Primary prevention studies and the healthy elderly: evaluating barriers to recruitment.

Although primary prevention studies are important tools in helping the healthy elderly stay healthy, recruiting from a community-based cohort of healthy elderly individuals for a primary prevention study involves numerous barriers. To better identify and understand these barriers, we conducted and evaluated a comprehensive recruitment strategy for a primary prevention study testing aspirin in an HMO population. In the recruitment phase, we identified healthy individuals (65 years of age or older) who were members of a large, group-model HMO in Oregon and Washington, and used computerized medical database screening, statistical sampling, health plan mailings, e-mail communication with primary care providers, and the experience of a well-established research clinic in an effort to enroll health elderly in this primary prevention trial. Among a random sample of 47,453 eligible patients over the age of 65, 44% responded to recruitment efforts, but only 3% were enrolled--an overall yield of slightly less than 2%. To evaluate these results, we then conducted focus groups with 225 randomly selected "eligible refusers." We determined that healthy elders were hesitant to give up their choice to use aspirin, unwilling to travel to the research center, and reluctant to risk their tenuous hold on good health to participate in a study of primary prevention. Awareness of these attitudes is an indispensable step toward designing effective recruitment strategies for primary prevention studies involving the healthy elderly.

A comparison of Medicare fee-for-service and a group-model HMO in the inpatient management and long-term survival of elderly individuals with syncope.

OBJECTIVE: To compare the management and survival of elderly patients hospitalized with syncope in 2 healthcare delivery systems. STUDY DESIGN: Retrospective cohort. PATIENTS AND METHODS: Using hospitalization records from Medicare and a group-model health maintenance organization (HMO) in Oregon, we identified individuals with an admission or discharge diagnosis of syncope between 1992 and 1994. Among patients 65 years or older (median age = 79 years), we randomly selected a sample of the standard Medicare patients (primarily fee-for-service; n = 473) and all of the group-model HMO patients (n = 583). Use of inpatient testing and consultation was ascertained by chart review; all-cause mortality was obtained from the National Death Index. RESULTS: Median diagnostic testing and consultation costs were the same (P = .35) in the standard Medicare population ($643) and the HMO population ($619), although the use of specific tests differed. More cardiovascular syncope was identified in the Medicare population (23% vs 18%; P = .02). Inpatient, 30-day, 1-year, and 4-year mortality rates were higher in the Medicare population (1.7%, 3.8%, 16.7%, and 50.6% respectively) than in the HMO population (0.7%, 1.5%, 13.2%, and 41.8%). After adjusting for age, gender, comorbidity, diagnostic testing, and administrative factors, the relative risk (RR) of dying was lower for group-model HMO patients (RR = 0.74; 95% confidence interval = 0.60, 0.91) than for standard Medicare patients (RR = 1.0). CONCLUSIONS: The inpatient management of these elderly patients with syncope was similar in the group-model HMO and standard Medicare settings, but survival was better for the individuals in the HMO. The reason for the differential survival is not obvious and warrants additional study.

Epidemiology of syncope in hospitalized patients.

OBJECTIVE: To describe the etiologies of syncope in hospitalized patients and determine the factors that influence survival after discharge. DESIGN: Observational retrospective cohort. SETTING: Department of Veterans Affairs hospital, group-model HMO, and Medicare population in Oregon. PATIENTS: Hospitalized individuals (n = 1,516; mean age +/- SD, 73.0 +/- 13.4 years) with an admission or discharge diagnosis of syncope (ICD-9-CM 780.2) during 1992, 1993, or 1994. MEASUREMENTS AND MAIN RESULTS: During a median hospital stay of 3 days, most individuals received an electrocardiogram (97%) and prolonged electrocardiographic monitoring (90%), but few underwent electrophysiology testing (2%) or tilt-table testing (0. 7%). The treating clinicians identified cardiovascular causes of syncope in 19% of individuals and noncardiovascular causes in 40%. The remaining 42% of individuals were discharged with unexplained syncope. Complete heart block (2.4%) and ventricular tachycardia (2. 3%) were rarely identified as the cause of syncope. Pacemakers were implanted in 28% of the patients with cardiovascular syncope and 0. 4% of the others. No patient received an implantable defibrillator. All-cause mortality +/- SE was 1.1% +/- 0.3% during the admission, 13% +/- 1% at 1 year, and 41% +/- 2% at 4 years. The adjusted relative risk (RR) of dying for individuals with cardiovascular syncope (RR 1.18; 95% confidence interval [CI] 0.92, 1.50) did not differ from that for unexplained syncope (RR 1.0) and noncardiovascular syncope (RR 0.94; 95% CI 0.77, 1.16). CONCLUSIONS: Among these elderly patients hospitalized with syncope, noncardiovascular causes were twice as common as cardiovascular causes. Because survival was not related to the cause of syncope, clinicians cannot be reassured that hospitalized elderly patients with noncardiovascular and unexplained syncope will have excellent outcomes.

Basic epidemiology of tuberculosis in Peru: a prevalence study of tuberculin sensitivity in a Pueblo joven.

Tuberculosis continues to cause significant morbidity and mortality in developing nations. As a first step in defining the magnitude of the problem in Peru, we determined the prevalence of tuberculin sensitivity in an age-stratified, community-based population on the outskirts of Lima in December 1990. The overall prevalence of 10 mm or more induration in 368 individuals was 34%. When stratified by age, the prevalence was 12% in the 0-1-year-old group, 18% in the 2-4-year-old group, 24% in the 5-14-year-old group, 60% in the 15-24-year-old group, and 68% in the > or = 25-year-old group. Vaccination with bacillus Calmette-Guerin (87% of the study population) caused significant increases in weak (5-9 mm) reactions to purified protein derivative, but did not cause strong (10 mm or more) reactions. The prevalence of tuberculous infection in this population is higher than that previously reported in Peru and in most other high-risk populations. Unfortunately, the current political and economic situation in Peru makes it difficult to implement public health measures to prevent infection and progression of infection to disease.