Use of a Pandemic Preparedness Drill to Increase Rates of Influenza Vaccination Among Healthcare Workers
Objective. To determine the effect of a pandemic influenza preparedness drill on the rate of influenza vaccination among healthcare workers (HCWs).
Design. Before‐after intervention trial.
Setting. The University of Iowa Hospitals and Clinics (UIHC), a large, academic medical center, during 2005.
Subjects. Staff members at UIHC.
Methods. UIHC conducted a pandemic influenza preparedness drill that included a goal of vaccinating a large number of HCWs in 6 days without disrupting patient care. Peer vaccination and mobile vaccination teams were used to vaccinate HCWs, educational tools were distributed to encourage HCWs to be vaccinated, and resources were allocated on the basis of daily vaccination reports. Logit models were used to compare vaccination rates achieved during the 2005 vaccination drill with the vaccination rates achieved during the 2003 vaccination campaign.
Results. UIHC vaccinated 54% of HCWs (2,934 of 5,467) who provided direct patient care in 6 days. In 2 additional weeks, this rate increased to 66% (3,625 of 5,467). Overall, 66% of resident physicians (311 of 470) and 63% of nursing staff (1,429 of 2,255) were vaccinated. Vaccination rates in 2005 were significantly higher than the hospitalwide rate of 41% (5,741 of 14, 086) in 2003.
Conclusions. UIHC dramatically increased the influenza vaccination rate among HCWs by conducting a pandemic influenza preparedness drill. Additionally, the drill allowed us to conduct a bioemergency drill in a realistic scenario, use innovative methods for vaccine delivery, and secure administrative support for future influenza vaccination campaigns. Our study demonstrates how a drill can be used to improve vaccination rates significantly.
Received July 5, 2007; accepted October 23, 2007; electronically published December 31, 2007.
Healthcare workers (HCWs) are at risk of contracting influenza in the healthcare environment and transmitting it to patients under their care.1‐5 The influenza vaccine is the most effective means for preventing both the nosocomial spread of influenza and illness‐related absenteeism among HCWs. For these reasons, the Centers for Disease Control and Prevention (CDC) recommends annual influenza vaccination for all HCWs.
Despite the CDC’s recommendation, on average, less than 40% of HCWs receive influenza vaccination each year.6 However, influenza vaccination rates for HCWs are not low at all healthcare institutions, and some facilities have implemented interventions to increase vaccination rates.7‐9 Most of the successful interventions require substantial resources and administrative support to achieve and sustain higher vaccination rates.8
In 2005, the University of Iowa Hospitals and Clinics (UIHC) conducted a pandemic influenza drill, which used the Hospital Emergency Incident Command System (HEICS), to distribute vaccine at the start of its annual influenza vaccine campaign. UIHC had 3 goals for the drill: (1) to increase vaccination rates among HCWs without disrupting patient care, (2) to increase emergency preparedness, and, (3) to carry out 1 of the 2 mandatory annual drills to test emergency management plans that are required by the Joint Commission on Accreditation of Healthcare Organizations.10 This paper describes the drill and its effect on influenza vaccination rates among HCWs at UIHC.
Methods
Setting
UIHC is a 680‐bed, academic tertiary medical center with 281 specialty clinics for ambulatory patients. The hospital staff consists of more than 700 staff physicians, 450 resident physicians, and 180 physicians who are fellows at the institution. UIHC also employs more than 2,000 nurses and more than 4,500 other professional staff members. UIHC has historically provided influenza vaccination at no charge to all hospital employees during a campaign that lasts for approximately 4 weeks each fall. Vaccination was available during this period at the employee health clinic, Monday through Friday from 8:00 am to 5:00 pm. Prior to 2005, vaccination was not routinely available to employees outside of the employee health clinic.
Intervention
During the last week of October 2005, UIHC implemented a pandemic influenza preparedness drill in anticipation of the 2006 requirement to test the response phase of its emergency management plan. Instead of being a tabletop exercise, this drill included a component for immunizing many HCWs quickly (ie, mass vaccination). When the drill was initiated, hospital administrators were presented with the following scenario and directive: “A strain of influenza has caused serious illness and several deaths in other states; therefore, the Iowa Department of Public Health directs that healthcare workers providing direct patient care be immunized immediately.” Because this directive required the hospital to mobilize extensive resources, administrators activated the HEICS component of UIHC's disaster and emergency preparedness plan.
The HEICS approach organizes disaster response from 5 different functional areas: administration, logistics, planning, finance, and operations.11 These 5 areas are each led by an incident commander.12 The incident commander then assigns prioritized activities to various people in leadership positions.
At the beginning of the 6‐day drill, UIHC’s HEICS leadership established an operational goal of vaccinating 70% of HCWs who had direct contact with patients before the end of the drill. The organizers of the drill chose this vaccination rate because it approaches the minimum level of vaccination needed to achieve herd immunity.13 UIHC purchased influenza vaccine for 2005 before the drill was organized, so the necessary vaccine supply was estimated on the basis of the number of vaccinations given in past campaigns. For this reason, HEICS leaders were concerned that the initial vaccine supply might not be sufficient, and during the 6‐day drill, UIHC vaccinated direct caregivers and other persons who routinely worked within 1 m of patients; employees and volunteers who did not provide direct patient care were asked to wait until after the 6‐day drill to receive their influenza vaccination. This recommendation was made the first day of the drill and was relaxed in subsequent days as it became apparent that more vaccine would be readily available.
The drill included several methods for dispensing vaccine. Nurse “champions” coordinated a peer vaccination and dispensing program for the staff on their units. Nurse champions also educated their staff about influenza vaccination, promoting vaccination as the most important preventive measure. The University Employee Health Clinic provided the nurse champions with packets containing vaccine, necessary supplies, and consent forms. These supplies were replenished by staff in the Program of Hospital Epidemiology as needed. Additionally, nurse champions were given a pager number they could use to place an immediate request for more vaccine or supplies. The Department of Pharmacy helped coordinate the vaccine supply in the distributing pharmacies and also at the unit level. Mobile vaccination teams—nurses equipped with wheeled carts carrying vaccination supplies and laptop computers—complemented the peer vaccination and dispensing program’s activities by responding to special requests to vaccinate staff at different venues (eg, at medical staff meetings or during grand rounds). During the drill, HCWs also could be vaccinated by staff of the University Employee Health Clinic. For 2 weeks after the drill, all hospital employees and volunteers could receive influenza vaccination at no charge from the University Employee Health Clinic during the clinic’s usual hours (ie, Monday through Friday from 8:00 am to 5:00 pm). After this 2‐week period, vaccination was still available, but it was no longer free.
Throughout the drill, hospital epidemiology and employee health staff distributed educational and promotional materials describing the benefits of influenza vaccination. New educational materials included: (1) “Flu Facts” screensavers displayed on computers throughout clinical areas; (2) posters placed at the main entrance of the hospital and in staff dining areas, nursing units, and clinics; (3) e‐mail broadcasts sent to hospital staff; and (4) educational talks given at leadership meetings. HCWs who had been vaccinated were given stickers that said “I got it” to place on their identification badges.
Staff in the Program of Hospital Epidemiology used information from consent forms and personnel records to calculate vaccination rates for the whole hospital, vaccination rates by department and division, and vaccination rates for specific groups, such as nurses, medical residents, and physicians. These rates were subsequently reported during daily meetings with the HEICS leadership and distributed over a shared Web site to midlevel leaders (eg, nurse managers). The HEICS leadership used the vaccination rates as immediate feedback to direct resources to specific areas (eg, mobile vaccination team routes), in an effort to increase vaccination levels on specific units and to assess overall vaccine supply.
UIHC did not repeat the emergency influenza drill in 2006. However, UIHC used the educational approaches introduced during the drill (eg, screensavers, posters, and stickers) and also used similar dispensing methods including the peer vaccination and dispensing program and the mobile vaccination teams. UIHC continued to track HCW vaccination rates but provided feedback on rates to unit‐level leadership, such as nurse managers, rather than to administrators and did not provide this feedback in real time. Additionally, the daily administration meetings held during the drill did not occur.
Statistical Methods
We compared the influenza vaccination rates for the vaccination campaigns in 2003, 2005, and 2006. We excluded the results of the 2004 season from our analysis because the vaccine supply was severely limited. We used logit models to compare the vaccination rates for 2005 and 2003, for both nurses and residents. The P values for the comparisons were based on Wald test statistics. For nurses, rates were available for 54 clinical units. Thus, we included a random effect in this model to account for the associations between the subgroup‐specific rates for 2005 and 2003, and we fit the model using generalized estimating equations.
Results
During the 6 days of the 2005 vaccination drill, we vaccinated 54% of the HCWs (2,934 of 5,467) who had direct contact with patients. Over an additional 2 weeks, we increased the vaccination rate to 66% (3,625 of 5,467). The overall vaccination rate for all hospital employees (ie, those with and those without direct patient contact) was 51% (6,539 of 12,873) in 2005, compared with a vaccination rate of 41% (5,741 of 14,086) in 2003. For nurses on 54 clinical units, vaccination rates increased from 39% (793 of 2,025) to 63% (1,429 of 2,255) (
), and among resident physicians, vaccination rates increased from 46% (197 of 423) to 66% (311 of 470) (
).
In certain subgroups (eg, clinical units or departments), the vaccination rate increased substantially. Among neurosurgery residents and physicians who were fellows of the institution, the rate increased by 39% (in absolute terms), compared with their 2003 rates, although substantial increases also occurred among resident physicians in internal medicine (28%), pediatrics (25%), and anesthesia (17%). Overall, the median increase in vaccination rates for the 54 nursing units studied was 22%; changes in the vaccination rates on these units ranged from a decrease of 12% to an increase of 60%. In the 10 nursing units with the highest vaccination rates, the rate increased by a mean of 35% during the 2005 campaign, compared with the rates in 2003. Examples of notable increases include a 60% increase in the vaccination rate among nurses in the transplant clinic, a 50% increase among nurses on the adult bone marrow transplant unit, a 47% increase among nurses on a pediatric unit, and a 32% increase among nurses on an adult intensive care unit.
The peer vaccination and dispensing program administered 82% of the vaccine, the University Employee Health Clinic staff administered 12%, and the mobile vaccination teams administered 6%. Additionally, 41% of the vaccinations administered during the entire campaign were given during the first 2 days of the drill.
In 2006, the overall vaccination rate for the campaign was 64% (4,168 of 6,438) for HCWs with direct patient contact and 49% (5,999 of 12,191) for hospital employees overall (ie, those with and those without direct patient contact). Vaccination rates were 66% (1,446 of 2,205) for nurses and 69% (464 of 670) for resident physicians and physicians who were fellows at the institution.
Discussion
Influenza vaccination rates for HCWs at UIHC increased substantially after an influenza preparedness drill that included a vaccination component. Additionally, the pandemic influenza drill addressed 2 important needs in a single exercise: the drill allowed UIHC both to test different vaccination delivery methods and to conduct a bioemergency drill in a realistic scenario while simultaneously fulfilling a requirement established by the Joint Commission on Accreditation of Healthcare Organizations.10
Innovations in vaccination delivery methods were essential to the success of this drill. We were motivated to implement new methods for delivering influenza vaccination because we realized that the old approaches at UIHC did not work. For example, the new peer vaccination program allowed UIHC to deliver vaccination quickly without disrupting patient care. Because we vaccinated HCWs where they worked, they did not have to walk to the University Employee Health Clinic during their lunch hour or break. In addition, nurse champions, who vaccinated their peers, had closer working relationships with their coworkers than the most visible proponents of previous influenza vaccination campaigns (eg, infection control professionals). Mobile vaccination teams administered only a small percentage of vaccinations, perhaps because other delivery approaches were so successful.
Although the drill was associated with improved vaccination rates, the extensive resources needed to conduct the drill (eg, daily meetings with numerous senior hospital administrators) precluded us from running another drill in 2006. However, the drill required the cooperation of personnel from many different departments, thereby demonstrating to hospital administrators the effort needed to develop a more effective vaccination program. In 2006, the year after the drill, UIHC devoted significantly more resources to the influenza vaccination campaign than it did in the years prior to the drill. For example, in 2006, staff members from many groups helped plan and execute the vaccination campaign. These planning groups included hospital administration, pharmacy, nursing, and information systems. Most importantly, we continued the peer vaccination program in 2006, which likely sustained the higher vaccination rates first achieved in 2005. Although the higher vaccination rates were sustained in 2006, the time that was needed to vaccinate a substantial proportion of HCWs increased significantly without the drill. During the 2005 drill, UIHC vaccinated 41% of staff members with direct patient contact in 2 days, whereas the 2006 campaign required more than a week to achieve a similar vaccination rate. However, this difference is probably not clinically significant (except in the event of an early influenza season).
UIHC will need to conduct subsequent influenza vaccination campaigns before we can determine whether the additional resources will sustain the higher vaccination rates. Unfortunately, despite a dramatic improvement, vaccination rates at UIHC still remain suboptimal and additional effort and resources may be needed if rates are to continue to increase.
Besides improving vaccination rates, the drill was associated with other process improvements. Because our drill was a “real‐world” rather than “tabletop” vaccination exercise, we identified logistical problems (eg, deficiencies in the hospital’s personnel database and in its ability to track vaccine supply) that we otherwise would not have recognized. For example, we had difficulty collecting and interpreting data throughout the drill. To track vaccination delivery and generate daily vaccination reports, we had to identify which HCWs did and did not provide direct patient care and identify which HCWs had and had not been vaccinated. Our information management systems did not allow us to quickly identify HCWs who provide direct patient care at UIHC. Because UIHC is an academic medical center, not all persons classified as “physicians” or “nurses” provide direct patient care. For example, some physicians and nurses are engaged entirely in research activities. We may have underestimated the actual vaccination rates for direct healthcare providers because we inadvertently included some research staff in the denominator count data. Additionally, we could not track the influenza vaccination status of HCWs who received vaccination at a location other than UIHC, which may have caused us to further underestimate actual vaccination rates.
We determined the vaccination status for most HCWs by manually entering information from vaccination consent forms into a database. At least once a day, volunteers picked up completed forms and replaced them with a new supply. Throughout the drill, HEICS leadership met and reviewed vaccination rates to determine how best to allocate resources. Therefore, we needed to enter the data as quickly as possible to generate useful reports. However, the data collection process was labor‐intensive and required an extensive time commitment from drill organizers. As a result, we learned that our information‐management system needs to be able to better identify staff members involved in patient care and to quickly update and retrieve information about these individuals (eg, vaccination status).
The Healthcare Infection Control Practices Advisory Committee and the Advisory Committee on Immunization Practices recently issued evidence‐based recommendations for influenza vaccination programs.6 The Infectious Diseases Society of America’s Emerging Infections Network conducted a survey among its members to assess the implementation of these recommendations and to determine the components of a successful influenza vaccination campaign.14 On the basis of the survey, the Emerging Infections Network determined that administrative support was one of the most important components of successful hospital influenza vaccination programs.14 However, the current literature lacks studies that demonstrate how hospital epidemiologists and infection control professionals can convince healthcare administrators to increase support and resources for influenza vaccine campaigns. Although other groups previously used drills to test their capacity to vaccinate large numbers of people quickly,15 our study demonstrates how an emergency preparedness drill can improve vaccination rates significantly and motivate hospital administrators to implement and sustain evidence‐based practices that increase influenza vaccination rates among HCWs. Given the Joint Commission on Accreditation of Healthcare Organizations requirement (EC.4.20) that hospitals conduct drills to test their emergency management plans twice per year,10 other hospitals might be able to use a similar drill to improve influenza vaccination rates.
Acknowledgments
We would like to thank Dr. John H. Staley for his leadership and all of the administration and staff who contributed to the 2005 UIHC pandemic influenza preparedness drill.
Potential conflicts of interest. All authors report no conflicts of interest relevant to this article.
References
- 1. Elder AG, O’Donnell B, McCruden EA, et al. Incidence and recall of influenza in a cohort of Glasgow healthcare workers during the 1993‐94 epidemic: results of serum testing and questionnaire. BMJ 1996; 313:1241‐1242.
- 2. Horcajada JP, Pumarola T, Martinez JA, et al. A nosocomial outbreak of influenza during a period without influenza epidemic activity. Eur Respir J 2003; 21:303‐307.
- 3. Salgado CD, Farr BM, Hall KK, Hayden FG. Influenza in the acute hospital setting. Lancet Infect Dis 2002; 2:145‐55.
- 4. Harrison J, Abbott P. Vaccination against influenza: UK health care workers not on‐message. Occup Med 2002; 52:277‐279.
- 5. Gardner P, Court S, Brocklebank J, et al. Virus cross‐infection in pediatric wards. Br Med J 1973; 9:571‐575.
- 6. Bridges CB, Harper SA, Fukuda K, Uyeki TM, Cox NJ, Singleton JA; Advisory Committee on Immunization Practices (ACIP). Prevention and control of influenza: recommendations of the ACIP. MMWRRecomm Rep 2003; 52:1‐34.
- 7. Fedson DS. Influenza vaccination of medical residents at the University of Virginia: 1986‐1994. Infect Control Hosp Epidemiol 1996; 17: 431‐433.
- 8. Pottinger JM, Herwaldt LA. Improving health care worker compliance with influenza immunization. In: Poland GA, Schaffner W, Pugliese G, eds. Immunizing Health Care Workers: A Practical Approach. Thorofare, NJ: Slack; 2000.
- 9. Salgado CD, Giannetta ET, Hayden FG, et al. Preventing nosocomial influenza by improving the vaccine acceptance rate of clinicians. Infect Control Hosp Epidemiol 2004; 25:923‐928.
- 10. Joint Commission on Accreditation of Healthcare Organizations (JCAHO). 2006 Hospital Accreditation Standards For Emergency Management Planning, Emergency Management Drills, Infection Control, and Disaster Privileges. Oakbrook Terrace, IL: JCAHO; 2003.
- 11. The San Mateo County Health Services Agency, Emergency Medical Services. The Hospital Emergency Incident Command System. 3rd ed. San Mateo, CA: The San Mateo County Health Services Agency; 1998.
- 12. Arnold, JL, Dembry L‐M, Tsai M‐C, et al. Recommended modifications and applications of the Hospital Emergency Incident Command System for hospital emergency management. Prehosp Disaster Med 2005; 20:290‐300.
- 13. Patriarca PA, Weber JA, Parker RA, et al. Risk factors for outbreaks of influenza in nursing homes: a case‐control study. Am J Epidemiol 1986; 124:114‐119.
- 14. Polgreen PM, Chen Y, Beekmann S, Cavanaugh JE, Herwaldt LA. Components of influenza vaccination programs that predict higher vaccination rates: results of an EIN survey. In: Program and abstracts of the 17th Annual Scientific Meeting of the Society for Healthcare Epidemiology of America (Baltimore). 2007. Abstract 210.
- 15. Pur S, Vestal A, Connolly P, Segreti J, Trenholme G. Mass vaccination in a hospital setting: the push and pull approaches. In: Program and abstracts of the 15th Annual Scientific Meeting of the Society for Healthcare Epidemiology of America (Los Angeles). 2005:72. Abstract 46.