Background: On May 1, 1954, a prospective mortality study was instituted in the Medical Department of the New England Mutual Life Insurance Company for all persons on whom an electrocardiogram (ECG) was made. Details were coded on an 80-column, mark-sense punched card for the ECG interpretation, for clinical findings, and for demographic/insurance data.
Results: Mortality is based on the experience of 231 policy-holders with complete right bundle branch block (RBBB) and 45 policyholders with complete left bundle branch block (LBBB). These were drawn from 28,687 interpretation records 1954-1966, if there was some follow-up (FU) exposure between 1954 and 1975. Mortality data are for all ages and all durations combined. In cases with associated rated cardiovascular (CV) impairment, there were 22 observed vs 7.72 expected deaths in RBBB, and 6 observed vs 2.72 expected deaths in LBBB. Exposures and deaths were smaller when there was no rated CV impairment associated with the ECG abnormality: 11 observed vs 8.11 expected deaths in RBBB, and 3 observed vs 1.78 expected deaths in LBBB.
Conclusions: In complete RBBB, excess mortality was significant at the Poisson 95% confidence level when a rated CV impairment was associated with the RBBB, but the excess was minimal and not significant when the RBBB was essentially an isolated finding with no associated CV impairment. In LBBB the numbers of deaths were too few to permit even a 90% confidence level of significance when there was an associated CV impairment (the 2.72 expected deaths were just above the lower limit of 2.6 deaths at the 90% level).