Methodologic issues in conducting research on hospitalized older people

OBJECTIVE: To describe challenges in conducting research with hospitalized geriatric patients. DESIGN: Quasi-experimental longitudinal study with in-person interviews at baseline and discharge and a three-month postdischarge telephone interview. Study protocol required baseline interviews within 48 hours of admission for patients or 72 hours for proxies. SETTING: 813-bed urban teaching hospital. PARTICIPANTS: 240 patients, mean age 83.8 years, at moderate to high risk for functional decline during hospitalization, admitted from the emergency room to a general medical unit. MEASUREMENTS: Delays starting or interruptions continuing interviews; patient and environmental conditions potentially affecting data quality; and information on proxy use. RESULTS: Timely access to patients or proxies was a major difficulty, resulting in the screening of only 53.1% of 867 potentially eligible patients. Multiple patient contacts and visits were required to complete interviews. Only 61.3% of baseline and 28.1% of follow-up interviews were completed on the initial day of contact. Despite having on-site interviewers 7 days a week, 29% of discharge interviews were conducted by telephone. Interviews were >90% complete in 487 of 503 of encounters (97%). Baseline delays and interruptions were usually due to the presence of medical staff, off-unit tests, patient illness, nurse unavailability for interview, and need for a proxy. Most in-hospital interviews were conducted with others present. Proxies were required for approximately one-third of patients at all three interviews. CONCLUSION: Conducting clinical research with older adults in the current inpatient setting, where patients are more severely ill yet have shorter lengths of stay now than in the past, proves a challenging yet achievable goal. Effective procedures for negotiating the acute care environment are critical to successful studies.