Project SMART


Go to content

MEETINGS AND CONFERENCE

Meetings and Conference


SMART First Expert Meeting (Munich, 12 -13 March 2009)

Minutes of the meeting (PDF document)
Introduction

A review of the main activities that have been undertaken since Barcelona meeting was made. They included "homework" from three working groups on three topics:

  • Heavy drinking, binge drinking, drunkenness, dependence and problems,
  • Consumption, unrecorded consumption, drinking context,
  • Survey administration, sampling, quantitative and qualitative methods.

In addition, two instruments on literature review and survey review were elaborated and circulated among participants for piloting and comments

Discussion on the literature review and survey review including review instruments
Presentation on initial literature search based on two key words (alcohol survey, drinking survey) was pre-circulated. Following questions were raised in the discussion:

  • Are the keywords sufficient and appropriate to identify the kinds of publications we need? The databases and keywords miss publications in languages other than English. Government sponsored documents (many of the national surveys) are not available in the published (journal) literature but are published in report format or available electronically.
  • The list of papers generated by the search is in English (majority of American studies) and a much wider European literature base needs to be identified.
  • What is the purpose of the literature review? Are we looking at methodological approaches/ discussions, instruments used/ results? Would we get the kind of information we need from conducting a literature review? Two papers were considered as possible outcomes of the literature review:

A review of the surveys - methodology and measures (accepted as related to the main focus of the work).
A review of the results (the viability of this was questioned, especially to make comparisons across countries).

Following
decisions were taken to:

  • At this stage priority should be given to the review of alcohol surveys in Europe at the expense of overall literature review.
  • It was agreed that the instrument for survey review itself worked quite well. A final version after cosmetic revisions would be circulated before we carry out further reviews
  • Each partner should review (published) national population surveys
  • Partners undertake to contact experts in other European countries aiming at completion alcohol surveys review in all EU countries
  • Jacek to produce a standard letter to experts from non-SMART countries explaining the SMART project and asking for assistance (partners can personalise this)
  • Consider how we can best acknowledge the help provided by outside experts
  • The review of surveys should cover the survey methods and procedures, measures and results - using the questionnaire - which might be modified after further consideration.
  • At country level, we look at the wider (mainly European with some wider international) literature to provide context for the review of surveys.


Development of the questionnaire. General issues, survey administation
Over the course of the morning, a number of issues were raised concerning the development of the population survey instrument and the survey methodology. Issues raised:

  • What will be the outcome measures of pilot survey carried out in 9 participating countries?
  • Again the question about how far we could compare countries.
  • Should we aim to produce a set of 'core' questions and 'optional' questions which countries could include or not, depending on resources, interest etc.?
  • Should we aim to produce an instrument for insertion into a more general survey (as many existing national surveys are of this type) or for use as an alcohol specific survey?
  • Should we produce an instrument AND a methodology which describes best practice and provides recommendations?
  • Should the questionnaire be administered face-to-face or by self-completion?


After long discussion it was agreed that relevance or feasibility of different consumption measures in the pilot survey will be assessed against three major criteria:

  • comparing several annual consumption measures ("the higher the better")
  • comparing predictive qualities of different consumption measures "the higher association with problems the better"
  • comparing consumption distribution "the more similar between countries the better"

A decision was also taken that the output would be a common best methodology and a common instrument for an alcohol specific survey.

It was hoped that this would complement current policy directions by providing an appropriate methodology and instrument should member states be required to conduct alcohol specific national surveys.

The question of whether to have a core set of questions and other optional sets was not finally decided but combination of a core set and optional questions respecting cultural specificity seemed to be the most appropriate.

It was agreed that the questionnaire would be administered face to face for the pilot. If this form of administration will be advised for general population use is to be decided based on existing literature. Discussion highlighted the expense of this option but the advantage in terms of response and completeness of data.

It was repeatedly stressed that a general population survey should not be too long (30-40 minutes) and it was acknowledged that countries would differ (for cultural as well as other reasons) in the length of questionnaire/ type and number of questions on alcohol which would be seen as acceptable and/or likely to be used.

Development of the questionnaire. Four methods of measuring consumption
Four alcohol consumption measures were presented. Their advantages and disadvantages were considered and discussed.

  • FxQ method - last 30 days, than last 12 months
  • FxQ method separately for weekdays and weekend days - last 30 days, then ordinary FxQ last 12 months
  • Exact review of last week and then preceding last 3 drinking episodes
  • Frequencies of drinking occasions x quantity for last time of each occasion


It was suggested that graduated quantity/frequency measure could be considered as an extra option, even though it is difficult to respond
The difficulty to capture by one measure the irregular consumption patterns as well as the regular ones was noted.
Respondents in FxQ report a mode consumption intake rather than average which leads to underestimation.
Two important questions were raised:

  • What is the evidence for using one or other of the 4 methods?
  • Is it worth testing the methods? -What are the chances that countries will adopt/ implement a new set of questions as there are many reasons why they may prefer to use existing measures.


It was felt that the survey/literature reviews would help to address the issue of evidence and that which methods to test would be finally decided taking account of the reviews.
The number of methods which should be tested was also considered and
it was suggested that, taking account of country contexts, it might be more useful to test 2 options plus a country specific method (a method already in use in the country). This latter suggestion was adopted as optional.
Context questions might be identified by looking at available sets of questions e.g. GENASIS. This should be discussed at the next meeting.

Q/F will be used as one method but the precise format is still to be decided.

Decisions taken:

  • To test no more than 2-3 methods - 2 'global' and one national consumption measure (the latter one - optional).
  • The questionnaire will be structured.
  • A qualitative approach will be used to examine understanding of concepts and experience of completing the questionnaire.
  • The questionnaire should take no more than 40 minutes to administer


Measures of Binge drinking / Drunkenness / Problems
Presentations on binge drinking and drunkenness indicated the problems of defining and measuring these concepts and offered some alternatives for consideration.

Binge drinking: It was suggested that:

  • The term 'risky single occasion drinking' might be adopted by SMART.
  • We should adopt the measure of 60+ grams in a single occasion as a measure of 'binge' drinking (The decision should depend on evidence emerging from the literature review).
  • The time period needs to be added: 2 hours / 6 hours / or a continuous - 2,3,4,...etc. hours.
  • The qualitative work could ask about whether the cut off point (risk) was meaningful for people.
  • We might possibly include a body weight question.


Drunkenness:
It was agreed that although this is a subjective measure and open to wide individual and cultural variation, it would be worthwhile including a question(s) and possibly look at this measure in relation to the measure of binge drinking.

The problem of attribution was discussed especially when questions included attribution as part of the question. It was suggested that perhaps questions could be asked without mention of the relationship to drinking - but this was not decided.

It was agreed that qualitative questioning could be used to examine what respondents meant by 'drunk'. It was also suggested that we could ask, 'How many drinks it takes you to get drunk?' The measure still has to be decided.

Other health and social problems
A list was circulated showing measures used in a sample of European surveys for different domains of problems - dependency, chronic, acute, social.
It was felt that existing measures (e.g.DSM4) were available for chronic alcohol related conditions.
Harm to self and harm to others were suggested as two important dimensions for inclusion.
Both frequency and severity of harm should be considered.
Again attribution needs to be considered.
Acute problems needed to be included and the best indicators identified.
One suggestion was to choose indicators of harm in line with EU and/or WHO priority areas e.g. children, workplace, injuries etc.

Social problems - again there is a long list of possibilities and
Group 1 was asked to circulate a proposal for measures of acute harm/ harm to third party / social harm.
The questions in this section should take no longer than 7 minutes to administer. Time frames of 30 days and the last 12 months were suggested as suitable.
In qualitative pilot assessment, respondents could be asked open questions re: what they think are alcohol-related problems and what they think are harms to others resulting from alcohol.


Administration of the piloting survey
The questionnaire is intended to pilot and test different methods of acquiring data on alcohol consumption and drinking patterns. The intention is to find a common instrument to get the best data.
Some suggestions for sampling were circulated - 200 interviews per country; 30% 'problematic' drinkers/ 70% 'regular' drinkers; urban/ rural; 50% age 18-39/ 50% age 40-64; 100 women/ 100 men; abstainers to be excluded.
Discussion centred on the scientific rigour of attempting to do this with so many variables and on the practicalities of sampling and administering the pilot questionnaire.

Agreed that we need to:
1. Pilot the instrument to identify optimum design and measures
2. Examine qualitatively individual responses and experiences to completing the questionnaire - interviewer keeps notes and discusses understanding of concepts/ experience of completing the questionnaire etc. with a sub-sample of respondents
3. Discuss with a group of respondents - this raised questions regarding the viability of this exercise - Is it practical? What will we achieve by way of information? Would it be more useful to use a group of 'experts'? The decision was left up to individual teams/


A screening method or instrument was considered to identify appropriate respondents before completing the questionnaire.

Economic Impact of Alcohol Policies
A paper was pre circulated and Peter Anderson presented this work package. It was agreed that development of the cost/benefit analysis should take into account work completed or in progress (e.g. CHOICE/ AMPHORA project). Three possible models were presented and discussed:
1. Cost-of-illness studies
2. The CHOICE model
3. Avoidable burden of disease/illness

The third one was likely to be used for SMART but other options were also under consideration.

The importance of considering the impact of different policies on different social groups was discussed as well as looking at the unpredicted costs of different policy options. It was agreed that this should be taken into account.

It was suggested that some questions be added to measure public support for different alcohol policy options. Measures will need to be developed.

Short report from the EC Committee on Data Collection, Indicators and Definitions
Following indicators of alcohol-related harm were suggested:

  • Total consumption of alcohol - per capita or per population aged 15+, beverage specific sub-categories.
  • Harmful pattern of consumption - 60 grams or more on one occasion, monthly or more often in the past 12 months - same for men and women.
  • Alcohol-related harm - alcohol-attributable years of life lost based on WHO attributable years lost due to chronic disease and injury.


Website
Points raised:
Fully active members of partner teams could be named in a section on the country teams.
Partners to send Jacek additional appropriate links for the website.
Partners to send information (if available) on how to count 'hits' from different countries.

Other issues - organisational etc.
A timetable for completion of the work was discussed. It was agreed that collection of the data may take longer than expected and should start this Fall.
Partners need to complete time sheets. Reporting period is Sept.2008 - end August 2009.
External evaluation has been arranged.
Agreed that a final version of the survey reviews would be available by end May 2009 for discussion at KBS meeting.

Next meeting will be in Florence end Sept./ early Oct., most likely 5-6 October -
to be confirmed.



Back to content | Back to main menu