This study in a large nation-wide claims dataset showed significant reductions in utilization of mental health services in Germany during the first two lockdown phases of the COVID-19 pandemic, foremost in the inpatient and day clinic care sectors and in outpatient incident diagnoses. Prescriptions of psychopharmacological medications and outpatient psychotherapy provision remained stable. In the period between the two lockdown phases, no complete recovery of utilization was observed.

Mental health inpatient care admissions were significantly reduced, in line with results from other regions of Europe and reports from local German provider networks [1, 5, 10,11,12, 20]. However, specific diagnostic groups were affected to different degrees: The first lockdown phase showed stistically significant reductions for substance use and anxiety disorders; the second lockdown phase for all diagnostic groups but they were more pronounced for e.g. substance-use, anxiety, obsessive-compulsive and stress-associated disorders or affective disorders but to a lesser degree psychotic disorders. Reduced indicators of standard care but unchanged indicators of psychiatric intensive care are in line with reports of relative increases in more acute cases during the lockdown phases [5, 10, 12]. The day clinic care setting exhibited not only significantly reduced admission numbers in both LDPs but additionally significantly reduced length of stay in the first lockdown phase. Thus, this less life-disrupting and more integrative treatment mode was even more impacted by service reductions.

The reasons for these significant reductions cannot be found out by routine data studies alone. Results of surveys of psychiatric inpatient institutions and office-based outpatient psychiatrists suggested that reduced demand by patients and a lack of staff played minor roles. Mainly, institutions themselves seemed to have changed their admission policies for creating capacities for comorbid, severely mentally ill and infectious patients, and for hygiene and social distancing measures. In the second LDP premature discharges of patients with SARS-CoV-2 infections might have played a role, too. Additionally, financial incentives that were thought to provide capacities for SARS-CoV-2-patients might have been an important factor [5, 21]. Unfortunately, studies of user perspectives are lacking, but the existing results hint at a reduced healthcare provision being a significant factor. This raises questions, if and to what degree these patients were treated by the outpatient system instead or if the reductions resulted in an underprovision of services for people with mental disorders or – as one might argue because of Germanys unique large share of inpatient mental health care – in a normalization of a previous overprovision.

Outpatient care showed no significant change in the number of patients in psychotherapy and prescriptions of psychotropic drugs. Patients “stocking” prescriptions in order to be prepared for access problems can explain probably slight increases in prescriptions immediately before and small dips during the lockdowns. However, the data showed indications neither for supply gaps nor for an overall increased demand during the first year of the pandemic. In the United Kingdom drops in prescriptions for antidepressants were reported for the first lockdown [22], while psychotherapy services were rarely examined on a national level within Europe. Psychotherapy service provision in one British regional network did not change significantly during 2020 [23].

Significant reductions within the outpatient system were found in the number of incident diagnoses. In line with these results, a survey of outpatient psychiatrists reported maintenance (partly by telemedicine) of offerings for known but reductions for new patients due to capacity restraints by e.g. social distancing measures [21]. Additionally, some outpatient psychiatrists reported a slightly reduced demand in the first lockdown phase, but increased demand in later phases of the pandemic due to catch-up effects, social isolation and economic hardships [21]. However, (financial) caps limit the outpatient system’s ability to provide large capacity increases. Taken together, patients impacted by reductions in inpatient and day clinic services were probably not to a large degree absorbed by outpatient system offerings. However, as no individual patient’s treatment sequences were examined, the currently available evidence cannot definitely answer this question.

Due to this lack of studies of individual patient’s treatment sequences and outcomes it cannot be answered neither, if the observed reductions in mental health service utilization let to negative consequences. Indirect evidence is inconclusive: Surveyed psychiatric inpatient departments and outpatient psychiatrists reported exacerbations, contact breakdowns, a lack of integration into the patients’ living environment, and suicide attempts and saw them – without proven causality – linked to reduced inpatient capacities and insufficient outpatient treatment alternatives [5, 21]. Pandemic-related reductions in maintenance electroconvulsive therapies resulted in exacerbations [24]. However, no general increase in suicide mortality was reported for neither Germany nor Europe for the first year of the pandemic [25, 26], but longterm results are lacking, some departments reported increases in suicidality for certain disease groups [27] and no comprehensive statistics on suicide attempts exist.

This study utilizes a large set of claims data covering a total of 8.8 million insured individuals across Germany, however, with a bias towards the region of Saxony. Nonetheless, since the majority of the results were also confirmed in the subset of the BKK data that is representative for Germany [40] (see Supplement III), the results can be generalized to the whole of Germany. Claims data can offer complete and unbiased information on health care utilization and provision [14, 28]. However, it is restricted to broad indicators only and does not allow to distinguish between changes in utilization (in a narrow sense) and provision. The data itself allows no inference about the causes of changes in health care utilization. Additionally, while claims data offers much information on the provider level, it lacks information on the user perspective. Finally, our study covers only the first year of the pandemic and follow-up studies on later changes and possible catch-up effects would be of great interest.

During the first year of the pandemic significant reductions in mental health care service utilization took place, probably mostly as a consequence of (in the light of the imminent threads of COVID-19 well-intentioned) changes in political guidance, necessary hygiene measures, and financial incentives. Some studies suggest important negative consequences of these reductions, but due to a lack of a systematic monitoring or studies of trans-sectoral treatment sequences and routine treatment outcomes, no final conclusion can be drawn.

When during the crisis of the pandemic rapid decisions like e.g. changes in financial incentives and hygienic isolation measures had to be made without an evidence base it was impossible to guide these decisions by relevant up-to-date data and to monitor their effects transparently and in a timely manner – neither in Germany nor in many other regions of Europe. Infrastructures for monitoring and oversight are among the central recommendations for resilient health care systems by the WHO and OECD [7, 8, 29]. We therefore propose that a transparent public mental health system surveillance is needed, including indicators of trans-sectoral treatment-sequences and routine treatment outcomes. Indicators like those in this study could be the nucleus for such a comprehensive surveillance that could contribute to a better crisis preparedness and a more resilient mental health care systems in Europe [8].