In a new round of projections, the Scenario Modeling Hub evaluated the trajectory of COVID-19 in the coming year under different assumptions about the scale of the fall 2022 booster campaign and the emergence of a hypothetical new variant in fall 2022

Our main findings include:

• Absent a new variant, COVID levels may continue to rise nationally and stay elevated through early fall. 50% projection intervals suggest that hospitalizations are unlikely to exceed 60% of the Omicron peak. However, this national pattern masks significant variation in projected trajectories between states
• Not all models project a significant peak to result from the introduction of a new variant, however those that do project peaks in late fall or winter.
• Regardless of the presence of a new variant, over the first fifty weeks of the projection period an extended booster campaign is projected to lead to a greater than 15% reduction in hospitalizations and 10% reduction in deaths relative to the restricted booster scenario. This represents reductions of 348,000 (95% PI: -64,000-759,000) hospitalizations and 43,000 (95% PI: -23,000-108,000) deaths in the no variant scenario at the national level.
• Absent a new variant or change in vaccination policy, over the next fifty weeks 1.7 million (95% PI: 503,000-4.5 million) hospitalizations and 230,000 (95% PI: 29,000-873,000) deaths are projected, with very broad uncertainty.
• We are still in a period of substantial uncertainty about the future and there are significant qualitative differences between individual models, particularly in the long-term. The ensembling methodology may blunt the magnitude of the projected peaks when individual models differ in peak timing.
• As of the release of this report, hospitalization have tracked well with the ensemble projections while deaths have tended to be on the lower end of the projection interval.
• These key takeaways are subject to several limitations/caveats:
  • Not all models include BA4/5, and those that do are fit to limited observed data on their circulation.
  • Radical changes in case ascertainment and reporting are making models increasingly hard to fit and clouding the comparison between model projections and observed outcomes.
  • The efficacy of reformulated vaccines is unknown.
  • The new variant introduced in these scenarios is purely hypothetical, and any new variant that does emerge may have very different characteristics.

Table 1. COVID-19 Scenario Modeling Hub round 14 scenarios. More detailed scenario definitions and model characteristics can be found at https://github.com/midas-network/covid19-scenario-modeling-hub.

In a new round of projections, the Scenario Modeling Hub evaluated the trajectory of COVID-19 in the coming year under different assumptions about the durability of the immune response to SARS-CoV-2 and the emergence of a hypothetical new variant in late Spring 2022.

Our main findings include:

• There is high uncertainty in the future trajectory of the pandemic over the 52-week projection period March 2022-March 2023, although outbreaks are projected to be less intense than in the acute phase of the pandemic. There may be occasional periods of increased incidence, although the timing of these periods varies depending on the model considered and assumptions about immunity and variants.
• Moderate rises in hospitalizations and deaths are plausible in the next year, relative to March 2022 levels. A shorter duration of immunity against infection (a median of 4 months vs 10 months) and the arrival of an immune escape variant would hasten these rises.
• Based on a subset of models that include the rise of BA.2 Omicron variant and/or a reduction in social distancing, we anticipate a moderate rise in incidences between April-June 2022.
• During March 2022- 2023, 95,800 cumulative deaths (95% projection interval [PI] 9,000 to 324,000) are projected to occur in the most optimistic scenario (slow waning of immunity and absence of a new variant). This corresponds to 1.06M cumulative deaths (95% PI 0.98-1.29M) since the start of the pandemic, March 2020-March 2023. In the most pessimistic scenario (fast waning and with emergence of an immune escape variant), 211,000 cumulative deaths would occur during March 2022-2023 (95% PI 52,000-466,000).
• See our companion statement regarding the date at which we expect to reach 1M deaths

These projections have several caveats and limitations:

• There is little agreement between models in the timing of future outbreaks.
• The epidemiological characteristics and timing of emergence of future variants is particularly uncertain. Our analysis is limited to the case of a moderate immune escape variant that arises in May 2022.
• It is likely that SARS-CoV-2 testing behavior and reporting of cases and deaths will change in the coming year as COVID19 settles into an endemic pattern. These changes are difficult to fully anticipate and model.

As part of the thirteenth round of COVID-19 scenario projections, the COVID-19 Scenario Modeling Hub has aggregated projections from 7 modeling teams and generated ensemble estimates for the date at which the United States will cross 1 million cumulative COVID-19 deaths, for multiple scenarios spanning uncertainties about the persistence of immunity to the virus. While we may be seeing the end of the acute phase of the pandemic, we feel it is important to take a moment and reflect on this somber milestone and the lives lost during two years of the pandemic.

The COVID-19 Scenario Modeling Hub estimates that, absent the emergence of a new variant, the United States will most likely cross 1 million deaths in June or early July 2022 (see Figure). However, there is substantial uncertainty as to when we will reach this milestone, with the range of dates this event is likely to occur (i.e., the 50% projection interval) spanning May 5 to November 14 at the time projections were made (March 13, 2022). This range captures both the uncertainty between individual models, uncertainty within models, and uncertainty about exactly how long immunity against COVID-19 persists (as detailed in Scenarios A and C).

The emergence of a new variant or other unforeseen event could accelerate this timeline. While it is impossible to know what the properties and timing of such a variant would be, in this round the COVID-19 Scenario Modeling Hub did consider scenarios where a hypothetical variant with moderate immune escape emerged and started circulating in the United States on May 1, 2022. While the considered hypothetical variant did not substantially impact the earliest or most likely time we might see deaths cross 1 million, it did make a longer wait until this milestone less plausible.

These projections come with the caveat that there also continues to be high variability in the reporting of COVID-19 deaths. There have been notable changes in death reporting in recent months in US states which makes model calibration difficult, and may affect our projections. There have also been substantial backfilling and negating of reported deaths in recent months, so we may be closer (or even past) or further from this milestone than we currently realize. We also should acknowledge that it is likely we have already passed this milestone given that many deaths from COVID-19 may not be identified as such, and 1.1 million excess deaths have been estimated to have already occurred (https://www.economist.com/graphic-detail/coronavirus-excess-deaths-tracker).

Figure. Estimates of the date at which the United States crosses 1 million cumulative deaths nationally, from round 13 of the COVID-19 Scenario Modeling Hub.

Table 1. Ensemble estimates of the date to crossing 1 million reported deaths nationally in the US for four scenarios. Estimates are reported as median and interquartile range (in parentheses).

Table 2. COVID-19 Scenario Modeling Hub round 13 scenarios. More detailed scenario definitions and model characteristics can be found at https://github.com/midas-network/covid19-scenario-modeling-hub.

Round 12 is focused on the impact of the Omicron variant in the United States and updates an earlier “emergency” round. We aim to provide a set of planning scenarios around this variant. With updated severity information that has become available since our emergency round, we now provide 4 plausible scenarios that address different levels of severity and immune escape for Omicron. To reflect residual uncertainty in the characteristics of the Omicron variant, we focus on the projection interval rather than the central estimates from the ensemble.

We note several key takeaways from this round:

• Most models project that both cases and hospitalizations peak before the end of January 2022 in most states. The peak is expected to occur earlier in the North East than in the rest of the country.
• At the end of projection period in early April 2022, incidence is projected to drop to low levels, assuming no new immune escape variant.
• In the scenario with low immune escape and optimistic severity, we expect incident national hospitalizations to peak at 146,000 per week (95% PI 94,000-255,000) in the week of January 15th, 2022. In the scenario with high immune escape and high severity we expect national hospitalizations to peak at 193,000 per week (95% PI 79,000-507,000) in the week of Jan 22nd, 2022. Assumptions about immune escape have a low impact on all projections.
• Nationally during the projection period (January 15 to Apr 2, 2022), we expect between 250,000 and 2,036,00 cumulative hospitalizations and 13,000-97,000 cumulative deaths resulting from the Omicron wave in the low immune escape and optimistic severity scenario. There are variations in cumulative estimates between scenarios.
• Substantial uncertainty remains, notably:
  • The intrinsic severity of Omicron and the protection afforded by full vaccine schedules and boosters remain debated.
  • Data are scarce on the serial interval for the Omicron variant. A shorter serial interval would result in a lower transmissibility advantage over Delta; hence a smaller Omicron wave.
  • Case projections should be considered with caution due to potential changes in case ascertainment in the Omicron era. Issues include higher rates of asymptomatic infections, unreported positive home tests, and saturation in testing due to the sheer volume of Omicron infections.
  • There is uncertainty in the duration of protection from reinfection and the possible persistence of the Delta variant, which may affect estimates towards the end of the projection period.

Round 11, focused on the impact of the Omicron variant in the United States, is an “emergency” round aimed at providing a timely set of planning scenarios around this variant. Because of the developing situation, these scenarios were necessarily based on the best evidence at the time of their specification, and new information may deviate from main assumptions. Notably, at this point only the “low-severity” scenarios are considered plausible (scenarios A & B, the scenarios selected by default). Additionally, to reflect the larger uncertainty of this round, we focus on the projection intervals rather than the central estimates from the ensemble.

Despite the inherent uncertainty in this developing situation, their remain several key takeaways from this round:

• The Omicron wave is projected to be sharp and fast in all scenarios, with most models projecting both cases and hospitalizations to peak before the end of January 2022 in every state (all models project peaks before the end of February 2022).
• Cases, hospitalizations, and deaths will likely have receded substantially from the peak by the end of the projection period (March 12, 2022), but are projected to remain elevated compared to June 2021 (the lowest levels seen in the pandemic so far).
• Weekly national hospitalizations could peak at substantially higher levels than reported since the start of the pandemic, with the upper bound of the 95% projection interval reaching 280% of the winter 2020-2021 peak, and 370% of the Delta peak (August- November 2021) in low severity scenarios.
• Nationally during the projection period (December 19, 2021 to March 12, 2022) we expect to see between 409,000-2,380,000 cumulative hospitalizations and 54,000-304,000 deaths. Expected hospitalizations and deaths vary by scenario, with an ensemble median estimate of 832,000 hospitalizations in scenario A and 1,547,000 in scenario B.
• Substantial uncertainty remains, notably:
  • Contributing models show a high level of qualitative and quantitative heterogeneity, reflecting high levels of uncertainty (though models agree on key points as noted above).
  • There is still uncertainty around the epidemiology and severity of Omicron, some of which (e.g., a possible shortened generation time and risk of severe outcomes in naive individuals) could substantially impact projections.
  • Case projections, particularly, should be considered with caution due to potential changes in the definition and identification of a case during the Omicron wave.

We are a consortium of scientists who contribute to the COVID-19 Scenario Modeling Hub. Currently, we are working to produce a round of planning scenarios to help policymakers and the public prepare for the potential impact of the Omicron variant of SARS-CoV-2 that is spreading rapidly in the United States and throughout the world. This process entails careful scenario specification, running of complex models by multiple independent teams, comparison and review of projections, and summarizing the results using formal techniques. Although we have not completed this round (Round 11), preliminary projections from multiple models raise deep concerns about the speed and severity of the coming Omicron wave. This urgency is further fueled by the unprecedented speed at which the Omicron variant has become the dominant strain in many areas. We believe it is important to share these preliminary results to help individuals and institutions prepare for potential surges in cases and healthcare demand.

The planning scenario projections produced so far universally show a large wave of COVID-19 cases that will likely exceed numbers seen nationally at the peak of the Delta wave by the first week of January 2022. While it is clear that the wave of COVID-19 infections will be large, it is less clear what Omicron’s impact will be in terms of hospitalizations and deaths, as much remains uncertain regarding the severity of primary, secondary and breakthrough infections with Omicron compared to what was witnessed for Delta and prior variants. The scenarios considered by the COVID-19 Scenario Modeling Hub capture a range of possible severities for the Omicron variant consistent with current evidence. Depending on severity, preliminary projections of weekly hospitalizations range from levels twice as high as any previously observed in the United States during the pandemic, to levels around half of what was seen at the peak of the Delta wave. Regardless of where the relative severity of Omicron falls, the sheer number of cases projected means that even a relatively mild severity of the Omicron variant has the potential to severely stress, if not overwhelm, already strained health care systems across the country. This includes straining our ability to provide COVID-19 testing to all who need it. There is substantial uncertainty underlying these projections, but they remain one of the best tools at our disposal to prepare for the effects of Omicron in the face of significant uncertainty.

The work of the COVID-19 Scenario Modeling Hub teams, multiple other modeling groups world wide, and the epidemic situation in countries where Omicron is already widespread, all indicate that the United States must be prepared to face an unprecedented wave of COVID-19 cases in the coming weeks and months. However, we are not helpless in the face of the coming wave, and have proven tools to blunt its impact. Physical distancing, limiting gatherings and masking have repeatedly proven themselves effective in slowing the spread of the virus. Vaccination, even if somewhat less effective due to Omicron’s immune escape, remains one of the best ways to reduce the risk of severe outcomes if infected and still reduces the chances of passing on the virus to others. Reassuringly, boosters have been shown to reduce the risk of symptomatic Omicron infections, and hence will likely provide high levels of protection against hospitalization and death; yet only 30% of those eligible in the US have received a booster shot. Antigen testing also provides an important tool to help prevent spread at holiday gatherings, though the availability of these tests has become limited in recent weeks. To the extent to which individuals and communities proactively use these tools, many of the worst outcomes projected by our models may not come to pass.

The COVID-19 Scenario Modeling Hub is working towards a rapid release of more formal planning scenarios assessing the threats posed by the Omicron wave. It will be important to understand the potential impact of the Omicron wave in different US jurisdictions that have various levels of population immunity. These COVID-19 Scenario Modeling Hub projections will undoubtedly be augmented by other critical information that will emerge in the coming weeks pertaining to viral characteristics, immune response, severity, and population level impact. However, the best information we have at the moment indicates the threat posed by Omicron is substantial and imminent, and individuals and governments should be prepared to respond accordingly.

Signed

• Justin Lessler, UNC Gillings School of Public Health
• Katriona Shea, The Pennsylvania State University
• Rebecca Borchering, The Pennsylvania State University
• Cecile Viboud, Fogarty International Center, NIH
• Shaun Truelove, Johns Hopkins Bloomberg School of Public Health
• Spencer Fox, The University of Texas at Austin
• Bryan Lewis, University of Virginia
• Srini Venkatramanan, University of Virginia
• Przemyslaw Porebski, University of Virginia
• Alessandro Vespignani, Northeastern University
• Matteo Chinazzi, Northeastern University
• Ajitesh Srivastava, University of Southern California
• Harry Hochheiser, University of Pittsburgh
• Emily Howerton, The Pennsylvania State University
• Claire P. Smith, Johns Hopkins Bloomberg School of Public Health
• Shi Chen, University of North Carolina at Charlotte
• Thomas Hladish, University of Florida
• Alexander N. Pillai, University of Florida
• Jiangzhuo Chen, University of Virginia
• Guido Camargo España, Univerasity of Notre Dame