As the Medical Director of the Kansas Business Group on Health I’m sometimes asked to weigh in on topics that might affect employers or employees. This is a reprint of a blog post from KBGH:

In a perspective piece in this week’s New England Journal of Medicine, Dr. Marc Larochelle proposes a three-component strategy for returning to work: 1) a framework for counseling patients about the risks posed by continuing to work, 2) urgent policy changes to ensure financial protections for people who are kept out of work, and 3) a data-driven plan for safe re-entry into the workforce.

Let’s go through his framework for work-related risk first. He summarizes it with this diagram:

The occupational risk on the vertical y-axis is defined by OSHA standards. The horizontal x-axis is based on age and the presence of high-risk chronic conditions identified by the CDC, like diabetes and heart disease.

How this could work

Let’s consider a couple hypothetical employees to see how this rubric might work: “Matt” is a 65 year-old man with no chronic medical problems who takes no medications other than an occasional ibuprofen for joint pain. By the CDC risk stratification rubric, then, he is at high-risk based on his age alone. Matt works as a radiology technician with no direct patient contact, but he is within six feet of patients with confirmed or suspected SARS-CoV-2 infections daily. By the OSHA standard, then, he is at high, but not “very high,” risk. Regardless, by the proposed Larochelle risk stratification above, Matt would be category “C” and should not go back to work.

“Shelley” is a 25 year-old woman with a history of type 1 diabetes complicated by neuropathy and kidney disease. So she is medium-risk in spite of her complex medical history, in spite of her young age. She works in a retail setting in Wichita, a city with currently relatively modest levels of community transmission. Therefore her occupational risk would be considered “medium.” This would put her in category “A,” in which Dr. Larochelle recommends that she be able to return to work, albeit likely with fastidious use of personal protective equipment (PPE).

But beyond a strategy to determine the safety of the workplace, we owe it to people of elevated risk, even those working in risky jobs, to work on two additional goals: first, we should conceive strategies to allow people to live a dignified life without hardship while away from work. This is largely a policy position we can support through elected officials at the state and federal level.

Second, we owe it to high-risk people to develop strategies to allow them to eventually return to their jobs, preferably even before the risk of those jobs drops due to decreased community prevalence of the virus. In a widely read piece in The Atlantic a couple weeks ago, Dr. Julia Marcus analogized our current predicament to the HIV/AIDS epidemic of the 1980s and 1990s. It would have been easy, in theory, to stop HIV cold in its tracks: people just needed to stop having sex. But people like having sex, just like they like going to work and eating at restaurants and watching baseball. So public health officials were forced to come up with alternative, innovative strategies like promoting condom use.

Similarly, COVID-19 could be stopped cold, much as Mongolia has accomplished, by instituting strict limits on social interactions. But we’re at a point of quarantine fatigue in which further efforts at social distancing in the immediate future are likely to be met with resistance. And a vaccine is months, if not years, away. This is where testing comes in, especially in regard to risk stratifying people for return to work and social interaction.

The role of testing

In listening or watching the news on testing, you are likely to think that risk is binary: a positive swab test means you have COVID-19, and a negative test means you don’t. Likewise, one might believe, a positive IgG antibody test means that you’re immune to COVID-19, and a negative antibody test means you’re still at risk. But neither of these assertions are true. The “positive predictive” value of a test, meaning the likelihood of a positive test predicting the presence of an actual disease state, depends on the “pre-test probability.” So a person who lives in a community with low prevalence of COVID-19 and has had no known exposure to someone with the disease is unlikely to have immunity, regardless of what her antibody test says. Even with a test that is 90% sensitive and 95% specific, that person likely has only about a 25% chance of having immunity.

This doesn’t only apply to infectious disease testing, by the way, and testing for other conditions has real implications for your employees. It is popular for doctors to routinely check the thyroid blood tests of patients as part of routine medical testing. It is not uncommon for mildly abnormal results of such testing to result in the patient being put on thyroid hormone for life. But mildly abnormal thyroid blood testing in someone who feels well and has no physical signs of thyroid disease does not mean that person has a thyroid problem. It only means the person has about a one in three chance of having a thyroid problem. As with COVID-19 antibody testing, the initial abnormal test result should prompt additional evaluation, not a definitive diagnosis.

So what do we do with the results of COVID-19 antibody testing? CDC suggests that we use them to “risk stratify” people on a population level, not as a marker to indicate safe return to the workplace.

Instead, we should aggressively use nasal testing for the virus itself to determine the status of people with exposure to persons with known or suspected COVID-19, much as we’ve discussed in the past.