Correspondence
Racial Differences in the Treatment of Early-Stage Lung Cancer
N Engl J Med 2000; 342:517-519February 17, 2000
- Article
To the Editor:
We believe that racial differences in social and economic factors and in coexisting illness may have confounded the finding, reported by Bach et al. (Oct. 14 issue),1 that black patients with early-stage non–small-cell lung cancer were less likely to receive surgical treatment and to survive than similar white patients. We offer four suggestions for dealing with these potential confounding variables more effectively.
First, Bach et al. did not include marital status in their statistical models. In the U.S. population of persons 65 to 84 years old, relatively fewer blacks than whites are married and living with a spouse,2 and in a prior study of treatment for lung cancer, being married was strongly associated with undergoing surgery.3 Since the Surveillance, Epidemiology, and End Results (SEER) cancer registries record marital status, possible confounding could be assessed.
Second, the authors used a proxy measure of individual economic status based on median income for the postal code of residence. An alternative measure of individual economic need is whether a patient was eligible for the Medicaid Buy-in program, which pays for Medicare Part B coverage from state welfare funds. Medicare has recorded Medicaid Buy-in data since 1985, and these data may be more useful than aggregate measures of income.
Third, the authors controlled for coexisting illness by using the Romano–Charlson comorbidity index or the number of hospitalizations in the previous year. The comorbidity index was calculated for only about 24 percent of the patients in the study (those who were 66 years of age or older, had indemnity insurance, and had recently been hospitalized), and it is unclear whether the reasons for hospitalization were similar among black patients and white patients. Thus, one cannot assume that the assessment of coexisting illness was unbiased. Moreover, these measures are poor surrogates for contraindications to treatment for lung cancer. A more appropriate measure could be derived from data in Medicare claims files on reduced pulmonary function, hypercarbia, or cor pulmonale, which are specific contraindications to pulmonary resection. But the problem of missing data would still be daunting.
Finally, since 1988, the SEER program has collected data on why patients were not treated (i.e., whether surgery was contraindicated or the patient refused treatment). Bach et al. studied patients who received diagnoses between 1985 and 1993, so reasons for not performing surgery may be recorded for most of them. Analysis of these data may help shed some light on the many plausible explanations for the apparent undertreatment of black patients with cancer as compared with white patients and for the subsequent lower survival rate among black patients.
3 ReferencesDiane E. Campbell, Ph.D.
E. Robert Greenberg, M.D.
Norris Cotton Cancer Center, Lebanon, NH 037561
Bach PB ,Cramer LD ,Warren JL ,Begg CB . Racial differences in the treatment of early-stage lung cancer. N Engl J Med 1999;341:1198-1205
Full Text | Web of Science | Medline2
Current Population Survey Reports. Unpublished tables — marital status and living arrangements: March 1998 (update). Washington, D.C.: Bureau of the Census, 1998. (See: http://www.census.gov:80/population/www/socdemo/ms-la.html.)
3
Greenberg ER ,Chute CG ,Stukel T , et al. Social and economic factors in the choice of lung cancer treatment: a population-based study in two rural states. N Engl J Med 1988;318:612-617
Full Text | Web of Science | Medline
To the Editor:
Bach et al. report a lower rate of surgical resection among blacks than whites in a group of patients 65 years of age or older who were given a diagnosis of early-stage non–small-cell lung cancer between 1985 and 1993 and who resided in SEER study areas. The authors note that “it is not clear whether there is similar variability in the care provided to younger patients with lung cancer.”
On the basis of SEER data1 for patients with stage I or II non–small-cell lung cancer2 diagnosed between 1988 and 1995, the difference between the proportions of black patients and white patients who underwent surgery was smaller (albeit statistically significant) in the group of patients under the age of 65 years than in the group of patients who were 65 years of age or older (Table 1Table 1
Surgical Status of Black Patients and White Patients with Stage I or II Non–Small-Cell Lung Cancer Diagnosed between 1988 and 1995 in SEER Areas.).Preferences for treatment may differ between black patients and white patients, black patients may be offered treatment less often than white patients, or both factors may explain the differences.3,4 The SEER data (Table 1) show that the absence of a physician's recommendation for surgery was more frequent in the group of black patients than in the group of white patients; refusal of surgery and contraindications (due to other medical conditions) were uncommon. The results of an analysis that excluded patients who survived for less than two months (those with a limited opportunity to undergo surgery) were similar (data not shown). Research should focus on why black patients, especially those who are elderly, receive a recommendation for surgery less often than white patients.
4 ReferencesAnthony P. Polednak, Ph.D.
Connecticut Department of Public Health, Hartford, CT 061341
SEER program public-use CD-ROM (1973–1995) (based on the August 1997 submission). Bethesda, Md.: National Cancer Institute, April 1998.
2
National Cancer Institute. SEER program: comparative staging guide for cancer. Bethesda, Md.: National Institutes of Health, 1993. (NIH publication no. 93-3640.)
3
King TE Jr ,Brunetta P . Racial disparity in rates of surgery for lung cancer. N Engl J Med 1999;341:1231-1233
Full Text | Web of Science | Medline4
Bach PB ,Cramer LD ,Warren JL ,Begg CB . Racial differences in the treatment of early-stage lung cancer. N Engl J Med 1999;341:1198-1205
Full Text | Web of Science | Medline
To the Editor:
The article by Bach et al. highlights a pervasive, disturbing, and clearly intolerable feature of our health care system. There are several disparities in health care between blacks and whites in the outcomes of certain diseases. This disparity results in at least 60,000 excess deaths in the black population every year. Compare this figure with the 58,000 deaths that occurred over a 10-year period as a result of the Vietnam War. Our outrage over the needless deaths in Vietnam should be no greater than our outrage at the even larger loss of life in American communities every year. Even after one accounts for some differences in socioeconomic, genetic, social, or cultural factors between blacks and whites, an unexplained factor seems to remain. Overt and subtle racism has been postulated as a possible explanation.1 The study by Bach et al. appears to confirm that there is a difference in how patients are treated according to their race. Whatever the cause of this difference, the health care system is clearly disempowering and dangerous for black patients.
1 ReferencesCharles DeShazer, M.D.
7234 Muncaster Mill Rd., Suite 549, Rockville, MD 208551
Charatz-Litt C . A chronicle of racism: the effects of the white medical community on black health. J Natl Med Assoc 1992;84:717-725
Web of Science | Medline
Author/Editor Response
The authors reply:
To the Editor: The correspondents raise some important issues that contribute to our understanding of the disparities in resection rates between black patients and white patients with stage I or II non–small-cell lung cancer. Greenberg et al. previously demonstrated an association between marital status and surgery for lung cancer 1 — this factor was not considered in our first analysis. In our cohort, married patients were more likely than others to undergo surgical resection (P<0.001), but marital status did not confound the primary results (Table 1Table 1
Resection Rate and Relative Risk, According to Race and Marital Status, for Medicare Beneficiaries (65 Years of Age or Older) with Stage I or II Non–Small-Cell Lung Cancer, 1985 to 1993.). Disparities in treatment were present in all marital categories.We used a validated, claims-based measure of coexisting illness,2 and explained why the methods we used would be more likely to bias the results in favor of the null hypothesis than to overstate an observed disparity in treatment. The proposal by Campbell and Greenberg to use pulmonary-specific measures of coexisting illness is appealing but cannot be accomplished effectively with codes from the International Classification of Diseases, 9th Revision. Similarly, an array of adjustments for socioeconomic status is available, each with its own limitations. We chose a validated method based on U.S. census data that is frequently used in population-based studies.3,4 There are two problems with using data from the Medicaid Buy-in program as a surrogate variable for poverty. First, the program does not include all persons who are poor, only those who qualify for selected state-specific programs. Second, less than 6 percent of the patients in our sample were enrolled in this program.
Polednak, as well as Campbell and Greenberg, proposes an analysis of the SEER variable for the reason that surgery was not recommended. Because of limited documentation in hospital records, it is not possible to know what factors resulted in the decision not to recommend surgery (Ries L, SEER program: personal communication). However, the analysis Polednak presents is intriguing and suggests that further research on physician–patient interactions will be fruitful. This type of analysis may help explain the treatment disparities that we observed. It should be emphasized that Polednak's analysis, performed with the SEER public-use data base, contains data on patients who were excluded from our analysis because there was no documentation of nodal status, the patient had received a diagnosis of another cancer within two months of receiving the diagnosis of primary lung cancer, or there was no link to Medicare data.5 The inclusion of these patients in the analysis performed by Polednak does not alter our primary finding but does result in the lower overall surgical rates that he reports.
5 ReferencesPeter B. Bach, M.D.
Laura D. Cramer, Sc.M.
Colin B. Begg, Ph.D.
Memorial Sloan-Kettering Cancer Center, New York, NY 10021Joan L. Warren, Ph.D.
National Cancer Institute, Bethesda, MD 208921
Greenberg ER ,Chute CG ,Stukel T , et al. Social and economic factors in the choice of lung cancer treatment: a population-based study in two rural states. N Engl J Med 1988;318:612-617
Full Text | Web of Science | Medline2
Romano PS ,Roos LL ,Jollis JG . Adapting a clinical comorbidity index for use with ICD-9-CM administrative data: differing perspectives. J Clin Epidemiol 1993;46:1075-1079
CrossRef | Web of Science | Medline3
Garg PP ,Frick KD ,Diener-West M ,Powe NR . Effect of the ownership of dialysis facilities on patients' survival and referral for transplantation. N Engl J Med 1999;341:1653-1660
Full Text | Web of Science | Medline4
Wolfe RA ,Ashby VB ,Milford EL , et al. Comparison of mortality in all patients on dialysis, patients on dialysis awaiting transplantation, and recipients of a first cadaveric transplant. N Engl J Med 1999;341:1725-1730
Full Text | Web of Science | Medline5
Potosky AL ,Riley GF ,Lubitz JD ,Mentnech RM ,Kessler LG . Potential for cancer related health services research using a linked Medicare-tumor registry database. Med Care 1993;31:732-748
CrossRef | Web of Science | Medline
- Citing Articles (4)
Citing Articles
1
Robert A. Hiatt, Nancy Breen. (2008) The Social Determinants of Cancer. American Journal of Preventive Medicine 35:2, S141-S150
CrossRef2
Mitchel Berger, Mary Jo Lund, Otis W. Brawley. (2007) Racial Disparities in Lung Cancer. Current Problems in Cancer 31:3, 202-210
CrossRef3
Jérôme Adda, Francesca Cornaglia. (2006) Taxes, Cigarette Consumption, and Smoking Intensity. American Economic Review 96:4, 1013-1028
CrossRef4
Miquel Porta, Esteve Fernandez, Joan Alguacil. (2003) Semiology, proteomics, and the early detection of symptomatic cancer. Journal of Clinical Epidemiology 56:9, 815-819
CrossRef
