Abstract
PURPOSE
The quality of breast cancer care in sub-Saharan Africa contributes to the region’s dismal breast cancer mortality. ASCO has issued quality measures focusing on delivery of adjuvant chemotherapy, radiotherapy, and endocrine therapy. We applied these measures in five South African public hospitals and analyzed factors associated with care concordance.
MATERIALS AND METHODS
Among 1,736 women with breast cancer who were enrolled in the South African Breast Cancer and HIV Outcomes study over 24 months, we evaluated care using ASCO’s three measures. We also evaluated adjuvant chemotherapy receipt in 957 women with an indication. We used logistic regression to estimate associations between measure-concordant care and patient factors.
RESULTS
Of 235 women with hormone receptor–negative cancer, 173 (74%) began adjuvant chemotherapy within 120 days from diagnosis. Of 194 patients who received breast-conserving surgery, 73 (37%) began radiotherapy within 365 days from diagnosis. Of 999 women with hormone receptor–positive cancer, 719 (72%) initiated endocrine therapy within 365 days from diagnosis. Chemotherapy and radiotherapy measure-concordant care were more common among women residing < 20 km from the hospital (odds ratio [OR], 1.79; 95% CI, 1.32 to 2.44 and OR, 3.17; 95% CI, 1.57 to 6.42). Endocrine therapy measure-concordant care was more common among English-speaking women (OR, 2.12; 95% CI, 1.12 to 4.02). Participating hospitals varied in care concordance. HIV infection did not affect care quality.
CONCLUSION
More timely delivery of chemotherapy, radiotherapy, and endocrine therapy is needed in South Africa, particularly for women living > 20 km from the hospital or not speaking English. Focused quality improvement efforts could support that goal.
INTRODUCTION
Breast cancer (BC) is the most common cancer among women in sub-Saharan Africa (SSA).1 Unfortunately, resource constraints limit access to surgery, radiotherapy, and systemic treatments, and mortality rates are much higher than in the United States and Europe.2
In 2007, ASCO published three measures for evaluating the quality of BC care3:
Proportion of women age 18-70 years with American Joint Committee on Cancer (AJCC) stage II-III disease and estrogen receptor (ER)– and progesterone receptor (PR)–negative histology who receive chemotherapy within 120 days from diagnosis
Proportion of women age 18-70 years with AJCC stage I-III disease treated with breast-conserving surgery (BCS) who receive radiation therapy to the breast within 365 days from diagnosis
Proportion of women aged ≥ 18 years with AJCC stage I-III disease, tumor size > 1 cm, and ER- or PR-positive histology who receive tamoxifen or an aromatase inhibitor within 365 days from diagnosis
These measures were based on evidence of clinical benefit from each therapy.4-6 They were endorsed by the National Quality Forum and used in ASCO’s Quality Oncology Practice Initiative.7 Although the three measures were designed for the United States, they have previously been used to assess BC care in middle-income countries, including Brazil and Malaysia.8,9
CONTEXT
Key Objective
To determine the degree to which breast cancer care in South Africa’s public hospitals complies with quality measures developed by the American Society of Clinical Oncology.
Knowledge Generated
Among breast cancer patients from five South African hospitals, concordance with ASCO quality metrics assessing adjuvant chemotherapy, radiotherapy and endocrine therapy was seen in 74%, 37% and 72% of patients, respectively. Patients living less than 20 kilometers from their hospital and primarily speaking English were significantly more likely to receive measure-concordant care.
Relevance
The quality of breast cancer adjuvant therapy delivery in the South African public hospital system requires improvement, particularly the timely delivery of adjuvant radiotherapy. Quality improvement initiatives may have the greatest impact by focusing on high risk populations, such as patients residing far from their treating facility and non-native English speakers. In addition, future quality metrics should specifically address the resource-constraints seen in South Africa’s public system.
The South Africa (SA) National Department of Health’s BC treatment guidelines, which overlap significantly with guidelines issued by ASCO and the European Society for Medical Oncology (ESMO), recommend treatment consistent with the ASCO measures.10 All three modalities are offered within SA’s public health care system, but high patient volumes, provider shortages, and other resource constraints limit their availability. Little has been published regarding the extent to which actual BC care in SA’s public hospitals aligns with national guidelines.
Given their consistency with SA’s national BC guidelines, ASCO’s quality metrics may be appropriate for describing the quality of BC care in SA. However, the feasibility of their use in SA and their relevance to patients in SA have not been evaluated.
In this study, therefore, we used those measures to describe the quality of BC care in five SA public hospitals and examined the role of patient factors in measure-concordant care. Through our analyses, we also hoped to gain insight into the applicability of the ASCO measures to SA’s public health care system.
MATERIALS AND METHODS
Data Source and Setting
Our study population was drawn from the South African Breast Cancer and HIV Outcomes (SABCHO) study cohort. The primary aim of SABCHO, which has been enrolling women from five public SA hospitals since July 2015, is to characterize the impact of HIV infection on BC outcomes.11 Women were eligible for SABCHO if they were > 18 years of age, newly diagnosed with BC, had no history of other cancers, received their BC care at a study hospital, and provided consent.
The five study hospitals—all part of the same public system—were Chris Hani Baragwanath Academic Hospital (CHBAH), Charlotte Maxeke Johannesburg Academic Hospital (CMJAH), Inkosi Albert Luthuli Central Hospital (IALCH), Ngwelezana Hospital (NH), and Grey’s Hospital (GH). CHBAH and CMJAH serve Soweto and Johannesburg, respectively, and are affiliated with the University of Witwatersrand. IALCH, NH, and GH are affiliated with the University of KwaZulu-Natal and located in the cities of Durban, Empangeni, and Pietermaritzburg, respectively. IALCH and NH share facilities and providers and were analyzed as a single site. Participants’ BC care is centralized at the hospitals, where study staff enter data on patient demographics, risk factors, household wealth, pathology, treatments, and outcomes into a custom-built, Web-based electronic medical record (EMR) system originally developed for clinical use but adapted to serve as the SABCHO study database. A few participating providers continue to use paper records; study staff regularly extract their patients’ data into the electronic database.
Although SA is an upper middle income country, it has tremendous income inequality. The mean household income of white families is 6 times that of black households.12 HIV prevalence in black women is 25%, 20 times higher than in white women.13 Per capita spending in the public health care system is < 15% of that in the coexisting private system; BC survival differs significantly between the two.14,15 National policy specifies that low- or no-cost cancer surgery, chemotherapy, radiotherapy, and endocrine therapy be available at tertiary-level public hospitals, but timely access to these treatments is inconsistent.
Study Design and Participants
We analyzed women enrolled in the SABCHO study between July 1, 2015, and July 1, 2017, with follow-up through August 2018. We used data on age, American Joint Committee on Cancer 7th edition stage, ER/PR status, and type of surgery to establish three cohorts, each corresponding to the denominator described by the ASCO measures for delivery of chemotherapy (ASCO-C cohort), radiotherapy (ASCO-R cohort), or endocrine therapy (ASCO-E cohort). Patients were excluded from a given cohort if they died during that measure’s follow-up period. We also created an additional cohort of women < 70 years old with an indication for adjuvant chemotherapy according to ESMO 2015 guidelines for BC management (ESMO-C cohort; ie, luminal B, human epidermal growth factor receptor 2 [HER2]–enriched, and triple-negative [TNBC] tumors or luminal A tumors with metastases in ≥ 4 lymph nodes, ≥ T3 tumor stage, or grade 3 tumor histology).16 Thus, the ESMO-C cohort included all women in the ASCO-C cohort and additional women with aggressive luminal-type cancers. The molecular subtype definitions used for SABCHO analyses have been published.17 Laboratories accredited by the South African National Accreditation System conducted all immunohistochemistry.
Variables
We compiled participant data on age, race, primary language, relationship status (ie, married/cohabitating v not), employment status (ie, full- or part-time employment v unemployed or retired), HIV status, and other comorbidities. Addresses were used to calculate straight-line distance from the treating hospital. Date of diagnosis was the date of first biopsy confirming invasive disease.
Adapting a strategy used by the Demographics and Health Surveys Program to create a single variable approximating socioeconomic status, we performed principal component analysis on items from each patient’s baseline household wealth survey, including water sources, toilet facilities, and physical amenities.18 We assigned participants to quintiles using the value of the first principal component.
Outcomes
For each ASCO measure cohort, our primary outcome was initiation of the relevant therapy within the specified time frame. For the ESMO-C cohort, our primary outcome was initiation of chemotherapy within 120 days from diagnosis. We therefore collected the dates of surgery, chemotherapy initiation, radiotherapy initiation, and endocrine therapy initiation. We reviewed paper records as well as the electronic study database to minimize underestimation of delivered therapies.
Statistical Analysis
We categorized the care of participants who received the relevant therapy within the time frame as measure concordant and that of those who did not as discordant. We report rates of concordance with each measure. We computed crude odds ratios for the association of each the above-mentioned patient characteristics with measure-concordant care using bivariate logistic regression. Factors showing a P-value ≤ .1 on Wald testing for crude association were included as covariates in multivariate logistic regression models used to calculate adjusted odds ratios (ORs). All calculations were performed using SAS Studio, version 3.6 (Cary, NC).
Ethics
This work was approved by the institutional review boards of Columbia University, the University of Witwatersrand, and the University of KwaZulu-Natal. All participants provided informed consent for inclusion in the SABCHO study cohort.
RESULTS
Participants
Of the 1,795 women enrolled in the SABCHO study, 59 either lacked invasive disease or had a second cancer diagnosis. The remaining 1,736 were eligible for the 4 study cohorts. Median age at diagnosis was 55.5 (interquartile range [IQR], 44.9-66.2) years, median distance from home to the hospital was 20.7 (IQR, 8.7-42.9) km, and 1,320 (76%) women were black (Table 1). The most common primary language was Zulu (42%). At diagnosis, 743 (43%) patients had stage III disease, and 353 (20%) had stage IV. ER/PR expression and HER2 overexpression were detected in 1,356 (79%) and 467 (27%) women, respectively, and 367 (22%) women were HIV infected (Table 1). The ASCO-C, ESMO-C, ASCO-R, and ASCO-E cohorts included 235, 957, 194, and 999 patients, respectively (Fig 1).
TABLE 1.
Overall Demographic and Clinical Characteristics Among All Eligible SABCHO Study Participants and Within Each Quality Measure Cohort
FIG 1.
Women enrolled in the SABCHO (South African Breast Cancer and HIV Outcomes) study from July 2015 to July 2017 and eligibility for individual measure cohorts. Listed reasons for exclusion are not mutually exclusive. C, chemotherapy; BCS, breast-conserving surgery; E, endocrine therapy; ER, estrogen receptor; ESMO, European Society for Medical Oncology; PR, progesterone receptor; R, radiotherapy.
Chemotherapy
Among ASCO-C patients, 173 (74%) received chemotherapy within 120 days from diagnosis (ie, measure-concordant care). An additional 33 (14%) patients received chemotherapy after 120 days; overall median time to treatment was 66.5 (IQR, 50-105) days (Table 2). Among ESMO-C patients, 642 (67%) received chemotherapy within 120 days. Another 166 (17%) patients were treated later; median time to treatment was 75 (IQR, 49.5-112.5) days (Table 2; Fig 2A).
TABLE 2.
Rates of Metric-Concordant Care and Median Times to Treatment Initiation for All Measure Cohorts, by Hospital
FIG 2.
Days from diagnosis to first receipt of (A) neoadjuvant or adjuvant chemotherapy in the European Society for Medical Oncology–chemotherapy cohort, (B) radiotherapy in the ASCO-radiotherapy cohort, and (C) endocrine therapy in the ASCO-–endocrine therapy cohort.
The ESMO-C cohort’s multivariate model included age; distance from the hospital; employment status; stage; molecular subtype; concurrent hypertension, diabetes, and arthritis; treating hospital; and receipt of radiotherapy (Table 3). Concordant care was more likely in patients residing < 20 km from the hospital (OR, 1.79; 95% CI, 1.32 to 2.44) or who had received radiotherapy (OR, 2.81; 95% CI, 2.06 to 3.85). Concordant care was less likely in patients with stage II disease (OR, 0.47; 95% CI, 0.34 to 0.64), with luminal A versus TNBC subtype (OR, 0.32; 95% CI, 0.18 to 0.55), or treated at IALCH/NH versus CHBAH (OR, 0.38; 95% CI, 0.26 to 0.57; Table 3).
TABLE 3.
Sociodemographic and Clinical Factors in Relation to Chemotherapy Receipt Within 120 Days From Diagnosis Among Women in the ESMO-C Cohort
Radiation Therapy
In the ASCO-R cohort, 73 patients (37%) received radiotherapy within 365 days. An additional 69 (36%) patients received radiotherapy after 365 days; median time to treatment was 358 (IQR, 296-425) days (Table 2; Fig 2B). The multivariate model included distance from the hospital, stage, molecular subtype, and receipt of chemotherapy (Table 4). Patients residing < 20 km from the hospital (OR, 3.17; 95% CI, 1.57 to 6.42) or having stage I versus stage III disease (OR, 6.74; 95% CI, 1.83 to 24.88) were more likely to receive radiotherapy measure-concordant care, and those with luminal B subtype versus TNBC (OR 0.30, 95% CI 0.12-0.78) were less likely.
TABLE 4.
Sociodemographic and Clinical Factors in Relation to Radiotherapy Receipt Within 365 Days From Diagnosis by Women in the ASCO-R Cohort
Endocrine Therapy
Of the 999 eligible patients, 719 (72%) patients initiated endocrine therapy within 365 days. Another 117 (12%) patients started after 365 days; median time to initiation was 238 (IQR, 157-308.5) days (Table 2; Fig 2C). Age; race; primary language; stage; molecular subtype; HIV infection, hypertension, heart disease, diabetes, arthritis, or asthma/chronic obstructive pulmonary disease; hospital; and receipt of radiotherapy were included in the multivariate model. Measure-concordant care was increased in women whose primary language was English (OR, 2.12; 95% CI, 1.12 to 4.02), with stage I versus stage III (OR, 3.17; 95% CI, 1.18 to 8.50), treated at GH versus CHBAH (OR, 2.25; 95% CI, 1.39 to 3.65), and those who received radiotherapy (OR, 2.32; 95% CI, 1.69 to 3.18). It was decreased in those age < 45 years or 45-65 years versus ≥ 65 years (OR, 0.44; 95% CI, 0.27 to 0.73; OR, 0.51; 95% CI, 0.34 to 0.77; Table 5).
TABLE 5.
Sociodemographic and Clinical Factors in Relation to Endocrine Therapy Receipt Within 365 Days From Diagnosis by Women in the ASCO-E Cohort
DISCUSSION
In this cohort from 5 public SA hospitals, we found variations in the concordance of BC care with ASCO’s quality measures. Regarding chemotherapy, 73% of patients in the ASCO-C cohort and 67% with an ESMO guideline indication for chemotherapy received their first dose within 120 days from diagnosis. In the ASCO-R and the ASCO-E cohorts, 37% and 72% of patients initiated radiotherapy and endocrine therapy, respectively, within 365 days. Women living < 20 km from the treating hospital were less likely to receive care concordant with the chemotherapy and radiotherapy measures, and those who primarily spoke English were more likely to receive care concordant with the endocrine therapy measure. HIV infection showed no association with care quality, and study hospitals’ provision of measure-concordant care varied.
Measure performance is often lower than expected on initial evaluation in US hospitals. Baseline chemotherapy, radiotherapy, and endocrine therapy compliance rates at hospitals from the National Cancer Institute Community Centers Cancer Program were 85%, 79%, and 58% but improved to 93%, 92%, and 92% with just implementation of real-time reporting.19 Increased data capture with routine measurement of performance likely contributed to the rapid improvement, suggesting that baseline reports may underestimate actual care quality. Investigators at Parkland Memorial Hospital in Dallas saw drastic improvement in performance when data were drawn from all available clinical documents rather than from medical records alone.20 Our study used all available clinical documents though, decreasing the likelihood that performance is significantly underestimated.
The performance of our sites on the radiotherapy measure was poor. Radiotherapy equipment is in short supply in SA’s public health care system. This constraint is a barrier to the use of BCS throughout SSA. SA’s national guidelines include lack of access to radiotherapy as a contraindication to BCS, and until access improves, breast conservation should be used cautiously.
The relationship between measure-concordant care and distance from the hospital is revealing. In high-income countries (HICs), proximity to a health care facility affects BC stage at diagnosis, treatment, and outcomes.21-23 In an earlier breast cancer cohort, greater distance from CHBAH was associated with later stage at diagnosis.24 In SSA, decentralized services for HIV and noncommunicable diseases have improved access and survival, but specialized and multidisciplinary cancer care is harder to make available in remote areas.25,26 Interventions to instead reduce burdens related to frequent travel may significantly improve care quality. Poorer performance among women whose primary language is not English may reflect miscommunication between these patients and their English-speaking providers. Multilingual providers or standardized communication tools may also improve care.
Hospital-level variations persisted after adjustment for patient factors, so they do not seem to be entirely due to differences in populations served. Anecdotal reports of differences in available resources may explain our findings. The medical oncology division at IALCH/NH experienced high provider attrition during the study period and provided less chemotherapy measure-concordant care. GH serves a small population and had an especially well-organized clinical support staff, which may have allowed more consistent provision of endocrine therapy. These differences, although hard to quantify, illustrate the instability of resource-limited care systems.
We did not find associations between care quality and race, household wealth, education, relationship status, or employment. Socioeconomic factors certainly affect access to SA’s private health care system, but we found no evidence of racial or socioeconomic disparities in public system care quality. This finding may signal a true equivalence of care, a limited ability of our wealth indicator to represent socioeconomic status, or low variability in socioeconomic status within our cohort preventing any measurable effect.
A Lancet Global Health commission recently reported that, in low- and middle-income countries (LMICs), low-quality health care is responsible for more deaths than lack of access.27 Our findings similarly suggest that quality improvement will be needed to decrease BC mortality and must start with the identification of meaningful quality measures. Appropriateness of a process measure can be evaluated according to several characteristics: the feasibility of measurement, the existence of variability and substandard performance, the possibility of improving performance, and the potential to affect clinical outcomes (both the strength of the scientific evidence supporting the process under evaluation and the proportion of patients to whom it applies).28
Performance variability and room for improvement were clearly demonstrated. The ASCO measures also relied on data routinely captured from the EMR at our 5 hospitals. Such EMR systems allow for real-time quality monitoring with regular reporting at the institutional level, just as in the United States. Reporting itself can prompt quality improvement by increasing providers’ attention to the monitored processes and their own performance.
Multiple randomized trials have confirmed that the ASCO measures’ therapies improve survival. Early Breast Cancer Trialists’ Collaborative Group meta-analyses estimate the relative risks of death with appropriate use of adjuvant chemotherapy, radiotherapy after BCS, and endocrine therapy as 0.75 at 4 years, 0.82 at 15 years, and 0.71 at 4 years, respectively.4-6 There is less real-world evidence that measure-concordant care leads to improved outcomes. A study of high- versus low-performing regions in the US National Cancer Database did not show differences in survival.29 However, even low-performing regions provided > 80% concordant care, and variation was slight. Stronger associations with survival might be observed in health care systems with poorer baseline performance and less access to advanced treatments.
The ESMO chemotherapy and endocrine therapy measures were applicable to > 50% of our patients with BC. However, the radiotherapy measure was less relevant because only 20% of patients received BCS.
The ASCO measures may also overlook challenges specific to SSA. They focus on treatment initiation, ignoring long-term adherence. Moreover, patients may receive only a portion of indicated treatment types. A “treatment completion” measure used in Rwanda found that only half of patients initiate all indicated treatment types (surgery, chemotherapy, and endocrine therapy).30 Furthermore, these measures examine only care processes. Patient experience, quality of life, and safety are important quality considerations but rarely studied in LMICs.31
This study provides the largest and, as far as we are aware, first prospective, multicenter description of BC care quality in SSA. SA’s public health system is better resourced than most SSA systems but serves many impoverished patients, providing a unique context for establishing quality baselines. We believe our findings are broadly representative of SA’s public system. The South African National Cancer Registry reports 8,230 new BC diagnoses in 2014.32 Of these, approximately 3,500 were in the public system, suggesting the SABCHO study enrolled nearly 25% of the public system’s incident patients with BC over the 2-year period.
Limitations include uncertainty that our findings are generalizable to other nations. Care across SSA is drastically heterogeneous.33 In addition, we captured only patient-level factors and, thus, are unable to assess the potentially significant effects of system-level factors (eg, number of providers, chemotherapy chairs and radiotherapy machines per facility, drug stock-outs) on care quality. Data on care received outside our study sites were not available, although we believe that this limitation had little effect on our findings; public BC care is centralized at tertiary hospitals, and public patients are unlikely to seek private care for these expensive services.
In summary, in our cohort of patients with BC receiving care in five public hospitals in SA, baseline care was reasonably concordant with the ASCO BC care quality measures for chemotherapy and endocrine therapy but poor for radiotherapy. Measure-concordant care was associated with proximity to the hospital and with speaking English, and study hospitals varied in performance. The measures, designed for HICs, likely require adaptation for use in LMICs. Although the measures proved informative about care quality in SA’s public system, they call for quality improvement initiatives that target both high-risk patient groups and system-level barriers to the provision of high-quality care. Such initiatives are necessary, although not sufficient, to eliminate the disparities in outcomes between SA patients with BC and those from HICs.
Footnotes
This work was supported by the Conquer Cancer Foundation of the American Society of Clinical Oncology (2018 Endowed Young Investigator Award in memory of Evelyn H. Lauder [D.S.O]); the National Cancer Institute of the National Institutes of Health (grant number NCI 1R01CA192627 [J.S.J, M.J., A.I.N. and P.R.] and NCI 3P30CA13696 [A.I.N.]); a University of Witwatersrand/South African Medical Research Council / University of the Witwatersrand Common Epithelial Cancer Research Centre grant (P.R.).
AUTHOR CONTRIBUTIONS
Conception and design: Daniel S. O’Neil, Sarah Nietz, Maureen Joffe, Judith S. Jacobson, Alfred I. Neugut, Paul Ruff, Herbert Cubasch
Financial support: Daniel S. O’Neil, Maureen Joffe, Alfred I. Neugut, Paul Ruff
Administrative support: Maureen Joffe, Alfred I. Neugut, Paul Ruff
Provision of study material or patients: Daniel S. O’Neil, Urishka Singh, Laura Stopforth, Maureen Joffe, Alfred I. Neugut, Paul Ruff
Collection and assembly of data: Daniel S. O’Neil, Wenlong Carl Chen, Oluwatosin Ayeni, Sarah Nietz, Ines Buccimazza, Urishka Singh, Sharon Čačala, Laura Stopforth, Maureen Joffe, Alfred I. Neugut, Herbert Cubasch
Data analysis and interpretation: Daniel S. O’Neil, Sarah Nietz, Sharon Čačala, Maureen Joffe, Katherine D. Crew, Judith S. Jacobson, Alfred I. Neugut, Paul Ruff, Herbert Cubasch
Manuscript writing: All authors
Final approval of manuscript: All authors
Accountable for all aspects of the work: All authors
AUTHORS’ DISCLOSURES OF POTENTIAL CONFLICTS OF INTEREST
The following represents disclosure information provided by authors of this manuscript. All relationships are considered compensated unless otherwise noted. Relationships are self-held unless noted. I = Immediate Family Member, Inst = My Institution. Relationships may not relate to the subject matter of this manuscript. For more information about ASCO’s conflict of interest policy, please refer to www.asco.org/rwc or ascopubs.org/jgo/site/misc/authors.html.
Open Payments is a public database containing information reported by companies about payments made to US-licensed physicians (Open Payments).
Daniel S. O’Neil
Honoraria: Ipsen
Alfred I. Neugut
Stock and Other Ownership Interests: Stemline Therapeutics
Consulting or Advisory Role: Otsuka Pharmaceutical, United BioSource, EHE International
Expert Testimony: Hospira
Paul Ruff
Research Funding: MSD (Inst), Janssen Oncology (Inst), Roche (Inst), AstraZeneca (Inst), Novartis (Inst), Pfizer (Inst)
Travel, Accommodations, Expenses: Merck Serono, MSD Oncology, Pfizer, Dr. Reddy’s Laboratories, AstraZeneca, Roche
No other potential conflicts of interest were reported.
REFERENCES
- 1.Ferlay J, Shin HR, Bray F, et al. Estimates of worldwide burden of cancer in 2008: GLOBOCAN 2008. Int J Cancer. 2010;127:2893–2917. doi: 10.1002/ijc.25516. [DOI] [PubMed] [Google Scholar]
- 2.Pace LE, Shulman LN. Breast cancer in sub-Saharan Africa: Challenges and opportunities to reduce mortality. Oncologist. 2016;21:739–744. doi: 10.1634/theoncologist.2015-0429. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Desch CE, McNiff KK, Schneider EC, et al. American Society of Clinical Oncology/National Comprehensive Cancer Network quality measures. J Clin Oncol. 2008;26:3631–3637. doi: 10.1200/JCO.2008.16.5068. [DOI] [PubMed] [Google Scholar]
- 4.Early Breast Cancer Trialists’ Collaborative Group (EBCTCG) Peto R, Davies C, et al. Comparisons between different polychemotherapy regimens for early breast cancer: Meta-analyses of long-term outcome among 100,000 women in 123 randomised trials. Lancet. 2012;379:432–444. doi: 10.1016/S0140-6736(11)61625-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Early Breast Cancer Trialists’ Collaborative Group (EBCTCG) Darby S, McGale P, et al. Effect of radiotherapy after breast-conserving surgery on 10-year recurrence and 15-year breast cancer death: Meta-analysis of individual patient data for 10,801 women in 17 randomised trials. Lancet. 2011;378:1707–1716. doi: 10.1016/S0140-6736(11)61629-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Early Breast Cancer Trialists’ Collaborative Group (EBCTCG) Davies C, Godwin J, et al. Relevance of breast cancer hormone receptors and other factors to the efficacy of adjuvant tamoxifen: Patient-level meta-analysis of randomised trials. Lancet. 2011;378:771–784. doi: 10.1016/S0140-6736(11)60993-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Neuss MN, Desch CE, McNiff KK, et al. A process for measuring the quality of cancer care: The Quality Oncology Practice Initiative. J Clin Oncol. 2005;23:6233–6239. doi: 10.1200/JCO.2005.05.948. [DOI] [PubMed] [Google Scholar]
- 8.Bergerot CD, Philip EJ, Bergerot PG, et al. Quality Oncology Practice Initiative can guide and improve oncology providers’ training in Brazil. J Glob Oncol. 2016;3:189–193. doi: 10.1200/JGO.2016.006148. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Lim GC, Aina EN, Cheah SK, et al. Closing the global cancer divide–performance of breast cancer care services in a middle income developing country. BMC Cancer. 2014;14:212. doi: 10.1186/1471-2407-14-212. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Republic of South Africa National Department of Health: Breast Cancer Control Policy. South Africa, National Department of Health, 2017. http://www.health.gov.za/index.php/2014-08-15-12-53-24?download=2533:breast-cancer-policy.
- 11.Cubasch H, Ruff P, Joffe M, et al. South African breast cancer and HIV outcomes study: Methods and baseline assessment. J Glob Oncol. 2017;3:114–124. doi: 10.1200/JGO.2015.002675. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Statistics South Africa: Census 2011 Statistical release (Revised) - P0301.4, 2012. http://www.statssa.gov.za/publications/P03014/P030142011.pdf.
- 13. Human Sciences Research Council: The Fifth South African National HIV Prevalence, Incidence, Behavior and Communication Survey, 2017: HIV Impact Assessment Summary Report. Cape Town, HSRC Press, 2018. http://www.hsrc.ac.za/uploads/pageContent/9234/SABSSMV_Impact_Assessment_Summary_ZA_ADS_cleared_PDFA4.pdf.
- 14.Coovadia H, Jewkes R, Barron P, et al. The health and health system of South Africa: Historical roots of current public health challenges. Lancet. 2009;374:817–834. doi: 10.1016/S0140-6736(09)60951-X. [DOI] [PubMed] [Google Scholar]
- 15.Coetzee WC, Apffelstaedt JP, Zeeman T, et al. Disparities in breast cancer: Private patients have better outcomes than public patients. World J Surg. 2018;42:727–735. doi: 10.1007/s00268-017-4187-0. [DOI] [PubMed] [Google Scholar]
- 16.Senkus E, Kyriakides S, Ohno S, et al. Primary breast cancer: ESMO clinical practice guidelines for diagnosis, treatment and follow-up. Ann Oncol. 2015;26(suppl 5):v8–v30. doi: 10.1093/annonc/mdv298. [DOI] [PubMed] [Google Scholar]
- 17.McCormack VA, Joffe M, van den Berg E, et al. Breast cancer receptor status and stage at diagnosis in over 1,200 consecutive public hospital patients in Soweto, South Africa: A case series. Breast Cancer Res. 2013;15:R84. doi: 10.1186/bcr3478. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Rutstein SO, Johnson K: The DHS wealth index. 2004. http://dhsprogram.com/publications/publication-cr6-comparative-reports.cfm.
- 19. Spain P, Teixeira-Poit S, Halpern MT, et al: The National Cancer Institute Community Cancer Centers Program (NCCCP): Sustaining quality and reducing disparities in guideline-concordant breast and colon cancer care. Oncologist 22:910-917, 2017. [DOI] [PMC free article] [PubMed]
- 20.Lodrigues W, Dumas J, Rao M, et al. Compliance with the commission on cancer quality of breast cancer care measures: Self-evaluation advised. Breast J. 2011;17:167–171. doi: 10.1111/j.1524-4741.2010.01047.x. [DOI] [PubMed] [Google Scholar]
- 21.Ambroggi M, Biasini C, Del Giovane C, et al. Distance as a barrier to cancer diagnosis and treatment: Review of the literature. Oncologist. 2015;20:1378–1385. doi: 10.1634/theoncologist.2015-0110. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Schroen AT, Brenin DR, Kelly MD, et al. Impact of patient distance to radiation therapy on mastectomy use in early-stage breast cancer patients. J Clin Oncol. 2005;23:7074–7080. doi: 10.1200/JCO.2005.06.032. [DOI] [PubMed] [Google Scholar]
- 23.Jensen LF, Pedersen AF, Andersen B, et al. Distance to screening site and non-participation in screening for breast cancer: A population-based study. J Public Health (Oxf) 2014;36:292–299. doi: 10.1093/pubmed/fdt068. [DOI] [PubMed] [Google Scholar]
- 24.Dickens C, Joffe M, Jacobson J, et al. Stage at breast cancer diagnosis and distance from diagnostic hospital in a periurban setting: A South African public hospital case series of over 1,000 women. Int J Cancer. 2014;135:2173–2182. doi: 10.1002/ijc.28861. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Boyer S, Eboko F, Camara M, et al. Scaling up access to antiretroviral treatment for HIV infection: The impact of decentralization of healthcare delivery in Cameroon. AIDS. 2010;24(suppl 1):S5–S15. doi: 10.1097/01.aids.0000366078.45451.46. [DOI] [PubMed] [Google Scholar]
- 26.Newman PM, Franke MF, Arrieta J, et al. Community health workers improve disease control and medication adherence among patients with diabetes and/or hypertension in Chiapas, Mexico: An observational stepped-wedge study. BMJ Glob Health. 2018;3:e000566. doi: 10.1136/bmjgh-2017-000566. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Kruk ME, Gage AD, Joseph NT, et al. Mortality due to low-quality health systems in the universal health coverage era: A systematic analysis of amenable deaths in 137 countries. Lancet. 2018;392:2203–2212. doi: 10.1016/S0140-6736(18)31668-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Rubin HR, Pronovost P, Diette GB. From a process of care to a measure: The development and testing of a quality indicator. Int J Qual Health Care. 2001;13:489–496. doi: 10.1093/intqhc/13.6.489. [DOI] [PubMed] [Google Scholar]
- 29.Kantor O, Wang C-H, Yao K. Regional variation in performance for commission on cancer breast quality measures and impact on overall survival. Ann Surg Oncol. 2018;25:3069–3075. doi: 10.1245/s10434-018-6592-5. [DOI] [PubMed] [Google Scholar]
- 30.O’Neil DS, Keating NL, Dusengimana JMV, et al. Quality of breast cancer treatment at a rural cancer center in Rwanda. J Glob Oncol. 2018;4:1–11. doi: 10.1200/JGO.2016.008672. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31. Ganz PA, Hassett MJ, Miller DC: Challenges and Opportunities in Delivering High-Quality Cancer Care: A 2016 Update. Am Soc Clin Oncol Educ Book e294-300, 2016. [DOI] [PubMed]
- 32. National Cancer Registry: Cancer in South Africa: 2014 Full Report. South Africa, National Institute for Communicable Diseases, 2017. http://www.nicd.ac.za/wp-content/uploads/2017/03/2014-NCR-tables-1.pdf.
- 33.Sutter SA, Slinker A, Balumuka DD, et al. Surgical management of breast cancer in Africa: A continent-wide review of intervention practices, barriers to care, and adjuvant therapy. J Glob Oncol. 2016;3:162–168. doi: 10.1200/JGO.2016.003095. [DOI] [PMC free article] [PubMed] [Google Scholar]