© 1999 by Oxford University Press
Journal of the National Cancer Institute, Vol. 91, No. 9, 802-803,
May 5, 1999
© 1999 Oxford University Press
CORRESPONDENCE |
RESPONSE: Re: Race, Prostate Cancer Survival, and Membership in a Large Health Maintenance Organization
Affiliations of authors: A. S. Robbins, A. S. Whittemore, Division of Epidemiology, Department of Health Research and Policy, Stanford University School of Medicine, CA; S. K. Van Den Eeden, Division of Research, Kaiser Permanente Medical Care Program (Northern California Region), Oakland, CA.
Correspondence to: Anthony S. Robbins, M.D., Division of Epidemiology, Department of Health Research and Policy, HRP Redwood Bldg., Rm. T223, Stanford University School of Medicine, Stanford, CA 94305 (e-mail: robbinsA{at}leland.stanford.edu).
We appreciate the interest of Roach et al. in our analysis of survival among black and white prostate cancer patients in the San Francisco Bay Area (1). The investigators have raised important issues, and we welcome the opportunity to respond to them. They argue against biologic factors as an explanation for poorer stage-specific survival of black prostate cancer patients compared with white patients. Instead, they argue that the observed higher death rates in blacks can be explained completely by racial differences in socioeconomic barriers, such as inadequate access to health care, to good treatment, and to health education leading to earlier diagnosis. We would like to believe this argument because it means that preventive strategies could eliminate the observed survival differences. Unfortunately, however, it is not supported by the data we analyzed.
The hypothesis that the black survival disadvantage can be explained completely by socioeconomic barriers predicts that the disadvantage is decreased when some of these barriers are removed. Since financial access to medical care is an important barrier that is eliminated in a health maintenance organization (HMO) such as the Kaiser Permanente Medical Care Program (KPMCP), the hypothesis predicts a smaller survival disadvantage for black KPMCP members than for black nonmembers. In contrast, the data showed similar stage-adjusted black/white death rate ratios of 1.28 for KPMCP members and 1.22 for KPMCP nonmembers. Thus, the data do not support the hypothesis.
Roach et al. fault us for not adjusting our analysis for tumor grade. Requiring that two patients (one black and one white) have tumors with the same grade puts them on artificially equal footing with respect to an important indicator of tumor aggressiveness. If black men have poorer survival because their tumors are more aggressive, adjusting for grade will mislead us into believing that eliminating all socioeconomic barriers will eliminate all survival differences. The question we raised was "Can the poorer stage-specific survival of blacks compared to whites be explained by differences in access to medical care, or is it due at least in part to biologic differences in tumor characteristics"?
To support their argument that black versus white survival differences are unrelated to black versus white tumor characteristics, Roach et al. cite data from KPMCP prostate cancer patients showing no black versus white differences in prediagnostic prostate-specific antigen measurements (2). However, these measurements were taken many years before diagnosis and so do not pertain to tumor characteristics or survival. In contrast, Roach et al. fail to cite data from an equal health care setting showing substantial differences in tumor features between black and white patients (3). They also fail to cite autopsy data suggesting that prostate cancers in blacks are more aggressive than those in whites of the same age (4,5). It is difficult to explain the autopsy findings on the basis of racial differences in socioeconomic factors.
Roach et al. criticize us for failing to comment on randomized trials that show similar survival for blacks and whites, citing a paper by Bolla et al. (6). It is not clear why this paper is relevant, because it does not mention the participants' ethnicities. Roach et al. argue that randomized trials are superior to population-based data, because the latter do not control adequately for confounding factors. Yet, the strengths of randomized trials stem from their freedom from confounding with respect to comparison of the treatment arms under study and not from confounding with respect to comparisons of auxiliary, nonrandomized variables, such as ethnicity. Moreover, the participants in these trials may have been selected for factors related to prognosis, limiting generalization of ethnic comparisons to more general populations.
Roach et al. also mistake our conclusion as being genetically based. Although there may indeed be genetic factors involved, we also believe that nongenetic factors, for example, diet or exercise, may influence tumor aggressiveness.
Roach et al. accuse us of assuming that all HMO members utilize health care equally. Yet, we clearly stated: "Systematic racial differences in survival within stages could plausibly arise, even in equal health care systems, if there were important racial differences in factors such as frequency of contact with health care providers. An important point is that equal-access health care systems can only guarantee equality of covered benefits for enrolled members, not equality in the frequency of use of all forms of care." Roach et al. also state that we did not address the possibility of residual confounding by stage. Yet we clearly acknowledged ". . . within each of the three broader Surveillance, Epidemiology, and End Results (SEER) stages used in the present study, racial differences in survival still could exist if white men in a given stage were diagnosed on average at an earlier point in the history of their cancers."
Roach et al. complain that we do not attribute survival differences between whites in and out of the KPMCP to biologic factors. But comparing black versus white survival among men with uniform financial access to care is quite different from comparing survival of whites with unequal financial access. It is precisely this distinction between an ethnic comparison within HMO members and a member versus nonmember comparison within ethnic groups that motivated our analysis.
Roach et al. compare our analysis to the infamous Tuskegee syphilis trials and cite their earlier conclusion that any interpretation of the observed black versus white survival differences as due to biologic factors is "racist" (7). We believe that such a conclusion, a priori of a full evaluation of the data, constrains studies to be politically correct and does a disservice to the scientific community and the public. Indeed, any restraints on objective analysis of data do a disservice to all prostate cancer patients.
In conclusion, we agree with Roach et al. that much of the survival disadvantage of black men may be due to their disproportionate burden of barriers to the best health care. We need to address these inequities. However, the data suggest that some of the disadvantage may also be due to biologic differences in tumor characteristics. We believe to dismiss this possibility would be a disservice to good science and the community.
REFERENCES
1
Robbins AS, Whittemore AS, Van Den Eeden SK. Race, prostate
cancer survival, and membership in a large health maintenance organization. J Natl Cancer
Inst 1998;90:986-90.
2
Whittemore AS, Lele C, Friedman GD, Stamey T, Vogelman JH,
Orentreich N. Prostate-specific antigen as predictor of prostate cancer in black men and white
men. J Natl Cancer Inst 1995;87:354-60.
3 Moul JW, Douglas TH, McCarthy WF, McLeod DG. Black race is an adverse prognostic factor for prostate cancer recurrence following radical prostatectomy in an equal access health care setting. J Urol 1996;155:1667-73.[CrossRef][ISI][Medline]cancerlit;96186051
4 Sakr WA, Grignon DJ, Haas GP, Schomer KL, Heibrun LK, Cassin BJ, et al. Epidemiology of high grade prostatic intraepithelial neoplasia. Pathol Res Pract 1995;191:838-41.[ISI][Medline]cancerlit;96188488
5 Guileyardo JM, Johnson WD, Welsh RA, Akazaki K, Correa P. Prevalence of latent prostate carcinoma in two U.S. populations. J Natl Cancer Inst1980 ;65:311-6.
6
Bolla M, Gonzalez D, Warde P, Dubois JB, Mirimanoff RO,
Storme G, et al. Improved survival in patients with locally advanced prostate cancer treated with
radiotherapy and goserelin. N Engl J Med 1997;337:295-300.
7 Roach M 3d. Is race an independent prognostic factor for survival from prostate cancer? J Natl Med Assoc 1998;90:S713-9.[Medline]cancerlit;99046028
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||