Across sub-Saharan Africa, countries are responding to generalized HIV/AIDS epidemics, particularly through the scale-up of antiretroviral treatment (ART) to eligible patients. Malawi, estimated to have approximately 1 million people currently infected with HIV and 170,000 in need of ART,1 has made tremendous progress toward its goal of universal access to ART by 2010, having initiated ART in more than 57,000 patients in the public sector by June 30, 2006 (unpublished data, HIV Unit, Ministry of Health, Malawi, 2006).
High early mortality has been observed in a number of ART programs in sub-Saharan Africa, including Malawi.1-5 The reasons for high early mortality in patients starting ART are not completely understood, and effective interventions to decrease these death rates have not yet been tested.
Cotrimoxazole (CTX) prophylaxis is associated with a 25% to 46% reduction in mortality for HIV-infected individuals in sub-Saharan Africa who are not on ART, even in areas with high bacterial resistance to the antibiotic.5-18 CTX prophylaxis also decreases the frequency of clinic visits and hospitalizations, reduces weight loss, may slow the decline in CD4+ lymphocyte counts and the rise in HIV viral loads, and is cost-effective.5-19 Efficacy seems to be maintained over the 1- to 2-year follow-up periods of the studies.6,15 CTX prophylaxis in Africa has few adverse reactions.6,15 In summary, CTX is a safe, inexpensive, and readily available antimicrobial agent that can extend and improve the quality of life of HIV-infected individuals.
In developed countries, data have clearly demonstrated the effectiveness of CTX at reducing morbidity and mortality in HIV-infected individuals, including those on ART, particularly from Pneumocystis jiroveci (formerly Pneumocystis carinii).20 Unique factors in sub-Saharan Africa, however, warrant continued evaluation of CTX for HIV-infected individuals on ART: the ongoing scale-up of ART, differences in prevalent opportunistic infections such as malaria and diarrheal and respiratory diseases that may be affected by CTX, and concern about resistance to CTX in settings with limited therapeutic options.
The Malawi Ministry of Health established clinical guidelines for CTX prophylaxis in July 2005.21 These guidelines stipulate that the CTX should be offered to the following HIV-positive adults (aged 13 years and older): all persons with symptomatic HIV disease (World Health Organization [WHO] stages II, III, and IV), all persons who have a CD4 lymphocyte count <500 cells/mm3 regardless of symptoms, and pregnant women after the first trimester who are symptomatic or have a CD4 count <500 cells/mm3. For various reasons, however, including a lack of adequate CTX supplies, scale-up plan, and operational training, health facilities have been slow in implementing this policy.
To inform and provide momentum to the scale-up of CTX prophylaxis for patients receiving ART in Malawi, we undertook an evaluation of the impact of CTX prophylaxis on the 6-month mortality of patients receiving ART and factors associated with patient survival. To do so, we conducted a retrospective cohort study in a programmatic setting.
Available Routine Program Data
The Malawi Ministry of Health began providing ART in June 2004, using an approach similar to a tuberculosis (TB) directly observed therapy-short course (DOTS) model, employing case finding, provision of free standardized drug regimens, and routine quarterly clinical monitoring of standard treatment outcomes.22 The first-line regimen is a fixed-dose combination of stavudine, lamivudine, and nevirapine, with zidovudine substituted for stavudine and efavirenz substituted for nevirapine in case of adverse drug reactions. Patient registers and individual patient (master) cards are used at every ART site in the country. Patients are registered on the day of ART initiation, whereas follow-up information, including primary (eg, alive, dead, default) and secondary (eg, pill count, side effects) outcome measures, is documented on the master cards, usually on a monthly basis. Follow-up intervals are sometimes extended in instances in which patients are determined to be clinically stable and adherent to ART for more than 6 months.
Baseline demographic and clinical information that is recorded in ART monitoring tools includes age at the time of ART initiation, gender, and reason for ART, including WHO clinical stage III or IV or CD4 count <200 cells/mm3. Routine laboratory monitoring, including CD4 testing, is not a mandatory requirement according to national ART guidelines. Some patients undergo CD4 testing to determine ART eligibility, however, particularly in instances in which WHO clinical staging is uncertain or patients are in WHO clinical stage II.23 In instances in which a CD4 count <200 cells/mm3 is the indication for ART initiation, CD4 results are not routinely documented on the patient master card. Treatment outcomes are routinely updated by ART clinic staff and in preparation for quarterly supervisory visits by the Ministry of Health (MOH) HIV/AIDS Unit, which have been previously described. Patients who have not returned for follow-up evaluation or medication refill for more than 90 days are considered to have defaulted. When patients receiving ART are determined to have died, defaulted, stopped treatment, or transferred out to another ART clinic, the date of the event is ascertained and recorded in the ART register. Most ART sites have no formal patient tracing systems to determine the specific outcome of default patients.
Site Selection and Data Collection
In mid-2005, 5 public sector mission hospitals in Malawi that were providing ART began providing CTX to all patients receiving ART. These sites are hereafter referred to as “CTX sites.” All these CTX sites were initiating approximately 25 new patients on ART per month, were rural or periurban, and received no or limited external financial or technical support. We identified 6 comparison ART sites, hereafter referred to as “non-CTX sites,” that had not implemented CTX prophylaxis but were similar to the CTX sites in the number of patients on ART, time of initiating an ART program, type of facility (all were mission hospitals), and geographic location (Table 1). In May and June 2006, data were abstracted from the medical records of patients at these 11 sites, and outcomes were compared.
Records were abstracted for patients who initiated ART after June 1, 2005, and no later than 6 months before the date of abstraction. Abstracted data elements included ART site name; indication for ART, WHO clinical stage III or IV, or CD4 count <200 cells/mm3 (actual figures were not available); whether active TB disease or a history of TB influenced clinical eligibility for treatment; date of ART initiation; patient age at ART initiation; gender; date of CTX initiation if applicable; clinical status 6 months after ART initiation, which included “alive and on ART,” dead, defaulted, stopped ART completely, or transferred out to another ART facility; and the dates of outcomes for patients who were not alive and on ART.
Data were entered into a Microsoft Access (Microsoft Corp., Seattle, WA) database using Epi-Info (version 3.4; CDC, Atlanta, GA). Simple and standardized data collection tools facilitated reliable data abstraction. The accuracy of the data was verified by entering 25% of the abstracted patient records a second time. The error rate was <0.1% of fields, and no systematic errors in keying were found.
Statistical analyses were performed using SAS (version 9.1; SAS Institute, Cary, NC). Differences in mortality risk and reduction between the CTX and non-CTX groups at 6 months were assessed using relative risk reduction (RRR). Kaplan-Meier and life-table survival methods were used to examine survival probabilities for the CTX and non-CTX groups across the 6-month assessment period; differences in survival were evaluated using the log-rank test. Additionally, a Cox regression model, with adjustments for age when ART was initiated, gender, active or prior TB at ART initiation, and indication for ART (including WHO clinical stage III or IV and CD4 count <200 cells/mm3), was fitted to the data to test for differences in survival between the 2 groups.
Because the status of default patients (ie, whether they were alive and on ART, dead, or had another treatment outcome) was impossible to ascertain, outcomes were initially analyzed with all defaults excluded. In addition, analyses were performed while treating all 100% of defaults as deaths in one group and, at the same time, reclassifying a range of percentages of default patients who may actually represent deaths in the other group, and vice versa. Moreover, because we were concerned with the accuracy of the recorded dates of adverse events (ie, death, lost to follow-up, stopped treatment), we used the life-table method to assess interval-based mortality over the 6-month assessment period. Given the possibility that the recorded dates for some adverse events, particularly death and default, may be inaccurate by a few days, treating event times as occurring within a discrete monthly interval rather than on an exact date is a more conservative method.
Data for patients who were children <15 years old (considered pediatric in the ART program), transferred out of treatment, were initiated on CTX at non-CTX sites or not initiated on CTX at CTX sites, were initiated on CTX after 14 days (or within 30 days of initiating ART), or had unverifiable outcomes were excluded from the analyses. These exclusions were made to increase the likelihood that differences in event status between the 2 groups were attributable to the intervention and not to some other factor.
The evaluation was exempted from ethical review by the National Health Sciences Research Committee in Malawi because this study was designed for program monitoring and evaluation. The evaluation was also determined to be program evaluation by the US Centers for Disease Control and Prevention.
A total of 1295 records were abstracted; after exclusions were made based on the previously discussed criteria, 1052 patient records (574 from CTX sites and 478 from non-CTX sites) were analyzed, including 6-month outcomes for all patients. Overall, 83 records were excluded because of a lengthy ART CTX start interval, 129 records because of CTX being given at non-CTX sites, 25 patients <15 years old, 5 patients not given CTX at CTX sites, and 1 patient with an unverified outcome. Patient age at ART initiation ranged from 15 to 71 years (mean of 38 years for patients at CTX sites and 37 years for patients at non-CTX sites; Table 2). Most patients were female (62% in both groups). The CTX group had a significantly higher mean age, percentage of patients with a CD4 count <200 cells/mm3 as the indication for ART, and percentage of patients with prior or active TB at the time of ART initiation (see Table 2). The distribution of patients according to clinical status of patients on ART at 6 months differed between CTX and non-CTX groups (see Table 2).
Mortality With Default Patients Excluded
Mortality at 6 months was first examined among 941 patients with outcomes of alive and on ART, dead, and stopped treatment; patients with an outcome of “default” were excluded. In this analysis, mortality at 6 months was 10.7% in the CTX group compared with 18% in the non-CTX group. The 6-month mortality risk reduction was 40.7% (P = 0.0013) for patients receiving ART at clinics offering CTX prophylaxis as compared with patients receiving ART at clinics not offering CTX. A stratified analysis showed that the difference in mortality between the CTX and non-CTX groups remained even after adjusting for ART indication (P = 0.007). Similarly, when patients who transferred out were included (n = 979), the analysis showed a significantly lower mortality between patients receiving ART at clinics offering CTX prophylaxis compared with clinics not offering CTX prophylaxis (P = 0.0020).
Mortality With Default Patients Included
Because death is a possible cause of default, sensitivity analyses were performed varying the proportion of defaults that were considered to be deaths. Given that the proportion of defaults attributable to death may have differed between the 2 groups, we also conducted analyses assuming 100% of defaults as death in the first group with varying percentages of defaults as deaths in the second group, and vice versa. Mortality differences between the CTX and non-CTX groups were noted in all scenarios of default as death, except in instances when 100% of defaults were considered as deaths in the CTX group and 0% or 10% of defaults were considered as deaths in the non-CTX group (Table 3). Similarly, when patients who transferred out were included in the analysis, the association did not change.
Survival estimates based on the life-table method did not differ substantially from the results based on the Kaplan-Meier method and Cox regression; the results from the Kaplan-Meier and Cox methods are reported. Kaplan-Meier survival curves for patients receiving CTX and patients not receiving CTX were significantly different (P = 0.002) when all default patients were excluded from analysis (Fig. 1) and when defaults were included. When all defaults were excluded from the analysis, mortality differences were noted at approximately 40 days after ART initiation.
The Cox model demonstrated that the lower mortality of patients on ART at sites providing CTX remained statistically significant after controlling for age when ART was initiated, gender, active or prior TB at ART initiation, and indication for ART (including WHO clinical stage III or IV and CD4 count <200 cells/mm3; Table 4). Throughout the 6-month follow-up period, male patients receiving ART were approximately 1.5 times more likely to die than female patients receiving ART. Patients with WHO stage IV disease when they started ART were 2 times more likely to die at any point in time compared with those with WHO stage III disease. TB status, age, and CD4 count <200 cells/mm3 when ART was initiated were not associated with an increased risk of death.
The major finding of this study was that patients receiving ART at clinics offering CTX prophylaxis had 40.7% lower mortality during the first 6 months of ART as compared with patients on ART at comparable clinics not offering CTX prophylaxis in Malawi. This finding builds on existing data demonstrating that CTX leads to substantial mortality reductions in HIV-infected individuals not on ART. In Malawi and other resource-limited settings, CTX prophylaxis may have the potential to lengthen the lives of patients on ART because CTX is readily available and comparatively inexpensive and, as this study demonstrates, may significantly reduce patient mortality, at least during the first 6 months of ART.
Many countries in sub-Saharan Africa are successfully scaling up ART on a national level. Currently, however, in many countries, CTX prophylaxis is viewed primarily as a component of non-ART HIV care and as a means of bridging patients who are ineligible for ART into treatment programs. Although some countries, such as Uganda, have achieved high-level coverage of CTX in non-ART and ART settings, the scale-up of such interventions continues to be a challenge elsewhere in sub-Saharan Africa. Delay in the dissemination of national CTX guidelines, problems with drug procurement and supply, lack of discrete health care infrastructure for non-ART HIV care, and persistent concerns about drug resistance have limited the implementation of scale-up. These data confirm that CTX prophylaxis should be encouraged not only in the context of non-ART HIV care but as a basic intervention that should be prioritized within ART programs. Increasingly, ART programs have ample infrastructure, drug procurement and supply chains, and monitoring and evaluation systems that can facilitate the scale-up of CTX prophylaxis to patients receiving ART. Ideally, scale-up within ART programs should also augment the capacity of non-ART HIV care programs and coverage of CTX to HIV-infected individuals not yet eligible for ART.
Limitations of the evaluation primarily stem from the retrospective design. Additionally, we lacked laboratory data on immune function that would have helped us to address potential confounding by differences in the immunologic characteristics of the CTX and non-CTX groups. Because the CTX and non-CTX groups received care at different sites, it is possible that the mortality reduction found in the CTX group was attributable to higher quality care, despite the fact that CTX and non-CTX sites shared several important characteristics. Differences were noted between CTX and non-CTX groups in terms of age, TB status, and indication for ART. Age distributions in both groups were consistent with the demographics of the HIV epidemic and national ART reports in Malawi. The age differences and potential for age-associated variation in morbidity and mortality are not likely to have had a significant effect on the results, because the mean age of the CTX group was only 1 year older than that of the non-CTX group.
It is difficult to compare the quality of ART provision at CTX and non-CTX sites directly because of limited outcome data on patients receiving ART available through routine data collection other than mortality, which was the primary outcome of interest in this study. Site level data regarding adherence in patients receiving ART were, however, available for quarter 3 (July 1 to September 30, 2005) and quarter 4 (October 1 to December 31, 2005). These quarterly cohort data show that the CTX and non-CTX groups had pill counts recorded for 93% and 87% of alive patients, respectively. Furthermore, 97% of patients receiving CTX were determined to have ART adherence >95% compared with 96% in the non-CTX group.
Variation in the prevalence of opportunistic infections, such as TB, could have led to differences in mortality rates between the 2 groups. A higher percentage of patients on ART at CTX sites (15%) were found to have had active or prior TB as part of the WHO clinical staging that determined eligibility for ART compared with non-CTX sites (9%). In the Cox regression model, however, TB status was not associated with risk of death in the 6-month assessment period.
Data recently published from Uganda found that HIV-infected adults given ART and CTX had a 64% lower incidence of malaria compared with those given CTX alone, although mortality was not assessed.23 Differences in the incidence of malaria between the 2 groups in this study were unlikely to have contributed to the mortality reductions noted, because ART sites were in similar geographic areas and malaria is a relatively uncommon cause of death among HIV-infected adults in malaria-endemic areas.24
A CD4 count <200 cells/mm3 was a more common indication for ART in the CTX group. According to the same national quarterly report, the distribution of indications for ART initiation in Malawi was as follows: WHO stage III (68%), WHO stage IV (24%), and CD4 count <200 cells/mm3 (8%). Thus, the CTX group included a higher percentage (17%) of patients whose ART eligibility was determined by CD4 cell count, whereas the non-CTX group had a rate (7%) comparable with the national average. This may point to a general difference in the quality of care provided by CTX sites. A CD4 count <200 cells/mm3 as an indication for ART was not, however, associated with risk of death in the proportional hazard model.
Malawi has made significant progress toward scale-up of CTX since these findings were shared with stake holders last year. In August 2006, a national implementation and scale-up plan was disseminated. The plan called for a massive increase in the procurement of CTX to supply all ART sites in the country by mid-2007. Drug forecasting and orders have already been completed at the site level based on the number of patients alive on ART by the end of 2006 and the number of new patients anticipated to access treatment in 2007. CTX distribution is to include convenient pill containers of 120 tablets per bottle to allow for 2-month dispensing. The ART patient register and the ART treatment master cards have been modified to include monitoring of CTX, and systems have also been set up to monitor the use of CTX in patients who may be started on CTX outside the ART clinic. All site staff were briefed on the evaluation findings at clinical refresher trainings in 2006, and additional in-service training is to be included in the routine second-quarter 2007 supervision.
In summary, our analysis demonstrated a significant reduction in 6-month mortality for patients receiving ART at clinics offering CTX prophylaxis. These findings support prioritization of expanded coverage of CTX within ART programs in sub-Saharan Africa, in synergy with broader efforts to increase coverage of CTX prophylaxis among all symptomatic HIV-infected individuals not yet eligible for or unable to access ART.
The authors thank the patients receiving ART and clinic staff who were involved with the evaluation and Dr. Scott Filler for his helpful review of the manuscript.
1. National AIDS Commission. National Estimate of HIV/AIDS in Malawi in 2003
. Government of Malawi; Lilongwe, Malawi; 2003.
2. Lawn SD, Myer L, Orrell C, et al. Early mortality among adults accessing a community-based antiretroviral service in South Africa: implications for programme design. AIDS
3. Wools-Kaloustian K, Kimaiyo S, Diero L, et al. Viability and effectiveness of large-scale HIV treatment initiatives in sub-Saharan Africa: experience from western Kenya. AIDS
4. Stringer JSA, Zulu I, Levy J, et al. Rapid scale-up of antiretroviral therapy at primary care sites in Zambia. JAMA
5. ART-LINC Collaboration and ART Cohort Collaboration Groups. Mortality of HIV-infected patients in the first year of antiretroviral therapy: comparison between low-income and high-income countries. Lancet
6. Wiktor SZ, Sassan-Morokro M, Grant AD, et al. Efficacy of trimethoprim-sulphamethoxazole prophylaxis to decrease morbidity and mortality in HIV-1-infected patients with tuberculosis in Abidjan, Cote d'Ivoire: a randomised controlled trial. Lancet
7. Badri M, Maartens G, Wood R, et al. Co-trimoxazole and HIV-1 infection. Lancet
8. Zachariah R, Spielmann M-PL, Chinji C, et al. Voluntary counselling, HIV testing and adjunctive cotrimoxazole reduces mortality in tuberculosis patients in Thyolo, Malawi. AIDS
9. Zachariah R, Harries AD, Arendt V, et al. Compliance to cotrimoxazole for the prevention of opportunistic infections in HIV infected tuberculosis patients in Thyolo, Malawi. Int J Tuberc Lung Dis
10. Mwaungulu FBD, Floyd S, Crampin AC, et al. Cotrimoxazole prophylaxis reduces mortality in human immunodeficiency virus-positive tuberculosis patients in Karonga District, Malawi. Bull World Health Organ
11. Chimzizi R, Gausi F, Bwanali A, et al. Voluntary counselling, HIV testing and adjunctive cotrimoxazole are associated with improved TB treatment outcomes under routine conditions in Thyolo District, Malawi. Int J Tuberc Lung Dis
12. Chimzizi RB, Harries AD, Manda E, et al. Counselling, HIV testing and adjunctive cotrimoxazole for TB patients in Malawi: from research to routine implementation. Int J Tuberc Lung Dis
13. Mermin J, Lule J, Ekwaru JP, et al. Effect of cotrimoxazole prophylaxis on morbidity, mortality, CD4-cell count, and viral load in HIV infection in rural Uganda. Lancet
14. Grimwade K, Sturm AW, Nunn AJ, et al. Effectiveness of cotrimoxazole prophylaxis on mortality in adults with tuberculosis in rural South Africa. AIDS
15. Chintu C, Bhat GJ, Walker AS, et al. Co-trimoxazole as prophylaxis against opportunistic infections in HIV-infected Zambian children (CHAP): a double-blind randomised placebo-controlled trial. Lancet
16. Creese A, Floyd K, Alban A, et al. Cost effectiveness of HIV/AIDS interventions in Africa; a systematic review of the evidence. Lancet
17. Watera C, Todd J, Muwonge R, et al. Feasibility and effectiveness of cotrimoxazole prophylaxis for HIV-1-infected adults attending an HIV/AIDS clinic in Uganda. J Acquir Immune Defic Syndr
18. Guidelines on Co-Trimoxazole Prophylaxis for HIV Infections Among Children. Adolescents and Adults in Resource-Limited Settings: Recommendations for a Public Health Approach
. World Health Organization, Geneva, Switzerland; 2006.
19. Goldie SJ, Yazdanpanah Y, Yosina E, et al. Cost effectiveness of HIV treatment in resource-poor settings-the case of Cote d'Ivoire. N Engl J Med
20. McNaghten AD, Hanson DL, Jones JL, et al. Effects of antiretroviral therapy and opportunistic infection chemoprophylaxis on survival after AIDS diagnosis. AIDS
21. Malawi National Cotrimoxazole Guidelines
. Malawi Ministry of Health, Lilongwe, Malawi; October 2005.
22. Libamba E, Makombe S, Harries AD, et al. Scaling up antiretroviral therapy in Africa: learning from tuberculosis control programmes-the case of Malawi. Int J Tuberc Lung Dis
23. Mermin J, Ekwanu JP, Liechty CA, et al. Effect of co-trimoxazole prophylaxis, antiretroviral therapy, and insecticide-treated bednets on the frequency of malaria in HIV-1-infected adults in Uganda: a prospective cohort study. Lancet
24. Laufer M, van Oosterhout J, Thesing P, et al. Impact of HIV-associated immunosuppression on malaria infection and disease in Malawi. J Infect Dis