Happening RIGHT NOW!

We'll be live tweeting the highlights of this talk. #COVID19 #DataScience

1/ https://twitter.com/SigmaXiSociety/status/1319005860232941569

Infectious disease modeling? Translates assumptions into quantitative estimates and provides synthetic lab for experiments.

Can be applied to ecology, math, stats, sociology, etc.

Policy ← Modeling → Science

2/

Examples?
- How does individual-level disease progression affect population-level disease dynamics?
- How pathogen characteristics affect disease establishment?
- Role of population structure?
- etc.

Data are FUNDAMENTAL for disease modeling.

3/

Big Data include for #InfectiousDisease modeling:
- Administrative health data, social media, mobile records.

Opportunities?
- Improved resolution, timeliness, volume

BUT, there are limitations such as:
- Representativeness, noise, ethics/privacy

4/

Important to keep in mind data such as relationships between disease dynamics with covariates (including demographics, environment, policy, behavior, healthcare access/seeking/coverage, physician reporting, etc.)

So, how to optimize disease surveillance?

5/

What about disparities in disease burden?
- Differences socioeconomics, healthcare, racism, etc shape disease burden differences across different areas/states of the USA.

Using models to identify which of these factors are key drivers in dictating disease outcome!

6/

Also using data to estimate how behavior changes across the country. Think of, for example, vaccine refusal in different geographical settings and what contributes to these behaviors (e.g., likeminded people living in same area, etc.)

7/

Now, what aboud #COVID19?

Use data for syndromic surveillance (eg. differentiates between suspected vs confirmed cases, relies on symptoms)

Incorporate the covariate factors (demography, employment, #health, environment) + policy responses (masks, #SocialDistancing, travel)

8/

Trying to address:
- what does the pandemic look like beyond testes cases?
- early dynamics of #COVID19?
- effective mitigation strategies to control transmission?
- socio-environmental-health factors impact on #COVID_19?
- interaction between #SARSCoV2 and other viruses

9/

Now to the Q&A w @bansallab, hosted by @drkiki:

On data source(s):
- State Health Department is the 'source' of the data here to compile it from different sources in a standardized manner. This makes it more uniform regardless of, for example, where the testing was done.

10/

Q: Why is this important if we can't make predictions based on these models?

A: Infectious diseases are based on ~human behavior~ and there are many limitations in predicting the future because of this disease-behavior intricate relationship that we can't directly control.

11/

Q: worried about the upcoming season?

A: Looking at the predictions we had for flu. It was originally as expected but far fewer cases upon introduction #COVID19 (changes in human behavior as driver?) But, depends on sticking to best practices and % of flu vaccination.

12/

Range of clinical symptoms is very broad (mild, severe, systemic issues, death, etc.) compared to flu.

Another big/serious issue? some folks are showing symptoms weeks-to-months after #COVID19 infection!

We can't undermine the disease, we are still learning its dynamics.

13/

Q: social media - beneficial or destructive on these issues?

A: mixed. Pros: source of data, people's views, education, etc. Cons: misinformation is still a big problem.

Overall, it is really helping information dissemination as platforms restrict spread of misinformation.

14/

Q: occupations in this field for nurses/public health? how can we help?

A: healthcare providers are one of the users of these models. A partnership is needed for the final outcome to be useful (policy decisions, etc.)

15/

Q: herd immunity as a possibility?

A: NO. We use concepts of herd immunity to understand VACCINE efficacy & epidemiology in populations. But, in here we are talking about herd immunity WITHOUT a vaccine (so: spread, not efficacy) We have tools (masks, distancing) till then.

16/

Wrapping things up, but make sure to follow @bansallab's work for updates in the field of integrating data into surveillance models to inform decision making for #InfectiousDdiseases.

And check out @SigmaXiSociety for updates on upcoming presentations!

17/end