Re: Effect of breakfast on weight and energy intake: systematic review and meta-analysis of randomised controlled trials
Is the search strategy in the article "Effect of breakfast on weight and energy intake: systematic review and meta-analysis of randomised controlled trials" comprehensive? The authors should be recognized for including the complete search strategy for one of the databases they searched as supplemental material, in compliance with the PRISMA reporting guidelines. However, the search strategy may have some weaknesses, such as the use of subheadings in conjunction with subject headings, the use of the title index (instead of the textword or keyword indexes), and the use of phrases (instead of and or adjacency statements). The result of these choices is that some articles which may meet inclusion criteria are not retrieved by the reported search and may not have been screened. For example, Nas 2017 appears to be relevant, but does not appear in the search results of the reported search. Why not? I will explain with reference to lines 1-9 of the reported search strategy (the cluster of lines retrieving papers that address the concept of skipping breakfast).
Line 1: 1 exp Breakfast/ph -- while Nas 2017 is indexed with the MeSH term breakfast and the subject heading+subheading feeding behavior/physiology, line 1 requires that the physiology subheading be applied to the breakfast subject heading in particular. (Incidentally, the MEDLINE indexing date for Nas 2017 is 2017/08/02; it had been indexed at the time of the reported search.)
Line 2: morning meal.ti. -- while Nas 2017 does include the work "breakfast" in the title, it doesn't have the phrase "morning meal" in the title.
Line 3: meal timing.tw. -- while Nas 2017 does include the phrase "timing of meal skipping" in the abstract, it does not include the exact phrase "meal timing."
Line 4: Omitting breakfast.mp. -- while Nas 2017 does include the phrase "breakfast skipping," it does not include the exact phrase "omitting breakfast."
Line 5: skipping meals.mp. -- while Nas 2017 does include the phrases "breakfast skipping" and "meal skipping," it does not include the exact phrase "skipping meals."
Lines 6-9: eating habits.mp. or eating habits.tw. or eating habits.ti. or eating patterns.ti. -- Nas 2017 doesn't include the phrases "eating habits" or "eating patterns."
As a result, Nas 2017 is not among the results of the summary statement in line 10, and is not among the results of the overall summary statement in line 23.
I should note that the trial reported in Nas 2017 was indeed registered with clinicaltrials.gov as NCT02635139 and so could theoretically have been discovered by the authors when they searched trials registries for the word "breakfast"; that word is included in the clinicaltrials.gov study description and arms and interventions section of NCT02635139. Depending on which index the authors used when searching clinicaltrials.gov, NCT02635139 may or may not have been retrieved. This uncertainty underscores the importance of fully reporting even a simple, one-term query.
I mention Nas 2017 as an example of a larger class: papers that are potentially relevant and which should have been screened, but are not retrieved by the reported search strategy. This individual paper may or may not meet the authors' inclusion criteria; the larger point is that the reported MEDLINE search strategy excluded it from consideration.
Moher, D., Liberati, A., Tetzlaff, J., & Altman, D. G. (2009). Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. Annals of internal medicine, 151(4), 264-269.
Nas A, Mirza N, Hägele F, Kahlhöfer J, Keller J, Rising R, Kufer TA, Bosy-Westphal A. Impact of breakfast skipping compared with dinner skipping on regulation of energy balance and metabolic risk, 2. The American journal of clinical nutrition. 2017 May 10;105(6):1351-61.
Competing interests: No competing interests