gms | German Medical Science

Background/Research question/Problem: Many studies are influenced by the Corona virus disease 2019 (COVID-19) pandemic, both in conduct and evaluation. This also applies to the project isPO (integrated cross-sectoral psycho-oncology) [1]. The basic project idea is to achieve success in combating anxiety and depression in cancer patients through psychological, psychotherapeutic and social therapeutic measures. To assess the anxiety and depression level the Hospital Anxiety and Depression Scale (HADS) is used in combination with a quasi-experimental study design, the regression discontinuity-design (RDD) while providing psychological care to cancer patients according to their individual needs. For this, patients are divided into different levels based on their personal HADS score and a predefined threshold as a quasi-randomization. The different levels serve to compare patients with scores under and above the threshold and examine the effectiveness of intensive psychological care in those with more need to those with less need. For this purpose, the HADS is collected and evaluated at two other points in time in addition to the time of allocation. Through the feedback from the hospitals, i.e. the practice, it is shown that due to the Covid-19 pandemic the stress situation for cancer patients has worsened and the potential for anxiety and depression increased. This gave rise to the question of how to consider, if necessary, the deterioration due to the pandemic in the evaluation and proof of efficacy in the already ongoing study?!

Solution and suggestion: This question led to the idea of including an additional categorical variable in the database that would consider at what point in the study a patient completed the HADS questionnaires and include this variable in the secondary analyses. For maximum information content, the categorical variable is created on a very small scale to detect any differences in the course of the pandemic due to wave-like spread and to allow for easy coarsening. The timepoints for a start of a new wave are pre-defined.

Conclusion/Discussion/Lessons learned: This variable is then to be used to stratify patients and to possible identify differences between anxiety and depression levels between patients before, before and within as well as completely within the pandemic and to explore possible differences during the chronological trend. The aim is to account for potential different variabilities over time as it is suspected that the structure of the patients´ needs changes with the ongoing of COVID-19 pandemic and the resulting psychological strains. Therefore, it is suspected that more patients need psychological care or needing more care, which leads to a potential bias in the structure of the drop-outs as well as the before-after-differences found within a patient in the different strata. While being fully aware that this approach will not solve all the issues risen as COVID-19 came up in the ongoing study, we try to account for the potential multifold biases in the above described manner.