Using a search filter to improve literature searching efficiency for clinical question updating

Article type
Authors
Amos Q1, Chan W2, Tom G1
1Kaiser Permanente, Care Management Institute, USA
2Kaiser Permanente, NW Permanente, USA
Abstract
Background: Guideline updates require comprehensive search strategies to identify all relevant studies. The number of studies identified initially can be quite large and requires an efficient process to quickly identify the most relevant studies.

Objectives: Describe the methods used by a US health care delivery organization to create comprehensive search strategies and filter results to improve efficiencies in literature searching.

Methods: A comprehensive search strategy was constructed for a systematic review addressing a clinical question within a cardiovascular disease guideline. Identified abstracts were evaluated using a filter consisting of the top general medicine and disease specific journals. If filtered abstracts met inclusion/exclusion criteria, the systematic review was evaluated for possible updating using key questions (e.g. do the newly identified studies change the recommendation, confidence interval, or point estimate?). If yes, the systematic review was updated necessitating review of all identified abstracts. If no, the systematic review was not updated resulting in no change to the clinical question.

Results: Search strategy results identified 396 abstracts; 98 were identified in key journals and did not meet inclusion/exclusion criteria thereby warranting no update. Reviewing 98 abstracts compared to 396 reduced review time by 75%.

Conclusions: Using a filter to review relevant abstracts saves time by reviewing only those abstracts identified in top journals allowing the updating process to proceed more rapidly.