Kopparberg (County W of Two-County Trial), Sweden 1977 [6,7,8]
Age at entry: 40 to 74 years.
Randomization: Geographic cluster, with stratification for residence (urban or rural), socioeconomic factors and size. The process for randomization has not been described. The study women were older than the control women, P < .0001, but this should not have had a major effect on the outcome of the trial.
Exclusions: Women with pre-existing breast cancer were excluded from both groups, but the numbers are reported differently in different publications.
Sample size: Variably reported, ranging from 38,562 to 39,051 in intervention and from 18,478 to 18,846 in control.
Consistency of reports: Variable.
Intervention: Three single-view MMGs every 2 years for women younger than 50 years and every 33 months for women aged 50 years and older.
Control: Usual care, with MMG at study end.
Compliance: 89% participation.
Contamination: 13% of women in the Two-County trial had MMG as part of routine care, mostly between 1983 and 1984.
Cause of death attribution: Determined by a team of local physicians (see Ostergotland).
External audit: No. However, breast cancer cases and deaths were adjudicated by a Swedish panel that included the trial's investigators.
Follow-up duration: 12 years.
Relative risk of breast cancer death, screening versus control (95% CI): 0.68 (0.52-0.89).
Randomization: Cluster by physician practices, though many randomization assignments were changed after study start. Within each practice, there was inconsistent recruitment of women, according to the physician's judgment about each woman's suitability for the trial. Large differences in socioeconomic status between practices were not recognized until after the study end.
Exclusions: More women (338) with pre-existing breast cancer were excluded from the intervention group than from the control group (177).
Sample size: 23,226 study and 21,904 control.
Consistency of reports: Good.
Intervention: Initially, two-view MMG and CBE; then annual CBE, with single-view MMG in years 3, 5, and 7.
Control: Usual care.
Compliance: 61% screened.
Cause of death attribution: Cancer Registry Data.
External audit: No.
Follow-up duration: 10 years.
Relative risk of breast cancer death, screening versus control (95% CI): 0.84 (0.63-1.12).
Comments: Randomization process was flawed. Socioeconomic differences between study and control groups probably account for the higher all-cause mortality in control women compared with screened women. This difference in all-cause mortality was four times greater than the breast cancer mortality in the control group, and therefore, may account for the higher breast cancer mortality in the control group compared with screened women. Although a correction factor was used in the final analysis, this may not adjust the analysis sufficiently.