|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
© 2003 American Society for Clinical Oncology Economic Evaluation of Breast Cancer Treatment: Considering the Value of Patient Choice
From the University of Pennsylvania, Philadelphia, PA; Georgetown University, Washington, DC; Dana-Farber Cancer Institute, Boston, MA; and the Center for Clinical and Genetic Economics, Duke Clinical Research Institute, Durham, NC. Address reprint requests to Daniel Polsky, PhD, Division of General Internal Medicine, University of Pennsylvania, Blockley Hall 1212, Philadelphia, PA 19104-6021; email: polsky{at}mail.med.upenn.edu.
Purpose: To use 5 years of primary data to compare the incremental cost-effectiveness of breast conservation and radiation versus mastectomy with the restriction of choice to a single therapy versus providing a choice of either therapy. Patients and Methods: We evaluated a random retrospective cohort of 2,517 Medicare beneficiaries treated for newly diagnosed stage I or II breast cancer from 1992 through 1994. The outcome measures were quality-adjusted life-years (QALYs) and 5-year medical costs. Risk and propensity score adjustments were used in the analysis. Results: A breast conservation and radiation regimen has significantly higher costs than mastectomy in the first year after surgery; the adjusted 5-year costs are $14,054 (95% confidence interval, $9,791 to $18,312) greater than those of mastectomy. The adjusted incremental cost-effectiveness ratio comparing breast conservation and radiation to mastectomy was $219,594 per QALY for the comparison of the two strategies. If the possibility of patient choice from maintaining the availability of multiple treatments versus restricting choice to mastectomy alone provides a quality-of-life gain of 0.031 QALYs, then the cost-effectiveness ratio of this choice option is $80,440 per QALY. Conclusion: The current system of providing a choice between mastectomy and breast conservation surgery is economically attractive when the economic analysis includes the benefit of patient choice of treatment.
BREAST CANCER is the second leading cause of cancer death in American women.1 Most women receiving a diagnosis of breast cancer are 65 years of age or older (hereafter called "older"). With advances in screening and early detection, most cases of breast cancer are diagnosed in early stages, when survival is excellent.1 However, improvements in breast cancer outcomes come at a cost: the annual economic cost for breast cancer treatment in the United States has been estimated at $6 billion.2 Local therapies for early-stage breast cancer include breast-conserving surgery with radiation (BCSRT) and mastectomy. Large clinical trials have studied the efficacy of these treatments.3,4 The consensus from these trials is that BCSRT and mastectomy are equivalent procedures with respect to long-term survival.5 Although BCSRT was once thought to be underused,6,7 more recently the number of women receiving BCSRT as an alternative to mastectomy has increased.8,9 Radiation therapy is recommended after breast-conserving surgery for early-stage breast cancer patients because it has been shown to decrease local recurrence even though it does not improve survival.10 Although none of the efficacy trials included quality-of-life end points, findings from the many observational studies comparing the quality of life after breast conservation and mastectomy have shown no statistically meaningful differences in quality of life, except for minor differences in body image.11–14 Several cost studies have compared surgical treatments for early-stage breast cancer. These studies indicate that BCSRT may be more expensive than mastectomy, but the evidence is not conclusive.15–19 Hayman et al20 compared quality of life and costs between BCSRT and breast-conserving surgery without radiation treatment. They found BCSRT to be cost-effective because of the quality-of-life gains from diminished local recurrence. However, this analysis did not assess the cost-effectiveness of BCSRT versus mastectomy, the two most commonly used approaches to the local therapy of early breast cancer. To compare the relative effect of mastectomy and BCSRT on 5-year quality-adjusted life-years (QALYs), costs, and cost-effectiveness, we performed a retrospective cohort study using data from a national sample of older women with early-stage breast cancer who were treated between 1992 and 1994. We conducted two cost-effectiveness analyses, which included a traditional analysis and an alternative formulation that may be more relevant to health policy. In the traditional cost-effectiveness analysis, we compared mastectomy as the standard of care with BCSRT as the new intervention. In the alternative cost-effectiveness analysis, we assumed that the standard of care is an open regimen in which both BCSRT and mastectomy are widely available and the treatment ultimately delivered is a function of the option most preferred by the patient (or her physician), whereas the new intervention is a restricted regimen with only one procedure available. This alternative formulation allows physicians to consider whether a cost-effectiveness analysis comparing widely used treatment options should be judged simply by the incremental effect on costs and outcomes or whether system values, such as the benefits of being given a choice between different treatment options, should affect the adoption decision.
Study Sample The data were collected as part of the Outcomes and Preferences for Treatment in Older Women Nationwide Survey (OPTIONS) project. The sampling design has been described elsewhere.21 Briefly, we constructed the sample in four steps. First, we obtained Medicare claims for persons with a breast cancer diagnosis or surgery procedure code for calendar years 1992 to 1994 from the Centers for Medicare and Medicaid Services (formerly the Health Care Financing Administration) national claims database of a 5% random sample of all Medicare beneficiaries. Second, to limit the sample to women for whom BCSRT and mastectomy would be considered equivalent, we excluded cases for the following reasons: history of previous cancer diagnosis (4.5%), carcinoma-in-situ diagnosis (3.0%), metastasis diagnosis (0.6%), or bilateral procedure (0.1%). We also excluded cases that lacked a surgical procedure code (4.0%), were missing a physician identifier (1.7%), or occurred in patients younger than 67 years of age (14.2%). As was done in earlier studies,7,22,23 we deleted cases for which breast surgery was not the primary procedure code (16.2%) or for which breast cancer was not the primary diagnosis (13.9%). We also excluded 5.5% of cases because the surgeon could not be identified. Applying these exclusion criteria resulted in a preliminary sample of 10,695 women. Third, we surveyed the identified surgeons to verify study eligibility based on the presence of primary stage I and II invasive disease (ie, tumor size [T], node involvement [N], and metastasis status [M] combinations of T1N0M0, T0–1N1M0, T2N0M0, T2N1M0, and T3N0M0) and the absence of the preceding exclusion criteria. Among respondents, 80.7% provided information for at least one patient. There were 2,907 eligible women from the surgeon survey. Fourth, we applied additional exclusions. We excluded 86 patients who were in a Medicare health maintenance organization in the month of the surgery because their cost data were not available in the claims files. Because the goal of the cost-effectiveness analysis was to compare the two strategies recommended by the National Cancer Institute,5 we also excluded the 304 patients who had breast-conservation surgery who did not receive radiation therapy. The final sample consisted of 2,517 patients.
Key Variables The preference assessment interviews were conducted 3 to 5 years after surgery on 1,320 of the surviving subjects. To address the issue that 32% of the surviving subjects had no preference assessment, we imputed values for the noninterviewed women using a least squares regression of the VAS as a function of treatment group, surgery year, the interaction of surgery year and treatment group, and baseline health status variables.25 To address the issue that no repeated measures were elicited, we estimated an annual value for health-state preferences for each individual by adjusting individual values by changes over time in the mean VAS by treatment group.
Direct Medical Costs Imputations were performed for medical costs in the intervals with missing cost data; that is, for patients not followed by the Medicare claims for all months of the 5-year study period because they moved to health maintenance organization status or because follow-up ended in December 1997. First, we imputed costs assuming survival by using a repeated-measures analysis of variance regression of interval costs estimated among patients who were alive during the interval in which the independent variables were treatment group, interval, interaction between interval and treatment group, and a standard set of explanatory variables. Second, we adjusted for the fact that patients who are no longer observed may not survive by multiplying imputed costs in the interval by the patient’s predicted survival in that interval.
Independent Variables The other independent variables used in this study came from several sources. Patient age, race, Charlson comorbidity index score, and presurgery costs came from claims data. Breast cancer stage was determined from review of physician medical records. Because claims do not contain socioeconomic data, we used the following ZIP code–level data from the 1990 United States Census: the median household income, the percentage of the population with a college education, and the percentage of the population below the poverty level. The Area Resource File was the source of county-level data on health system characteristics, such as hospital admissions, number of nursing homes, and urban area.
Statistical Analysis For each of the outcomes, we report three estimates: one that is unadjusted and two that are adjusted. For each estimate, we report mean 5-year QALYs and medical costs by treatment group, the difference between groups, and the 95% confidence interval (CI) pertaining to the difference. The differences were considered significantly different from 0 if the 95% CI did not include 0. For the first adjusted estimate, we used a least squares regression for 5-year QALYs and costs. The covariates chosen for these regressions, as well as all other multivariate analyses, are variables that are both measurable and theoretically predictive of the outcome. In addition to treatment group, we included age at the time of surgery; cancer stage; Charlson comorbidity index; health care costs in the year before surgery; race; and market-area characteristics of hospital admissions, nursing homes, urban area, college graduates, median income, and poverty level. For the second adjusted estimate, we implemented a propensity score analysis that controls for observable confounders when the impact of treatment may differ for women with a different probability of receiving a particular treatment.26–28 The propensity score (ie, the predicted probability of receiving a particular treatment) was estimated from the logistic regression estimating the treatment received. The sample was stratified into four equal-size propensity groups based on the propensity score. Then the difference in outcomes and costs between mastectomy and BCSRT was estimated for each propensity group in a multivariate least squares regression. The propensity score treatment effect is the average of these propensity group estimates. We estimated separate cost-effectiveness ratios for the traditional analysis and for our alternative formulation of the cost-effectiveness problem. For the traditional analysis, the incremental cost-effectiveness ratio comparing mastectomy and BCSRT was defined as the difference in the mean costs of BCSRT (CBCSRT) and the mean costs of mastectomy (CMST) divided by the difference in the mean QALYs of BCSRT (EBCSRT) and the mean QALYs of mastectomy (EMST). This ratio was estimated from unadjusted, adjusted, and propensity score–adjusted differences in means. We assessed the statistical uncertainty of the cost-effectiveness ratio using CIs estimated by use of the nonparametric bootstrap.29 We also assessed the sensitivity of the cost-effectiveness ratio to the range of estimates of the cost difference between mastectomy and BCSRT that are available in the literature.15,16,19 For the alternative formulation, which estimated the incremental cost-effectiveness ratio for the comparison of the open treatment regimen (in which mastectomy and BCSRT are both available and patients can choose between therapies) versus the restricted regimen (in which surgery is restricted to mastectomy for all patients), we used the following equation:
where CCHOICE is costs incurred (or negative costs saved) because of the availability of choice, ECHOICE is QALYs gained (or lost) because of the availability of choice, and p is the proportion receiving BCSRT. For the restricted regimen, the costs and QALYs are the mean costs and QALYs of mastectomy. For the open regimen, the costs are the weighted average of costs for BCSRT and mastectomy, with weights based on the proportion of the population receiving each option [p and (1 - p)]. Similarly, the QALYs are the weighted average of QALYs for BCSRT and mastectomy using the same weights. For this formulation, we will use the propensity score adjusted estimate of costs and QALYs estimated in the traditional analysis, which assumes no additional advantage to matching a patient to the best option for that patient. We assume that given free choice, a nontrivial percentage of women will select each procedure.30,31 For this analysis, we used our data to estimate p at 28%. The two new terms in this equation, additional cost and QALYs associated with the choice, regardless of the choice taken (CCHOICE and ECHOICE), have been estimated in prior research.32 Polsky et al32 found that choice increased QALYs by 0.03 and decreased costs by $4,000. Because CCHOICE and ECHOICE are more conceptual than concrete, an exact measure is difficult. Thus, in addition to our best estimate (ie, CCHOICE = 0 and ECHOICE = 0.03), we also estimated the range of values for CCHOICE and ECHOICE that result in an incremental cost-effectiveness ratio for the regimen of $25,000, $50,000, $100,000, $150,000, and $219,000 per QALY. In addition, as a sensitivity analysis, we are interested in how sensitive the cost-effectiveness ratio of the comparison of regimes is to p, the proportion of patients selecting BCSRT in the open regimen. Starting with our best estimate, we compute new values for the cost-effectiveness ratio using double the value of p (56%) and using half the value of p (14%).
Patient Characteristics In this study, 1,813 women received mastectomy, and 704 women received BCSRT (Table 1  ). The patients receiving BCSRT were significantly (P < .05) more likely to be younger, have earlier-stage cancer, have lower costs before surgery, have a lower score on the Charlson comorbidity index (ie, fewer comorbid conditions), and have BCSRT in the later years of the study period. They were also more likely to live in areas that were more urban, more affluent, and more highly educated.
QALYs Unadjusted 5-year QALYs are 3.075 for mastectomy and 3.326 for BCSRT. In the regression of QALYs on treatment and baseline characteristics (Table 2 ), we see that older women and women with higher-stage cancer, more medical costs in the previous year, and more comorbid conditions all have lower QALYs. This adjustment results in BCSRT having an effect relative to mastectomy of 0.083 QALYs (P = .093), which represents a substantial reduction from the unadjusted effect of BCSRT (Table 3). The results of the propensity score adjustment (Table 3) are similar to but slightly lower than those of the multivariate regression.
The health-state preferences estimated from the VAS and survival, which make up the components of QALYs, are displayed with their adjusted values in Fig 1 . The suggestion from this figure is that the statistically insignificantly higher QALYs are driven by the slightly higher survival for women who undergo BCSRT and not by any difference in health-state preferences among survivors.
Medical Costs Figure 2 shows the timing of costs for BCSRT and mastectomy. The source of the cost difference occurs in the first year after surgery, with BCSRT incurring first-year costs more than $10,000 higher than mastectomy. After year 1, costs stabilize for both groups at approximately $6,000 per year. The similarity in costs between groups after year 1 suggests no cost offset of one treatment versus the other in this period. The 5-year unadjusted total costs are $40,130 for mastectomy and $48,519 for BCSRT. Unadjusted 1-year differences in total costs available in the literature range from roughly $2,76019 to $5,33515 or 33% to 64% of the estimated cost difference in this study.
The results of the regression analysis predicting 5-year direct medical costs are shown in the last set of columns in Table 2  . The coefficient of BCSRT is $13,775 (P = .000). Other covariates also significantly affect costs, including age 80 to 84 years, higher-stage cancer, higher previous medical costs, and more comorbid conditions.
The differences in 5-year medical costs are summarized in the second panel of Table 3
Cost-Effectiveness Analysis
Figure 3
The results of the alternative formulation of the cost-effectiveness analysis using the estimates of CCHOICE = -$4,000 and ECHOICE = 0.03 produces an incremental cost-effectiveness ratio in which the open regimen dominates over the restricted regimen (ie, the open regimen costs less and yields more QALYs). Using ECHOICE = 0.03 but setting CCHOICE = 0 yields a cost-effectiveness ratio of $80,440. These results are sensitive to the value for p, the percentage of people who choose BCSRT over MST. As more people choose the less efficient alternative (ie, BCSRT), the open regimen becomes less cost-effective relative to the closed regimen (ie, MST only). Starting with the cost-effectiveness ratio of $80,440 at a p = 28%, if p doubles to 56% the cost-effectiveness ratio goes up to $117,747 per QALY, an increase of 46%. If p is halved to 14%, the cost-effectiveness ratio goes down to $49,238, a decline of 39%.
We estimated the costs and quality-adjusted survival for a national random sample of older women with early-stage breast cancer. Over a 5-year period, patients receiving BCSRT incurred greater costs and no significant differences in quality-adjusted survival than patients receiving mastectomy, with a cost-effectiveness ratio of $219,594 per QALY. This indicates that, according to traditional cost-effectiveness analysis, treatment with BCSRT would not be considered an economically attractive option relative to mastectomy for older patients with early-stage breast cancer. However, traditional cost-effectiveness analysis may not produce the most policy relevant cost-effectiveness ratio when both treatment options are widely used as the standard of care, because the ratio indicates that a treatment option should be removed rather than introduced. If BCSRT is removed because it is not cost-effective, surgery for early-stage breast cancer would be restricted to mastectomy only. The alternative formulation of the cost-effectiveness ratio compares the following two policy-relevant scenarios: standard care of multiple options versus the alternative of mastectomy only. This formulation enables analysis of the economic effect of removing one of two therapies from the marketplace. In this case, patients would no longer have a choice between two treatments that widely differ in how patients perceive and prefer them. This results in a utility loss resulting from a more restricted choice; choice itself is of value to consumers because it creates an opportunity to match treatments to individual values and to have control over the course of therapy. Because the traditional approach to cost-effectiveness does not incorporate values for choice and thus fails to account for this utility loss, it underestimates the benefits of providing BCSRT along with mastectomy. Polsky et al32 estimated this utility loss to be 0.031 QALYs over the first year of recovery. When this value is added, the cost-effectiveness ratio of maintaining the current system of providing both BCSRT relative to the system of restricting therapy to mastectomy alone becomes $80,440. Polsky et al32 also found a cost savings from choice of $4,000. When this value is added on top of the QALY gain from choice, the cost-effectiveness ratio indicates that the current regimen dominates the restricted regimen alternative. Estimates of the cost difference between mastectomy and BCSRT are higher in this study than most available estimates.15,16,19 These alternative estimates are not directly comparable with our estimates because our study covers a longer time frame and is a representative national sample of clinically confirmed early-stage breast cancer patients. Nevertheless, if these alternative estimates are incorporated into our cost-effectiveness analysis, the cost-effectiveness ratio drops to a more attractive range (between $72,469 and $140,547 per 5-year QALY). This sensitivity analysis reinforces the conclusion that breast cancer surgery should not be restricted to mastectomy only based on this analysis. This study used observational data instead of clinical trial data; therefore, risk adjustment is critical because of the possibility of confounding by indication. The risk adjustment approaches used here, ordinary least squared (OLS) regression and propensity score adjustment, did account for the substantial amount of observed difference between groups. The negligible change between the OLS-adjusted result and the propensity score result indicates that there is little heterogeneity in treatment effects across different levels of severity. The validity of the risk adjustments is confirmed by the fact that the risk-adjusted, quality-adjusted survival of mastectomy and BCSRT did not statistically differ, a finding consistent with evidence on survival derived from clinical trials3,4 and the observational study evidence on quality of life.11–14 Despite these risk adjustments for observed differences between treatment groups, unobserved risk differences could still influence the differences between groups. We tried to account for unobserved differences using instrumental variables analysis, but we found that the OLS estimates were the preferred estimates.33 If the adjustment of the observational data is not large enough to account for both observed and unobserved differences between treatment groups, the bias that remains is likely to make BCSRT a less favorable option. Thus, the conclusion that BCSRT is not economically attractive would only be strengthened if the bias remained in our estimates. There are other limitations. We did not consider the relative effect of these therapies beyond 5 years. We did not make a distinction between cancer-related costs and non-cancer-related costs. Finally, we did not attempt to measure patient utilities directly. Instead, we proxied for patient utilities using the VAS, which gives a preference value for the current health state rather than a measure of the utility of the current health state. Our study benefited from its ability to analyze a large, national, random sample of Medicare beneficiaries with confirmed local stages of breast cancer, supplemented by preference data collected from survivors. In addition to our finding that both BCSRT and mastectomy should continue to be offered to patients despite the higher costs per QALY of BCSRT versus mastectomy, this study also makes an important methodologic contribution. To make cost-effectiveness analysis relevant for policy, not only does the comparator (that is, the standard of care) need to accurately reflect the true standard of care, but the intervention considered must also reflect the true policy under consideration. Our study highlights the importance of adding the value of choice when approaching a cost-effectiveness analysis of a condition for which it is known that well-informed patients will vary in their preferences for two treatment strategies for the same condition.
We thank Lynne Eggert, FNP, MPH, Project Coordinator for field implementation of the protocol, the Outcomes and Preferences for Treatment in Older Women Nationwide Study (OPTIONS) support staff who made data collection possible, Mathematica Policy Research, Inc, for data management and collection, Mary Cummings, DrPh, of the Agency for Healthcare Research and Quality for administrative and programmatic support and oversight, and the OPTIONS National Advisory Committee for helpful suggestions on the conduct of the project. The OPTIONS team includes: Deborah Axelrod, MD; Anthony Cahan, MD; Marilyn Halper, MPH; and Bert Petersen, Jr, MD, Beth Israel Medical Center, Boston, MA; Frederick Barr, MD, and Luther Gray, Jr, MD, Sibley Memorial Hospital, Washington, DC; Christine Berg, MD, Suburban Hospital; Caroline Burnett, RN, ScD; Lynne Eggert, CRNP, MPH; Claudine Isaacs, MD; Jeanne Mandelblatt, MD, MPH; Jean Mitchell, PhD; Julia Rowland, PhD; Robert Siegel, MD; Theodore Tsangaris, MD; and Qin Wang, Georgetown University Medical Center, Washington, DC; Stephen Edge, MD, Roswell Park Cancer Institute, Buffalo, NY; Karen Gold, PhD, Abt Associates; Jackie Dunmore-Griffith, MD, Howard University Hospital, Washington, DC; Ed Guadagnoli, PhD, Harvard Medical School, Boston, MA; Jack Hadley, PhD, Urban Institute; Nuhad Ibrahim, MD, and Rodger Winn, MD, M.D. Anderson Cancer Center, Houston, TX; Jon Kerner, PhD, National Cancer Institute, Bethesda, MD; Jack Lynch, MD, Washington Hospital Center Cancer Institute, Washington, DC; Neal Meropol, MD, Fox Chase Cancer Center, Philadelphia, PA; Kevin Schulman, MD, MBA, Duke Clinical Research Institute, Duke University, Durham, NC; Ruby Senie, PhD, Mailman School of Public Health, New York, NY; Brenda Shank, MD, PhD, Doctors Medical Center, San Pablo, CA; Rebecca Silliman, MD, MPH, PhD, Boston University Medical Center, Boston, MA; Juliana Simmons, MD, Washington Hospital Center Cancer Institute, Washington, DC; and Jane Weeks, MD, Dana-Farber Cancer Institute, Boston, MA. The OPTIONS National Advisory Committee includes Judith Baigis, RN, PhD, Georgetown University, Washington, DC; Harold Freeman, MD, Harlem Hospital Center, New York, NY; Mary Jo Gibson, MA, and Joanne Lamphere, American Association of Retired Persons, Washington, DC; Bruce Hillner, MD, Medical College of Virginia, Richmond, VA; Amy Langer, MBA, National Alliance of Breast Cancer Organizations, New York, NY; Marc Lippman, MD, Georgetown University Medical Center, Washington, DC; Monica Morrow, MD, Prentice Women’s Hospital, Chicago, IL; Jeanne Petrek, MD, Memorial Sloan-Kettering Cancer Center, New York, NY; Rashida Muhammad, National Council of Negro Women, Washington, DC; and Natalie Davis Spingarn and David Winchester, MD, American College of Surgeons, Chicago, IL.
Supported by grant no. HS08395 from the Agency for Health Care Policy and Research (Agency for Healthcare Research and Quality), United States Department of Health and Human Services grant no. 17-94-J-4212 from the Department of the Army, and cooperative agreement no. U01/CA88283A from the National Cancer Institute, National Institutes of Health, Department of Health and Human Services, Bethesda, MD.
1. Ries LAG, Eisner MP, Kosary CI, et al (eds): SEER Cancer Statistics Review, 1973–1997. Bethesda, MD, National Cancer Institute, 2000 2. Brown ML, Lipscomb J, Snyder C: The burden of illness of cancer: Economic cost and quality of life. Annu Rev Public Health 22:91–113, 2001[CrossRef][Medline] 3. Fisher B, Bauer M, Margolese M, et al: Five years of a randomized clinical trial comparing total mastectomy and segmental mastectomy with or without radiation in the treatment of breast cancer. N Engl J Med 312:665–673, 1985[Abstract] 4. Fisher B, Redmond C, Poisson R, et al: Eight-year results of a randomized clinical trial comparing total mastectomy and lumpectomy with or without irradiation in the treatment of breast cancer. N Engl J Med 320:822–828, 1989[Abstract]
5. National Institutes of Health Consensus Conference: Treatment of early-stage breast cancer. J Am Med Assoc 265:391–396, 1991
6. Lazovich D, White E, Thomas DB, et al: Underutilization of breast-conserving surgery and radiation therapy among women with stage I or II breast cancer. J Am Med Assoc 266:3433–3438, 1991 7. Nattinger AB, Gottlieb MS, Hoffman RG, et al: Minimal increase in the use of breast conserving surgery from 1986–1990. Med Care 34:479–489, 1996[CrossRef][Medline] 8. Arnold RE, Frykberg ER, Kilkenny JW, et al: Trends in surgical treatment of breast cancer at an urban teaching hospital: A six-year review. Am Surg 64:107–111, 1998[Medline]
9. Lazovich D, White E, Thomas DB, et al: Change in the use of breast-conserving surgery in western Washington after the 1990 NIH Consensus Development Conference. Arch Surg 132:418–423, 1997
10. Clark RM, McCulloch PB, Levine MN, et al: Randomized clinical trial to assess the effectiveness of breast irradiation following lumpectomy and axillary dissection for node-negative breast cancer. J Natl Cancer Inst 84:683–689, 1992
11. Pozo C, Carver CS, Noriega V, et al: Effects of mastectomy versus lumpectomy on emotional adjustment to breast cancer: A prospective study of the first year postsurgery. J Clin Oncol 10:1292–1298, 1992 12. Shimozuma K, Ganz PA, Petersen L, et al: Quality of life in the first year after breast cancer surgery: Rehabilitation needs and patterns of recovery. Breast Cancer Res Treat 56:45–57, 1999[CrossRef][Medline] 13. Ganz PA, Schag AC, Lee JJ, et al: Breast conservation versus mastectomy: Is there a difference in psychological adjustment or quality of life in the year after surgery? Cancer 69:1729–1738, 1992[CrossRef][Medline]
14. Rowland JH, Desmond KA, Meyerowitz BE, et al: Role of breast reconstructive surgery in physical and emotional outcomes among breast cancer survivors. J Natl Cancer Inst 92:1422–1429, 2000 15. Desch C, Penberthy L, Hillner B, et al: A sociodemographic and economic comparison of breast reconstruction, mastectomy, and conservation surgery. Surgery 125:441–447, 1999[Medline] 16. Norum J, Olsen J, Wist E: Lumpectomy or mastectomy? Is breast conserving surgery too expensive? Breast Cancer Res Treat 45:7–14, 1997[CrossRef][Medline] 17. Given C, Bradley C, Luca A, et al: Observation interval for evaluating the costs of surgical interventions for older women with a new diagnosis of breast cancer. Med Care 39:1146–1157, 2001[CrossRef][Medline]
18. Barlow WE, Taplin SH, Yoshida CK, et al: Cost comparison of mastectomy versus breast-conserving therapy for early-stage breast cancer. J Natl Cancer Inst 93:447–455, 2001
19. Warren JL, Brown M, Fay MP, et al: Costs of treatment for elderly women with early-stage breast cancer in fee-for-service settings. J Clin Oncol 20:307–316, 2002 20. Hayman JA, Hillner BE, Harris JR, et al: Cost-effectiveness of routine radiation therapy following conservative surgery for early-stage breast cancer. J Clin Oncol 16:1022–1029, 1998[Abstract]
21. Hadley J, Mitchell JM, Mandelblatt J: Medicare fees and small area variations in breast-conserving surgery among elderly women. Med Care Res Rev 58:334–360, 2001 22. Farrow DC, Hunt WC, Samet JM: Geographic variation in the treatment of localized breast cancer. N Engl J Med 326:1097–1101, 1992[Abstract] 23. Nattinger AB, Gottlieb MS, Veum J, et al: Geographic variation in the use of breast-conserving treatment for breast cancer. N Engl J Med 326:1102–1107, 1992[Abstract] 24. The EuroQol Group: EuroQol: A new facility for the measurement of health-related quality of life. Health Policy 16:199–208, 1990[CrossRef][Medline] 25. Lin DY: Linear regression analysis of censored medical costs. Biostatistics 1:35–47, 2000[Abstract] 26. Rosenbaum PR, Rubin DB: Reducing bias in observational studies using subclassification on the propensity score. J Am Stat Assoc 79:516–524, 1984[CrossRef]
27. Connors AF Jr, Speroff T, Dawson NV, et al: The effectiveness of right heart catheterization in the initial care of critically ill patients. J Am Med Assoc 276:889–897, 1996
28. Rubin DB: Estimating causal effects from large data sets using propensity scores. Ann Intern Med 127:757–763, 1997 29. Polsky D, Glick HA, Willke R, et al: Confidence intervals for cost effectiveness ratios: A comparison of four methods. Health Econ 6:243–252, 1997[CrossRef][Medline]
30. Guadagnoli E, Weeks JC, Shapiro CS, et al: Use of breast-conserving surgery for the treatment of stage I and stage II breast cancer. J Clin Oncol 16:101–106, 1998 31. Mandelblatt JS, Hadley J, Kerner JF, et al: Patterns of breast carcinoma treatment in older women: Patient preference and clinical and physical influences. Cancer 89:561–573, 2000[CrossRef][Medline] 32. Polsky D, Keating N, Weeks JC, et al: Patient choice of breast cancer treatment: Impact on health state preferences. Med Care 40:1068–1079, 2002[CrossRef][Medline] 33. Hadley J, Polsky D, Mandelblatt JS, et al: An instrumental variable analysis of the outcomes of localized breast cancer treatments in a medicare population. Health Econ (in press) Submitted March 22, 2002; accepted November 25, 2002.
This article has been cited by other articles:
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||
|
|||||||||||
|
|
Copyright © 2003 by the American Society of Clinical Oncology, Online ISSN: 1527-7755. Print ISSN: 0732-183X
|
TOP
ABSTRACT
INTRODUCTION
2 tests were used for the categorical variables. Using a nominal P value of .05 as the threshold for significance, we report tests that indicated a statistical difference between treatment groups. 
