Research Article: Does prior recall of past week alcohol use affect screening results for at-risk drinking? Findings from a randomized study

Date Published: June 4, 2019

Publisher: Public Library of Science

Author(s): Andreas Staudt, Jennis Freyer-Adam, Christian Meyer, Gallus Bischof, Ulrich John, Sophie Baumann, Joel Msafiri Francis.


Underreporting of alcohol consumption is one of the major challenges in survey research including self-reports. The aim of this study was to test whether underreporting can be reduced by prompting respondents to first reflect on their drinking in the past week and then answer quantity-frequency based screening questions on their typical alcohol use. Data come from 2,379 adults (54% female; mean age = 31.8 years, SD = 11.4 years) consecutively recruited at a local registration office in northeastern Germany. Participants responded to an electronic, self-administered questionnaire on different health behaviors. They were randomized to receiving the Alcohol Use Disorders Identification Test—Consumption (AUDIT-C) either before or after the assessment of past week timeline follow-back questions. Logistic regression models were calculated predicting positive screening results for at-risk drinking. Potential interaction effects with gender, age and educational background were explored. Results show that the assessment of past week alcohol consumption prior to the assessment of the AUDIT-C reduced the odds of obtaining positive screening results (OR = 0.83; 95% CI = 0.70–0.99). There were no interaction effects with gender, age and educational background. As a secondary finding, participants reported consistently lower alcohol consumption in the alcohol measure that was administered later in the questionnaire. Preceding questions about alcohol consumption in the past week reduced the probability of positive screening results for at-risk drinking. Our findings suggest that prompting people to recall past week alcohol use prior to screening may not be a solution to reduce underreporting.

Partial Text

Misreporting of alcohol consumption is a problem in population surveys: people struggle in reporting their actual alcohol consumption accurately [1]. Misreporting on self-report measures may compromise the detection of individuals with hazardous alcohol consumption. Furthermore, misreporting attenuates the quality of evidence that is used to estimate the risk for alcohol-related problems and to derive low-risk drinking guidelines [2, 3]. While both over- and underreporting of alcohol consumption have been observed, the gap between self-reported alcohol consumption and estimates derived from alcohol sales data indicates that underreporting is the more severe problem [4]. Underreporting has been found to vary by gender and age [5, 6], drinking pattern [4, 7] as well as by features of the research instrument [8]. If quantity-frequency items and retrospective diary methods were compared within the same sample, the latter would yield substantively higher estimates of average alcohol consumption [9–11].

Data were collected as part of the randomized controlled trial “Testing a proactive expert system intervention to prevent and to quit at-risk alcohol use” (PRINT, German Clinical Trials Register: DRKS00014274, date of registration: 2018/03/12) described in more detail elsewhere [32]. The ethics committee of the University Medicine Greifswald approved the study (protocol number BB 147/15).

Our study revealed two main findings. First, respondents who were asked to recall their past week alcohol use before responding to the AUDIT-C had reduced odds of receiving a positive screening result. Furthermore, in both conditions the alcohol measure that was assessed first revealed higher values in comparison to the other condition, in which the respective measure was assessed second.