Coronary Heart Disease Incidence, Mortality, and Risk Factor Relationships

BACKGROUND:

Many factors may have contributed to the decline in coronary heart disease including revised coding of death certificates, behavioral modification of risk factors, improvements in emergency and other hospital care, and use of new medications and surgery. The Rochester Heart Study contributed trend data covering the time before and during the decline in coronary heart disease mortality. At that time, medical care for residents of Rochester and Olmsted County was almost exclusively provided by the Mayo Clinic and Olmsted Medical Group. Under the Rochester Epidemiology Program Project supported by program project grants from the National Institute of General Medical Sciences and the National Institute of Arthritis and Musculoskeletal and Skin Diseases, the medical records of the Olmsted Medical Group and Community Hospital were indexed for retrieval by the same system as the Mayo Clinic. Those of other outside sources such as the University of Minnesota Hospital, Rochester State Hospital, and Veterans Hospitals which pertained to the occasional admission of Olmsted County residents were also added to the Mayo Clinic maintained diagnostic index. The resulting central diagnostic file gave virtually complete case assignments for diagnosed cases of myocardial infarction, sudden unexpected death, angina pectoris, and noncoronary heart disease.

DESIGN NARRATIVE:

The centralized case index of the Mayo Clinic and the Rochester Epidemiology Program Project were used to identify any Rochester resident with a diagnosis between 1950 and 1987 of any condition suggesting that coronary heart disease might be present. Included were diagnoses of coronary heart disease, myocardial infarction, angina pectoris, coronary insufficiency, or arrhythmias. All relevant index codes were specified and a computer generated list was provided of the unit record numbers of all candidate cases. The medical histories were then retrieved and reviewed for inclusion. The screened histories were assigned to nurse abstractors whose duty it was to identify from the screening list those patients who satisfied the diagnostic criteria in the protocol as well as the incidence and residence requirements. In case finding of coronary heart disease incidence, the following had to be true: the diagnosis of angina pectoris, myocardial infarction or sudden unexpected death must have satisfied the protocol's diagnostic criteria; the diagnosis must have been a first diagnosis; the patient, at the time of first diagnosis, must have been free of diagnoses of congestive heart failure and valvular heart disease; the patient must have been a resident of Rochester at the time of diagnosis. If all of these criteria were met, the electrocardiograms were read by a physician for case verification. A precoded abstract of the history was prepared for each incidence case. The abstract included: identification and demographic data; information on smoking, hypertension, hypertension therapy, lipids, and use of estrogens; diagnostic factors such as ECG, treatment tests, angiography, and selected laboratory test results. The first myocardial infarction following the first coronary heart disease diagnosis was coded; if the first manifestation of coronary heart disease was a myocardial infarction, the second myocardial infarction was coded to provide a basis for determining reinfarction rates. In the earlier study done from 1950-1969, little data on treatment were abstracted. From 1970 to 1991t, drugs and surgery were abstracted. All death certificates for Rochester and Olmsted County residents for the period 1960-1979 were recoded to the 8th revision of the ICDA in order to obtain comparability over the study period. All deaths suspected of being due to some form of coronary heart disease were reviewed by a group of cardiologists using a clinical classification of death. The cardiologists used medical, hospital, and autopsy records to determine cause of death. Follow up was 99 percent complete. Date of last follow-up and cause of death were recorded.

The medical records for all Olmsted County residents ages 30 and over coming to autopsy for the period 1950-1978 were reviewed. Data abstracted included date of death, age at death, sex, weight, heart weight, grade of lesions in the coronary arteries, and evidence of recent or old myocardial infarction. All autopsy records for 1980 through 1984 were reviewed for coronary heart disease. Beginning in 1986, valvular heart disease data were also collected for the period 1950 through 1987.

The effect of the Diagnostic Related Groups (DRG) on incidence rates for coronary and valvular heart disease was ascertained for all cases hospitalized after January 1, 1984. The Mayo Medical Center DRG system which captured and retained diagnoses on hospitalized patients was compared with the study medical index files.

Incident cases were also followed for specific diagnostic and therapeutic procedures. This was accomplished by record review and by matching lists of case patient numbers with identification numbers of patients who underwent the following: radionuclide ejection fractions and radionuclide exercise tests beginning in 1980; arteriography, ventriculography, valve evaluation, thrombolytic infusion, percutaneous transluminal coronary angioplasty or balloon angioplasty, and electrophysiologic testing beginning in 1976; coronary artery bypass surgery, myocardial resection, and valve repair or replacement beginning in 1978; coronary intensive care unit data beginning in 1976. New procedures were monitored as they came into use.

The study completion date listed in this record was obtained from the "End Date" entered in the Protocol Registration and Results System (PRS) record.