Estimates of the relative importance of unsafe medical injections in the AIDS pandemic vary across orders of magnitude. This is because the probability an individual unsafe medical injection will transmit HIV is not known, and estimates supported in the peer reviewed literature range from 0.1% to 6.9%. These estimates are drawn from four types of empirical evidence: (1) rates of HIV infection from needle-stick injuries (any accidental scratch or jab commonly injuring a health worker while administering an injection to an HIV infected patient) [63
]; (2) HIV incidence among IDU who share needles [63
]; (3) retrospective analysis of large iatrogenic HIV outbreaks [66
]; and (4) laboratory examinations of used syringes collected in the field [67
]. Although interpretation of the available evidence is divided, these four types of estimates of the probability a medical injection will transmit HIV all include the range from 1.9–2.3%. The WHO models the probability of transmitting HIV as 1.2% [69
HIV prevalence is stabilizing in much of sub-Saharan Africa, but the AIDS burden on health care is still increasing as more patients progress to advanced HIV disease, unfortunately outpacing the availability of antiretroviral drugs. Updating the WHO's model of the global burden of disease from unsafe injections (describing the epidemic in 2000), to account for the elevated clinical prevalence of HIV, an estimated 12–17% of new HIV infections in 2007 could be attributed to unsafe medical injections alone [70
]. Hospital acquired infections from other invasive procedures have not been estimated, but assisted delivery has been linked to excess HIV infections across Africa and visible blood has been observed on arterial forceps, sutures and other equipment that contacts patients in maternity and pediatric wards [71
Hundreds of recorded cases of HIV positive children with HIV negative mothers indicate that the harm to children has been substantial [72
]. Today most African countries use only auto-disable (self-destructing, non-reusable) syringes for immunizations, but other risks to children that persist include invasive procedures, dental care, and non-immunization injections. In South Africa auto-disable syringes are not required for immunizations, and the HIV prevalence in children is too high to be explained by mother-to-child transmission alone [73
]. Moreover the incidence of HIV in children no longer breastfeeding and already immunized (ages 2–14) is 0.5% per year in South Africa [74
The WHO's model of injection risks in the year 2000 estimates that African adults receive on average 2.1 injections per year, and that almost one in five injections are unsafe [69
]. More recent data on unsafe injection frequency, available from 12 of the 14 countries in Table , demonstrate significant improvement. The probability that an adult will receive an unsafe medical injection in a year varies from 0.1% to 22% (lowest and highest in Lesotho and Rwanda), but the median is only 4.4% [8
]. In these countries unsafe injection risks are generally greater for men, for the poor, and in rural areas [8
Unsafe injection frequency and sterilization equipment coverage in sub-Saharan Africa 2002–2007
More than a third of the population of sub-Saharan Africa (living in Nigeria, Uganda, Malawi, the Democratic Republic of Congo, and Burkina Faso) should be at much lesser risk of unsafe injections, as they are protected by national bans on the use of disposable syringes that can be unsafely reused. Other injection safety interventions have been funded under the President's Emergency Plan for AIDS Relief (PEPFAR) through Making Medical Injections Safer projects. These interventions and those of the WHO's Safe Injection Global Network (SIGN) have reduced the frequency of unnecessary injections, reduced the risks posed by improper disposal of sharp medical waste, and produced and disseminated infection control guidelines to improve clinical practice. However these programs ignore larger problems with infection control capacity in African health care settings, as reported in Table [8
AIDS researchers and health workers under rationing pressures face a conflict of interest in acknowledging and investigating risks to transmit HIV from patient to patient, as this may undermine public confidence in the competence and motivation of researchers and health workers, leading to under-utilization of essential health services and to preventable morbidity and mortality [75
]. Ministries of Health have a duty to resolve this ethical dilemma while scaling up primary health care services. Informing patients and health workers of the seriousness of HIV transmission risks in minor blood exposures and equipping the health care system to cope with the full demands of infection control will be necessary to avert further iatrogenic HIV transmission. These responsibilities go beyond injection safety interventions such as using only auto-disable (self-destructing, non-reusable) syringes.
WHO assurances that medical injection risks are minimal are not credible, and reflect a pattern of suppressing evidence that heterosexual sex explains less than 90% of HIV transmission in Africa [1
]. Where evidence of harm is egregious, leading AIDS researchers have invoked a relativistic standard, characterizing a 1% prevalence of HIV positive children with HIV negative mothers (in six major African cities) as representing a "low" risk of patient-to-patient HIV transmission [76
]. In a crude irony concerning the social construction of disease, the WHO is defending a 90% estimate that was arrived at by a process of elimination; that is, not on the basis of positive evidence that 90% of HIV infections can be traced to sex in Africa [77
]. In fact infection tracing has been consistently avoided in cases of reportedly non-sexual HIV transmission identified in epidemiological research. Self-reported virgins with HIV, and research subjects with incident infections who claim not to have had sex over the study interval, have been classified as evincing "social desirability bias," by denying epidemiologically implicit sexual behavior.
Blood exposures were of interest to HIV epidemiologists in the 1980s, before a consensus focusing on heterosexual transmission was established for Africa, but even transfusion risks were considered intractable at an early stage. Early Western experts' statements concerning the place of infection control in HIV prevention efforts in Africa were highly pessimistic [1
]. For example, "one cannot hope to prevent reuse of disposable injection equipment when many hospital budgets are insufficient for the purchase of antibiotics." This statement appears in an important 1986 article whose authors include the heads of WHO's Global Programme on AIDS and later UNAIDS for most of the next 21 years [77
]. The problem has not worked itself out, and cannot wait for the day when rationing does not limit the options of health workers in sub-Saharan Africa.