OBJECTIVES : This review aimed to investigate the effectiveness and value of citation pearl mining for the identification of relevant studies.

METHODS : We searched for randomized controlled trials (RCTs) that evaluated an exercise program alone, or as an adjunct to a smoking cessation program, for achieving long-term smoking cessation. A conventional systematic search was undertaken in Embase^® and Medline from database inception to May-2020. The pearl growing method included identification of primary pearl, backward and forward citation mining through searching for its bibliography and ‘cited by’ section in PubMed, respectively. The ‘similar articles’ section in PubMed was also explored.

RESULTS : The systematic searching of Emtree and MeSH terms resulted in the identification of 4,029 citations to screen. In comparison, 332 citations were identified from the first pearl (278: similar articles; 8: cited by; 46: bibliography). Overall, 22 relevant RCTs were identified through systematic database searching, while 27 trials were uniquely identified by the pearl growing method. These additional RCTs were either not listed in the biomedical database (n=3) or were inadequately indexed (n=2). Two trials were judged to be at high risk, one at low risk, and two at unclear risk of bias indicating an overall low methodological quality. All additional RCTs quantitatively reported improvements in abstinence rates with smoking cessation and physical activity measures underlying its value; however, it did not add new knowledge to the synthesis. Methodologically, the time taken to screen 4,029 citations was significantly higher in the traditional systematic approach (24 days work/screening at a rate of 400 citations/day) compared with the pearl growing method (7 days work).

CONCLUSIONS : This case study found the pearl growing method to be more effective in numerical terms by maximizing the location of evidence and reduced workload. It should be considered an essential addition to the arsenal of systematic searching rather than a substitute approach.