Our society is at a critical juncture in the fight against COVID-19. We’ve seen some successes: Broad social-distancing policies have helped to level off or decrease infection rates in most states. And hospitals remain far below capacity in all but a few hot spots. However, these intensive steps have also produced dire societal consequences: unemployment, isolation, heightened stress and uncertainty abound. And treatment of medical conditions — including some cancer surgeries — has taken a back seat. It’s critical that we remobilize society— and it’s critical that we establish appropriate strategies, neither too lenient nor too strict, for doing so.

Appropriately, current planning efforts by federal, state, and academic leaders offer a combination of testing, tracking and hygiene that will allow gradual relaxing of the strictest social distancing. But overly strict preconditions — including one lynchpin of the “Roadmap to Pandemic Resilience” released by a group at Harvard University — risk hampering our progress.

This report contains numerous well-conceived ideas, including a ramp up of testing and tracking, which are key to managing and mitigating the virus. However, it also calls for near-universal testing: “5 million tests per day by early June to deliver a safe social reopening,” increasing to “20 million a day (ideally by late July) to fully remobilize the economy.” It states that even this may not be enough. Such assertions are repeated eight times in the document.

To put this into perspective: 20 million tests per day would mean testing every individual in the U.S. every two weeks or so. It is not clear how long it would take to accomplish this, even with a simple test such as blood pressure, much less with tests in limited supply and in high demand internationally, which require personal protective equipment to administer. This would be a greater-than-100-fold increase over our current daily testing numbers; we already have by far the most robust raw testing in the world and a per capita rate on par with most other countries. The report provides no explicit argument for why these numbers are necessary conditions to begin reopening society. And if followed strictly, they would effectively keep current lockdown procedures in place indefinitely.

While increased testing clearly must be an important component of our strategy, I believe that we can use what we have learned in the past month to implement more targeted social distancing measures that retain much of the benefit of the current lockdowns. While we still have much to learn, we now know a lot about this virus. It seems to spread by close personal contact or contaminated surfaces, and “super-spread” via large, dense gatherings is a key risk, especially given COVID-19’s communicability from asymptomatic carriers. Recent data suggest that possibly more than half of infected people remain asymptomatic, and among symptomatic infected, about 80% only have mild symptoms. Most critical patients have known co-morbidities, and most deaths occur in these high-risk populations.

Based on this knowledge, perhaps for many populations we can immediately relax some of the most extreme measures and focus on more targeted mitigation steps: rigorous disinfecting, limited large gatherings, basic social distancing, and public mask wearing. And we can apply more restrictive practices for vulnerable populations and high-density areas. Hybrid epidemiological-statistical modeling tools that measure disease spread in relation to social movement, density, and demographics may be useful in helping identify which communities need to take more caution. And if overall containment and control are still not feasible, these steps may be ample to keep the infection rate sufficiently low in most places, enabling our healthcare system to manage a limited number of severe cases.

It is clear our society needs to do a better job with testing and tracking. We need to vigorously develop supplies and infrastructure for broad viral and antibody-based testing and contact tracing, perhaps advanced by data science and mobile apps, to help limit spread and track and manage risk. However, we can reopen certain elements of society and ramp up our testing in parallel. I believe this is the best course to manage the virus, retain most of the benefits of lockdowns, and mitigate societal damage.

Jeffrey S. Morris is director of the biostatistics division in the department of biostatistics, epidemiology, and informatics at the University of Pennsylvania’s Perelman School of Medicine. He can be reached at jeffrey.morris@pennmedicine.upenn.edu. His website links to key covid research results, data sets, and applications and models with commentary.