How do Covid-19 policy options depend on end-of-year holiday contacts in Mexico City Metropolitan Area? A Modeling Study

Abstract

Background: With more than 20 million residents, Mexico City Metropolitan Area (MCMA) has the largest number of Covid-19 cases in Mexico and is at risk of exceeding its hospital capacity in late December 2020.

Methods: We used SC-COSMO, a dynamic compartmental Covid-19 model, to evaluate scenarios considering combinations of increased contacts during the holiday season, intensification of social distancing, and school reopening. Model parameters were derived from primary data from MCMA, published literature, and calibrated to time-series of incident confirmed cases, deaths, and hospital occupancy. Outcomes included projected confirmed cases and deaths, hospital demand, and magnitude of hospital capacity exceedance.

Findings: Following high levels of holiday contacts even with no in-person schooling, we predict that MCMA will have 1·0 million (95% prediction interval 0·5 - 1·7) additional Covid-19 cases between December 7, 2020 and March 7, 2021 and that hospitalizations will peak at 35,000 (14,700 - 67,500) on January 27, 2021, with a >99% chance of exceeding Covid-19-specific capacity (9,667 beds). If holiday contacts can be controlled, MCMA can reopen in-person schools provided social distancing is increased with 0·5 million (0·2 - 1·0) additional cases and hospitalizations peaking at 14,900 (5,600 - 32,000) on January 23, 2021 (77% chance of exceedance).

Interpretation: MCMA must substantially increase Covid-19 hospital capacity under all scenarios considered. MCMA's ability to reopen schools in mid-January 2021 depends on sustaining social distancing and that contacts during the end-of-year holiday were well controlled.

Funding: Society for Medical Decision Making, Gordon and Betty Moore Foundation, and Wadhwani Institute for Artificial Intelligence Foundation.

Research in context: Evidence before this study: As of mid-December 2020, Mexico has the twelfth highest incidence of confirmed cases of Covid-19 worldwide and its epidemic is currently growing. Mexico's case fatality ratio (CFR) - 9·1% - is the second highest in the world. With more than 20 million residents, Mexico City Metropolitan Area (MCMA) has the highest number and incidence rate of Covid-19 confirmed cases in Mexico and a CFR of 8·1%. MCMA is nearing its current hospital capacity even as it faces the prospect of increased social contacts during the 2020 end-of-year holidays. There is limited Mexico-specific evidence available on epidemic, such as parameters governing time-dependent mortality, hospitalization and transmission. Literature searches required supplementation through primary data analysis and model calibration to support the first realistic model-based Covid-19 policy evaluation for Mexico, which makes this analysis relevant and timely.Added value of this study: Study strengths include the use of detailed primary data provided by MCMA; the Bayesian model calibration to enable evaluation of projections and their uncertainty; and consideration of both epidemic and health system outcomes. The model projects that failure to limit social contacts during the end-of-year holidays will substantially accelerate MCMA's epidemic (1·0 million (95% prediction interval 0·5 - 1·7) additional cases by early March 2021). Hospitalization demand could reach 35,000 (14,700 - 67,500), with a >99% chance of exceeding current capacity (9,667 beds). Controlling social contacts during the holidays could enable MCMA to reopen in-person schooling without greatly exacerbating the epidemic provided social distancing in both schools and the community were maintained. Under all scenarios and policies, current hospital capacity appears insufficient, highlighting the need for rapid capacity expansion.Implications of all the available evidence: MCMA officials should prioritize rapid hospital capacity expansion. MCMA's ability to reopen schools in mid-January 2021 depends on sustaining social distancing and that contacts during the end-of-year holiday were well controlled.

Figure 1.

Observed (red dots) and model-predicted (green lines) Covid-19 incident detected cases (A), deaths (B), cumulative cases (C), and deaths (D) in MCMA between February 24, 2020 and March 7, 2021. Left column plots assume compliance with social distancing during end-of-year holiday period. Right column plots assume substantially less compliance with social distancing during the end-of-year holiday period. Double-dashed vertical line represents the last day used for calibration. The green shaded area shows the 95% posterior model-predictive interval of the outcomes and the green lines show the posterior model-predicted mean based on 1,000 simulations using samples from posterior distribution.

Figure 2.

Observed (red area) and model-predicted (green lines) total hospital occupancy and demand in MCMA between February 24, 2020 and March 7, 2021. Left plot assumes compliance with social distancing during end-of-year holiday period. Right plot assumes substantially less compliance with social distancing during the end-of-year holiday period. Double-dashed vertical line represents the last day used for calibration. The green shaded area shows the 95% posterior model-predictive interval of the outcomes and the colored lines show the posterior model-predicted mean based on 1,000 simulations using samples from posterior distribution. The horizontal black lines show total Covid-19-specific hospital capacity.

Figure 3.

Estimated model-predicted daily incident cases (A) and deaths (B) by scenario in MCMA between December 7, 2020 to March 7, 2021. Left column plots assume compliance with social distancing during end-of-year holiday period. Right column plots assume substantially less compliance with social distancing during the end-of-year holiday period. Vertical dashed line represents the day of policy implementations.

Figure 4.

Estimated model-predicted daily hospitalization demand in MCMA between December 7, 2020 and March 7, 2021. Left column plots assume compliance with social distancing during end-ofyear holiday period. Right column plots assume substantially less compliance with social distancing during the end-of-year holiday period.

Figure 5.

Daily estimated likelihood of hospitalization demand exceeding Covid-19-specific capacity in MCMA between December 7, 20201 and March 7, 2021 by levels of compliance with social distancing during end-of-year holiday period. Top panel assumes compliance with social distancing during end-of-year holiday period. Bottom panel assumes substantially less compliance with social distancing during the end-of-year holiday period.