From 1976 to 2007, influenza was associated with an average of 23,607 deaths per year (min 3,349; max 48,614; Center for Disease Control and Prevention [CDC], 2010a). The Advisory Committee on Immunization Practices (ACIP) now recommends influenza vaccination as an effective preventive measure for over 99% of the population (CDC, 2010b). As its recommendation for children has evolved from risk-based to age-based, and, eventually, to a universal recommendation, school-located influenza vaccination (SLV) has been proposed as a potentially viable and cost-effective supplement to vaccinations delivered in health care provider offices (Cawley, Hull, & Rousculp, 2010; Hull, 2011; Hull, Frauendienst, Gundersen, Monsen, & Fishbein, 2008; Schmier, Li, King, Nichol, & Mahadevia, 2008). We define SLV clinics as clinics primarily designed to vaccinate school children, on school grounds, during or outside of school hours (CDC, 2009a).

Prior to the 2009 H1N1 pandemic, influenza SLV clinics were held in some states at the local level sporadically, and state-wide in Hawaii for multiple years (Effler et al., 2010; Hull, 2010; King et al., 2005; King et al., 2006; Schmier et al., 2008; White, Lavoie, & Nettleman, 1999). During the 2009 H1N1 influenza pandemic, SLV clinics were used by states and local jurisdictions to vaccinate school-age children, one of the priority groups for the 2009 H1N1 influenza vaccine (CDC, 2009b). In Maine’s 2009 state-wide SLV program for both H1N1 and seasonal vaccines, over 95% of schools conducted or participated in SLV clinics (Lorick et al., 2011). Maine’s state health department (Maine Center for Disease Control and Prevention), promoted SLV clinics and offered limited funding and some supplies and materials to schools that conducted SLV clinics (Thomas & Mills, 2010). However, schools and communities were asked to independently organize or form partnerships with other health entities to make the SLV clinics feasible. Maine achieved 60.2% coverage for H1N1 vaccine among children and adolescents 6 months to 17 years (national median coverage, 6 months to 17 years: 36.8%; Lu et al., 2010).

The planning process for SLV can be challenging because of the need to minimize school and classroom disruption. Coordination of SLV clinics also introduced new challenges—bringing in new partnerships and agencies (Figure 1). In Maine, each district’s SLV program varied in the types of partnerships between schools, public health, and other community entities (Cawley et al., 2010). Who vaccinated was also left up to the school districts; some districts partnered with other health entities (e.g., Visiting Nurses Associations or hospitals) and at other districts, school nurses vaccinated on their own. Maine’s SLV clinics also varied in their implementation approach; the primary approach was onsite vaccination during school hours, although some SLVs vaccinated off-site and outside of school hours (Lorick et al., 2011).

In Maine, many school nurses (66%) serve at more than one school in their district (Oostveen-Buterbaugh, 2009). Maine also has public health nurses who serve schools without designated school nurses in the most rural parts of the state. Consent form processes were primarily handled by school nurses and staff.

Because of the planning and consent form processes required for SLV clinics, we expect there to be considerable labor hours allocated to outside-clinic coordination activities (OCA). We define OCA as coordination activities conducted outside of the operating hours of the SLV clinic. OCA include planning time, but also time spent reviewing consent forms, contacting parents outside of clinic hours, or billing insurers. OCA costs exclude regular clinic costs, which are the labor, materials, and vaccine needed to run clinics (from setup to breakdown). Past research has focused on regular clinic costs and some studies have evaluated the time spent planning clinics (Carpenter, 2007; Hull, 2010). We build on these studies by conducting a detailed analysis of OCA costs by time period (before, during, and after SLV clinics). We do not include regular clinic costs in this analysis; they were analyzed and reported separately (Cho et al., 2011).

There were eight respondents to the survey, representing seven school districts, 16,645 students, 45 schools, 84 clinics, and 22,596 doses of vaccine (14,714 of 2009 H1N1 and 7,882 of seasonal influenza vaccine). Of the seven school districts in the sample, one was located in a city, two were in suburbs, two were in towns, and two were in rural areas (U.S. Department of Education). For job titles, four respondents were “lead nurses,” one nurse was a “parttime nurse,” one nurse held the title of “School Nurse Coordinator,” and another nurse had the title of “Health Services Coordinator.” Because one district had a lead and non-lead nurse respond, we included the non-lead nurse’s time in the “other district staff” category. School districts varied greatly in size and socioeconomic status. The median school district population was 1606 students (min 838, max 7,023; U.S. Department of Education, 2010). The percentage of students eligible for free and reduced lunch ranged from 7% to 58% (Table 2; Maine Department of Education, 2011).

The mean number of clinics per nurse was 12 (min 4, max 30; Table 2) and the mean total hours per clinic was 51 (Table 3). Of mean total hours, nurses reported a mean of 22 extra hours per clinic (42% of total nurse time). For non-lead nurse district staff, labor hours for all periods averaged 18 hr per clinic. In total, for all school district staff (including lead nurses), the mean time for OCA was 69 hr per clinic (Table 3). Lead nurse time per clinic was greatest during the planning period (27 hr), then the implementation period (17 hr), and last, the post-implementation period (7 hr). Only four school districts reported time spent by other staff. Activities of high level administrators were related to meetings to discuss executive decisions for location of SLV clinics, staffing schedules, and press releases. Secretaries, volunteers, and other staff helped with all aspects of clinic preparation including, but not limited to consent process, Vaccine Information Statement forms, and writing letters to parents.

The State of Maine achieved 60.2% coverage for 2009 H1N1 influenza vaccine among children and adolescents 6 months to 17 years; a level much higher than the national median (36.8%; Lu et al., 2010). It is likely that this high level of coverage was due in part to the success of their SLV program; which was the method by which almost all school-age children received their vaccine in Maine. In fact, literature suggests that SLV programs may vaccinate more children than would have otherwise been vaccinated (Hull, 2011; Humiston et al., 2011). Our results show that school nurses were an integral part of this effort. The majority of their time was spent on critical activities such as communications and consent processes. Further, a large portion of lead nurse time was spent outside of regular working hours, which, although not uncommon during a public health emergency response (CDC, 2011) places significant burden on one individual and can create challenges for the sustainability of SLV clinics during non-pandemic seasons.

Given the results of a related evaluation conducted on the labor and materials cost to run SLV clinics in the state of Maine from 2009 to 2010 ($13.51 per dose; Cho et al., 2011), adding our OCA cost ($15.36) yields a total per dose cost of $28.87 excluding vaccine and syringe costs. This cost is higher than other recent studies on SLV clinic costs, which have suggested a range of $20 to $27 per dose during non-pandemic seasons (including vaccine costs; Cawley et al., 2010; Hull, 2011). Nevertheless, nurses did report an expected 76% reduction in planning time if they were to conduct SLV clinics in the following year. Assuming other costs stay the same, this would reduce the OCA cost estimate from 15.36 to a total of $10.08 per dose (the planning period represented $6.95 per dose, 76% reduction in planning costs implies a reduction of $5.28 per dose: 15.36–5.28 = 10.08).

During 2009–2010 SLV program, nurses reported delays in “usual” activities such as hearing and vision and BMI screenings. In lower income areas, where primary care doctors are less likely to be present (Shipman, Lan, Chang, & Goodman, 2011), schools may serve as a “public health safety net,” making routine screenings by nurses more important. Finally, because the 2009–2010 H1N1 season was a pandemic influenza year with two influenza vaccines, SLV activities were less predictable than in a regular season. If influenza SLV clinics become routine, other school nurse activities would likely be scheduled more efficiently.

Superintendents, principals, secretaries, and non-lead nurses were also involved in OCA. Though they contributed fewer hours to the SLV clinic effort, according to initial interviews and survey comments, they were also important in making sure that the school district would have a successful vaccination program. Regarding lead school nurses, even if the regular clinic model changed, where school nurses were not the primary coordinators, school nurses can be an important bridge between schools and public health stakeholders.

Our study is subject to several limitations. First, we focus on clinics that were primarily planned and implemented by school nurses and were held during school hours, with parental presence not required; other approaches would likely affect the amount of nurse time required. However, many of the activities would still need to be conducted (e.g., communication and consent form process) regardless of approach used. Second, we evaluated SLV clinics during a pandemic year, which could have increased the coverage in schools (potentially reducing the average cost), but also increased the amount of time necessary to plan clinics (e.g., delays in vaccine shipments, more communication time). Additionally, our cost per dose would be lower if we were able to account for the revenue generated from health insurance reimbursement. Third, the results are subject to recall bias because our survey was detailed and conducted 6 months after the last SLV clinic; however, some nurses kept detailed time logs and/or consulted their payroll records. Fourth, lead nurses reported time for other staff members, which was likely to be less accurate than if other staff members had reported the time themselves. Finally, several districts also reported the need for copy and translation services; however, none of the respondents reported quantities and costs for translation services. For these reasons, we excluded materials costs from the analysis, making our analysis an underestimate. It is reassuring to note that many respondents stated anecdotally that there were no major materials costs associated with planning and coordinating SLV clinics.

The H1N1 influenza pandemic of 2009 was the first time that many practicing school nurses and other partners in Maine conducted or participated in school vaccination clinics. This likely presented a steep learning curve and potentially increased time needed for implementing SLV programs. The 2009 influenza season also included two doses, increasing the complexity.

Our study provides a previously unpublished detailed breakdown of costs related to planning and coordination activities outside of clinic operations. Activities such as consent and communications processes were very resource intensive; however, these activities can become more streamlined over subsequent, non-pandemic influenza seasons. There are many synergies and benefits that can be gained from collaboration between public health and schools; however, schools, whose primary mission is education, may need additional support in order to run school clinics especially if clinics are held in school, during school hours, with school nurses vaccinating.