Hospitalist Perspective: Rethinking Our Hospitalization Criteria

At multiple hospitals where I practice, I’ve seen the same thing. Lots of patients with COVID-19, but many fewer patients with all of the usual hospital medicine problems we manage – heart attacks, strokes, diverticulitis, syncope, rapid A-Fib, and so forth. Why are we seeing fewer of these patients? 

Best guess is that people are staying home and managing as an outpatient to avoid the risk of exposure to COVID-19 at the hospital. The real question then is, do these people need to be hospitalized? Would their outcomes be better with routine hospital care? Many of them appear to be having higher mortality by staying home. If we look at the CDC website for national numbers, or at The New York Times website for breakdown by state, we can see that there are excess deaths starting at the end of March. These excess deaths are beyond those known to be from COVID-19. For example, in New York City, between 3/15 and 5/2, there have been 23,000 excess deaths, but only 18,706 are attributable to COVID-19. This leaves another 4,294 excess deaths that are NOT attributable to COVID-19.

Where are these excess deaths coming from? We can make a guess – maybe people with chest pain from a heart attack are staying home and therefore dying without receiving medical attention. These people would surely have had better outcomes with hospitalization. But among these excess deaths, there may be a population of people with some of those problems mentioned above (e.g. diverticulitis) that didn’t come to the hospital, and did just fine with outpatient treatment. 

This is an area that requires significant research over the next 12-18 months. Hopefully some enterprising data scientists and epidemiologists are going to look at the data from this time period and break it down by diagnosis. We can then look at changes in hospitalization by diagnosis, and then by mortality the same way. This may provide us a window into the mortality associated with not hospitalizing these patients, which could revolutionize our approach. 

For example, when I first finished training, patients with deep venous thrombosis (DVT) or pulmonary embolism (PE) were always brought into the hospital to initiate anti-coagulation, and only sent home after several days. With the rollout of new medications, we were able to start sending patients with DVT home from the Emergency Department with appropriate outpatient follow-up. We would still admit the patients with PE though. Now, with more time and data, we no longer even admit the patients with PE unless there is some other reason for them to be high risk and require hospitalization. My hope is that, with good data analysis, we may find more diagnoses that we can safely manage as outpatients rather than requiring hospitalizations and all of the risk that entails.