Association of Previous Clinical Breast Examination With Reduced Delays and Earlier-Stage Breast Cancer Diagnosis Among Women in Peru

Key Points

Question  Is previous clinical breast examination associated with earlier presentation of self-detected disease and earlier breast cancer stage at diagnosis?

Findings  In this cross-sectional investigation of 113 women studied at a regional cancer center in Peru, 105 had self-detected breast cancers. Women who underwent previous clinical breast examination were more likely to have shorter delays from symptom development to presentation and be diagnosed with early-stage cancer.

Meaning  In a health care system that relies on women presenting with symptoms of breast cancer, interaction with a health care professional, including clinical breast examination, may improve awareness, reduce delays, and lead to earlier-staged disease at diagnosis.

Importance  Mammographic screening is impractical in most of the world where breast cancers are first identified based on clinical signs and symptoms. Clinical breast examination may improve early diagnosis directly by finding breast cancers at earlier stages or indirectly by heightening women’s awareness of breast health concerns.

Objective  To investigate factors that influence time to presentation and stage at diagnosis among patients with breast cancer to determine whether history of previous clinical breast examination is associated with earlier presentation and/or earlier cancer stage at diagnosis.

Design, Setting, and Participants  In this cross-sectional analysis of individual patient interviews using a validated Breast Cancer Delay Questionnaire, 113 (71.1%) of 159 women with breast cancer treated at a federally funded tertiary care referral cancer center in Trujillo, Peru, from February 1 through May 31, 2015, were studied.

Main Outcomes and Measures  Method of breast cancer detection and factors that influence time to and stage at diagnosis.

Results  Of 113 women with diagnosed cancer (mean [SD] age, 54 [10.8] years; age range, 32-82 years), 105 (92.9%) had self-detected disease. Of the 93 women for whom stage was documented, 45 (48.4%) were diagnosed with early-stage disease (American Joint Committee on Cancer [AJCC] stage 0, I, or II), and 48 (51.6%) were diagnosed with late-stage disease (AJCC stage III or IV). Mean (SD) total delay from symptom onset to initiation of treatment was 407 (665) days because of patient (mean [SD], 198 [449] days) and health care system (mean [SD], 241 [556] days) delay. Fifty-two women (46.0%) had a history of clinical breast examination, and 23 (20.4%) had undergone previous mammography. Women who underwent a previous clinical breast examination were more likely to have shorter delays from symptom development to presentation compared with women who had never undergone a previous clinical breast examination (odds ratio, 2.92; 95% CI, 1.30-6.60; P = .01). Women diagnosed with shorter patient delay were more likely to be diagnosed with early-stage disease (AJCC stage 0, I, or II) than those with longer patient delay (31 [58.5%] vs 11 [30.6%], P = .01). Women who underwent a previous clinical breast examination were more likely to be diagnosed with early-stage disease compared with women who had never undergone previous clinical breast examination; this relationship remained significant after controlling for insurance and household income (odds ratio, 2.44; 95% CI, 1.01-5.95; P = .048).

Conclusions and Relevance  In a population in which most breast cancers are self-detected, previous clinical breast examination was associated with shorter patient delay and earlier stage at breast cancer diagnosis. In regions of the world that lack mammographic screening, the routine use of clinical breast examination may provide a resource-appropriate strategy for improving breast cancer early diagnosis.

In their updated guidelines, the American Cancer Society “does not recommend clinical breast examination (CBE) for breast cancer screening among average-risk women at any age.”^{1(p 1608)} This recommendation assumes that women have access to and are undergoing mammographic screening and should not be presumed to apply in settings where mammographic screening is lacking.^2,3 Because population-based mammographic screening is impractical in most of the world, most breast cancers in low- and middle-income countries (LMICs) are initially identified on the basis of clinical signs and symptoms.^4,5 In LMICs, the systematic implementation of CBE may play a crucial role in the early diagnosis of breast cancer by finding breast cancers at earlier stages or heightening women’s awareness of breast health concerns so that women present for evaluation earlier in their disease.

To understand barriers to early diagnosis of breast cancer in LMICs, we investigated factors that influence time to and stage at diagnosis among patients with breast cancer at a public cancer hospital in northern Peru. Our goals were to determine in this population whether (1) most breast cancers are self-detected, (2) previous CBE is associated with earlier presentation of self-detected disease, and (3) earlier presentation is associated with earlier cancer stage at diagnosis.

All patients with breast cancer who visited a surgeon at Instituto Regional de Enfermedades Neoplásicas (IREN Norte) from February 1 through May 31, 2015, were eligible to participate. Patients were recruited and written informed consent obtained after a clinical encounter. Data were deidentified, and each study participant was assigned a number. This cross-sectional investigation was conducted at the IREN Norte, a federally funded referral cancer center in Trujillo, Peru. Approval was obtained from the institutional ethics review committee at IREN Norte, and institutional review board exemption was granted from the University of Washington, Seattle.

A Breast Cancer Delay Questionnaire previously validated in Mexico⁶ was modified for Peru. Trained personnel administered patient questionnaires in Spanish, collecting information regarding demographic characteristics, breast cancer detection method, initial perception of the breast problem, symptoms, patient-identified factors that contribute to delays in receipt of medical attention, past use of health care services, diagnosis, treatment, and general breast health knowledge. Hospital staff provided breast cancer stage at diagnosis, when available.

Patients were categorized as having shorter delays (<90 days) or longer delays (≥90 days). Patients with available cancer staging data were categorized as having early-stage disease (American Joint Committee on Cancer [AJCC] stage 0, I, or II; n = 45) or late-stage disease (AJCC stage III or IV; n = 48). No significant differences were identified when comparing women who had a documented stage (n = 93 [81.4%]) with those who did not (n = 20 [17.7%]). Previous CBE was defined as any CBE by a trained health care professional before development of the current breast cancer episode that did not lead to a breast cancer diagnosis. In bivariate analysis, continuous variables were assessed using independent unpaired, 2-tailed t tests; categorical variables were assessed using χ² tests or Fisher exact tests. On multivariate analysis, the association between stage at diagnosis and CBE and between patient delay and CBE was assessed with logistic regression comparing these unadjusted associations with adjusted models, including monthly income and insurance.

On multivariate analysis, the test used to calculate the P values was logistic regression, and the cutoff for significance used was .05. STATA/IC software, version 14.2 (StataCorp), was used for all calculations.

Of the 159 patients who qualified for study inclusion, 113 participated (71.1% enrollment). Mean (SD) age was 54 (10.8) years (range, 32-82 years). Patients lived a mean of 3 hours from the oncology institute (range, 0-36 hours). Most households (75 [66.4%]) were covered by Seguro Integral de Salud, the government-funded medical insurance for Peruvian households earning less than minimum wage.

Of 113 diagnosed cancers, 105 (92.9%) were self-detected by the patient. One cancer (0.9%) was detected by a screening CBE, and 7 (6.2%) were initially detected by screening mammography. Fifty-two patients (46.0%) reported having ever undergone previous CBE, whereas 23 (20.4%) reported having ever undergone previous mammography.

Delay information was available for 109 of the 113 women (96.5%). Mean (SD) total delay (symptom onset to initiation of treatment) was 407 (665) days for the 74 women who had this data point available. Mean (SD) patient-attributable delay (symptom onset to first medical visit) was 198 (449) days for the 109 women who had this statistic recorded. Mean (SD) health system delay (initial medical consultation at any facility to initiation of treatment) was 241 (556) days for the 76 women who had this data point available.

Women with shorter patient delay were compared with those with longer patient delay (Table 1). Shorter patient delay was associated with married status, higher household income, and employer or private insurance. Women with shorter patient delay were more likely than those with longer delay to have undergone a previous CBE (38 [56.7%] vs 13 [31.0%], P = .009). Neither history of previous screening mammography nor routine performance of breast self-examination by the patient influenced patient delay. Women diagnosed with shorter patient delay were more likely to be diagnosed with early-stage disease (AJCC stage 0, I, or II) than those with longer patient delay (31 [58.5%] vs 11 [30.6%], P = .01). Women who underwent a previous CBE were more likely to have shorter delays from symptom development to presentation compared with women who had never undergone a previous CBE (odds ratio, 2.92; 95% CI, 1.30-6.60; P = .01).

Stage at diagnosis included 45 women with early-stage disease and 48 women with late-stage disease (stage 0: 2 women [1.8%]; stage I: 3 [2.7%]; stage II: 40 [35.4%]; stage III: 45 [39.8%]; stage IV: 3 [2.7%]). Twenty women (17.7%) did not have a recorded stage at diagnosis.

In a comparison of early- and late-stage disease (Table 2), age, self-reported household income, and educational level were not associated with stage at diagnosis. Early stage at diagnosis was associated with being married, having employer-provided or private health insurance, and living within 3 hours of the cancer hospital. History of breast self-examination was not associated with stage at diagnosis. Previous CBE and previous mammography were associated with early stage at diagnosis on univariate analysis, but previous mammography did not retain significance on multivariate analysis after controlling for socioeconomic variables. On multivariate analysis, after adjusting for income and insurance, women who underwent a previous CBE were more likely to be diagnosed with early-stage breast cancer compared with women who had never undergone a previous CBE (odds ratio, 2.44; 95% CI, 1.01-5.95; P = .048) (Table 3).

Only 7% of our study population had their breast cancer diagnosed by screening before the disease became apparent to the patient based on self-detected symptoms. Although breast cancer screening through CBE or mammography exists to a limited degree in LMICs, its use is generally opportunistic and sporadic rather than organized and population based, which contributes to the advanced stages at which breast cancers are detected.⁷

This study suggests that the value of CBE goes beyond its utility as a cancer screening tool. In this patient population, women who had previously undergone a CBE presented at an earlier time point after the development of symptoms and were then found to have less advanced disease at presentation compared with women who never underwent a previous CBE. In addition, CBE appears to play an educational role in improving breast health awareness that translates into cancer down-staging in an otherwise unscreened population.

Of note, previous exposure to CBE had as much of an association with staging as did health insurance. This finding is encouraging and suggests that making available a simple clinical service (CBE) to low-income women may have a positive effect on down-staging disease. Widespread implementation of CBE as an educational and screening tool is feasible.⁸ The observations from this study are likely to be applicable to other LMICs where delay is a common obstacle to improving breast cancer outcomes.

In northern Peru, where population-based mammographic screening is unavailable and advanced-stage breast cancer presentation is common, prior exposure to CBE was associated with reduced patient delay and earlier-stage breast cancer at diagnosis. In LMICs, CBE may function as a screening method and an educational tool, promoting breast health awareness and improving early diagnosis. Widespread application of the routine CBE in LMICs has the potential to reduce delays, downstage disease, and improve survival for women with breast cancer.

Corresponding Author: Benjamin O. Anderson, MD, Department of Surgery, University of Washington, PO Box 356410, Seattle, WA 98195 (banderso@uw.edu).

Accepted for Publication: March 9, 2017.

Published Online: May 25, 2017. doi:10.1001/jamaoncol.2017.1023

Author Contributions: Drs Romanoff and Anderson had full access to all the data in the study and take responsibility for the integrity of the data and accuracy of the data analysis.

Study concept and design: All authors.

Acquisition, analysis, or interpretation of data: Romanoff, Hayes Constant, Johnson, Cedano Guadiamos, Anderson.

Drafting of the manuscript: Romanoff, Hayes Constant, Johnson, Zunt, Anderson.

Critical revision of the manuscript for important intellectual content: All authors.

Statistical analysis: Hayes Constant, Cedano Guadiamos, Burga Vega.

Obtained funding: Romanoff, Zunt.

Administrative, technical, or material support: Romanoff, Hayes Constant, Zunt.

Study supervision: Romanoff, Zunt, Anderson.

Conflict of Interest Disclosures: Dr Anderson has reported receiving grant support from the Susan G. Komen for the Cure and unrestricted educational grants from Pfizer and Roche Pharmaceuticals. No other disclosures were reported.

Funding/Support: This project was supported by National Institutes of Health Research Training Grant 3R25TW009345 awarded to the Northern Pacific Global Health Fellows Program by the Fogarty International Center and Institutes at the National Institutes of Health; the Fulbright Scholar Program, a program of the US Department of State Bureau of Educational and Cultural Affairs; and grant R25CA092408 from the National Cancer Institute Biobehavioral Cancer Prevention and Control Training Program.

Role of the Funder/Sponsor: National Institutes of Health Research Training Grant 3R25TW009345 provided funding for the design and conduct of the study; collection, management, and interpretation of the data; and preparation of the manuscript. National Cancer Institute Biobehavioral Cancer Prevention and Control Training Program R25CA092408 funding was used in management, analysis, and interpretation of the data and preparation and review of the manuscript.

Additional Contributions: The staff volunteers of Alianza de Apoyo al Instituto Regional de Enfermedades Neoplásicas Nórte (ALINEN Nórte) (specifically,Monica Celis, Talu Diaz Claros, Sonia Elizabeth Reyes Guerra, Paquita Parodi, Ryta Rivera, Gloria Velásquez Vergara, and Maria López Vilca) provided invaluable contributions to the data collection. None of these contributors received compensation.