Abstract
Alcohol use has been associated with increased risk and poorer prognosis of various types of cancer, but the extent of actual screening for alcohol use in hospitalized cancer patients has received limited attention. In a random sample of 618 Connecticut patients diagnosed in 2002 with invasive cancer at age 20 + years, ascertained from a population-based registry, some information on usual frequency of drinking was found in hospital records for 478 (77.3%), but 143 (57.2%) of the 250 current drinkers were poorly described (e.g., “social,” “occasional”). Alcohol use disorders were rarely mentioned. Gamma glutamyl transpeptidase was found for only 21 (3.4%). Mean corpuscular volume (MCV) was found for 434 (70.2%) and was associated with heavier drinking, independent of other predictors of MCV. Better screening is needed for all hospitalized cancer patients, to plan for interventions and conduct follow-up studies of prognosis.