Developing and Implementing Context-Specific Quantity Frequency (CSQF) for Assessing Drinking Patterns and Alcohol Consumption in a General Population

Abstract

Objective: This thesis aimed to develop a questionnaire with a contextual approach to accurately measure alcohol drinking patterns and alcohol consumption in a community of southern Thailand Methods: This was a mixed-method study that consisted of both qualitative with in-depth interview techniques and quantitative approaches using a community-based cross-sectional survey conducted among adults aged ≥15 years in Songkhla Province, Thailand. There were two parts in this study. The context-specific quantity frequency (CSQF) instrument was developed from a literature review and the results of in-depth interviews and expert comments (part 1). The CSQF was then validated by the traditional beverage-specific quantity frequency (BSQF) instrument using the same 3- month retrospective time frame and applied in a random order with each participant. The implications of the CSQF were then examined (part 2). Results: In the qualitative approach, 15 current drinkers were interviewed. There were four themes of Thai drinking behavior: (i) group drinking with at least one friend who is a buddy drinker who knows the drinker's limits; (ii) social and cultural drinking; (iii) seasonal drinking; and (iv) drinking and driving. Regarding seasonal drinking and the culture of drinking in a group, a CSQF questionnaire using probing questions on drinking context was developed and experts from the four main regions of Thailand revised the CSQF. The final version of the CSQF questionnaire elicited information on location, partner or partners, beverage, quantity, and frequency for five common drinking situations. The CSQF questionnaire allowed for a description of 15 types of 'drinking events'. Each drinking event was a unique combination of one specified drinking situation, location, drinking partner(s), beverage, and volume consumed. For convergent validity testing, 183 current drinkers in the last three months were identified from a total of 804 participants. At the individual level, total alcohol consumption of almost all types of beverage according to the CSQF questionnaire was higher than based on the BSQF questionnaire in approximately 50% of current drinkers and was mainly due to the higher report of average quantity. At the sample level, there were no significant differences in the average daily intake, 3-month intake per drinker or per capita consumption between the instruments. The interview duration and burden of answering the questions by the participants for the CSQF were not significantly higher than for the BSQF. This thesis revealed two implications of the use of the prototype CSQF questionnaire. First, the drinking context associated with different drinking intensities using the CSQF instrument can be described. There were 412 drinking events (215 low- ,79 medium-, and 118 high-intensity) from the 183 current drinkers. More than half of these events occurred in special situations (i.e., holiday, party, and cultural drinking). About half of the drinking events occurred outside the drinker's house, usually in the homes of other people, and most drinking events occurred among friends. The multilevel analysis showed that higher drinking intensity was associated with a higher level of education (adjusted odds ratio [OR] 4.74 for medium- and aOR 5.23 for high- intensity) and with a special drinking situation (aOR 2.46 for medium- and aOR 2.78 for high-intensity). Non-beer beverages (aOR 7.27) were associated with medium risk of acute harm. Second, the hypothesis that drinking alcohol even at a low level is associated with low-to-moderate health-related quality of life (HRQOL) was tested and compared to abstainers. A case-control study was conducted as part of the large survey. The adult participants (15 years) with low-to-moderate HRQoL scores (EuroQol [EQ- 5D] index ≤0.8, n = 108) were assigned to be the case group and those with high HRQOL (EQ-5D index >0.8, n=443) were the control group. Compared with lifetime abstention, light drinking (0.1-7 drinks/week) was significantly associated with low- moderate HRQOL (odds ratio (OR) = 3.16, 95% confidence interval (CI) 1.08-9.20); however, no significant associations were found for moderate to heavy drinking (OR = 3.55, 95% CI 0.49-25.49) or abstinence during the previous 12 months (OR = 0.86, 95% CI 0.45-1.63). Conclusion: The contextual approach appears to be appropriate for an alcohol survey. Comprehensive assessment of the drinking context (i.e., drinking situation, location, drinking partner(s), beverage, and sharing of beverages in a group) provides valuable information for clinical practice and for alcohol policies and helps to more clearly understand drinking behavior. However, the current version of the CSQF instrument needs to be explored further for reliability (e.g., inter-interviewer reliability, parallel- form test-retest reliability) and revised again to be more user-friendly.