0
Clinical Investigations: SURGERY |

Evaluation of the Causes for Racial Disparity in Surgical Treatment of Early Stage Lung Cancer* FREE TO VIEW

Jennifer McCann, MD; Vasken Artinian, MD; Lisa Duhaime, MD; Joseph W. Lewis, Jr, MD, FCCP; Paul A. Kvale, MD, FCCP; Bruno DiGiovine, MD, MPH, FCCP
Author and Funding Information

*From the Division of Pulmonary and Critical Care (Drs. Artinian, Kvale, and DiGiovine), Department of Internal Medicine (Drs. McCann and Duhaime), and Division of Thoracic Surgery (Dr. Lewis), Department of Surgery, Henry Ford Health System, Detroit, MI.

Correspondence to: Bruno DiGiovine, MD, MPH, Division of Pulmonary and Critical Care, Henry Ford Health System, 2799 W Grand Blvd, K-17, Detroit, MI 48202; e-mail: bdigiov1@hfhs.org



Chest. 2005;128(5):3440-3446. doi:10.1378/chest.128.5.3440
Text Size: A A A
Published online

Study objectives: Black patients undergo surgical treatment for early stage lung cancer less often than whites. We wanted to determine the causes for the racial difference in resection rates.

Design: We studied a retrospective cohort of patients who presented to our institution with potentially resectable lung cancer (stage I or II) between the years 1995 and 1998, inclusive.

Setting: A tertiary-referral hospital and clinic with a cancer database of all lung cancer patients seen.

Patients: A total of 281 patients were included: 97 black patients (35%) and 184 white patients (65%).

Measurements and results: The surgical rate was significantly lower in blacks than in whites (56 of 97 patients [58%] vs 137 of 184 patients [74%], p = 0.004). We could not find evidence that the rate at which surgical treatment was offered was different between the two racial groups (68 of 97 black patients [70%] and 145 of 184 white patients [79%], p = 0.11). After controlling for preoperative pulmonary function, tumor stage, history of smoking, and significant comorbidities, we were unable to show that race was a predictor of being offered surgical treatment (odds ratio, 0.46; 95% confidence interval, 0.18 to 1.14; p = 0.09). The difference in surgical rates was mainly due to the fact that blacks were found to decline surgical treatment more often than their white counterparts (12 of 68 patients [18%] vs 7 of 145 patients [5%], p = 0.002).

Conclusions: Our analysis suggests that the lower surgical rate among black patients with early stage lung cancer is mainly due to low rates of acceptance of surgical treatment.

Figures in this Article

Lung cancer is responsible for more cancer deaths in the United States than the next three most common cancers combined.1Only 20 to 25% of patients with non-small cell lung cancer (NSCLC) have stages I or II.23 Surgical treatment of the early stages of NSCLC is considered likely to be curative and is the standard of treatment for this disease. Patients who undergo surgical treatment for stages I and II NSCLC will have approximately a 40 to 50% chance of surviving for 5 years, compared to a < 1-year median survival in patients who do not undergo surgical treatment.4 Thus, performing surgical resection in a timely manner is the only potential curative therapeutic modality that impacts survival in early stage lung cancer.24

Mortality from lung cancer is higher in the black population.57 Given the striking difference in survival between patients with early stage lung cancer who undergo surgical resection and those who have alternative treatments, it is not surprising that research has shown that much of the higher mortality in blacks with early stage NSCLC can be attributed to differences in surgical rates.8 In fact, there are now multiple studies810 that have shown that black patients receive surgical resection for lung cancer less often than whites. However, the underlying reason for this disparity in surgical rates has not been directly addressed, with some authors11 suggesting that physician bias may be playing a role.

We therefore undertook this study to evaluate the factors that might account for the difference in surgical rates of stages I and II NSCLC between blacks and whites. We designed our study so that we could address several possible explanations for the discrepancy in surgical rates. Specifically, we assessed the following: (1) whether blacks had a higher incidence of coexisting illnesses that made them less fit for surgical treatment, (2) whether physicians offered surgical treatment less frequently to blacks, and/or (3) whether blacks were less likely to accept surgical treatment.

We retrospectively analyzed all black (African-American) and white patients enrolled in the Josephine Ford Cancer Center registry with potentially localized lung cancer. We explicitly excluded other races, as the numbers of such patients was believed to likely be too small for statistical analysis. We limited our search to those patients seen in the pulmonary clinic at Henry Ford Health System (HFHS) in Detroit, MI, between January 1995 and December 1998. From these patients, we identified all patients who were presenting for the first time with stage I or II NSCLC (based on preoperative clinical evaluation) according to the staging system of the American Joint Committee on Cancer.12 Thus, patients were excluded if they were found to have a more advanced stage of lung cancer, had small cell lung cancer, or had metastatic cancer from a separate primary. Also, patients were excluded if they died prior to definitive evaluation or refused workup at such an early point that they were unable to be clinically staged (no chest CT was completed). Patients were also excluded from analysis if they were unavailable for follow-up (Table 1 ).

Pertinent data were obtained from the Josephine Ford Cancer Center registry and from the patient’s medical record. The HFHS has an electronic medical record for outpatient visits that allowed for high reliability in obtaining the information necessary for the completion of this study. The study was reviewed and approved by the Institutional Review Board of the HFHS.

Data on stage, treatment, demographics, lung functional status, physician recommendations, patient’s acceptance of physician recommendations, comorbidities, and survival were collected from chart review and compiled in a database. Age and marital status were coded at the time of diagnosis. All patients were staged clinically at the time of their visit to the pulmonary clinic. Those patients who underwent surgical resection may have also been staged pathologically. However, to allow valid comparisons between patients, only clinical staging is reported in this article. Race was self-determined by the patient at the time of initial presentation to the HFHS. Socioeconomic status was estimated based on the patient’s address using the block group median income from the 1990 US census. Information on the following comorbidities were collected: COPD, diabetes mellitus, coronary artery disease (CAD), myocardial infarction within 6 months prior to cancer diagnosis, stroke, smoking history, hypertension, dementia, and congestive heart failure as confirmed by a low ejection fraction or a mention of the diagnosis in the medical record. All patient diagnoses were obtained from chart review. In addition, whether the patient had been hospitalized in the 6 months prior to diagnosis of lung cancer was included. Pathologic grade and histology were obtained on all the patients.

Patients were considered to be offered surgical treatment if this treatment modality was recommended and documented in the medical record by the treating physician at the time of diagnosis. The outcome variables were acceptance of physician’s recommendation for surgical resection and actual resection. Patients were believed to have accepted surgical resection if that acceptance was documented in the medical record and/or if they underwent an operation. Patients were labeled as having undergone surgical resection if they underwent a procedure attempting to resect the tumor whether or not that resection could ultimately be completed. Patient charts were reviewed for any treatment documented outside of the HFHS, and those patients who were noted to have an operation out of HFHS were included in this group of patients who accepted surgical resection. Patients were labeled as declining surgical treatment if there was explicit documentation in the medical record of that refusal or if they failed to follow up with treatment recommendations for surgical resection but were not unavailable for follow-up.

Statistical Analysis

Differences in patient characteristics and the rates at which patients were offered, accepted, and underwent surgical resection for lung cancer were assessed by means of χ2 test for categorical variables. To test for an association between race and being offered surgical treatment, all potentially important variables were assessed in a univariate logistic regression model to determine if they were significantly associated with the outcome of interest. Those variables were then added to a multivariate model in a stepwise fashion based on the variable with the highest Wald χ2 value based on the recommendations from Hosmer and Lemeshow.13 In building the multivariate model, an α level of 0.05 was used as a cutoff to enter a new variable in a forward stepwise analysis. After further variables were entered, a variable was removed if the p value associated with it was > 0.10. After we defined which factors should be in the model, we checked the assumption of linearity in the logit. If it was not present, we changed the variable to a categorical variable, according to the methods of Hosmer and Lemeshow.13 A similar analysis was done to assess factors that were significantly associated with a patient’s decision to accept surgical resection.

The final model was then investigated for possible problems. All linear variables were assessed for lack of conformity to a linear gradient, using design variables as described by Hosmer and Lemeshow.13All final variables were assessed for interactions and collinearity. Goodness of fit was assessed using the Hosmer-Lemeshow statistic.14All data from these analyses are presented in Tables 234 using the style suggested by Lang and Secic.15

Odds ratios (ORs) and 95% confidence intervals (CIs) were computed for all significant variables. With the exception of the multivariate model noted above, an α level of 0.05 is used throughout the article to indicate statistical significance. Statistical software (Version 6.1; SAS Institute; Cary, NC) was used to conduct the analysis.

A total of 388 patients with presumed localized lung cancer were seen in the pulmonary clinic during the study period. Of these, 107 patients were excluded (12 were stage IIIA, 39 were stage IIIB, 19 were stage IV, 26 had small cell lung cancer, 2 had lung metastases from a separate primary, 2 were unavailable for follow-up, and 7 were unable to be staged because they died prior to the work-up or refused work-up and entered hospice).1 Thus, 281 patients with stage I or II NSCLC met our inclusion criteria and had sufficient data in their medical record to abstract the main variables of whether surgical treatment was offered and/or accepted. Of these 281 patients, 97 were black (35%) and 184 were white (65%). The mean (± SD) age for the population was 68.3 ± 9.6 years, with no difference noted by race. There was also no difference in sex, marital status, tumor histology, or comorbidities (Table 1). Blacks were of a lower socioeconomic status and were more likely to present with a higher stage of cancer.

Only 56 black patients (58%) underwent surgical resection, as opposed to 137 white patients (74%; p = 0.004) [Fig 1] . This was true despite the fact that > 90% of the population had insurance and all patients had access to health care through the HFHS. This difference in surgical rates was observed despite the fact that the rates at which patients were offered surgical resection were not significantly different (145 of 184 white patients [79%] vs 68 of 97 black patients [70%], respectively; p = 0.11). Therefore, much of the difference in the rate of surgical resection was related to a differential likelihood to decline surgical treatment. Of those patients offered surgical resection, blacks were over three times more likely to decline this recommendation (12 of 68 patients [18%] vs 7 of 145 patients [5%], p = 0.002). It is possible that a patient would be more likely to accept surgical intervention if they saw a thoracic surgeon than if they only saw a pulmonologist. Thus, we investigated the rate of referral to a thoracic surgeon among those who refused surgical intervention. Overall, 68% (13 of 19 patients) who refused surgical treatment saw a thoracic surgeon. Importantly, the percentage of black patients (75%; 9 of 12 patients) who saw a thoracic surgeon but refused surgical resection was higher than for whites (57%; 4 of 7 patients), although refusal rates were not significantly different (p = 0.42).

We also wanted to examine whether the race of the treating physician(s) was a factor in the decision to refuse surgical treatment. There were only five black patients in our database who saw a black physician. Only three of these patients were offered surgical resection, and all three had an operation. These small numbers do not allow us to draw any conclusions as to whether the race of the treating physician impacts the decision to accept surgical treatment among black patients.

Although there was no statistical difference in the rate of offering surgical treatment, there was a trend toward fewer blacks being offered surgical treatment. To evaluate the significance of this trend, we did a number of further analyses. First, we evaluated the documentation in the medical record to see if there were stated reasons to explain the reason for the patients not being offered surgical treatment. Poor pulmonary function was the most likely reason a patient was not offered surgical treatment (40 of 68 patients, 59%), followed by poor overall performance status/multiple comorbidities (18 of 68 patients, 26%), followed by poor cardiac function (7 of 68 patients, 10%). The remaining three patients were not offered surgical treatment for a variety of reasons. In summary, there was no significant difference in the documented reasons why patients were not offered resection. Also, in no case was a reason not found for not offering surgical treatment.

We then performed univariate and multivariate analyses to discover which factors were associated both with the likelihood of being offered surgical resection and of declining it. Univariate logistic analyses found multiple measures of pulmonary function, comorbidities, and demographic data to be associated with the likelihood of being offered surgical resection (Table 2). In multivariate analysis, only a history of COPD, CAD, expiratory lung volumes (FEV1 and FVC), and stage of lung cancer were associated with the likelihood of being offered surgical resection. In evaluating the continuous variables, we found that the likelihood of being offered resection was not linearly related with FEV1. Instead, the likelihood was best modeled by changing FEV1 into a group variable as follows: FEV1 < 1.31, FEV1 > 1.31 but ≤ 1.75, and FEV1 > 1.75. Thus, for the variable FEV1 group, the OR shown is associated with a 1-U increase in the group variable (Table 3). The multivariate model predicting the offering of surgery fit the data well (Hosmer-Lemeshow statistic, 5.01; p = 0.76). After controlling for stage, COPD, CAD, and expiratory lung volumes, we were unable to show that race was significantly associated with the likelihood of being offered surgical resection (OR, 0.46; 95% CI, 0.18 to 1.14; p = 0.09).

The likelihood of declining surgical resection was associated only with age, race, and the presence of CAD in a univariate analysis. In the multivariate analyses, only age and race were significant predictors of the likelihood of declining resection (Table 4). Interestingly, neither socioeconomic status nor marital status was a significant factor. However, after controlling for age, we found that blacks were more than four times more likely to decline surgical resection (OR, 4.19; 95% CI, 1.48 to 11.83; p = 0.007). Again, the model fit the data well (Hosmer-Lemeshow statistic, 11.93; p = 0.15), and there was no evidence for interactions or collinearity.

Our study confirms prior research that black patients are less likely to undergo surgical treatment for early stage lung cancer. To further investigate this disparity in surgical resection, we looked at the rates at which black patients were offered and accepted surgical treatment. After adjusting for age, black patients were more likely to decline lung resection. We also could not find a statistically significant relationship between patient race and a physician’s recommendation for surgical treatment. Thus, our data do not support the hypothesis that racial discordance in surgical rates is related to racial bias in the offering of surgical resection, as argued by some.11

What factors could potentially explain the lower rate of acceptance of surgical treatment in black patients? It is possible that blacks weigh the risks of an operation differently than whites. This is supported by previously observed racial differences in surgical treatment and survival from different cancers. For example, blacks were found to undergo surgical resection less frequently for esophageal cancer despite the presence of equal access to the medical system.16Black patients were also found to have surgical resection less frequently for colorectal cancers after adjusting for age, tumor stage and comorbidities.17

In addition to individual factors such as patient perception and preference, racial and cultural differences between the patient and the physician may influence surgical acceptance among blacks. In one study, black patients seeing white physicians rated their physician’s decision-making styles as less participatory.18In a separate study,19black patients with black physicians were more likely to report receiving necessary medical care than black patients with other-race physicians. Distrust of the health-care system and suspicion of being mistreated can also be a barrier for partnership, communication, and delivery of appropriate treatment. Racial and other ethnic minority group members report less satisfaction and trust of physicians compared to white patients.20 Thus, a number of patient- and physician-related factors could account for the observed lower surgical acceptance in blacks.

However, we believe that the most likely root cause of this racial difference may have been recently elucidated by Margolis et al,21 who performed a cross-sectional survey of pulmonary and lung cancer clinic attendees of various races. They surveyed the patients as to whether they had heard and believed the notion that lung cancer spreads if exposed to air during an operation. They found that 29% of white patients and 61% of black patients believed that the notion was true. Importantly, 5% of whites and 19% of blacks said they would oppose surgical treatment for this reason.21 It is striking that these percentages (5% of whites and 19% of blacks) corresponds quite closely to the percentage of patients who actually declined surgical resection in our study (5% of whites and 18% of blacks). Although we have no data on our patients’ beliefs, we would hypothesize that the notion of an operation leading to lung cancer metastasis likely had a very strong impact on the likelihood of the patients declining surgical treatment.

Previous research2225 evaluating differences in cancer outcomes across races has highlighted the importance of socioeconomic status. This relationship also holds in lung cancer, in which patients with a low median family income and without private medical insurance were shown to receive surgical treatment less often.10,26 Socioeconomic class had no influence on the likelihood to decline surgical treatment in our population. Importantly, an equally high percentage (approximately 93%) of each racial group had medical insurance and access to a specialized medical care center. The selection of a subgroup with medical insurance might account for the lower impact of socioeconomic status.

In addition to race, patient age was found to be a predictor for surgical acceptance. This adds to a growing body of research indicating that the choice of cancer therapy varies with patient age. Elderly patients are less likely to receive potentially curative surgical treatment for different cancers including early stage lung cancer.2728 It has been suggested that this lower rate of surgical treatment may be because physicians are less likely to recommend specialist consultation for elderly patients.29 Conversely, we did not find age to be an independent factor in the offering of surgical resection. Instead, we found age to be an independent marker of patients who were more likely to decline surgical treatment.

Our study has several potential limitations. First, the use of median income according to the address of residence as a surrogate for socioeconomic status does not exclude residual confounding due to heterogeneity of income within different neighborhoods. However, our measure of income is likely the best that is possible in such a retrospective study. Second, given the design of our study, we were unable to examine potential interactions between the patient and the treating physician. Problems in communication due to racial discordance between the patient and the physician might have affected the adherence with the recommendation to undergo surgical resection. We could not independently assess these interactions in this study; however, we do agree with those who have suggested that increased education in “cultural competence” may lead to significant improvements in the care of racial and ethnic minorities.11,30 In fact, our results further confirm the importance of racial and cultural factors in the patient-physician relationship. Third, while we were unable to show a statistically significant relationship between patient race and the offering of surgical resection, there was a trend toward a lower rate of offering surgical resection to blacks. In fact, our multivariate analysis suggested that whites were approximately twice as likely to be offered surgery. However, because of the imprecision of our statistical test (related mainly to sample size), we could not prove that the likelihood of being offered surgery was different for blacks and whites. Nevertheless, we do not think that our data should be used to argue that the rate of the offering of surgery is the same. The lack of a statistically significant relationship does not prove its absence. Given the magnitude of the OR, one is left to wonder whether a significant relationship would have been found if more patients were enrolled. This is certainly possible. However, there were clear limitations on sample size in this study. The complexity of the chart review undertaken for this study made it necessary to limit the analysis to one center. Also, in attempts to limit the potential confounding effect of different surgeons having different practice patterns, we tried to limit our study time period. This was quite successful as approximately 90% of the patients who underwent surgical resection or refused surgical resection despite seeing a thoracic surgeon were seen by the same surgeon (J.W.L.). Despite these limitations, our study size was in fact quite impressive. This can be better appreciated by reflecting on the study size from the study by Bach et al8 on rates of surgical resection by race. In that 8-year study of all the patients in the Surveillance, Epidemiology, and End Results registry, 860 black patients were included.8 In our study, we were able to completely evaluate the medical records of 98 black patients from one institution. This total is rendered more impressive when one realizes that our study contained more black patients than were included in the Surveillance, Epidemiology, and End Results registry from Los Angeles County in an 8-year period (n = 86).8 Nevertheless, further research should continue to examine whether there are differences in the rate of offering surgical resection by race. However, given the much more striking difference in rates of surgical acceptance, we believe programs that wish to increase surgical rates in blacks with NSCLC should focus primarily on the issue of surgical acceptance, perhaps by addressing patients’ beliefs about cancer spread.21

In conclusion, we have confirmed the lower rate of surgical resection for early stage lung cancer in blacks. Our study design has allowed us to narrow the search for the root cause by identifying patient acceptance of surgical treatment as a key factor. Further research now should be done to understand what factors are most important in causing this low rate of surgical acceptance. This hopefully will lead to greater rates of surgical acceptance in black patients and ultimately lower lung cancer mortality rates in the black population.

Abbreviations: CAD = coronary artery disease; CI = confidence interval; HFHS = Henry Ford Health System; NSCLC = non-small cell lung cancer; OR = odds ratio

Presented at the annual meeting of the American College of Chest Physicians, November 2002, San Diego, CA.

Funding was provided by American Cancer Society grant CRTG-98–284-01.

Table Graphic Jump Location
Table 1. Patient Characteristics*
* 

Data are presented as mean ± SD or % unless otherwise indicated.

Table Graphic Jump Location
Table 2. Factors Associated With the Likelihood of Being Offered Surgery in a Univariate Analysis*
* 

Variables are listed in order of decreasing Wald χ2. Dlco = diffusion capacity of the lung for carbon monoxide.

 

OR represents the change in likelihood associated with 1-U change in the continuous variable.

 

OR represents the change in likelihood associated with the presence of the variable.

§ 

OR represents the risk associated with any one step increase in stage from IA through IIB (stages are IA, IB, IIA, and IIB).

Table Graphic Jump Location
Table 3. Variables Significantly Associated With the Offering of Surgical Resection in a Multivariate Model*
* 

Variables are listed in order of decreasing Wald χ2.

 

OR represents the change in likelihood associated with 1-U change in the continuous variable.

 

OR represents the change in likelihood associated with the presence of the variable.

§ 

We found that the likelihood of being offered resection was not linearly related with FEV1. Instead, the likelihood was best modeled by changing FEV1 into a group variable as follows: FEV1 < 1.31, FEV1 > 1.31 but ≤ 1.75, and FEV1 > 1.75. Thus, for the variable FEV1 group, the OR shown is associated with a 1-U increase in the group variable.

 

OR represents the risk associated with any one step increase in stage from IA through IIB (stages are IA, IB, IIA, and IIB).

Table Graphic Jump Location
Table 4. Variables Significantly Associated With the Declining of Surgical Resection in a Multivariate Model*
* 

Variables are listed in order of decreasing Wald χ2.

 

OR represents the change in likelihood associated with 1-year change in age.

 

OR represents the change in likelihood associated with being black.

Figure Jump LinkFigure 1. Rates of being offered surgical resection and rates among those patients offered surgical resection of declining resection or having resection stratified by race. In each case, the percentages are reflecting different denominators. Thus, for the surgical rate and the offered rate, the denominator is the total number of patients of each race (184 white patients and 97 black patients). However, for the rate of declining surgical resection, the denominator is the number of patients offered resection (145 white patients and 68 black patients).Grahic Jump Location

The authors thank Martin Tammemagi and Christine Cole Johnson for their thoughtful critique of the manuscript.

. American Cancer Society. (2004)Cancer facts and figures, 2004. American Cancer Society. Atlanta, GA:
 
Scott, WJ, Howington, J, Movsas, B Treatment of stage II non-small cell lung cancer.Chest2003;123,188S-201S. [CrossRef] [PubMed]
 
Smythe, WR Treatment of stage I non-small cell lung carcinoma.Chest2003;123,181S-187S. [CrossRef] [PubMed]
 
Deslauriers, J, Gregoire, J Surgical therapy of early non-small cell lung cancer.Chest2000;117,104S-109S. [CrossRef] [PubMed]
 
Cooley, ME, Jennings-Dozier, K Lung cancer in African Americans: a call for action.Cancer Pract1998;6,99-106. [CrossRef] [PubMed]
 
Wingo, PA, Ries, LA, Giovino, GA, et al Annual report to the nation on the status of cancer, 1973–1996, with a special section on lung cancer and tobacco smoking.J Natl Cancer Inst1999;91,675-690. [CrossRef] [PubMed]
 
Wu, LY, Semenya, KA, Hardy, RE, et al Cancer rate differentials between blacks and whites in three metropolitan areas: a 10-year comparison.J Natl Med Assoc1998;90,410-416. [PubMed]
 
Bach, PB, Cramer, LD, Warren, JL, et al Racial differences in the treatment of early-stage lung cancer.N Engl J Med1999;341,1198-1205. [CrossRef] [PubMed]
 
Smith, TJ, Penberthy, L, Desch, CE, et al Differences in initial treatment patterns and outcomes of lung cancer in the elderly.Lung Cancer1995;13,235-252. [CrossRef] [PubMed]
 
Greenwald, HP, Polissar, NL, Borgatta, EF, et al Social factors, treatment, and survival in early-stage non-small cell lung cancer.Am J Public Health1998;88,1681-1684. [CrossRef] [PubMed]
 
King, TE, Jr, Brunetta, P Racial disparity in rates of surgery for lung cancer.N Engl J Med1999;341,1231-1233. [CrossRef] [PubMed]
 
Mountain, C Revisions in the international system for staging lung cancer.Chest1997;111,1710-1717. [CrossRef] [PubMed]
 
Hosmer, DW, Lemeshow, S Applied logistic regression 2nd ed.2000 John Wiley and Sons. New York, NY:
 
Lemeshow, S, Hosmer, DW, Jr A review of goodness of fit statistics for use in the development of logistic regression models.Am J Epidemiol1982;115,92-106. [PubMed]
 
Lang, T, Secic, M. How to report statistics in medicine. 1997; American College of Physicians. Philadelphia, PA:.
 
Dominitz, JA, Maynard, C, Billingsley, K, et al Race, treatment, and survival of veterans with cancer of the distal esophagus and gastric cardia.Med Care2002;40,I14-I26. [PubMed]
 
Cooper, GS, Yuan, Z, Landefeld, CS, et al Surgery for colorectal cancer: race-related differences in rates and survival among Medicare beneficiaries.Am J Public Health1996;86,582-586. [CrossRef] [PubMed]
 
Cooper-Patrick, L, Gallo, JJ, Gonzales, JJ, et al Race, gender, and partnership in the patient-physician relationship.JAMA1999;282,583-589. [CrossRef] [PubMed]
 
Saha, S, Komaromy, M, Koepsell, TD, et al Patient-physician racial concordance and the perceived quality and use of health care.Arch Intern Med1999;159,997-1004. [CrossRef] [PubMed]
 
Doescher, MP, Saver, BG, Franks, P, et al Racial and ethnic disparities in perceptions of physician style and trust.Arch Fam Med2000;9,1156-1163. [CrossRef] [PubMed]
 
Margolis, ML, Christie, JD, Silvestri, GA, et al Racial differences pertaining to a belief about lung cancer surgery: results of a multicenter survey.Ann Intern Med2003;139,558-563. [PubMed]
 
Wegner, EL, Kolonel, LN, Nomura, AM, et al Racial and socioeconomic status differences in survival of colorectal cancer patients in Hawaii.Cancer1982;49,2208-2216. [CrossRef] [PubMed]
 
Page, WF, Kuntz, AJ Racial and socioeconomic factors in cancer survival: a comparison of Veterans Administration results with selected studies.Cancer1980;45,1029-1040. [CrossRef] [PubMed]
 
Dayal, HH, Polissar, L, Dahlberg, S Race, socioeconomic status, and other prognostic factors for survival from prostate cancer.J Natl Cancer Inst1985;74,1001-1006. [PubMed]
 
Dayal, H, Polissar, L, Yang, CY, et al Race, socioeconomic status, and other prognostic factors for survival from colo-rectal cancer.J Chronic Dis1987;40,857-864. [CrossRef] [PubMed]
 
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 Med1988;318,612-617. [CrossRef] [PubMed]
 
Jazieh, AR, Kyasa, MJ, Sethuraman, G, et al Disparities in surgical resection of early-stage non-small cell lung cancer.J Thorac Cardiovasc Surg2002;123,1173-1176. [CrossRef] [PubMed]
 
Samet, J, Hunt, WC, Key, C, et al Choice of cancer therapy varies with age of patient.JAMA1986;255,3385-3390. [CrossRef] [PubMed]
 
Hoekstra, HJ Cancer surgery in the elderly.Eur J Cancer2001;37(suppl),S235-S244
 
Loudon, RF, Anderson, PM, Gill, PS, et al Educating medical students for work in culturally diverse societies.JAMA1999;282,875-880. [CrossRef] [PubMed]
 

Figures

Figure Jump LinkFigure 1. Rates of being offered surgical resection and rates among those patients offered surgical resection of declining resection or having resection stratified by race. In each case, the percentages are reflecting different denominators. Thus, for the surgical rate and the offered rate, the denominator is the total number of patients of each race (184 white patients and 97 black patients). However, for the rate of declining surgical resection, the denominator is the number of patients offered resection (145 white patients and 68 black patients).Grahic Jump Location

Tables

Table Graphic Jump Location
Table 1. Patient Characteristics*
* 

Data are presented as mean ± SD or % unless otherwise indicated.

Table Graphic Jump Location
Table 2. Factors Associated With the Likelihood of Being Offered Surgery in a Univariate Analysis*
* 

Variables are listed in order of decreasing Wald χ2. Dlco = diffusion capacity of the lung for carbon monoxide.

 

OR represents the change in likelihood associated with 1-U change in the continuous variable.

 

OR represents the change in likelihood associated with the presence of the variable.

§ 

OR represents the risk associated with any one step increase in stage from IA through IIB (stages are IA, IB, IIA, and IIB).

Table Graphic Jump Location
Table 3. Variables Significantly Associated With the Offering of Surgical Resection in a Multivariate Model*
* 

Variables are listed in order of decreasing Wald χ2.

 

OR represents the change in likelihood associated with 1-U change in the continuous variable.

 

OR represents the change in likelihood associated with the presence of the variable.

§ 

We found that the likelihood of being offered resection was not linearly related with FEV1. Instead, the likelihood was best modeled by changing FEV1 into a group variable as follows: FEV1 < 1.31, FEV1 > 1.31 but ≤ 1.75, and FEV1 > 1.75. Thus, for the variable FEV1 group, the OR shown is associated with a 1-U increase in the group variable.

 

OR represents the risk associated with any one step increase in stage from IA through IIB (stages are IA, IB, IIA, and IIB).

Table Graphic Jump Location
Table 4. Variables Significantly Associated With the Declining of Surgical Resection in a Multivariate Model*
* 

Variables are listed in order of decreasing Wald χ2.

 

OR represents the change in likelihood associated with 1-year change in age.

 

OR represents the change in likelihood associated with being black.

References

. American Cancer Society. (2004)Cancer facts and figures, 2004. American Cancer Society. Atlanta, GA:
 
Scott, WJ, Howington, J, Movsas, B Treatment of stage II non-small cell lung cancer.Chest2003;123,188S-201S. [CrossRef] [PubMed]
 
Smythe, WR Treatment of stage I non-small cell lung carcinoma.Chest2003;123,181S-187S. [CrossRef] [PubMed]
 
Deslauriers, J, Gregoire, J Surgical therapy of early non-small cell lung cancer.Chest2000;117,104S-109S. [CrossRef] [PubMed]
 
Cooley, ME, Jennings-Dozier, K Lung cancer in African Americans: a call for action.Cancer Pract1998;6,99-106. [CrossRef] [PubMed]
 
Wingo, PA, Ries, LA, Giovino, GA, et al Annual report to the nation on the status of cancer, 1973–1996, with a special section on lung cancer and tobacco smoking.J Natl Cancer Inst1999;91,675-690. [CrossRef] [PubMed]
 
Wu, LY, Semenya, KA, Hardy, RE, et al Cancer rate differentials between blacks and whites in three metropolitan areas: a 10-year comparison.J Natl Med Assoc1998;90,410-416. [PubMed]
 
Bach, PB, Cramer, LD, Warren, JL, et al Racial differences in the treatment of early-stage lung cancer.N Engl J Med1999;341,1198-1205. [CrossRef] [PubMed]
 
Smith, TJ, Penberthy, L, Desch, CE, et al Differences in initial treatment patterns and outcomes of lung cancer in the elderly.Lung Cancer1995;13,235-252. [CrossRef] [PubMed]
 
Greenwald, HP, Polissar, NL, Borgatta, EF, et al Social factors, treatment, and survival in early-stage non-small cell lung cancer.Am J Public Health1998;88,1681-1684. [CrossRef] [PubMed]
 
King, TE, Jr, Brunetta, P Racial disparity in rates of surgery for lung cancer.N Engl J Med1999;341,1231-1233. [CrossRef] [PubMed]
 
Mountain, C Revisions in the international system for staging lung cancer.Chest1997;111,1710-1717. [CrossRef] [PubMed]
 
Hosmer, DW, Lemeshow, S Applied logistic regression 2nd ed.2000 John Wiley and Sons. New York, NY:
 
Lemeshow, S, Hosmer, DW, Jr A review of goodness of fit statistics for use in the development of logistic regression models.Am J Epidemiol1982;115,92-106. [PubMed]
 
Lang, T, Secic, M. How to report statistics in medicine. 1997; American College of Physicians. Philadelphia, PA:.
 
Dominitz, JA, Maynard, C, Billingsley, K, et al Race, treatment, and survival of veterans with cancer of the distal esophagus and gastric cardia.Med Care2002;40,I14-I26. [PubMed]
 
Cooper, GS, Yuan, Z, Landefeld, CS, et al Surgery for colorectal cancer: race-related differences in rates and survival among Medicare beneficiaries.Am J Public Health1996;86,582-586. [CrossRef] [PubMed]
 
Cooper-Patrick, L, Gallo, JJ, Gonzales, JJ, et al Race, gender, and partnership in the patient-physician relationship.JAMA1999;282,583-589. [CrossRef] [PubMed]
 
Saha, S, Komaromy, M, Koepsell, TD, et al Patient-physician racial concordance and the perceived quality and use of health care.Arch Intern Med1999;159,997-1004. [CrossRef] [PubMed]
 
Doescher, MP, Saver, BG, Franks, P, et al Racial and ethnic disparities in perceptions of physician style and trust.Arch Fam Med2000;9,1156-1163. [CrossRef] [PubMed]
 
Margolis, ML, Christie, JD, Silvestri, GA, et al Racial differences pertaining to a belief about lung cancer surgery: results of a multicenter survey.Ann Intern Med2003;139,558-563. [PubMed]
 
Wegner, EL, Kolonel, LN, Nomura, AM, et al Racial and socioeconomic status differences in survival of colorectal cancer patients in Hawaii.Cancer1982;49,2208-2216. [CrossRef] [PubMed]
 
Page, WF, Kuntz, AJ Racial and socioeconomic factors in cancer survival: a comparison of Veterans Administration results with selected studies.Cancer1980;45,1029-1040. [CrossRef] [PubMed]
 
Dayal, HH, Polissar, L, Dahlberg, S Race, socioeconomic status, and other prognostic factors for survival from prostate cancer.J Natl Cancer Inst1985;74,1001-1006. [PubMed]
 
Dayal, H, Polissar, L, Yang, CY, et al Race, socioeconomic status, and other prognostic factors for survival from colo-rectal cancer.J Chronic Dis1987;40,857-864. [CrossRef] [PubMed]
 
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 Med1988;318,612-617. [CrossRef] [PubMed]
 
Jazieh, AR, Kyasa, MJ, Sethuraman, G, et al Disparities in surgical resection of early-stage non-small cell lung cancer.J Thorac Cardiovasc Surg2002;123,1173-1176. [CrossRef] [PubMed]
 
Samet, J, Hunt, WC, Key, C, et al Choice of cancer therapy varies with age of patient.JAMA1986;255,3385-3390. [CrossRef] [PubMed]
 
Hoekstra, HJ Cancer surgery in the elderly.Eur J Cancer2001;37(suppl),S235-S244
 
Loudon, RF, Anderson, PM, Gill, PS, et al Educating medical students for work in culturally diverse societies.JAMA1999;282,875-880. [CrossRef] [PubMed]
 
NOTE:
Citing articles are presented as examples only. In non-demo SCM6 implementation, integration with CrossRef’s "Cited By" API will populate this tab (http://www.crossref.org/citedby.html).

Some tools below are only available to our subscribers or users with an online account.

Related Content

Customize your page view by dragging & repositioning the boxes below.

CHEST Journal Articles
  • CHEST Journal
    Print ISSN: 0012-3692
    Online ISSN: 1931-3543