Screening for alcohol problems: what makes a test effective?

Alcohol Research & Health, Wntr, 2004 by Scott H. Stewart, Gerard J. Connors

Analogous to the positive likelihood ratio, the negative likelihood ratio can be used to calculate a specific patient's probability of having a disease (e.g., being an at-risk drinker) based on the patient's history and a negative result on the screening test. For instance, assuming that the two patients in the previous situation both had negative screening results on the AUDIT, calculations to determine the postscreen probability of at-risk drinking would yield values of 11 percent for Patient 1 and 5 percent for Patient 2. (5)

This example demonstrates that likelihood ratios can be useful for predicting the risk in individual patients for a certain disorder; however, there are some limitations to their use. For example, a clinician must have good clinical acumen to accurately predict a given patient's probability of having the disorder based on the patient's history (i.e., the patient's pretest probability), which is needed to estimate the posttest probability of having the disorder as accurately as possible.

SUMMARY

Screening for AUDs and other drinking problems is warranted in a variety of settings and contexts because these conditions have a relatively high prevalence, can lead to substantial adverse consequences for individuals and society, and can be significantly improved by appropriate treatment. Screening tests should be validated in populations similar to the one being tested and for the specific disorder or group of disorders of interest. The selection of appropriate cutoff scores that balance sensitivity and specificity is a key consideration when using screening tests. With the help of appropriately conducted screening tests, clinicians can better predict the probability that individual patients do or do not have a given disorder. Specific examples of screening tests for AUDs and other alcohol-related problems, as well as of subsequent brief interventions, are highlighted in the remaining articles in this and the companion issue of Alcohol Research & Health.