A, Surgeon volume. B, Surgeon percentage; 2 surgeons with a percentage greater than 60% were excluded. The sum of the first 4 percentages (24%, 30%, 17%, and 8%) shows that 79% of surgeons performed less than 10% of their operative cases for treatment of breast cancer. The remaining 21% of surgeons performed 10% or more of their operative cases for treatment of breast cancer.aPercentage values less than 1%.
Annual Medicare surgeon volume and the OR of undergoing SLNB vs ALND are depicted relative to the baseline odds of undergoing SLNB for a surgeon who performed 6 annual incident Medicare breast cancer operations and 3% of all surgical procedures for treatment of breast cancer. At each of the surgeon volumes (6, 12, and 24 cases), 4 surgeon percentages are shown. For each box, the black line represents the OR estimate, the box represents the 50% CI, and the vertical error bars represent the 95% CI.
Yen TWF, Laud PW, Sparapani RA, Nattinger AB. Surgeon Specialization and Use of Sentinel Lymph Node Biopsy for Breast Cancer. JAMA Surg. 2014;149(2):185-192. doi:10.1001/jamasurg.2013.4350
Sentinel lymph node biopsy (SLNB) is the standard of care for axillary staging in patients with clinically node-negative breast cancer. It is not known whether SLNB rates differ by surgeon expertise. If surgeons with less breast cancer expertise are less likely to offer SLNB to these patients, this practice pattern could lead to unnecessary axillary lymph node dissections and lymphedema.
To explore potential measures of surgical expertise (including a novel objective specialization measure: percentage of a surgeon’s operations performed for breast cancer determined from Medicare claims) on the use of SLNB for invasive breast cancer.
Design, Setting, and Population
A population-based prospective cohort study was conducted in California, Florida, and Illinois. Participants included elderly (65-89 years) women identified from Medicare claims as having had incident invasive breast cancer surgery in 2003. Patient, tumor, treatment, and surgeon characteristics were examined.
Main Outcome and Measure
Type of axillary surgery performed.
Of 1703 women who received treatment by 863 surgeons, 56.4% underwent an initial SLNB, 37.2% initial axillary lymph node dissection, and 6.3% no axillary surgery. The median annual surgeon Medicare volume of breast cancer cases was 6.0 (range, 1.5-57.0); the median surgeon percentage of breast cancer cases was 4.5% (range, 0.4%-100.0%). After multivariable adjustment of patient and surgeon factors, women operated on by surgeons with higher volumes and percentages of breast cancer cases had a higher likelihood of undergoing SLNB. Specifically, women were most likely to undergo SLNB if the operation was performed by high-volume surgeons (regardless of percentage) or by lower-volume surgeons with a high percentage of breast cancer cases. In addition, membership in the American Society of Breast Surgeons (odds ratio, 1.98; 95% CI, 1.51-2.60) and Society of Surgical Oncology (1.59; 1.09-2.30) were independent predictors of women undergoing an initial SLNB.
Conclusions and Relevance
Patients who receive treatment from surgeons with more experience with and focus on breast cancer are significantly more likely to undergo SLNB, highlighting the importance of receiving initial treatment by specialized providers. Factors relating to specialization in a particular area, including our novel surgeon percentage measure, require further investigation as potential indicators of quality of care.
Over the past 2 decades, sentinel lymph node biopsy (SLNB) has been the single most important advance in the primary surgical treatment of clinically node-negative breast cancer. Compared with axillary lymph node dissection (ALND), which involves the removal of most axillary lymph nodes, SLNB involves the removal of only a few lymph nodes (on average, 2-3 nodes).1- 5 Several large randomized clinical trials demonstrated that SLNB is associated with similar disease-free survival when compared with ALND1,4- 9 as well as a reduced likelihood of developing lymphedema and other arm/shoulder morbidity.1,4- 6,10- 17 For these reasons, SLNB is now the standard of care for axillary staging in patients with clinically node-negative breast cancer18- 20 and is considered a surgical quality measure by certain organizations.21- 23
Many women with breast cancer in the United States are cared for by surgeons with low volumes of breast cancer cases.17,24- 26 However, it is not known whether patients of these surgeons are as likely to undergo SLNB. If surgeons with less breast cancer expertise are less likely to offer SLNB to patients with clinically node-negative breast cancer, this practice pattern could lead to unnecessary ALNDs and cases of lymphedema.
In this report, we evaluate the relationship between surgeon characteristics and axillary surgery for invasive breast cancer by exploring several potential measures of surgeon expertise. One measure, surgeon volume of breast cancer cases, has been used as a measure of technical proficiency with regard to margin status and reexcision rates after breast-conserving surgery, use of axillary surgery, and adequacy of ALND.27- 30 For the present study, we developed a novel objective measure of surgical specialization: the percentage of a surgeon’s operative cases for treatment of breast cancer as determined from Medicare claims data. Prior studies29- 32 have evaluated the relationship of self-reported percentages of a surgeon’s practice focused on breast surgery to the type of breast surgery performed, adequacy of ALND, and patient satisfaction. Other measures of expertise investigated in the present study included the number of years in practice and membership in surgical oncology and breast-specific societies. We explored these 5 measures of surgical expertise and receipt of initial axillary surgery in a population-based cohort of older breast cancer survivors.
The study sample consisted of a population-based cohort of elderly breast cancer survivors who participated in a National Cancer Institute–sponsored survey study examining breast cancer care outcomes. Women between the ages of 65 and 89 years were identified from Medicare claims as having had an incident breast cancer surgery in 2003, using a validated claims-based algorithm.33 Details regarding study recruitment and assessments have been previously described.17,34 Briefly, potential participants were contacted by mail in September 2005, and 4 subsequent annual structured telephone interviews were conducted. For the present study, the cohort consisted of 1703 women with invasive breast cancer who completed all 4 waves of the survey, had complete information regarding pathologic nodal status, and had a surgeon identifiable by Medicare claims.
This study was approved by the Centers for Medicare and Medicaid Services Privacy Board, as well as our institution’s and each tumor registry’s institutional review board. Participants provided written informed consent and received $20 for each completed survey.
Data on tumor characteristics and stage were provided by state cancer registries.35 Medicare claims information was collected from inpatient, outpatient, and carrier Standard Analytical Files. Surgeon characteristics were obtained from the American Medical Association’s professional database (Physician Masterfile)36 (Table 1).37,38 Society membership was determined from online directories of the Society of Surgical Oncology (SSO)39 and the American Society of Breast Surgeons (ASBrS).40 Surgeons were not contacted directly for this study.
Table 1 summarizes the various patient, tumor, treatment, and surgeon characteristics examined and the source of information. Medicare claims data were used to determine the type of axillary surgery performed.41- 43 Patients were classified into 1 of 3 groups: (1) no axillary surgery; (2) initial SLNB, which includes patients who underwent only SLNB and those who went on to completion ALND either during the same operation or in subsequent surgery; and (3) initial ALND. Procedural codes in 2003 for SLNB included injection of the radioactive tracer or dye (Current Procedural Terminology [CPT] codes A9520, 38790, and 38792) and removal of sentinel lymph nodes (CPT codes 38500 and 38525). The ALND procedural codes included those that referred only to axillary surgery (CPT codes 38740 and 38745) and those that included combined breast and axillary procedures (CPT codes 19162, 19200, 19220, and 19240).
Provider codes from Medicare claims of the surgeons operating on cohort patients were linked to the 2004 American Medical Association’s Physician Masterfile to obtain surgeon characteristics. Surgeons were considered to have an academic affiliation if they performed most of their breast operations in an Association of American Medical Colleges Council of Teaching Hospitals and Health Systems facility,44 were employed by a medical school, or primarily practiced medical research or education. Medicare claims were used to determine annual Medicare surgeon volume of breast cancer cases, as previously described,17,45 based on claims for all breast cancer cases treated in each state, not solely for cohort patients.
The novel objective measure (surgeon percentage of practice devoted to breast cancer cases) was created. For each surgeon, the annual number of patients who underwent an initial breast cancer surgery was divided by the number of patients who underwent a general surgery operation performed that same year. General surgery operations were defined by CPT codes 10021 to 69990 and all patient-refined diagnosis related groups surgical P codes.42,46 This specialization measure was based on Medicare claims for all surgical procedures performed in each state, not solely for cohort patients.
The conceptual framework was to model the decision to perform axillary surgery and, if the procedure was done, which type of axillary surgery (initial SLNB or ALND) was performed. We hypothesized that surgeon characteristics would affect the type of axillary surgery in a fashion incremental to clinically important preoperative factors. Variables included in these analyses were determined a priori: 5 surgeon characteristics (surgeon percentage, surgeon volume, membership in ASBrS or SSO, and years since medical school graduation), 3 patient characteristics (age; body mass index [BMI], calculated as weight in kilograms divided by height in meters squared; and comorbidities), and geographic location. Tumor characteristics (clinical stage, hormone receptor, and ERBB2 [formerly HER2/neu] status) could not be included because this information was not consistently recorded in the state tumor registries.
The 2 axillary surgery outcomes were analyzed by separate multiple logistic regression modeling. The first model included the entire cohort and examined the independent effects of various surgeon and patient characteristics on the likelihood of undergoing axillary surgery (initial SLNB or ALND vs no axillary surgery). In the subset of women who underwent axillary surgery, the second model examined the relationship between these same characteristics and the likelihood of undergoing initial SLNB vs ALND. Data analyses were conducted using SAS, version 9.3 (SAS Institute Inc).
The mean (SD) age of the 1703 women at the time of surgery was 72.9 (5.6) years; 90.7% were white. At the time of surgery, 34.9% of the women were underweight or normal weight (BMI, <25), 26.8% were overweight (BMI, 25 to <30), 19.4% were obese (BMI, ≥30), and the BMI in 18.9% of the patients was unknown. Most (62.6%) women were healthy, with a National Cancer Institute combined comorbidity score of 0; 31.5% had a score greater than 0, and the score in 5.9% was unknown. Most women had early-stage disease: 69.9% had category T1 tumors (tumor size <2 cm); 18.8% had category T2 tumors (2-5 cm), and 2.6% had tumors larger than 5 cm; 8.6% had no reported tumor size. There were 21.7% of the women with lymph node involvement. Almost two-thirds (62.5%) underwent breast-conserving surgery; 37.5% underwent mastectomy. No axillary surgery was performed in 6.3% of the participants; 56.4% underwent an initial SLNB and 37.2% received an initial ALND. The mean number of lymph nodes removed was 7.1 (6.7), with a range of 0 to 44 (interquartile range, 9 [2-11]). Women undergoing ALND had a mean of 9.2 (6.8) nodes removed compared with 3.0 (3.2) for those undergoing SLNB (P < .001). Approximately two-thirds (67.1%) of the cohort received radiotherapy, 23.0% received chemotherapy, and 66.9% received hormonal therapy.
Table 2 details the characteristics of the 863 surgeons who operated on these 1703 women. Most were male, graduated from a US medical school, and were general surgeons. Few were surgical oncologists, had an academic affiliation, or were ASBrS or SSO members. Figure 1 shows the distribution of surgeon volume and surgeon percentage. The median annual Medicare volume was 6.0 cases (range, 1.5-57.0; interquartile range, 7.5 [3.0-10.5]). Overall, 40.7% of surgeons performed fewer than 6 incident Medicare breast cancer cases annually, which equates to fewer than 15 cases in women of all ages.
Among all Medicare operations performed by the 863 surgeons during 1 year, the median percentage of breast cancer cases was 4.5% (range, 0.4%-100.00%; interquartile range, 6.2% [2.6%-8.8%]). Twenty-one percent of the surgeons (n = 181) conducted 10% or more of their operations to treat breast cancer (Table 2 and Figure 1B) and performed more than twice as many breast cancer operations annually as surgeons devoting less than 10% of their operations to breast cancer (mean, 19.7 vs 7.5) (Table 2). The surgeons with the higher percentage of breast cancer operations were more likely to be female, to be a surgical oncologist, to be a member of the ASBrS and SSO, and to have an academic affiliation. Only 8% and 4% of surgeons performed more than 20% and 30% of their operations to treat breast cancer, respectively.
Of the entire cohort, 93.7% (n = 1595) of women underwent either SLNB or ALND and 6.3% (n = 108) underwent no axillary surgery. In the multiple logistic regression model, when adjusting for surgeon and patient characteristics, the only independent predictor of undergoing any axillary surgery (SLNB or ALND) was patient age. Compared with women aged 65 to 69 years, women aged 80 years or older were significantly less likely to undergo any axillary surgery (odds ratio [OR], 0.21; 95% CI, 0.12-0.36). Patient BMI, comorbidity, and all 5 surgeon characteristics were not predictors of a woman’s likelihood of undergoing axillary surgery.
Among the 1585 women who underwent axillary surgery, 60.4% underwent initial SLNB and 39.6% received ALND. Women operated on by surgeons who were members of the ASBrS or SSO were more likely to undergo SLNB (Table 3). Because there was no interaction between ASBrS and SSO membership (P = .92), the effects of society membership are additive. Women undergoing an operation by surgeons who were members of both the ASBrS and SSO had a 3.14 OR of undergoing SLNB (95% CI, 2.02-4.87).
Women operated on by surgeons who had higher breast cancer volumes and higher percentages of breast cancer cases were also more likely to undergo SLNB. These 2 effects, however, were not additive, as indicated by an interaction between these 2 variables (P < .001). Figure 2 displays the relationship between surgeon volume, surgeon percentage, and the OR of undergoing SLNB vs ALND. All ORs are in reference to the baseline odds for a surgeon who performs 6 breast cancer cases annually, with 3% of all surgical procedures for treatment of breast cancer (baseline surgeon). For a low-volume surgeon (annual volume of 6 Medicare cases), the likelihood of a patient undergoing SLNB increased significantly with increasing surgeon percentage of breast cancer cases. Compared with the OR of the baseline surgeon, the OR for a patient undergoing an initial SLNB ranged from 1.05 (95% CI, 1.02-1.09) if her surgeon’s percentage of breast cancer operations was 5% to 1.96 (1.03-3.05) if the surgeon’s percentage was 30%. As surgeon volume increased, the incremental effect of surgeon percentage on the OR of receiving an SLNB diminished. If a surgeon performed 12 Medicare breast cancer cases annually, the OR of a patient receiving an SLNB ranged from 1.40 (95% CI, 1.25-1.56) to 2.09 (1.44-3.05) as her surgeon’s percentage rose from 5% to 30%. At a high surgeon volume of 24 annual Medicare cases (which equates to approximately 60 annual cases of all ages), the effect of surgeon proportion was negligible.
Similarly, the incremental effect of surgeon volume on surgeon percentage was the greatest at the lower surgeon percentages (5% and 10%). As the surgeon percentage increased, the effect of surgeon volume became minimal. When the surgeon percentage was 30% (only 4% of the surgeons in this cohort), the incremental effect of surgeon volume was negligible. With increasingly higher surgeon volume and surgeon percentage values, the CIs widened significantly; caution must be used in interpreting the model results at the high end of surgeon volume and surgeon percentage values.
In this population-based cohort of more than 1700 elderly invasive breast cancer survivors treated by 863 predominantly community-based general surgeons, 56.4% of the women underwent an initial SLNB procedure, 37.2% underwent an initial ALND, and 6.3% underwent no axillary surgery. The surgeon characteristics evaluated in this study were measurable factors that we believed would reflect experience with breast cancer (case volume, years since medical school graduation) or degree of focus or specialization in breast cancer (percentage of total surgical procedures performed for breast cancer, ASBrS and SSO membership). None of these 5 surgeon factors was predictive of a woman’s likelihood of undergoing axillary surgery. However, if a woman underwent axillary surgery, surgeon percentage of breast cancer surgical procedures, case volume, and membership in either ASBrS or SSO were associated with the likelihood of a woman undergoing an SLNB, controlling for patient age, BMI, and comorbidity.
Specialization has been found to be beneficial in other areas of health care delivery (eg, critical care intensivists47,48). Our study results would support some degree of specialization as benefitting patients with breast cancer. Surgeons who dedicate a larger percentage of their practice to breast cancer or are members of breast and surgical oncology societies may have a greater motivation to become facile with treatment advances such as SLNB. However, it would be difficult from a health policy perspective to base provider recommendations on professional society membership because it is a surgeon’s choice to join professional societies and membership requirements vary by society. Therefore, we favor the objective measure of percentage of a surgeon’s practice that is focused on breast cancer (specialization construct), in addition to surgeon volume of breast cancer cases (experience construct), as a potential surgical quality marker.
Both higher surgeon volume and higher surgeon percentage of breast cancer cases were associated with a higher likelihood of a patient undergoing an initial SLNB. However, this relationship between surgeon volume and surgeon percentage is somewhat complex (Figure 2). As one variable increases (eg, surgeon volume), the effect of the other variable (eg, surgeon percentage) decreases. The incremental effect of higher surgeon percentage on a high-volume surgeon’s likelihood of performing an SLNB is marginal, but the incremental effect of higher surgeon percentage on a low-volume surgeon’s likelihood of performing an SLNB is substantial. This finding could partially explain why some low-volume surgeons have good outcomes if they are high-percentage surgeons. Similarly, the incremental effect of surgeon volume on a high-percentage surgeon’s likelihood of performing an SLNB is minor, but the incremental effect of surgeon volume on a low-percentage surgeon’s likelihood of performing SLNB is significant. We believe that this novel surgeon percentage measure further refines the surgeon volume measure and vice versa and warrants further investigation as a potential quality marker for surgical outcomes.
This study has several limitations. First, several tumor characteristics are typically known preoperatively and affect a surgeon’s decision to recommend axillary surgery. Information regarding the clinical tumor stage prior to surgery, hormone receptor, and ERBB2 status were not routinely recorded in the 2003 state tumor registries; therefore, we could not conclusively determine which patients were appropriate candidates for SLNB. However, these factors are unlikely to vary systematically by surgeon characteristics. Second, misclassification of the outcome is possible because this axillary surgery variable was claims based and defined by billing codes available in 2003. However, Medicare claims are generally accurate for breast cancer surgical billing codes.49,50
Finally, because this study involved Medicare patients who received treatment in 2003, a period when SLNB was considered an option to ALND for women with early-stage breast cancer51,52 but was already widely used in clinical practice,53- 55 our findings may not be generalizable to younger and more contemporary populations. Surprisingly, only 6.3% of the women in this elderly cohort did not undergo any axillary surgery, and if axillary surgery was performed, older women were more likely to receive ALND rather than SLNB. Studies in more current cohorts of elderly women should be performed because these findings are in contrast to the National Comprehensive Cancer Network guidelines, which consider both SLNB and ALND optional in elderly patients.19,56,57 However, the fact that all patients were Medicare beneficiaries is a strength because we could identify surgeons on the basis of claims information, calculate surgeon volume and percentage of breast cancer cases, and perform our analyses of various surgeon characteristics by linkage with other sources. An observational study of this magnitude affords us an opportunity to evaluate practice patterns in the community, where most breast cancer treatment occurs.
Several previously reported29- 32,58,59 measures of surgeon specialization in breast cancer have been based on subjective measures: local reputation, opinion, and surgeon self-reported proportion of their total practice focused on breast cancer–related procedures. We believe that our novel objective measure, computed from health care claims, is likely to be a more reliable and sensitive surgeon-specific measure than these subjective measures and warrants further evaluation as a potential indicator of other aspects of breast cancer quality of care. Finally, this surgeon percentage/specialization measure may help refine studies that include surgeon volume in evaluating outcomes.
In summary, among this population-based large cohort of elderly invasive breast cancer survivors, women operated on by surgeons who perform more breast cancer cases, focus their practice more on breast cancer, or are members of breast and surgical oncology societies were more likely to undergo an initial SLNB. Because women who undergo SLNB are at lower risk of developing lymphedema than are those who undergo ALND, the importance of receiving initial treatment at centers that have experienced SLNB teams19 and surgeons who specialize in breast cancer must be emphasized to decrease lymphedema rates and its associated complications, which significantly affect the quality of life and medical costs of survivors.13,16,60,61 Studies addressing provider specialization are needed to determine whether this novel objective surgeon percentage measure is an indicator of improved outcomes and quality of cancer and noncancer care.
Accepted for Publication: June 29, 2013.
Corresponding Author: Tina W. F. Yen, MD, MS, Division of Surgical Oncology, Center for Patient Care and Outcomes Research, Medical College of Wisconsin, 9200 W Wisconsin Ave, Milwaukee, WI 53226 (firstname.lastname@example.org).
Published Online: December 25, 2013. doi:10.1001/jamasurg.2013.4350.
Author Contributions: Dr Yen had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of data analysis.
Study concept and design: All authors.
Acquisition of data: Sparapani.
Analysis and interpretation of data: All authors.
Drafting of the manuscript: Yen, Laud.
Critical revision of the manuscript for important intellectual content: All authors.
Statistical analysis: Laud, Sparapani.
Obtained funding: Yen, Nattinger.
Study supervision: Yen.
Conflict of Interest Disclosures: None reported.
Funding/Support: This research was supported by a Career Development Award and supplement K07CA125586 and K07CA125586-03S1 (Dr Yen) from the National Cancer Institute and by National Cancer Institute research grants R01CA81379 and R01CA127648 (Dr Nattinger).
Role of the Sponsors: The National Cancer Institute had no role in the design and conduct of the study; collection, management, analysis, or interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.
Disclaimer: The content does not necessarily represent the official views of the National Cancer Institute or the National Institutes of Health.
Previous Presentation: This study was presented in part as an oral presentation at the Sixth Annual Academic Surgical Congress; February 3, 2011; Huntington Beach, California.
Additional Contributions: Changbin Guo, PhD, Mikesh Shivakoti, MS, and Jianing Li, MS, conducted data analyses, and Thomas Chelius, MS, assisted with the acquisition and collection of data. They received no financial compensation for these services.