NOTTINGHAM: African HIV trials fudged registration dates

So what else did they play fast and loose with?

Ripe Tomatoes, September 13, 2013


Circumcision and HIV

Two out of three key trials registered late

Does circumcision of adult men reduce HIV transmission? Initial scepticism, surely the foreskin reduces friction between penis and vagina, and prevents abrasions and blood transfer, and conflicting observational data, were eventually overcome by three randomised trials, one each from South Africa (click here), Kenya (click here) and Uganda (click here). In each trial men were randomly allocated to immediate circumcision or to wait for two years. All three trials showed about half the number of new HIV infections in the immediate circumcision group. The relevant Cochrane review (click here) summarises them as follows:

“The resultant incidence risk ratio (IRR) was 0.50 at 12 months with a 95% confidence interval (CI) of 0.34 to 0.72; and 0.46 at 21 or 24 months (95% CI: 0.34 to 0.62). These IRRs can be interpreted as a relative risk reduction of acquiring HIV of 50% at 12 months and 54% at 21 or 24 months following circumcision.”

The Cochrane reviewers judged “the potential for significant biases affecting the trial results [as] low to moderate”. Referring to selective reporting they wrote “All three trials clearly stated in their protocols that the primary outcome was HIV incidence. The risk of bias due to incomplete outcome reporting is therefore low in all three trials.”

The trial protocols have not been published so we have to trust that the Cochrane reviewers interpreted them correctly. But if they relied on trial registration documents (click here, for South Africa, here for Kenya and here for Uganda) they would have been misled. They they all indicate more of less the same primary outcomes and planned sample sizes as in the respective papers. But according to the date of first entry, only the Kenyan trial was prospectively registered. The South African trial was registered a year and a half after recruitment ended and seven days before the results were published! The Ugandan one a month after recruitment ended,and a month before publication.

	Recruitment period	Date of 1st publication of results	1st trial registration date	Registry planned sample size	Primary outcome in registry	Primary outcome in paper
South Africa	July 2002 – Feb 2004	26 July 2005	July 19 2005	3274 planned. 3274 in paper	HIV infection at 3, 12 and 21 months	All HIV infections at 3, 12 and 21 months
Kenya	Feb 2002 – Dec 12 2006	24 Feb 2007	April 23 2003	Initially 3,000. Altered to 2887 in registry. 2784 in paper	HIV incidence at 2 years Complications of circumcision	HIV incidence after three interim analyses
Uganda	August 2002 – Dec 2006	24 Feb 2007	Jan 23 2007	5,000 planned 4996 in paper.	HIV acquisition no time point specified	HIV incidence

It looks like PLOS One and The Lancet wanted to publish sexy high impact trials, found they weren’t registered, so got the authors to do it retrospectively, and hoped no-one would notice! All three trials were funded by the public sector.

Jim Thornton (Professor of Obstetrics and Gynaecology at Nottingham University)