Summary of Key Characteristics of Randomized Controlled Trials of Mammography Screening.
| Study (year began) | Age (yr) | Screening Provided | Rounds (of Mammography) (No) | Screening Interval (mo) | Views | Compliance Rounds (%) | Crossovers (Controls Who Had Mammography) (%) | Follow-up (yr) | Breast Cancer Mortality* RR | Absolute Risk Reduction in Breast Cancer Mortality (per 1,000) | Number Needed to Screen 10 yr† |
|---|---|---|---|---|---|---|---|---|---|---|---|
| HIP11 ,12 (1963) | 40–64 | M + CBE vs UC | 4 | 12 | 2 | 1 = 67 | Data not available | 13 | 0.83 | 1.4193 | 916 |
| N = 61,000 | 2–4 = 54–46 | (0.7–1.0) | |||||||||
| Malmo13 (1976) | 45–70 | M vs UC | 9 | 18–24 | 1–2 | 1 = 74 | 25 | 11–13 | 0.81 | 1.0127 | 1,185 |
| N = 68,000 | 2–5 = 70 | (0.62–1.07) | |||||||||
| Two-Country14 (1977) | 40–74 | M vs UC | 3 | 24–33 | 1 | 1 = 89 | 13 | 10 | 0.68 | 1.8095 | 553 |
| N = 126,000 | 2–3 = 83–84 | (0.5–0.93) | |||||||||
| Stockholm15 ,16 (1981) | 40–64 | M vs UC | 2 | 24–28 | 1 | 1 = 81 | Controls offered screening yr 5 | 7.4 | 0.71 | 0.5369 | 1,378 |
| N = 60,000 | 2 = 81 | (0.4–1.2) | |||||||||
| Gothenburg17 (1982) | 39–59 | M vs UC | 5 | 18 | 1–2 | 1 = 85 | 20 | 8 | 0.86 | 0.3286 | 2,435 |
| N = 40,000 | 2–5 = 77–78 | (0.54–1.37) | |||||||||
| Edinburgh18 ,19 (1979) | 45–64 | M + CBE vs UC | 4 | 24 | 1–2 | 1 = 66 | Data not available | 14 | 0.71 | 0.945 | 1,482 |
| N = 45,000 | 7 = 44 | (0.53–0.95) | |||||||||
| Canada 120 ,21 (1980) | 40–49 | M + CBE vs single CBE | 4–5 | 12 | 2 | 1 = 100 | 25 | 13 | 0.98 | No reduction | Not applicable |
| N = 50,000 | 2–4 = 85–89 | (0.73–1.31) | |||||||||
| Canada 221 ,22 (1980) | 50–59 | M + CBE vs CBE | 4–5 | 12 | 2 | 1 = 100 | 16 | 10.5 | 1.14 | No reduction | Not applicable |
| N = 40,000 | 2–5 = 87–90 | (0.83–1.56) |
RR = relative risk, N = number of participants, M = screening mammography, CBE = clinical breast examination, UC = usual care.
* Length of follow-up reported for various times up to 20 years. Data estimates were closest to 10-year point.
† Number needed to screen corrected for 10 years of screening.