In order to fulfill the purposes of the study, a quantitative and a qualitative study were conducted. An overview of the study is depicted in Fig. 1.
Quantitative study
In order to examine the effectiveness of the web-based aftercare an intervention study is conducted, comparing a web-based aftercare program with traditional aftercare program and a control condition.
Intervention
In this study, two rehabilitation aftercare programs are implemented: the traditional face-to-face aftercare program (IRENA; Intensified Rehabilitation Aftercare) and the web-based aftercare program (digIRENA; digitized Intensified Rehabilitation Aftercare). Both aftercare programs consist of 24 sessions and are tailored to the orthopaedic conditions of the rehabilitants [24]. The program has to be conducted by a qualified physiotherapist who selects the contents of the aftercare program. In the web-based aftercare program, the subjects receive a comparable individualized program with the same number of sessions as the traditional face-to-face IRENA. The web-based program is supervised by an external team of physiotherapists, employed by the provider of the web-based platform. Moreover, the web-based platform provides tools such as relaxation exercises or cooking recipes equivalent to the multimodal approach of a traditional face-to-face aftercare program. In both programs, if patients do not attend a session for at least six subsequent weeks, they are considered dropouts. If inactive, the intervention provider in the web-based aftercare program can encourage the patient to remain in the intervention. Patients in both aftercare groups are continuously supervised by qualified therapists during the interventions. In case of possible harm from the intervention, patients can be referred to appropriate treatment. The participants in the control group receive neither IRENA nor digIRENA. However, they have the possibility to organize their own rehabilitation activities. Finally, all patients can receive additional therapies depending on their health status.
Sampling and participants
Patients with orthopedic complaints who have applied for and been approved for rehabilitation on the prescription of a primary care physician are recruited from five rehabilitation centers throughout Germany. The inclusion criteria are that patients are at least 18 years old, have a basic knowledge of the German language, and their primary diagnosis concerns an orthopaedic problem. Furthermore, only patients who are insured with the public pension insurance Knappschaft Bahn See can participate in the study, as the included rehabilitation centres belong to Knappschaft Bahn See. When patients are informed about the possibility to take part in an aftercare program during their stay, they are also asked to participate in the study. In case they give their consent, they are assigned to three groups: (1) traditional face-to-face aftercare program (IRENA), (2) web-based aftercare program (digIRENA), or a control group. For ethical reasons, a partial randomized controlled trial was used. All patients are first asked whether they want to take part in the traditional aftercare program (IRENA). Only if they deny, they are randomly assigned to the web-based aftercare program or the control group. This means that the IRENA condition was offered preferentially and the randomization referred only to the digIRENA and control condition. Sealed envelops are used for the randomization and patients only know their group after opening the envelop. The envelops are distributed with the goal to obtain a ratio of 1:1 between the two groups. Due to nature of the study neither participants nor staff can be blinded to allocation. As an incentive to take part in the study, all patients who complete the study with all measurement occasions receive a 50 Euro Amazon voucher.
Because only a small number of studies exists in relation to web-based aftercare interventions, we chose to take a conservative approach in the power analysis with a small estimated effect size [8]. Conservatively estimating the effect size by Cohen’s f = .15, α = .05, 1-β = .80, with three groups, a correlation among repeated measures of r = .30 and a design with three repeated measurement occasions, on average, using an ANOVA with repeated measures (within-between interaction), the calculated total sample size is 573 participants, nearly equally distributed over three groups. Since we expect a dropout rate of about 50% [25], the plan is to recruit 1150 patients.
Data collection of quantitative study
The data collection started in January 2020. Due to the pandemic, the rehabilitation centers did not receive any patients from mid-March until about mid-May. To increase the number of recruited patients, two more rehabilitation centers joined the project in August 2020 (initially there were only three centers). The recruitment of the patients is planned to finish in October 2021. In total, the data collection includes four measurement occasions. The baseline measurement takes place at the rehabilitation center before the start of the intervention, the first follow-up measurement 13 weeks after the baseline, the second follow-up 26 weeks after the baseline, and the third follow-up 43 weeks after the baseline (see Fig. 2). The first three measurement occasions represent the development of the outcome variables, approximately, before, during and at the end of the intervention. Due to the limited duration of the study, the last follow-up measurement is intended to illustrate the sustainability of developments over the next 4 months. For the measurements 2–4, the questionnaires are directly sent to the patients, in accordance to their preference via post or e-mail. A reminder is sent in case the patients have not returned the questionnaires after 2 weeks.
Measures
The attendance, work ability, subjective health, and motivation will be measured to compare the development of the three groups (i.e., web-based aftercare, traditional face-to-face aftercare, control group). While the work ability, subjective health, and motivation are assessed at every measurement occasion, the attendance is not assessed at the first measurement occasion. Furthermore, attendance and motivation are only assessed in the web-based aftercare and the traditional face-to-face aftercare groups.
Attendance
This part of the questionnaire was developed for this study. Patients in the traditional face-to-face aftercare group and the web-based aftercare group are asked (a) how often they participated in their program and (b) how many minutes within a week on average they trained with their program. Furthermore, patients of all three groups will be asked about their average weekly participation in other sport activities.
Work ability
The German version of the Work Ability Index (WAI) as an internationally established instrument is used to measure the work ability of the patients [26]. The WAI consists of ten questions with categorical and continuous response formats, resulting in a score from a minimum of 7 to a maximum of 49. This score indicates the current assessment of one’s own work ability to deal with the work demands. The WAI has been shown to have an acceptable Cronbach᾽s α = 0.78 [27] and to predict an early work termination and long-term work disability [18, 19].
Subjective health
The German version of the SF-12 is used to measure the subjective health [28]. The SF-12 consists of 12 items with dichotomous and continuous response formats, which comprise the two dimensions physical and mental health. Studies have shown good Cronbach᾽s α = 0.83 for physical health and α = 0.87 for mental health [28]. Moreover, the SF-12 can be used independent of the current health status, and the questions have been deemed as understandable and relevant [29].
Motivation
The Behavioural Regulation Exercise Questionnaire-2 [BREQ-2 [30]], translated into German, is used to measure motivation. The questionnaire consists of 19 items with a five-point Likert scale, measuring five dimensions of motivation: intrinsic motivation, identified regulation, introjected regulation, extrinsic motivation, and amotivation. The reliability has been shown to range from acceptable to good for the different dimensions: intrinsic motivation (α = 0.88), identified regulation (α = 0.83), introjected regulation (α = 0.77), extrinsic motivation (α = 0.77), and amotivation [α = 0.60; 30]. In addition, the subscales show inter-correlations in line with theoretical considerations, and more autonomous types of motivation are associated with a higher probability of health behaviours [31].
Data analysis of quantitative data
Data are entered by research assistants and checked for accuracy by an independent research assistant. All personal information is pseudo-anonymized. Missing data are treated using either the multiple imputation or the full-information maximum likelihood approach. A first analysis will include all patients who intended to participate in the interventions and a second study will include only those patients who actually completed the interventions. A multivariate analysis of variance with repeated measurements will be used to analyze the data. Additionally, the development of work ability and subjective health will be analyzed by latent growth curve models [32]. These models allow the analysis of linear and nonlinear developments over time. A mediation analysis will be conducted to test whether motivation mediates the effects of the intervention on the attendance rate. The effects of age and gender will be controlled.
Qualitative study
Parallel to the quantitative study, a qualitative study using interviews will be conducted to examine the applicability, acceptance and usability of the web-based aftercare program.
Data collection of qualitative study
Semi-structured interviews will be conducted with 13 therapists and 15 patients using the interview guidelines developed in accordance with the criteria of qualitative research (Tong, Sainsbury, & Craig, 2007). The content of the interview guidelines for both patients and therapists target applicability, acceptance, and usability of the web-based aftercare program. Particularly, the questions are formulated to assess the general experience with the program, its evaluations, and suggestions for improvement. It is intended to include patients who are still actively participating in the web-based aftercare program as well as patients who had dropped out of this program. The interviews will be conducted via Skype or over the phone due to the current Covid-19 pandemic.
Data analysis of qualitative data
For data transcription and data analysis, the use of the software MaxQDA is planned, which is explicitly designed to support transcriptions and analyses of qualitative data.
In order to specify the questions about the applicability, acceptance and usability of the web-based aftercare program, the data analysis is based on the Qualitative Content Analysis [33]. The categories will be developed as a combination of inductive and deductive category formation. As a result, strategies for improving the applicability, acceptability and usability of the web-based aftercare program will be developed. To ensure the rigour of the qualitative study, all steps are carried out by two independent researchers, implying an investigator triangulation [34]. In addition, there will be meetings with the other researchers of the team who will act as critical friends helping to reflect on alternative interpretations of the data [35].