Geriatric patients are more likely than younger patients to be admitted to the hospital when they present to the emergency department (ED). Identifying trends in geriatric short-stay admission may inform the development of interventions designed to improve acute care for the elderly.

To evaluate trends in US geriatric short-stay hospitalizations from 1990 to 2010.

Retrospective study using the National Hospital Discharge Survey (NHDS). Trends in short-stay hospitalizations were analyzed from 1990 to 2010 for age groups 22 to 64, 65 to 74, 75 to 84, and at least 85 years using linear regression.

A total of 4.5 million survey visits representing 580 million adult hospitalizations were available for analysis; 250 million (43%) were among patients 65 years or older. Of these, 12%, 25%, and 40% were ≤1, ≤2 and ≤3 days' short-stay admissions, respectively. Between 1990 and 2010, short-stay admissions increased as a percentage of total hospitalizations for each geriatric age group but remained relatively constant for younger adults. Admissions from NHDS were similar to admissions from the ED for years where ED-specific data were available. The older a patient was (age >65 years), the more likely their admission was to have started in the ED.

For all elderly patients, short-stay admissions represented a growing proportion of total admissions, regardless of the definition of short stay. These trends were identified despite the NHDS exclusion of observation status hospitalizations. The increase in short-stay admissions was the most pronounced in the extreme elderly (age ≥85 years). Future research is needed to optimize treatment for geriatric patients presenting to the ED, some of whom, with brief observation and appropriate follow-up, may be better cared for without hospitalization.