Work-Related Medical Rehabilitation for Cancer Patients Receiving Non-postacute Rehabilitation

A post-hoc Subgroup Analysis

More than one third of all cancers affect persons of working age. In the first two years after the diagnosis, nine in every 10 of those affected successfully return to work (1), but persons with cancer are at higher risk of unemployment and occupational restrictions than healthy control groups (2).

Multidisciplinary interventions with a clear occupational focus can support the return after cancer disease (3).

Our cluster randomized multicenter study compared the effects of work-related medical rehabilitation with those of medical rehabilitation in cancer (German Clinical Trials Register: DRKS00007770; ethics approval granted by the University of Lübeck No 14–289) (4). In both groups, improvements in physical and mental functioning were observed after 12 months. Between-group comparison showed short-term and medium-term benefits of work-related medical rehabilitation (4), but no effectiveness after 12 months in spite of well and elaborately implemented programs. In our explorative analysis we looked for indications of effectiveness in specific subgroups of the sample. Because of the immediate consequences of treatment and illness, such a complex program directly after concluding primary treatment (for example, as postacute rehabilitation) is probably not able to achieve the desired effect to a satisfactory degree. We report our re-analysis of the data of the cluster randomized study (4) while considering a partial sample that included only persons receiving non-postacute rehabilitation and evaluated the differences one year after the end of the rehabilitation for these persons.

Methods

Cancer patients of working age were recruited in three rehabilitation clinics and, in groups with the same date of admission, were randomly allocated to the intervention group (work-related medical rehabilitation) or the control group (medical rehabilitation) (4). For the post-hoc analysis we evaluated exclusively the data of study participants who had applied for non-postacute rehabilitation—that is, a rehabilitation program that can be repeated if functional impairments persist. Participants receiving postacute rehabilitation were excluded.

The primary outcome was self-assessed role functioning (0–100 points) one year after the end of the rehabilitation measure, which was captured using the questionnaire of health related quality of life of the European Organization for Research and Treatment of Cancer (EORTC QLQ-C30) (4, 5). The role functioning scale (two items) captures the extent to which those affected feel restricted in their jobs and leisure time activities. Secondary endpoints were further scales of the EORTC QLQ-C30 relating to physical and emotional functioning, the fatigue module (EORTC QLQ-FA13 (0 to 100 points), and gainful employment one year after the end of the rehabilitation. Treatment effects were tested in mixed linear and logistic multilevel models. In the model estimates, the baseline values and the clinic were used as control variables (4). Absolute differences, standardized mean differences (SMDs; small effect: ≥ 0.2; moderate effect: ≥ 0.5; large effect: ≥ 0.8), and odds ratios were reported as effect estimators.

Results

For the cluster randomized trial, a total of 484 persons were recruited and randomly allocated in 165 groups to the intervention or control groups (intervention group: 80 groups). For three clinics (n=297; 61.4%) data were available as regards the type of rehabilitation (non-postacute rehabilitation: n=101; postacute rehabilitation: n=196). Of the 101 participants receiving non-postacute rehabilitation, complete data were evaluated for 73 (72.3%) persons who participated in the follow-up survey. No relevant differences were seen between groups (intervention group: n=39; control group: n=34) in terms of the sample characteristics (Table 1).

Table 1

Sample parameters at the start of rehabilitation

Differences with small to moderate effects sizes were observed in favor of the intervention group for role functioning, general health, physical functioning, physical and emotional fatigue, and problems in everyday life (Table 2). One year after the end of the rehabilitation program, 58.7% (95% confidence interval [44.0; 73.4]) in the intervention group and 49.3% [33.7; 64.8] in the control group were in gainful employment or fit for work (OR=1.71; [0.49; 6.01]). Results were similar in the intention to treat analysis, with missing values substituted by means of the last observation carried forward method (role functioning: difference 11.6 [1.7; 21.5]; SMD=0.37; n=101).

Table 2

Outcomes at the start of rehabilitation and one year after the end of rehabilitation

Discussion

One year after the end of the rehabilitation, clinically relevant differences with small to moderate effect sizes were observed in favor of work-related medical rehabilitation in persons receiving non-postacute rehabilitation. Compared with the publication of the complete study (4) the differences are primarily due to the fact that the control group did not experience any relevant improvement. A limitation lies in the fact that the subgroup analysis evaluated data originally collected for the cluster randomized multicenter study and the comparison of the groups is no longer protected by randomization. This has important implications for future research projects into work-related medical rehabilitation programs. Firstly, future studies should differentiate by the type of rehabilitation applied for, as especially persons receiving non-postacute rehabilitation obviously require additional support and benefit from an occupationally focused approach. Secondly, the model of repeated treatment after postacute rehabilitation might represent an option for implementing work-related medical rehabilitation programs for other disorders too (for example, myocardial infarction).

David Fauser, Julian Wienert, Jan Schmielau, Matthias Bethge

Universität zu Lübeck, Institut für Sozialmedizin und Epidemiologie, Lübeck (Fauser, Bethge), [e-mail]

IU Internationale Hochschule, Bad Reichenhall (Wienert)

AMEOS Reha Klinikum Ratzeburg (Schmielau)

Conflict of interest statement
The study was funded by the Federal German Pension Insurance (Deutsche Rentenversicherung [DRV] Bund). The University of Lübeck also received funding for studies with similar subjects from the German Pension Insurance North and from the Federal Ministry of Labour and Social Affairs.

MB is a board member of the German Society of Rehabilitation Sciences.

JS is a senior physician in one of the participating rehabilitation centers and was deputy chair of the Working Group “Oncology and Rehabilitation” in the German Association of Hematology and Oncology until 2020.

The remaining authors declare that no conflict of interest exists.

Manuscript received on 6 June 2023, revised version accepted on 20 September 2023.

Translated from the original German by Birte Twisselmann, PhD.

Cite this as:
Fauser D, Wienert J, Schmielau J, Bethge M: Work-related medical rehabilitation for cancer patients receiving non-postacute rehabilitation—a post-hoc subgroup analysis. Dtsch Arztebl Int 2023; 120: 841–2. DOI: 10.3238/arztebl.m2023.0217

References

1. Mehnert A: Employment and work-related issues in cancer survivors. Crit Rev Oncol Hematol 2011; 77: 109–30

2. de Boer AGEM, Taskila T, Ojajärvi A, van Dijk FJH, Verbeek JH: Cancer survivors and unemployment: a meta-analysis and meta-regression. JAMA 2009; 301: 753–62

3. de Boer AGEM, Taskila T, Tamminga SJ, Feuerstein M, Frings-Dresen MHW, Verbeek JH: Interventions to enhance return-to-work for cancer patients. Cochrane Database Syst Rev 2015; 9: CD007569

4. Fauser D, Wienert J, Zomorodbakhsch B, et al.: Work-related medical rehabilitation in cancer-a cluster-randomized multicenter study. Dtsch Arztebl Int 2019; 116: 592–9

5. Aaronson NK, Ahmedzai S, Bergman B, et al.: The European Organization for Research and Treatment of Cancer QLQ-C30: a quality-of-life instrument for use in international clinical trials in oncology. J Natl Cancer Inst 1993; 85: 365–76