We don’t know how best to reopen the country, but we have a playbook

The primary strategy to stem the tide of the COVID-19 pandemic across the country has been the implementation of strict social distancing measures to avoid overwhelming hospitals and causing unimaginable mortality. Now, as data is accumulating that social distancing is reducing new cases, attention is turning to reopen the economy while minimizing the chances of reigniting uncontrolled viral spread. 

Federal and public policy groups and think tanks are actively engaged in plotting exit strategies centered on three pillars — testing, contact tracing, and slow and selective reopening of workplaces. We cannot afford to settle on policies that do not work in the coming months before a vaccine becomes available, and absent a coordinated effort, we risk ending the summer with no better idea of how to exit the stay-at-home period safely than when we started.

How should our government make vital policy decisions when there is no agreed-upon (let alone tested) best solution?  

It’s prudent to draw on the lessons learned by the medical field as we sail through uncharted policy waters over the next few months. 

Be skeptical of conclusions based on observations 

Spurious correlations based on observations abound in medicine. Just as stress does not cause ulcers and polio isn’t caused by ice cream, we must be skeptical of observations in our current search for policy solutions to combat the spread of COVID-19. It has been observed that countries pursuing aggressive contact tracing measures including South Korea, Singapore and China have been more successful at reducing mortality, but is contact tracing the reason for this success? Maybe. Alternatively, perhaps, countries that have pursued more extreme contact tracing (primarily in Asia) are doing other things differently from countries that haven’t (the U.S. and Europe). A great number of structural, policy, and cultural differences, and not contract tracing per se, may make these countries better suited to dealing with the pandemic.  

The solution can be worse than the problem

Medical history is full of well-intended treatments that turned out to cause more harm than good (radium-containing water was a widely touted cure-all in the 1920s and much blame for the current opioid crisis can be attributed to over-prescription). 

So how do we evaluate a treatment or policy objectively so that we neither overreact nor underperform? In randomized controlled trials, cohorts that are equivalent with respect to trial-relevant criteria are randomly assigned an experimental method and outcomes are directly compared between these groups. This is a central feature of the medical playbook and has time and time again overturned what seemed like solid common sense observations and promoted implementation of effective medical interventions.

Controlled trials are complicated to carry out, and there is a tantalizing allure to ignore time-honored lessons to come up with quick fixes even in the medical field. In the rush to devise a COVID-19 therapeutic regimen, thousands of patients have received the anti-malarial drug hydroxychloroquine (HCQ) based on a small number of promising observations by clinicians without using controlled trial methods. Some of these patients have died of heart attacks, a known risk factor of this drug. Now, several early controlled trials have suggested that HCQ may be ineffective at treating COVID-19. It is understandable that severely ill patients would accept the risk of any medicine that offers a glimmer of hope, yet the medical field has learned repeatedly through conducting controlled trials that the majority of treatments simply do not work. 

Similarly, if we want to come out of the next few months with some understanding of what policies are truly effective in withstanding this and future pandemics, the value of controlled studies cannot be overstated. Setting up true randomized controlled policy trials presents considerable, perhaps insurmountable challenges. Nonetheless, there is a set of lessons from the medical field that can be applied to the design of policy, each of which should be considered as this country embarks on a set of high-stakes societal experiments for which there is no clear best path forward. 

There is no panacea

A key step in medical trial design is to match treatment options with people who are most likely to benefit from them. Just as there are different treatments for each type of cancer, certain policies to combat COVID-19 may benefit only those municipalities with certain infrastructure or a particular population density. Policies should be tailored to municipalities of different compositions, such as rural vs. urban locations. 

Limit the number of test conditions

A typical medical trial evaluates only a handful of variable conditions. Evaluating only a singular COVID-19 testing, tracing, and reopening policy may leave better policies unexplored, but trialing bespoke policies in every community will prevent us from disentangling the efficacy of specific policies from sheer luck or other factors that leave some communities predestined to better or worse outcomes. Policymakers should agree on a limited menu of specific testing and contact tracing policies for each set of comparable municipalities to trial on a controlled backdrop of standardized statewide policy and public messaging.  

Evaluate the results evenly and carefully 

Standardizing the metrics of success is one of the most crucial aspects of medical trial design and it may be the most important lesson the medical field can apply to inform societal policy. Standardized measurement and reporting on COVID-19 metrics allow direct comparison of data, such as positive diagnoses, which is of immeasurable benefit to assessing the effectiveness of policies in containing the spread of the virus.

But the impact of this pandemic reaches far beyond the immediate health risk of viral infection. The shockwaves will impact our collective employment opportunities, mental health, and social interactions for years to come, and the policies we implement from scratch over the coming months will have ripple effects on every aspect of American life. If we ask the right questions at scale, we have an opportunity to identify policies that provide the greatest holistic benefit to society. 

Our normal life is on hold and communities across our nation are facing a singular challenge for which there is no cure. This shared baseline and shared goal creates a rare setting for us to measure wide-ranging impacts of policy changes, the likes and importance of which we are unlikely to experience again in decades. Any information we glean on policies that succeed or fail, and the infrastructure we build to better evaluate policies, in general, will not only help us over the upcoming year but will help us build a better policy arsenal for the next pandemic. 

In all, the most important lessons the medical field can offer as we proceed through this critical juncture in history are lessons in humility. We must acknowledge that we do not and cannot know the best path forward, that our observations can lead us to draw incorrect conclusions, and that well-meaning but untested solutions can make problems worse.

Coordinating how these policies are devised, implemented, and evaluated according to tried and tested methods from the medical playbook will allow us to converge on the most effective routes to keep the economy open while minimizing mortality and morbidity of COVID-19.

Richard Sherwood, Ph.D., is an assistant professor of Medicine at Brigham and Women’s Hospital and Harvard Medical School. He runs a research lab developing ways to understand and treat disease through gene editing. Follow him on Twitter: @SherwoodLab. Mandana Arbab, Ph.D., is a postdoctoral fellow at the Broad Institute of MIT and Harvard University. She researches applications of CRISPR-Cas9 gene therapy. The opinions expressed in this article are solely those of the authors and do not reflect the views and opinions of Brigham and Women’s Hospital.