DOI: 10.1097/01.aids.0000199015.73762.97

,  

Issn Print: 0269-9370

Publication Date: 2006/01/02


Print

Problems in medical injection study

Matthew Hogan

    loading  Checking for direct PDF access through Ovid

Excerpt

Priddy et al. [1] have performed a great service by attempting an all too rare serious investigation of the medical injection risk of transmitting HIV. The authors did a further service by noting that ‘heterosexual’ (the conventional and dangerously imprecise term for ‘penile–vaginal intercourse’) [2] transmission of HIV in Africa is merely the ‘assumeD' vector. Better still, their expressed concern about prioritizing needle use hygiene is undeniably valuable.
Nevertheless, the authors are grossly premature in declaring that their study ‘suggests that medical injection practices in this region do not contribute significantly to HIV transmission’. At best, the study was only suggestive that more study needs to be done. Their needle sample was concededly low (212) and their test's sensitivity for needle HIV-RNA was also concededly low (> 40 000 copies/ml). Another problem was the probable suppression of the true state of injection hygiene because the behavior of the health institution workers was occurring under observation; World Health Organization (WHO) Injection Safety Technical Tools require onsite visits and other procedures not likely wholly to eliminate distortion and the Hawthorne effect [3].
Other major additional problems can be seen in sampling. The sample of 72 patients to deduce the source rate of HIV (with a wide 95% confidence interval of 3.9–17%) in the Ethiopian health institutions is stated as being taken only from patients in ‘health centres’, which seems to exclude patients in the ‘health stations’. Yet it is the latter institutions where the overwhelming incidence of observed needle reuse took place. The prevalence of HIV in the sources of the needles tested for HIV RNA was thus undetermined. In addition, the reusable EPI syringes and needles were not tested for HIV RNA at all, although they accounted for 30% of the total observed injections, and over two-thirds of those were found to have been not properly sterilized according to WHO standards. In fact, no syringes (as opposed to needles) were reported as having been tested at all.
Furthermore, no confidence interval was provided for the 0 result of the low-sensitivity test for HIV RNA.
In any event, it is patently unwise to argue (and to incorporate into the abstract) that the study's findings ‘suggest that medical injection practices in this region do not contribute significantly to HIV transmission’. First, it is best to slam the brakes on a readiness to dismiss a universally proved vector in deference to a mere assumption (however prevailing) of a geographically unique, histologically improbable [4], and imprecisely described, sexual vector.
Second, it runs foul of common sense and the encouragement of responsible healthcare practice even timidly to downplay the risk of a bloodborne virus's spread via unsafe medical injections as a result of a significantly limited study in the course of which it was also found 1) that ‘50% (of the medical facilities) had open wastebaskets for sharps disposal that were dirty or overflowing’, 2) that in ‘75% (of the facilities exposed, used needles were found on the grounD', 3) that ‘12% of therapeutic injections were given with previously used disposable needles’, 4) that ‘70% of the vaccinations administered using EPI needles/syringes … had not been properly sterilizeD' [because ‘81% (of the facilities) did not operate the sterilizers according to WHO guidelines’], and 5) that, in looking forward there remains ominously ‘a continued high rates of medical injections in Ethiopia’.
    loading  Loading Related Articles