Build an integrated public health surveillance system

Nigam Shah and Rusty Hofmann on establishing a robust public health surveillance system

A parable succinctly makes the case for building a robust public health surveillance system. An angler noticed someone drowning and jumped into the river, swam the victim to shore and resuscitated them. Minutes later, the angler saw another person drowning, then another and another. Recognizing that saving so many people one by one was unsustainable, the exhausted angler walked upstream and discovered a broken bridge that was causing people to fall into the water. After the angler fixed the bridge, the drownings stopped.

Stanford Medicine experts Nigam Shah, MBBS, PhD, associate professor of medicine and biomedical data science, and Rusty Hofmann, MD, professor of radiology, have much in common with the angler in the parable. They aim to collect, analyze and interpret data to help in solving public health challenges. Using the latest advances in data science, artificial intelligence, machine learning and other technological breakthroughs, they hope to identify at-risk regions and populations — essentially predicting and preventing the spread of disease. Both Shah and Hofmann have launched projects to proactively respond to the COVID-19 pandemic. Below, Shah and Hofmann discuss how to establish an even more effective public health surveillance system.

What attributes does a public health system need to reach its potential, and how has Stanford Medicine facilitated that?

Shah: On the data side, there are three parts to a robust data-driven public health surveillance system. Part one is monitoring what is happening in our institution and region. At the beginning of the COVID-19 pandemic, Stanford Medicine researchers set up dashboards for our region as well as internal dashboards for the two adjacent counties before they developed their own.

Part two is responding based on the information gleaned from the monitoring. And there are two essential response mechanisms. One is providing situational awareness in the form of a daily update, then making forecasts about what will happen tomorrow and the following week. We developed mechanisms for that pretty well.

The third step is enablement, setting up data so that public health researchers can start using what is collected for surveillance and research. This might take the form of provisioning our own data as well as creating consortia that would get us access to regional data. So, if we’re building a data-driven public health surveillance system, it needs to have these three functionalities: monitor, respond and enable.

Before the pandemic, what challenges has Stanford Medicine faced related to public health surveillance?

Shah: In the United States, public health surveillance varies drastically county by county and state by state. One reason for that is that regulation around public health surveillance gets set at the state level, but the mechanism of compliance gets set at the county level. Health systems are required to comply with all of those regulations — both the what and the how. So, for example, if Stanford sees patients from nine Bay Area counties, each of those nine can have a different way by which we have to report a reportable condition to them. Some will require a paper form. Some a fax. Some an electronic case reporting system via Digital Bridge, and so on.

Hofmann: Building on that, the data we do have on COIVD-19 is so fragmented that it’s difficult to track. Even with the data, you don’t have an opportunity to be proactive. What is currently in place is a very reactive system in which it takes days to get the data, analyze it and formulate a response. It’s hard to know whether this is a structural issue, a messaging issue or a leadership issue. For substantive change to occur in public health surveillance, we need a consistent message on what the problem is, how we are going to measure the problem, what we are going to do to solve the problem and, most importantly, what success looks like.

How has Stanford Medicine’s response to the pandemic had a broader impact on public health?

Hofmann: One thing that Stanford Medicine has done exceptionally well is to lead by example. Stanford Health Care not only provided outstanding care for patients, but it also kept people informed, protected clinicians and staff, and ramped up testing. Additionally, many of our scientists pivoted their research to respond to this pandemic. As a health care provider and as a scientific community, Stanford Medicine’s strategic response has provided innovative insights and served as an example of how to keep communities safe.

In the midst of the COVID-19 pandemic, what role does Stanford have in creating a more proactive public health surveillance system?

One of the many things that Stanford does very well is building proofs of concept or pilots

Shah: The role that an institution like Stanford can play is to provide an outline of how a highly effective public health surveillance system could look. My colleague Bonnie Maldonado [MD, senior associate dean of faculty development and diversity, and professor of pediatrics and of health research and policy] and I, along with collaborators at UC San Francisco, proposed the Bay Area SARS-CoV-2 Information Commons, or BASIC. Through BASIC, we hope to anticipate the evolution of the COVID-19 pandemic better. Whereas existing predictions of the pandemic lack accuracy due to many factors, including data fidelity, BASIC combines diverse inputs to quantify the effects of the shelter-in-place order in the Bay Area. This ability to create more accurate inputs for models about the pandemic would serve as a step in the right direction. However, to reach its full potential, it will require participation from many competing health systems. Though BASIC essentially bypasses the official public health reporting systems, it can be a proof of concept of what could be done with modern technology that the public health infrastructure could then adopt.

Hofmann: One of the many things that Stanford does very well is building proofs of concept or pilots. To do surveillance of COVID-19, my team developed the National Daily Health Survey, which tracks COVID-19 symptoms nationally using self-reported data and creates a heat map that can inform institutional and governmental public health response. We have seen engagement with these projects, but you need a regulatory body’s firepower to take it to the next level. So, Stanford’s role is to identify solutions and share them with public health officials who can scale them as part of their response. We’re very privileged to be part of Stanford University, which has the financial and intellectual resources and outstanding leadership required to help contribute to solving in some way, shape or form some piece of this global pandemic.