Health disparities research is full of low quality work

It often does a disservice to marginalized groups and often merely furthers the careers of mediocre academics

Recently, I saw a tweet from World Lung Conference. A presenter lamented differences in Lung cancer screening rates by race. Black and Hispanic patients had lower rates of lung cancer CT screening than whites, and the speaker argued that we needed targeted efforts to improve this disparity.

But lung cancer screening doesn't work, as I described in a prior post.

It does not improve all cause mortality. No one has shown an improvement in lung cancer morbidity. The follow up to NLST failed to sustain the initial, spurious OS gain, the Nelson trial was null for OS or even all cancer death. CT screening for lung cancer failed in pooled analysis of all the available studies.

Why should Black and Hispanics be coerced into participation in a screening program that is itself unproven?

This is the sad state of disparities research.

Researchers want to correct a *difference* in health utilization, but they are too ignorant to understand the evidence for that intervention. They assume it must be good, and ask why some racial groups are being cheated out of it, rather than grapple with the fact that whites are probably disproportionately the victim of harmful medicalization. The same is true of papers lamenting booster uptake in young men, avastin use in colorectal cancer patients, and Paxlovid use in vaccinated people.

Disparities research in medicine often feels hopeless. Overly concerned with generating low quality papers, alternating between obvious and uninspired conclusions, and outlandish and irreproducible ones.

In the latter category comes many papers claiming that Black doctors have half the rate of death as white doctors when they care for neonates in Florida. This is a preposterous effect size and the methods are poor. Papers that show that as the fraction of PCPs who are Black increases in an area, black mortality is reduced massively. I tore apart both these papers on my substack and in these pages. It's important to have Black doctors. It's almost important to have truthful research.

Recently I had the pleasure of listening to Roland Fryer on Econtalk. Fryer is an economist at Harvard, who does robust, high quality work on racial discrimination. In the talk, he outlines the difference between disparity and discrimination. Noting that the former is a difference in outcome, and the latter is treating someone differently based solely on consideration of race. He outlined statistical methods to separate the two, and stressed why it was important to have an idea of the root cause of the issue. You can't fix what you don't understand.

Ultimately he argued that the modern focus of DEI education, which has dominated major corporations and universities, is largely comprised of ineffective programs. To truly address the real systematic failures in America, we have to leverage data. We can't be lazy. Fryer offers one provocative solution where Ivy league institutions invest small portions of their endowment to bolster specific high schools that cater to underrepresented minorities. Instead of this investment, schools prefer to pursue changes in admission policies that do not deal with the root issue. Some of these policies were recently struck down by the Supreme Court. He considers the lack of creativity of universities to be emblematic of laziness and indifference to the problem of discrimination.

His words could not be more true in medicine. Entire careers and labs claim to care about health disparities, but they don't want to use the best methods to ascertain the modifiable causes. They seem indifferent to using randomization to test what actually works. Instead, their work product merely serves to further the careers of mediocre academics, performing irreproducible, low quality and often unhelpful research.

In one paper (now on preprint), we found academics in medicine studying disparities were 3 times more likely to study their own career promotion than issues that pertain to young children. Of course both matter, but the relative prioritization reveals a narcissism among investigators.

https://www.medrxiv.org/content/10.1101/2023.03.05.23286652v2

What can be done?

If you are in training and you care about these issues, do not work with investigators who report obvious findings, or have weak causal methods. You will be much better served to work in a different field entirely, and develop stronger methods. Or work with Roland Fryer. Later you can bring these methods over to this space of health disparities.

If you are lamenting that a medical service has poor uptake among some racial groups, do a deep dive into whether that service actually improves outcomes or not. It's hard to complain about CT screening for lung cancer when the data are so poor. Same is true for many other interventions.

Universities will continue to promote researchers doing weak disparities work— for e.g. claiming that Black doctors halve the rate of death among black neonates (an effect size that can't be true) using data that does not actually establish which doctor provided the most care for the individual patient— because university administrators are lazy. It is harder for universities to actually reward scholarship that works to correct the fundamental disparities in society than papers that make the press for sensational and irreproducible claims.

Ironically bad health disparities research will do the very thing the researchers claim to be working against. It will further polarize people and create a backlash. It delays racial justice, rather than promotes it. It makes it harder for the poor black kid in Baltimore to succeed, rather than easier. We have forgotten entirely about entrenched patterns of discrimination and power, and do nothing to address it. Instead opportunistic researchers chase promotion to Associate Professor with terrible papers, and disparity becomes a buzzword for grant applications. The whole affair is sad for anyone who can see it for what it is.

I feel sorry for the poor minority child in America born today whose purported advocates are too incapable or narcissist to do the work that might help her. Our field fails her every day.