The impact of the Covid-19 pandemic on the effectiveness of psychosomatic rehabilitation in Germany

Abstract

Background: The aim of the present study is to investigate the impact of the Covid-19 pandemic on the effectiveness of psychosomatic rehabilitation.

Methods: Between April 2019 and March 2022, a total of 18,388 patients from 7 rehabilitation centres could be included in the study. For each patient, score values from the HEALTH-49 and ICF AT-50 Psych questionnaires were calculated at the beginning and at the end of rehabilitation and the effectiveness of the rehabilitation program was determined by comparing the scores at the beginning and at the end of the rehabilitation programme. Using risk adjusted linear mixed models, three time intervals were compared: a pre-pandemic episode (April 2019 to March 2020), the first year of the pandemic (April 2020 to March 2021) and the second year of the pandemic (April 2021 to March 2022).

Results: Overall, it can be stated that the pandemic has substantially impaired the effectiveness of psychosomatic rehabilitation measures. This phenomenon can be observed across a wide range of psychosocial markers and even two years after the start of the pandemic there is no end to the limited effectiveness. With regard to 'psychological and somatoform disorders', for example, there was a relative decrease in the effectiveness of the rehabilitation measure by 11.29% in the first year of the pandemic compared to the pre-pandemic episode, p < 0.001. In the second year of the pandemic, the effectiveness of the rehabilitation measure was still decreased by 8.8% compared to the pre-pandemic episode, p < 0.001. In addition, the evaluations show that a division of the pandemic effect into direct effects (on the individual) and indirect effects (via further complication of the occupational problem environment) can be made and that the pandemic-related complication of the occupational problem environment are still prevalent more than two years after the start of the pandemic.

Discussion: The Covid-19 pandemic has had a significant impact on the psychosomatic rehabilitation programs reducing the effectiveness of treatment not only for a short period of time but constantly until March 2022.

Trial registration number: DRKS00029669; Date of registration: 02/08/2022.

Keywords: Covid-19 pandemic; HEALTH-49; ICF AT-50 Psych; PAREMO; Psychosomatic rehabilitation; SIMBO-C.

Fig. 1

Relative effect of the pandemic on the effectiveness of the rehabilitation Effectiveness of the rehabilitation is defined as difference between the questionnaire scores at the beginning and at the end of rehabilitation. The pre-pandemic episode (April 2019 to March 2020) acted as reference and relative effects of the first year of the pandemic (April 2020 to March 2021) and the second year of the pandemic (April 2021 to March 2022) are shown using mixed-effects generalized linear models with a Gaussian family and log link. Due to the log link, the resulting coefficients may be interpreted as semi-elasticities. A semi-elasticity represents the percentage change in the dependent variable after a 1-fold absolute change in the independent variable. Possible confounding factors such as age, gender, duration of rehabilitation and the respective questionnaire scores at the beginning of rehabilitation were included as fixed effects and different rehabilitation centres were included as a random intercept

Fig. 2

Absolute effect of the pandemic on the effectiveness of the rehabilitation Effectiveness of the rehabilitation is defined as difference between the questionnaire scores at the beginning and at the end of rehabilitation. For these vizualizations, the rigid categorisation into three episodes (pre-pandemic episode, first pandemic year and second pandemic year) was broken down and time (in quarters) was modelled as a non-linear variable. The non-linearity was implemented using restricted cubic splines with 5 knots and knot location based on Harrel’s recommended percentiles. Once again, the remaining specification of the multilevel mixed-effects generalised linear model with a Gaussian family and identity link was retained.