First let me say that I don’t know why your 2nd comment required my approval to be seen, so apologies for the delay in posting it.

Next, you should note that we’re arguing the same point but merely from different angles. While I believe you are correct in all that you say, my point for this site is to showcase that a deep dive into the facts is not required for understanding, merely a different dive, one with charts and numbers. Specifically some US commentators have cited the Lancet report and merely state that the US is #1 in cancer survival. But until you see the data that supports that (the bar charts here) the reader doesn’t really get a sense of what that means. Assuming the article is true, something that the lay public has to do, the article itself indicates that the US’s lead is not that substantial. When you reflect on this lead in terms of the dollars the US spends on health care, this slim lead hardly seems worth the effort.

Regarding some of your other points…
The registry data used in the US did include a breakout in terms of race. And, while there were regional differences, overall the fraction of blacks in the registries was about the US average.

You should note the original Lancet article included error bars for the various countries. It appears to me that these error bars are the standard statistical thing and do not represent any additional information about the quality of data in the registries. But when the data with the error bars is included it’s easy to see that the US is merely tied for first place with about 4-5 other countries (something I mentioned in my article).

You correctly point out a feature that I cut out of my original draft of this article. Namely how much of the US ranking here is due to social and cultural features and not directly tied to the health care system? Your example of the Scots drinking and smoking themselves to death and the US possibly having a higher screening rate are things that are not going to change under the proposed bills currently being considered in the US.

Finally, thanks for the pointers to the additional articles. I am interested in revisiting this topic in a future article and will undoubtedly use the articles you cite when I do so.

http://www.eurocare.it/Results/tabid/79/Default.aspx

you can see that the UK is the only large population country that is a 100% registry reporter, and is pretty brutally honest about the stats too.

I’m not suggesting that the US is not ahead for a large section of its population, thanks to aggressive screening and top notch comprehensive cancer centres. The Lancet paper makes this clear. But the size of the effect is likely to be smaller than we can see now. A few factors being:

– lag time – a lot has happened in the laggard countries (eg the UK) to improve cancer care in the 2000s. This registry data we are looking at is mostly from the 1990s
- health inequalities are particularly acute in the US and as data from the American Cancer Society makes clear, blacks, Hispanics and the uninsured suffer worse outcomes. They may be underrepresented in the registry data (although there are also biological factors in play too). The UK also has a steep inequality curve too, but it is counting everyone so there’s nowhere to hide
- if we were to count in 100% from Germany, Netherlands and France these countries may well beat out the US (Germany only has 1% in Eurocare!).

And overall a key point: from what I can see, much of the difference between the US and elsewhere is down to access and use of diagnostics, not treatment. That of course is part of a healthcare system but is also a cultural factor. It’s well known that the Scots, for example, drink and smoke themselves to early graves and avoid doctors but there are superb hospitals in Scotland. So the claims for the overall superiority of American healthcare – if we take into account all the outlying hospitals and the undiagnosed – and not thinking it’s all MD Anderson and Sloan Kettering – may be overinflated.

This article on Eurocare says of the US:

‘European patients need to know that there is no particular reason to think that cancer treatment in the US is better than can be obtained in Europe. It is also important to stress that in both Europe and the US there are large survival differences between the rich and poor. Also, the survival differences between European populations for all cancers combined would decrease after exclusion of prostate cancer, and, to a lesser extent, breast cancer, whose survival is artificially increased by lead time due to screening…’

http://www.istge.it/fp_acc/index_file/Comparative%20cancer.pdf

M.

The original Lancet graphs included more information than my version of them. Theirs identified the countries that had 100% coverage from those that didn’t, and as you say the US had only 42% coverage. The other countries with less than 100% coverage were Canada, Japan, France, Italy, Spain, Netherlands, Switzerland, Germany, Austria, Portugal, Poland, Czech Republic, Brazil and Algeria, about half of the countries.

I didn’t see any reference in the article to the US portion being more or less skewed toward those with insurance. (I suspect you may be referring to articles I haven’t seen.) But regarding data coverage, there was another whole section of the article devoted to specific cities and states in the US where close to 100% of the registries were used. For example, the US fraction included Atlanta but not Georgia, all of Colorado and all of California with additional detail for Los Angeles and San Francisco, just to name a few. In contrast, the article did mention that the portion of Italy studied was a more affluent part than the country as a whole and it mentioned Cuba as having some quality control issues with their registries (and for that reason I excluded Cuba from my version of their graphs).

The article did mention the race variation you refer to above. However either it or other articles I researched also pointed out that the US is much more heterogeneous than the average European country. While this may weigh more favorably for the US (we’re #1 despite having a more heterogeneous population), I chose to present the original Lancet article data as is. After all, there are only 2 ways to account for this, either compare countries as a whole (what was done) or compare the predominant race in each country to that of other countries (eg, Caucasian US to Caucasian France). Fortunately data wasn’t available for the latter comparison and besides it seems off-topic to consider race specific data when trying to set policy for an entire country. (I’m not suggesting you were implying this, but you raised race so I needed to give all of my reasoning behind ignoring racial differences.)