Some states are now insisting on mandatory testing for the coronavirus in nursing homes and other senior-care facilities. New York state reportedly will require all nursing facilities to test staff twice per week, Maryland has mandated the testing of all nursing home residents and staff, and Massachusetts recently announced that long-term care facilities must perform baseline testing on 90 percent of residents and staff by May 25.
Unfortunately, mandating testing without the right logistics and a sensible strategy is unlikely to stabilize the situation in senior-care facilities. There are three problems: testing capacity, what exactly is being tested, and how test results are used.
In requiring testing for senior-care facilities, some states seem to be assuming that sufficient testing capacity is available and accessible. In fact, nothing could be further from the truth.
In our experience over the past month working on a pilot surveillance program run by the Massachusetts Senior Care Association, we have run into numerous binding constraints, including test kits, specimen collection (the process of going to nursing homes and swabbing for virus), and test processing.
This pilot is for only six long-term care facilities — a mere fraction of the facilities across the state. The idea that individual nursing homes can buy enough test kits, collect swabs, and secure processing capacity to cover their residents and staff in even a medium-sized state remains a fantasy. We have had to secure some swabs directly from Chinese manufacturing plants. Connections from our affiliations with Harvard and MIT made this possible. The average nursing home does not have such connections. If the mandate is to test every person, the state must ensure sufficient supplies. The burden cannot be placed on the individual facilities.
Similarly, the logistics of testing all nursing home staff are daunting. The idea that this could be done for all employees twice per week is unrealistic — unless and until the state either provides the test kits and personnel to collect specimen itself (even the National Guard is too small in most places) or allows appropriately qualified private emergency medical services to do this at scale. The pilot program was made possible by grants from Schmidt Futures and another foundation, and also received in-kind support from several other entities. Sustaining and scaling up such efforts is not possible without thoughtful state support.
On what exactly should be tested, there seems to be a widespread misunderstanding. The so-called gold standard polymerase chain reaction test is effective in determining whether a person is infected with COVID-19. But such virology tests are not enough. For example, if a nursing home has a 5 percent infection rate, does that mean that the disease has just broken into the building, or that COVID-19 has largely run its course among the residents?
Recently, 13 percent of residents in a local nursing home tested positive for COVID-19. But what does this mean? We also have a preexisting cross-sectional virology sample of this facility from three weeks earlier. At that time, over 35 percent of the residents were positive. So the notion of using a viral test alone for a mandatory single-time-point baseline survey does not make sense.
The best way to understand the situation within any community more precisely is to also conduct serology tests — these are blood tests that look for the presence of antibodies. Most people who have survived COVID-19 likely develop antibodies.
Antibodies are currently routinely measured with a significant error rate, but implications can still be drawn based on the extent to which antibodies are present in any communal living space (such as a senio-care facility). In the example above, where a single virological prevalence (current infections) was found to be 13 percent, the true fraction with antibodies (and therefore previously or still infected) was over 60 percent. Thus, the 13 percent positive rate suggests that the outbreak may be on the downswing.
As a tool of an organized surveillance program, serology tests are essential — but like the limitations of swabs for viral testing, it’s hard to get the necessary numbers of fingerprick blood collection cards suitable for surveillance. Lab capacity for these tests is also a major issue. Laboratories in Massachusetts aren’t set to report out serological results to individuals and into the public health information system. The development of high throughput facilities able to process thousands of these specimens per day is needed. Governors need to take this on as a priority.
On the impact of test results, COVID command centers need to do some more hard thinking. When a nursing home staff member tests positive, obviously they should not be allowed to come into contact with residents who are not COVID-19 positive. But who then will take care of the senior-care residents? There are already severe staff shortages, as a result of infections, quarantines, school closures, and staff fearful of coming in to work. Recognizing the issue, Maryland has created so-called bridge teams, and Massachusetts has offered clinical response teams and deployment of the National Guard. Unfortunately, the initially announced scale (260 registered nurses in Maryland and 120 nurses and certified nursing assistants for the clinical response teams in Massachusetts) is too small to make enough of a difference.
Governors need to step up and provide nursing homes with the tools to protect their staff and residents, rather than only mandating testing. State testing programs should: ensure adequate capacity and supplies are in place for polymerase chain reaction testing; add scaled-up serology testing as an important part of the monitoring toolkit; and provide bridge teams for facilities facing a staffing shortage as a result of COVID-19 testing.
Simon Johnson is cochair of the COVID-19 Policy Alliance and a professor at MIT’s Sloan School of Management. Michael Mina is assistant professor of epidemiology at Harvard T. H. Chan School of Public Health and a physician at Brigham and Women’s Hospital. Tess Cameron is a consultant and incoming principal at RA Capital. Eric Friedman is senior research scientist at the International Computer Science Institute.