This essay is featured in Boston Review’s new book, Thinking in a Pandemic.

Half a year into the COVID-19 pandemic, more than 150,000 Americans have died from the disease. Over one percent of the population has had a confirmed infection, with roughly 90 percent of infections missed. Several other nations have brought transmission under control, but the United States is facing a rapid uptick in the number of new cases. Despite our desperation for life to go back to normal, the end is not yet in sight. How should we think about the ongoing challenge we face?

At this point, we know the problems. What are we doing to fix them?

At this point, we know the problems. The novel coronavirus, SARS-CoV-2, can spread before infected people develop symptoms—between 4 and 41 percent of people never develop symptoms, in fact—yet we are nowhere near herd immunity. At the same time, the United States continues to lack sufficient capacity for testing and contact tracing. Countries like Germany, Greece, Italy, and Australia conduct on average two hundred tests to find a single confirmed case; in the United States, every twelve tests uncovers a new infection. Widespread lockdowns imposed in March and April dramatically slowed epidemic growth, but they came with enormous costs, not only economic—leaving tens of millions unemployed—but also social and medical. Vital medical services have been disrupted, including infant vaccination, cancer screening, and HIV treatment programs. The impacts are far-reaching and severe. On the other hand, where lockdowns have been lifted too quickly, transmission has resurged, leading some states to reclose businesses.

It is natural to react to seemingly impossible situations by focusing on the problems. Why is this happening? Why were we so unprepared? Who is to blame? Our collective experience during the pandemic has been likened to the stages of grief; hundreds of millions of people have collectively experienced denial, anger, bargaining, and depression. The last, critical stage is acceptance. Given that the virus is here to stay, how do we learn to live with it safely?

This shift in framing may seem simple, but it is important. Energy focused on arguing about problems is energy not spent on developing and implementing solutions. Of course, characterizing problems can be an important first step in finding answers. Recognizing the role of severe clotting in COVID-19 cases, for example, has helped clinicians to explore anti-thrombotic treatments. But when it comes to discussion of broader policies, we must be careful not to get stuck at the problem-finding phase. There is clear evidence that the pandemic is disproportionately harming Black, Latinx, and Indigenous communities in the United States. Now that we have identified that problem, what are we doing to fix it?

As we continue the work of implementing sorely needed solutions, there are four principles we can use to guide our action.

Be Specific

First, we must constantly work to shift the public discussion from the general to the specific, forcing ourselves into the proverbial weeds. Much of the national conversation surrounding lockdowns, for example, has operated at an unhelpful level of generality. Policies of this magnitude have large and broad-reaching impacts—medical and economic, of course, but also legal and social. They also are not implemented uniformly: they impact different geographic regions and different employment sectors in different ways. The reality is far more complex than a simple dualism: “lockdown good!” or “lockdown bad!”

Moreover, we cannot evaluate the appropriateness of a policy in isolation without a clearly defined alternative. The value of many “exit plans” that have been reported on is their specificity and detail—the pains they take to indicate fine points and to search for proactive strategies that can replace harsher restrictions without putting populations at an unacceptable level of risk. A similar framing for our discussions around school reopening is useful. The options we have do not fall neatly into two poles, distance learning versus a pre-pandemic return to schools. What does social distancing for elementary school students look like? Can some instruction be moved outside? How can indoor activities be conducted more safely? The act of laying out alternatives is instructive, as it gives us something concrete to debate.

Energy focused on arguing about problems is energy not spent on developing and implementing solutions.

In examining specifics, our strategies need not only be about reducing COVID-19 infections and deaths. They can also be about minimizing the broader impacts of the pandemic. Schools provide many critical services to our communities, for example—not just education, of course, but also meal programs. Where schools have closed, some areas have continued to make regular meals available for pick up, thus mitigating one of the consequences of this particular policy. Another example is the interruption of routine medical care like cancer screening and infant vaccination. As some patients are reluctant to visit hospitals for fear of infection, patients may prefer decentralized care at remote offices, or even home-based care. By breaking diffuse larger problems into specific smaller problems, a seemingly intractable problem becomes more manageable.

Compare Globally

As we continue to search for solutions, we would be wise to look globally. Each country, with its own unique circumstances, is tackling the same basic set of challenges. This generates a wealth of data points for us to study, lessons we can learn from alternative approaches. A shift from hospital-based to home-based care was an early lesson from northern Italy’s outbreak, for example, as described in a case study comparing hard-hit Lombardy with neighboring Veneto.

One element of this is considering the areas that are having the greatest success and trying to learn from them. Even if New Zealand has the epidemiological advantage of geographic isolation, what can we learn from the clear crisis communications employed by Prime Minister Jacinda Ardern? Likewise, what can we take away from the massive testing campaigns conducted in China? Following their 2015 Middle East Respiratory Syndrome (MERS) outbreak, South Korea dramatically improved their early detection surveillance systems and infection prevention protocols. In Japan, emphasis has been on reducing close-range conversations in crowded places (3Cs) and focused detection of large transmission clusters using teams of public health nurses. In South Africa, community healthcare workers have gone door-to-door to ask about symptoms and raise awareness. In Kerala, India, housing and meals have been provided to stranded migrant workers. Dedicated hotlines and apps in Vietnam have kept the public up to date on the latest information. The Democratic Republic of the Congo has repurposed contact tracing protocols originally developed for Ebola.

We must not give in to the pressure to oversimplify the conversation or rest content with false dichotomies.

Some have sensibly taken a systematic approach to learning from other countries. On the website Our World in Data, readers can study Exemplars in Global Health to learn from South Korea, Vietnam, and Germany about their successful COVID-19 responses. In April the UK’s Royal Society of scientists launched the DELVE (Data Evaluation and Learning for Viral Epidemics) initiative, whereby multidisciplinary teams gather evidence from other countries and synthesize these into detailed reports and policy recommendations. Their May report on test, trace, isolate compared policies across successful countries to identify commonalities. They highlighted the short turnaround times of testing results and the use of apps to supplement manual tracing by performing automated follow-up symptom checks. Importantly, for test trace isolate and for other policies, not all countries have pursued the same approach. This diversity of response demonstrates that the same playbook is not needed everywhere.

Think Local

Once we have identified strategies to pursue, we must think carefully about how to make them work in local conditions. Adaptation is critical; not all solutions will work equally well everywhere. Drive-thru testing sites may work well in Ohio, for example, but not in New York City. In neighborhoods with many undocumented immigrants, traced contacts may not want to share personal information. One-size-fits-all solutions do not exist, and top-down approaches can falter when unexpected problems arise in implementation. Mathematical models can help us anticipate how quickly labs need to turn around test results, how many contacts need to be traced, and how many need to adhere to quarantine, but they do not tell us how to reach this level of success.

To tailor strategies to the local context, we must look for solutions from within, those doing the most successful work in the field: the contact tracers no one hangs up on, or the nursing homes with no outbreaks. To figure out how these outliers have succeeded, the best approach involves engaging those on the frontlines of doing this work, here, contact tracers or nursing home staff, in discovering what works. When team members at all levels are engaged in the solution, it can yield the most creative ideas, and the team is invested in carrying these ideas forward. This can yield generalizable lessons, like improving caller ID so that more people answer calls from contact tracers, but for many other problems there will be no quick fixes. Contact tracing, in particular, relies heavily on the skills of interviewers to earn trust of patients so that they feel comfortable sharing important but sensitive information. A successful tracer may have a particular way of structuring the interview, or even use a particular tone of voice. Outside consultants can facilitate but not replace the learning process. Rather than being told what works, participants benefit from the discovery process itself—measuring performance, finding the outliers, and seeing firsthand what makes them more effective. This is what can inspire lasting change.

Build to Learn

Finally, we need to create an architecture for learning—a plan for gathering insights to inform our evidence-based decisions. Over the last several months many online tools have sprung up to help us monitor the data, including the COVID Tracking Project, COVID Act Now, COVID Exit Strategy, and Johns Hopkins Testing Insights Initiative, to name only a few. These reporting dashboards give us a way to assess at a glance what is going on across the country but can be undermined by missing data across the reported metrics. Statistics are generally provided at the state level rather than the preferred county or zip code level. Furthermore, they are rarely broken out by age and race or ethnicity, primarily because states have not adopted a standardized format for reporting. A simple example is how data are broken out into age categories. Where age ranges are defined differently—where one state lumps people in 18–34 year olds but another uses the range 18–49, say—they cannot be compared in a meaningful way. Without standardization that enables data to be combined, we are not getting full use of the data that our clinical and public health teams work hard to collect.

We should welcome even incremental improvements as moving us closer to this goal. And we should reject the wishful thinking that holds out for a simple fix such as a vaccine.

Contact tracing is another extraordinarily valuable learning tool. By identifying the likely source of infections, we have a data-driven way to evaluate the riskiness of different activities. For example, though we have evidence that transmission clusters are linked to nightclubs, churches, and Zumba classes, there remains uncertainty about movie theaters, subways, and airplanes. Finding a safe path forward involves identifying low-risk venues as much as about identifying high-risk venues. With businesses re-opening or implementing new mitigation measures, the landscape of settings we need to examine is constantly changing. We need greater investment in time and resources to collect and study complex tracing data to generate new and actionable insights, so that we can continue to learn as our lifestyles evolve.

Prospective cohort studies, where populations are carefully studied over time, can generate very high-quality evidence. As these studies take time to set up, we need to start planning now for the information we want to have in the future. This effort can pay dividends, though, in that their results are more likely to be definitive. At the start of the pandemic, some complained that randomized clinical trials of different COVID-19 treatments were too slow and that therapies should be approved on limited evidence. Still, those who forged ahead with randomized trials produced high-quality data that produced both wins (the steroid dexamethasone) and losses (the anti-malarial hydroxychloroquine). Not all policy questions can be addressed in a randomized trial, of course. While Norwegian researchers proposed a randomized study of school reopening, it was not approved by the government. But the act of describing what such a trial would look like— the alternatives being compared, the key outcomes to measure—moves the conversation forward. And where randomized studies may not be feasible, we must pursue a range of well-designed observational cohort studies as the next best option.

• • •

Unlike problem-based thinking, which looks back on—and often gets stuck in—the past, solution-based thinking looks forward to a safer and healthier future for all. It welcomes even incremental improvements as moving us closer to this goal. And it rejects the wishful thinking that holds out for a simple fix such as a vaccine. Scientists are working tirelessly to develop, test, and manufacture a safe form of immunization, but meanwhile there is so much more we can and must do to protect our communities. Increase testing capacity, eliminate bottlenecks, scale up contact tracing, increase mask wearing, improve public health messaging, and carefully study the data to inform targeted policies—all these have an essential and urgent role to play, even after months of lost opportunities. We must not give in to the pressure to oversimplify the conversation or rest content with false dichotomies.

We need more emphasis on what Devi Sridhar has called the “hard slog of public health.”

In the end, we need more emphasis on what Devi Sridhar, chair of global public health at the University of Edinburgh, has called the “hard slog of public health” and less on silver bullets, in part because waiting for some future panacea may deter us from learning how to do things better now. It is thanks to the public health slog, for example, that millions in Africa are now receiving lifesaving anti-retroviral therapy at thousands of community care sites in the aftermath of the explosive HIV epidemic in the early 2000s. As a result, for so many HIV is a chronic illness rather than a death sentence. For COVID-19, the countries that are doing the best right now worked hard to achieve their results, and it is because their communities feel safe that they are closest to the “normal” we desire. Solution-based thinking involves rolling up our sleeves to put in the hard work needed to get us there—and appreciating why this hard work on the small details is worthwhile.