Monitoring COVID-19 related changes in population mental health

Abstract Background The COVID-19 pandemic, and its consequences in terms of control measures and restrictions to normal life, has affected the population mental health. One of the four case studies from the Population Health Information Research Infrastructure (PHIRI) for COVID-19 is focused on mental health with the objective to measure changes in incidence of mental health problems associated with the COVID-19 pandemic in several European countries. Methods Using electronic health records (EHR), data on new episodes of depression or anxiety, prescription of antidepressants and anxiolytics, and visits to primary care, specialized care or emergency units with an episode of depression/anxiety, were collected by participant data hubs at national/regional level for the period 2017-2021. A common data model to collect the data was defined for all participating data hubs and analysis of status prior and during the COVID-19 pandemic was performed using R. Results Data hubs from Austria, Finland, Spain (Aragon), and United Kingdom (Wales) were able to provide aggregated results from raw individual-level data. Preliminary analysis of trends suggests a decrease in new cases of depression and anxiety in the pandemic period (2020-2021) in comparison with previous years. Different trends were observed between data hubs regarding prescription of drugs and the number of primary/specialized care visits due to depression or anxiety. Issues in the access to data in some of the participating data hubs were observed, related to ethical and legal matters, and the lack of centralized registers and of private consultations statistics. Conclusions The results of this use case show that EHR for the secondary use can be retrieved in a common way across Europe to analyse and compare the impact of COVID-19 in population mental health in European countries. However, the process is more complicated and time consuming than expected.


Background:
Healthcare systems across Europe reorganized services to provide attention to COVID-19 patients. In the event of the surge of cases, countries were forced to cancel or postpone non-urgent care. The objective of this work is to investigate whether there were time-to-treatment delays in breast cancer due to April-May 2020 restrictions, and whether the delays were permanent and different across countries. Methods: Design: Quasi-experimental pre-post study with a historical control. Population: Virtually the universe of breast cancer patients receiving elective surgery, radiotherapy, hormonal therapy or chemotherapy since January 2017 (until December 2021) in the participant regions -Belgium, Marché (IT), Riga (LV), Portugal, Wales, and Aragon (ES). The main endpoint is the change in the median time-to-treatment before and after an empirical joint-point. The study variables are detailed here https://doi.org/10.5281/zenodo.5148022. Analysis: Distributed generalized additive models using https://cran.r-project.org/ package = mgcv.

Results:
Preliminary results show that the impact in March-April 2020 time-to-treatment evolved differently across countries. For instance, while the median time from diagnosis to surgery, as the first treatment, increased from approximately 39 days (2018-2019) to more than 45 days (2020)(2021) in Wales, in the Marche region (IT) the median time decreased from 52 days in 2017-2019 to 47 days in 2020. Complete analyses for the rest of the participant countries are currently undergoing.

Conclusions:
We have observed differences in time to treatment in women with breast cancer across countries; however, the magnitude and direction of the effect has been uneven across countries.

Background:
The COVID-19 pandemic and lockdowns may adversely affect pregnancy outcomes due to disrupted healthcare provision and increased stress, anxiety and economic hardship. We assessed changes in perinatal outcomes in 2020 using population birth data in Europe.

Methods:
25 Countries in the Euro-Peristat Network implemented a federated analysis using routine national data. Countries generated anonymised aggregate data files using R scripts from individual-level data formatted to a common data model with 22 variables. We compared preterm birth, stillbirth, neonatal death and caesarean delivery rates in 2020 to 2015-2019 for 2 periods: full-year (FY) and pandemic (March-September [MS]). Data from October onward were not included in the MS period because potentially declining pandemic-related fertility may affect perinatal indicators. Country-specific relative risks (RR) for the periods, adjusted for linear trends, overall and by socio-economic (SES) group, were calculated and pooled using random effects metaanalysis.

Conclusions:
In 2020, there was an unexpected decline in preterm birth in some countries, while increases in stillbirths and caesarean occurred in others. High country-level heterogeneity suggests that some government policies to mitigate the pandemic might have been more protective of pregnant women and newborns than others.

Background:
The COVID-19 pandemic, and its consequences in terms of control measures and restrictions to normal life, has affected the population mental health. One of the four case studies from the Population Health Information Research Infrastructure (PHIRI) for COVID-19 is focused on mental health with the objective to measure changes in incidence of mental health problems associated with the COVID-19 pandemic in several European countries.

Methods:
Using electronic health records (EHR), data on new episodes of depression or anxiety, prescription of antidepressants and anxiolytics, and visits to primary care, specialized care or emergency units with an episode of depression/anxiety, were collected by participant data hubs at national/regional level for the period 2017-2021. A common data model to collect the data was defined for all participating data hubs and analysis of status prior and during the COVID-19 pandemic was performed using R.

Results:
Data hubs from Austria, Finland, Spain (Aragon), and United Kingdom (Wales) were able to provide aggregated results from raw individual-level data. Preliminary analysis of trends suggests a decrease in new cases of depression and anxiety in the pandemic period (2020)(2021) in comparison with previous years. Different trends were observed between data hubs regarding prescription of drugs and the number of primary/ specialized care visits due to depression or anxiety. Issues in the access to data in some of the participating data hubs were observed, related to ethical and legal matters, and the lack of centralized registers and of private consultations statistics.

Conclusions:
The results of this use case show that EHR for the secondary use can be retrieved in a common way across Europe to analyse and compare the impact of COVID-19 in population mental health in European countries. However, the process is more complicated and time consuming than expected.

Background:
Low birth weight (LBW) and preterm birth are associated with an increased risk of neonatal death and chronic conditions across the life course. Reducing LBW is a global public health priority and requires strategies to improve healthcare during pregnancy. We aimed to assess the effect of a health policy providing full coverage of illness-related costs from 13 weeks of gestation through 8 weeks postpartum on birth outcomes and neonatal mortality in Switzerland.

Methods:
We applied a regression discontinuity design to administrative data gathered as part of a Swiss research program (NCCR on the Move). We included all children (N = 166,709) born between March 1, 2013 and February 28, 2015. The outcomes were birth weight (BW), gestational age (GA), LBW (<2,500 g) and very low birth weight (VLBW; <1,500 g), preterm (<37 weeks of gestation), and extremely preterm (<28 weeks), and neonatal ( 28 days) death. Children were exposed to the policy if they were born from March 1, 2014 onwards. We estimated the intention-to-treat effect of the policy using parametric regression models. Results: Children had a mean BW of 3,291 g and mean GA of 275 days. The prevalence of LBW was 6.4%, VLBW 1%, preterm 7.2%, and extremely preterm 0.4%, respectively. Some 0.3% newborn died within one month. The policy increased BW (mean difference = 13 g [95% confidence interval (CI): 1, 25]) and decreased the risk of LBW (odds ratio [OR] = 0.89; 95% CI: 0.82, 0.98) and VLBW (OR = 0.81; 95% CI: 0.64, 1.01). Additionally, the policy slightly decreased the risk of preterm birth (OR = 0.94; 95% CI: 0.87, 1.03), while it did not affect GA. Effect estimates for extremely preterm and neonatal mortality were imprecise and inconclusive.

Conclusions:
This quasi-experimental and population based-study of 166,709 live births between 2013 and 2015 in Switzerland provides evidence of a reduction in the risk of LBW, VLBW and preterm birth thanks to a health policy that fully covered healthcare services during maternity. Key messages: Free access to healthcare during pregnancy may mitigate adverse newborn health outcomes. A Swiss health policy that fully covered healthcare services during pregnancy reduced the risk of low birth weight and preterm births.