Academic journal article Alcoholism and Psychiatry Research

Unrecorded Alcohol Consumption: Quantitative Methods of Estimation

Academic journal article Alcoholism and Psychiatry Research

Unrecorded Alcohol Consumption: Quantitative Methods of Estimation

Article excerpt

(ProQuest: ... denotes formula omitted.)


Unrecorded alcohol consumption is one of the key indicators of alcohol consumption in a country, since it is linked to the level of alcohol-related problems. 1_6 Information about such consumption is therefore necessary when developing comprehensive alcohol policies and monitoring alcohol-related behavior and outcomes. Having taken into account the many sources and types of unrecorded alcohol, assessing its prevalence appears to be a complex problem. Many assessment methods exist, but they all have advantages and limitations.7-13 It is possible to determine the approximate level of unrecorded alcohol consumption from police and customs records on seizures of home-made, counterfeit, and smuggled alcohol and travelers' imports, from estimates of ethanol use for industrial and medicinal purposes, from records on the total area of cultivated vineyards, and from sales data of ingrethents necessary for home alcohol production, such as sugar.9,14-18


Ih general, currently available approaches to assessing the level of unrecorded alcohol consumption can be divided into direct and indirect methods. Direct methods include population surveys via national screenings.7.19-20 The many types of questionnaires and survey techniques used worldwide considerably restrain data comparison.8·21-24 The simplest and most used method of evaluating total alcohol consumption - including unrecorded consumption, calculated as the difference between the total level of consumption and the level of recorded alcohol sales - is a population survey on volume and frequency of drinking.7·8 There is also the graduated quantity-frequency method, during which respondents are offered options from large doses to gradually diminishing doses.9·10·20 Screenings have certain advantages in comparison to data based on alcohol sales statistics, as they contain information about the patterns of drinking - i.e., the frequency of consumption, quantity consumed during each drinking session, and frequency of excessive drinking. Furthermore, such studies make it possible to obtain data on the nature of alcohol consumption by men and women, as well as different age groups. One of the drawbacks of screenings is that respondents tend to underreport their drinking levels. Thus, Pernanen found that the total level of alcohol consumption obtained from population surveys represented 40 to 60% of consumption estimates based on sales statistics.25 Another shortcoming of this method is an insufficient representativeness of samples, as surveys rarely include heavy drinkers, a group that is more likely to consume unrecorded alcohol than the general population. However, if one assumes that the error in estimates obtained from population surveys on alcohol consumption is a constant value, the application of standard survey procedures at different time points will make it possible to trace the dynamics of unrecorded alcohol consumption. A variation of the screening method is the use of questionnaires that address the type and volume of unrecorded alcohol consumed.7·8 To improve the reliability of responses, Warner developed a randomized response method, which allows participants to choose between answering two questions, one of which is neutral, and the other sensitive.26 With the help of this technique, it was estimated that approximately 12% of the population in Norway purchased smuggled alcohol.4 These results were two-fold higher than data obtained from standard surveys. Nordlung proposed a model of assessing total alcohol consumption in Norway, based on the assumption that systematic underreporting in surveys is proportional in character.13 In this case, the real level of alcohol consumption, C, may be calculated using the following linear regression equation:

C = f c + e

where c is the average level of individual alcohol consumption, determined using quantity-frequency questionnaire; fis the proportionality fraction, or coverage factor, and e is the randomized error approaching zero. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed


An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.