Main article: Reports from the I3th ICASA meeting in Nairobi

This article originally appeared in HIV & AIDS treatment in practice, an email newsletter for healthcare workers and community-based organisations in resource-limited settings published by NAM between 2003 and 2014.
This article is more than 20 years old.

Summary

By Julian Meldrum

This report covers a selection of topics and inevitably misses out some important themes. Those covered, with key messages, are listed below in the order in which they are reported in the article.

ARV ACCESS PREPARATIONS

  • Collection and maintenance of clinical records is going to be a vital element in expanded treatment programmes, not least because international funders are going to need hard evidence that the treatment is reaching people who need it and is delivering benefits to them.
  • Many aspects of healthcare systems and practices will need to be improved, and these need not wait until ARVs are available to receive attention.
  • Considering how to deliver treatment to children, and how to meet community needs in respect of TB, malaria, clean water, nutrition, need to be included in this process.

ARV PROGRAMME RESULTS

  • Considering the limitations of many initial treatment programmes, in terms of the drugs used, the outcomes have been impressive. The challenge will be to maintain this as the scale increases.

Glossary

sputum

Material coughed up from the lungs, which can be examined to help with diagnosis and management of respiratory diseases.

VCT

Short for voluntary counselling and testing.

herpes simplex virus (HSV)

A viral infection which may cause sores around the mouth or genitals.

sensitivity

When using a diagnostic test, the probability that a person who does have a medical condition will receive the correct test result (i.e. positive). 

paediatric

Of or relating to children.

CO-TRIMOXAZOLE AND ISONIAZID PROPHYLAXIS

  • Countries that, until now, have not used isoniazid (INH) prophylaxis are moving towards doing so, in view of evidence that it does reduce mortality and active TB incidence among people with HIV.
  • Evidence for co-trimoxazole (CTX) may be weaker, but probably due to poor adherence which points to the need for active community education for this to have public health benefits.

DIAGNOSTICS

  • Rapid testing kits, when properly chosen and used in appropriate combinations, can give excellent results. Otherwise, they can be disastrous with high rates of false positives.
  • Rapid tests on oral fluids can also be applied to sputum samples, but (a) this does not remove the need for confirmatory testing and (b) great care must be taken, not to spread misunderstandings about HIV transmission.

HSV TREATMENT

  • It may be time to add aciclovir (to treat Herpes Simplex Virus 2) to the protocols for syndromic management of genital ulcer disease and a study from Mali suggests how this might be done.

OCCUPATIONAL SAFETY

  • Training on universal precautions can help to ensure that people with HIV and their families do not suffer stigmatisation as a result of accessing health care. There are online information resources that can help.

Internet Resources

The ICASA conference website here has both the abstract book (2.4 Mbytes) and a programme supplement (2.9 Mbytes) including late breaker abstracts. Both have been well produced, with impressive efficiency, and the email addresses of most presenters are included.

Extensive reporting of the conference by Health and Development Networks (HDN) is available through their website here and a final published report of the conference is promised for later.

ARV Access Preparations

In Burkina Faso, community organisations are building up medical records for people with HIV, to develop a profile that can be used for planning future access to ARVs. People are being trained in collecting the necessary information, which is then being entered into databases that can provide a more detailed basis for those plans than has previously been available.

In Mombasa, Kenya, an international consortium of NGOs (Shikely, Adungosi) is working with the Kenyan government to to prepare for ARV introduction with a systematic analysis of the strengths and weaknesses of the existing healthcare system. Their process may well be of interest to others doing the same, and has a lot in common with that outlined in HATIP #9, based on resources produced by John Snow, Inc. A very similar process is being undertaken in Rwanda too (Ostyn) and Ghana (Kwasi, Field-Nguer). Could this be what WHO has in mind for the emergency teams that it is preparing to offer to priority countries for ARV access?

Researchers in Mombasa carried out interviews with a number of people with HIV, some of whom had received ARVs, some of whom had received TB treatment, to assess the level of interest in and acceptability of a proposed modified DOTS system for delivering ARV treatment in the community. The main conclusion was that most would prefer to receive such treatment at a clinic rather than at home, since it would enhance their feeling of being in control, including control over disclosure of their HIV status. Another conclusion may be that one size does not fit all people, and that treatment which can support individual preferences and strategies for adherence which people choose for themselves may be more successful in the longer term.

Zambia has seen some extensive community-based discussion of the prerequisites for effective ARV programmes, in which the need for greater efficiency in various systems has been identified by everyone concerned (Dhaliwal). The test will be to see what happens when they actually receive some resources to start implementing effective treatment.

In Kisumu, Kenya, which apparently has the highest HIV prevalence in the country at around 25% of the adult population action to strengthen the health care services is seen as the clear priority before ARVs can be introduced (Otieno).

Similarly, in Kibera a slum district of Nairobi where volunteers work with health workers to provide HIV/AIDS related services, there was a clear need for improvement in training on recognising TB symptoms and access to diagnostics for malaria and other conditions, alongside any action that will be needed when ARVs are introduced there (Marum). Nonetheless, plans to do so are well advanced (Macharia, Muhenje, Njoroge).

In Botswana, which now has Africas highest-profile ARV treatment programme, efforts are now being made to ensure that children can be treated with ARVs by non-paediatric HIV specialists, with the development of dosing charts and schedules for the purpose (Ncube). No sign yet, however, of promised cheaper generic paediatric formulations of the drugs needed.

One of the many practical challenges now being met in Botswana is the establishment of efficient medical records. A hybrid computer-plus-paper system has been chosen for maximum resilience in the face of operating problems and seems to be working well (Hermann).

A proposed electronic system based on handheld computers and wireless connections, at the doubtless soon-to-be-renamed Moi Teaching and Referral Hospital in Kenya, sounds wonderful except that it would be reassuring to know that they, too, had some hard copy back-ups, just in case (Siika).

  • Adungosi J. Assessment of health care services before introducing ART in Mombasa, Kenya. 13th ICASA, Nairobi, abstract 959883, 2003.
  • Dhaliwal M et al. Community preparedness for ARV treatment in Zambia. 13th ICASA, Nairobi, abstract 375096, 2003.
  • Field-Nguer ML et al. Making antiretroviral therapy a part of comprehensive care in Ghana, Kenya and Rwanda: the Start Initiative. 13th ICASA, Nairobi, abstract 398344, 2003.
  • Hawken M et al. PLHA views on modified DOT strategy to promote adherence to HAART. 13th ICASA, Nairobi, abstract 493470, 2003.
  • Hermann B et al. Developing and implementing a hybrid electronic records management system. 13th ICASA, Nairobi, abstract 291500, 2003.
  • Kwasi T et al. Setting up ART program in a district based setting in Ghana. 13th ICASA, Nairobi, abstract 432735, 2003.
  • Macharia D et al. Developing comprehensive AIDS care including ARVs in a Nairobi slum. 13th ICASA, Nairobi, abstract 345508, 2003.
  • Marum E et al. Assessment of AIDS care and support for slum residents in Nairobi, Kenya. 13th ICASA, Nairobi, abstract , 2003.
  • Muhenje O et al. Needs, attitudes and beliefs related to adherence to ARV in Nairobi slum. 13th ICASA, Nairobi, abstract 744637, 2003.
  • Ncube P et al. Pediatric antiretroviral treatment dosing guidelines for non-pediatricians. 13th ICASA, Nairobi, abstract 567210, 2003.
  • Nguyen V-K et al. Burkina: a community cohort to prepare for expanded access to treatment. 13th ICASA, Nairobi, abstract 295886, 2003.
  • Njoroge A et al. Integration of TB/HIV services into a slum community program in Nairobi. 13th ICASA, Nairobi, abstract 438014, 2003.
  • Ostyn B et al. Assessment of health care services before introducing ART program in Rwanda. 13th ICASA, Nairobi, abstract 502923, 2003.

* Otieno J et al. Hospital and community health services in anticipation of ARV programs. 13th ICASA, Nairobi, abstract 608866, 2003.

* Shikely K. Process for introducing a district-based ART program in Mombasa, Kenya. 13th ICASA, Nairobi, abstract 199396, 2003.

* Siika A et al. Establishing an electronic medical records system for outpatient care of HIV infected patients at Eldoret, Kenya. 13th ICASA, Nairobi, abstract 300019, 2003.

ARV Programme Results

A report from Nigeria is a little worrying, as it suggests that the 50 patients described were only treated with nevirapine 200mg daily, while getting d4T (stavudine) 40mg and 3TC (lamivudine) 150mg twice daily. This may be the correct lead-in dosage, for patients with a body weight of 60kg or over, but after the first two weeks the nevirapine dosage should have risen to 200mg twice daily. Still it was only a preliminary report and it did find positive results, with adherence described as good in 85% of patients treated and clear evidence of improved health. Hopefully, the abstract does not tell the complete story about the dosages used (Oni Idigbe).

Another report of the same programme, from the treatment centre at Jos, says that of 176 patients three had to stop due to severe reactions to the nevirapine (Idoko). This implies that they are treating through minor nevirapine rash, which is almost certainly the right course of action given the absence of other realistic treatment options in their setting.

In Chiradzulu district, Malawi, MSF continues to treat hundreds of adults and dozens of children with HIV, in a programme that continues to expand. As of January 2003, 340 (81.1%) of their patients are still on ARV. 93% of patients with an adherence assessment available at their last visit reported taking >80% of their prescribed ARV doses (Durier)

In Mozambique, a programme that has relied on use of generic fixed dose combination drugs (AZT or d4T with 3TC and nevirapine) has achieved a substantial fall in death rates (by 76-83%) and reports limited and manageable problems with drug toxicity. ARVs are combined with nutritional supplementation, cotrimoxazole prophylaxis, treatment of opportunistic infections and malaria. After treating several hundred patients, the programme is now in the course of being scaled up (de Luca).

In Uganda, one of the first clinics to provide ARVs reported respectable levels of virological suppression in patients between 45 and 50% - among a relatively small number of patients who commenced treatment between 1998 and 2000, especially considering that a significant minority began on dual therapy with nucleoside analogue drugs (Bahendeka).

Their counterparts in Cote dIvoire also seem to have done relatively well, with more than 60% achieving viral load suppression below 400 copies despite starting on regimens that would not be todays first choices. The majority were still suppressing viral load a year later (Lago, Toure). There are also encouraging results from one of the first paediatric HIV treatment programmes to report (Diomande).

However, another programme in Abidjan, Cote dIvoire, has begun to measure the rate of drug resistance among circulating viruses. Early results show that there are already detectable ARV mutations, probably associated with unregulated and inconsistent ARV access in the city. Further monitoring is essential (Daquin). High rates of drug resistance reported among a small number of people with HIV on drug treatment in Kinshasa, Democratic Republic of the Congo, were clearly associated with past experience of dual therapy with nucleoside analogue drugs (Ostyn).

A presentation from Pointe Noire, Congo (Brazzaville), reported 98% survival over 6 months of a cohort of 23 men and 26 women on ARV treatment, a majority of regimens including a protease inhibitor (Tran-Minh).

Reasons for switching treatment were identified in Khayelitsha, Western Cape Province, South Africa: The majority of patients began treatment with zidovudine, lamivudine and one of efavirenz or nevirapine& Highest rates of change due to intolerance at one year were for nevirapine (5.2%) and zidovudine (3.6%). Incident tuberculosis and pregnancy occasioned additional switches from nevirapine and efavirenz respectively. This kind of information will undoubtedly be of great value in making further plans for large-scale treatment access based on standardised first-line regimens (Boulle).

In Senegal, the outcomes of treatment for 20 patients with HIV-2 were reported (Ndiaye). Since HIV-2 does not respond to treatment with nevirapine or efavirenz, and viral load tests have not been available for monitoring it, this sets particular challenges. However, going by evidence of rising CD4 counts, weight gain (20 kg in a year), reduced opportunistic infections, and reported improvements in quality of life, treatment with triple nucleosides or two nucleosides with a protease inhibitor can be successfully implemented for the majority of patients. Criteria for ARV treatment were set as a CD4 count of 200 if asymptomatic, 350 if symptomatic. New HIV-2 viral load assays are being developed, which may improve the ability to monitor what is happening (Alabi).

  • Alabi A et al. Development, evaluation and potential applications of an in-house viral load assay for quantification of HIV-1 and HIV-2 in human plasma. 13th ICASA, Nairobi, abstract 490491, 2003.

Bahendeka S et al. Long-term virologic & CD4 count response to antiretroviral drugs in Uganda. 13th ICASA, Nairobi, abstract 390888, 2003.

  • Boulle A et al. Durability of regimens and individual drugs in a public sector ARV program. 13th ICASA, Nairobi, abstract 886025, 2003.
  • D'aquin TT et al. Primary HIV-1 ARV resistance observatory in Côte d'Ivoire (ANRS study). 13th ICASA, Nairobi, abstract 153634, 2003.
  • De Luca et al. Efficacy of free HAART in HIV 1 infected subjects in Mozambique. 13th ICASA, Nairobi, abstract 665186, 2003.
  • Diomande VK. Virologic and immunologic response to highly active antiretroviral therapy among HIV infected children in Côte dIvoire. 13th ICASA, Nairobi, abstract 1098761, 2003.
  • Durier N. A program of triple antiretroviral therapy in Chiradzulu District, Malawi. 13th ICASA, Nairobi, abstract 569320, 2003.
  • Idoko J et al. Antiretroviral therapy among HIV infected adults in Nigeria: experience from the Jos centre of the national antiretroviral programme. 13th ICASA, Nairobi, abstract 672204, 2003.
  • Lago H et al. Viral load suppression in adults on antiretroviral therapy in Cote d'Ivoire. 13th ICASA, Nairobi, abstract 412145, 2003.
  • Ndiaye I. Immunological and clinical efficacy of ARV treatment of patients living with HIV-2 in Senegal. 13th ICASA, Nairobi, abstract 468224, 2003.
  • Oni Idigbe E et al. A preliminary report on the Nigerian ARV programme. 13th ICASA, Nairobi, abstract 485145, 2003.
  • Ostyn B et al. Resistance patterns in patients on ART in Kinshasa, RDC. 13th ICASA, Nairobi, abstract 910766, 2003.
  • Toure S et al. Survival and morbidity in HIV+ adults receiving ARV therapy, Abidjan. 13th ICASA, Nairobi, abstract 756828, 2003.
  • Tran-Minh T et al. Bilan de 6 mois de tritherapie au Congo. 13th ICASA, Nairobi, abstract 833805, 2003.

Co-trimoxazole and isoniazid prophylaxis

Lukwiya in Uganda reported on a systematic programme based first on consultation and then education of people with HIV in Uganda, to introduce CTX prophylaxis at a cost of US $20 per year per person, which is being offered to tens of thousands of people in contact with TASO. The idea of taking medicines to stay well is unfamiliar to people, but its acceptance will be vital for future effective access to ARVs.

Also in Uganda, Nakiyingi reported a cohort study of people with HIV who were positive on a PPD skin test for TB. 6 months of isoniazid, given to 96 patients, definitely reduced TB risk and cut death rates from 22% a year in the control group to 8% in the treated group in the year when they were treated with INH. However, when CTX was offered to all members of this cohort, instead of INH, the death rate rose to 14% in those who had received INH compared to 20% in the control group, which was not a significant difference. The conclusion was that the INH effect was fairly short-lived, and that CTX had little or no effect in this population. However, this was an intent to treat analysis and the level of adherence to treatment is not reported in the abstract.

At Kendu Bay on Lake Victoria in Kenya, CTX prophylaxis has been introduced but has run up against other priorities in the lives of a poverty-stricken community (Aulo). People understandably are more concerned to treat or avoid recurrence of illnesses they may have experienced and find it hard to grasp the idea of preventing illnesses they have not yet experienced.

In Rwanda, a programme to offer 9 months of INH prophylaxis, and/or lifelong CTX prophylaxis, through hospital-based HIV voluntary counselling and testing sites, seems to have achieved high levels of uptake and adherence, with more than 60% completing at least 6 months of INH in the group offered that treatment (Karibushi).

The same idea has been adopted in Mombasa, Kenya, where CTX prophylaxis is being offered through VCT based in two primary health care clinics to local residents. After careful screening for anaemia and any history of drug sensitivity, treatment was uneventful, with no reports of unwanted effects. However, follow-up rates were disappointingly low (Chai).

Yet another Ugandan study (Lule) looked at the effects of measures to improve water safety, alongside the use of co-trimoxazole prophylaxis. Households were randomised to receive a safe water system intervention based on storage of water in plastic containers and use of hypochlorite tablets. This reduced diarrhoea rates (recorded weekly) by 30%, and a further 30% reduction was achieved when co-trimoxazole was provided to HIV positive household member aged 5 years or over. So perhaps CTX has its uses after all.

A Nairobi hospital study (Siika) of patients with chronic cough who were HIV positive and smear-negative for TB found that a high proportion of them had Pneumocystis pneumonia (PCP). This is obviously a highly selected population, but may still point to underdiagnosis of PCP and underestimation of the potential value of CTX prophylaxis in some African settings. Another report from the same setting also highlighted PCP (Bii).

  • Aulo T. Integrated community based Bactrim prophylaxis, Kendu Bay. 13th ICASA, Nairobi, abstract 862653, 2003.
  • Bii CC et al. Opportunistic fungal infections in HIV/AIDS in Nairobi, Kenya. 13th ICASA, Nairobi, abstract 884774 , 2003.
  • Chai K et al. Integration of co-trimoxazole prophylaxis into VCT in Mombasa, Kenya. 13th ICASA, Nairobi, abstract 842013 (and also 440993), 2003.
  • Karibushi B et al. Feasibility of interventions to prevent opportunistic infections in Rwanda. 13th ICASA, Nairobi, abstract 790781, 2003.
  • Lukwiya M et al. Co-trimoxazole prophylaxis (CP) programme at The AIDS Support Organisation. 13th ICASA, Nairobi, abstract 249882, 2003.
  • Lule JR et al. Effect of safe water and co-trimoxazole on diarrhea among people with HIV. 13th ICASA, Nairobi, abstract 776099, 2003.
  • Nakiyingi J et al. INH and co-trimoxazole prophylaxis in PPD positive HIV infected Ugandans. 13th ICASA, Nairobi, abstract 701254, 2003.
  • Siika A et al. Aetiology of chronic cough in HIV infected adults with negative sputum smears for acid fast bacilli: a bronchoscopic evaluation of 62 patients at Kenyatta national hospital, Nairobi. 13th ICASA, Nairobi, abstract 205385, 2003.

Diagnostics

A Nigerian study in which stored plasma samples were re-tested found that the poorly managed introduction of rapid HIV antibody tests, used in contravention of WHO guidelines as a single test, had led to as many as 10% of those testing positive being misdiagnosed as HIV positive when in fact they were HIV negative. If anything positive can emerge from this miserable tale, it should be a determination that no-one, no matter what the HIV prevalence in a population, should be given an HIV diagnosis on the basis of an unconfirmed test.

A reported evaluation of Unigold and Biotec rapid tests for use in Nigeria, which found Unigold satisfactory and rejected the Biotec ones, shows that some lessons have been learned (Aniedobe).

In Namibia, a group has evaluated Capillus (Trinity Biotech), Determine (Abbott), OraQuick (OraSure Technologies), Hema-strip (ChemBio), StatPak (ChemBio), and UniGold (Trinity Biotech). 291 HIV negative and 217 HIV positive samples were tested by the six rapid tests. All rapid tests (except UniGold) yielded 100% sensitivity and 100% specificity. One false negative was observed by UniGold (99.5% sensitivity; 100% specificity) (Weiss).

In Kenya, KEMRI has been undertaking its own evaluation of the confirmatory system to be used for rapid test kits in VCT systems, and seems to be getting excellent results (Kemunto). Similarly, the Uganda Virus Research Institute (UVRI) has evaluated rapid tests used on fresh blood and concluded: A sequential algorithm based on Determine and Unigold with Hemastrip as a `tie-breaker' offers a cost-effective HIV rapid testing algorithm for finger-stick blood (Mujurizi). Similarly, a group working in Northern Uganda found that rapid tests, when used in combination, could match the performance of ELISA tests (Nattabi).

UVRI has also looked at the use of antibody testing systems on dried blood spot samples, as a basis for quality assurance for field-based testing (Mujurizi). The results are very convincing and if blood spots are going to be collected for other purposes, e.g. viral load tests, then they can be used to check on the accuracy of antibody tests too.

In Tanzania, a public-private partnership is evaluating the use of rapid test kits by primary health care workers as part of a strategy to expand access to HIV testing. With thorough training in the use of two different tests and careful management, including a quality assurance scheme, they have demonstrated that this can in fact give excellent results (Saba). This is already in place in Zimbabwe, where non-health-care staff are carrying out the tests, using Determine and Unigold tests in parallel. Where results disagree, a lab-based ELISA test is done (Osewe).

An interesting idea from Kenya, which complements efforts to integrate HIV care with TB services, is to use sputum samples as a basis for HIV testing (Odhiambo). Using the OraQuick test designed for samples of oral fluid and comparing the results to rapid tests applied to blood gave encouraging preliminary results. Out of 143 sputum specimens examined so far, 5 specimens were excluded from analysis on account of sputum "thickness" obscuring valid reading. Based on 138 valid readings, the OraQuick test has a sensitivity of 97.0% (97/100) and a specificity 86.8% (33/38). While this is not as good as blood testing, and would obviously require confirmation if it were ever used as a screening test, it could certainly be informative for surveillance purposes. Further results based on larger numbers of tests are promised later.

One hazard of introducing such novel testing strategies, highlighted in a Kenyan evaluation of the same test used on oral samples (Ngure), is that it may spread misconceptions about how HIV is transmitted. Therefore, if sputum samples are to be tested for HIV it will be essential to make it clear that HIV is NOT an airborne disease!

  • Aniedobe M et al. Evaluation of Unigold and Biotec HIV diagnostic kits. 13th ICASA, Nairobi, abstract 658230, 2003.
  • Kemunto O. Monitoring of quality assurance in HIV testing in VCT facilities. 13th ICASA, Nairobi, abstract 326360, 2003.
  • Mujurizi T et al. Operational evaluation of rapid HIV test algorithms for fingerstick blood. 13th ICASA, Nairobi, abstract 101779, 2003.
  • Mujurizi T et al. Quality control testing algorithm for HIV on finger-stick dried blood spots. 13th ICASA, Nairobi, abstract 894855, 2003.
  • Nathan E et al. Quality assurance of HIV screening tests in Lagos, Nigeria. 13th ICASA, Nairobi, abstract 125539, 2003.
  • Nattabi B et al. Evaluating the performances of an HIV diagnostic algorithm based on rapid tests in resource limited settings. 13th ICASA, Nairobi, abstract 508014, 2003.
  • Ngure P et al. Perceptions of rapid oral testing for HIV in Kenya. 13th ICASA, Nairobi, abstract 521046, 2003.
  • Odhiambo J et al. Evaluating a sputum-based test for HIV among TB patients and suspects. 13th ICASA, Nairobi, abstract 491984, 2003.
  • Osewe et al. Accuracy of rapid HIV testing algorithm within new start VCT in Zimbabwe. 13th ICASA, Nairobi, abstract 788408, 2003.
  • Saba J et al. Taking VCT out of the laboratory: the Rungwe experience in Tanzania. 13th ICASA, Nairobi, abstract 413637, 2003.
  • Weiss C et al. HIV rapid test evaluation in Namibia. 13th ICASA, Nairobi, abstract 643301, 2003.

HSV Treatment

In Mali one of the worlds poorest countries syndromic treatment for genital ulcer disease in women and men should now include aciclovir treatment of episodes, according to an intensive diagnostic study of 37 cases (27 female, 10 male) from clinics. Only 33% of ulcers could be attributed to a particular pathogen, but of these more than 90% were due to HSV-2, with less than 3% due to chancroid, and none to syphilis, the conditions generally targeted through syndromic management strategies. Multiple ulcers were especially likely to be due to HSV-2.

  • Doumbia S et al. HSV2 as major component of genital ulcers syndromic management in Mali. 13th ICASA, Nairobi, abstract 234718, 2003.

Occupational safety

The importance of good infection control practice in health care settings is clear, not least because bad practice such as limiting protective measures to people known to be HIV positive can be stigmatising and harmful to the goal of expanding access to treatment and care. Middleberg reported on training resources developed by Engender Health, an NGO focussed on womens reproductive health, including a free downloadable course (in English and Spanish) which they have used in Malawi, Guinea, Ghana, South Africa, Senegal, Nigeria, India, Nepal, Egypt and other countries.

In Ibadan, Nigeria, HIV tests on samples routinely handled in a hospital laboratory showed that 13% were HIV positive and highlighted the need for stringent infection control precautions to protect the technicians (Odaibo).

  • Middleberg M et al. Practical strategies for preventing HIV transmission in medical settings. 13th ICASA, Nairobi, abstract 476814, 2003. http://www.engenderhealth.org/res/onc/index.html
  • Odaibo G et al. Risk of nosocomial HIV infection of laboratory. 13th ICASA, Nairobi, abstract 975434, 2003.