Literature review > Issue_2 > Review Behets et al. 

The question of whether or not to treat women presenting with the complaint of vaginal discharge for gonorrhoea (Ng) and chlamydial (Ct) infection (in settings where it not possible to diagnose these infections definitively) has been much debated. One school of thought maintains that, since many studies in different settings have shown that the prevalence of these infections is no higher in women with the symptom of discharge than in those without, it is not logical to give treatment for them to women presenting with discharge. The other school of thought believes that women presenting with discharge should be treated for Ng and Ct, both to prevent complications in those who have the infections, and to prevent onward transmission. In their view, failure to treat women who present to the health services with symptoms possibly related to these infections represents a dereliction of duty both to the patient, and to the public. It is true that many women will be treated unnecessarily, but the expense and, possibly, side effects induced are justified by the public health benefit.

Various compromises have been sought between these two extreme positions. It has been suggested that women complaining of discharge should be treated in the first instance for vaginal infections (bacterial vaginosis, T. vaginalis, and perhaps Candida), and asked to return if their symptoms have not resolved after one week; women who return should be treated for Ng and Ct. This proposal is not entirely logical, since the high proportion of women in whom Ng and Ct infection are asymptomatic will not be treated if this algorithm is used.

An alternative is to assess to what extent such women are at risk of Ng and Ct infection, by asking about their own sexual behaviour, and about the presence of symptoms in their male partners. Such risk assessment, which should be an integral part of good clinical care, has not met with universal approval, since several studies have shown that the "risk score" suggested by WHO, based on responses to 5 specific questions, is neither sensitive nor specific in identifying women with these infections. It is a little unfair that risk assessment should be thrown out on the basis of these findings. The WHO did not propose that the risk score, developed on the basis of one study in Kinshasa, Zaire, should be applied uncritically in other settings; the suggestion was, rather, that locally appropriate risk scores should be developed in different countries and settings.

Whether or not to treat women presenting with vaginal discharge for Ng and Ct depends ultimately on the prevalence of these infections in the clinic population. Risk assessment is an attempt to identify a sub-group of women with a higher prevalence of Ng and/or Ct infection. The conclusion of the authors of this paper, who studied a population with prevalences of Ng and Ct of 13% and 11%, respectively, is that all women presenting with discharge should be treated for both these infections. The conclusions of other experts, working with low prevalence populations, will be the opposite. WHO is in the difficult position of proposing guidelines that will satisfy both groups. Their current recommendation, still under consideration, is that all women should be treated if the prevalence of Ng and/or Ct exceeds 15%, and none should be treated if the prevalence is less than 5%. Where the prevalence is between 5 and 15% - in practice this will include most clinic populations - risk assessment should be used to decide who should be treated.

about SDI | newsletters | grants | publications | literature reviews

WHO Home - WHO Search - TDR Home - SDI Home - SDI Contact us
(c) WHO/OMS 2001