A photograph of the researcher on the invitation letter did not affect the participation rate of a postal survey: a randomized study within a trial (SWAT) | BMC Medical Research Methodology


This SWAT was registered in the Queens University Belfast SWAT Store as SWAT 104 (04/10/2019).


As part of a population-based survey on the attitude of the general German population towards medical second opinions, we conducted a SWAT with a parallel-group-based, randomized controlled study with a 1:1 assignment. Please refer to the original paper for details [15]. The survey is part of a mixed-method study on medical second opinions in Germany [16]. We briefly describe the design of the original study: We randomly identified 9990 people from 25 randomly selected registration offices in the states of Berlin and Brandenburg. We chose these two federal states because they cover a settlement structure with both very high and very low population densities. The participants were selected using a disproportionately stratified sample, with the settlement structure (cities, communities and surrounding areas, rural areas) serving as a stratification variable, which led to three samples of equal size. Settlement patterns are defined by Eurostat’s ‘degree of urbanisation’ [17]. The registry offices belonged to cities (n = 5), cities and suburbs (n = 10) and rural areas (n = 10). There are only 5 urban areas in these states, which is why we chose all of them. See Fig. 1 for the detailed process. In addition, the registry offices contained data on gender and most also on age. The questionnaire contained 47 items on 14 pages. It comprised 8 parts: health-related items (2 items), on-site care situation (5 items), patient needs regarding second opinions (7 items), possible experiences with second opinions (8 items), design of the second opinion procedure (8 items), experiences with and knowledge of second opinion programs of health insurance companies (3 items), experiences with and knowledge of the offer of second opinion providers (3 items). The last part included sociodemographic characteristics (11 items). We used the 16-item European Health Literacy Survey to assess health literacy [18].

Fig. 1
illustration 1

Participant Flow Chart. *Defined by Eurostat. ‡Unable to determine if a questionnaire was successfully delivered to the recipient


As this was a population-based survey, we had no selection criteria other than age (≥ 18 years) and residence in Berlin or Brandenburg. The questionnaire was mailed twice (first and 6 weeks later with a reminder) together with the invitation to participate. The questionnaire can be sent back to us in a reply envelope. When they returned the questionnaire to us, the participants could also enclose their declaration of consent to participate in a prize draw for 125 Amazon vouchers (€50 each) in a separate envelope.

Also Read :  KalVista Pharmaceuticals Presents New Patient-Centric Data at 2022 HAEi Global Leadership Workshop


The intervention group received a letter of invitation with a color photo of two researchers, while the control group received a letter of invitation without a photo. Otherwise the letters were identical. See Supplementary File 1 for details.

Initially, we planned to conduct a factorial design with two different interventions in a 1:1:1:1 ratio. The second intervention was a teaser on the cover (“Join now and help shape healthcare”) compared to a blank cover. Unfortunately, the printer only sent blank envelopes, so we didn’t investigate.


The result was the participation rate. Participation rate is defined as the number of people who answered at least one question in relation to the total number of people who received a survey invitation. Whereas the response rate is defined as the number of people who responded to the survey in some way in relation to the total number of recipients of the survey invitation. The response rate can always be higher than the participation rate, a refusal to participate would be a response, but not a participant.

Also Read :  Landmark trial begins over Arkansas' ban on trans youth care

sample size

All addressees of our survey on medical second opinions have been included in this SWAT. We chose a disproportionately stratified sample of equal size (cities, towns and suburbs, rural areas) because we hypothesized that there might be differences between settlement patterns regarding our initial study question (second opinions).


For the randomization, one researcher used random numbers generated by Microsoft Excel. Participants were originally allocated 1:1:1:1 with no stratification. In group 1: cover letter without a photo and without a teaser on the cover, group 2: cover letter with a photo without a teaser on the cover, group 3: cover letter without a photo with a teaser on the cover, group 4: cover letter with a photo with a teaser on the cover . The same researcher who determined the assignment sent the address data of the four groups to the printer. Once we found the envelopes failed to print with a teaser, we merged Group 1 and Group 3 into Group 1 and Group 2 and Group 4 into Group 2.


Participants were unaware that we conducted this SWAT as part of our survey. At the bottom of the questionnaire there was only a small mark “I”, “II”, “III”, “blank space “for groups 1, 2, 3, and 4 so the researchers could determine which group the participants were placed into. The researchers were not blinded to the data analysis.

Also Read :  New Study IDs Risk Factors for Acute Exacerbation in ANCA-ILD | More Inflammation, Poor Lung Function Tied to Worse Symptoms

data collection

All data obtained in the survey was extracted into an Excel spreadsheet. And the data from the entire sample (gender, zip code and age for most registration offices) received from the registration offices was included in a second Excel spreadsheet. We have categorized postal codes according to the “degree of urbanization” from Eurostat and the regional statistics of the Federal Statistical Office, which correspond to the settlement patterns previously described [17, 19]. In order to create an international comparability, we used the Comparative Analysis of Social Mobility in Industrial Nations (CASMIN) categorization for abstract education. CASMIN contains three main categories, primary, secondary and tertiary education with various sub-categories. The categories used were tertiary education (academic qualification, independent of school education and type of qualification), upper secondary level (no or no vocational training with at least intermediate school education) and lower secondary level (no or any vocational training with at least general school education). ) [20]. Due to the study design, we could not systematically collect the reasons for non-participation.

statistical methods

We calculated the risk difference and the odds ratio for the primary outcome using contingency tables. Differences between the characteristics of the participants in both groups were tested by performance χ2 Analyzes for binary data and t-test for continuous data. We performed exploratory subgroup analyzes in females, males, urban, urban-suburban, and rural individuals, and compared participants and non-participants.

In the subgroup analyzes we also use χ2 Analysis for binary data. We planned to use logistic regressions but were unable to do so because individual participant data were not fully assignable. All analyzes were performed in SPSS version 28.0 and Microsoft Excel.

Source link