Many sub-Saharan countries have rapidly expanded HIV care and treatment services over the past years, increasing the ART coverage in this region to 33% in 2009.1 To identify evidence-based interventions associated with better outcomes, several leading organizations, including the World Health Organization (WHO), the Global Fund, and the US Institute of Medicine have called for an increased emphasis on operations research in HIV care and treatment scale-up.2–4
Mozambique, with an estimated national HIV prevalence of 11.5% in 2009,5 has rapidly scaled-up HIV care and treatment in the past years. By the end of 2009, approximately 170,000 people, including an estimated 13,500 children, had initiated antiretroviral therapy (ART) at more than 200 clinics, for an estimated national coverage of 42% of those estimated to be in need of treatment.6 We embarked on an effort to utilize existing HIV care and treatment service delivery data from 28 clinics in Mozambique that were receiving support from ICAP-Columbia University through funding from the President's Emergency Plan for AIDS Relief to inform design and quality of the programs as part of the Identifying Optimal Models of HIV Care Collaboration. The aim of this collaboration is to evaluate the variation in program outcomes across sites and settings and to identify clinic and contextual-level factors associated with key outcomes. To set the stage for future analyses, the objective of this analysis is to describe characteristics of these clinics, baseline characteristics of the cohort at enrollment, and ART initiation along with the frequency of key outcomes such as loss to follow-up (LTFU) and death.
Study Design and Study Population
We conducted a descriptive analysis of routinely collected service delivery data on HIV-infected patients enrolled at 28 HIV care and treatment clinics within the national HIV care and treatment program in Mozambique. These clinics were receiving support from ICAP-Columbia University through funding from the President's Emergency Plan for AIDS Relief and had an electronic patient-level database as of December 2009. These clinics, located in 5 Mozambican provinces, represented 13% of all clinics providing ART in the country at the end of December 2009 (Fig. 1). The study population includes all patients (adult and pediatric) who were ever enrolled in HIV care at these clinics by the end of December 2009.
Patient information routinely collected during each clinic visit was documented by clinicians on national patient forms and included in patient charts. This included sociodemographic information (ie, sex, marital status, point of entry) and clinical and laboratory parameters (ie, medications, opportunistic infections, WHO stage, CD4 cell count, etc.). Information from these forms was entered on a daily basis by trained data clerks into an electronic patient tracking system database that was deployed at all 28 sites. The electronic patient tracking system is a Microsoft Access electronic database developed by ICAP that is based on the Mozambique national forms, which are used for data entry into the system. Data quality assessments were done every 6 months to check the completeness and accuracy of the database compared with the paper forms.
Clinic and Program Characteristics
Data on clinic and program characteristics were derived from routinely conducted structured site assessments completed annually since 2007 by ICAP field staff. These assessments capture information through observation and through interview of site staff regarding the characteristics and availability of services at the time of the survey: the setting (urban, rural), the type of facility (primary, secondary, tertiary), presence at the clinic or facility of programmatic services [eg, prevention of mother to child transmission program (PMTCT), voluntary counseling and testing (VCT), tuberculosis treatment], patient support services (eg, peer educators or outreach for patients who miss clinic visits), laboratory services (eg, availability of CD4 testing), and staffing characteristics (eg, number/type of providers).
Guidelines for HIV Care and ART
National guidelines for care of HIV patients in place during the study period recommended that all HIV patients at their first clinic visit should have a comprehensive history, receive a physical examination, and get CD4 count to assess ART eligibility. Patients >5 years old were considered eligible for ART initiation if they had the following: (1) WHO stage IV, (2) WHO stage III and CD4 count <350 cells per microliter, or (3) CD4 count <200 cells per microliter irrespective of WHO stage. Patients <5 years old were considered ART eligible if they had WHO stage III/IV, CD4 percentage <25% (<12 months), CD4 percentage <20% (12–35 months), or CD4 percentage <15% (36–59 months). For patients not eligible for ART, the guidelines recommended semiannual visits with clinical and CD4 assessment. Once ART eligibility is established, patients needed to attend 3 ART readiness pretreatment counseling sessions and disclose their HIV status to a treatment partner, upon which a physician reevaluated the patient to confirm ART eligibility and readiness and prescribed ART medication. Pediatric patients went through a similar process involving their main caregivers. At the time of ART initiation, guidelines recommended obtaining information on age, weight, and WHO clinical stage. First-line ART regimens were prescribed by a physician according to national protocols, and at the time of this study, this included lamivudine (3TC) plus nevirapine (NVP) plus either zidovudine (AZT) or stavudine (d4T). The initial follow-up schedule for those starting ART included 2 clinical visits in the first month to assess adherence and detect possible adverse effects and then monthly visits for clinical assessments and medication pick up. For patients with CD4 count <350 cells per microliter or clinical WHO stage III/IV, cotrimoxazole was indicated.
Patient characteristics at enrollment into care and at ART initiation were compared by sex using the χ2 test for categorical variables and Wilcoxon test for medians. For patients' characteristics at enrollment into care (ie, weight, WHO stage, CD4 cell count), any value within 6 months of enrollment was used. For patients' characteristics at ART initiation, we considered the closest value using a window period of 6 months before and 1 month after ART initiation. Age distribution at ART initiation for adult and pediatric patients was examined using population age pyramids. For children <5 years old, HIV-associated immunodeficiency was defined based on the CD4 percentage according to the age-adjusted CDC classification [severe suppression (<15%), moderate suppression (15 to <25%) and no evidence of suppression (≥25%)].7 Weight-for-age Z scores were calculated based on the CDC standards.8 A substantial proportion of patients had missing information on CD4 count and WHO stage at enrollment and ART initiation; they were included in our analyses with CD4 count and WHO stage assigned to “missing” categories. LTFU was defined as not having had a visit to the clinic or pharmacy for more than 12 months for patients in care and for 6 months in patients who had initiated ART. Only patients with enough follow-up time were included in the LTFU analysis (12 months for pre-ART and 6 months for ART patients). Kaplan–Meier product limit estimates are reported for known deaths and LTFU at 6 and 12 months. Data were analyzed using SAS 9.2 (SAS Institute, Cary, NC), and estimates of mortality corrected for LTFU were derived using a web calculator (http://www.iedea-sa.org) for a recently developed nomogram method by Egger et al.9 This method calculates a corrected 12-month mortality estimate based on the observed 12-month death rate, total number of patients initiating ART, and total number of patients LTFU using outcome data (mortality rates) from tracing studies among sub-Saharan African ART patients who were LTFU.
The study utilized data that are routinely collected by sites for service delivery, which was anonymized for analysis purposes. This study is part of the Identifying Optimal Models of HIV Care and Treatment in Mozambique Collaboration, which was approved by the Mozambican National Ethics Committee, the Columbia University Medical Center Institutional Review Board, the US Centers for Disease Control and Prevention, and the President's Emergency Plan for AIDS Relief's Office of the Global AIDS Coordinator.
Clinic and Program Characteristics
Of the 28 HIV care and treatment clinics included in the analysis, 24 clinics (85.7%) were located in a city/town and 4 (14.3%) in rural settings. There were 12 (42.9%) at primary facilities, 11 (39.3%) at secondary facilities, and 5 (17.9%) at tertiary facilities (Table 1). All facilities offered free HIV care and treatment services and adherence support services, and 24 (85.7%) had PMTCT services at the same facility. Fifteen (53.6%) had CD4 cell count testing available at the facility. Twenty (71.4%) clinics indicated that they had an active outreach program to track patients who missed clinic visits (defaulters). Three-fourths (75%) of the clinics had peer educator services, and approximately half (53.6%) offered nutritional support services.
Adult Patient Characteristics at Enrollment into Care and ART Initiation
Overall, 154,188 patients were enrolled in HIV care at the 28 clinics between 2003 and 2009 (Fig. 1A). Among the 144,024 adults, 94,261 (65.4%) were women, and they were significantly younger than men (median age; 30.0 years vs. 36.0 years, respectively, P value <0.001) (Table 2). Eleven percent of women were documented as being pregnant at enrollment, increasing from 2.4% in 2005 to 6.4% in 2006 and remaining constant at around 13% in subsequent years. It also varied greatly by site, with a median proportion of 12.4% [interquartile range (IQR): 3.4%–18.6%]. Approximately, one-third (32.4%) were referred to HIV care from VCT services and 8.2% entered through PMTCT, with 37.3% having unknown point of entry. The proportion of patients referred from VCT was stable over time, whereas the proportion referred PMTCT increased from 1.3% in 2005 to 5.3% in 2006 and then remained constant at around 11% in subsequent years. Only 0.2% of adults had a documented history of having received prior ART at other clinics. Forty-eight percent of the adult patients reported being married and 31.2% being single, and this information was missing for 14.1% of patients, with the proportion who reported being single significantly higher among women compared with men (33.6% vs. 26.7%, P value <0.001). Almost half of adult patients had primary education, and 17.5% of women and 25.7% of men had 8 or more years of education (P value <0.001), however, data were missing on years of education for 24.2% of patients. Among patients for whom HIV disease stage data were available (WHO stage or CD4 cell count), women entered care with less advanced disease stage than men, with a lower proportion with WHO stage III or IV (31.3% vs. 39.4%, P value <0.001), CD4 count <200 cells per microliter (20.0% vs. 25.9%, P value <0.001) or tuberculosis (TB) treatment at enrollment (2.8% vs. 5.5%, P value <0.001). However, both WHO stage and CD4 cell count measurements were missing at enrollment into care for a high proportion of adult patients (32.7% and 43.9%, respectively), and overall 22.2% were missing both WHO staging and CD4 cell count. The proportion missing both declined considerably over time (from 52.9% in 2004 to 33.4% in 2005), stabilizing after 2006 to 15%–17%. The median proportion missing both WHO stage and CD4 cell count by site was 15.8% (IQR: 12.5%–20.5%). Among those who had a CD4 count, the median CD4 count at enrollment in care was 262 (IQR: 123–452) cells per microliter.
Of those enrolled in care, 51,269 (35.6%) of adults started ART by December 2009, of which 62.8% were women (Table 3). The number of patients who started ART over time is represented in Figure 2B. There was a rapid increase in the number of patients starting ART in 2005, and then the number stabilized by 2007. At ART initiation, women were significantly younger than men (median age; 32.5 vs. 38.0, P value <0.001), and 6.9% of the women were documented as pregnant at ART initiation. The age distributions are shown in Figure 3A for the overall population (including the percentage of females that were pregnant at ART initiation). WHO stage and CD4 cell count measurements at ART initiation were missing for 33.7% and 35.1% of adults, respectively, and 21.2% were missing both in the database. Of those whose ART eligibility could be determined at the time of enrollment in care [74,756 (51.9%) of all patients], 43,695 (58.5%) were eligible to start ART according to national guidelines. Among them, the proportion initiating ART within 3 months of ART eligibility was 73.6% [95% confidence interval (CI): 73.1% to 74.1%] [median time: 34 days (IQR: 15–67) between ART eligibility and ART initiation]. Among those who had a CD4 count, the median CD4 count at ART initiation was 144 (IQR: 69–221) cells per microliter, and 35.1% had a CD4 count <100 cells per microliter. Fourteen percent had WHO stage IV at ART initiation. The proportion documented to be receiving TB treatment at ART initiation was lower in women (4.5% vs. 8.0%, P value <0.001). The initial ART regimens for adult patients were D4T + 3TC + NVP (80.8%), D4T + 3TC + EFV (10.6%), AZT + 3TC + NVP (6.9%) among others (1.7%). Of those who started ART, 2.9% were known to have died within 6 months, and 23.4% were LTFU (Table 4).
Of those patients who were eligible for ART at enrollment in care but who were not initiated on ART, 7.3% (95% CI: 6.5 to 8.1%) were documented to have died [median most recent CD4 = 66 cells/μL (IQR, 22-144.5)] and 81.1% (95% CI: 80.4% to 82.0%) were LTFU at 12 months [median most recent CD4 = 113 cells/μL (IQR: 42–192.5)]. Of those with unknown eligibility at enrollment due to missing information, 81.3% did not start ART by December 2009, of which 1.3% (95%CI: 1.2% to 1.5%) were documented to have died, and 53.9% (95% CI: 53.4% to 54.4%) were LTFU at 12 months.
Pediatric Patients at Enrollment into Care and ART Initiation
Of the total population enrolled in care, 10,164 (6.6%) were children <15 years with a median age of 2 years (IQR: 0.9–5.0) and a similar distribution of females and males (52.3% and 47.7%, respectively) (Table 5). A total of 5,398 children (53.1%) had WHO stage at enrollment; of those, 38.7% were WHO stage III and 8.0% were WHO stage IV. In children ≥5 years old with a CD4 count at enrollment (61.2%), 24.4% had a CD4 count <200 cells per microliter. The results of CD4 percentage were missing for 61.1% of pediatric patients <5 years old. Of children <5 years old with CD4 percentage at enrollment, 35.0% had severe immunosuppression (CD4% <15%). Among those who had a weight measurement at enrollment (50.4%), 35.3% were severely malnourished (weight-for-age Z score <−3). Finally, 1.9% of children were documented as being on TB treatment at enrollment into care.
Among those enrolled in care during 2003–2009, 3745 (36.8%) started ART, with an equal distribution of females and males (50.7% and 49.3%, respectively) (Table 2). The median age at ART initiation was 2.7 (IQR: 1.3–6.1) years, and the age distributions of pediatric patients (Fig. 3B) show that approximately one-third were <2 years old. Of those with a WHO stage at ART initiation (55.8%), 1071 (51.2%) children were WHO stage III and 273 (13%) were WHO stage IV. In children younger than 5 years, who had a CD4 percentage result (53.7%), more than half were severely immunosuppressed. The initial ART regimens for pediatric patients were D4T + 3TC + NVP (54.7%), AZT + 3TC + NVP (28.1%), and D4T + 3TC + EFV (6.9%) among others. Among those who had a weight measurement at ART initiation (71.6%), 30.0% were severely malnourished (weight-for-age Z score <−3). Among the children who initiated ART, 6.5% (95% CI: 5.6% to 7.5%) were known to have died, and 20.5% (95% CI: 19.1% to 22.0%) were LTFU at 12 months (Table 4). Among those enrolled in HIV care who did not initiate ART, 6.3% (95% CI: 5.5% to 7.1%) were known to have died at 12 months [14.1% (95% CI: 11.3% to 17.4%) among those known to be eligible for ART] and 54.1% (95% CI: 55.6% to 52.6%) were LTFU at 12 months [71.3% (95% CI: 68.1% to 74.3%) among those known to be eligible for ART].
Corrected Mortality Estimates
The estimates of mortality at 12 months after ART initiation, corrected for unascertained mortality among those lost to follow-up,9 were 13.1% (IQR: 11.5%–16.0%) for adults and 13.5% (IQR: 11.4%–17.4%) for children.
Analysis of routinely collected data from a subset of HIV care and treatment clinics in Mozambique documents the rapid scale-up HIV care and treatment services in the country during the 2003–2009 period. The majority of adult patients enrolled in care and those who initiated ART were women (65.4% and 62.8%, respectively), a slightly higher proportion than in the larger HIV epidemic in Mozambique (57%).10 This may reflect higher testing and referral rates for women in Mozambique. For example, at least 10.5% of women included in this analysis were pregnant at enrollment, likely reflecting availability of PMTCT programs and linkages between PMTCT programs and HIV care. Children were also enrolled and initiated on ART, however, only a small proportion of children initiating ART (15.5%) were <1 year old. This is likely due to very high death rates, only recent availability of early infant diagnostic tests, and enrollment before new guidelines which recommend that all HIV-infected infants younger than 1 year be initiated on ART irrespective of HIV disease stage.11 Additionally, a high proportion of children were severely malnourished at enrollment and ART initiation.
We observed high rates of nonretention among ART adult patients at 12 months (27%), although they were similar to those described in a systematic review of 39 cohorts from sub-Saharan Africa [median attrition at 12 months was 23% (range 7%–45%)].12 Among children on ART, 21% were LTFU at 12 months, which was higher than those reported in other cohorts.13,14 High rates of LTFU were observed among patients enrolled in care who had not initiated ART, especially among a substantial proportion of patients who despite being ART eligible at enrollment were not initiated on ART. The high proportion of 1-year LTFU in this group (81% in adults and 71% in children) may have hindered initiation of ART and could also represent a substantial number of unascertained deaths.15–17 For example, a recent study conducted in South Africa with 44,844 patients enrolled in HIV care observed that 23% of eligible patients died before starting ART with a median time to death of 92 (IQR: 33–216) days.18 Some of the LTFU could also be due to overburdened clinics (due to waiting times and prioritization of ART patients to receive adherence support and outreach services), undocumented transfer of patients to other facilities,19 or due to patients not completing the nonclinical national requirements to start ART (ie, failing to attend 3 ART readiness pretreatment counseling sessions and disclosing HIV status to a treatment partner). Our findings around LTFU underscore the importance of conducting active tracing of all patients enrolled in HIV care to engage pre-ART patients in HIV care to help ensure timely ART initiation20 and reduce the risk of mortality and morbidity.21,22 The recent changes in ART initiation guidelines in Mozambique [CD4 count <250 cells/μL (vs. the previous <200 cells/μL), WHO stage IV and WHO stage III, and CD4 count <350 cells/μL] will likely result in a 10% increase in the number of eligible patients at enrollment into care. Nevertheless, additional efforts at the community and clinic levels will be needed to ensure that this larger population of ART eligible patients do in fact start ART earlier.
Among those who initiated ART, we also observed a relatively high proportion of patients initiating ART at advanced stages of HIV disease (CD4 count <100 cells/μL or WHO stage IV), although 21% had missing WHO stage and CD4 count at ART initiation. Late ART initiation has been associated with high early mortality rates in sub-Saharan Africa.22,23 A cross-sectional study from a clinic in rural Uganda observed that 40% of the 2,311 patients initiating ART were late presenters and found that male sex, lower education level, and unemployment, among other factors, were associated with a higher likelihood of late ART initiation.24 A recent analysis of 268 HIV care clinics in sub-Saharan Africa found both program level factors (provider to patient ratio, adherence support services, linkages with PMTCT programs, and outreach services) and AIDS knowledge and testing coverage in the general population to be important determinants of low median CD4 counts in populations of persons initiating ART.20 In addition to diminishing the successes of scale-up and its future potential to save lives, the problem of late ART initiation also adds significant burden and costs to care,25 and missed opportunities for secondary HIV prevention due to late diagnosis.26–29 It is therefore critical to understand more about the upstream determinants of late ART initiation (eg, determinants of HIV testing and late diagnosis among HIV-positive persons).
The 12-month reported mortality rate of 5% of adults and 7% of children who initiated ART in this study are very high, and almost certainly underestimated due to high numbers of persons LTFU, among whom death rates have been shown to be quite high in sub-Saharan African scale-up programs.19,21 Application of a recently described method which attempts to correct mortality estimates for LTFU suggest that the true 12-month mortality rate among patients initiating ART in our study population is probably closer to 13% and 14%, for adults and children, respectively. Taken together, these data underscore that efforts are urgently needed in Mozambique to identify and enroll patients at earlier stages of HIV infection.
The strengths of our study include the fact that the data were derived from routine clinical care from multiple HIV care and treatment clinics from a variety of facility types and from various regions in the country. It also includes information on adult and pediatric patients on ART and those in HIV care. This analysis included 55,014 patients from 28 sites in 5 Mozambican provinces, which represent one-fourth of the approximately 229,000 patients who ever initiated ART in Mozambique across more than 200 sites.6,30
The study also has limitations that motivate caution in interpretation of the findings. First, the clinics included in this analysis were selected for establishment of electronic database and are therefore not representative of all clinics providing HIV care and treatment in Mozambique. Second, some children <18 months who had not yet initiated ART may have been entered into the database with unknown HIV infection status, overestimating the number of HIV-infected children who were receiving care at these sites. Finally, an important limitation is the quality of the data that were derived from service programs including a large proportion of missing date and inability to validate the information. Despite periodic data quality assessments, there was a high proportion of patients with missing data on WHO stage (32.7% and 33.7%) and CD4 cell count (43.9% and 35.1%) at enrollment in HIV care and ART initiation, respectively. This precluded us from assessing their ART eligibility status, among other things. The reasons for these missing data may include the following: lack of clinical staging, poor documentation of information in the medical chart, poor data entry, lack of CD4 cell count testing (including reagent shortages or machine downtime), or some combination of these factors. Although missing data are a common challenge in large scale service delivery programs, it nonetheless limits the ability to accurately determine a variety of estimates and examine determinants of key outcomes,31 requiring the use of missing data methods such as multiple imputation32 (for an example of this technique for baseline CD4 cell count in HIV scale-up). Focused efforts to improve data quality are ongoing and will be critical to provide further insights into HIV care and treatment scale-up in the future. The importance of obtaining complete information during clinical assessments and documentation of all findings in medical records must be emphasized during the training of health care workers.
In conclusion, our analysis documents the success in the enrollment of large numbers of patient in HIV care and on ART. However, several challenges remain. These include high rates of loss to follow-up and death and initiation of ART in the advanced stages of HIV disease among both adults and children. Further efforts are also needed to ensure that accurate and complete data are obtained on all individuals enrolled in the HIV programs. When possible, future analyses of HIV scale-up programs should include an analysis of patients from the pre-ART phase of care, even if key clinical data may be missing on a large proportion of patients.
We are grateful to all patients and staff at the HIV care and treatment clinics included in this analysis and to Carla Xavier, of ICAP Mozambique, who designed the Electronic Patient Tracking System used by the clinics included in this analysis, and ICAP Clinical Officers who mentor site providers to deliver the services described in this analysis. We would also like to thank Lisa Nelson, from the US Center for Disease Control and Prevention, Mozambique for technical support and for reviewing this article.
1. WHO. Towards Universal Access: Scaling up Priority HIV/AIDS Interventions in the Health Sector: Progress Report 2010. 2010 Geneva, Switzerland WHO
2. Institute of Medicine. Scaling up Treatment for the Global AIDS Pandemic. 2005 Washington, DC The National Academies Press
3. The Global Fund to Fight AIDS, Tuberculosis and Malaria. Framework for Operations and Implementation Research in Health and Disease Control Programs. 2008 Geneva, Switzerland The Global Fund to Fight AIDS, Tuberculosis and Malaria
4. WHO. HIV Testing, Treatment and Prevention: Generic Tools for Operational Research. 2009 Geneva, Switzerland WHO
5. Mozambique Ministry of Health (MISAU). INSIDA, Inquérito Nacional de Prevalência, Riscos Comportamentais e Informação sobre o HIV e SIDA em Moçambique (National Seroprevalence, Behavioral Risk and HIV/AIDS Information Survey. Preliminary Results. 2010 Mozambique Ministry of Health
6. United Nations General Assembly Special Session on HIV and AIDS. Mozambique Progress Report 2008–2009. March 2010 UNAIDS
7. CDC. . Revised classification system for human immunodeficiency virus infection. MMWR Recomm Rep. 1994;43(No. RR-2):1–10
8. Kuczmarski R, Ogden C, Guo S, et al. 2000 CDC growth charts for the United States: methods and development. Vital Health Stat. 2002;11:1–203
9. Egger M, Spyche B, Sidle J, et al. Correcting mortality for loss to follow up: a nomogram applied to antiretroviral treatment programmes in sub-Saharan Africa. PLoS Med. 2011;8:e1000390
10. UNAIDS/WHO. Epidemiological Fact Sheets on HIV and AIDS, 2008 Update. 2008 World Health Organization
11. WHO. Antiretroviral Therapy for HIV Infection in Infants and Children: Towards Universal Access: Recommendations for a Public Health Approach. 2010 World Health Organization
12. Fox MP, Rosen S. Patient retention in antiretroviral therapy programs up to three years on treatment in sub-Saharan Africa, 2007 –2009: systematic review. Trop Med Int Health. 2010;15(suppl 1):1–15
13. Arrive E, Kyabayinze DJ, Marquis B, et al. Cohort profile: the paediatric antiretroviral treatment programmes in lower-income countries (KIDS-ART-LINC) collaboration. Int J Epidemiol. 2008;37:474–480
14. Davies MA, Keiser O, Technau K, et al. Outcomes of the South African National Antiretroviral Treatment Programme for children: the IeDEA Southern Africa collaboration. S Afr Med J. 2009;99:730–737
15. Brinkhof MW, Dabis F, Myer L, et al. Early loss of HIV-infected patients on potent antiretroviral therapy programmes in lower-income countries. Bull World Health Organ. 2008;86:559–567
16. Geng EH, Emenyonu N, Bwana MB, et al. Sampling-based approach to determining outcomes of patients lost to follow-up in antiretroviral therapy scale-up programs in Africa. JAMA. 2008;300:506–507
17. Geng EH, Nash D, Kambugu A, et al. Retention in care among HIV-infected patients in resource-limited settings: emerging insights and new directions. Curr HIV/AIDS Rep. 2010;7:234–244
18. Ingle SM, May M, Uebel K, et al. Outcomes in patients waiting for antiretroviral treatment in the Free State Province, South Africa: prospective linkage study. AIDS. 2010;24:2717–2725
19. Geng EH, Nash D, Kambugu A, et al. Retention in care among HIV-infected patients in resource-limited settings: emerging insights and new directions. Curr HIV/AIDS Rep. 2010;7:234–244
20. Nash D, Wu Y, Elul B, et al. Program and contextual-level determinants of low median CD4+ cell count in cohorts of persons initiating ART in 8 sub-Saharan African countries. AIDS. 2011:25 E-pub ahead of print (doi:10.1097/QAD.1090b1013e32834811b32834812)
21. Brinkhof MW, Pujades-Rodriguez M, Egger M. Mortality of patients lost to follow-up in antiretroviral treatment programmes in resource-limited settings: systematic review and meta-analysis. PLoS One. 2009;4:e5790
22. Lawn SD, Harries AD, Anglaret X, et al. Early mortality among adults accessing antiretroviral treatment programmes in sub-Saharan Africa. AIDS. 2008;22:1897–1908
23. Bekker LG, Myer L, Orrell C, et al. Rapid scale-up of a community-based HIV treatment service: programme performance over 3 consecutive years in Guguletu, South Africa. S Afr Med J. 2006;96:315–320
24. Kigozi IM, Dobkin LM, Martin JN, et al. Late-disease stage at presentation to an HIV clinic in the era of free antiretroviral therapy in Sub-Saharan Africa. J Acquir Immune Defic Syndr. 2009;52:282–289
25. Krentz HB, Auld MC, Gill MJ. The high cost of medical care for patients who present late (CD4 <200 cells/microL) with HIV infection. HIV Med. 2004;5:93–98
26. Branson BM, Handsfield HH, Lampe MA, et al. Revised recommendations for HIV testing of adults, adolescents, and pregnant women in health-care settings MMWR Recomm Rep. 2006;55(RR-14):1–17 quiz CE11–CE14
27. Bunnell R, Mermin J, De Cock KM. HIV prevention for a threatened continent: implementing positive prevention in Africa. JAMA. 2006;296:855–858
28. Cohen MS, Hellmann N, Levy JA, et al. The spread, treatment, and prevention of HIV-1: evolution of a global pandemic. J Clin Invest. 2008;118:1244–1254
29. WHO. Towards Universal Access: Scaling up Priority HIV/AIDS Interventions in the Health Sector: Progress Report 2009. 2009 Geneva, Switzerland WHO
31. Greenland S, Finkle WD. A critical look at methods for handling missing covariates in epidemiologic regression analyses. Am J Epidemiol. 1995;142:1255–1264
32. Nash D, Katyal M, Brinkhof MW, et al. Long-term immunologic response to antiretroviral therapy in low-income countries: a collaborative analysis of prospective studies. AIDS. 2008;22:2291–2302
Keywords:© 2011 Lippincott Williams & Wilkins, Inc.
antiretroviral treatment; HIV/AIDS; HIV care; implementation science; Mozambique; operations research; PEPFAR; scale-up