Background: Artificial Intelligence (AI) and data science research are promising tools to better inform public policy and public health responses, promoting automation and affordability. During the COVID-19 pandemic, AI has been an aid to forecast outbreak spread globally. The overall aim of the study is to contribute to the ongoing public health, socioeconomic, and communication challenges caused by COVID-19. Protocol: COLEV is a five-pronged interdisciplinary mixed methods project based on AI and data science from an inclusive perspective of age and gender to develop, implement, and communicate useful evidence for COVID-19-related response and recovery in Colombia. The first objective is identification of stakeholders’ preferences, needs, and their use of AI and data science relative to other forms of evidence. The second objective will develop locally relevant mathematical models that will shed light on the possible impact, trajectories, geographical spread, and uncertainties of disease progression as well as risk assessment. The third objective focuses on estimating the effect of COVID-19 on other diseases, gender disparities and health system saturation. The fourth objective aims to analyze popular social networks to identify health-related trending interest and users that act as ‘super spreaders’ for information and misinformation. Finally, the fifth objective, aims at designing disruptive cross-media communication strategies to confront mis- and dis-information around COVID-19. To understand stakeholders’ perspectives, we will use semi-structured interviews and ethnographic work. Daily cases and deaths of COVID-19 reported from the National Surveillance System (INS) of Colombia will be used for quantitative analysis, and data regarding the online conversation will be obtained from Facebook and Twitter. Conclusions: COLEV intends to facilitate the dialogue between academia and health policymakers. The results of COLEV will inform on the responsible, safe and ethical use of AI and data science for decision-making in the context of sanitary emergencies in deeply unequal settings.

The COVID-19 pandemic has impacted the well-being of millions of people around the globe. During the COVID-19 pandemic, the mental health of the population was affected, which means that governments would need to implement different actions to mitigate and treat mental health disorders result of the pandemic.   This study aims to estimate the prevalence of anxiety and depression for female and male adolescents and adults in Colombia before the COVID-19 pandemic. It also aimed to estimate the potential increase of the prevalence in each group as a result of the COVID-19 pandemic in 2020. We used the Individual Registry of Health Services Delivery data from 2015 - 2021 to estimate the observed prevalence of anxiety and depression. Using the National Mental Health Survey 2015, we simulated the expected prevalence of anxiety and depression for adolescents (12 to 17 years) and adults (18 or older) from 2016 to 2020. We used an arithmetic static Monte Carlo simulation process to estimate the expected prevalence. The results of the analysis using revealed an important increase in the observed prevalence of these disorders for adults and adolescents and men and women between 2015 and February 2020. When we simulated different scenarios using the National Mental Health Survey and estimated the prevalence of both depression and anxiety for adults and adolescents, we found that the prevalence of depression and anxiety has had an important increase in the last five years for all groups and had an important increase during the 2020. This increase has been greater for women than for men, and for adolescents than adults.  Our results show the number of people who need potential attention from the health system in Colombia and highlight the importance to think about how to avoid and detect potential cases of anxiety and depression especially in female adolescents.

Mapping the municipal results allows the identification of mortality hotspots in Colombia as well as contiguous areas with low levels of mortality (see Figure 3 for all-cause mortality and Figures S1-S7 in this document for the specific causes of death). As the general level of mortality is the product of different cause-specific trends, the overall pattern conceals many interesting panoramas, including territorial contrasts in the specific causes that we analysed. These are succinctly described in the article.