. Additionally, disasters can exacerbate depression and lead to posttraumatic stress disorder among dialysis patients.

Introduction
Natural hazards are elements of the physical environment that are caused by forces extraneous to human intervention and may be harmful to human beings. Natural hazards, such as weather events, can lead to natural disasters (hereinafter referred to as disasters), which are serious societal disruptions. Disasters can lead to dis-ruption of dialysis provision, a life-threatening event for dialysisdependent people. People with end-stage renal disease (ESRD) who are dialysis-dependent constitute a medically vulnerable population with high rates of health care use, morbidity, and mortality (1)(2)(3). Missed dialysis sessions exacerbate these adverse outcomes and correlate with a higher patient-perceived burden of kidney disease, higher mortality and hospitalization rates, increased emergency department (ED) visits, and worse general and mental health (4-7).
Disasters can affect access to dialysis by disrupting transportation, electricity, and water supply (8). Lack of transportation can leave dialysis patients immobile and unable to receive treatment. Similarly, loss of electricity and contamination of water systems can force dialysis centers to close, requiring dialysis patients to seek care elsewhere or miss treatments (9,10). The immediate threats from disasters are compounded by long-term stressors and mental health effects (11).
Just 577 miles south of Florida, the Cayman Islands is home to more than 68,000 people (12) and has more than 2 million visitors annually (13). As of 2018, the Cayman Islands had 4.1 physicians per 1,000 residents and fewer than 250 inpatient hospital beds (14,15). In addition to government health care services, the Cayman Islands have 100 private health care facilities (most of which are outpatient clinics) and 2 private hospitals; both hospitals are located on Grand Cayman, although neither provides dialysis services nor operates an ED (15). Hurricanes can disrupt dialysis provision, and dialysis patients may be flown overseas to receive care (16). However, patient transport is costly, and the dialysis population is growing; therefore, effective emergency preparedness programs are important in the Cayman Islands and other island settings.
The effect of disasters on dialysis patients living in geographically vulnerable areas such as the Cayman Islands is largely understudied. The objective of this systematic review was to describe the scope and effects of disasters on dialysis patients and the unique needs of dialysis patients during and after a disaster, to inform planning and effective emergency preparedness.

Data sources
From January 29, 2019, through February 1, 2019, we searched PubMed, Scopus, CINAHL, and the Cochrane Library to identify peer-reviewed studies published from January 1, 2009, through January 31, 2019, that reported on the effects of disasters on dialysis patients. We selected the search terms in consultation with a research librarian; they were a combination of Medical Subject Headings (Box) and keyword terms (full search string available in Appendix A). This review followed PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidelines (17).

Box. Medical Subject Headings (MeSH) Search Terms Used in a Systematic
Review of Natural Disasters in the Americas, Dialysis Patients, and

Study selection
We reviewed studies that met the following inclusion criteria: they reported on the effects of disasters on dialysis patients; they were published in English from January 1, 2009, through January 31, 2019; and they were conducted in the Americas. We excluded review articles, editorials, and commentaries. However, we examined the reference sections of these articles for potentially relevant articles meeting our inclusion criteria.
For this review, "disasters" refer to all naturally occurring hazardous events of the physical environment such as hurricanes, tornadoes, and earthquakes, that can lead to human, material, economic, and/or environmental losses or impacts (18,19). The effects of disasters on patients can be direct or indirect. Direct effects include harm to the physical, mental, or social well-being of patients, and indirect effects include damage to health care facilities, dialysis centers, dialysis apparatus, water supply, electricity, or transportation.
Two authors (R.S.S. and R.J.Z.) performed independent reviews of the identified titles and abstracts to assess whether they met the inclusion criteria for full-text review. Next, these authors reviewed full-text articles and independently determined which articles to include for data extraction. They reviewed bibliographies to identify additional relevant articles and resolved discrepancies by consensus.

Data extraction
Two reviewers (R.S.S. and R.J.Z.) independently extracted data from each study in the sample. They extracted the following in- formation: author names, publication year, study objectives, study design, participant demographic characteristics, sample size, and relevant findings. We did not pre-identify outcome summary measures for data extraction because we considered multiple outcomes for inclusion. However, when a quantitative study reported an outcome of interest by using a summary measure, such as an odds ratio or hazard ratio, we extracted these data. Additionally, the 2 reviewers independently identified the direct and indirect effects of disasters on dialysis patients and categorized them as indirect effects, direct effects, mental health effects, and others.
Finally, the 2 reviewers independently assessed the quality of each study by using the following tools: the Newcastle-Ottawa Scale, a measurement tool for assessing the quality of observational cohort studies (20), the Critical Appraisal Skills Programme Qualitative Checklist (21), and the Joanna Briggs Checklist for Analytical Cross Sectional studies (22). Neither the Critical Appraisal Skills Programme Qualitative Checklist nor the Joanna Briggs checklist includes a scoring system. Therefore, the reviewers discussed and agreed on the overall value for qualitative studies and overall appraisal for cross-sectional studies. We converted the Newcastle-Ottawa Scale to good, fair, or poor quality categories by using a method described previously (23,24). The reviewers resolved discrepancies by consensus.
The study was not registered before data extraction, and the study design was developed in consultation with a research librarian.

Results
The initial search yielded 434 articles published; we removed 56 duplicates and screened 378 titles and abstracts ( Figure). After eliminating 357 articles that did not meet inclusion criteria, we assessed 21 full-text articles for eligibility. We excluded 2 review articles, and we removed 4 more articles that did not meet the inclusion criteria after closer review. Fifteen articles met the selection criteria for full-text data extraction (Table 1).

Figure.
The study selection process for a systematic review of natural disasters in the Americas, dialysis patients, and implications for emergency planning. The search was conducted from January 29, 2019, through February 1, 2019.

Indirect effects
Seven studies reported on the indirect effects of disasters on dialysis patients (25,28,30,(33)(34)(35)37), including loss of electricity (25,33), lack of clean water (25), blocked roads ( Loss of electricity and clean water can result in the closure of dialysis centers (25,33), which can lead to missed dialysis sessions, treatment delay (later in the day or next day), or referral to other centers (33). Another consequence was the development of bacterial peritonitis in 3 peritoneal dialysis patients who manually forced the fluid exchange (because of lack of electricity) or used river water (because of disrupted water supply) to clean the catheter exit site (25). Blocked roads and the lack of transportation presented challenges to transporting dialysis patients, and these challenges led to missed sessions (25,34).
Disruptions to living situations and the requirement for evacuation from residences can interrupt a dialysis patient's usual source of care, which can place a strain on other centers as they face an increased patient load (28,30,35). Similarly, center closures and evacuation can have a ripple effect. When a center closes (or patients are evacuated), patients are shunted to another facility, where staff are forced to shorten treatments to meet the increased demand on units (28,35). When relocation is not an option, patients can miss 1 or more sessions, which can lead to electrolyte imbalances or ED visits (35). Closures can be complicated by disrupted communication, which can limit a center's ability to communicate with patients or staff members about emergency plans (25).
Missed dialysis sessions among dialysis patients after a disaster were found to be associated with patients being on dialysis fewer than 2 years, living alone before the storm, and being unaware of the emergency plans of their dialysis center (37).
ED use and number of hospitalizations increased among dialysis patients in the week after the storm (26,27,29,32,38). However, effects on mortality were inconclusive. In 1 study, the 30-day mortality rate was higher among patients living in areas affected by a hurricane than either comparison group (32), whereas, in another study, the hurricane was not associated with excess mortality of dialysis patients (36).

Mental health effects
Only 1 study addressed mental health among dialysis patients after a disaster (39). In a sample of patients with ESRD, after Hurricane Katrina, 92 (24%) reported symptoms consistent with a dia-gnosis of PTSD (posttraumatic stress disorder), and 178 (46%) reported symptoms consistent with a diagnosis of depression. Positive screening for depression was associated with higher risks for all-cause and cardiovascular-related hospitalization and mortality in the year after the storm (39).

Other effects
Two studies reported on predisaster activities and their effects on dialysis patients postdisaster (30,31). Predisaster activities included dialysis-specific preparedness and early dialysis (receiving a session ahead of schedule). Dialysis-specific preparedness was associated with a significantly lower incidence of missed sessions (30). Similarly, receiving early dialysis was associated with a significantly smaller number of missed sessions (30) and reduced odds of 30-day mortality (31).

Quality assessment
In our quality assessment (Appendix B), observational cohort studies met 5 to 8 of the possible 9 criteria of the Newcastle-Ottawa Scale. The criterion "outcome not present at start of study" was not met by any study because all studies assessed explored the exacerbation of an existing event (such as increases in ED use and hospitalization). The criterion "adequacy of follow-up of cohorts" was met by only 2 studies. Although no review complied with all 9 criteria assessed with the Newcastle-Ottawa Scale, after we converted the scale to good, fair, or poor quality categories, we determined that all but 1 study was of good quality.
We determined that 2 studies assessed by using the Critical Appraisal Skills Programme Qualitative Checklist were somewhat valuable, and 1 study was determined to be valuable. Additionally, the overall appraisal for all studies assessed by using the Joanna Briggs checklist was that these studies should be included in the review.

Discussion
Findings from the 15 studies examined show that disasters have indirect, direct, mental health, and other effects on dialysis patients. The emergency preparedness recommendations identified in the study ( and 1 ED, patients are easily susceptible to treatment disruptions. Furthermore, key stakeholders (the director of public health, the deputy epidemiologist, and a nephrologist), expressed concerns about the growing dialysis population. Therefore, preparing to address the complex needs of ESRD patients is important.
This review highlighted several implications for emergency planning in the island setting. The indirect effects of disasters -lack of electricity, clean water, and transportation; damage to communication systems; mass evacuation and disturbed living situations -resulted in the closure of dialysis centers, ESRD patient surge in host dialysis centers, missed dialysis sessions, difficulties communicating with providers and patients, and difficulties moving patients (25,28,30,(33)(34)(35)37). These findings suggest that emergency preparedness planners and dialysis centers should have a contingency plan to transport patients to another center if dialysis units are rendered nonfunctional after storms. Lack of transportation, blocked roads, and dialysis patient surge are also significant concerns. Early evacuation can serve as a proactive approach for dialysis patients living in vulnerable areas and for those patients with limited mobility (25).
Similarly, having dialysis providers readily available in alternate locations and other plans to accommodate demand surges in host dialysis centers can help to address surge issues. Because the Cayman Islands has only 2 dialysis centers, managing patient surge is particularly important, because the closure of 1 dialysis center could likely overwhelm the remaining dialysis center. Emergency planners could prepare for this by keeping the functioning dialysis center open for extended hours to care for the increasing patient load and have dialysis providers readily available to address the needs of the dialysis community.
Another challenge is communicating with providers and patients when systems are disrupted. Information that is given to dialysis patients before a hurricane should include contact information for alternative dialysis centers, information on an emergency renal diet, copies of their dialysis orders, and a list of their medications and comorbidities (41). Providing patients with this information ahead of time can allow receiving centers to deliver care more efficiently to nonregular dialysis patients (28).
The direct effects of disasters include increases in ED use, hospitalizations, and mortality (26,27,29,32,36,38). Providing early dialysis in advance of a disaster is a proactive approach to curb these adverse outcomes. Receiving early dialysis was associated with lower odds of ED visits and hospitalizations in the week of the storm and reduced odds of 30-day mortality (31).
PTSD and depression symptoms are prevalent in the dialysis population post-disaster (39). Therefore, emergency planning for dialysis patients should include the identification and treatment of depression, PTSD, and other mental illnesses after disasters.
Our study has several limitations. The outcomes of interest were limited to the study's definition of effects. Although we consulted with a research librarian to fully capture "effects" in our search, we may have missed terms that could have provided more value to our study. Also, only 1 study addressed the effects of disasters on peritoneal dialysis patients; therefore, findings may not be generalizable to this population. Most studies in this review addressed hurricanes; so, research exploring the effects of other types of disasters on dialysis patients is needed, particularly no-notice events such as earthquakes. Such disasters would preclude evacuation or opportunities for early dialysis. Publication and language bias are also possible limitations because we did not search the gray literature, and we included only articles in English. Finally, all but 1 study (25) reported findings in the continental United States. Dialysis patients living on islands may encounter additional challenges not present in nonisland settings.
Our study also has several strengths. We consulted with a research librarian; 2 reviewers independently searched the databases and screened the articles; and we searched 4 databases. These strengths helped to reduce selection bias and improve the scope of the studies included. Additionally, emergency preparedness recommendations are generalizable to other island settings with similar disasters.
Elucidating the effects of disasters on people whose lives depend on dialysis is of critical importance because the risk for adverse health outcomes increases when dialysis care is disrupted. The effects of disasters on dialysis patients have several implications for emergency planning. However, the topic is inadequately studied, especially in the island setting. The geographic isolation of islands can hamper the timely provision of resources to the dialysis community and presents a unique context to study the effects of disasters on dialysis patients.
Many islands have a tenuous health care system and lack economic safety nets, which can exacerbate the adverse outcomes of disrupted dialysis care. Efforts to mitigate the effect of disasters on dialysis patients will require coordination among public health professionals and other key personnel, carefully designed emergency preparedness plans, and education and training of all involved. From the day of the hurricane (day 0) through day 5, categories of primary ICD-9 diagnosis codes with significant (P < .001) increases among ED patients were chronic kidney disease, dialysis dependence, electrolyte abnormality, and renal failure.

PREVENTING CHRONIC DISEASE
• The significant increase in dialysis dependence lasted the longest of the 4 increases: it was significant (P < .001) from day 0 through day 5.
• The frequency of ED use significantly (P < .001) increased among patients with a secondary ICD-9 diagnosis category of dialysis dependence and chronic kidney disease. • The percentage of participants who had ED visits was greater in the study group (4.1%) than in comparison group 1 (2.6%) and comparison group 2 (1.7%), both P < .001.

•
The percentage of participants who were hospitalized during the week of the storm was greater in the study group than in comparison groups: 4.5% in study group, 3.2% in comparison group 1 (P < .001), and 3.8% in comparison group 2 (P = .003).
• 23% of study group participants who visited the ED received dialysis, compared with 9.3% in comparison group 1 (P < .001) and 6.3% in comparison group 2 (P < .001).
• Primary discharge diagnoses for patients visiting the ED or being hospitalized were for dialysis or ESRD. • Factors significantly associated with increased surge capacity were the average number of patients per day during nondisaster operations (P = .04), having affiliated outpatient dialysis centers (P = .03), use of extra dialysis machines (P = .01), and having extra staffing (P = .007).
• Storm-related challenges prevented the efficient operation of dialysis units; 7 of 14 operating hospital dialysis facilities reported a staff shortage due to transportation issues in getting to the facilities.
• All 5 affiliated outpatient dialysis centers cited communication challenges with ambulette service providers, which resulted in delays in transferring patients from EDs to outpatient dialysis centers.  Electricity and clean water are critical for dialysis; emergency planners could compensate for the loss of electricity by using generators and lack of clean water by making preparations to have extra storage of potable water; additionally, emergency planners and dialysis providers can make arrangements to transport patients to affiliate sites.
Lack of clean water Leads to closure of dialysis facilities and missed dialysis sessions. Use of unclean water by peritoneal dialysis patients can lead to bacterial peritonitis Blocked roads and lack of transportation Creates challenges in transporting dialysis patients and leads to missed dialysis sessions. Problems in the commute of staff members and providers to dialysis facilities can lead to a shortage of dialysis providers.
Emergency planners and dialysis centers should have a contingency plan to transport patients to another center; proactively evacuate dialysis patients living in vulnerable areas or those with limited mobility; make preparations for dialysis staff members and providers to shelter in place at dialysis units.

Disrupted communication system
Presents challenges in communicating with patients or staff members about emergency plans.
Develop an action plan of how to communicate with staff members ahead of disasters; provide dialysis patients with pertinent information before a hurricane, such as contact information for alternative dialysis centers, information on an emergency renal diet, copies of their dialysis orders, and a list of their medications and comorbidities.

Mass evacuation and disturbed living situation
Interrupts usual source of care for dialysis patients, leading to a strain on other centers as they face an influx of dialysis patients. Onset or exacerbation of posttraumatic stress disorder In addition to preparing to manage the medical and social needs of dialysis patients after disasters, clinicians should prepare to screen dialysis patients for signs of depression, posttraumatic stress disorder, and other mental health conditions, and develop an action plan to address and treat the mental health needs of dialysis patients, such as referral to counseling and support groups.

Depression
Onset or exacerbation of depression Others Dialysis-specific preparedness Lower the incidence of missed dialysis sessions Periodically review dialysis-specific preparedness and awareness with dialysis patients, especially during the hurricane season; providers can assess the readiness of dialysis patients by using the disaster preparedness checklist provided by the National Kidney Foundation.
Early dialysis Lower odds of missed dialysis sessions Emergency planners should consider offering preemptive dialysis to curb adverse outcomes associated with missed dialysis sessions, such as emergency department visits and hospitalizations.