Comparison of Using a PubMed Comprehensive Search Approach vs. Clinical Queries

To compare the searching strategies, let's conduct a PubMed comprehensive search (See Figure 1) to answer Eric's question. This is the traditional or typical approach that individuals use if they are not familiar with the Clinical Queries feature. Beginning on the PubMed homepage, chewing tobacco cessation is typed into the search box resulting in the retrieval of 1227 citations (Figure 23). A disadvantage to conducting this type of search is that you have no idea as to what type of studies you have found, e.g., systematic reviews, RCTs, animal or laboratory studies, personal opinions, etc.

Figure 23. Traditional PubMed Search for Chewing Tobacco Cessation.

Image of Traditional PubMed Search for Chewing Tobacco Cessation.

In this case, someone not familiar with the PubMed Limits feature would either spend considerable time reviewing the 1227 titles and abstracts to determine which ones might be useful in answering the question, or more realistically, just look at the first 10 to 20 and hope that one answers the question. Fortunately, for those who are familiar with PubMed features, this process becomes easy to manage. As previously illustrated, to find the highest levels of evidence using the filters, under Article types, click on the higher levels of evidence to put them on the Search results page. You can also click on Human and English. Start with the highest level of evidence, Practice Guidelines. This results in 3 Practice Guidelines, however of the 3, only 1 addresses smokeless tobacco (vs. smoking cessation) and it was published in 2004. Subsequently, re-run the search by clicking on “Clear all” to remove Practice Guideline as a filter and then click on Meta-Analysis. This results in 24, a reasonable starting point to determine if relevant information has been found. Of the 24, however, the majority address smoking cessation and those that do address our topic, use the term ‘smokeless tobacco’. By adding “NOT smoking cessation” the 24 results are then limited to 5, with the Cochrane 2015 review listed, as it was when conducting the Clinical Queries search (Figure 19).

Figure 24. Results of Limiting the 1227 Citations for Chewing Tobacco to 24 Meta-Analysis.

Image of Results of Limiting the 1227 Citations for Chewing Tobacco to 24 Meta-Analysis.

Again, if this was the approach and the user notices that the term being used is ‘smokeless tobacco’ rather than chewing tobacco, a new search query is run using ‘smokeless tobacco cessation NOT smoking cessation’ resulting in 1,103 citations. Limiting these using Meta-Analysis identifies 8, a much more reasonable number to review. Of the 8 citations, one includes the most recent Cochrane publication in 2015, and two previous ones which the 2015 updates (Figure 25).

Figure 25. Results of Limiting 1,103 Results by Selecting Meta-Analyses and Adding “NOT smoking cessation.”

Image of Results of Limiting 1,103 Results by Selecting Meta-Analyses and Adding “NOT smoking cessation”.