Author (year) | Number of included studies | Focus only on CBE | AMSTAR 2b classification | Reports on outcomes | Conclusions on CBE |
---|---|---|---|---|---|
a. RCTs b. NRSI c. Systematic reviews | a. Mortality b. Downstaging c. False positive rate d. Sensitivity | Specificity | Positive predicted value | ||||
Fletcher SW (1993) [30] | a. 2 b. 0 c. 0 | No | Low | a. No difference in mortality rate between MMR + CBE vs CBE b. Did not report c. Did not report d. 46–64% | 99.1–99.7% | 2.9–4% | + Mammography and CBE detect breast cancer in a complementary manner + Careful CBE may be as effective as mammography regarding mortality reduction |
Barton MB (1999) [32] | a. 4 b. 4 c. 0 | Yes | Low | a. No difference in mortality rate between MMR + CBE vs CBE b. Did not report c. Did not report d. Pooled results: 54.1% | 94% | 10.6% | + A well-conducted CBE can detect at least 50% of asymptomatic cancers and may contribute to mortality rate reduction in women screened -- > Screening CBE should be conducted |
Humphrey LL (2002) [33] | a. 4 b. 2 c. 1 | No | Moderate | a. 14–29% mortality reduction in trials of MMR + CBE. Mortality reductions in trials of MMR + CBE were similar to trials of CBE only b. Did not report c. 13.4% d. 40–69% | 88–99% | 4–50% | + MMR has little additive benefit in the setting of a careful, detailed CBE + No direct evidence that CBE decreases mortality |
Kosters JP (2003) [31] | a. 1 b. 0 c. 0 | No | High | a. Did not report b. Did not report c. Did not report d. Did not report | The only trial investigated CBE vs no screening was discontinued due to poor compliance -- > CBE cannot be recommended |
Elmore JG (2005) [37] | a. 4 b. 3 c. 2 | No | Low | a. Did not report b. Did not report c. 20% d. 28–54% | 94% | NR | + CBE detects some cancers missed by MMR |
a. 4 b. 1 c. 0 | No | High | a. No difference in mortality rate between MMR + CBE vs CBE (RR = 1.02, 95% CI: 0.78–1.33) b. Did not report c. Did not report d. 25.6% | NR | 1% | + Trials of CBE are ongoing -- > no benefit on mortality has been shown at this point | |
a. 4 b. 2 c. 0 | No | High | a. No evidence was found to show that CBE reduced mortality due to BC or all-cause mortality b. Did not report c. Did not report d. Did not report | No evidence was found to support the benefit of CBE, either alone or in conjunction with mammography | |
a. 3 b. 4 c. 0 | No | Moderate | a. No effect of CBE alone on mortality (based on only 1 US case-control study which also found no effect of mammography on mortality) b. Did not report c. 0.9–5.7% d. Did not report | + Lack of evidence showing benefits of CBE alone or in conjunction with mammography + No studies assessing other critical outcomes | |
Hamashima C (2016) [7] | a. 1 b. 6 c. 1 | No | Moderate | a. Based on 1 Japanese case-control study, among asymptomatic women, 1 CBE within 5 years: OR = 0.45 (95% CI: 0.22–0.89) b. Did not report c. Did not report d. 46–63% | 94.3–97.3% | NR | + CBE is not recommended for population-based screening program due to insufficient evidence |
IARC (2016) [36] | a. 6 b. 10 c. 1 | No | Moderate | a. No difference in mortality rate between MMR + CBE vs CBE (RR = 0.97, 95% CI: 0.62–1.52) b. Mumbai trial: Significant shift to a lower stage in the screening arm compared with the control arm (RR, 1.45; 95% CI: 1.09–1.93). Kerala trial: early-stage breast cancer was 43.8% in the intervention group versus 25.4% in the control group (P = 0.023) c. 5.7% d. 52–85% | 93.4–96% | 1–4% | + There is sufficient evidence that screening by CBE alone shifts the stage distribution of tumours detected towards a lower stage + There is inadequate evidence that screening by CBE alone reduces breast cancer mortality |
Mandrik O (2019) [4] | a. 0 b. 0 c. 10 | No | Moderate | a. No solid evidence of mortality reduction b. Acknowledged but did not summarise the evidence c. Higher rate of false-positive rates (did not report how higher) d. 28–36% in the community, 47–69% in RCTs in all except 1 review | > 88% in all reviews | NR | + The review could not summarise evidence on down-staging but IARC report concluded there are sufficient evidence for this outcome + More original research on benefits and harms of CBE is required + Lack of research in LMICs -- > evidence cannot be generalized to these settings |