Abstract
OBJECTIVES
To explore the relationship between level and type of comorbidity and guideline-concordant care for early-stage breast cancer.
DESIGN
Cross-sectional.
SETTING
National Program of Cancer Registry (NPCR) Breast and Prostate Cancer Patterns of Care study, which re-abstracted medical records from 2004 in seven cancer registries.
PARTICIPANTS
Individuals with stage 0–III breast cancer.
MEASUREMENTS
Multicomponent guideline-concordant management was modeled based on tumor size, node status, and hormone receptor status, according to consensus guidelines. Comorbid conditions and severity were measured using the Adult Comorbidity Evaluation Index (ACE-27). Multivariate logistic regression models determined factors associated with guideline-concordant care and included overall ACE-27 scores and 26 separate ACE comorbidity categories, age, race, stage, and source of payment.
RESULTS
The study sample included 6,439 women (mean age 58.7, range 20–99; 76% white; 44% with no comorbidity; 70% estrogen- or progesterone-receptor positive, or both; 31% human epidermal growth factor receptor 2 positive). Care was guideline concordant in 60%. Guideline concordance varied according to overall comorbidity burden (70% for none; 61% for minor; 58% for moderate, 43% for severe; P < .05). In multivariate analysis, the presence of hypertension (odds ratio (OR) = 1.15, 95% confidence interval (CI) = 1.01–1.30) predicted guideline concordance, whereas dementia (OR = 0.45, 95% CI = 0.24–0.82) predicted lack of guideline concordance. Older age (≥50) and black race were associated with less guideline concordance, regardless of comorbidity level.
CONCLUSION
When reporting survival outcomes in individuals with breast cancer with comorbidity, adherence to care guidelines should be among the covariates.
Keywords: breast cancer, comorbidity, guideline concordance
In individuals with breast cancer, comorbidity is linked to lower overall and disease-specific survival.1,2 This association may be through effects on disease biology, receipt of or response to treatment, or to other factors. Treatment of breast cancer requires multicomponent care, including surgery, radiation therapy (RT), endocrine therapy (ET), and chemotherapy (CTX). With adjuvant CTX,3,4 toxicity is a concern.5 Adjuvant RT after breast-conserving surgery (BCS) is less often used in the presence of comorbidity.6 Certain comorbidities may also have different effects on survival, disease course, and treatment.1,7 Prior studies examining the relationship between guideline care and comorbidity did not study dose-response or specific comorbidities.1,3,8,9 The use of the Adult Comorbidity Evaluation Index (ACE-27), which includes 26 comorbidities and disease severity, adds breadth and power to the analysis.10
Decades of clinical research led to multidisciplinary guidelines that minimize recurrence and extend survival in women with breast cancer,11,12 but guidelines do not account for comorbidity, and the relationship between guideline concordance and comorbidity is understudied. The authors of the current study are aware of only one report of multidisciplinary care, which used the Charlson Comorbidity Index derived from administrative data.13 Furthermore, it is important to try to disentangle the independent effects of age and comorbidity and not just assume that it is morbidity burden in elderly adults that effects concordance. This study aimed to examine the relationship between comorbidity, with attention to overall burden and specific comorbid conditions, and receipt of multicomponent guideline-concordant care. It was hypothesized that age and comorbidity burden would have independent effects on guideline concordance.
METHODS
Data Source
The National Program of Cancer Registry (NPCR) Breast and Prostate Cancer Data Quality and Patterns of Care Study (POC-BP) study14,15 included individuals with breast cancer diagnosed in 2004 from seven population-based cancer registries (California, Georgia, Kentucky, Louisiana, Minnesota, North Carolina, Wisconsin). Human subjects committee approval was obtained from each participating institution. Re-abstraction of hospital and physician office records supplemented registry data on initial course of treatment. Additional data on adjuvant therapy and comorbidities were obtained from physicians and outpatient facilities. Cases were randomly selected across strata of race and ethnicity and state-specific factors (e.g., Appalachian vs non-Appalachian region, type of facility, and patient volume of the facility) and oversampled for racial and ethnic minorities in some states. Cases from Veterans Affairs hospitals and identified solely from death certificates or autopsies were excluded.
Study Sample
The initial sample included 9,142 women with breast cancer. Exclusion criteria were missing stage (n = 1,601); metastatic cancer (n = 358); missing tumor size, if nodal status negative or hormone receptor status missing (n = 94); missing nodal status (n = 185); no primary breast tumor (T0, n = 22); and missing hormone receptor status and did not receive ET (n = 418). Cases in which RT (n = 105), ET (n = 257), and CTX (n = 218) and breast (n = 15) or node (n = 14) surgeries were unknown were also excluded. This excluded 2,703 women, leaving a final sample of 6,439. Characteristics of included women were different from those excluded with regard to payor (more Medicare, less Medicaid), race (more blacks, fewer Hispanics) and stage (more Stage 0, fewer Stage 2 and 3).
Hormone receptor (estrogen (ER) and progesterone (PR) status was defined as positive (ER+ or PR+, or both), negative (ER− and PR−), or unknown. Human epidermal growth factor receptor 2 (HER2) status was defined as positive (3 + according to immunohistochemistry (IHC) or amplified using fluorescent in situ hybridization (FISH)), negative (0 or 1 + according to IHC or 2 + according to IHC and not amplified using FISH), or unknown, although HER2 status was not consistently recorded for early-stage breast cancer cases in 2004.
Outcomes
The primary outcome was receipt of guideline-concordant care or not. Guideline-concordant breast cancer treatment was defined based on tumor size, nodal status, and hormone receptor status, similar to previous work.13 For each of 10 categories, National Comprehensive Cancer Network16 and St. Gallen Conference17 guidelines were consulted to create management requirements, which included components of surgery (BCS or mastectomy), adjuvant RT (before, during, or after surgery), lymph node surgery (LN), CTX, and ET. Minimum acceptable components of therapy for each group are as follows:
Group A (ductal carcinoma in situ (DCIS)), BCS with RT or mastectomy
Group B (T1a-b N0), BCS with RT and LN or mastectomy and LN
Group C+ (T1cN0, >1 cm, ER with PR+), BCS with RT or mastectomy, LN, ET, and CTX if younger than 70
Group C− (T1c N0, ER with PR−), BCS with RT or mastectomy, LN, and CTX
Group D+ (T2N0, ER with PR+), BCS with RT or mastectomy, LN, ET, and CTX
Group D− (T2N0, ER with PR−), BCS with RT or mastectomy, LN, and CTX
Group E+ (T1–3 N1–3 or T3N0, and ER with PR+), BCS with RT or mastectomy, LN, ET, and CTX
Group E− (T1–3N1–3 or T3N0 and ER with PR−), BCS with RT or mastectomy, LN, and CTX
Group F+ (T4NX and ER with PR+), mastectomy, LN, ET, and CXT
Group F− (T4NX and ER with PR−), mastectomy, LN, and CTX
Receipt of more than guideline management was considered guideline concordant.
Comorbidity
The ACE-27,10 a robust, validated, chart-based instrument developed specifically for individuals with cancer, was used. The index includes 26 conditions relevant to cancer therapy choice and outcome, and their severity, measured according to three levels of decompensation. After training using a tested and validated Internet-based program, abstractors reviewed records and assigned comorbidity categories and level of decompensation based on diagnoses, medical history, and laboratory and clinical tests.10,18 Comorbidities present at or before diagnosis were included; complications of treatment were excluded. An overall comorbidity index (none, low, moderate, or severe) was assigned to each participant based on the comorbidity with the highest level of decompensation. Severe level of decompensation for the overall index was assigned in cases in which there was a moderate level of decompensation of two or more comorbidities in different body systems (e.g., cardiovascular and nervous), even if none were severe.18
The 26 categories were also mapped into 12 groups, based on organ system: cardiovascular disease (myocardial infarction, coronary artery disease, congestive heart failure, arrhythmia, hypertension, venous disease, peripheral artery disease), respiratory diseases, gastrointestinal diseases (hepatic disease, stomach or intestinal diseases, pancreatic disease), renal disease, diabetes mellitus, nervous system (stroke or cerebrovascular accident, dementia, paralysis, neuromuscular disorders), psychiatric, rheumatological, acquired immunodeficiency syndrome (AIDS), cancer (solid tumor, leukemia, lymphoma) excluding the index cancer, substance abuse (alcohol abuse, illicit drugs), morbid obesity.
Statistical Analysis
Cases were stratified according to overall comorbidity burden (none, minor, moderate, severe). Receipt of guideline-concordant breast cancer care, for each group across the aggregated and disaggregated measures of burden, was compared using chi-square statistics. Similar analyses were conducted for each of the 12 mapped comorbidity categories (as above). Finally, multivariate logistic regressions were estimated using three comorbidity classifications: 26 ACE-27 comorbidity categories (Model 1), 12 collapsed ACE-27 categories (Model 2), and four levels of comorbidity (Model 3). Three models were created to maximally explore specific comorbidities and groups of comorbidities and to study the dose-response relationship overall. With these regressions, the statistics are reported for four age categories (<50, 50–64, 65–74, ≥75) and four racial ethnic groups (white, black, Hispanic, other), because these are factors known to be related to comorbidity and to guideline concordance and control for the 10 stage groups and source of payment. SAS version 9.2 was used throughout the analyses (SAS Institute, Inc., Cary, NC).
RESULTS
The study population was primarily Caucasian (76.0%), with mean age of 58.7 (range 20–99; Table 1). ACE comorbidity level was severe in 3.9%, moderate in 9.9%, mild in 42.0%, and none in 44.3%. The most common comorbidities were cardiovascular disease (n = 2,952, 45.8%), diabetes mellitus (n = 802, 12.5%), morbid obesity (n = 534, 8.3%), psychiatric disease (n = 381, 5.9%), and respiratory disease (n = 380, 5.9%).
Table 1.
Characteristics of the Sample (N = 6,439)
| Characteristic | N | Weighted Percentage |
|---|---|---|
| Age | ||
| <50 | 1,830 | 26.7 |
| 50–64 | 2,419 | 37.7 |
| ≥65 | 2,190 | 35.6 |
| Race or ethnicity | ||
| White | 3,643 | 76.0 |
| Black | 1,910 | 15.3 |
| Hispanic | 488 | 5.2 |
| Other | 396 | 3.5 |
| Source of payment | ||
| Medicare plus other public insurance | 1,432 | 23.1 |
| Medicaid | 835 | 9.1 |
| Private | 3,765 | 61.6 |
| Not insured or self-pay | 177 | 1.8 |
| Unknown | 230 | 4.4 |
| Adult Comorbidity Evaluation scorea | ||
| None | 2,784 | 44.3 |
| Mild | 2,729 | 42.0 |
| Moderate | 654 | 9.9 |
| Severe | 272 | 3.9 |
| Stage | ||
| 0 | 1,290 | 19.9 |
| I | 2,873 | 47.1 |
| II | 1,773 | 26.3 |
| III | 503 | 6.7 |
| Hormone receptor status | ||
| Positive (ER+ or PR+) | 4,398 | 70.0 |
| Negative (ER− and PR−) | 1,470 | 21.4 |
| Unknown | 571 | 8.6 |
| HER2 status | ||
| Positive | 1,962 | 30.5 |
| Negative | 2,713 | 42.1 |
| Unknown | 1,764 | 27.4 |
| Type of surgery | ||
| BCS | 3,645 | 57.9 |
| Mastectomy | 2,701 | 40.9 |
| None | 93 | 1.2 |
| Lymph node assessment | 5,293 | 82.7 |
| Radiation after BCS | 2,967 | 80.6b |
| Adjuvant chemotherapy | ||
| Single agents | 73 | 1.1 |
| Doxorubicin and cyclophosphamide | 704 | 9.9 |
| Doxorubicin, cyclophosphamide, and a taxane | 1,236 | 18.6 |
| Other multiple agents | 663 | 9.4 |
| None | 3,765 | 61.1 |
| Endocrine therapy | 3,428 | 54.7 |
| Trastuzumabc | 169 | 2.9 |
Diseases included myocardial infarction, angina pectoris or coronary artery disease, congestive heart failure, arrhythmias, hypertension, venous disease, peripheral arterial disease, respiratory system disease, hepatic disease, stomach or intestinal disease, pancreatic disease, end-stage renal disease, diabetes mellitus, stroke, dementia, paralysis, neuromuscular disease, psychiatric disorder, rheumatologic disease, acquired immunodeficiency syndrome, solid tumor including melanoma, leukemia or myeloma, lymphoma, alcohol abuse, illicit drugs, obesity.
Percent who received adjuvant RT among those who had BCS.
Trastuzumab was delivered to 2.9% overall and 8.1% of human epidermal growth factor receptor 2 (HER2)-positive cancers.
ER = Estrogen Receptor; PR = Progesterone Receptor; BCS = Breast-Conserving Surgery.
Care was guideline concordant in two-thirds of participants (69.5% without comorbidity, 59.5% with). Surgery was BCS in 57.9% and mastectomy in 40.9%; 82.7% had had a LN assessment. RT occurred in 80.6% of participants who underwent BCS. CTX was delivered to 38.9%. ET was used in 54.7% overall and in 71% of women with ER+ or PR+ tumors or both.
Table 2 shows the unadjusted rate of guideline-concordant care according to stage and comorbidity categories. Guideline concordance varied between stage groups (unreported chi-square P < .001), and level of comorbidity (P < .05). Concordance was inversely proportional to comorbidity burden: 69.5% for none, 61.4% for minor, 58.0% for moderate, and 42.7% for severe, although this inverse relationship is statistically significant for only some individual stage groups (e.g., individuals with DCIS (A)). Table 2 also shows statistically significant relationships between individual comorbidity categories and guideline-concordant treatment for a number of conditions, including cardiovascular disease, end-stage renal disease, diabetes mellitus, nervous system disorders, psychiatric disorders, and rheumatological disease. For all but psychiatric disorders, the presence of the condition was always associated with less-concordant care. In some cases, such as diabetes mellitus and nervous system disorders, there is a clear inverse relationship between severity and concordance.
Table 2.
Percentage Guideline Concordance According to Stage Group, Comorbidity (Collapsed Adult Comorbidity Evaluation (ACE-27) Comorbidity Categories), and Comorbidity Level in Individuals with Breast Cancer (N = 6,439)
| None | Minor | Moderate | Severe | Any Comorbidity | |
|---|---|---|---|---|---|
| n (% Yes) | |||||
| Comorbidity level | |||||
| A (ductal carcinoma in situ)a | 554 (81.8) | 577 (80.4) | 101 (74.0) | 42 (46.2) | 720 (77.7) |
| B (T1a/b N0) | 449 (88.2) | 480 (84.9) | 114 (82.5) | 43 (83.2) | 637 (84.4) |
| C+ (T1c N0, ER/PR+)a,b | 447 (37.7) | 447 (26.7) | 110 (27.6) | 47 (15.3) | 604 (26.1) |
| C− (T1c N0, ER/PR−)a | 127 (72.3) | 125 (56.5) | 24 (85.7) | 11 (19.6) | 160 (58.6) |
| D+ (T2N0, ER/PR+)a,b | 208 (53.2) | 224 (34.4) | 57 (18.8) | 27 (12.2) | 308 (29.5) |
| D− (T2N0, ER/PR−)a,b | 139 (75.0) | 119 (51.6) | 26 (70.0) | 10 (27.5) | 155 (53.1) |
| E+ (T1–3N1–3 or T3N0, ER/PR+)a,b | 581 (67.5) | 502 (57.6) | 136 (57.7) | 47 (43.7) | 685 (56.8) |
| E− (T1–1N1–3 or T3N0, ER/PR−) | 209 (84.9) | 182 (83.9) | 56 (80.7) | 30 (62.1) | 268 (80.0) |
| F+ (T4NX, ER/PR+) | 37 (57.2) | 37 (48.6) | 21 (60.9) | 9 (40.0) | 67 (51.6) |
| F− (T4NX, ER/PR+) | 33 (52.2) | 36 (75.6) | 9 (77.3) | 6 (48.5) | 51 (71.3) |
| Alla,b | 2,784 (69.5) | 2,729 (61.4) | 654 (58.0) | 272 (42.7) | 3,655 (59.5) |
| Morbidity | All Levels | ||||
| Cardiovascular diseasec,d,e | 3,487 (68.8) | 2,559 (59.9) | 324 (43.6) | 69 (47.9) | 2,952 (57.6) |
| Respiratory disease | 6,059 (64.1) | 339 (61.3) | 23 (63.3) | 18 (44.9) | 380 (60.7) |
| Gastrointestinal diseasef | 6,268 (64.1) | 153 (52.2) | 16 (77.1) | 2 (100.0) | 171 (55.0) |
| End-stage renal diseasec,d | 6,378 (64.1) | 42 (38.9) | 13 (12.1) | 6 (62.7) | 61 (37.4) |
| Diabetes mellitusc,d | 5,637 (65.1) | 669 (55.3) | 99 (46.9) | 34 (40.7) | 802 (53.7) |
| Nervous system diseasec,d,g | 6,170 (65.3) | 196 (37.1) | 52 (33.4) | 21 (15.0) | 269 (34.8) |
| Psychiatric diseasec,d | 6,058 (63.4) | 354 (71.0) | 24 (35.4) | 3 (89.5) | 381 (70.0) |
| Rheumatological disease,d | 6,305 (64.2) | 104 (50.3) | 28 (46.5) | 2 (0.0) | 134 (49.0) |
| Acquired immunodeficiency syndrome | 6,434 (63.9) | 5 (41.8) | 0 (0.0) | 0 (0.0) | 5 (41.8) |
| Cancerh | 6,395 (63.9) | 25 (59.3) | 3 (12.4) | 16 (51.5) | 44 (52.3) |
| Substance abusei | 6,386 (63.8) | 35 (71.7) | 14 (82.0) | 4 (0.0) | 53 (71.0) |
| Obesity | 5,905 (63.8) | 203 (57.5) | 318 (69.1) | 13 (68.3) | 534 (64.5) |
For comorbidity categories, weighted percentage with guideline-concordant care was used.
Statistically significant comparing four levels of comorbidity, chi-square, P < .05.
Statistically significant comparing any comorbidity versus none, chi-square, P < .05. (for group C−, all levels, P = .05).
Statistically significant three levels of comorbidity and none, chi-square, P < .05.
Statistically significant all levels of comorbidity versus none, chi-square, P < .05.
Myocardial infarction, coronary artery disease, congestive heart failure, arrhythmia, hypertension, venous disease, and peripheral artery disease.
Hepatic disease, stomach and intestinal diseases, and pancreatic disease.
Stroke or cerebrovascular accident, dementia, paralysis, and neuromuscular disorders.
Solid tumor, leukemia, or lymphoma.
Alcohol abuse and illicit drugs.
ER = Estrogen Receptor; PR = Progesterone Receptor.
Any comorbidity, versus none, was associated with less-guideline-concordant care (P < .05) in four stage strata (stage I, tumor >1 cm and hormone receptor positive (C+), 26.1% vs 37.7%; stage IIA and hormone receptor positive (D+), 29.5% vs 53.2%; stage IIA and hormone receptor negative (D−), 53.1% vs 75%; and stage IIB or III and hormone receptor positive (E+), 56.8% vs 67.5%) and overall (59.5% vs 69.5%). The presence of comorbidity did not significantly affect the rate of guideline-concordant care for DCIS or less than 1 cm, node-negative cancers, for which only local therapy met guideline concordance. Few cases of higher-stage cancer may have limited the ability to find significant differences in rates of guideline-concordant care according to comorbidity level.
The effect of specific diagnosis and severity on receipt of guideline-concordant care was examined (Table 2). In unadjusted analyses, comparing any comorbidity with none, receipt of guideline-concordant care was less likely (P < .05) with cardiovascular diseases (58% vs 69%), renal disease (37% vs 64%), diabetes mellitus (54% vs 65%), nervous system disease (35% vs 65%), and rheumatological disease (49% vs 64%). Delivery of guideline-concordant care was more likely (univariate P < .05) with psychiatric illness (70% vs 63%). In univariate analysis, with comorbidity categorized according to level of severity, significant differences in guideline-concordant care (P < .05) were found for diabetes mellitus and for cardiovascular, renal, and nervous system diseases. For AIDS, other cancers, substance abuse, and obesity, small numbers limited detection of a dose-response relationship.
Table 3 shows results of multivariate analyses using three models. In Model 1, less guideline concordance was associated with dementia (OR = 0.45, 95% CI = 0.24–0.83), older age, and black race (vs white OR = 0.87, 95% CI = 0.76–0.99) and higher likelihood of guideline concordance with hypertension (OR = 1.15, 95% CI = 1.01–1.30). In Model 2, less guideline concordance was associated with renal system disease (OR = 0.57, 95% CI = 0.33–0.99), nervous system disease (OR = 0.62, 95% CI = 0.47–0.82), and older age. In Model 3, less guideline concordance was associated with severe comorbidity (OR = 0.67, 95% CI = 0.51–0.89), older age, and black race.
Table 3.
Multivariate Logistic Regressiona of Guideline-Concordant Care According to Comorbidity, Age, and Race in Individuals with Breast Cancer (N = 6,208b)
| Comorbidity Group (Model) | Model 1 (Includes Each 26 Separate ACE-27 Comorbidity Categories in the Model) | Model 2 (Includes 12 Collapsed ACE-27 Categories in the Model) | Model 3 (Includes 4 Levels of Comorbidity in the Model) | |
|---|---|---|---|---|
|
|
|
|||
| 1 (1) | 2 (2) | Odds Ratio (95% Confidence Interval) | ||
| Myocardial infarction | 0.94 (0.64–1.37) | |||
|
| ||||
| Coronary artery disease | 0.97 (0.74–1.28) | |||
|
| ||||
| Congestive heart failure | 0.74 (0.52–1.07) | |||
|
| ||||
| Arrhythmia | 0.90 (0.64–1.26) | |||
|
| ||||
| Hypertension | 1.15 (1.01–1.30)c | |||
|
| ||||
| Venous disease | 1.13 (0.67–1.92) | |||
|
| ||||
| Peripheral arterial disease | Cardiovascular | 0.67 (0.36–1.22) | 1.13 (1.00–1.28) | |
|
| ||||
| Respiratory disease | Respiratory | 1.18 (0.93–1.51) | 1.14 (0.89–1.44) | |
|
| ||||
| Hepatic disease | 0.62 (0.33–1.19) | |||
|
| ||||
| Stomach or intestinal disease | 1.01 (0.68–1.51) | |||
|
| ||||
| Pancreatic disease | Gastrointestinal | 0.88 (0.63–1.23) | ||
|
| ||||
| Renal system disease | Renal system disease | 0.64 (0.36–1.12) | 0.57 (0.33–0.99)c | |
|
| ||||
| Diabetes mellitus | Diabetes mellitus | 0.91 (0.76–1.08) | 0.88 (0.75–1.06) | |
|
| ||||
| Stroke or cerebrovascular accident | 0.78 (0.55–1.11) | |||
|
| ||||
| Dementia | 0.45 (0.24–0.83)c | |||
|
| ||||
| Paralysis | 0.88 (0.14–5.66) | |||
|
| ||||
| Neuromuscular disease | Nervous system | 0.62 (0.34–1.15) | 0.62 (0.47–0.82)c | |
|
| ||||
| Psychiatric disease | Psychiatric disease | 1.28 (1.00–1.63) | 1.27 (0.99–1.62) | |
|
| ||||
| Rheumatological disease | Rheumatological disease | 0.79 (0.54–1.15) | 0.80 (0.55–1.17) | |
|
| ||||
| AIDS | AIDS | 0.25 (0.04–1.52) | 0.25 (0.04–1.53) | |
|
| ||||
| Solid tumor | 0.54 (0.27–1.09) | |||
|
| ||||
| Leukemia | ||||
|
| ||||
| Lymphoma | Cancer | 0.623 (0.328–1.183) | ||
|
| ||||
| Alcohol abuse | 1.006 (0.487–2.077) | |||
|
| ||||
| Illicit drugs | Substance abuse | 0.827 (0.296–2.310) | 1.021 (0.547–1.904) | |
|
| ||||
| Obesity | Obesity | 1.108 (0.900–1.363) | 1.108 (0.901–1.363) | |
|
| ||||
| Comorbidity index (vs none) | ||||
|
| ||||
| Mild | 1.124 (0.990–1.277) | |||
|
| ||||
| Moderate | 1.117 (0.919–1.359) | |||
|
| ||||
| Severe | 0.673 (0.507–0.893)c | |||
|
| ||||
| Age (vs <50) | ||||
|
| ||||
| 50–64 | 0.653 (0.563–0.757)c | 0.657 (0.566–0.762)c | 0.652 (0.563–0.756)c | |
|
| ||||
| 65–74 | 0.394 (0.325–0.478)c | 0.395 (0.326–0.480)‡ | 0.385 (0.318–0.465)c | |
|
| ||||
| ≥75 | 0.135 (0.109–0.167)c | 0.131 (0.106–0.162)c | 0.126 (0.102–0.155)c | |
|
| ||||
| Race (vs white) | ||||
|
| ||||
| Black | 0.872 (0.763–0.996)c | 0.878 (0.769–1.002) | 0.873 (0.767–0.993)c | |
|
| ||||
| Hispanic | 0.924 (0.743–1.149) | 0.932 (0.749–1.159) | 0.907 (0.730–1.126) | |
|
| ||||
| Other | 0.944 (0.741–1.203) | 0.949 (0.745–1.208) | 0.933 (0.733–1.186) | |
Controlling for stage and source of payment; pancreatic disease, lymphoma, and leukemia excluded from Model 1 because of too few observations (<5).
Observations excluded (n = 231) for missing payer or race, also omitted metastatic, missing stage, if both PR and ER receptors missing, and if tumor size missing and nodes less than N1.
Statistically significant, three levels of comorbidity and none, chi-square, P < .05.
ACE = Adult Comorbidity Evaluation; AIDS = acquired immunodeficiency syndrome.
DISCUSSION
Using this large, retrospective database, with systematically collected, detailed information about comorbid diagnoses and tumor registry and record-based information about multicomponent breast cancer management, the authors of this study confirmed that higher comorbidity is associated with a lower rate of multidisciplinary guideline-concordant care. Older age and black race also predicted less guideline concordance.19,20 This adds to the literature by exploring the effect of specific comorbid diagnoses on receipt of guideline-concordant breast cancer care. In unadjusted analyses, guideline concordance was lower with cardiovascular, renal, nervous system, and rheumatological diseases and diabetes mellitus; this held true in adjusted analyses for renal and nervous system diseases. With regard to specific diagnoses, dementia was associated with less-guideline-concordant care whereas hypertension was associated with small, but statistically significantly greater guideline concordance.
It may be that a diagnosis of hypertension implies more access to health care and, therefore, greater likelihood of guideline concordance and better outcomes. Individuals with a diagnosis of hypertension are likely to receive antihypertensive medication and are more likely to receive guideline-concordant care, either of which might improve outcomes. Thus, use of guideline-concordant breast cancer care may have confounded previous reports that antihypertensive use improves breast cancer prognosis.7,21–23 Future research studying the effects of comorbid diagnoses and their treatment on breast cancer outcome should include an assessment of guideline-concordant cancer treatment.
In general, higher comorbidity burden led to less guideline-concordant care, which may be justified, given the known inverse relationship between comorbidity and life expectancy24 and higher financial or quality-of-life burdens, especially in those with high comorbidity burden. Comorbidity may decrease the benefits or increase the risk of certain components of care. For instance, the current study found that individuals with stage IIA disease (T2N0), regardless of hormone receptor status, and stage IIB and III disease with hormone receptor–positive disease, in which the absolute benefit of adding chemotherapy is small, were significantly less likely to receive guideline-concordant care.
Strengths of this study include the large, geographically diverse database, with tumor registry data augmented with chart abstraction, and use of the ACE-27 comorbidity measure. Record re-abstraction enabled capture of detailed information about tumor characteristics, treatment, and comorbid conditions. Whereas most studies of comorbidity were limited to older adults with Medicare, this study included individuals aged 20–99. Large sample size and availability of well-defined variables pertaining to demographic, participant, tumor, and treatment characteristics facilitated study of multicomponent management. Whereas other studies have primarily focused on single treatment components, the rich database allowed study of multicomponent, coordinated care. Use of the ACE-27, a robust, validated, chart-based instrument, developed specifically for individuals with cancer to study presence and severity of comorbidity, is a tremendous strength of the study.10 Other studies have used less-powerful measures of comorbidity.1,3,8,9 The widely used Charlson Comorbidity Index25,26 can be used with administrative databases but includes fewer comorbidities and no information about disease severity. The Cumulative Illness Rating Scale, perhaps the most-comprehensive index, requires detail only available in prospective studies.27 A claims-based version of the ACE has been proposed.28 Detailed information from the ACE-27 also allowed associations with specific comorbid diagnoses, rather than only a summary score of comorbidity, to be explored.5
There are limitations to this study. As a retrospective study, what was recorded in the medical record, which is a poor source of information about physician and individual decision-making, actual referrals, and patient adherence, limited it. Lack of record review from each involved physician may have led to missed care components, but the abstractors sought records from hospital and physician offices. Inclusion criteria for the analysis excluded approximately 25% of the database and may limit external validity of the findings. Small numbers in some subcategories limited power. Last, the study was structured to detect treatment meeting minimum standards; because less than guideline care is known to be associated with worse outcomes, it was felt that overclassification in this direction was acceptable.
In summary, this study found that greater comorbidity burden is associated with a lower rate of guideline-concordant multidisciplinary care. The noted association between guideline care and hypertension is worthy of further study. Because less than guideline-concordant care for breast cancer, regardless of age or race, leads to inferior outcomes,29,30 where possible in studies of cancer outcome, guideline concordance should be included in analyses.
Acknowledgments
An abstract based on these results was presented at the annual meeting of the American Society of Clinical Oncology in June 2012.
The Centers for Disease Control and Prevention supported the Breast and Prostate Cancer Data Quality and Patterns of Care Study through cooperative agreements with the California Cancer Registry (Public Health Institute) (1-U01-DP000260), Emory University (1-U01-DP000258), Louisiana State University Health Sciences Center (1-U01-DP000253), Minnesota Cancer Surveillance System (Minnesota Department of Health) (1-U01-DP000259), Medical College of Wisconsin (1-U01-DP000261), University of Kentucky (1-U01-DP000251), and Wake Forest University (1-U01-DP000264). The findings and conclusions in this report are those of the authors and do not necessarily represent the official position of the Centers for Disease Control and Prevention.
Conflict of Interest: Dr. Kimmick has served on speakers boards; been a consultant for AstraZeneca, Pfizer, and Novartis; and has received research funding from Astra-Zeneca, Pfizer, Roche, Bionovo, Wyeth, Bristol-Myers Squibb, and GlaxoSmithKline. Dr. Anderson has had research funding from AstraZeneca and Roche. None of the other authors have financial interests, activities, relationships, or affiliations that would pose a conflict of interest with the content of this manuscript.
Author Contributions: Study concept and design: Kimmick, Fleming. Acquisition of data: Fleming, Sabatino, Wu, Wilson, Cress, Anderson. Analysis and interpretation of data: Kimmick, Fleming, Sabatino, Wu, Hwang, Wilson, Lund, Cress, Anderson. Preparation of the manuscript: Kimmick, Fleming, Sabatino, Wu, Hwang, Wilson, Lund, Cress, Anderson.
Sponsor’s Role: The sponsor had a significant role in the construction of the database and oversight of the data use.
References
- 1.Louwman WJ, Janssen-Heijnen MLG, Houterman S, et al. Less extensive treatment and inferior prognosis for breast cancer patient with comorbidity: A population-based study. Eur J Cancer. 2005;41:779–785. doi: 10.1016/j.ejca.2004.12.025. [DOI] [PubMed] [Google Scholar]
- 2.Land LH, Dalton SO, Jorgensen TL, et al. Comorbidity and survival after early breast cancer. A review Crit Rev Oncol Hematol. 2012;8:196–205. doi: 10.1016/j.critrevonc.2011.03.001. [DOI] [PubMed] [Google Scholar]
- 3.Hawfield A, Lovato J, Covington D, et al. Retrospective study of the effect of comorbidity on use of adjuvant chemotherapy in older women with breast cancer in a tertiary care setting. Crit Rev Oncol Hematol. 2006;59:250–255. doi: 10.1016/j.critrevonc.2005.12.004. [DOI] [PubMed] [Google Scholar]
- 4.Garg P, Rana F, Gupta R, et al. Predictors of toxicity and toxicity profile of adjuvant chemotherapy in elderly breast cancer patients. Breast J. 2009;15:404–408. doi: 10.1111/j.1524-4741.2009.00745.x. [DOI] [PubMed] [Google Scholar]
- 5.Lee L, Cheung WY, Atkinson E, et al. Impact of comorbidity on chemotherapy use and outcomes in solid tumors: A systematic review. J Clin Oncol. 2011;29:106–117. doi: 10.1200/JCO.2010.31.3049. [DOI] [PubMed] [Google Scholar]
- 6.Schonberg MA, Silliman RA, McCarthy EP, et al. Factors noted to affect breast cancer treatment decisions of women aged 80 and older. J Am Geriatr Soc. 2012;60:538–544. doi: 10.1111/j.1532-5415.2011.03820.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Braithwaite D, Tammemagi CM, Moore DH, et al. Hypertension is an independent predictor of survival disparity between African-American and white breast cancer patients. Int J Cancer. 2009;124:1213–1219. doi: 10.1002/ijc.24054. [DOI] [PubMed] [Google Scholar]
- 8.Siegelmann-Danieli N, Khandelwal V, Wood GC, et al. Breast cancer in elderly women: Outcome as affected by age, tumor features, comorbidities, and treatment approach. Clin Breast Cancer. 2006;7:59–66. doi: 10.3816/CBC.2006.n.014. [DOI] [PubMed] [Google Scholar]
- 9.Field TS, Bosco JL, Prout MN, et al. Age, comorbidity, and breast cancer severity: Impact on receipt of definitive local therapy and rate of recurrence among older women with early-stage breast cancer. J Am Coll Surg. 2011;213:757–765. doi: 10.1016/j.jamcollsurg.2011.09.010. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Piccirillo JF, Creech C, Zequeira R, et al. Inclusion of comorbidity into oncology data registries. J Registry Manag. 1999;26:66–70. [Google Scholar]
- 11.Chevarley F, White E. Recent trends in breast cancer mortality among white and black US women. Am J Public Health. 1997;87:775–781. doi: 10.2105/ajph.87.5.775. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Boyer-Chammard A, Taylor TH, Anton-Culver H. Survival differences in breast cancer among racial/ethnic groups: A population-based study. Cancer Detect Prev. 1999;23:463–473. doi: 10.1046/j.1525-1500.1999.99049.x. [DOI] [PubMed] [Google Scholar]
- 13.Maskarinec G, Pagano IS, Yamashiro G, et al. Influences of ethnicity, treatment, and comorbidity on breast cancer survival in Hawaii. J Clin Epidemiol. 2003;56:678–685. doi: 10.1016/s0895-4356(03)00079-9. [DOI] [PubMed] [Google Scholar]
- 14.German RR, Wike JM, Bauer KR, et al. Quality of cancer registry data: Findings from CDC-NPCR’s Breast and Prostate Cancer Data Quality and Patterns of Care Study. J Registry Manag. 2011;38:75–86. [PubMed] [Google Scholar]
- 15.Wu XC, Lund MJ, Kimmick GG, et al. Influence of race, insurance, socioeconomic status, and hospital type on receipt of guideline-concordant adjuvant systemic therapy for locoregional breast cancers. J Clin Oncol. 2011;30:142–150. doi: 10.1200/JCO.2011.36.8399. [DOI] [PubMed] [Google Scholar]
- 16.Carlson RW, Edge SB. Update: NCCN breast cancer clinical practice guidelines. J Natl Compr Canc Netw. 2004;2(Suppl 3):S1–S4. [PubMed] [Google Scholar]
- 17.Aapro MS. Oncologist; Adjuvant therapy of primary breast cancer: A review of key findings from the 7th international conference; St. Gallen. February 2001; 2001. pp. 376–385. [DOI] [PubMed] [Google Scholar]
- 18.Johnson AS, Piccirillo JF, Creech CM, et al. Validation of a comorbidity education program. Journal of Registry Management. 2001;28:125–131. [Google Scholar]
- 19.Kimmick G, Hughes K, Muss HB. Breast cancer in older women. In: Harris JR, Lippman ME, Morrow M, et al., editors. Diseases of the Breast. Vol. 4. Philadelphia, PA: Lippincott Williams & Wilkins; 2009. pp. 1059–1072. [Google Scholar]
- 20.Wojcik BE, Spinks MK, Optenberg SA. Breast carcinoma survival analysis for African American and white women in an equal-access health care system. Cancer. 1998;82:1310–1318. doi: 10.1002/(sici)1097-0142(19980401)82:7<1310::aid-cncr14>3.0.co;2-9. [DOI] [PubMed] [Google Scholar]
- 21.Powe DG, Voss MJ, Zanker KS, et al. Beta-blocker drug therapy reduces secondary cancer formation in breast cancer and improves cancer specific survival. Oncotarget. 2010;1:628–638. doi: 10.18632/oncotarget.197. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Ganz PA, Habel LA, Weltzien EK, et al. Examining the influence of beta blockers and ACE inhibitors on the risk for breast cancer recurrence: Results from the LACE cohort. Breast Cancer Res Treat. 2011;129:549–556. doi: 10.1007/s10549-011-1505-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Barron TI, Connolly RM, Sharp L, et al. Beta blockers and breast cancer mortality: A population-based study. J Clin Oncol. 2011;29:2635–2644. doi: 10.1200/JCO.2010.33.5422. [DOI] [PubMed] [Google Scholar]
- 24.Ahern TP, Lash TL, Thwin SS, et al. Impact of acquired comorbidities on all-cause mortality rates among older breast cancer survivors. Med Care. 2009;47:73–79. doi: 10.1097/MLR.0b013e318180913c. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Charlson ME, Pompei P, Ales KL, et al. A new method of classifying prognostic comorbidity in longitudinal studies: Development and validation. J Chronic Dis. 1987;40:373–383. doi: 10.1016/0021-9681(87)90171-8. [DOI] [PubMed] [Google Scholar]
- 26.D’Hoore W, Bouckaert A, Telquin C. Practical considerations on the use of the Charlson Comorbidity Index with administrative data bases. J Clin Epidemiol. 1996;49:1429–1433. doi: 10.1016/s0895-4356(96)00271-5. [DOI] [PubMed] [Google Scholar]
- 27.Wedding U, Roehrig B, Klippstein A, et al. Comorbidity in patients with cancer: Prevalence and severity measured by cumulative illness rating scale. Crit Rev Oncol Hematol. 2007 Mar;61:269–276. doi: 10.1016/j.critrevonc.2006.11.001. [DOI] [PubMed] [Google Scholar]
- 28.Fleming ST, Sabatino SA, Kimmick G, et al. Developing a claim-based version of the ACE-27 comorbidity index: A comparison with medical record review. Med Care. 2011;49:752–760. doi: 10.1097/MLR.0b013e318215d7dd. [DOI] [PubMed] [Google Scholar]
- 29.Bouchardy C, Rapiti E, Fioretta G, et al. Undertreatment strongly decreases prognosis of breast cancer in elderly women. J Clin Oncol. 2003;21:3580–3587. doi: 10.1200/JCO.2003.02.046. [DOI] [PubMed] [Google Scholar]
- 30.Schonberg MA, Marcantonio ER, Ngo L, et al. Causes of death and relative survival of older women after a breast cancer diagnosis. J Clin Oncol. 2011;29:1570–1577. doi: 10.1200/JCO.2010.33.0472. [DOI] [PMC free article] [PubMed] [Google Scholar]