In their paper 'Burden of HIV among female sex workers in low-income and middle-income countries: a systematic review and meta-analysis', the authors list a number of limitations. But the study may be more notable for a limitation they do not allude to in any way: there is no mention of the risks sex workers face from unsafe healthcare, and perhaps even cosmetic services.
The main risks examined are sexual risks, though there is brief reference to the fact that intravenous drug use may drive or exacerbate HIV transmission. This is surprising considering that sex workers in some of the countries studied must face considerable non-sexual risks, perhaps not only from intravenous drug use. Yet health facilities, especially in low income countries, are not safe places.
It's quite difficult to assess the possible contribution of health facilities to HIV epidemics directly because it is a little-studied field. But there is plenty of indirect evidence, such as a recent study which found that adverse events are common in some 'developing and transitional economies' and quite a few result in the death of the patient. There are also the somewhat cryptic messages found in a UN publication which allude to non-sexual HIV transmission, but only seem to be directed at people who don't generally have to rely on 'non-UN approved' health facilities. Here's a good example:
"The most efficient means of HIV transmission is the introduction of HIV-infected blood into the bloodstream, particularly through transfusion of infected blood. Most blood-to-blood transmission now occurs as a result of the use of contaminated injection equipment during injecting drug use. Use of improperly sterilized syringes and other medical equipment in health-care settings can also result in HIV transmission. We in the UN system are unlikely to become infected this way since the UN-system medical services take all the necessary precautions and use only new or sterilized equipment. Extra precautions should be taken, however, when on travel away from UN approved medical facilities, as the UN cannot ensure the safety of blood supplies or injection equipment obtained elsewhere. It is always a good idea to avoid direct exposure to another person’s blood — to avoid not only HIV but also hepatitis and other bloodborne infections."
There is even fairly direct evidence from publications such as Service Provision Assessments and similar documents, some of which are researched and published regularly and are publicly accessible. These show that equipment to ensure safety in health facilities is often in short supply, or completely unavailable, that staff, who are also in short supply, often lack training and/or awareness of infection control procedures and that protocols for infection control are often unavailable or unused.
But sex workers and their clients, or those assumed to be sex workers and clients of sex workers, may have additional exposure to potentially unsafe healthcare. Since before HIV was identified, these groups were the targets of STI treatment programs. There is additional indirect evidence of non-sexual exposure where HIV positive infants have been identified whose mothers are not infected, women without any sexual risks have been infected although their partners are uninfected and population based surveys, such as the Demographic and Health Surveys, have shown that there quite significant numbers of people who are HIV positive despite having no identified sexual risks.
It's bad enough that HIV may be transmitted in health facilities and through other non-sexual routes; but the biggest worry is that people are not being warned of these possibilities. They don't know what to watch out for in health or cosmetic facilties, whether they are being vaccinated, treated, visiting an antenatal clinic, having an operation, going to a dentist, getting a tattoo or pedicure or even being circumcised on the grounds that it may give some protection to men against HIV infection.
Indeed, when people in African countries are tested for HIV and found to be positive, it is generally assumed that they were infected sexually, despite all the other ways they could have been infected. If, upon being asked about their sexual exposures, they report none, they tend not to be believed. The report on adverse events in hospitals mentioned above notes that the hospitals involved in the study were likely to be among the best available in those countries and that, for a variety of reasons including poor record keeping, the reported rate of 8% "probably represents an underestimate of the true rate".
It is quite extraordinary that sex workers in Africa have been so much studied, yet their non-sexual risks for HIV have been ignored, despite all the evidence that HIV is not always sexually transmitted. HIV rates among sex workers in some countries are exceptionally high. But, as the report on sex work above finds, in other countries rates are very low. Often where rates are high, they can be accounted for by intravenous drug use. So why should HIV transmission be so high among sex workers in some African countries, and why should we believe it is mostly sexually transmitted? How can health facility transmitted infections be ruled out until this issue has been properly investigated?