Scenario Background: You are a disease modeler supporting the Los Angeles County Department of Public Health, at the beginning of the original Omicron wave. The LA County Board of Supervisors is concerned about what the next few months will look like, and what level of intervention will be required to manage what is shaping up to be a large winter Covid-19 wave. Vaccines were broadly available during this time period and vaccination should be accounted for in the modeling.
Scenario Setting/Situation:
Time = December 28th, 2021 (right around upswing of Omicron wave), Location = LA County
Scenario Asks:
Find a model capable of forecasting Covid cases and hospitalizations (these don’t need to be broken down by vaccination status, but the model should account for vaccination in some way). Parameterize model either using data from the previous two months (October 28th – December 28th, 2021), or with relevant parameter values from the literature. Forecast Covid cases and hospitalizations over the next 3 months under no interventions.
Based on the forecast, do we need interventions to keep total Covid hospitalizations under a threshold of 3000 on any given day? If there is uncertainty in the model parameters, express the answer probabilistically, i.e., what is the likelihood or probability that the number of Covid hospitalizations will stay under this threshold for the next 3 months without interventions?
Assume a consistent policy of social distancing/masking will be implemented, resulting in a 50% decrease from baseline transmission. Assume that we want to minimize the time that the policy is in place, and once it has been put in place and then ended, it can't be re-implemented. Looking forward from “today’s” date of Dec. 28, 2021, what are the optimal start and end dates for this policy, to keep projections below the hospitalization threshold over the entire 3-month period? How many fewer hospitalizations and cases does this policy result in?
Independent from question 3, assume there is a protocol to kick in mitigation policies when hospitalizations rise above 80% of the hospitalization threshold (i.e. 80% of 3000). When hospitalizations fall back below 80% of the threshold, these policies expire.
a. When do we expect these policies to first kick in?
b. What is the minimum impact on transmission rate these mitigation policies need to have the first time they kick in, to (1) ensure that we don't reach the hospitalization threshold at any time during the 3-month period, and (2) ensure that the policies only need to be implemented once, and potentially expired later, but never reimplemented? Express this in terms of change in baseline transmission levels (e.g. 10% decrease, 50% decrease, etc.).
Now assume that instead of NPIs, the Board wants to focus all their resources on an aggressive vaccination campaign to increase the fraction of the total population that is vaccinated. What is the minimum intervention with vaccinations required in order for this intervention to have the same impact on cases and hospitalizations, as your optimal answer from question 3? Depending on the model you use, this may be represented as an increase in total vaccinated population, or increase in daily vaccination rate (% of eligible people vaccinated each day), or some other representation.
Question | Tasks | TA Workflow Tested | Metrics
-- | -- | -- | --
1 | Find a model that meets the scenario requirements (or can be updated to meet the requirements). Model requirements: · Needs to output infections and hospitalizations (or can be modified to do so) · Needs to support NPIs that effect a change in baseline transmission levels, including support for interventions implemented at different points in time, and for various lengths of time, which may depend on the values of output variables (hospitalizations in particular). The model should allow the user to choose the specific start date of the intervention, and length of time in place, etc., or set triggers for intervention start and end times. · Needs to represent vaccination with some way to increase level of vaccination in population | TA1: Model Search and Discovery | Time: How long does search for appropriate model take? Relevance (qualitative): How suitable is the selected model for the scenario described?
1 | [Optional] If relevant model is not already in the ASKEM system, do model extraction from paper/code | TA1: Model Extraction | Time: How long does knowledge extraction take? How long does it take to get model into executable form? Quality (qualitative): Qualitative score on metadata quality (correctness, relevance, completeness), based on human inspection of the equations, variables, parameters, etc.
1 | [Optional] If required, do necessary model extension/transformation to meet scenario requirements | TA2: Model Extension/ Transformation | Time: to extend/modify model
1 | Parameterize model according to time and location. This may require a literature search, or search for data to inform model parameters. | TA1: Search and Discovery (for parameters) | Time: How long does search for information required to fully parameterize the model?
1-5 | Answer questions 1-5, with supporting output to justify the answers. | TA3: Simulation Workflows; Answers to Scenario Questions | Time: For each question, measure time to set up simulation workflows, get final answer and prepare supporting output. Quality (qualitative): Does output seem reasonable given the scenario?
| [Optional] If at any point, you need to search for parameter values, do a literature review, or find datasets, please track time spent, approach taken (e.g. what were the keywords or key concepts you searched by), and sources/databases you searched across. | TA1: Search and Discovery | Time: How long does search for required information take?
Scenario 2: Limiting Hospitalizations
Scenario Background: You are a disease modeler supporting the Los Angeles County Department of Public Health, at the beginning of the original Omicron wave. The LA County Board of Supervisors is concerned about what the next few months will look like, and what level of intervention will be required to manage what is shaping up to be a large winter Covid-19 wave. Vaccines were broadly available during this time period and vaccination should be accounted for in the modeling.
Scenario Setting/Situation:
Time = December 28th, 2021 (right around upswing of Omicron wave), Location = LA County
Scenario Asks:
<html xmlns:o="urn:schemas-microsoft-com:office:office" xmlns:w="urn:schemas-microsoft-com:office:word" xmlns:m="http://schemas.microsoft.com/office/2004/12/omml" xmlns="http://www.w3.org/TR/REC-html40">
Question | Tasks | TA Workflow Tested | Metrics -- | -- | -- | -- 1 | Find a model that meets the scenario requirements (or can be updated to meet the requirements). Model requirements: · Needs to output infections and hospitalizations (or can be modified to do so) · Needs to support NPIs that effect a change in baseline transmission levels, including support for interventions implemented at different points in time, and for various lengths of time, which may depend on the values of output variables (hospitalizations in particular). The model should allow the user to choose the specific start date of the intervention, and length of time in place, etc., or set triggers for intervention start and end times. · Needs to represent vaccination with some way to increase level of vaccination in population | TA1: Model Search and Discovery | Time: How long does search for appropriate model take? Relevance (qualitative): How suitable is the selected model for the scenario described? 1 | [Optional] If relevant model is not already in the ASKEM system, do model extraction from paper/code | TA1: Model Extraction | Time: How long does knowledge extraction take? How long does it take to get model into executable form? Quality (qualitative): Qualitative score on metadata quality (correctness, relevance, completeness), based on human inspection of the equations, variables, parameters, etc. 1 | [Optional] If required, do necessary model extension/transformation to meet scenario requirements | TA2: Model Extension/ Transformation | Time: to extend/modify model 1 | Parameterize model according to time and location. This may require a literature search, or search for data to inform model parameters. | TA1: Search and Discovery (for parameters) | Time: How long does search for information required to fully parameterize the model? 1-5 | Answer questions 1-5, with supporting output to justify the answers. | TA3: Simulation Workflows; Answers to Scenario Questions | Time: For each question, measure time to set up simulation workflows, get final answer and prepare supporting output. Quality (qualitative): Does output seem reasonable given the scenario? | [Optional] If at any point, you need to search for parameter values, do a literature review, or find datasets, please track time spent, approach taken (e.g. what were the keywords or key concepts you searched by), and sources/databases you searched across. | TA1: Search and Discovery | Time: How long does search for required information take?