The willingness of personnel to provide health care services in a disaster has serious implications for access to health care and the quality of the care provided in a disaster. This issue has been identified as an “untilled or undertilled” research area.Reference Abramson, Morse and Garrett1 Thus, a review of the current literature regarding willingness to work in disaster can provide critical information for policymakers, planners, and health care personnel as well as identify gaps for researchers to explore.
Auf der HeideReference Auf der Heide2 stated “disaster planning is only as good as the assumptions on which it is based.” The reality of disaster response is that health care personnel may face competing duties to employers, patients, family, loved ones, and self.Reference Chaffee3 It is important to understand how personnel may make decisions when facing competing duties so that plans, policies, and organizational decisions can be based on the best evidence available. Willingness to work in disaster is one area in which hospitals and communities may be making assumptions not based on evidence, and this could have serious consequences. For example, Gershon et alReference Gershon, Qureshi and Stone4 reported that both the New York city and state avian influenza plans depend on home health care workers to provide home care to patients in an outbreak. However, in a recent study they found that only 11% of home health workers and 37% of registered nurses would be willing to provide care for patients during such an outbreak.Reference Gershon, Qureshi and Stone4 The plans do not reflect the evidence—and would likely falter due to lack of personnel to provide care as expected.
Although many definitions of disaster have been proposed, there is general agreement that disaster is an inherently social phenomenon involving disruption of an organized human social system.Reference Perry5 Disasters are likely to cause a sudden and possibly massive demand for services in health care facilities.Reference Alexander6 Surge response (often referred to as surge capacity) is defined as the ability of a health care facility or system to rapidly expand its operations to safely treat an abnormally large influx of patients in response to an incident.Reference Bonnett, Peery and Cantrill7 Surge response depends on the availability of 3 elements: personnel, equipment and supplies, and structure (facilities and organization).Reference Kaji, Koenig and Bey8 All 3 are important, however, response to a surge in a hospital or other health care organization cannot be successful without adequate personnel to provide patient care and other personnel to support care delivery. Thus it becomes imperative to better understand how hospital personnel may actually behave when faced with a disaster.
Studies over time have shown that people tend to demonstrate responsible, positive, and adaptive behavior in disaster including compliance, orderliness, and altruism.Reference DiGiovanni9, Reference Tierney10 The concept of role abandonment, including how individuals with disaster-related responsibilities would perform in light of family obligations, raised interest and was examined in the 1950s and 1960s and in a later study by Dynes.Reference Dynes11 It was found that some individuals do experience role conflict but almost never abandon their occupational roles due to family conflicts.Reference Webb, Beverly and McMichael12 Quarantelli13 studied more than 6000 emergency workers in natural disasters that occurred between 1964 and 1974 and found no evidence of role abandonment. However, recently the vulnerability of societies to disaster has increased,Reference Bankoff, Frerks and Hilhorst14 the nature of disaster has evolved to include the potential for industrial and technological catastrophe,Reference Perrow15 the risk of terrorism has emerged,Reference Alexander16, Reference Garwin17 and the direct risks to health care personnel in biological outbreaks such as severe acute respiratory syndrome became evident.Reference Koh, Lim and Chia18 It is unclear whether previously supported findings remain accurate in a contemporary world environment that includes new threats and risks.Reference Chaffee19 Early studies of the willingness of health care personnel to work in certain disaster situations indicate that role abandonment may now be a valid concern.
The purpose of this integrative review is to describe and analyze the state of evidence concerning the willingness of health care personnel to work during disaster. The research question for this review is, “What is known about the willingness of health care personnel to work in disaster?” The main objective is to present the best available evidence about the willingness of health care personnel to work in a disaster.
METHODS
Identification of Existing Reviews
A search for existing literature reviews on this topic revealed one that reviewed 8 studies.Reference Smith20 Although the review was published in 2007, this review proceeded because additional studies were known to exist.
Inclusion Criteria
1 Research design: Original quantitative or qualitative designs.
2 Primary study participants: Clinical, administrative, and support personnel involved in the delivery of health care services; employment site (eg, hospital, clinic, rescue squad) was not a factor.
3 Outcome measure: Willingness to work in a disaster or public health emergency; variations of wording were accepted including “willingness to report to work,” “willingness to care for patients,” “willingness to work,” “willingness to respond,” “willingness to report,” “willingness to report to duty,” and “willingness to provide care.”
4 Research methodologies: Reports of published quantitative and qualitative studies were included in this review. Unpublished reports such as conference proceedings and dissertations were included that met specific criteria.Reference Benzies, Premij and Hayden21
5 Language of studies: English only.
6 Time period: 1950–2007.
7 Dates of review: September 1 to December 1, 2007.
8 Type of research reports: Peer-reviewed research articles, peer-reviewed research abstracts, peer-reviewed summaries of research findings, and nonpublished reports that were subject to a review and approval process.
Search Strategy
Six methods were used to identify potential candidates for inclusion in this review:
1 Examination of a systematic review database: The Cochrane Library was searched although reports meeting the criteria for this review were not expected; none were found. The Cochrane Library evaluates reports of clinical outcomes.
2 Electronic database searching: Electronic databases were searched: PubMed, LILACS, CRISP (Computer Retrieval Information on Scientific Projects), PsycINFO, EBSCO Psychological and Behavioral Sciences Collection, and SSCI (Social Sciences Citation Index). Search terms used were “willingness,” “work,” “disaster,” “willingness disaster,” “willingness work disaster,” “willingness to report,” “willingness research,” “disaster research,” “willingness psychometrics,” and “willing*,” with the asterisk indicating any search keyword beginning with “willing.”
3 Ancestry searching: The references of each publication selected for inclusion were examined for additional relevant publications.
4 Internet searching: A search of publicly available Internet content was conducted using Google and Google Scholar. The same search terms were used as those used with electronic database searches.
5 Hand-searching: Journals with a specific disaster focus were analyzed issue by issue when available. These included Australasian Journal of Disaster and Trauma Studies, Disasters, Disaster Medicine and Public Health Preparedness, The Internet Journal of Rescue and Disaster Medicine, and Disaster Management and Response.
6 Networking: Individuals with a known interest in the topic were contacted.
Screening
Research reports were screened using a 3-stage process described by Gifford et al.Reference Gifford, Davies and Edwards22 Articles titles were reviewed for potential relevance and either retained for additional review or discarded. Next, abstracts appearing relevant were reviewed using the inclusion criteria and then retained for further evaluation or discarded. In the third step, full articles were reviewed with the inclusion criteria being applied. The studies included in this review were evaluated using a checklist of specific aspects of quality, reliability, and validity developed by the researcher.
Integration of Quantitative and Qualitative Studies
Current methods of evidence synthesis tend to favor quantitative data only and systematic reviews generally do not include qualitative data.Reference Dixon-Woods, Agarwal and Jones23 However, in complex public health issues, quantitative findings may be insufficient to provide clinicians and others with adequate evidence on which to base decisions.Reference Jack24 To gain the broadest perspective on willingness to work, quantitative and qualitative studies were included in this review. Narrative summary as described by Dixon-Woods et alReference Dixon-Woods, Agarwal and Jones23 was used to select, chronicle, and order evidence from the qualitative studies to present it alongside the quantitative findings (Table 1).
TABLE 1 Key Findings in Studies of Willingness of Health Care Personnel to Work in a Disaster (1991–2007)
TABLE 1 Key Findings in Studies of Willingness of Health Care Personnel to Work in a Disaster (1991–2007) (Continued)
Study Quality
The studies were evaluated to determine whether an adequate rationale was made to justify the study and whether an explicit purpose was stated.Reference Girden50 Each study was found to make a clear and logical case for the investigation. Purpose statements were consistently clear. Research questions or hypotheses were articulated in only 3 of the research reports evaluated. Locke et alReference Locke, Silverman and Spiraduso51 identified reasons to suspend trust in research, including technical problems brought to the reader's attention, conflicts of interest, carelessness, sampling inadequacy, lack of replication, poor scholarship, and lack of review by a refereed journal. The studies under review did not appear to have any substantive reasons to suspend trust in their findings, although a number did not report all of the study characteristics that are desirable.
Justification for Inclusion of Gray Literature
The peer-reviewed scientific literature is viewed as the most widely accepted source of researchReference Conn, Valentine and Cooper52 and these reports were most strongly desired for inclusion in this review. There are circumstances under which including research reports from sources other than the peer-reviewed scientific literature may be advantageous in a systematic review. Benzies et alReference Benzies, Premij and Hayden21 defined criteria for including “gray” (non–peer-reviewed) sources to determine when to include gray literature. Benzies et alReference Benzies, Premij and Hayden21 recommend using gray literature if the outcome is complex, if there is lack of consensus concerning the measurement of the outcome, and if there is a low volume of evidence. These criteria were met by this systematic review; thus, gray literature was included. Advantages of using gray literature include that its use may increase the likelihood of a comprehensive search and may offset bias against the null hypothesis.Reference Conn, Valentine and Cooper52 Evaluating the quality of gray literature may be difficult if it is not easily accessible or does not fully describe all aspects of the study. This review includes research abstracts, a doctoral dissertation, and a master's thesis. The research abstracts, although limited in scope, were published in peer-reviewed journals. The thesis and dissertation were evaluated through an academic review and approval process. Both are available online. Table 2 presents a summary of each report in this review; the type of report (peer-reviewed or not) is noted in the table.
TABLE 2 Research Report Type, Data Type, Study Purpose, and Sample Characteristics of 27 Reviewed Studies
TABLE 2 Research Report Type, Data Type, Study Purpose, and Sample Characteristics of 27 Reviewed Studies (Continued)
TABLE 2 Research Report Type, Data Type, Study Purpose, and Sample Characteristics of 27 Reviewed Studies (Continued)
RESULTS
A total of 27 studies of health care personnel's willingness to work in disaster were identified that met all inclusion criteria for review. Twenty-five were quantitative studies (N = 20,325; range 50–6428) and 2 used qualitative approaches (N = 90; range 30–60). The first report found was a study published in Israel in 1991.Reference Shapira, Marganitt and Roziner25 No further studies on the topic appeared until 2002, when the first qualitative report appeared. The methodology used most frequently in the quantitative studies was paper-based, self-administered survey (N = 21). Interviews and focus groups were used in the 2 qualitative studies; methodology was not identified in 2 research reports. Twenty-three of the studies that met search criteria were conducted in the United States, and the remainder in Israel, Canada, and Australia. Physicians, nurses, and paramedics were the personnel categories studied most frequently (Figure 1).
FIGURE 1 Health care populations included in reports of willingness to work in a disaster
Reliability of Quantitative Studies
Research reports in this review were evaluated for test-retest, alternate form (parallel form), internal consistency, split-half, and interrater reliability. Reference Nunnally and Bernstein53–Reference Waltz, Strickland and Lenz55 Two research reports included evidence of reliability: CraneReference Crane38 and Young and PersellReference Young and Persell35 report internal consistency reliability. Young and PersellReference Young and Persell35 indicated that interrater reliability was evaluated, although findings were not reported (Table 3).
TABLE 3 Variables, Evidence of Reliability, and Evidence of Validity in 27 Quantitative Studies of Willingness to Work in Disaster
TABLE 3 Variables, Evidence of Reliability, and Evidence of Validity in 27 Quantitative Studies of Willingness to Work in Disaster (Continued)
Validity of Quantitative Studies
Evidence of validity provides the reader with greater confidence that the instrument used in the study did indeed measure what it was expected to measure.Reference Polit and Beck56 (Evidence of validity appears in Table 3.) The types of validity sought were face validity, content validity, construct validity, and criterion validity.
Face Validity
Not considered evidence of true validity, its presence may encourage participants to respond and may therefore increase response rate.Reference Waltz, Strickland and Lenz55 CraneReference Crane38 and Young and PersellReference Young and Persell35 report that they evaluated face validity.
Content Validity
Content validity is assessment by content experts, content validity index analysis, content validity ratio, and so forth. Alexander and WyniaReference Alexander and Wynia29 report that they assessed content and construct validity a priori but do not describe how. Three investigations used content experts to evaluate the relevance of specific instrument items.
Construct Validity
Construct validity is hypothesis testing anchored in a conceptual framework, factor analysis, multitrait–multimethod approach, contrasted or known groups approach, and so forth. Balicer et alReference Balicer, Omer and Barnett40 reported their findings fit well with their theoretical framework (indicating evidence of construct validity). There were no reports of the use of convergent or discriminant validity, multitrait–multimethod analysis, or factor analysis.
Criterion Validity
No measure of concurrent or predictive validity—for example, criterion-related validity—was noted in these studies.
Evaluation of Instrument Development
New instruments were developed for use in 20 of the studies included in this review, thus evidence of psychometric evaluation was sought. This included evidence of reliability and validity, reading level, cognitive interviews or focus groups to evaluate comprehension, and expert panel evaluation of survey items. Overall, little description of the instrument development process appeared in the reviewed reports.
Additional Factors Evaluated
Pretesting
Eight investigators reported doing a pretest or pilot test. Five of these describe the test or that the results led to instrument revision. Three do not report the outcome of the pretest or whether it was used to revise or refine the instrument or administration methods. Mackler et alReference Mackler, Wilkerson and Cinti47 and Qureshi et alReference Qureshi, Merrill and Gershon28 indicate that their studies were pilot tests.
Use of a Theoretical or Conceptual Framework
Two research reports described the use of a conceptual or theoretical framework (Balicer et alReference Balicer, Omer and Barnett40 and CraneReference Crane38). Not using theory to guide research can leave a gap in the scientific process. Theories predict the presence of new phenomena and generate hypotheses that can be translated into questions that can be answered through scientific study.Reference Wilson57
Sampling Strategy
Sampling procedures were evaluated to discern whether the sample groups were representative and were not distorted by the selection process.Reference Locke, Silverman and Spiraduso51 Five of the 27 studies reviewed used random sampling, a sampling strategy that improves external validity; 14 used convenience sampling and 7 did not report sampling procedures.
Statistical Power
One group of investigators reported a power analysis (Katz et alReference Katz, Nekorchuk and Holck41).
Selection Associated With Nonresponse
Alexander and WyniaReference Alexander and Wynia29 evaluated nonresponse bias and found nonresponders differed from responders on a few variables. Balicer et alReference Balicer, Omer and Barnett40 and SchecterReference Schechter46 evaluated nonresponse and found no differences. DiGiovanni et alReference DiGiovanni, Reynolds and Harwell30 acknowledge that sample bias may have occurred in their study as did DiMaggio et al.Reference DiMaggio, Markenson and Loo37
Social Desirability Bias
Alexander and WyniaReference Alexander and Wynia29 discussed the potential for social desirability bias in their study.
Missing Data Management
There were no discussions of analysis or management of missing data in the research reports in this review.
Data Synthesis
Although studies of willingness to work have appeared largely since 2002, a body of research findings has emerged that is beginning to describe the phenomenon of willingness to work in disaster.
Influence of Type of Disaster
Willingness to work was found to be influenced by the type of disaster.Reference Smith20,Reference Lanzilotti, Galanis and Leoni27,Reference Martens, Hantsch and Stake32,Reference Qureshi, Gershon and Sherman39,Reference Cone and Cummings44 In general, respondents indicated they were more willing to work in weather-related disasters and mass casualty events than in radiological, nuclear, biological, or chemical disasters. Biological outbreaks appear to be a significant barrier to willingness to work. In the Balicer et alReference Balicer, Omer and Barnett40 study of public health department employees, 53.8% indicated willingness to work in a pandemic influenza outbreak. This is consistent with other reports. Qureshi et alReference Qureshi, Gershon and Sherman39 found 48% of health care workers in New York City indicated they would be willing to work in a severe acute respiratory syndrome outbreak. Cone and CummingsReference Cone and Cummings44 found that 58% of hospital employees would be willing to work in a biological event. DiMaggio et alReference DiMaggio, Markenson and Loo37 found 64.8% of emergency medical services workers would be willing to report to work in a smallpox outbreak (whereas 87.7% would be willing to respond to an explosion). Fifty percent of hospital personnel in the Irvin et alReference Irvin, Cindrich and Patterson49 study indicated willingness to work in an avian influenza outbreak. The lowest willingness data to date were the Gershon et alReference Gershon, Qureshi and Stone4 study of New York home health workers in a hypothetical avian flu outbreak, in which 11% of home health aides and 37% of registered nurses indicated that they were willing to care for an infected, quarantined patient.
Concern for Family and Loved Ones
In both the quantitative and qualitative studies, concern for the well-being of family and loved ones emerged as a potentially powerful barrier to willingness to work in disaster.Reference Smith20,Reference French, Sole and Byers26,Reference Young and Persell35,Reference DiMaggio, Markenson and Loo37,Reference Qureshi, Gershon and Sherman39,Reference Schechter46,Reference Kruus, Karras and Seals48 The importance of family in influencing willingness was identified in both the first quantitative studyReference Shapira, Marganitt and Roziner25 and the first qualitative studyReference French, Sole and Byers26 and has continued to be a factor in subsequent studies. Concern for family was the most frequently cited reason for unwillingness to respond in the Qureshi et al study of New York City hospital personnelReference Qureshi, Gershon and Sherman39 and the DiMaggio et al study of emergency medical technicians.Reference DiMaggio, Markenson and Loo37 In Young's study of nursing students, 90% of respondents indicated that they would not work with contagious patients if their families did not receive prophylaxis.Reference Young and Persell35 Kruus et al found that family support and concern about family support influenced willingness to work.Reference Kruus, Karras and Seals48
Concern for Pets
The 2002 qualitative study by French et al appears to be the first evidence that concern for pets may influence willingness to work.Reference French, Sole and Byers26 Concern for pets emerged again in the Qureshi et al study of New York City hospital personnel.Reference Qureshi, Gershon and Sherman39 Thirty-three percent of Cone and Cummings'sReference Cone and Cummings44 respondents desired pet care if they were to be on duty for a prolonged period. The impact of the need for safe pet care became evident during the Hurricane Katrina response in 2005. James Montgomery, chief executive officer of Tulane University Hospital and Clinic in New Orleans, had to set up a veterinary care center in the hospital's parking deck to care for his employees' 79 dogs, cats, and birds. Montgomery said, “Don't underestimate unwillingness to leave them behind.”58
Relation of Education/Training to Willingness
The influence of education on willingness to work has been examined, although not in a consistent manner. Lanzilotti et alReference Lanzilotti, Galanis and Leoni27 found that in a sample of physicians and nurses, those with greater knowledge of biological agents reported greater willingness to work in a field medical facility during a biological event. DiMaggio et alReference DiMaggio, Markenson and Loo37 found that emergency medical technicians were more willing to respond to chemical, radiological, or smallpox events if they had recently undergone hands-on training. GullionReference Gullion33 found a correlation between school nurses' education concerning their response role and willingness to respond when at risk. Qureshi et alReference Qureshi, Merrill and Gershon28 found in a pilot study of nurses that intention to work in an emergency increased by 12% after a training program. Several studies report descriptive data that summarizes disaster education, but the studies do not report associations between education and willingness to work.
Influence of Personal Obligations
Personal responsibilities to others were identified as barriers to willingness to work. Qureshi et alReference Qureshi, Gershon and Sherman39 found that child and eldercare obligations were related to decreased willingness to work. Child care was identified as a service employees desired so that respondents could work.Reference Cone and Cummings44
One's Value in the Response
The perception of one's effectiveness or importance in disaster response was a significant factor identified by Balicer et alReference Balicer, Omer and Barnett40 and Kruus et al.Reference Kruus, Karras and Seals48 In the Balicer et al study of Maryland public health department personnel, respondents indicated that the single most important factor that influenced their willingness to work was the perception of the importance of their role in the agency's response.Reference Balicer, Omer and Barnett40 Kruss et al reported that “importance of working” was a significant predictor of willingness to work in a riot situation and a power outage (although the variable's conceptual definition is not offered).Reference Kruus, Karras and Seals48
Belief in Duty to Care
Alexander and WyniaReference Alexander and Wynia29 found that willingness to treat in their sample of physicians was associated with belief in a duty to treat. Of the emergency medical service workers who indicated to DiMaggio et alReference DiMaggio, Markenson and Loo37 that they were willing to respond to terrorism, 83.3% reported their willingness was due to a sense of responsibility.
Availability of Personal Protective Equipment
The availability of personal protective equipment (PPE) emerged in several studies as a factor that would influence willingness to work. Reference Mackler, Wilkerson and Cinti42,46,47 A study of paramedics by Mackler et alReference Mackler, Wilkerson and Cinti47 highlighted the influence of PPE. More than 80% of paramedics surveyed indicated they would not remain on duty in a smallpox outbreak if PPE and vaccine were not available. If fully protected, then 91% indicated they would remain on duty (although that number dropped to 38% if the family was not protected). All 60 participants in the Shaw et alReference Shaw, Chilcott and Hansen42 study of Australian physicians indicated that they would be willing to work in a pandemic influenza outbreak, but nearly 92% stated they would cease to work if PPE were not available. Shapira et alReference Shapira, Marganitt and Roziner25 found that willingness to work in a chemical attack would increase from 42% to 86% if safety measures were provided to hospital personnel.
Support for Basic Needs
Nurses in a study by French et alReference French, Sole and Byers26 raised concerns about having basic needs met during hurricane response (eg, water, food, rest, shelter). Cone and CummingsReference Cone and Cummings44 found respondents desired telephone service and e-mail access if they were to be on duty for a prolonged period.
Length of Response
Steffen et alReference Steffen, Masterson and Christos34 examined the variable of willingness to work additional hours and the potential factors associated with it. They found emergency department personnel were more willing to work extra hours to respond to an air crash than to a biological or radiological event. They also found respondents with children were willing to work the same number of extra hours as those without children. Cone and CummingsReference Cone and Cummings44 attempted to quantify how long hospital employees would be willing to remain on duty assuming basic needs were met. The mean was 3.6 days.
DISCUSSION
Each study in this review has revealed certain facets of the complexities of human behavior in disaster. The seminal study by Shapira et alReference Shapira, Marganitt and Roziner25 of willingness to work in disaster was an important one that, surprisingly, did not seem to initially generate interest. This study, however, was the spark that initiated a growing body of evidence, although it was more than a decade before the next study appeared in the literature. Shapira et al found that in a sample of Israeli hospital personnel, only 42% would be willing to report to work in a missile attack with a chemical weapon. Equally important was their finding that willingness to work could be manipulated. Of the respondents who indicated unwillingness to work, 33% reported they would change their mind if provided adequate protective gear. The willingness rates increased when gas mask lenses, transportation, or an “all clear” announcement was made. These findings are the first that demonstrate how hospital (or government) leadership may be able to influence willingness to work through both actions taken in advance to prepare for disaster response and during a disaster. The studies reported following Shapiro et al have uncovered additional evidence that indicates willingness to work can be enhanced. An example of evidence being used to improve willingness to work appeared at a Florida hospital that frequently dealt with hurricanes and personnel problems. Cape Canaveral Hospital, located on Florida's east coast, evacuated 5 times for hurricanes during a 9-year period. In 1999, during Hurricane Floyd, about 100 employees failed to work—and 30 were terminated.Reference McCoy and Stackpoole59 The hospital examined the problem, assessed employee needs, revised policies and training, and clearly communicated employer and employee expectations in a hurricane, and significant improvements were found in personnel response.
Although the sociological literature has not demonstrated large-scale abandonment of duties in disaster, the nature of disasters is evolving to include situations that may be personally threatening to health care personnel and may threaten their loved ones. This review identifies the type of disaster as an important influence on willingness to work. Respondents indicated a willingness to work that varied from a low of 11% in home health aides in the Gershon et al study of New York City home health personnel in a hypothetical pandemic flu outbreak to a high of 95% in DiGiovanni's study of a Rift Valley fever outbreak. DiGiovanni's high reported willingness to work was contingent, however. Respondents indicated their willingness was based on receiving adequate information and protection at their workplace and no further acts of bioterrorism.Reference DiGiovanni, Reynolds and Harwell30 This review indicates biological events may seriously reduce willingness to work, although the provision of protective gear and education may improve willingness to work in these incidents. This is an area that would benefit from additional investigation.
The influence of education, knowledge, and competency on willingness to work in disaster is unclear, and is an important gap in current knowledge. There have not been systematic studies of the influence of disaster preparedness training nor has the level of education been evaluated in relation to willingness to work. DiMaggio et alReference DiMaggio, Markenson and Loo37 did find increased willingness to work after training. Emergency medical technicians who had received terrorism-related training in the 2 years before the study were twice as likely to be willing to respond to a smallpox outbreak or chemical incident.
A number of findings that have appeared are intriguing, but have only been reported once or twice and deserve more attention. Sex may influence willingness. Shapira et al found female respondents less willing to work than males.Reference Shapira, Marganitt and Roziner25 It is not yet clear whether other demographics (eg, age, race, occupation, distance to work, years of experience, type of employment) may predict willingness to work. There may be a link between immunization status and willingness to work in certain situations. Some areas have not yet been examined. There have been no reported studies of military personnel or of federal employees; all studies to date have sampled civilian employees (and 1 study of county public health department employees). There has not yet been an investigation of the role of organizational culture on willingness to work. There is a need for the exploration of the ethical and legal implications of not working in disaster and how this may influence decision making.
The body of evidence in this review should be considered in relation to studies that examine how health care personnel have behaved in actual disasters, the literature concerning duty to care in disaster, and the literature concerning vicarious traumatization (or secondary traumatization) of health care personnel who work in disaster. Disaster responders experience a broad range of physical and mental health consequences as a result of disaster exposure.Reference Benedek, Fullerton and Ursano60 It is unclear how the knowledge that a person may sustain these sequelae from working during disaster may influence willingness to work, and does not appear to have been investigated. The influence of compensation has been examined only in 1 quantitative studyReference Irvin, Cindrich and Patterson49 and was described in 1 qualitative study; it deserves further investigation.
Based on this integrative review, future methods used in studies of willingness to work in disaster could be strengthened. Consumers of research findings on willingness to work in disaster would benefit when they include a description of sampling strategy, evidence of reliability and validity, descriptions of the psychometric evaluation of new instruments, conceptual frameworks, and descriptions of pretesting procedures and results. Concept analyses of “willingness” and “willingness to work” will assist researchers in understanding these concepts and in defining variables. A description of the data collection instrument is valuable for investigators considering replication of a study. Only a few of the studies in this group included publication of their instrument, part of the instrument, or a detailed explanation of the scoring scheme.
Willingness to work in disaster is a phenomenon that has no political or disciplinary boundaries. It has been recognized by multiple disciplines (eg, medicine, nursing, public health, emergency management) as a problem worthy of scientific scrutiny and future studies may be strengthened through studies undertaken by multidisciplinary teams. Although most studies of the phenomenon have taken place in the United States, others have taken place in Canada, Australia, and Israel—indicating early global interest in how willingness to work may affect disaster response. Multination studies may be a future strategy that uncovers important evidence.
The primary purpose of studying the willingness of health care personnel to work in disaster was not always well described in the studies in this review. The word “patient” rarely appears, but it is indeed the needs of vulnerable individuals seeking care during a disaster that drives these investigations. Hospital surge capacity depends on the availability of 3 elements: personnel, equipment and supplies, and structure (facilities and organization).Reference Kaji, Koenig and Bey8 This review has found that there will likely be disaster situations in which hospital personnel are not willing to work, and that will directly degrade surge capacity. This review also reveals, however, that there appear to be actions that organizations can take to improve willingness to work. It is these actions that have the potential to improve surge capacity and ultimately the quality and availability of care in times of disaster.
Limitations
Despite extensive searching, it is possible a study may have been missed. The topic is multidisciplinary and studies could appear in a wide variety of journals. Studies in languages other than English may exist that are not indexed in major English databases. Research reports that have not been fully peer reviewed were included in this integrative review because of the low volume of studies on this topic to date. Caution must be used in the use of research findings in the gray (unpublished) literature.
CONCLUSIONS
This integrative review of the literature uncovered a number of valuable findings. The review revealed that the phenomenon has been the subject of scientific investigation only since 1991, with nearly all studies being reported since 2002. In this 17-year period, 27 studies were reported, although only 18 of these appeared in the peer-reviewed scientific literature. The science concerning the phenomenon of willingness to work in disaster can be said to be “immature” or “emerging.” Nonetheless, important data already have been produced that may have value for policymakers, emergency planners, and others involved in disaster response. The early findings have uncovered numerous areas ripe for further exploration—especially the influence of family, education, personal obligations, concerns for personal safety, and the type of disaster.
In the early stages of examining a phenomenon, qualitative exploration can be a particularly useful method. Only 2 qualitative studies have been completed that explored willingness to work; additional qualitative exploration as well as mixed methods designs should be considered. It is critical that evidence of the reliability and validity of the instruments being used to measure willingness to work be obtained and reported. Only then can we have confidence in the quality of the data being produced.
The frequency of disaster is increasing and vulnerable populations will continue to experience the health effects of disaster. The demand for health care services will continue to be significant in many disaster situations, and due to economic trends surge capacity is expected to be an ongoing challenge. Future explorations of willingness to work in disaster have the potential to improve the effectiveness of hospital disaster response. The critical factor will be merging the evidence with policymaking and planning. Future research that builds logically upon the body of work that has emerged will enhance understanding of the phenomena and will support evidence-based policymaking, leadership, and planning—and ultimately the care available for those who are harmed in disasters.
