Bulletin of the World Health Organization

Lost opportunities to complete CD4+ lymphocyte testing among patients who tested positive for HIV in South Africa

Bruce A Larson a, Alana Brennan a, Lynne McNamara b, Lawrence Long c, Sydney Rosen a, Ian Sanne b & Matthew P Fox a

a. Center for Global Health and Development, Boston University School of Public Health, 801 Massachusetts Ave, Boston, MA, 02118, United States of America.
b. University of the Witwatersrand, Clinical HIV Research Unit, Johannesburg, South Africa.
c. University of the Witwatersrand, Health Economics and Epidemiology Research Office, Johannesburg, South Africa.

Correspondence to Bruce A Larson (e-mail: blarson@bu.edu).

(Submitted: 25 July 2009 – Revised version received: 08 January 2010 – Accepted: 18 January 2010 – Published online: 16 April 2010.)

Bulletin of the World Health Organization 2010;88:675-680. doi: 10.2471/BLT.09.068981


Although access to antiretroviral therapy (ART) for human immunodeficiency virus (HIV) infection and acquired immunodeficiency syndrome (AIDS) in South Africa has increased dramatically since 2004, the majority of patients with HIV infection initiate care and treatment with very low CD4+ lymphocyte (CD4 cell) counts, often after they are already symptomatic and long after they are first eligible for ART. During 2009 in South Africa, eligibility for ART was determined by a CD4 cell count ≤ 200 cells/μl or by World Health Organization (WHO) Stage IV condition.1 A recent analysis in the Free State province showed that more than half of the patients enrolled in the public sector HIV/AIDS treatment programme began ART with a CD4 cell count < 100>2 Similarly, in an analysis at a large public-sector treatment facility in Johannesburg, 53% of a random sample of patients who started ART in 2005 had a baseline CD4 cell count < 100>3 More recent data from the same facility showed that the median CD4 cell count was 98 cells/μl for all patients who started therapy during 2007–2008, with 25% having a CD4 cell count < 34 cells/μl="" (a="" brennan,="" unpublished="" data,="">

Starting care and treatment for HIV infection and AIDS in patients with a low CD4 cell count imposes unnecessary costs on patients and society.4 Costs to patients include additional morbidity and mortality from HIV- and AIDS-related illnesses and a worse prognosis after initiating ART. Numerous studies have shown that a low CD4 cell count (≤ 100 cells/μl) and WHO Stage IV at the start of treatment are major predictors of mortality.57 Immune system recovery after 3 years on ART is also positively correlated with a patient’s CD4 cell count when ART is begun.8 Access to care and treatment among patients with very low CD4 cell counts is associated with additional health care utilization costs to society through outpatient and inpatient services for opportunistic infections and AIDS-related illnesses. A portion of these costs could be avoided through earlier access to HIV care and treatment, since a substantial share of health care is paid for by the government.

For an HIV+ individual to initiate ART as early as eligibility criteria allow, a specific sequence of events must be completed, starting with HIV testing to diagnose the infection, a baseline CD4 cell count and enrolment in a care programme with regular CD4 cell monitoring, and ending with initiation of ART. In recent years access to HIV testing has increased dramatically, and a recent national survey found that more than half of all South African adults have been tested at least once.9 This expansion of testing, however, has not translated into earlier treatment initiation.57 One likely reason for this is that people who test positive do not always complete CD4 testing after their HIV test. Because of delays in the completion of CD4 testing, further delays occur in obtaining appropriate care and treatment.

CD4 testing involves two steps. First a blood sample is obtained from the patient at a clinic. In South Africa the blood sample is typically then sent to an external laboratory for analysis, and the results are usually available to the clinic in 1 week. For the purposes of the analysis in this article, CD4 testing was considered complete when a patient actually returned to the clinic and obtained the results. At combined voluntary HIV counselling and testing (VCT) and HIV/AIDS clinics, patients are typically tested for a baseline CD4 cell count on the day they test positive for HIV, although they may return for CD4 testing at a later date. When a stand-alone HIV testing centre does not perform CD4 cell counts (e.g. mobile VCT units and VCT provided through youth centres and social organizations), a patient with HIV infection must go to a referral clinic to initiate and complete CD4 testing.

A first step in solving the problem of post-VCT loss to initiation of appropriate HIV care and treatment is to obtain accurate estimates of the magnitude of the problem in public-sector facilities that serve most South African patients. Because patients who test positive for HIV infection at VCT programmes need to complete CD4 testing to enrol in an appropriate care or treatment programme, the primary objective of this study was to estimate the proportion of walk-in VCT patients who completed CD4 testing within 6 and 12 weeks of testing positive for HIV infection at a large, urban, public-sector, comprehensive HIV/AIDS clinic in South Africa.


Study site and population

Themba Lethu Clinic (TLC) is the comprehensive care management and treatment facility for HIV infection and AIDS at Helen Joseph Hospital, a large academic hospital in Johannesburg. This clinic was one of the first public-sector ART roll-out sites in South Africa and remains one of the largest treatment sites in the country. As of May 2009, more than 20 000 patients had initiated ART at the clinic. Services provided by TLC include VCT, CD4 cell testing, pre-ART care and wellness management, tuberculosis care and treatment, ART and adherence support. The clinic is funded by the Gauteng Department of Health with additional support from the United States President’s Emergency Plan for AIDS Relief and the United States Agency for International Development through a South African nongovernmental organization called Right to Care.

This study focused on individuals who are “walk-ins” to TLC for HIV testing. Walk-in patients are those who come to the clinic of their own accord and request an HIV test. Other patients who receive VCT services at the clinic are referred for HIV testing from hospital wards or by other clinics or private physicians because they are already symptomatic. These patients have already experienced delays in access to HIV care and therefore were not the focus of this study.

Walk-in VCT clients receive standard pretest counselling, have an HIV test and then receive post-test counselling that is differentiated according to their test result (positive or negative). If a patient tests positive and states an intention to receive care at TLC (rather than some other facility), the standard procedure is to initiate CD4 testing during the same visit. Patients are then counselled to return in 1 week to collect their CD4 cell count result. The blood sample is sent to a separate laboratory for processing and the CD4 cell count results are available to the clinical staff within 1 week. CD4 testing is considered completed when the patient returns to collect the results.

When patients complete CD4 cell testing, they are enrolled in the general HIV care programme (called the Wellness Program), CD4 cell count results are discussed and future care and treatment plans are reviewed with the patient. Care, wellness and monitoring are initiated if the baseline CD4 cell count is > 250 cells/μl. A similar care and wellness programme as well as a series of ART adherence training activities in preparation for the initiation of ART are started if the baseline CD4 cell count is ≤ 250/μl.

Data collection

We conducted a retrospective record review in June 2009 at TLC to identify all walk-in patients in the VCT programme who tested positive between 1 January 2008 and 28 February 2009. Censoring of the data set as of June 2009 provided a minimum of 12 weeks (1 March through 31 May 2009) for all subjects to have completed CD4 testing after the last date of HIV testing included in the study.

Patients included in the current analysis were identified through the VCT log books used to register patients. For each HIV-positive VCT patient we recorded testing date, sex, age, employment and marital status from the VCT log book, pretest and post-test counselling forms and ART programme medical records. The latter source was used to determine whether a patient initiated and completed CD4 testing after the HIV test date.

Data analysis

At the study site, CD4 cell counts are available within 1 week and patients are requested to return 1 week after their HIV test to collect their CD4 cell count results and complete CD4 cell testing. Waiting for 1, 2 or 3 weeks to complete CD4 testing may have few negative consequences for their health, but a delay of several months or more could involve substantial risks to patients who already have very low CD4 cell counts. To allow for an adequate period to complete CD4 cell testing, we conservatively defined our primary outcome as whether or not a person completed CD4 testing within 12 weeks (84 days) of testing positive for HIV infection. As a secondary analysis, we stratified this time interval into two categories: testing completed within 6 weeks (≤ 42 days) or between 6 and 12 weeks (43–84 days). We used these definitions to estimate the proportion of patients who completed CD4 testing within each period.

We compared baseline demographic characteristics (sex, age, marital status and employment status) of the group of patients who returned to collect their results versus those who did not return within 12 weeks by using simple proportions and medians, where appropriate. We also identified predictors of returning for CD4 cell count results with multivariate logistic regression. All statistical analyses were done with STATA software version 10.1 (STATA Corporation, College Station, United States of America).

The study was approved by the Human Research Ethics Committee (Medical) of the University of Witwatersrand and Institutional Review Board of Boston University’s Medical Campus.


All HIV+ patients

A total of 416 walk-in patients tested positive for HIV at the study site during the study period. Each patient was identified as belonging to one of three groups. The first group included 27 patients (6.5%) who had previously tested positive for HIV infection and received CD4 testing at TLC or elsewhere. We had no data to explain why these 27 patients completed another HIV test, and the majority of these repeat testers (20 of 27) were already eligible for ART (CD4 cells ≤ 200/μl) after their previous HIV test. The second group was composed of 37 patients (8.9%) who did not initiate CD4 testing at TLC during the study period and for whom a baseline CD4 cell count was consequently not available. It is possible that some of these patients requested to go elsewhere for further HIV care. The third group, made up of 352 patients (84.6%), initiated CD4 testing at TLC on or after the day of HIV testing, and the remainder of this analysis focuses on this group.

Initiated CD4 cell testing

Patients who initiated CD4 testing on or after the day when they tested positive for HIV had a median CD4 cell count of 184.5/μl (interquartile range, IQR: 67.5–353.5). As shown in Fig. 1, 191 (54.3% of 352) had a CD4 cell count ≤ 200/μl and were already eligible for ART. The other 161 (45.7% of 352) had a CD4 cell count > 200/μl and were thus eligible for wellness counselling and pre-ART medical care.

Fig. 1. Patients with high and low CD4+ lymphocyte counts who completed CD4 testing within 6 or 12 weeks of voluntary HIV counselling and testing, Johannesburg, South Africa, 2008–2009
Fig. 1. Patients with high and low CD4+ lymphocyte counts who completed CD4 testing within 6 or 12 weeks of voluntary HIV counselling and testing, Johannesburg, South Africa, 2008–2009
HIV, human immunodeficiency virus; IQR, interquartile range; VCT, voluntary HIV counselling and testing.

Among the 191 patients eligible for ART at the time of HIV diagnosis, 51.3% completed CD4 testing within 12 weeks and the remaining 48.7% did not (Fig. 1). Among the 161 patients not eligible for ART at the time of HIV diagnosis (CD4 cells > 200/μl), only 14.9% completed CD4 testing within 12 weeks. Among all 352 patients, only 35% completed CD4 testing within 12 weeks of HIV testing.


There were no differences in basic demographic characteristics between those who returned for their CD4 cell count results within 6 weeks, those who returned within 12 weeks, and those who did not return (Table 1). Most patients were women; median age ranged between 36 and 40 years, less than one third were employed and approximately one half were married.

A substantial majority of patients who completed CD4 testing within 6 weeks (82.4%) or between 6 and 12 weeks (75.7%) after HIV testing had a baseline CD4 cell count ≤ 200 cells/μl. However, 40.4% of all patients who did not complete CD4 testing were also already eligible for ART based on their CD4 cell counts. The baseline median CD4 cell count of 77 cells/μl in patients who completed CD4 testing within 6 weeks was significantly lower (P < 0.01) than="" the="" median="" count="" of="" 163="" cells/μl="" in="" patients="" who="" completed="" testing="" between="" 6="" and="" 12="" weeks.="" the="" median="" cd4="" cell="" count="" of="" 263/μl="" in="" patients="" who="" did="" not="" complete="" testing="" within="" 12="" weeks="" was="" significantly="" higher="">P < 0.01) than="" in="" both="" of="" the="" groups="" who="" completed="" testing="">

Multivariable analysis (Table 2) showed that the baseline CD4 cell count was negatively associated with the odds of a patient completing CD4 testing. Compared to the group with the lowest baseline CD4 cell count (0–100 cells/μl), those whose baseline CD4 cell count was between 101 and 200 cells/μl had 0.41 times the odds of completing their tests within 6 weeks. Those with higher CD4 cell counts (201–300 cells /μl or > 300 cells/μl) had an even lower odds of completing testing. When we used the 12-week criterion, we found that the odds of patients with baseline CD4 cell counts from 100 to 199 cells/μl completing testing were similar to the odds for those in the lowest CD4 cell count category (< 100 cells/μl),="" whereas="" those="" in="" the="" higher="" categories="" continued="" to="" be="" substantially="" less="" likely="" to="" complete="" testing="" compared="" with="" those="" having="" the="" lowest="" cd4="" cell="">


The UNAIDS/WHO policy statement on HIV testing from 2004 concludes that, “access to integrated prevention, treatment and care services” is a cornerstone of HIV testing scale-up.10 Nevertheless, scale-up of HIV testing since 2004 in South Africa has not resulted in earlier initiation of treatment.2,3,11 At the major urban, public care and treatment site in South Africa involved in this study, 54% of walk-in patients who tested positive for HIV infection were already eligible for ART on the day of testing (CD4 cells ≤ 200/μl), yet 48% of these patients did not complete CD4 testing within 12 weeks and therefore did not obtain appropriate medical care for their illness. Some of these individual are likely to have died, given that some patients who are known to be eligible for ART die before initiating therapy.12

To increase substantially the proportion of patients who initiate ART as soon as they are eligible, a large proportion of patients need to complete CD4 testing before they become eligible (i.e. while CD4 cells are still > 200/μl). Among such patients in the current study, however, 85% did not complete CD4 testing within 12 weeks of being diagnosed as HIV+. Under current conditions, these patients are likely to return for HIV care and treatment at a later date with substantially lower CD4 cell counts.

This study had several limitations. First, we had CD4 cell count data only from the site at which patients were tested for HIV. It is possible that some patients who did not complete CD4 testing at TLC went to another clinic, were tested again for HIV and then completed CD4 testing there. We had no way of tracking HIV+ clients of the VCT who were seen at other clinics. Second, our results are based on data from only one comprehensive HIV care and treatment facility in South Africa, albeit one of the largest HIV clinics in the country. The results show the experience at one site, and caution should be used in generalizing the results to populations of patients walking in for VCT services at other facilities. Finally, we have no information about the reasons why VCT clients did or did not complete CD4 testing.

The results in this study contribute to a growing literature that explores the problem of post-VCT loss to initiation of HIV/AIDS care and treatment in South Africa and other sub-Saharan countries.11,1315 In a semiprivate clinic in Durban with a high prevalence of HIV infection (63% of all patients tested), about 53% of the patients with newly-diagnosed HIV infection did not complete CD4 testing within 8 weeks. In a small community outside Cape Town, 62% of patients who tested positive for HIV infection completed CD4 testing within 6 months of testing positive. Given the relatively long period for completion in this study (24 weeks) and that about 30% of these patients had a CD4 cell count < 200 l="" at="" the="" time="" of="" hiv="" testing,="" it="" remains="" likely="" that="" some="" of="" these="" patients="" completed="" cd4="" testing="" as="" part="" of="" their="" access="" to="" other="" medical="" care="" for="" aids-related="" conditions.="" an="" analysis="" of="" the="" data="" from="" two="" towns="" in="" central="" mozambique="" showed="" that="" approximately="" half="" of="" all="" patients="" who="" tested="" positive="" for="" hiv="" in="" multiple="" testing="" centres="" did="" not="" initiate="" cd4="" testing="" within="" 60="" days="" of="" their="" hiv="">14 In a rural area in the northern part of the United Republic of Tanzania, where patients who tested positive for HIV at VCT sites were referred to another clinic in a nearby town about 20 km away, 72% of men and 66% of women completed an initial referral visit to the clinic within 6 months of their HIV test, although it is not known whether these patients actually completed CD4 cell testing.15

Significant improvements are needed to integrate HIV testing into the full continuum of HIV/AIDS care and treatment in South Africa. Before interventions can be implemented and evaluated, however, improved referral systems with a more structured process of patient tracking are needed in all stages of pre-ART medical care and monitoring (i.e. HIV testing, CD4 testing and enrolment in care to initiation of ART). The current system, which is based on separate systems to manage and monitor patients at the VCT and HIV care levels, constrains the monitoring and evaluation of linkages to HIV care.

Patients who test positive for HIV infection must overcome several barriers if they wish to complete CD4 cell testing. We suspect that these obstacles are similar to barriers to treatment adherence in general. Such barriers include (but are not limited to) transport costs and time, loss of earnings, lack of financial resources, stigma, lack of family support and perceived need when individuals are asymptomatic.1622 Future research must look for ways to overcome these barriers, either directly or through other interventions, to improve post-VCT medical care.


The authors thank the staff and management of Themba Lethu Clinic for their assistance with understanding patient flows and patient file management. This analysis could not have been completed without their generous support. The authors thank Melinda Wilson of the United States Agency for International Development for her recognition of the importance of this topic, leadership in securing funding and advice to the study team. The authors also thank the journal reviewer for very helpful comments and suggestions. The opinions expressed herein are those of the authors and do not necessarily reflect the views of the Themba Lethu Clinic, USAID, the National Institute of Allergy and Infectious Diseases or the National Institutes of Health.


Funding for this study was provided by the South Africa Mission of the United States Agency for International Development under the terms of Cooperative Agreement GHSA-00-00020-00, Country Research Activity (G ⁄PHN⁄HN⁄ CS). Matthew Fox was supported by Award Number K01AI083097 from the National Institute of Allergy and Infectious Diseases.

Competing interests:

None declared.