What if EPOC reviews only included studies from MEDLINE?

ID: 

290

Session: 

Poster session 2

Date: 

Monday 17 September 2018 - 12:30 to 14:00

All authors in correct order:

Johansen M1, Paulsen E1, Sporstøl Fønhus M1, Miller P2
1 Cochrane Effective Practice and Organisation of Care (EPOC), Norway
2 Cochrane Effective Practice and Organisation of Care (EPOC), UK
Presenting author and contact person

Presenting author:

Marit Johansen

Contact person:

Abstract text
Background:
Cochrane Effective Practice and Organisation of Care (EPOC) prepares systematic reviews about non-clinical interventions targeting health personnel, health services and health systems. According to the MECIR standards for the conduct of Cochrane Reviews, "searches for studies should be as extensive as possible in order to reduce the risk of publication bias and to identify as much relevant evidence as possible". As a minimum, all Cochrane Review authors are required to search the bibliographic databases CENTRAL, MEDLINE and Embase (if available). It is also highly desirable to search additional resources, such as databases relevant to the review topic and grey literature. One way to try to minimize the often time-consuming process related to searching for studies to include in systematic reviews is to investigate the contribution from these resources.

Objectives:
To assess the impact on EPOC reviews of studies not indexed in MEDLINE.

Methods:
We looked at 104 EPOC reviews published in the Cochrane Library (Issue 9, 2016) and grouped them according to the four EPOC topics: delivery, implementation, governance and finance. All reviews had searched more widely than MEDLINE to find studies to include. We used MEDLINE Ovid to find out whether the included studies were indexed in MEDLINE or not. We only looked at reviews that included one or more non-MEDLINE studies that had been critically appraised using GRADE and included in a 'Summary of findings' (SoF) table. For each SoF table, we identified all outcomes that completely or partly based their findings on non-MEDLINE studies. In order to find any impact on reviews from non-MEDLINE studies, we removed them from the SoF tables and re-analyzed the results. We further investigated each of the non-MEDLINE studies that had an impact on one or more SoF outcome, regarding type of publication, study design and source of indexing.

Results:
One hundred and four EPOC reviews included 1871 primary studies of which 119 were not indexed in MEDLINE. Sixty-one of these were included in SoF tables across 30 reviews. Preliminary results indicate that non-MEDLINE studies have a limited impact on EPOC reviews. We will present the final results.

Conclusions:
To be reported.

Patient or healthcare consumer involvement:
None.

Relevance to patients and consumers: 

As our research is on searching for studies to include in systematic reviews about non-clinical interventions targeting health personnel, health services and health systems, it is less relevant for patients and consumers.