Design and study overview
The pilot study was conducted with a cross-sectional design from August to September 2017. All four participating schools were approached with the Prevention Bus for an entire school week (see Fig. 1). The Prevention Bus team on-site at each school consisted of a bus driver and a medical team, including two physicians and two nurses. The bus, which has been converted into a doctor's office in 2016, enables vaccinations to be carried out directly on the school ground to the full extent of medical quality standards (for more information about the Prevention Bus see [28]). Information material and consent documents were distributed to all students and their parents by the school staff one week in advance. During the week on-site, students were addressed in their respective class contexts with two different on-site procedures diverging in degree of information and interactivity (intervention conditions: Educational Class vs. Low Intensity Information Condition). Both procedures ended with a joint visit of the class of the Prevention Bus and the opportunity to be vaccinated with MMR or Tdap-IPV. However, vaccinations could also be carried out on any other day of the week while bus and medical team were on-site.
Participant recruitment and informed consent
Schools in the pilot study were eligible for participation when they fulfilled two out of three inclusion criteria of the schools eligible for the main trial. Inclusion criteria were being an upper secondary school i.e. a high school, or a vocational school (for more information see Appendix 2 (a)), with at least 200 students in grades 9 to 11, and located in the city center of Berlin. This approach was chosen in order to achieve maximum comparability without interfering with schools eligible for the main trial. School administrations from selected schools were contacted in advance by letter and email and asked whether they wanted to participate. When an initial interest had been expressed, further questions were clarified by telephone or within a personal appointment at the respective school. In the pilot study, schools were not randomized to conditions.
If the school administration had agreed to participate, parents and students were informed one week in advance about the Prevention Bus project. Students within the selected schools were eligible for participation for the educational/basic information components of the intervention if they were currently attending grades 9 to 11. For on-site vaccination, if the vaccination card together with a consent form for the vaccination was completed. All minors had to present a consent form signed by their parents. For this purpose, the school staff provided the students with take home materials consisting of an information letter on the Prevention Bus project and consent forms for vaccination (hard copies). The parents were asked to complete and hand the documents back to their children to bring cards along to school on the day of the intervention. On the day of the on-site intervention all vaccine recipients, including minors, were educated and asked to sign a consent form for vaccination by themselves. In case of missing vaccination documents on the day of the intervention, these could be presented during the course of the on-site week and necessary vaccinations could be carried out at any time in the bus on the following days.
Intervention Conditions
At all four schools included in the study the Prevention Bus was present for an entire school week (see Fig. 1), which comprises five days. One intervention week comprised only three days due to public holidays (see Table 1). As described above, all students and parents were informed about the Prevention Bus project one week before the bus was at the respective school. The school secretary's office received take-home materials together with a declaration of consent form for the parents. The school administration was instructed to distribute the materials and to remind the school classes 1–2 days in advance to bring the materials, including their vaccination card, along on the day of the intervention.
At each school all potentially participating school classes were allocated lessons over the course of the week in which the students participated in the on-site intervention conditions, differing in information and educational content. While the possibility of being vaccinated was identical for all students at the participating schools, two planned intervention conditions for the main trial were tested, i.e., Educational Class or Low Intensity Information.
Educational Class Condition
At three schools (school 1, 2 and 3, see Fig. 1) a 90-minute procedure was carried out in individual school classes (Educational Class Condition [ECC]). This included an educational unit taught by a physician in the classrooms, the checking of vaccination cards, and the completion of the questionnaire by each student. The educational unit consisted of a digital PowerPoint presentation with interactive elements. Topics addressed were the immune system and infectious diseases, vaccination processes, herd protection, as well as risks and benefits specific to the measles vaccination. A group discussion as well as media elements such as newspaper articles and videos were also included. The educational unit was based on three theories of health behavior change, the SCT [21, 22], the PMT [20], and the HBM [23]. It was intended to increase knowledge by providing and discussing information [20–23], to address vaccination/disease-related risk perception by risk communication [20], and to use a testimonial by a fictional role model addressing self-efficacy [21, 22]. A detailed description of the educational unit and underlying theories can be found in the study protocol [28]. Immediately after the educational unit and the completion of the questionnaires, a guided tour through the Prevention Bus followed. During this tour, students had the opportunity to directly receive a vaccination if needed. Students could also have the vaccination carried out later in the week.
The ECC procedure largely corresponds to the planned procedure in the main trial. However, in schools 1 and 3 questionnaires were completed by students not only after but also before the educational unit (see Fig. 1) to test sensitivity to change of the items.
Low Intensity Information Condition
In school 4 (see Fig. 1), a 45-minute unit was conducted with the individual school classes (Low Intensity Information Condition [LIIC]). In each participating class physicians orally provided brief information about the Prevention Bus and vaccinations offered. Detailed vaccination questions were not discussed within the LIIC. Solely organizational information and information already given in the written vaccination consent form were raised. Subsequently, vaccination cards were checked, and the questionnaire was completed. Afterwards, the tour of the bus and the opportunity to receive vaccination directly at the tour or later on, if needed, were identical to the ECC-procedure.
Measures
To assess feasibility of the recruitment process, of on-site procedures and in order to evaluate measurement instruments, we collected data from seven process indicators and measures. The variables collected were: school recruitment log, vaccination documents, vaccinations delivered, rating of the educational unit, semi-structured interviews on the educational unit, a vaccination-related knowledge scale, and a vaccination-related perceived self-efficacy scale. In addition, covariates were assessed which could influence the vaccination behavior. This mixed-method data assessment facilitated a multitude of information feedback and potential target group-specific adaptation of processes and measurement instruments that were tested in the pilot study.
With the school recruitment log recruitment approaches and strategies were documented.
We assessed presence and contents of vaccination documents as an indicator of the feasibility of on-site procedures regarding requirement for vaccination in schools. The number of vaccination cards that were brought to school as well as the number of signed parental consent forms for vaccination were recorded. This allowed us to document who could potentially participate in a vaccination, as all required documents were brought along. In addition, we recorded the vaccination status (vaccinations: type and number of doses received) in order to determine the proportion of students with a need for vaccination. Vaccination status for all standard vaccinations recommended of the German Vaccination Committee “Ständige Impfkommission am Robert-Koch Institut” (STIKO) was documented [29]. All indicators were documented by a physician or nurse, during the ECC/LIIC before vaccination on-site was offered.
For the vaccinations delivered on-site the medical team documented how many students received a vaccination in the Prevention Bus. In this study, combination vaccination against mumps, measles, rubella (MMR [30]) and tetanus, diphtheria, pertussis and polio (Tdap-IPV [31]) were offered. Whether students had an indication/need for vaccination or not was determined again in accordance with the recommendations of the STIKO [29].
The rating of the educational unit by all students in the ECC was documented. The students evaluated the course and content of the educational unit with a single item on a 3-point Likert scale (scale: very interesting, interesting, boring), applied after the ECC took place. Further feedback was recorded with semi-structured interviews on the educational unit with selected students (single and group settings) in the ECC to ascertain how easily understandable the contents of the educational unit were and which components were of particularly high or low interest. In addition, students were asked to evaluate the interactive components of the educational unit. This reflects for example a guided group discussion, which was triggered by a fictional social media post. Students had to evaluate how realistic the scenario in the post was to them. Furthermore, students were asked whether their attitude towards vaccination had changed after the presentation. The interviews were conducted after completing the individual questionnaire, but before students entered the bus.
The vaccination-related knowledge scale indicated, with a maximum of 7 items, the level of immunization knowledge (see Appendix 2 (b)). Figure 2 provides an overview of which items were assessed at which school. Items covered facts about target groups of infectious diseases, prevention of the spreading of infectious diseases, herd immunity, vaccination and side effects of infections. Items were multiple choice, with one correct answer. Across all items an overall knowledge sum score could be generated.
Two different sum scores were calculated in the pilot study. The 4-item sum score included items 1–4 (Childhood Disease, Herd Immunity, Immunization, Measles Outbreak) and was assessed in all four schools (see Fig. 2). In addition, for school 3 and 4 a 6-item sum score could be computed as all final items were applied including the newly successively introduced items 6 and 7 (Infertility, Pregnancy Malformation). A low sum score (0) represents a low level of vaccination-related knowledge and a high score (4 or 6) a high level of vaccination-related knowledge. Item 5 (Bacteria) was not selected for the formation of scales2.
The vaccination-related perceived self-efficacy scale, reflected students’ beliefs about their own competency to understand, appraise, and apply vaccination-related health information (see Appendix 2 (b)). The scale consisted of a short-version of the health literacy questionnaire HLS-EU Q47 [32]. The short-version was created by selecting those items, which covered vaccination and information on prevention. Students were asked to rate five items on 4-point Likert type scales ranging between 1 (very difficult) and 4 (very easy). A mean score was computed, with low scores representing low and high scores high levels of perceived self-efficacy.
Additionally, covariates, including gender, socioeconomic, and migration status, were assessed on school and/or individual level. Regarding socioeconomic status (SES) two items were applied, assessing whether students had a room of their own [33, 34] and had access to reduced fares in public transport and leisure activities (i.e., “berlinpass”). Migration background was documented on school level, indicating the ratio of students with a foreign language background.
Analysis
Data were analyzed using IBM SPSS Statistics 25. Descriptive data of the school recruitment log was used to evaluate the feasibility of recruitment processes. Prior to scale analyses, missing values for all final vaccination-related knowledge items (ranging from 2.2% – 7.3%) and all perceived self-efficacy items (ranging from 9.5% – 12.0%) were imputed with a Multiple Imputation (MI) procedure using fully conditional specification and 20 imputed datasets.
According to the school recruitment log, a ratio between targeted schools and successfully recruited schools was used. Further, absolute numbers of vaccination documents brought to school (the number of vaccination cards present, consent forms returned) and vaccinations delivered on-site were reported. A relative ratio of students with a need for vaccination and vaccinations carried out in this study were reported at a descriptive level.
For the vaccination-related perceived self-efficacy scale, the internal consistency (i.e. Cronbach’s Alpha) and the factorial structure (varimax-rotated exploratory factorial analysis [EFA]) was calculated. To determine significant changes from pre to post intervention in vaccination-related knowledge and perceived self-efficacy, Linear Mixed Models (LMM) for continuous outcomes (covariance type = variance components, estimation = Restricted Maximum Likelihood) and Generalized Estimating Equations (GEE) for dichotomous outcomes (distribution = binomial, link function = logit) were used; these allow for an adjustment for the nested structure of students in classes and control for the influence of age, gender, and SES. The longitudinal pre-post-educational unit measurements of the questionnaire in the same school class (see Fig. 1, school 1 and 3) were used specifically for the pilot study to test the sensitivity to change of the vaccination-related knowledge and perceived self-efficacy measurements. In addition, we compared schools in which the questionnaire was collected only once, either post-educational unit (school 2) or post-basic information (school 4).
All tests of significance were based on p < 0.05 level and confidence interval of 95% was applied.
In order to evaluate the feasibility of the educational unit, frequencies of students rating of the educational unit were used. Furthermore, a research assistant coded the transcripts of the semi-structured interviews and interrater reliability was tested with another research assistant coding the transcript. Mean-level results and distributions are presented at a descriptive level.