New roles for hospital librarians: a benchmarking survey of disaster management activities

Hospitals are often at the center of activity when a disaster occurs. The well-known examples of the 9/11 terrorist attacks, Hurricane Katrina, and the 2011 Joplin, Missouri, tornado demonstrate the devastating impact that disasters have on local hospitals. In addition to maintaining vital services to patients during the response period, hospitals also support emergency personnel and manage a potential surge of disaster victims. As a result of high-profile catastrophes raising awareness of the responsibilities of hospitals, organizations from the Joint Commission to the Department of Health & Human Services have developed standards, initiatives, and resources to help hospitals with mitigation, preparedness, response, and recovery. These four phases can also be described as “disaster management activities” 1. 
 
Given that hospital librarians provide direct support to patient care and work closely with everyone from the emergency department to administration, it follows that they are likely to be involved in disaster management activities in their organizations. Understanding hospital librarian involvement and articulating their roles will help develop best practices for all medical librarians. To this end, the investigators decided to undertake a benchmarking survey to establish a baseline of current disaster management activities by hospital librarians and to develop an instrument that could be used to collect additional data in the future. The survey was designed to provide quantitative evidence for both existing and emerging roles beyond the anecdotal reports and qualitative investigations that currently exist in the literature on the topic of library roles in disasters.


INTRODUCTION
Hospitals are often at the center of activity when a disaster occurs. The well-known examples of the 9/11 terrorist attacks, Hurricane Katrina, and the 2011 Joplin, Missouri, tornado demonstrate the devastating impact that disasters have on local hospitals. In addition to maintaining vital services to patients during the response period, hospitals also support emergency personnel and manage a potential surge of disaster victims. As a result of high-profile catastrophes raising awareness of the responsibilities of hospitals, organizations from the Joint Commission to the Department of Health & Human Services have developed standards, initiatives, and resources to help hospitals with mitigation, preparedness, response, and recovery. These four phases can also be described as ''disaster management activities'' [1].
Given that hospital librarians provide direct support to patient care and work closely with everyone from the emergency department to administration, it follows that they are likely to be involved in disaster management activities in their organizations. Understanding hospital librarian involvement and articulating their roles will help develop best practices for all medical librarians. To this end, the investigators decided to undertake a benchmarking survey to establish a baseline of current disaster management activities by hospital librarians and to develop an instrument that could be used to collect additional data in the future. The survey was designed to provide quantitative evidence for both existing and emerging roles beyond the anecdotal reports and qualitative investigations that currently exist in the literature on the topic of library roles in disasters.

LITERATURE REVIEW
The investigators' literature search demonstrated the evolution of disaster management activities for hospital librarians. In 1976, Turner coauthored a paper with Kirz outlining a sample ''reference library'' collection for the emergency department in the Journal of the American College of Emergency Physicians [2]. As emergency department personnel are often those most closely associated with disaster management in a hospital, establishing services for them is an important first step, and several other hospital librarians have since continued the trend of involvement with emergency departments [3][4][5]. Additionally, a number of articles illustrated the importance of emergency plans for hospital libraries and undertaking of disaster management activities, specifically preparedness [6][7][8][9]. Finally, there is anecdotal and qualitative literature demonstrating the ways that hospital librarians can be and have been involved with disaster management at the organizational level: Featherstone et al. captured a number of oral histories from hospital librarians detailing experiences with different community disasters, while McKnight studied accounts of the reference services that hospital librarians provided in the aftermath of Hurricane Katrina [10,11]. In 2010, Reynolds and Tamahana described a pilot project from the National Library of Medicine (NLM) Disaster Information Management Research Center (DIMRC), where a librarian was embedded on the hospital emergency preparedness team [12]. Most recently, as Featherstone described in late 2012, the ''disaster information specialist'' has emerged as a professional subdiscipline across medical (including hospital) librarianship with a distinct Disaster Information Specialization continuing education program from the Medical Library Association (MLA) [1].

METHODS
The method of data collection for this study was a simple survey. The research was exempted from institutional review board (IRB) oversight as the Aurora Health Care (for whom the primary investigator worked during the data collection phase of the project) IRB determined the survey did not constitute human subject research. The majority of survey questions were designed to collect quantitative data, although three questions allowed ''open answers'' that the investigators analyzed using qualitative research methods. The authors collected and analyzed data in an Excel spreadsheet. Qualitative data were organized into categories agreed on by both authors. Preliminary results were presented in poster format at MLA's annual meeting in 2012.
One of the authors (a hospital librarian herself at the time) wrote the questions for this study's survey instrument, utilizing previous reports on medical librarians and disaster management activities to inform question development [10,13,14]. An initial draft of the survey was sent to a small number of librarian disaster management experts, and the author used their feedback to revise the original questions. She posted the survey as a Google Documents form and opened it to respondents from February 15 through February 29, 2012.
The primary means for distributing the survey was medical librarian email discussion lists. An invitation to the survey along with some brief information as to Supplemental Appendix and Appendix B are available with the online version of this journal. the purpose (Appendix A, online only) was sent to the MEDLIB-L, DISASTR-OUTREACH-LIB, and the MLA Hospital Libraries Section email discussion lists. The message to these lists also encouraged cross-posting to local lists and other organizations' lists. A reminder notice was sent out over the same lists a few days prior to the date the survey was scheduled to close. In addition, the survey link was disseminated via Twitter.
An initial question asked respondents if they considered themselves hospital librarians. Those who answered ''no'' to the initial question saw only a thank you message. Those who answered ''yes'' were taken to the rest of the survey, which included some demographic questions as well as the questions about disaster management activities (Appendix B, online only, has a complete copy of the survey instrument).

RESULTS*
One hundred forty-nine hospital librarians completed the Google Documents survey. The investigators estimated a response rate range of 6.2%-17.5%, based on 2 previously reported methods for determining the number of hospital librarians in the country [15]. The first method uses the total number of hospital libraries that are listed in the National Network of Libraries of Medicine (NN/LM) Membership Directory, which was 2,416 (1,529 full members; the remainder are affiliates) as of February 2013. The second method uses the number of individuals subscribed to the MLA Hospital Libraries Section email discussion list, which was 850 as of March 2012. Three respondents indicated that they were from outside the United States. The most represented geographic region was the Greater Midwest (Iowa, Illinois, Indiana, Kentucky, Michigan, Minnesota, North Dakota, Ohio, South Dakota, Wisconsin) (34%), and the least represented was the Pacific Northwest (Alaska, Idaho, Montana, Oregon, Washington) (5%). The hospital type for which the most respondents worked was medium-sized community hospitals (200-500 beds, 26%), followed by smaller community hospitals (0-200 beds, 17%). An additional 20% indicated they worked for ''Teaching hospitals, any size,'' and 9% selected ''Academic medical center-affiliated hospital.'' A majority (60%, n590) of respondents indicated they had not been personally involved with disaster management activities in their hospitals. While 38% (n534) of these uninvolved librarians were interested in assuming a role but had not been assigned one, 31% (n528) reported that they had neither an interest nor a role to play.
A total of 32% of librarians (n528) who had not been involved in emergency management activities reported that they would have an assigned role to play in their hospitals during a disaster, and the investigators asked respondents to describe what those roles would be. The majority (n521) wrote that they would ''report for further assignment,'' during the response phase of the emergency or disaster. Librarians wrote that they would report and ''do what needs to be done'' or ''do what I am told to do,'' but very few (n55) knew what they would be doing once they checked in. Only 1 indicated a role providing information-in this case, ''providing information about various chemicals/conditions.'' The other 4 wrote that they would fill roles as ''runners,'' ''delivering messages,'' ''calling doctors,'' ''attending to patient families,'' ''directing patients,'' or ''preparing food.'' Three indicated that they would have a role in food services. Five respondents, who would not be asked to report for general duties as assigned, wrote that they would remain in the library and provide information services. Two of the librarians who would continue working in the library indicated that they would fill ''rush information requests'' and ''[n]ormal lit search and document delivery work, just faster than usual.'' The other 40% (n559) of the total respondents who indicated they had been involved in disaster management activities in their hospital selected 1 or more of 7 role categories provided by the investigators. The percentages of the total respondents for each role category are shown in Figure 1. Fifteen librarians selected the ''Other'' role; unfortunately, only 5 gave further description in an open comments box. These open responses included descriptions of the roles of library spaces, and 1 librarian indicated that the library computer room was also the hospital's disaster command center. Another participant reported being the fire warden for an entire floor and attending meetings and training sessions on disaster preparedness. One participant described being involved in disaster simulation training as a ''[f]ake patient.'' Three common scenarios emerged from the qualitative data collected on how librarians became involved in disaster management: participants volunteered or asked to be involved (n517), they were required (n510), or they were invited (n524). Three participants, who became involved as a result of being required, indicated that they were either supervisors or department heads and that it was compulsory for all managers at their level to join disaster or emergency planning committees or be involved with writing organizational disaster plans. Six participants, who either volunteered or were invited to participate in organizational emergency management activities had prior related experiences. One responder reported having ''[p]revious experience as an EMT with a volunteer ambulance corps, experience as a firefighter * The following results do not summarize all information gathered by the survey. In order to focus on emerging roles for hospital librarians in disaster management, responses to some demographic questions have been left out of this report, as has some additional information related to participants' interest in participating in a follow-up interview and their possible involvement in international disaster recovery efforts. These results may be helpful in future investigations and will be communicated through additional publications. on a volunteer fire department, [and] instructor with American Red Cross for CPR and first aid.'' Two librarians reported that they were involved in disaster relief efforts following 9/11 and Hurricane Katrina, and that these experiences influenced their decision to participate in disaster management activities at their hospitals.
In summary, 121 hospital librarians from around the country reported involvement, interest, and/or concrete roles in emergency and disaster activities. A final question for the cohort of librarians who reported that they were involved in emergency management asked if they played a part in a disaster or emergency that affected their hospital. Forty-two of the 59 respondents indicated no; the other 17 answered yes.

DISCUSSION
The primary limitation of the survey was participant self-selection. As it is safe to assume hospital librarians who are interested in disaster and emergency issues were more likely to participate, the investigators suspect that survey respondents underrepresent the opinions and activities of hospital librarians who are uninterested in disaster management.
Even though librarians indicated involvement during disaster response and recovery, one of the interesting patterns to emerge from this survey was that activities performed during this phase were of a nonprofessional nature (e.g., delivering food). Additional research is needed to investigate the nature of the roles played during different disaster management phases and to establish when librarians are able to utilize professional skills. The investigators hypothesize that hospital librarians are best able to support the preparedness activities of their institutions through their unique information skills and knowledge of hospital policy and procedure. While this benchmarking survey established that roles exist during all disaster management phases, future research will focus on how activities differ during each phase and how educational opportunities can support hospital librarians in taking on roles that better fit their professional skill set.
Along with providing evidence of existing roles for hospital librarians in institutional disaster management activities, the survey also uncovered three separate ways that hospital librarians tend to get involved: volunteering themselves, being invited, or being assigned a role. Identifying best practices for each of these pathways is another area of potential research interest and a way that MLA could further support hospital librarians.

CONCLUSIONS
This benchmarking survey provided quantitative and qualitative evidence from hospital librarians about their roles in disaster management activities. With 40% of survey respondents indicating that they had played a role in disaster management in their hospitals, data suggested the existence of a subgroup of librarians who are developing disaster management experience in hospital librarianship. The survey also provided the investigators with a view of how hospital librarians feel about being involved in institutional disaster management. While there is still a significant population of hospital librarians who are uninterested and content to remain uninvolved with disaster management (approximately 19% of respondents), an increasing number of librarians are either already involved or are interested in becoming Answers to the question, ''How are you involved in disaster management activities?'' As respondents could choose more than one role, the percentages add up to more than 100%.
involved. Educational opportunities, such as MLA's Disaster Information Specialization Program, that are targeted to librarians who are already engaged in or who are interested in disaster management activities are warranted. The survey established a current state of hospital librarian activities in disaster management. Repeating the survey in the future will allow researchers to track how these activities are evolving or changing over time.