Public Comment

Better COVID-19 Tracking Data Needed to Plan for
Re-Opening Berkeley

Claire V. Broome, M.D. , Assistant Surgeon General, US Public Health Service (retired)
Sunday May 24, 2020 - 01:21:00 PM

Residents of Berkeley need better information to track Covid-19 in Berkeley and Alameda County

I worked in public health for 28 years, so I understand the challenges facing Health Departments in the City of Berkeley, Alameda County, and the state. The fundamental tool for health departments facing a new frightening disease like Covid 19 is data –how many new cases each day? Are case numbers stable, increasing, or decreasing? Where are cases happening? What are the risk factors (eg exposure to known case, essential worker, crowded facility)?

Many residents may not understand that with a new disease, Public Health Departments at all levels of government are learning as they go, doing what is urgent based on knowledge and past experience, and figuring out how to improve that response. Systems that worked for routine disease tracking may need to be adapted on the fly to track and respond to Covid-19.

But residents of Berkeley like me need to know what's happening in our city now, and how the City will make informed decisions as we move forward with re-opening. The essential questions: are NEW cases stable, going up, or down? As we re-open, where are new cases coming from? How will the City effectively control spread so we are safe but can get back to school and work?

The City of Berkeley and Alameda County provide the number of new cases each day confirmed by PCR testing. But the number of cases confirmed by testing may be difficult to interpret due to recent increased availability of testing. Testing may also not reach vulnerable populations. 

We also need to know the number of NEW cases each day hospitalized or seen in Emergency Departments (including both those confirmed by test, and suspect Covid-like illness). Persons ill enough to seek care represent a more consistent measure over time. Currently reported hospitalization numbers include all who are currently in hospital with Covid, or suspected Covid—but Covid-19 cases may be hospitalized for weeks, meaning that the hospitalization number reflects what was happening in the past, as well as counting the same case on multiple days. The figure is useful for monitoring health care resources—but virtually meaningless for tracking changing disease trends. Privacy of health care data is important—but so is residents’ trust that meaningful data will be available. Given that Berkeley residents can be hospitalized in many different locations, providing this data in some de-identified form (potentially a three day moving average?) should be within the capability of Health Departments. 

While the number of tests administered is an indicator for reopening, this is an indirect metric. One could administer a lot of tests to the “worried well”, a low-risk population, and miss disease spread in high-risk populations who have less access to testing. Reporting new hospitalizations/Emergency Department visits avoids that distortion, as it can identify ill persons who may not have consistent care providers. For example, Berkeley's case rate is low compared to the rest of Alameda County (64 vs 170 per 100,000 population), and tests are widely available in Berkeley. The case data are more informative than Berkeley’s lower rate of testing. 

There've also been a lot of calls for repeated random testing of asymptomatic persons. That violates Sutton's Law —when you have a complex test, limited personnel, and testing sites, you go where the money is –i.e. individuals with symptoms, and a systematic approach to high risk settings, like health care personnel, essential workers, unsheltered persons, and crowded facilities. 

Finally, as Berkeley moves cautiously into reopening, what is the Health Department's plan for identifying where transmission is occurring? The tracking systems discussed above identify cases. Then Berkeley needs to find people who may have been in contact with an infected person; test and isolate as needed; and understand if possible where their infection may have come from. We can’t wait for an app solution which may or may not be useful—there is no substitute for on the ground Public Health investigation by City employees. 

We are fortunate to have well trained dedicated public health workers at local, state, and national levels working 24-7 on the crisis. A critical part of their work requires public trust—trust which will be increased with transparent data sharing and descriptions of the work they are doing to keep us safe.