Variations in the treatment of early-stage breast cancer in Quebec between 1988 and 1994

Nicole Hébert-Croteau,* MD, PhD; Jacques Brisson,†‡ MD, ScD; Jean Latreille,¶ MDCM; Caty Blanchette,† MSc; Luc Deschênes,‡§ MD

CMAJ 1999;161(8):951-5

• Abstract
• Introduction
• Methods
• Results
• Interpretation
• References

Background: The influence of organizational factors on the process and outcomes of the treatment of breast cancer has been extensively investigated. Although the quality of care is presumed to be better in larger centres, evidence is inconsistent. This study was conducted to determine whether therapies for patients with breast cancer varied according to hospital caseload.

Methods: Women newly diagnosed between 1988 and 1994 with early-stage node-negative primary breast cancer were randomly selected from the Quebec tumour registry and the Quebec hospital discharge database. Data were collected from medical charts, and only women having undergone dissection of the axilla were included in the analyses. Logistic regression analysis was used to adjust for case mix and organizational variables.

Results: The final sample included 1259 patients with node-negative stage I or II pri-mary breast cancer. The proportion of women who underwent breast-conserving surgery increased significantly with hospital caseload (from 78.0% in hospitals admitting fewer than 25 new cases each year to 88.0% in those admitting 100 patients or more; p for trend < 0.001). This trend remained significant even after statistical adjustment for case mix and organizational factors (p for trend = 0.001). Of the 1039 women who underwent breast-conserving surgery 965 (92.9%) received radiotherapy. Use of systemic adjuvant therapy (tamoxifen or chemotherapy, or both) increased with the number of patients treated in a given centre (from 60.1% to 68.5%), but this trend disappeared after adjustment for case mix and other factors. The proportion of patients receiving systemic adjuvant therapy consistent with published consensus guidelines tended to increase with caseload for those treated in hospitals participating in multicentre clinical trials but decrease with caseload for patients in hospitals not involved in clinical research.

Interpretation: The care of patients in Quebec with early-stage breast cancer is characterized by a high prevalence of both breast-conserving surgery and systemic adjuvant therapy. Large centres, especially those actively involved in clinical research, rapidly adopt innovative therapeutic modalities.

[Contents]

Organizational factors related to the patterns of treatment and treatment outcomes in oncology and, more specifically, those related to breast cancer have been thoroughly investigated. Hospital characteristics that have been reported to significantly affect treatment and outcomes include the teaching status^{1,2,3,4,5,6,7,8,9,10,11} and location^{1,8,12,13,14,15} of the hospital, patient volume,^{5,7,8,12,14,16,17,18,19,20,21,22} and the onsite availability of specialized services needed for treatment.^{1,7,10,20,23,24} In addition, more favourable outcomes have been demonstrated among women treated by physicians who have larger caseloads or more experience with patients with breast cancer.^{16,18,25,26,27,28}

The influence of organizational factors on the management of breast cancer in Canada was investigated by Iscoe and colleagues;²⁹ they analysed the number of patients undergoing breast-conserving surgery by county and by hospital in Ontario and found a strong hospital effect but no effect of hospital caseload. In a survey of breast cancer cases in Ontario and British Columbia, Goel and associates¹¹ reported a significant association between breast-conserving surgery and radiation therapy and the academic affiliation of both the hospital and the attending physician, but not with the number of beds in a given hospital. Finally, in a report on systemic adjuvant therapy for breast cancer in British Columbia, Olivotto and colleagues²⁴ concluded that compliance with provincial practice guidelines was more likely to occur in a cancer centre than in a community hospital.

The goal of this study was to investigate whether treatment for breast cancer in hospitals in Quebec varied by hospital caseload and whether patients treated in larger centres were more likely to receive optimal treatment.

[Contents]

Details on sample selection and data collection are provided elsewhere.^30,31 In brief, patients with breast cancer living in various regions of Quebec were randomly selected from the Quebec tumour registry for the periods Apr. 1 to Mar. 31, 1988/89 and 1991/92, and from the Quebec hospital discharge database, Apr. 1 to Mar. 31, 1993/94. The sample was restricted to women who underwent dissection of the axilla and had node-negative stage I or II disease. The sample was stratified according to whether the hospital participated in multicentre clinical trials; trials sponsored by the pharmaceutical industry were excluded.

Data were obtained from medical charts and other sources (e.g., oncology registries and pharmacy databases) after the necessary approvals were obtained from the Commission d'accès à l'information du Québec, the directors of professional services and the research ethics committees of the collaborating institutions. Information was abstracted about the patient, the attending physician, the hospital of primary care, characteristics of the tumour and the treatments received within 6 months of diagnosis for patients who received chemotherapy or within 12 months for those who received radiation and tamoxifen therapy. Locoregional management consistent with consensus recommendations was defined according to the National Institutes of Health³² as total mastectomy with lymph node dissection or breast-conserving surgery with both dissection of the axilla and radiotherapy. We assessed the consistency of systemic adjuvant therapy with published standards using the St. Gallen 1992 consensus statement for node-negative cases³³ as a reference. Patients under experimental protocols were automatically classified as having had treatment consistent with standards. The criteria defined by the American Joint Committee on Cancer³⁴ were used for staging, and the highest grade assigned by any method was used in the analysis. This information was missing for 33.5% of the cases, however. Data on breast cancer caseload for each hospital were obtained from the Quebec tumour registry; caseload was divided into 4 categories: fewer than 25 newly diagnosed cases of breast cancer per year, 25­49 new cases, 50­99 new cases and more than 100 cases.

The association of each treatment with caseload was assessed by univariate methods and multivariate logistic regression analysis. Data were examined using successive categorizations of caseload into 2, 4, 5, 6 and 16 groups. Multivariate models also included patients' age, tumour size and grade, histology and estrogen receptor level, availability of onsite radiotherapy, number of years in practice of the attending physician and the date and location of treatment. Because of colinearity between hospital caseload and teaching status, the teaching status of the hospital was omitted from the analysis. Variances were adjusted for the large sampling fraction.³⁵ Results of categorical analyses were contrasted with those of generalized additive models with smoothing splines³⁶ using data from individual patients and adjusting for differences in case mix. Trends with caseload and the frequency of specific therapies were assessed by visual inspection of the curves. When results of both types of multivariate analyses were consistent aggregated data on treatment according to caseload are presented.

[Contents]

The study population included 1259 women with invasive breast cancer; 421 (33.4%) had their cancer diagnosed in 1988/89, 377 (29.9%) in 1991/92 and 461 (36.6%) in 1993/94. In all, 374 (29.7%) were less than 50 years of age, 653 (51.9%) were between 50 and 69 years, and 232 (18.4%) were 70 years or older. Two-thirds (66.7% [840]) had stage I cancer, and about one-third (31.3% [394]) had stage II cancer; staging could not be done for 25 (2.0%) patients because of insufficient information in the charts. Adjustment for the sampling scheme did not substantially modify the results and was therefore ignored.

The distribution of cancer patients by hospital caseload was as follows: 168 patients (13.3%) were admitted to hospitals that saw fewer than 25 new cases of breast cancer each year, 344 (27.3%) to centres that admitted between 25 and 49 cases, 287 (22.8%) to hospitals that admitted between 50 and 99 cases, and 460 (36.5%) to hospitals with 100 or more new cases of breast cancer each year. Of the 1259 women 186 (14.8%) were cared for by a physician with less than 10 years of experience, and 206 (16.4%) were cared for by one who had been in practice for 30 years or more. Overall, 1039 (82.5%) of the women had breast-conserving surgery, and among them 965 (92.9%) also received radiotherapy. Adjuvant systemic therapy with tamoxifen or chemotherapy, or both, was prescribed to 809 (64.3%) women.

The proportion of patients receiving each of the treatments considered in this study tended to increase with hospital caseload (Table 1). This trend was most significant for patients who underwent breast-conserving surgery: from 131 (78.0%) of 168 patients in centres with fewer than 25 new cases each year to 405 (88.0%) of 460 cases in hospitals admitting 100 patients or more (p for trend < 0.001). The trend remained highly significant even after differences in case characteristics and other organizational factors were taken into account (p for trend = 0.001). Among the women who had breast-conserving surgery the proportion who also received radiotherapy was very high and varied little with hospital caseload (Table 1).

Definitive locoregional treatment and the use of systemic therapy increased with caseload (from 92.9% to 96.1% and from 60.1% to 68.5%, respectively), but neither trend was significant in the multivariate analysis (p for trend = 0.10 and 0.08, respectively). The same was true when patients receiving systemic adjuvant therapy consistent with published guidelines were compared by hospital caseload (p for trend after adjustment = 0.70).

The association between treatment for breast cancer and hospital caseload was compared for hospitals that were and were not involved in multicentre clinical trials (Table 2). Of the 1259 patients 682 (54.2%) were admitted to hospitals involved in multicentre trials. Despite the small number of patients in some of the groups created when patients were compared across hospital status and caseload, the trends detected were similar, except for patients who received systemic treatment consistent with published guidelines. After adjustment for differences in case mix the proportion of patients receiving systemic treatment consistent with guidelines increased with caseload in research hospitals, whereas the opposite trend was observed in centres not involved in clinical research (Table 3).

[Contents]

Our results suggest that the care of patients with breast cancer in Quebec hospitals varies according to hospital caseload. However, except for patients who underwent breast-conserving surgery, trends with caseload were substantially reduced after statistical adjustment for differences in case mix and characteristics of the hospitals and physicians.

Many studies support the notion that a large hospital caseload is associated with the adoption of therapeutic innovations and discourages inappropriate interventions;^{5,8,12,16,17,19,20,21,22} not all studies support this notion, however.^{7,11,14,18,29,37} The strong attenuation of most effects observed in this study after adjusting for potential confounding variables highlights both the relation between caseload and other hospital- and physician-related variables and the need for appropriate analytic strategies.³⁸

Surgical management of breast cancer was the treatment most significantly associated with hospital caseload after adjustment. The clinical significance of these results and the associated policy implications can be debated. It has been shown that women receiving systemic adjuvant therapy often will agree to undergo treatments that may be of only small benefit to them.³⁹ We believe that a 10% difference (i.e., from hospitals with the smallest to those with the largest caseloads) in the probability of having breast- conserving surgery could be an important determinant in where treatment is sought.

One of the most striking observations of our study is the extremely high rate of breast-conserving surgery performed among these early-stage cancer patients: 78.0% to 88.0% of patients. This contrasts with data from the United States for all cases of breast cancer: 30% of patients underwent breast-conserving surgery between 1983 and 1986,¹ as did 21.2% of privately insured patients and 27.2% of Health Maintenance Organization patients between 1983 and 1990.⁴⁰ In 1991, 67.6% of node-negative patients in Ontario and 43.8% in British Columbia underwent breast-conserving surgery.¹¹

Limitations of our study include the restricted availability of the data recorded in medical charts, which was likely to be more substantial for systemic adjuvant treatment than for surgical interventions. In addition, comorbidity was not assessed, and the adjustment for differences in case mix between institutions may have been incomplete. Women admitted for primary treatment in smaller centres in Quebec were usually referred to larger hospitals for subsequent evaluation, and this probably reduced the differences reported between hospitals in the proportion of patients receiving systemic adjuvant therapy. Finally, we did not evaluate survival or disease-free survival as related to caseload because the duration of follow-up would still be too short. This, however, remains the ultimate "gold standard" against which practice patterns should be judged.

In conclusion, our data suggest that large centres, especially those actively involved in clinical research, rapidly adopt innovative therapeutic modalities for the treatment of breast cancer and support both the promotion of clinical research and the amalgamation of health care resources to ensure optimal care of women with breast cancer.

We thank Andrée Christen, Ghislaine Gascon, Aline Pelletier and Diane Roy for the supervision of data collection and review of medical charts, and Brigitte Simard for technical assistance.

This study was funded by the Fonds de la recherche en santé du Québec.

Competing interests: None declared.

Send a letter to the editor Envoyez une lettre à la rédaction

[Contents]

Reprint requests to: Dr. Nicole Hébert-Croteau, Direction de santé publique de Montréal-centre, 1301, rue Sherbrooke E, Montréal QC H2L 1M3; fax 514 528-2512; ncroteau@santepub-mtl.qc.ca

1. Nattinger AB, Gottlieb MS, Veum J, Yahnke D, Goodwin JS. Geographic variation in the use of breast-conserving treatment for breast cancer. N Engl J Med 1992;326:1102-7. [MEDLINE]
2. Basnett I, Gill M, Tobias JS. Variations in breast cancer management between a teaching and a non-teaching district. Eur J Cancer 1992;28A:1945-50.
3. Studnicki J, Schapira DV, Bradham DD, Clark RA, Jarrett A. Response to the National Cancer Institute alert. The effect of practice guidelines on two hospitals in the same medical community. Cancer 1993;72:2986-92. [MEDLINE]
4. Tanaka H, Hiyama T, Hanai A, Fujimoto I. Interhospital differences in cancer survival: magnitude and trend in 1975­1987 in Osaka, Japan. Jpn J Cancer Res 1994;85:680-5. [MEDLINE]
5. Lee-Feldstein A, Anton-Culver H, Feldstein PJ. Treatment differences and other prognostic factors related to breast cancer survival. Delivery systems and medical outcomes. JAMA 1994;271:1163-8. [MEDLINE]
6. Foster RS, Farwell ME, Costanza MC. Breast-conserving surgery for breast cancer: patterns of care in a geographic region and estimation of potential applicability. Ann Surg Oncol 1995;2:275-80. [MEDLINE]
7. Scorpiglione N, Nicolucci A, Grilli R, Angiolini C, Belfiglio M, Carinci F, et al. Appropriateness and variation of surgical treatment of breast cancer in Italy: when excellence in clinical research does not match with generalized good quality care. J Clin Epidemiol 1995;48:345-52. [MEDLINE]
8. Johantgen ME, Coffey RM, Harris R, Levy H, Clinton JJ. Treating early-stage breast cancer: hospital characteristics associated with breast-conserving surgery. Am J Public Health 1995;85:1432-4. [MEDLINE]
9. Richards MA, Wolfe CDA, Tilling K, Barton J, Bourne HM, Gregory WM. Variation in the management and survival of women under 50 years with breast cancer in the South East Thames region. Br J Cancer 1996;73:751-7. [MEDLINE]
10. Ballard-Barbash R, Potosky AL, Harlan LC, Nayfield SG, Kessler LG. Factors associated with surgical and radiation therapy for early stage breast cancer in older women. J Natl Cancer Inst 1996;88:716-26. [MEDLINE]
11. Goel V, Olivotto I, Hislop TG, Sawka C, Coldman A, Holowaty EJ, et al. Patterns of initial management of node-negative breast cancer in two Canadian provinces. CMAJ 1997;156(1):25-35. [MEDLINE]
12. Hand R, Sener S, Imperato J, Chmiel JS, Sylvester J, Fremgen A. Hospital variables associated with quality of care for breast cancer patients. JAMA 1991;266:3429-32. [MEDLINE]
13. Howe HL, Katterhagen JG, Yates J, Lehnherr M. Urban­rural differences in the management of breast cancer. Cancer Causes Control 1993;3:533-9.
14. Grilli R, Repetto F. Variation in use of breast-conserving surgery in Lombardia, Italy. Int J Technol Assess Health Care 1995;11:733-40. [MEDLINE]
15. Young WW, Marks SM, Kohler SA, Hsu AY. Dissemination of clinical results. Mastectomy versus lumpectomy and radiation therapy. Med Care 1996;34:1003-17. [MEDLINE]
16. Satariano ER, Swanson M, Moll PP. Non-clinical factors associated with surgery received for treatment of early-stage breast cancer. Am J Public Health 1992;82:195-8. [MEDLINE]
17. Grilli R, Mainini F, Penna A, Bertolini G, Scorpiglione N, Torri V, et al. Inappropriate Halsted mastectomy and patient volume in Italian hospitals. Am J Public Health 1993;83:1762-4. [MEDLINE]
18. Hynes DM. The quality of breast cancer care in local communities: implications for health care reform. Med Care 1994;32:328-40. [MEDLINE]
19. Kotwall CA, Covington DL, Rutledge R, Churchill MP, Meyer AA. Patient, hospital, and surgeon factors associated with breast conservation surgery. A statewide analysis in North Carolina. Ann Surg 1996;224:419-29. [MEDLINE]
20. Nattinger AB, Gottlieb MS, Hoffman RG, Walker AP, Goodwin JS. Minimal increase in use of breast-conserving surgery from 1986­1990. Med Care 1996;34:479-89. [MEDLINE]
21. Ma M, Bell J, Campbell S, Basnett I, Pollock A, Taylor I. Breast cancer management: Is volume related to quality? Br J Cancer 1997;75:1652-9. [MEDLINE]
22. Roohan PJ, Bickell NA, Baptiste MS, Therriault GD, Ferrara EP, Siu AL. Hospital volume differences and five-year survival from breast cancer. Am J Public Health 1998;88:454-7. [MEDLINE]
23. Raabe NK, Kaaresen R, Fossaa SD. Hospital-related differences in breast cancer management. Breast Cancer Res Treat 1997;43:225-35. [MEDLINE]
24. Olivotto IA, Coldman AJ, Hislop TG, Trevisan CH, Kula J, Goel V, et al. Compliance with practice guidelines for node-negative breast cancer. J Clin Oncol 1997;15:216-22. [MEDLINE]
25. Sainsbury R, Rider L, Smith A, MacAdam A. Does it matter where you live? Treatment variation for breast cancer in Yorkshire. Br J Cancer 1995;71:1275-8. [MEDLINE]
26. Sainsbury R, Haward B, Rider L, Johnston C, Round C. Influence of clinician workload and patterns of treatment on survival from breast cancer. Lancet 1995;345:1265-70. [MEDLINE]
27. Gillis CR, Hole DJ. Survival outcome of care by specialist surgeons in breast cancer: a study of 3786 patients in the west of Scotland. BMJ 1996;312:145-8. [MEDLINE]
28. Dixon JM, Ravisekar O, Cunningham M, Anderson EDC, Anderson TJ, Brown HK. Factors affecting outcome of patients with impalpable breast cancer detected by breast screening. Br J Surg 1996;83:997-1001. [MEDLINE]
29. Iscoe NA, Goel V, Wu K, Fehringer G, Holowaty EJ, Naylor CD. Variation in breast cancer surgery in Ontario. CMAJ 1994;150(3):345-52. [MEDLINE]
30. Hébert-Croteau N, Brisson J, Latreille J, Blanchette C, Deschênes L. Compliance with consensus recommendations for the treatment of early-stage breast cancer in elderly women. Cancer 1999;85:1104-13. [MEDLINE]
31. Hébert-Croteau N, Brisson J, Latreille J, Blanchette C, Deschênes L. Time trends in systemic adjuvant treatment for node-negative breast cancer. J Clin Oncol 1999;17:1458-66. [MEDLINE]
32. National Institutes of Health. Treatment of early-stage breast cancer. NIH Consensus Conference. JAMA 1991;265:391-5. [MEDLINE]
33. Glick JH, Gelber RD, Goldhirsch A, Senn HJ. Meeting highlights: adjuvant therapy for primary breast cancer. J Natl Cancer Inst 1992;84:1479-85. [MEDLINE]
34. American Joint Committee on Cancer. Breast. In: Beahrs OH, Henson DE, Hutter RVP, Kennedy BJ, editors. Manual for staging of cancer. 4th ed. Chicago: J.B. Lippincott Company; 1992. p. 149-54.
35. Levy PS, Lemeshow S. Sampling for health professionals. Belmont (CA): Life Learning Publications; 1980.
36. Greenland S. Dose­response and trend analysis in epidemiology: alternatives to categorical analysis. Epidemiology 1995;6:356-65. [MEDLINE]
37. Bonnet A, Roder D, Esterman A. Case-survival rates for infiltrating ductal carcinomas by category of hospital at diagnosis in South Australia. Med J Aust 1991;154:695-7. [MEDLINE]
38. Luft HS, Garnick DW, Mark DH, McPhee SJ. Hospital volume, physician volume, and patient outcomes. Ann Arbor (MI): Health Administration Press; 1990.
39. Ravdin PM, Siminoff IA, Harvey JA. Survey of breast cancer patients concerning their knowledge and expectations of adjuvant therapy. J Clin Oncol 1998;16:515-21. [MEDLINE]
40. Osteen RT, Winchester DP, Hussey DH, Clive RE, Freidman MA, Cady B. Insurance coverage of patients with breast cancer in the 1991 Commission on Cancer Patients Care Evaluation Study. Ann Surg Oncol 1994;1:462-7. [MEDLINE]

© 1999 Canadian Medical Association or its licensors