handsearching of 2 medical journals of Bahrain for reports of randomized controlled trials

ABSTRACT

To identify reports of randomized trials by handsearching 2 Bahrain medical journals, which are indexed in the biomedical database EMBASE and to determine any added value of the handsearching by comparing the reports found by handsearching with what would have been found by searching EMBASE to examine [i] the precision and sensitivity of the EMBASE index term Randomized Controlled Trial [RCT] and [ii] The Cochrane Collaboration's systematic electronic search of EMBASE [which uses 4 index terms and 9 free-text terms]. All issues of the Bahrain Medical Bulletin [BMB] [1979-2004] and the Journal of the Bahrain Medical Society [JBMS] [1989-2004] were handsearched in February 2005 for reports of RCTs or Controlled Clinical Trials [CCTs] according to Cochrane eligibility criteria. Out of 395 articles in BMB we found reports of 12 RCTs and 4 CCTs. Distribution by country of corresponding author Jordan [4 RCTs, one CCT], Bahrain [one RCT, one CCT], India [3 RCTs, one CCT], Kuwait [one CCT], Saudi Arabia [2 RCTs], USA/Bahrain [one RCT], and Oman [one RCT]; and by specialty Anesthesia [8], Surgery [1] Pediatrics [1], Radiotherapy [1], Community Medicine [1], Sports Medicine [1], Obstetrics/Gynecology [3]. The Journal of the Bahrain Medical Society included reports of 14 RCTs and 3 CCTs, out of 97 articles. Distribution by country of corresponding author Jordan [9 RCTs, 2 CCTs], Bahrain [3 RCTs], Egypt [one RCT], Kuwait [one RCT], and Saudi Arabia [one RCT]; and by specialty Anesthesia [7], General Surgery [3], Obstetrics/Gynecology [1], Radiotherapy [1], Pediatrics [1], Orthopaedic Surgery [1], Education [1] Ear Nose and Throat [1] Ophthalmology [1]. Overall, of the 33 reports of trials found by handsearching I both journals, only 23 were included in EMBASE of 1 which only 6 had been indexed with the term RCT. Off the 23 reports of trials included in EMBASE, 16 had been identified in the Collaboration's systematic search of EMBASE. Two reports of trials could have been retrieved by this search but there was insufficient information in the title and abstract to code these as trials. The EMBASE records for the remaining 5 reports of trials did not contain terms currently used by The Cochrane Collaboration to identify reports of randomized trials in this database. The handsearching of these journals will help minimize publication bias by locating randomized trials not previously identified and, through their inclusion in the Cochrane Central Register of Controlled Trials [CENTRAL] in The Cochrane Library, will ensure reports of randomized trials will not remain 'buried' through indexing bias