Epi in the News

Hi everybody....

The first article is about video games and mental health- the results are a little bit shocking to me.. I think its a cross sectional study. The researchers collection exposure details and outcome from gamers at the same time. The exposure is the number of hours playing video games, as documented from real time data. This is possible as most games are nowadays online and therefore can provide real data about time spent playing the game. They have reported a cut off of 4 hours- so maybe the comparison was between those who played 4 hours a day every day, vs those who played less than that. The outcome was measured using psychological questionnaires- specifically asking gamers how they felt and their well being. The study suggests that playing the particular game of animal crossing 4 hours a day every day made the gamers happier and they reported greater well being.

The second article is about the nocebo effect of statins- showing that side effects reported are generally due to the knowledge of taking a pill rather than the statins themselves. They have 60 participants. And it looks like the participants are their own controls. They are taking statin/ placebo/ no medication rotating every month in a random sequence. So its a randomized trial, and looks to me like a N-of-1 style- meaning one patient trials. The patients are their own controls and and this was the difference between this study and previously conducted studies addressing the same research question. Here the patients could see for themselves that the side effects were related to just taking a pill, and because they are expecting side effects and so mostly in their minds.

The second article about the nocebo effect is more informative and more details about the study is given. The results have also been stated using numbers that can easily be understood. I found the first article quite vague- it was not clear how much of a difference is the number of hours playing video games making. The comparison groups is not clear and as a reader I am not really impressed by the vagueness surrounding the results

Both articles are very interesting reads... What do you all think?

Fathima

For the first article about video games and mental health:
I assume this is a cross sectional study.
Exposure is ofcourse video games.
Outcome is great wellbeing (but they havent clarified "greater" in comparison to what controls ?)
So the finding is pretty shocking as it doubts the reports that say video gaming can harm mental health. Especially that they mentioned that if the research goes on there will be more findings of the toxicity in games with evidence! This finding is useful in bringing attention to an area with few data on but not sufficient to make any policy.

Regarding the second article, nocebo effect:
This is an intervention study.
No because the patients were not typical of all patients as they previously stopped the statin therapy before the study. The trial only gives good association of the nocebo effect with the symptoms but not a causation.

The reporting of the second study is more comprehensive clarifying the controls and the test and reporting the weaknesses were the sample might not be generalizable to the whole population. But the first study is pretty vague in those instances, especially that it doesnt show the comparison group.

Thanks Sujit.

It's very hard to draw much from the first study as there is no link to the original paper and we don't actually know the exposure or outcome definitions. The finding seems to be that using actual data on gaming rather than self-reported time spent gaming gives different findings and ones that are more positive for gaming.

The second study is an N of 1 double blind crossover study. The confidence intervals for the symptom score for patients taking placebo 12.1 - 18.7 vs 13.0 - 19.6 for those on the statin. This broad overlap means that there was no convincing evidence, in these patients, that statins caused more side effects than placebo.

Hi everyone i am glad to have joined this discussion, my sentiments below

In the first article,
1) Design: The study was cross sectional as both the exposure(hours spent gaming) and the outcome (psychological well-being) were assessed simultaneously.

2): Exposure measurement: Game play was assessed through an objective measure of tracked hours of game play. This is an advantage to the study as it brings objectivity and reduces the possibility of information bias in the form of reporter bias which may result in an over or under ascertainment of exposure (non differential miss-classification).

3): Outcome definition: It was not clear to me which psychological questionnaires were used. If these questionnaires and their administration was not standardised they possibly could have resulted in some form of information bias.
4) Finding: Game play was associated with greater well-being. This finding might be used to influence further epidemiological studies with better designs, accounting for other potential confounding factors, which then can help inform policy. Because there was no reporting of sample size in this study, i was unsure on the power to influence policy.

Second study
1) Study design: double-blind, three-group, n-of-1 trial. Participants in the trial acted as their own controls, being given first the statins/intervention and then the placebo
2) Causal relationship: The design might not demonstrate causal relationship. Firstly, the power of the study might be questioned, with just 60 participants, it is possible that these findings were due to chance, this is confirmed by the large confidence interval in the first analysis. The lack of a comparison group limits the ability to speak of causal associations between placebo exposure and reported side effects. It could be that the reported side effects during the placebo period could be due to the late effects of tacking the statins initially.
3) Quality of reporting: In the second article, the quality of reporting is much better and allows for tracking of the influence of chance and bias in the reported findings which lacks in the first article were. There is little description of the study methods, factors adjusted and sample size for before reaching the conclusions made in the first article and this is described to some extent in the second article

In the first article, the study described is a cross sectional study.

The exposure was time spent playing video games, specifically Nintendo’s Animal Crossing (AC) and EA’s Plants Vs Zombies (PVZ) They quantified exposure using gameplay data which gave an accurate record of actual time spent playing games.

The outcome was “wellbeing” as defined by “psychological questionnaires”.

As per the lead researcher this study “shows that if you play four hours a day of Animal Crossing, you’re a much happier human being”.

I don’t think the findings of this study can be used to inform policy at present. There are a number of sources of potential bias (I briefly skimmed some of the study which as yet is not peer-reviewed or published).
1. Selection bias; participants were invited to join the survey via email. Responders may be inherently different than non-responders and there was a low response rate. (re PVZ survey invitations were sent to 50,000 then 200,000 players in two waves and only 518 players finished the survey re AC sent to 342,835 players, 6011 players responded.)
2. Responder bias; Participants were aware that the aim of the study was to investigate how people play video games and how they feel over time. People who play 4 hours of video games daily may feel quite passionately about video games and so may modify their answers to show video games in a positive light.

The study design in the second article is a double-blind, three-group, n-of-1 trial. The researchers demonstrated that 90% of symptom burden experienced by a statin challenge was also elicited by placebo.

I found the quality of reporting poor in the first article. Firstly the headline is misleading. It gives the impression that playing any video game can benefit mental health and including “Oxford academics” in the headline seems tactical to give it more credibility. I think a more accurate headline (although unlikely to garner as many clicks) would be “Playing certain video games for particular lengths of time has been associated with scoring higher on psychological questionnaires which aim to assess your wellbeing”. Secondly, it would have been better if there had been quotes from alternative sources in addition to the lead researcher particularly as the study is not yet published or peer-reviewed. The article as a whole was quite vague and I found it difficult to pull specific information out of it.

The second article is far more transparent. It summarises the results clearly while making it clear that there is significant uncertainty as regards the generalisability of results.

Hello!
First, an article about video games and mental health.

1.it is a cross sectional study because it looks at number of hours spent playing games(actual play time data ) and the state of wellbeing(psychological questionnaires) at the sametime.

2. Comment on the exposure definition- the exposure definition only considers number of hours spent playing video games and this is narrow in the assessment of the study because this exposure may have it's shortfalls as not all people in the study may be actively playing even though they maybe logged on to the various platforms. In addition to that the exposure may be influenced by other factors such as schooling, chores etc.

3. Comment on the outcome definition- the outcome in itself is a subjective entity to measure as people's ideas of self may vary as well as a judgement of their own well being. the questionnaires may not be enough and participants may require a formal evaluation.

4. What was the finding? How can this finding be used to inform policy? It was found that playing at least 4 hours of video games was associated with and improved sense of well being. This finding can't be used to inform policy since there's no mention of other factors that could have been at play during the hours of playing video games. Example, contact with friends during play-time, a sense of identity and team work for multiplayer games, or even a sense of accomplishment with single player games. Secondly, the two video games used are not typical of the various classes of video games available and so the findings of the study may not be reliable .



Second, an article about the nocebo effect, with a quote from LSHTM's own Prof Liam Smeeth.

1. What is the study design? Intervention study(cohort??)

2. Are the researchers able to demonstrate a causal relationship? they are able to demonstrate a strong association between the nocebo effect and the development of side effects but no causality. This is because there was no information on certain factors that may have confounded the study. Such as the patient's current diagnoses and drug interaction with other medication the patient may be on already. The use of only patient's who had had side effects previously was also selection bias which reduced credibility of the study and the sample size was too small to rule out random error.


3. Comment on the quality of reporting for the first versus second article. First article's heading alluded to a possible association which was in line with the findings of the study while the second's heading spoke of a cause which had not yet been ascertained. Also, while the first article pandered to human interest aspects, there was no statistics to back the research. the second however gave raw data but this had no context as there was no comparison to the baseline sample