Even best medical reporting infected with “make do” bias

And Neugut said “even among gastroenterologists there’s variability in quality,” likely accounted for in part by the fact that those who have the opportunity to perform the highest volume of procedures tend to have their skills honed more.

Dr. Durado Brooks, the director of Prostate and Colorectal Cancers at the American Cancer Society, concurred, telling Remapping Debate that, “A lot of this has to do with…practice and technique and developing good skills.” He added that some practitioners simply start out and remain more proficient than others.

One way to approach the question of potential relative benefit would have been to ask, “How good can this procedure be in the hands of a true expert?” and then follow that up by asking, “How can we increase the skill level of more practitioners so that more people are getting the highest quality colonoscopy possible?”

That wasn’t the approach the Rosenthal article took, and that approach is anathema to the health care cost-cutters. If citizens are given a moment to think about potential benefits, there is a risk that they may say, “Perhaps we want to raise the standard of care that our society provides.”

Conflation, conflation, conflation

A particularly interesting argument regarding the relative merits of screening procedures has to do with “adherence.” As Baxter put it to us: “Screening is only effective if you actually will get it. And although when you talk to gastroenterologists, and surgeons, and people who’ve had colon cancer, they all promote colonoscopy, in fact many people don’t want it.”

But simply accepting what people currently “don’t want” is not the only way to go; here, again, there is a choice to be made. Instead of assessing the effectiveness of colonoscopy based on where people are now, one could imagine potential effectiveness if there were greater adherence, and public health policy could be shaped to help people overcome inhibitions about undergoing the procedure that are unrelated to actual risk.

As Dominitz pointed out, there is a need for “programmatic screening,” where it’s not just the good fortune of your doctor remembering to suggest a procedure. We have the ability to “build systems” to ensure those reminders, but we have to resolve to do so.

On another point, Dominitz suggested a circumstance where annual FOBT could detect a fast-growing cancer, whereas reliance on a colonoscopy not scheduled for another couple of years might mean that the cancer wouldn’t be detected in time.

Leaving aside the question of the relative infrequency of that scenario, his presentation was an important illustration of the tendency to set things up in an either-or fashion. Perhaps, instead, it would be useful to think about taking advantage — both on the individual level and the population level — of both FOBT and colonoscopy. But the health care world — ruled as it seems to be by health care economists — won’t do so, and this certainly wasn’t a type of approach or strategy considered by the Rosenthal article. Why? I think anything that suggests the prospect of greater cost makes that thing a possibility that dare not speak its name.

Indeed, the god before whom almost all health care policy currently bows is the god of cost-effectiveness. Baxter herself said that the question is the extent to which colonoscopy will be more effective and “does that make it worth the extra cost and the invasiveness of the procedure.” (Johnson said that, presuming a qualified expert performs the procedure, the relative risks associated with colonoscopy are small, in particular in relation to complications like perforation. Other experts we spoke to agreed that colonoscopy risks are relatively low.)

Cost-benefit can be an appropriate question, it’s just not the first question, and it shouldn’t be the case that a society can’t choose to assign a very high value to each life saved.

Bringing some nuance to the cost data

The interactive graphic used in the online version of the article, with the tag “The Cost of a Colonoscopy Varies Across the Country,” is startling: $8,577 in New York, for example.

(The fine print shows that the amounts displayed represent the highest amount paid in a metropolitan area, based on an analysis by Healthcare Blue Book. The print version of the article had the high price in big print and the low price in small print, with an explanation that outliers — those below the 5th percentile and those above the 90th percentile — were excluded.)

But Remapping Debate contacted Dr. Jeffrey J. Rice, chief executive officer of Healthcare Blue Book, to explore these data further. As requested, Rice forwarded what he said was a representative sample of data from three areas — Atlanta, New York, and Seattle — that were included in the Rosenthal story, explaining that the data reflected costs for commercially-insured patients (not those on Medicaid or Medicare).

We derived median prices (the 50th percentile), and the numbers don’t shout so loudly: New York came in at $2,083; Atlanta at $1,864; and Seattle at $1,542. As the chart to the right shows, there is certainly significant price variation, but the spread between the 25th and 75th percentiles — that middle 50 percent of procedures — is less than the article would lead one to believe.

None of this negates the concerns that Rosenthal identified about the market failure (and concomitant failure to regulate) that makes health care unnecessarily expensive, but the data should have been used in a way that represents better the mainstream of the cost of care.