While the review by Bobrovitz et al. [
4] was generally well conducted, there are several limitations worth mentioning. A systematic review is shaped by its inclusion criteria. By design, the authors may have excluded several studies that may have contained information relevant to the prevention of emergency hospitalizations using medication. While they searched several databases, including some that cover parts of the grey literature, there are several other databases that could have been searched which may have provided useful information, for example, EMBASE, CINAHL, and International Pharmaceutical Abstracts. Although these databases often overlap, several studies have shown that the overlap is far from complete [
6,
7]. Similarly, the review only included studies in English and, while there is conflicting evidence regarding language bias [
8‐
10], it remains possible that it missed certain medication interventions which affect emergency hospital admissions. For example, studies related to traditional Chinese medicine are more likely to be indexed in Chinese language databases [
10]. Searching the grey literature is an important component of comprehensive systematic reviews [
11]; however, Bobrovitz et al. [
4] examined sources that were limited to the UK. Expanding the search to international organizations or prominent organizations in other countries may have captured additional studies. Similarly, while the authors sought the input of experts, these were only approached at three conferences in the UK, limiting their reach. Further, for inclusion, studies needed to have searched more than one database and to have assessed and reported on the quality of the included studies. While these are criteria of comprehensive reviews, by excluding studies that failed to assess and report on quality or only searched a single database, some potential interventions may have been excluded. Finally, in addition to the limitations noted above, the authors note in their discussion that many medications which impact hospital admissions were excluded because the results were reported as a composite outcome.
For the prioritization of medications, Bobrovitz et al. [
4] first examined NICE guidelines and then expanded to relevant guidelines from Europe and America. To be considered ‘guideline based’, the medication had to be recommended by NICE and at least one other guideline. However, the list of guidelines they identified was not comprehensive. For example, infliximab was identified as having moderate evidence for the treatment of Crohn’s or ulcerative colitis, but was not flagged as guideline based in this article despite recommendations for use from NICE [
12] and the American Gastroenterological Association [
13].