A C Wagenaar et al (2009) Effects of beverage alcohol price and tax levels on drinking: a meta-analysis of 1003 estimates from 112 studies, Addiction, 104, 179–190 PDF
The full transcripts can be found on our archive page, however I’ve tried to summarise some of the key debates below.
1 – Were the aims of this study clear?
Although we all understood what that authors were hoping to achieve in this study, it was noted that the aims were not explicitely stated by the authors. An important observation was made by @DotBirch:
|@PHTwitJC as often seen with sys revs – the aim is in the abstract but not explicit in the paper (although I am fast reading!) #PHTwitJC|
@rebeccahams noted that the target population was not clearly defined. Also, some commented that the outcomes were complex and not clearly defined (e.g. the authors did not explain how they were measuring drinking behaviour). For example @rebeccahams tweeted:
|#PHTwitJC my concern was about the difference in alcohol intake/attitudes in different countries and how that might affect pricing polices|
2 – Was the systematic review comprehensive?
It was noted that the authors had not followed PRISMA guidelines for reporting methods of systematic reviews. For example @dotbirch tweeted:
|@PHTwitJC missing: flowchart and summary table? would have made reading much easier #PHTwitJC|
|@dotbirch @PHTwitJC #PHTwitJC And would have reduced the need to miss out detail in a lengthy narrative. (If word limits are an issue).|
Overall we agreed that the review was comprehensive from what was written, but noted that the presentation of such information could’ve been improved through abiding to the PRISMA guidelines.
3 – Were any adjustments made for study size or quality?
This wasn’t very clear. There was no information on data extraction nor quality assessment, and although some information was provided it was laid out in the narrative text rather than in tables or graphs.
|@PHTwitJC #PHTwitJC YES!!! I found it very difficult to read being embedded in the text. Tables, charts, graphs, please!|
|@Fibigibi13 I would echo this feedback – find formulae hard to judge, but charts / graphs much easier to understand and interpret #PHTwitJC|
As @DotBirch noted, adjustment for study size was done:
|@PHTwitJC they have checked homogenieity and used both i and Q tests would take into account study size #statsnightmare #PHTwitJC|
4 – Do you believe the results? Could anything else explain these findings?
The results of the paper show a strong statistically significant association, and can be summarised as follows:
|@PHTwitJC however the results are #lush #PHTwitJC|
|@PHTwitJC I’m not totally confident in that given the elisions etc and lack of clarity in data. But I’d like to think it’s right! #PHTwitJC|
Some further information would be useful to be able to more fully interpret these results. Suggestions included controlling for potential confounders such as: general living costs, GDP, sub-group analysis… and looking at potential bias, e.g. definition and measures of ‘drinking behaviour’.
5 – What implications do the findings have for public health practice & policy?
The results of the paper suggest a strong association between alcohol pricing and drinking behaviour, which adds to the evidence that a population-level intervention for minimising alcohol misuse. @DotBirch suggested:
|@PHTwitJC i hope cameron reads it and does a U U turn on minimum pricing! #PHTwitJC|
…to which @matandbec agreed:
|@PHTwitJC I’ll raise a glass (containing no more than 3-4 units) to that #PHTwitJC|
Following the lively discussion and debate we reflected upon the importance of communicating key messages from research, as well as being transparant in the methods used to derive at those messages. As @PHTwitJC tweeted:
|Interesting to reflect not just on what the results show and what that means, but how key #PublicHealth messages are communicated #PHTwitJC|