Summary of results

The studies identified in this scoping review highlighted that the COVID-19 pandemic appeared to exacerbate some of our pre-specified treatment gaps, particularly during the first wave of COVID-19 across the Spring to Summer of 2020. Across the seven countries yielded in our search, restriction measures were enforced around approximately the same time, initially across March, then again across November to December in response to the second COVID-19 wave. All identified studies were conducted within a similar timeframe and either followed or intersected with the introduction of the initial restriction measures; however, the full nature of these restrictions did vary between countries. For instance, the United Kingdom and Italy differed in severity of at-home isolation orders, while the timing of lockdown policies between the United Kingdom and Germany differed [Reference Meintrup, Nowak-Machen and Borgmann39]. There was nevertheless a common “lockdown” period of restricted contact shared across all countries in this review. Organisation of care, too, varied between countries, with the European Psychiatric Association reporting a significant difference in mental health service delivery between countries during the first wave [Reference Rojnic Kuzman, Vahip, Fiorillo, Beezhold and Skugarevsky40]. These differences pertained to variations in hospital-based care for mental health, to the creation of specialist units, and to the guidelines introduced by professional associations; all these factors are likely to affect (to some degree) our findings, though a full exploration of these disparities is beyond the scope. As this scoping review did not focus on capturing the full background context of government measures and care structures in each region, the trends we discuss below must therefore be interpreted with some degree of caution.

Investigations from the United Kingdom, Denmark, Spain, Italy, and France reported a reduction in access or perceived access to secondary or inpatient care, reflected in either reduced clinical contact or patient-reported decreases in clinical access [Reference Kolbaek, Jefsen, Speed and Ostergaard26-Reference Olié, Dubois, Benramdane, Guillaume and Courtet31]. Studies of primary care services also indicated a reduction in general practitioner visits for depression in the Netherlands and Germany during the pandemic, compared to pre-pandemic rates [Reference Bohlken, Kostev, Riedel-Heller, Hoffmann and Michalowsky32, Reference Splinter, Velek, Ikram, Kieboom, Peeters and Bindels33]. A single study assessing rates of self-reported pharmacological treatment for MDD [Reference Amerio, Lugo, Stival, Fanucchi, Gorini and Pacifici34], found a substantial increase in reported depressive symptoms that was not commensurate with a more modest increase in antidepressant usage. This discrepancy could imply a reduction in the proportion of patients receiving antidepressants during the pandemic. Although this evidence is patient-reported and indirect, it is supported by a study in the United Kingdom, which highlighted that rates of increasing antidepressant prescriptions in 2020 were not significantly different from pre-pandemic trends [Reference Rabeea, Merchant, Khan, Kow and Hasan41]. These results are consistent with other reports that underscore how COVID-19 has decreased overall healthcare utilisation, a phenomenon that can be explained by the restrictions imposed by pandemic-induced lockdowns, fear of exposure to the virus, and inadequate availability of healthcare resources, amongst others [Reference Moynihan, Sanders, Michaleff, Scott, Clark and To15, Reference Nunez, Sreeganga and Ramaprasad42].

We were cognisant to enforce relatively loose inclusion criteria for the definition of MDD in this review, given the limited available data addressing our treatment gaps. Despite this, only nine pertinent studies were identified, and only a few of the pre-specified treatment gaps were addressed across these reports. While some trends were evident in this scoping review, these cannot be interpreted with certainty, as there is heterogeneity among the outcomes evaluated across studies. Several of the studies employed patient-reported data to draw conclusions about the effect of the pandemic, but this only acts as a proxy for actual changes in relevant outcomes, with the accuracy of self-reported data varying heavily [Reference Hafferty, Campbell, Navrady, Adams, MacIntyre and Lawrie43, Reference Short, Goetzel, Pei, Tabrizi, Ozminkowski and Gibson44]. Moreover, several investigations were conducted at a very small number of clinical sites, such that the absolute numbers assessed were often very small; it is thus difficult to aver the generalisability of these conclusions. Rates of undetected MDD could only be crudely assessed by evaluating the number of positive diagnoses during the pandemic, while no identified articles explored follow-up contacts for MDD and delays to the detection of MDD as outcomes.

It is important to note that there were a number of studies identified in our systematic search that did not meet our inclusion criteria, but that did have relevant insights to support some of the findings. Several articles focused on treatment gaps broadly across psychiatric care rather than specifically for MDD. For instance, Sampson and Wright [Reference Sampson, Wright, Dove and Mukadam45] found a decrease in referrals to psychiatric liaison services across North Central London during the first pandemic wave, while Pignon and Gourevitch [Reference Pignon, Gourevitch, Tebeka, Dubertret and Cardot46] reported reduced emergency psychiatric consultations during the first 4 weeks of lockdown in Paris and its suburbs. A cross-sectional study in Austria further reported a decrease in the percentage of participants that self-reported treatment for broad mental disorders during lockdown [Reference Simon, Helter, White, van der Boor and Laszewska47]. While not specific to MDD, these findings altogether corroborate the notion of reduced access to acute care, secondary care or rates of treatment in MDD management during the pandemic. Rates of broader psychiatric treatment were also peripherally assessed in several articles: in a survey of university students in France, Wathelet and Duhem [Reference Wathelet, Duhem, Vaiva, Baubet, Habran and Veerapa48] reported that 6.8% were seeing a professional for mental health reasons, far less than the 16.1% that were assessed with MDD based on the Beck Depression Inventory. Across three Dutch case-control cohorts, Pan and Kok [Reference Pan, Kok, Eikelenboom, Horsfall, Jorg and Luteijn49] reported that only 55% of participants with a lifetime mental health condition were undergoing treatment during the pandemic. While these investigations for the rate of treatment are further suggestive of impaired healthcare provision for psychiatric conditions and thus MDD, these statistics were not compared to pre-pandemic statistics, such that direct comparison of treatment rates for MDD was not possible.

There were also a large number of articles that did not explore our outcomes of interest but did directly assess changes in MDD prevalence during COVID-19, generally highlighting increased prevalence in depressive disorders following the pandemic [Reference Babicki, Kowalski, Bogudzinska and Mastalerz-Migas50-Reference Winkler, Formanek, Mlada, Kagstrom, Mohrova and Mohr53]. This relationship has been further underscored in several comprehensive systematic reviews, which suggest worsened levels of overall psychological distress during COVID-19 [Reference Xiong, Lipsitz, Nasri, Lui, Gill and Phan19, 54]. This was concomitant with a greater need for therapeutic support and heightened public concern about treatment impairment [Reference Favreau, Hillert, Osen, Gartner, Hunatschek and Riese55]. Given the COVID-19-mediated impairment of MDD care provision identified in this review, this is a particularly serious public health concern. Altogether, this evidence emphasises the need to ameliorate these worsened treatment gaps and make MDD care more resilient to external factors.

To this end, we also attempted to collate evidence of mitigation strategies introduced during the pandemic to improve MDD care. However, given our inclusion criteria, only two studies were included in our synthesis [Reference Parolin, Benzi, Fanti, Milesi, Cipresso and Ascolto25, Reference Scholl, Kohls, Gorges, Steinbrecher, Baldofski and Moessner36]; both these studies trialled digital interventions, one to replace face-to-face group therapy sessions and the other to link service users with mental health services, indicating that participants were receptive to the usage of these tools, though their efficacy in managing MDD is unclear. Once again, our search did identify a number of other studies exploring digital interventions, but that did not strictly assess an outcome of interest or was too broad in nature. For instance, Bauerle and Jahre [Reference Bauerle, Jahre, Teufel, Jansen, Musche and Schweda37] evaluated an e-mental health intervention integrating mindfulness and cognitive behavioural therapy, highlighting some therapeutic efficacy without addressing implications on care provision. Olwill and Mc Nally [Reference Olwill, Mc Nally and Douglas56] surveyed psychiatrists about their experiences with phone consultations, but this was for broader mental health concerns. While these articles did not meet our inclusion criteria, they nonetheless illustrate how the pandemic has expedited a shift into digital or remote solutions for mental health care more generally, a trend likely to continue in the post-pandemic era given the reported efficacy of such solutions even in non-pandemic contexts [Reference Moshe, Terhorst, Philippi, Domhardt, Cuijpers and Cristea57, Reference Torous, Bucci, Bell, Kessing, Faurholt-Jepsen and Whelan58]. Clearly, more work needs to be done to identify how these innovations have affected, and continue to affect, MDD care during COVID-19. Our search strategy failed to identify articles that offered detailed evaluations on how remote care and the use of teleconsultations during COVID-19 affected MDD service provision in particular, though there was evidence for increased telemedicine usage across mental health [Reference Abraham, Jithesh, Doraiswamy, Al-Khawaga, Mamtani and Cheema59, Reference Patel, Irving, Brinn, Broadbent, Shetty and Pritchard60]; to this end, it would be useful to have more detailed evaluations on how the general shift to teleconsultations across healthcare has affected specific MDD care pathways.

In addition to the aforementioned mitigation strategies, it is important to consider the extent to which restructuring of mental health care services addressed COVID-19-mediated gaps in psychiatric care provision. This re-organisation of care arose due to a need to divert people away from acute hospitals (to reduce contact and relieve hospital burden) while maintaining some level of support for individuals suffering from mental health conditions. While our search strategy did not yield MDD-specific reports on this topic, several reports highlight how community mental health services across the world continued to provide broad psychiatric care and support throughout the pandemic, despite the many challenges they faced [Reference Kopelovich, Monroe-DeVita, Buck, Brenner, Moser and Jarskog61-Reference Vigo, Patten, Pajer, Krausz, Taylor and Rush63]. Furthermore, it is clear that these community services adopted digital technologies and telemedicine to ensure care delivery in a safe and optimal manner [Reference Burton, Wall and Perkins64, Reference Sklar, Reeder, Carandang, Ehrhart and Aarons65], with relatively favourable perceptions of this digital shift, which may have long-term implications for the how community mental health resolves to bridge the gaps in MDD, and psychiatric, care provision.

Strengths and limitations

There were several strengths to this scoping review. This study is the first to broadly investigate the effects of COVID-19 on pre-existing treatment gaps, and more broadly care provision, specific to MDD in Europe. Moreover, we conducted this review in accordance with rigorous PRISMA-ScR guidelines and adopted a systematic approach to our search and screening methodology, to ensure we properly synthesised evidence on this topic. We further registered a protocol prior to the conduct of this study, adhering to it to achieve objectivity and transparency in our review. Our eligibility criteria were also meticulously developed in several consensus meetings to ensure the review was comprehensive while relevant to MDD care provision, and we permitted studies that were not written in English for better coverage, though this did not facilitate the inclusion of additional studies.

Limitations of this study include the lack of critical appraisal for study quality and the broad nature of the data synthesis, though these limitations are a staple in scoping reviews. The eligibility criteria adopted in this review are concomitant with several other limitations: our MDD criteria were relatively loose to maximise scope, and certain subgroups were not investigated at length (e.g., the older adult population). Despite some methodological leniency, there were only a small number of articles eligible for analysis following our rigorous screening, which limits the interpretation of our findings. There are limitations to the generalisability of our findings across continental Europe, given the small number of countries reported on in this review, primarily concentrated across Northern, Western, and Southern Europe. Certain treatment gaps were left completely unaddressed, meaning we do not have a complete picture of how COVID-19 has affected MDD care provision in Europe. Furthermore, there was much methodological heterogeneity across studies and therefore a lack of standardisation, limiting the types of analysis that could be conducted with the results. We also acknowledge the possibility of missed relevant articles because of delays in indexing and because this is a fast-moving field. A final important limitation of this scoping review is the difficulty in directly comparing the extent of treatment gaps and mitigation strategies during the pre- and peri-pandemic periods. For example, while Butler and Delvi [Reference Butler, Delvi, Mujic, Broad, Pauli and Pollak30] found a 40.4% decrease in number of people accessing secondary care from 2019 to 2020, we cannot estimate comparable rates for the continued or post-pandemic period, rendering it difficult to assess the complete role of COVID-19 in moderating this particular treatment gap.

Future directions

In the peri- and post-pandemic era, it will be important to collate and re-evaluate several of the outcomes from the studies identified in this scoping review. There is a need to analyse longer-term observational data from MDD care services and wider public samples, in order to determine the nature and extent of treatment gaps in MDD care, particularly during the latter stages of and gradual transition from the pandemic. A fully developed understanding of how the pandemic impacted MDD care provision in Europe would enable comparison with the implementation of recommended guidelines, and importantly enable us to capture the effects of both positive (for instance, via the implementation of guideline recommendations) and negative events (like the pandemic) on MDD care.

This scoping review also highlighted several additional studies that could be conducted to help develop a full perspective on the impacts of COVID-19 on MDD care in Europe. There is still much scope to investigate the digital mitigation strategies implemented for management of MDD during the pandemic. The use of digital strategies in MDD has been a topic of interest even prior to the pandemic, with a large variety of tools deployed with mixed efficacy. For instance, evidence for internet-based cognitive behavioural therapy is fairly mature [Reference Karyotaki, Efthimiou, Miguel, Bermpohl, Furukawa and Cuijpers66], though the efficacy of other innovations, such as virtual reality, chatbots, and smartphone applications, has yet to be fully validated [Reference Torous, Bucci, Bell, Kessing, Faurholt-Jepsen and Whelan58] and requires more study. Telepsychiatry too has grown extensively in popularity during the pandemic, though there is still scope for refinement of its implementation [Reference Greenhalgh and Wherton67]. Our search found that many such digital interventions have been trialled and implemented during the pandemic; however, many investigations to date did not meet our inclusion criteria due to a lack of specificity for MDD service provision. This lack of specificity was a common occurrence for articles that did not pass screening for our eligibility criteria; while understandable given the comorbidity of MDD with other mental health disorders, it will be important to conduct additional investigations specific to MDD care, as treatment pathways in this subgroup already contain complexities. During our initial search, there were also several pertinent articles for specific subgroups (such as student populations) that did not meet eligibility criteria – it would thus be useful to conduct a similar review surveying MDD care pathways in these subgroup populations. For example, an investigation into how COVID-19 affected MDD care provision in the elderly population would be interesting, as they were likely impacted in different manners, due to various subgroup-specific factors, including the presence of comorbidities and differing access or skills with digitally enabled solutions that were trialled in the pandemic. Several of our pre-specified treatment gaps – especially delays to MDD detection and rates of follow-up – were not fully addressed by the current literature, further reinforcing that additional retrospective analysis would be useful. While there were longitudinal investigations, none included sufficient follow-up to enable a complete evaluation of COVID-19-related trends.

Despite the uncertainty around many of the conclusions from this scoping review, stemming from a limited amount of available data, treatment gaps for MDD appeared to be exacerbated by COVID-19. It will therefore be important to re-evaluate previous recommendations for MDD care to ensure they are robust in the face of unpredictable events. Treatment gaps in MDD care clearly remain extensive and problematic, so it is more critical than ever to implement standardised solutions throughout the pathway.