International Journal for Quality in Health Care 14:411-418 (2002)
© 2002 International Society for Quality in Health Care
Paper |
A first look at variations in use of breast conserving surgery at five teaching hospitals in Japan
Department of Healthcare Economics and Quality Management, School of Public Health, Kyoto University Graduate School of Medicine, Yoshida Konoe-cho, Sakyo-ku, Kyoto 606-8501, Japan
Objective. We assessed variations in the use of breast conserving surgery (BCS) for operable breast cancer patients among hospitals in Japan, and then compared the length of stay (LOS) and total charges during hospitalization between patients who received BCS and those who received mastectomy.
Setting and study participants. We used a database from the Voluntary Hospitals of Japan Quality Indicator Project that involved 10 teaching hospitals in Japan. We selected female operable breast cancer patients who were admitted to five of these hospitals from January 1996 to December 1999 (n = 614).
Main outcome measures. Proportion of the use of BCS, LOS, and total charges during hospitalization.
Results. Twenty-six percent of 614 subjects in the five hospitals received BCS. Proportions of the use of BCS varied from 9% to 51% across five hospitals during the 4-year period. Multiple logistic regression analysis revealed that when we selected as a reference one hospital that had the same proportion of the use of BCS (26%) as the average proportion among all hospitals, three hospitals were 0.3, 2.0, and 2.6 times more likely to use BCS than the reference (P < 0.05). LOS for BCS [mean 25.0 days, standard deviation (SD) 11.8 days] was significantly shorter than for mastectomy (mean 27.3 days, SD 8.6 days), and total charges during hospitalization for BCS (mean US$7771.5, SD 2676.7) were significantly lower than for mastectomy (mean US$8502.5, SD 2044.0). Linear mixed models confirmed that the use of BCS was significantly associated with shorter LOS (P < 0.001) and lower total charges (P < 0.001).
Conclusion. This preliminary description of breast cancer care in five teaching hospitals in Japan revealed variations in the use of BCS.
Keywords: breast cancer, breast conserving surgery, cost, Japan, length of stay, mastectomy
After several published articles showed no significant difference in prognosis for early breast cancer patients treated with mastectomy and breast conserving surgery (BCS) [1–3], the use of mastectomy has decreased and the use of BCS has increased in western countries [4–7]. In Japan, the type of surgical procedures for operable breast cancer has also changed during the past 20 years. According to data from the breast cancer registry of the Japan Breast Cancer Society [8], in which about one-quarter of Japanese hospitals that provided breast cancer surgery (313/1312) participated from 1995 to 1997 [9], the proportion of BCS has increased from 12.7% in 1991 to 29.2% in 1997, whereas the proportion of mastectomy has decreased from 86.7% in 1991 to 64.7% in 1997.
We would like to discuss whether or not there is any variation in the use of BCS among hospitals in Japan, and whether or not the use of resources during hospitalization, such as length of stay (LOS) and total costs, differs between BCS and mastectomy. To the best of our knowledge, there has been no study in Japan to assess variations in the use of BCS for breast cancer across hospitals, or to examine resource use for breast cancer surgery, although the use of BCS and its variations has been vigorously examined in the United States [10–12]. Thus, in this study, we preliminarily assessed variations in the use of BCS and examined factors associated with the use of BCS at five teaching hospitals in Japan. We then compared LOS and total charges for hospital care between patients who received BCS and those who received mastectomy in these hospitals.
Methods
Database
We used a database from the Voluntary Hospitals of Japan Quality Indicator Project, which contained data from 10 privately owned and leading teaching hospitals in Japan. These hospitals are located in urban cities in Hokkaido (in the north), throughout Honshu (the main island of Japan), and in Kyushu (in the south). Data on all discharged cases were collected from these hospitals from 1995. In this particular study, without previous information on the usage rate of BCS in each hospital, we selected five hospitals that had >50 breast cancer patients admitted from January 1996 to December 1999. All five hospitals provided community residents with tertiary care. The average number of general beds in the five hospitals was 671 (range 450–925) in the year 2000. All had full-time surgeons and four out of five had radiotherapy machines. The average number of surgeons who worked in the hospitals was 10.6 (range 7–13) in 2000. Each hospital in this study had similar characteristics to teaching hospitals. In particular, all five hospitals were teaching hospitals for the board certification for surgeons accredited by the Japan Surgical Society. Because the database involved all in-patient admissions, patients with both primary and recurrent breast cancer were included in the study population. To select the study population, we used the International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM). We also used the all-female breast neoplasm diagnostic code (174.0–174.9).
Definition of variables
Variables used in this study were hospital, age, surgical procedure, comorbidity status, LOS, and total charges during hospitalization. The surgical procedure was classified as either BCS or mastectomy. To determine the type of breast cancer surgery performed, we used ICD-9-CM procedure codes. We used BCS (85.21–85.23) and mastectomy (85.41–85.48), and used an adaptation of the Charlson Comorbidity Index to assess comorbidities [13,14]. A patient was identified as having comorbidity if she had any of the following diseases coded in her diagnosis: myocardial infarction, congestive heart disease, peripheral vascular disease, cerebrovascular disease, dementia, chronic pulmonary disease, rheumatological disease, peptic ulcer disease, mild liver disease, diabetes, diabetes with chronic complications, hemiplegia or paraplegia, renal disease, any malignancy (other than breast), moderate or severe liver disease, metastatic solid tumor, or acquired immunodeficiency syndrome. Total charges for the study subjects were calculated by summing any charge billed during hospital stay (US$1 = ¥120).
The five hospitals had a total of 1488 female breast cancer admissions over a 4-year period. We excluded patients who did not receive surgical treatment (862), patients who died during hospital stay (one), and patients with distant metastasis (11). Thus, a total of 614 patients who received surgical treatment were analyzed in this study.
Statistical analysis
Data were analyzed for continuous variables by either the Student’s t-test or the Mann–Whitney test, and for categorical variables by the 
2 test. We examined trends in the use of BCS between year classes by using the Cochran–Armitage test for trends. A multiple logistic regression analysis was performed to identify factors associated with the use of BCS. In the logistic regression analysis, a dependent variable was the type of surgical procedure. Independent variables were patient’s age, comorbidity status, time period of admission, and four dummy variables for the five hospitals. An association between the use of BCS and independent variables was described by an odds-ratio (OR) and a 95% confidence interval (CI). We performed the goodness-of-fit tests developed by Hosmer–Lemeshow to measure how well the model fits the data [15]. When we compared LOS and total charges between BCS and mastectomy, we first used univariate analyses, and then a linear mixed model to identify factors associated with LOS or total charges [16]. In the linear mixed model, a dependent variable was either LOS log-transformed or total charges log-transformed, and independent variables were age, type of surgical procedure, presence of comorbidity, and LOS as fixed variables, in addition to hospital as a random variable. All analytical procedures were performed using the SAS statistical package (SAS Institute Inc., Cary, NC) [17]. All reported P values were two-tailed, and the level of significance was P < 0.05.
Results
Among 614 subjects, 162 patients (26.4%) received BCS and 452 patients (73.7%) had a mastectomy. Table 1 shows subject’s characteristics, such as age and comorbidity status. Patients who received BCS were significantly younger than those who received mastectomy (P = 0.001). There was no patient with congestive heart failure, peripheral vascular disease, diabetes with chronic complications, hemiplegia or paraplegia, moderate or severe liver disease, or acquired immunodeficiency syndrome included in the comorbidity index.
|
Table 2 shows trends for the proportion of BCS use among operable breast cancer patients between 1996 and 1997, and 1998 and 1999. Among all five hospitals, the proportion of BCS use varied from 9% (hospital A) to 51% (hospital E). In addition, there was a significant increase in the proportion of BCS use (P < 0.001), from 18% in 1996–97 to 34% in 1998–99 among all hospitals; however, the proportion of the use of BCS in each hospital did not show a statistically significant increase. Table 3 shows the results of multiple logistic regression analysis to identify factors associated with provision of BCS among all operable breast cancer patients in the five hospitals. This analysis revealed that patients who were <50 years old (P = 0.008) and admitted into hospital in 1998 and 1999 (P = 0.004) were significantly more likely to receive BCS. In addition, the logistic model showed that compared with hospital B, hospital A (P < 0.001) was significantly less likely to use BCS, and hospital D (P = 0.030) and hospital E (P < 0.001) were significantly more likely to use BCS.
|
|
Table 4 shows differences in LOS and total charges during hospitalization, between BCS and mastectomy, by hospital. As for comparisons among all hospitals, patients who received BCS had significantly shorter LOS (P = 0.002) and incurred less total charges (P = 0.003) than those who received mastectomy. As for each comparison within each hospital, the average LOS in hospitals A and B for BCS was significantly shorter than that for mastectomy (P < 0.001), whereas the average LOS for BCS in hospital D was significantly longer than that for mastectomy (P = 0.025). The average LOS for mastectomy in each hospital was between 25 and 29 days. With respect to the distribution of the total charges, the average total charges for BCS in hospitals A and B were significantly lower than for mastectomy. Table 5 shows results from the linear mixed model to identify factors associated with LOS in the five hospitals. This analysis indicated that the use of BCS was significantly associated with shorter LOS (P < 0.001). Table 5 also shows factors associated with total charges during hospitalization at the five hospitals. Advanced age (P < 0.001) and the use of BCS (P < 0.001) were significantly associated with lower total charges, whereas longer LOS was significantly associated with higher total charges (P < 0.001).
|
|
Discussion
This study preliminarily identified variations in the use of BCS for operable breast cancer patients in five privately owned teaching hospitals located in urban cities in Japan; each hospital in this study had characteristics similar to a teaching hospital. Whereas one hospital provided BCS for only 9% of operable breast cancer patients admitted to the hospital, the other four hospitals had higher proportions of BCS (26–51%) during the 4-year period. The proportion of BCS use in the five hospitals increased significantly between 1996–97 and 1998–99. Multiple logistic regression analysis revealed that patients aged <50 years and patients admitted into hospital in 1998–99 were significantly associated with use of BCS. When comparing LOS between BCS and mastectomy among all subjects, LOS for BCS was significantly shorter than that for mastectomy, but the difference was only 2 days. Total charges during hospitalization for BCS in the five hospitals were significantly lower than those for mastectomy. The linear mixed model revealed that the use of BCS was significantly associated with shorter LOS. The linear mixed model also showed that patients aged 
50 years accounted for lower total charges during hospital stay.
This study confirmed variations in the use of BCS among five teaching hospitals in Japan. Hospitals A, D, and E were 0.3, 2.0, and 2.6 times more likely, respectively, to use BCS than hospital B (Table 3). Patient, hospital, and surgeon have also been reported to be associated with the use of BCS [18–20]. As one of the patient factors, this study showed that women aged <50 years were 1.7 times more likely to receive BCS than those aged 50 years. This result is concordant with several studies, mainly reported from the United States, which identified age as a strong factor associated with the use of BCS [6,7,18,19,21,22].
As for hospital factors, geographic variation has been reported in the use of BCS [6,10,18,19]. Athas et al. [23] reported a significant inverse relationship between travel distance and radiotherapy following BCS. In this study, because hospitals A, C, D, and E had radiation therapy machines, it is easy for us to conclude that patients had good access to BCS with radiation therapy, and this might have resulted in higher proportions of BCS use. However, this was not the case in hospital A. Although hospital A had a radiation facility, the proportion of BCS therapy in hospital A was <10% during the 4-year period.
Here, we hypothesize that hospital A’s surgeons tended to select surgical procedures for breast cancer: surgeons in hospital A might have prefered mastectomy to BCS. Tarbox et al. [24] investigated surgeons’ general beliefs that BCS gave an equal chance for survival as mastectomy for patients with stage I breast cancer, in Colorado in the early 1990s. They showed that not all surgeons believed that BCS and mastectomy had an equivalent survival rate. Surgeons who did not believe that BCS was equivalent to mastectomy in terms of survival performed less BCS than those who believed that BCS was equivalent to mastectomy. In Japan, because there have been no clinical trials to examine whether early breast cancer patients are equally well treated with BCS as with mastectomy, we believe that the decision to use BCS for early breast cancer patients might depend on the preferences of surgeons and/or patients.
This study also revealed a significant difference in average LOS for operable patients between BCS (25.0 days) and mastectomy (27.3 days); however, the difference in average LOS between the two procedures was only 2 days. Patients who received BCS stayed in hospital for about 4 days before surgery and they remained in hospital for about 20 days after surgery (Table 4). There are a few possible explanations for the approximately 4-week hospital stay among operable breast cancer patients in this study. Firstly, because Japanese hospitals provide post-operative patients with acute and post-acute care during the same hospital care episode, Japanese surgeons are likely to let post-operative patients stay in hospitals until they can recover their abilities to perform the same activities of daily living as they did before surgery. Secondly, because Japanese surgeons are frequently trained to examine patients’ surgical wounds every day by removing sutures, they may believe that it is beneficial to let post-operative patients stay in hospital until removing sutures. Therefore, one of the main reasons for post-operative treatment during hospitalization is simply to change dressings. Finally, there is no financial risk for Japanese surgeons in letting their patients stay in hospital for 4 weeks, because charges during admission are basically reimbursed on a fee-for-service basis.
This study showed that the average of total charges during admission for BCS was significantly lower than that for mastectomy. The results from the linear mixed model also indicated that the use of BCS and shorter LOS were correlated with lower total charges during hospitalization. Because in Japan total charges during hospitalization are calculated by adding up each point-fee of the procedure provided during hospitalization (e.g. diagnostic tests, imaging, prescriptions, injections, surgery, anesthesia, radiotherapy, in addition to room, board, nursing, and physician’s management on a daily basis), it is easy for us to assume that the shorter a patient stays in hospital, the lower the total charges during hospitalization will be. On the other hand, because the point-fees of lumpectomy (US$217 or 458) and subtotal mastectomy (US$717 or 1500) are lower than those for mastectomy (US$1325, 1500, or 2033), we assume that the difference in the point-fee for surgical procedure may also be associated with the difference in total charges during hospital stay between BCS and mastectomy. Because total hospital charges under the Japanese point-fee system do not reflect the actual costs of interventions, it would be much more helpful to know the actual costs of surgical treatment of breast cancer.
Some limitations must be considered when interpreting the results of this study. Firstly, as a general limitation of this study, the study subjects were women who received surgical treatment in five teaching hospitals in Japan, rather than simply those who were clinically eligible for either BCS or mastectomy. We do not have information on each subject with respect to tumor size, staging, or other pathological features of the breast cancer. In particular, we were unable to select out recurrent breast cancer patients in this study. Because a recurrent breast cancer probably has a higher rate of mastectomy, the proportion of patients who underwent BCS in this study may be biased toward underestimation. Thus, the data presented in this study are crude and give an overall picture of breast cancer treatment only in selected hospitals in Japan. Secondly, as these data were analyzed in a cross-sectional way, we were not able to follow each patient’s readmission and identify total medical charges for 6–12 months after surgical treatment. Finally, the hospitals selected in this study are not representative of all hospitals in Japan. However, we consider that the average LOS in the five teaching hospitals shown in this study was not biased towards a longer stay, whereas it was much longer than that found in other OECD countries [25]. Hasegawa [26] selected 136 hospitals, with an average LOS of <20 days for all discharged patients, from 6449 hospitals in the 1996 Patient Survey [27]. For operative breast cancer patients in those 136 hospitals, the average LOS and average stay before surgery were estimated as 30.3 and 5.3 days, respectively. These LOS indicators are similar to our study subjects.
In conclusion, this study confirmed variations in the use of BCS for breast cancer among five teaching hospitals in Japan. We need to continue to observe the use of BCS for the treatment of breast cancer over a longer period, and to examine both surgeon and patient behavior and beliefs regarding breast cancer treatment to explain variations in the use of BCS.
The authors are grateful to the participants of the Voluntary Hospitals of Japan Quality Indicator Project for their cooperation with this study: Teine Keijinkai Hospital (Sapporo, Hokkaido), Nikko Memorial Hospital (Muroran, Hokkaido), Takeda General Hospital (Aizu Wakamatsu, Fukushima), Tachikawa General Hospital (Nagaoka, Niigata), Kameda Medical Center (Kamogawa, Chiba), Kawakita General Hospital (Suginami, Tokyo), Keiju General Hospital (Nanao, Ishikawa), Rakuwa-kai Otowa Hospital (Kyoto, Kyoto), Hosoki Hospital (Kochi, Kochi), and Iizuka Hospital (Iizuka, Fukuoka). This study was supported in part by the Health Sciences Research Grants for the Research on Policy Planning and Evaluation from the Ministry of Health, Labour and Welfare of Japan (H13-Policy-030).
Address reprint requests to T. Ishizaki, Department of Healthcare Economics and Quality Management, School of Public Health, Kyoto University Graduate School of Medicine, Yoshida Konoe-cho, Sakyo-ku, Kyoto 606-8501, Japan. E-mail: tatsuro{at}pbh.med.kyoto-u.ac.jp 
Accepted for publication May 24, 2002.
REFERENCES
- Fisher B, Bauer M, Margolese R et al. Five-year results 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 1985; 312: 665–673.[Abstract]
- 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 1989; 320: 822–828.[Abstract]
- Fisher B, Anderson S, Redmond CK, Wolmark N, Wickerham DL, Cronin WM. Reanalysis and results after 12 years of follow-up in a randomized clinical trial comparing total mastectomy with lumpectomy with or without irradiation in the treatment of breast cancer. N Engl J Med 1995; 333: 1456–1461.
[Abstract/Free Full Text] - Nattinger AB, Gottlieb MS, Hoffman RG, Walker AP, Goodwin JS. Minimal increase in use of breast-conserving surgery from 1986 to 1990. Med Care 1996; 34: 479–489.[Web of Science][Medline]
- Lazovich D, White E, Thomas DB, Moe RE, Taplin S. Change in the use of breast-conserving surgery in western Washington after the 1990 NIH Consensus Development Conference. Arch Surg 1997; 132: 418–423.
[Abstract/Free Full Text] - Du X, Freeman JL, Freeman DH, Syblik DA, Goodwin JS. Temporal and regional variation in the use of breast-conserving surgery and radiotherapy for older women with early-stage breast cancer from 1983 to 1995. J Gerontol 1999; 54: M474–M478.
- Morris CR, Cohen R, Schlag R, Wright WE. Increasing trends in the use of breast-conserving surgery in California. Am J Public Health 2000; 90: 281–284.
[Abstract/Free Full Text] - Takashima S. Progress in treatment for breast cancer (in Japanese). J Jpn Med Assoc 2001; 125: 1707–1717.
- Health and Welfare Statistics Association. Health State of the Nation, Annual Report 2000. Tokyo, Japan: Kosei Tokei Kyokai, 2000 (in Japanese).
- Farrow DC, Hunt WC, Samet JM. Geographic variation in the treatment of localized breast cancer. N Engl J Med 1992; 326: 1097–1101.[Abstract]
- 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: 1102–1107.[Abstract]
- Sarr MG, Donohue JH, Grant CS, van Heerden JA, McIlrath DC. Breast-conserving surgery and radiation therapy: are they underused? J Am Med Assoc 1992; 268: 869–870.
[Abstract/Free Full Text] - Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis 1987; 40: 373–383.[Web of Science][Medline]
- Deyo RA, Cherkin DC, Ciol MA. Adapting a clinical comorbidity index for use with ICD-9-CM administrative databases. J Clin Epidemiol 1992; 45: 613–619.[Web of Science][Medline]
- Hosmer D, Lemeshow S. Applied Logistic Regression. New York, NY: John Wiley & Son, Inc., 1989, pp. 140–145.
- Verbeke G, Molenberghs G. Linear Mixed Models in Practice. New York, NY: Springer, 1997.
- SAS Institute Inc. SAS/STAT Software: Changes and Enhancements through Release 6.12. Cary, NC: SAS Institute Inc., 1997.
- Samet JM, Hunt WC, Farrow DC. Determinants of receiving breast-conserving surgery. The Surveillance, Epidemiology, and End Results Program, 1983–1986. Cancer 1994; 73: 2344–2351.[Web of Science][Medline]
- Johantgen ME, Coffey RM, Harris DR, Levy H, Clinton JJ. Treating early-stage breast cancer: hospital characteristics associated with breast-conserving surgery. Am J Public Health 1995; 85: 1432–1434.
[Abstract/Free Full Text] - 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: 716–726.
[Abstract/Free Full Text] - 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: 419–426.[Web of Science][Medline]
- Kotwall C, Covington D, Churchill P, Brinker C, Weintritt D, Maxwell JG. Breast conservation surgery for breast cancer at a regional medical center. Am J Surg 1998; 176: 510–514.[Web of Science][Medline]
- Athas WF, Adams-Cameron M, Hunt WC, Amir-Fazli A, Key CR. Travel distance to radiation therapy and receipt of radiotherapy following breast-conserving surgery. J Natl Cancer Inst 2000; 92: 269–271.
[Free Full Text] - Tarbox BB, Rockwood JK, Abernathy CM. Are modified radical mastectomies done for T1 breast cancers because of surgeon’s advice or patient’s choice? Am J Surg 1992; 164: 417–420.[Web of Science][Medline]
- Organization for Economic Cooperation and Development. OECD Health Data 2000. Paris: OECD Publication Service, 2000.
- Hasegawa T. The Strategy of Shortening Length of Hospital Stay for Hospital Management (in Japanese). Tokyo, Japan: Igaku Shoin Co. Ltd, 2001.
- Ministry of Health and Welfare. Patient Survey 1996 (in Japanese). Tokyo, Japan: Health and Welfare Statistics Association, 1998.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||