Tuberculosis (TB) continues to be the leading cause of morbidity and mortality among people living with HIV (PLHIV) in sub-Saharan Africa.1,2 Nearly one-third of the more than 30 million PLHIV in the world are estimated to have concomitant latent infection with Mycobacterium tuberculosis.3,4 Additionally, just over 10% of the nearly 9 million people who develop TB each year are PLHIV, approximately equal to 1 million new TB cases among PLHIV each year.5
To mitigate the dual burden of TB and HIV in populations at risk or affected by both diseases, the World Health Organization (WHO) published an interim policy on collaborative TB/HIV activities in 2004.6 The policy was intended to establish and strengthen mechanisms for integrated delivery of TB and HIV services, reduce the burden of TB among PLHIV, initiate early antiretroviral therapy (ART), and reduce the burden of HIV among people with presumptive TB or diagnosed TB. WHO advised its implementation worldwide, depending on the epidemiology of TB and HIV, guided by each country's specific health system context and evidence-based considerations. In 2011, WHO released an ambitious Global Plan to Stop TB, setting a target to screen for TB in 100% of PLHIV who are attending HIV services by 2015.7 This guidance focused on 3 major activities [intensified TB case finding (ICF), isoniazid preventive therapy (IPT), and infection control], which together have been shown to decrease incidence, transmission, and morbidity and mortality from TB in PLHIV.6 After this guidance, many countries developed specific plans to reach this global commitment.
Rwanda is a sub-Saharan African country with a generalized HIV epidemic, having a countrywide prevalence of 3%.8 According to a 2012 WHO report, the incidence of TB disease in Rwanda was estimated at 94 (95% confidence interval: 84 to 105) per 100,000 population and 28% of patients with TB tested HIV positive.9 In 2005, the Rwanda Ministry of Health (MOH) approved and implemented a national policy on TB/HIV collaborative activities. ICF was a focal area of that national plan.
TB screening in PLHIV was conducted through a standardized 5-question checklist, accompanied by appropriate diagnostic follow-up and treatment. The experience of TB screening in PLHIV in a single rural health site in Rwanda has previously been reported.10 In this study, 179 inpatients with HIV were screened for TB; 86 (48%) screened positive, and 39 (22%) were diagnosed with active TB. Similarly, 300 outpatients with HIV were screened for TB; 80 (27%) screened positive, and 11 (4%) were diagnosed with active TB. Since then, Rwanda has engaged in a concerted effort to scale-up TB screening activities in all HIV clinics in the country, where TB screening is done systematically at enrollment of every new PLHIV into care and treatment services and also at every contact with a health-care provider.
We describe trends of ICF in a population of persons newly enrolled in HIV clinics in Rwanda.
We used program data obtained from facility-based pre-ART and ART registers in HIV clinics in Rwanda. The study period spanned 6 years, from 2006 to 2011. Data were collected on a semiannual basis. In all health facilities, TB screening was done through questioning of patients using a 5-point checklist11 adapted to the Rwandan context:10
- cough for 2 weeks or more,
- fever for 3 weeks or more,
- night sweats for 3 weeks or more,
- loss of more than 3 kg in 4 weeks without any other cause, and
- close contact with a pulmonary TB case.
If at least 1 of these 5 questions is “yes,” TB screening is considered as positive and diagnostic follow-up was performed to rule out active TB. Diagnostic evaluation was standardized and consisted of the following steps: (1) if screening was positive with cough, sputum was taken for acid-fast bacillus smear or if the results were negative a second round of sputum was taken for acid-fast bacillus smear and chest radiography; (2) if screening was positive without cough, chest radiography was performed; and (3) if sputum and chest radiography were not conclusive and no improvement after empiric antibiotics, a medical decision was made to initiate anti-TB drugs. Cases were considered to be “confirmed” if sputum acid-fast smear or culture was positive or if sputum smear was negative, clinical history and chest radiography were suggestive, and the patient was initiated on anti-TB therapy.
This information was documented in patient files during consultations, abstracted from files, and recorded in the pre-ART and ART registers according to date of TB screening. Paper-based data collection from registers was done by staff from the Rwanda MOH and its implementing partners traveling to health facilities across the country.
There were 2 types of indicators: ICF at enrollment of new patients into HIV care clinics and TB screening done during follow-up visits in patients enrolled in the program for more than 6 months. The present study is concentrated on TB screening indicators at enrollment. Data abstraction focused on 6 indicators listed below:
- number of patients newly enrolled into HIV care and treatment during the reporting period;
- number of new patients screened for active TB at enrollment into HIV care during the reporting period;
- proportion of new patients screened for active TB at enrollment into HIV care and treatment during the reporting period (numerator: number of new patients screened for active TB at enrollment, denominator: number of patients newly enrolled);
- number of new patients who screened positive for active TB at enrollment into HIV care and treatment (suspect TB cases) during the reporting period;
- proportion of new patients who were screened for active TB at enrollment into HIV care who screened positive (numerator: number of new patients who screened positive for active TB, denominator: number of new patients screened for active TB); and
- number of newly enrolled patients into HIV care and treatment who were diagnosed with active TB and started treatment for active TB disease during the reporting period.
TB screening data were collected since 2006 in 20 health facilities. The screening questionnaire and diagnostic algorithm did not change during the study period. In the first semester of 2007, data collection was expanded to 95 total health facilities supported by the President's Emergency Plan For AIDS Relief (PEPFAR)–funded implementing partners. The scale-up in other non–PEPFAR-supported sites started in 2009 with 190 sites and increased up to 328 health facilities by the second semester of 2011. After a process of data cleaning and validation by the TB/HIV technical working group, the analysis was performed on aggregated data; therefore, this study was exempt from ethical approval by the Rwanda National Ethics Committee.
Statistical analyses were conducted using χ2 analysis for trends over time. P values <0.05 were considered significant. Stata 12 software (College Station, TX) was used for the analysis.
The characteristics of health facilities reporting on TB screening indicators are shown in Table 1. TB screening reporting was initiated in 20 health facilities in 2006 with 6 district hospitals (30%) and 14 health centers (70%); 8 (40%) of these facilities were in urban areas, whereas the other 12 (60%) were in rural locations. At the end of 2011, there were a total of 352 HIV care and treatment sites in the country; data were unable to be collected from 24 sites (6.8%) because of incorrect recording procedures (data recording from files to registers was incorrectly performed because of newly hired workers who had not yet been properly trained). Over the course of the study period, the number of facilities reporting TB screening data increased to 328 by the end of 2011. The distribution of reporting sites included 40 district hospitals and 288 health centers, and there were 46 urban sites (14%) and 282 rural sites (86%).
The number of new patients and the proportion screened for TB at enrollment into HIV care and treatment programs in the country increased progressively since the beginning of reporting in 2006 through 2011; in the first semester of 2006, among 3025 patients newly enrolled, 2326 (77%) were screened for TB. At the end of 2011, 11,986 of 12,739 newly enrolled patients (94%) were screened for TB (P < 0.001). The percentage of patients screened positive at enrollment declined over time. At the beginning of 2006, 546 of 2326 patients screened positive for TB (23%). At the end of 2011, only 1198 (10%) out of 11,986 patients screened positive (P < 0.001).
The incidence of confirmed TB cases remained relatively constant over the course of the study. In the first semester of 2006, 51 patients were diagnosed with active TB disease (2.2% of patients newly enrolled: 2192 persons per 100,000 population), and at the end of 2011, 259 patients (2.2%: 2161 persons per 100,000 population) were diagnosed with active TB (P = 0.9).
Year-by-year percentages of the number of newly enrolled patients who were screened and the number who screened positive and were started on anti-TB treatment are displayed in Figures 1A, B, respectively. In 2006, the true-positive rate of screening was 9.3% (51 confirmed TB cases/546 screening positive), and this increased to 21.6% (259 confirmed TB cases/1198 screening positive) by the end of 2011 (a trend that was statistically significantly different, P = 0.02 for trend).
This article has several important findings. First, from 2006 to 2011, the number of facilities screening for TB in newly enrolled PLHIV has increased 16-fold. Second, the percentage of patients screened for TB has increased from 77% to 94%. The percentage of patients screening positive for TB at enrollment significantly decreased from 23% in 2006 to 10% in 2011. On the other hand, the percentage of patients diagnosed with active TB and started TB treatment remained constant at just over 2%.
ICF is a key component of the WHO's “Three Is” recommendations for prevention and control of TB in PLHIV. Ultimately, the goal of these strategies is to improve morbidity and mortality outcomes in dually infected (TB/HIV) patients. Limited data from Rwanda have shown a reduction in relative mortality among PLHIV diagnosed with active TB,12 reinforcing the need to integrate and improve care in this area. Accordingly, the Rwanda MOH has emphasized a nationwide ICF among PLHIV upon enrollment for HIV care. Similar to the previously described scale-up of ART in the country,13 the number of facilities providing HIV care and treatment that are reporting on TB screening data has significantly increased from 2006 to 2011. ART scale-up efforts have generally contributed to improved provider training and education as Rwanda has leveraged investments in HIV/AIDS to drive progress in other areas of health care. Since 2009, more than 90% of newly enrolled HIV patients in Rwanda have undergone TB screening. In other sub-Saharan African settings, TB screening for PLHIV at enrollment has been reported to be at 64% in one program review of 360 sites in 8 countries.14 Another systematic review done in sub-Saharan Africa has shown that the percentage of PLHIV screened for TB was 88.7%, ranging between 48% and 100%.15 Rwanda has thus made significant progress in this area.
This study also found the proportion of patients screening positive for TB at enrollment has decreased from 23% to 10% over the 6-year study period. One important reason for this decline is that patients are presenting for HIV care at earlier WHO stages. Previous work from Rwanda over roughly the same period indicated a significant reduction over time in the proportion of patients presenting with WHO stage 3/4 disease and an increase in those presenting with stage 1/2 disease.13 As patients are less likely to have advanced HIV disease, they would be less likely to have active TB or other HIV-related comorbidities that would lead to false-positive screening. Concurrently, there was also an increase in the proportion of patients, who screened positive for TB, who were diagnosed with confirmed/active TB, from 9.3% in 2006 to 21.6% in 2011. Possible reasons for this increase include improved diagnostic techniques over the study period, better quality of TB screening by providers because of substantial investments in clinical mentorship, trainings, and supervisions, and programmatic progress in maintaining patients in care and decreasing patients lost to follow-up.
A meta-analysis and systematic review on the yield of HIV-associated TB during ICF in studies from sub-Saharan Africa found the median prevalence of newly diagnosed active TB in ART and medical clinics to be 8.6%, with a wide variation (range 3.6%–24.7%) among various countries.16 In our study, the number of PLHIV initiated on anti-TB treatment remained largely the same, 2.2% throughout the study period, suggesting that the prevalence of active TB was approximately stable during the study period. However, it is also possible that improved diagnostic techniques and increased confidence in making a TB diagnosis (through clinical mentorship and training) led to an increase in absolute number of patients diagnosed with active TB in the latter years of the study compared with the earlier years while the percentage diagnosed remained the same.
HIV testing for patients with TB is also important, and Rwanda has been making progress to increase this proportion in health facilities. Pevzner et al12 have recently reported that the proportion of patients with TB having documented HIV test results increased from 48% in 2005 to 97% in 2009 in Rwanda. According to WHO data, 69% of patients with TB were tested for HIV in the African Region in 2011, which is increased from only 3% in 2004.9
Limitations of this study relate to the use of routinely collected data, which may have incomplete data and potential transcription errors as data elements are entered from patient charts into pre-ART and ART registers. Because these data were collected at the programmatic level, individual patients were not followed to ensure that all patients who screened positive for TB received all subsequent testing to confirm or rule out active TB disease. Additionally, individual responses to the 5 screening questions are not routinely collected at the program level, so analyses of which questions were more likely to be positive or more likely to be associated with active TB disease were not possible. There may also be errors in medical record completion, with potential for missing data, which may have introduced some bias into the results. If clinics that failed to report their data also had lower rates of screening, the results reported may have overestimated the actual situation in the country. Data also did not cover all health facilities that should actually be reporting on these indicators as it was not possible to do data abstraction as their registers were not properly completed.
Several important lessons were learned from review of this programmatic ICF scale-up that may be applicable in other resource-limited settings. For example, strong collaboration between TB and HIV/AIDS programs at both the central and facility levels is a key to successful ICF scale-up. Countries can leverage the infrastructure that has been constructed for HIV/AIDS as scaffolding to build upon and benefit other important disease control programs. This program has demonstrated that ICF is feasible in resource-limited settings. ICF is one of at least 3 important pillars of successful TB control programs, along with IPT and infection control, as recommended by WHO. In Rwanda, IPT is being piloted at 3 sites with plans for potential scale-up based on the preliminary evaluation of the pilot program. Infection control measures have been included in TB trainings of health-care staff and include triage with separation of persons with cough in the outpatient setting, improving cough hygiene through education sessions and posters, isolated rooms for hospitalized patients with sputum smear positivity and TB suspects, and enhanced natural ventilation in all services receiving TB suspects or TB patients. Overall, the scale-up of the ICF program in Rwanda serves to illustrate how programs can be effectively expanded nationwide in a short time frame.
This study has shown that Rwanda has been able to increase the proportion of newly enrolled PLHIV who are screened for TB using a simple screening protocol. Rates of positive screens decreased over time, possibly as a consequence of patients presenting with earlier stage HIV disease. Countries with limited resources but high HIV and TB disease prevalence should implement WHO-recommended intensified case finding as part of their integrated HIV–TB treatment programs.
1. Wong EB, Omar T, Setlhako GJ, et al.. Causes of death on antiretroviral therapy: a post-mortem study from South Africa. PLoS One. 2012;7:e47542.
2. Gupta A, Nadkarni G, Yang WT, et al.. Early mortality in adults initiating antiretroviral therapy (ART) in low- and middle-income countries (LMIC): a systematic review and meta-analysis. PLoS One. 2011;6:e28691.
8. National Institute of Statistics of Rwanda (NISR) [Rwanda], Ministry of Health (MOH) [Rwanda], ICF International. Rwanda Demographic and Health Survey 2010. Calverton, MD: NISR, MOH, and ICF International; 2011.
10. Gasana M, Vandebriel G, Kabanda G, et al.. Integrating tuberculosis and HIV care in rural Rwanda. Int J Tuberc Lung Dis. 2008;12(suppl 1):39–43.
11. Mohammed A, Ehrlich R, Wood R, et al.. Screening for tuberculosis in adults with advanced HIV infection prior to preventive therapy. Int J Tuberc Lung Dis. 2004;8:792–795.
12. Pevzner ES, Vandebriel G, Lowrance DW, et al.. Evaluation of the rapid scale-up of collaborative TB/HIV activities in TB facilities in Rwanda, 2005-2009. BMC Public Health. 2011;11:550.
13. Nsanzimana S, Ruton H, Lowrance DW, et al.. Cell phone-based and internet-based monitoring and evaluation of the National Antiretroviral Treatment Program during rapid scale-up in Rwanda: TRACnet, 2004-2010. J Acquir Immune Defic Syndr. 2012;59:e17–e23.
14. Howard AA, El-Sadr WM. Integration of tuberculosis and HIV services in sub-Saharan Africa: lessons learned. Clin Infect Dis. 2010;50(suppl 3):S238–S244.
15. Uyei J, Coetzee D, Macinko J, et al.. Integrated delivery of HIV and tuberculosis services in sub-Saharan Africa: a systematic review. Lancet Infect Dis. 2011;11:855–867.
16. Kranzer K, Houben RM, Glynn JR, et al.. Yield of HIV-associated tuberculosis during intensified case finding in resource-limited settings: a systematic review and meta-analysis. Lancet Infect Dis. 2010;10:93–102.