The Coronavirus Hospital Data Debacle
Way back at the start of the pandemic, the state of Colorado was one of the very few that made a serious change to its reporting of deaths due to COVID. We began reporting how many people had died because of COVID versus with COVID. That was an essential public health decision that the rest of the nation generally didn’t follow, until now. Why and how has that problem distorted our public health response to this virus?
First, where are we on Omicron? As I reported a few weeks ago, this variant is MUCH less lethal than the original strain. For example, while cases are 235 percent higher than the last wave, deaths are only about 38% of that wave (1). So based on that data, about 1/7th as lethal. In addition, that data likely vastly overestimates the lethality of Omicron. Why? Because we have a serious hospital data reporting problem in a virus that’s more contagious which causes junk data reporting. Let’s dig in.
The Colorado COVID Death Correction
All public health responses require accurate data reporting. That’s not a statement that’s related to COVID, but the policy response for every disease since Smallpox and Measles. When that data is distorted, it’s tough to know how to deploy resources. More importantly, it’s almost impossible to know how to educate the public so that they can balance behavior modification with going about their lives.
That last part about educating the public is a BIG deal. In a public health sense, panic is not a net positive. Why? because it’s all about balancing the shut down of people’s lives with saving lives. Meaning, more shutdown equals more deaths due to things like missed serious diagnoses and deaths of despair (drug abuse, violence, alcoholism, suicide) so that must be balanced against the lives saved by isolating everyone in society.
In Colorado, our Governor realized this very early in the pandemic. After a news story highlighted several deaths labeled as due to COVID that were really traffic fatalities, he changed our COVID death certificate process by bifurcating it into “died with COVID” versus “died because of COVID” (2,3). The former is a death that should help morph public health policy, the latter, where someone died from a car crash but tested positive while in the hospital, should not.
The COVID Hospitalization Debacle
For our public health response to Omicron, we desperately need accurate hospitalization data. That’s because the early data out of South Africa from the discoverer of this variant was that its effects were less severe than Delta with few people being hospitalized. So are the COVID hospitalization numbers accurate? Nope.
In a recent report, we found out the following (4):
- At NYU Langone 65% of its COVID-19 patients were “incidentally” found to be infected after admission for other reasons. As a result, the NY governor asked hospitals to begin adjusting their reporting.
- 53% of 471 COVID-19 patients at Miami’s Jackson Health System were primarily admitted for other reasons
- At Johns Hopkins, that number was 20%
These numbers should startle and frankly upset you as well. Why? Because we know that hospitals have been testing every person admitted for COVID since the start of the pandemic. While some of this is because of their own public health response, some is also based on financial incentives. Why? At the start of the pandemic, Medicare placed a 20% COVID-19 payment kicker for people treated who had a COVID diagnosis (5). Hence, any hospital administrator worth their salt has been making sure that every admitted patient gets tested.
In response to this inaccurate data, the state of Massachusetts has also recently changed its policy on hospital reporting (6). That new, more accurate system begins today.
This Problem Is Now Magnified by Omicron’s Increased Infectivity
If we test every patient that walks through the door that’s admitted into the hospital and the ability of the virus to spread is increased while the lethality is decreased (i.e. Omicron), our data will be even more messed up. Why? Because the data is distorted by incidental positive tests and the percentage of people who will test positive without symptoms in the general population is higher for Omicron.
How Do We Fix This?
Colorado, New York, and Massachusetts Governors have demonstrated that this can be fixed. However, will the CDC step up to the plate and require hospitals to report those patients “with COVID” versus those admitted to treat COVID? Only time will tell.
The upshot? Omicron remains a MILD flu-like illness. Based on the data I’ve seen it’s MUCH less lethal than Delta. So get vaccinated, wash your hands, mask up where required, and live your lives, folks. There’s not much to see here.
(1) The Washinton Post. Four charts that analyze how omicron’s wave compares to previous coronavirus peaks. https://www.washingtonpost.com/health/interactive/2022/omicron-comparison-cases-deaths-hospitalizations/
(2) CBS 4 Denver. New COVID-19 Death Dispute: Colorado Coroner Says State Mischaracterized Death. https://denver.cbslocal.com/2020/05/14/coronavirus-montezuma-county-coroner-alcohol-poisoning-covid-death/ Accessed 5/16/20.
(3) Colorado Department of Public Health. Case Data. https://covid19.colorado.gov/data/case-data Accessed 5/16/20
(4) Beckers Hospital Review. Hospitals see more patients ‘with COVID-19’ vs. ‘for COVID-19’. https://www.beckershospitalreview.com/patient-safety-outcomes/hospitals-see-more-patients-with-covid-19-vs-for-covid-19.html
(5) US Department of Health and Human Services. CARES Act Provider Relief Fund. https://www.hhs.gov/coronavirus/cares-act-provider-relief-fund/index.html Accessed 5/5/20.
(6) Boston Globe. Mass. hospitals will begin reporting primary vs. incidental COVID-19 admissions on Monday, DPH says. https://www.bostonglobe.com/2022/01/06/nation/teachers-union-accuses-state-gross-incompetence-questions-swirl-around-masks-distributed-school-staff/