Those low UK breast cancer survival rates: Will conservative proposals for reform lead the US to the UK’s problems?
I am not a cancer specialist or epidemiologist, though I have worked on a few cancer prevention and screening projects. But I have not found anything in the US press that examines the issue of evaluating a whole nation’s health care system on the basis of breast cancer. I have not seen any articles examining what the implications of such a comparison would be for the health reform debate in the US. I am a health economist, for whatever that is worth (not much apparently, from what I have seen of their influence on the debate). In any case, I was curious, so I decided to look into it. This post is a report of my little investigation. If anyone thinks I have missed something, or an expert in the field happens by, please leave a comment.
But before that, let us debunk the whole idea of judging countries’ health care systems based on outcomes for just one disease. It is dishonest cherry picking, and sound-bites based on it deserve no place in an informed debate.
I think breast and prostate cancer were chosen to demonstrate the ‘absolute bestness’ of the US healthcare system in every respect by looking up recent studies (for example, the CONCORD study (1) which looked at six different cancer/sex combinations), picking the one or two where the US performed best, and then picking another country that did not perform so well to use as a bogeyman. That is a clearly biased and dishonest procedure. The results for breast cancer in the CONCORD study do not say that the financing of the health system is crucial in determining survival rates for common cancer, since Canada, France, Japan, the US, and perhaps even Cuba performed best in one category or another (though, the results for Cuba are uncertain because some of its data were incomplete).
A recent comparison of relative performance of health care systems including Australia, Canada, New Zealand, the UK and US evaluated them on survival rates for breast, cervical, and colorectal cancers, childhood leukemia and non-Hodgkins lymphoma. (2) The study found that no one country was the best at all of them. Australia, Canada, New Zealand and the US were all about equal on average. The UK clearly lagged in cancer care, overall. Out of the four non-US countries, the UK is clearly the least relevant model for health care reform here in the US. The dishonest cherry picking for this misinformation campaign about breast cancer clearly includes the choice of comparison country as well.
Finally, even if more appropriate comparisons were made, these sound bites misrepresent the statistical analysis, and give a false impression of dramatic differences where there may be none. The CONCORD Study (1) found that the UK had lower survival rates. Below are some of the top countries, and England (the UK was divided into England, Northern Ireland, Scotland and Wales, but all their results are similar, so England is used for simplicity).
Five year relative breast cancer survival rates from CONCORD study.
Rank….Country….mean (95% confidence interval)
1……….Cuba……….84.0 (82.9, 85.2)
2……….US………….83.9 (83.7, 84.1)
3……….Canada……82.5 (81.9, 83.0)
4………Sweden……82.0 (81.2, 82.7)
5………Japan……….81.6 (79.5, 83.5)
6………Australia.. ..80.7 (80.1, 81.3)
8………France……..79.8 (78.2, 81.4)
22…….England……69.8 (69.5, 70.2)
First thing to note here is that the differences between the top countries are very small. The CONCORD study report warns that the ranking of the countries maybe unstable. There are differences between these countries that cannot be controlled for, and aspects of statistical analysis that are imprecise. While the CONCORD study defends its methods and says that the resulting biases are small, it also warns that when the differences between countries are small, then a small bias might make a big difference in the ranking. So, the bottom line is that from these data and the warnings provided by the study itself, we do not really know whether Cuba, the US, Canada, Sweden or Japan is ‘the best’ at treating breast cancer.
Another thing the CONCORD study notes is that the results for the US may be inflated. The problem is that only a part of the population is covered by the cancer registries used for the US analysis, and there is evidence that these registries have higher survival rates than the population covered by other registries and the population as a whole. By contrast, the UK cancer registries used in the analysis cover 100% of the population.
Now we look at the process of treating breast cancer. My main source is a report examining the relative efficiency and productivity of the German, UK and US health care systems (3), which covers the practice during the period of screening and diagnosis from the mid 1980s to mid 1990s. Successful treatment of breast cancer is complex, and can be broken into several components.
First, early detection is crucial for high survival rates. There are several common methods for population screening: mammography, physical exam conducted by a health professional, and breast-self exam (physical exam conducted by the women herself).
Second, there must be some social and medical professional consensus for a national screening program, which includes age at which screening should be begin, and definition of “high risk’ for people who should consider earlier or more intensive screening.
Third, any suspicious symptoms must be clinically diagnosed. From the late 1980s through the mid 1990s, the McKinsey report provided a simplified description of the main protocols:
1. Inpatient Surgical biopsy and treatment, one-step: Patient given treatment options and gives prior consent on treatment should the diagnosis be positive, then put under anesthesia for general surgery. Diagnosis made and treatment provided during operation,
2. Inpatient surgical biopsy, two-step: Surgical biopsy under general anesthesia. Patient informed of results and makes decision on course of treatment. Treatment provided in second operation,
3. Outpatient surgical biopsy, two-step: Same as 2, except biopsy is performed in outpatient setting,
4. Fine needle aspiration (FNA), two step: Same as 3, except outpatient biopsy performed on cell sample extracted from suspicious tissue using a hollow needle. Restricted to suspicious tissue large enough to be targeted with a needle. Surgical procedure, if necessary, follows in step two.
Fourth, what kind of surgical treatment, if advisable, is used. The surgery may be either a mastectomy (removal of whole breast) or a Breast Conserving Surgery (BCS) that focuses on the offending tissue, and leaves as much of the breast as possible. During surgery, other decisions must be made, such as how to examine the lymph system draining the chest area, and how much of the lymph system to remove, should cancer be found there. The decisions here involve important quality of life issues for the rest of the patient’s life. Radical mastectomy removes at least some of the pectoral muscles, so far more than just the breast is involved. Disruption of the lymph system can cause serious and painful impairment. (4, 5)
Fifth, follow-up treatment and surveillance to reduce chance of fatal recurrence or metastasis. The treatment can include radiotherapy, chemotherapy, or hormonal therapy, or some combination.
Lower survival rates might be due to problems in any one of these steps, or some mismatch between the different steps.
A very brief summary of the McKinsey reports’ conclusions would be that the most important difference between the US and the UK was that the UK moved to a broad population screening program later than the US. It was far more narrowly targeted at specific high risk people than in the US. As a result, breast cancer cases were treated, on average, at a more advanced stage in the UK than in the US.
The UK moved more quickly to a more efficient and productive, and newer, approach to diagnosis (FNA two-step) compared to methods used in the US. The UK moved more quickly to common use of the newer BCS (versus mastectomy) during the time period. Masctomy was still strongly favored in the US. Several reasons for this difference were noted: patient preference for BCS in the UK, reimbursement practices that favored the US approach to clinical diagnosis and mastectomy in the US; fear of malpractice in the US; and better fit with ability of UK to provide appropriate follow-up care for BCS.
The McKinsey study’s conclusion about the importance of delayed diagnosis was recently reaffirmed during conferences that discussed the policy implications for the UK after it was found to have poorer performance than average compared to other European countries in another study (6):
The clear consensus was that the reported differences between countries were, to a large extent, real. Furthermore, findings from the high-resolution studies indicated that the poor results from the UK were attributable mainly to patients having more advanced disease at diagnosis than patients in other European countries. For policymakers, this conclusion is clearly of great importance, because it indicates that particular emphasis should be put on achieving earlier diagnosis.
The UK recently expanded its surveillance program for early detection and for those diagnosed in 2000-1 the five year survival through 2005-6 just exceeds 80%. (8) More than an third of the improvement has been attributed to earlier detection of breast cancer alone: through both expanded formal population surveillance programs, and initiatives to speed individual cases to clinical diagnosis. (9). This recent higher UK survival rate still trails that of the US, which is approximately 90% through 2005-6 (10), but the gap has closed significantly.
What are the lessons for the health reform debate in the US? Generally, that the impression left by sound bites based on a couple of cherry picked statistics may be very misleading and drive the US backward rather than forward in the quality and efficiency of care.
One misleading impression here is that treatment after a cancer diagnosis in the current US system is uniquely wonderful, much better than anywhere else. But as we have seen above, if you look at a range of cancers, several British style countries perform about as well as the US: Australia, Canada, and New Zealand, all with universal care.
Another misleading impression is the implication that the US is always the forerunner in adopting medical and surgical therapy. Everyone else is behind the times. But the UK was quicker to adopt what the McKinsey report finds to be a more efficient and productive method of diagnosis, and also was quicker to move from mastectomy to BCS.
The final lesson is that sound bite scare tactics that lead to misunderstandings about the reason for differences between countries can produce unintended consequences. One of the most popular reforms advocated by conservatives in the US is the move to more market oriented ‘consumer driven’ care that makes the patient feel every test and procedure and drug in the pocketbook. The idea is that consumers should bear more of the cost of care, and therefore they will be more careful and parsimonious about demanding “unneeded” tests, exams, and procedures.
So, the question is, what would be the consequences of this policy for breast cancer survival rates in the US if consumer drive care resulted in higher costs and more hurdles for breast cancer screening? Are the consumer driven care advocates saying that comprehensive care policies should not cover what some may deem to be wasteful preventive practices such as widespread and early breast cancer screening in the US?
Or, are we to assume that government provided public health programs would pick up important areas where the consumers’ perceived need may not conform to best standard of care? But isn’t wasteful government subsidy supposed to be the reason for our high costs in the first place? Or do they worry about it at all?
It seems dangerous to me to accuse a country (the UK) of providing callous and substandard care, and then advocate a policy that would seem to lead directly to the problems encountered by that very country (the UK).
This concern applies broadly to all preventive care that concerns dangerous disease that develop without symptoms, or easily missed symptoms. The UK ran into trouble not because it had no screening program at all, but because it had one that was not aggressive and widespread. Many women outside of the official high risk categories could get screening and diagnostic services, but there was no formal program to facilitate it: they had to go to an extra, and sometimes considerable, effort to obtain those services. It would seem to me that the results of the UK breast cancer screening policy in the 1980s and early 1990s and a US policy that increased the cost of preventive care, would lead to the same thing: delayed diagnosis and care for breast cancer.
To repeat, it seems extremely irresponsible (and either ignorant or malicious) to me, to use the UK as a whipping boy for the supposed evils of healthcare reform, and then advocate policies that may lead to the same results as recently experienced in the UK.
(1) Coleman MP, Quaresma M, Berrino F et al. Cancer survival in five continents: a worldwide population-based study (CONCORD). Lancet Oncology 2008; 9: 730–56.
(2) Peter S. Hussey Gerard F. Anderson, Robin Osborn, Colin Feek, Vvienne McLaughlin, John Millar, and Arnold Epstein (2004) How does the quality of care compare in five countries? Health Affairs 23(3): 89-99
(3) Health Care Productivity. 1996, McKinsey Global Institute, Los Angeles, CA
(4) Oshumi S, Shimozuma K, Kuroi K et al. Quality of life of breast cancer patients and types of surgery for breast cancer – current status and unresolved issues. Breast Cancer Breast Cancer 2007;14:66-73.
(5) Cody HS. Current surgical management of breast cancer. Curr Opin Obstet Gynecol 2002;14:45-52.
(6) Verdecchio A, Francisco S, Brenner H et al. Recent cancer survival in Europe: a 2000–02 period analysis of EUROCARE-4 data. Lance Oncology September 2007; 8: 784-796.
(7) Richards M. Reflection and Reaction: EUROCARE-4 studies bring new data on cancer survival. Lancet Oncology September 2007; 8: 752-3.
(8) Quinn M, Copper N, Rachet B et al. Survival from cancer of the breast in women in England and Wales up to 2001, British Journal of Cancer 2008; 99: S53-S55.
(9) Leary A and Smith I. Survival from breast cancer in England and Wales up to 2001. British Journal of Cancer 2008;99, S56 – S58.
(10) SEER Cancer Statistics Review 1975-2006. Section 4. Breast.
Below is the official citation it says to use:
Horner MJ, Ries LAG, Krapcho M, Neyman N, Aminou R, Howlader N, Altekruse SF, Feuer EJ, Huang L, Mariotto A, Miller BA, Lewis DR, Eisner MP, Stinchcomb DG, Edwards BK (eds). SEER Cancer Statistics Review, 1975-2006, National Cancer Institute. Bethesda, MD, http://seer.cancer.gov/csr/1975_2006/, based on November 2008 SEER data submission, posted to the SEER web site, 2009.