Baylor College of Medicine

The Wisdom of the Crowd


Aug. 26, 2020


Dear Members of the Baylor College of Medicine Community,
I am not going to write a weather-related piece today, as everyone is likely receiving plenty of information through the usual sources.  As I write this, it is mid-day Wednesday and it the path of Hurricane Laura is becoming more certain.  It appears it will not hit Houston directly, which if forecasts hold would help to mitigate the impact.  However, many are still directly in the path of a very dangerous storm, and our thoughts and/or prayers are with them.  Our support will be needed in the aftermath.
Obviously, there is concern about what impact this event could have on our COVID-19 outbreak.  If large numbers of people are displaced, and forced into congregate shelters, this will create a risk of spread of COVID-19.  Public officials have taken steps to limit the size of shelters and incorporate distancing principles into their design and operation, but any significant population displacement will place additional pressure on disease control efforts.
So with that brief nod to the issue that is on everyone’s minds today, I will return to an update on COVID-19 issues.  A couple of weeks ago, I posed the question to all of you: “If you were advising one of our presidential candidates, or local leadership, or a company executive on preparation for the next global pandemic, what is your priority?”  You emailed your responses to me, or posted them on Twitter using the hashtag #TheNextPandemicBCM where they are still available for review.  Following a brief review of our COVID-19 numbers, I want to report what you said.
The COVID-19 numbers continue to trend well, for the most part.  From prior posts, you know we are focused on three primary metrics: R(t), diagnostic test positivity rates among TMC facilities, and the number of new daily COVID-19 cases in Harris and surrounding counties.  
Recall the R(t) calculation is an estimate of viral infectivity, influenced by the properties of the virus and community dynamics.  Simply put, if I am infected with COVID-19, the R(t) is a statistical estimate of how many other people I am likely to infect.  If R(t) is less than 1.0, Houston is winning; greater than 1.0, SARS-CoV-2 is winning.
Even with a highly infectious virus (like measles), if we are truly and completely isolated from everyone else – a condition which only exists as a hypothetical – the R(t) would be effectively zero.  Conversely, if we all tend to congregate in crowded spaces with poor air exchange for prolonged periods of time, even a virus which is not inherently very infectious could have a very high R(t).  This is why attention to masking and distancing is so important.  The good news this week is the R(t) has been consistently less than 1.0 for 16 consecutive days.  As of now, we are winning.
Our second metric is the rate of positive tests among all tests ordered by Baylor and other TMC affiliate institutions.  This is a rough gauge of viral prevalence, which we would like to see sustainably below 5%.  The TMC positive test percentage is now 6.7%, and has been slowly trending downward.  This is not where it needs to be, but is headed in the right direction.  If you follow the city, county and state numbers, you will note that their percentage positive tends to be a few points higher.  This probably reflects some selection bias on the part of the Baylor and TMC facilities cohort.  Although we test symptomatic people in our ambulatory clinics, much of our testing is asymptomatic employees in surveillance programs, and asymptomatic patients admitted to hospitals, who would be expected to have a low positivity rate.  
However, we are very confident in the integrity of the data point we generate, which is why we have chosen this as our metric.
Finally, our most problematic metric: new daily COVID-19 cases in the community.  This number is (very) slowly drifting down, and on a 7-day rolling average basis is between 1,400 and 1,500.  This is good news is that in mid-July we were at about 2,400 cases.  However, we are still seeing 5-6 times the number of daily cases compared to May, and for public health officials to have a realistic chance of performing effective contact tracing it needs to be closer to 200.  This metric is the most methodologically flawed, for reasons I have previously reviewed, but should be increasingly reliable as state data issues are resolved.
In summary, the trends are encouraging, and hopefully future community stressors like school reopenings, the Labor Day holiday, flu season, etc. will not set us back.
Please continue to preach the masking and distancing message to your circles of influence.  
I hope we as a nation will get to a point soon – we are not even close yet – where we can stop managing the crisis in front of us, and begin to focus on how to be better prepared for the next pandemic.  That was the point of the five word #TheNextPandemicBCM exercise.  Roughly 100 people from the broad Baylor community responded.  Some people responded with tactical concerns, such as the need to stockpile PPE. A few made – mostly respectful – statements favoring one side of the political spectrum or the other.  Most responses addressed broad, fundamental key principles.  Taken individually, none of the principles is particularly surprising; however, there were four major themes when taken together constitute a thoughtful outline for how we can be better prepared next time – the BCM crowd-sourced pandemic preparedness plan:

  • Elevate the role of rigorous science in public policy.  Example of five words: Build public confidence in science; Clear communication improves scientific understanding; Trust scientists and not Facebook.  Fully one quarter of respondents reflected this theme in some way.  
  • Build reliable data and reporting systems.  Five words:  Worldwide electronic medical interconnectivity; Good data drives good decisions; Reliable information must unite us.  
  • Invest in public health infrastructure and education.  Your words: Comprehensive integrated public health network; public health in high school; Public health investment dollar well spent (technically, this last one is six words, but it was too good not to include)
  • Effective leadership is critical. Effective leadership: national, regional, local; Good leadership drives action; Clear messaging from leadership unites.

Our limited “Twitter poll” is obviously not a methodologically sound, comprehensive analysis of our current state nor an attempt at meaningful future planning.  However, the Baylor community does represent an interesting collection of researchers, educators, clinicians and learners.  Our exercise does represent a limited snapshot of the mood of this community at a time we were still actively engaged in our struggle with the virus.  Based on this snapshot, we should be advocating for public health curricula across the entire educational landscape: elementary to graduate school.  Our more informed populace will need reliable data presented in an accessible manner.  Our scientific community will continue its long-established commitment to academic integrity.  Leaders need to be able to promote an environment of trust and confidence.  These concepts may seem so obvious as to be trite.  However, if we do not actively work to address these – and no doubt other – issues now, our motivation and sense of urgency will flag as COVID-19 fades into memory.  As a learning health system, we have an obligation to hold our societal feet to the fire.
Stay well and stay safe.

James T McDeavitt, M.D.
Note:  please feel free to share these communications broadly, including outside of Baylor.  The following link can be copied and pasted into texts, emails or your social media venue of choice: