Differences in detection of alcohol use in a prenatal population (on a Northern Plains Indian Reservation) using various methods of ascertainment.

Although Fetal Alcohol Syndrome (FAS) rates have been reported to be higher in American Indian populations, no screening tool has been validated for alcohol use in American Indian women. The objectives of this study were to compare the detection of prenatal alcohol use by a self-administered questionnaire to detection by clinical interview; and to ascertain whether the screening tool would increase detection of pregnant women who are abusing alcohol. The hospital records of the women were reviewed for any history of alcohol-related illnesses or injuries to compare with results obtained from the questionnaire. Seventy women attending their first prenatal clinic visit on a reservation were screened for alcohol use. There was a wide range in detection of prenatal alcohol use (20%-71% of the sample detected) depending on the method used. There was a large variation in sensitivities (7%-93%) of the individual questions in identifying patients detected as "high risk" by the clinicians. The T-ACE screening questions significantly increased detection of alcohol use compared to detection by the clinicians (p = 0.04 Fisher's exact test). Due to the large variation between different methods of detection, it is recommended that screening tools that increase detection of alcohol use should be combined with methods of higher specificity such as using questions about quantity and frequency of alcohol intake, medical chart review and clinical interview. We also found that various interpretations of the screening questions by the patients highlighted the need to tailor the wording of individual questions to the particular patient population.