The available literature found a significant impact of the pandemic on the mental health of the general population, both in terms of access and use of services and of the possible onset of psychopathological distress. However, this impact differed significantly based on the phase of the pandemic, the kind of restrictive measures implemented by different governments, the incidence of COVID-19 in different areas (with the correlating different numbers of deaths and of people mourning a relative’s death), and the characteristics of the population affected (age, socioeconomic status, being a migrant, having a previous mental disorder, etc.) (16).
It should be noted that most of the available studies are limited by their cross-sectional research design and the extensive use of internet-based snowball procedures, which select opportunistic samples whose risk could differ from the general population. Indeed, a recent review and meta-analysis of longitudinal cohort studies found a high degree of unexplained heterogeneity in the mental health reactions to the pandemic and an increase in mental health symptoms soon after the outbreak of the COVID-19 pandemic, that was smaller than expected (25). Moreover, the authors found that mental health symptoms decreased to pre-pandemic levels by mid-2020 among most population sub-groups and symptom types (25). Accordingly, it seems that evidence from cross-sectional studies may have overemphasized the impact of the pandemic on mental health, meaning that more longitudinal studies are needed. Indeed, longitudinal population studies are the most powerful tools to investigate and monitor the mental health of the population as well as socioeconomic inequalities. Based on this, we set up a multicenter longitudinal cohort study in Italy with this aim.
In this work, we present the study design and the sample characteristics of the cohort before and during the COVID-19 pandemic, taking into account the role of socioeconomic and/or citizenship inequalities. Baseline data show that the overall crude incidence rate of access to mental health care was 3.3% in the pre-COVID-19 period and 2.6% during the pandemic. Drug prescription was the main reason for being included in the study (57.2%). Compared to the general population, people with mental health conditions were older and more often female. In general, we found a similar distribution of the deprivation index. Interestingly, immigrants from HMPCs were younger, socioeconomically more deprived, and more likely to be included into the study due to an admission to an ED.
One of the strengths of our project is its longitudinal design, a powerful approach that allow us to evaluate the large amount of information regarding mental health conditions and the use of mental health services.
Moreover, information regarding SARS-CoV-2 infections, rate of hospital admissions in non-critical and critical areas, and deaths caused by COVID-19 will be taken into account.
Finally, the OENES, which promoted and is coordinating the study, is part of a network that includes the main national and regional epidemiologic centers in Italy. It may facilitate the extension of the cohort to other centers, thereby making the cohort even more powerful as well as making it possible to create a national monitoring system for the inequalities in mental health care.
The main limitation of our study is that a psychiatric diagnosis is available for hospital discharges, ED accesses, and users of day care and residential mental health care but not for the drug prescription database, through which most of cohort subjects were enrolled.
Moreover, our choice to consider the use of psychopharmacological drugs as an inclusion criterion should increase the sensitivity of the study, as patients with mental health problems treated by family doctors and private physicians will be included in the study if psychopharmacological treatment was started. However, this criterion may reduce specificity, as these drugs are used not only in psychiatry but also in neurology, internal medicine, and other medical specialties.
Another limitation is that, without individual socioeconomic information such as education level, we must use an area-based indicator, i.e., the census tract deprivation index. This limitation may introduce residual ecological bias into the analysis.
Finally, the use of citizenship to identify immigrant status is subjected to residual information bias. According to Italian legislation, individuals born in Italy to non-Italian citizens are considered foreigners until the age of 18 years, while individuals born abroad can obtain Italian citizenship if they are descendants of Italian ancestors. These two facts influence the selection of the immigrant population: while boys and girls born in Italy, living in Italy, and speaking Italian are included as immigrants, people born abroad who do not necessarily speak Italian, did not attend Italian schools, or have any familiarity with cultural habits and customs are included in the Italian population.
The future steps of the project will focus on the impact of the pandemic through the evaluation of accesses to mental health care globally and for hospital discharge, emergency, outpatient, residential and day care services, and drug prescriptions. We will also evaluate socioeconomic inequalities through the use of census-based deprivation index and migration status. Finally, we will also analyze the impact of COVID-19 infection and outcome on the study cohort.