|
|
Abstract
The fate of African based HIV/AIDS clinical trials presented at conferences - do they get published in full?
N. Siegfried1, M. Clarke2
Background: Publication bias can cause over- or under-estimates of the true effect size when pooling data in meta-analyses in systematic reviews. This bias can occur when trials reported only in the so-called "grey literature" (e.g. conference abstracts) are excluded. A recent Cochrane methodology review found that about a third of trials presented first at conferences are never published in full. The review did not find previous research investigating HIV/AIDS or trials done in Africa. Our study is the first to assess the full publication of randomized trials of HIV/AIDS interventions done in Africa and presented at conferences.
Methods: We identified abstracts for randomized trials of HIV/AIDS interventions conducted in Africa and presented at national and international conferences before 2004 by searching the specialised conference electronic databases, AIDSearch and the AIDS Meeting Abstracts component of NLM Gateway. We searched for full articles reporting the same trials by using a specialised register of African HIV/AIDS trials and MEDLINE.
Results: A total of 64 conference abstracts, reporting 49 randomized trials, were found. Full reports were identified for 53 (82%) of the abstracts. Most (48/64) abstracts were presented at the International AIDS Conference. The mean time from conference presentation to full publication of results was 1.9 years, ranging from less than 1 to 7 years. As of October 2005, results of trials reported in 11 abstracts (from conferences held between 1990-2002) do not appear to have been published.
Conclusions: Our study is the first to assess the progression of African HIV/AIDS trials from conference presentation to full article. It shows that the proportion published in full is much higher than that in other areas of health care. However, the total number of African HIV/AIDS trials is indicative of a paucity of high quality intervention research, despite the scale of the healthcare crisis on the continent.
AIDS 2006 - XVI International AIDS Conference
Abstract no.
CDB0384
Suggested Citation
"N.Siegfried, et al.
The fate of African based HIV/AIDS clinical trials presented at conferences - do they get published in full?.
:
AIDS 2006 - XVI International AIDS Conference:
Abstract no.
CDB0384"
|
|
|