Pan American Journal Original research of Public Health

Access barriers to comprehensive care for people affected by tuberculosis and human immunodeficiency virus coinfection in , 2010–2015*

Lisset García-Fernández,1 Carlos Benites,1 and Byelca Huamán1

Suggested citation (Original Article) García-Fernández L, Benites C, Huamán B. Barreras para el acceso a la atención integral de las personas afectadas por la coinfección por tuberculosis y virus de inmunodeficiencia humana en Perú, 2010–2015. Rev Panam Salud Publica. 2017;41:e23.

ABSTRACT Objective. Identify the programmatic barriers that hinder access to comprehensive care of patients with tuberculosis and human immunodeficiency virus (TB/HIV) coinfection. Methods. This is a mixed-method study. Qualitative research was conducted via in-depth interviews with key actors and the quantitative component involved cross-sectional­ descriptive analysis of programmatic data from 2010-2015 on tuberculosis and HIV programs at health facilities in the cities of and . Results. Twenty-two key actors in seven establishments were interviewed. The iden­tified barriers were: little or no coordination between tuberculosis and HIV teams, separate manage- ment of tuberculosis and HIV cases at different levels of care, insufficient­ financing, limited or poorly trained human resources, and lack of an integrated information system. It was found that HIV screening in TB patients increased (from 18.8% in 2011 to 95.2% in 2015), isoniazid coverage of HIV patients declined (from 62% to 9%), and the proportion of deaths among TB/ HIV coinfection cases averaged 20%. Conclusions. There is poor coordination between HIV and TB health strategies. Management of TB/HIV coinfection is framented into different levels of care, which has an impact on comprehensive patien care. As a result of this research, a technical document was prepared to establish joint procedures that should be implemented to improve comprehensive care of TB/HIV coinfection.

Keywords Coinfection; tuberculosis; HIV; Peru.

People infected with the human im- Andean region of Latin America it is CD4 counts below 200, need to receive munodeficiency virus (HIV) are 29 118.4 per 100 000 population (2). In con- prophylaxis with isoniazid, as recom- times more likely to develop tuberculo- trast, the worldwide incidence of tuber- mended in various management guide- sis than those who do not have HIV (1). culosis in people without HIV is 98.7 lines (3–5). Globally, the incidence of tuberculosis per 100 000 population (2), while in the In 2013, there were 1.1 million people in people living with HIV (PLH) is Andean region of Latin America it is with tuberculosis/HIV (TB/HIV) coin- 105.2 per 100 000 population, and in the 117.9 per 100 000 population. Further- fection in the world (1). Deaths from tu- more, PLH who do not receive antiret- berculosis in PLH dropped from 540 000 * Official English translation provided by the Pan roviral therapy (ARVT) are at nine in 2004 to 360 000 in 2013 (6). However, American Health Organization. In the case of dis- times greater risk of contracting tuber- tuberculosis remains the leading cause of crepancy between the two versions, the Spanish original shall prevail. culosis than patients who do receive death in PLH around the globe (6). 1. National Health Strategy on the Control and treatment (2). The difference is 15 times In Peru, 1,094 cases of TB/HIV coinfec- Prevention of Sexually Transmitted Infections and greater in individuals with CD4 counts tion (7) were reported in 2014—an overall HIV/AIDS. Ministry of Health, Peru. Please direct correspondence to Lisset García-Fernández, of less than 200 cells/mL (2), so all TB/HIV coinfection rate of around 3% [email protected] AIDS patients, especially those with (8). The tuberculosis and HIV epidemics

Rev Panam Salud Publica 41, 2017 1 Original research García-Fernández et al. • Access barriers to comprehensive care for people affected by TB/HIV coinfection in Peru also have a shared epidemiological distri- comprehensive care of patients with TB/ key agents for research, to understand bution (8). Programmatic data indicate HIV coinfection. Information was gath- barriers in health systems that obstruct that HIV and tuberculosis are concen- ered from services that provide care to implementation as well as to identify trated mainly in urban areas along the tuberculosis and HIV patients, to iden- solutions to these barriers. Embedded coast and in the jungle region, and affect tify any programmatic gaps in TB/HIV ­research on program implementation in particular men in the economically ac- coinfection care. That information serves supports the effectiveness of existing tive population (7, 9). Accordingly, the as input for preparing and implementing processes and effective health policies proportion of TB/HIV coinfection is a joint national policy addressing the through the utilization of research con- higher in the Loreto (6.4%), (6.3%), TB/HIV problem. ducted as a part of the implementation and Northern Lima (5%) regions (8). process (Figure 1). A detailed description To reduce the problem of TB/HIV MATERIALS AND METHODS of implementation of the research meth- coinfection, since 2004, the World Health odology can be found in the iPIER con- Organization (WHO) (10) and other in- This study is part of a new initiative: cept paper on barriers to timely access to ternational organizations (11) have rec- “Improving Program Implementation comprehensive care for people affected ommended implementing collaborative through Embedded Research (iPIER)”, by TB/HIV coinfection in Peru. activities on TB/HIV. The WHO recom- developed by the Alliance for Health A mixed-method study was conducted. mendations focus on implementing (or Policy and Systems Research (AHPSR), It consisted of a qualitative component to strengthening) integrated services for in collaboration with the Pan American identify barriers in caring for patients TB/HIV coinfection, reducing the TB/ Health Organization (PAHO). The iPIER with TB/HIV coinfection, and a quantita- HIV disease burden, and providing early model makes program implementers tive component to identify programmatic ARVT (10). As a result, gains have been made around the globe, with access to ARVT for patients with TB/HIV coinfec- FIGURE 1. Research flow chart tion rising from 47% in 2012 to 65% in Health policy and program 2013 (6). In Latin America, ARVT cover- - Existence of national HIV and TB strategies age was 76%; however, in the countries - Known incidence and foci of TB/HIV coinfection of this region, collaboration between tu- berculosis and HIV programs has been hindered by the lack of joint national pol- icies and the absence of integration activ- Problems and barriers to implementation ities at the operational level (12). Problems: The Ministry of Health of Peru, through - Patients do not receive early comprehensive care the national health strategies for tubercu- - Low HIV screening coverage in PAT and INH chemoprophylaxis coverage losis and HIV/AIDS, is tasked with pre- Barriers: venting and controlling these health - Weak interaction between the two strategies for case diagnosis problems. The National Tuberculosis and clinical management Standard (13) and National HIV Standard - Poor programmatic and operational coordination among the (14) outline what interventions should be different levels of care (information, logistics, etc.) implemented for each of these diseases. - Separate technical standards for the diagnosis and management of TB and HIV cases. However, reports show low coverage in the coinfection indicators, which suggests difficulties in implementing and coordi- Implementation strategy nating care for patients with TB/HIV - Analyze access barriers to diagnosis and management of coinfection by the two programs (which coinfected patients are called “strategies” in Peru). - Address the training needs of TB and HIV health teams at the One of the main difficulties in caring intervention sites - Standardize programmatic information at the national, regional, for patients with TB/HIV coinfection is and local levels that care is provided separately for tu- - Hold workshop to review barriers and build consensus for berculosis and HIV. Furthermore, most coordinating operational activities among the different levels of care patients with TB/HIV coinfection are - Hold coordination/management meetings between the two strategies treated for tuberculosis in a primary care - Devise a joint work plan facility, but receive ARVT in another fa- cility, generally at the secondary or terti- Changes in the health policy or program ary level. There is not a detailed - Preparation of a single standard of care for patients with TB/HIV understanding of which elements pre- coinfection vent the tuberculosis and HIV health - Consensus-based programmatic and operational information at strategy teams from coordinating actions the national, regional, and local levels (continuing training, to improve care for patients with TB/ screening, treatment, information system, etc.) HIV coinfection. - Greater interaction between the TB and HIV programs - Preparation of an intervention plan The objective of this research was to identify programmatic barriers to HIV, human immunodeficiency virus; TB, tuberculosis; INH, isoniazid; PAT, people affected by tuberculosis.

2 Rev Panam Salud Publica 41, 2017 García-Fernández et al. • Access barriers to comprehensive care for people affected by TB/HIV coinfection in Peru Original research gaps. The study was conducted in Lima facilities’ tuberculosis and HIV strategy coding, using the ATLAS.ti version 7.2® and Iquitos, two cities with high inci- coordinators, physicians, nurses, obste- software. dence of tuberculosis and HIV (8). Seven tricians and people with TB/HIV coinfec- For the quantitative component, pro- health facilities in the district of San Juan tion were selected. Professionals on the grammatic data on the tuberculosis and de Lurigancho (Lima) and three districts technical teams for the health strategies HIV strategies in the health facilities were in Iquitos were selected (Figure 2), be- at the national level were also inter- inspected. Existing administrative and cause they had the greatest number of viewed. A semi-structured questionnaire clinical registries for each health facility tuberculosis or HIV patients. The charac- was administered by an interviewer, dif- were consulted for 2010 to 2015. Addi- teristics of the facilities are described in ferentiated for health providers and pa- tional information was requested from Table 1. tients. All the interviews were audio health workers to complement the fore- The qualitative component was con- recorded, then transcribed. The analysis going. The data collection instrument ducted via in-depth interviews with used an iterative process of reading the was prepared by adapting the checklist key actors in the care process. Health transcriptions, selecting subjects, and from a similar project in Honduras (15). It

FIGURE 2. Location of health facilities in the study

1 TUMBES A LORETO 4 M A Z Iquitos C 2 O A N J A 3 A S M A 1→ Regional Hospital of Loreto LAMBAYEQUE R 2→ Iquitos Hospital C SAN 3 Belén Health Center A MARTIN → 4→ Bellavista Nanay Health Center LA LIBERTAD

ÁNCASH PASCO UCAYALI HUÁNUC O

LIMA JUNÍN MADRE DE DIOS SJL

CUSCO

I HUANCAVELICA C APURIMAC A

5 7 6

MOQUEGUA

5→ Hospital San Juan de Lurigancho 6→ La Huayrona Health Center 7→ Sagrada Familia Health Post

Rev Panam Salud Publica 41, 2017 3 Original research García-Fernández et al. • Access barriers to comprehensive care for people affected by TB/HIV coinfection in Peru

TABLE 1. Characteristics of the health facilities visited

TB Strategy HIV Strategy Health facility City District Level Bacilloscopy DOT Screening ARVT “Felipe Santiago Arriola Iglesias” Regional Iquitos Punchana III-1 Yes Restricted a ELISA Yes Hospital of Loreto “César Garayar García” Iquitos Hospital Iquitos Iquitos II-2 Yes Yes Rapid test Yes Belén Health Center of Villa Belén Iquitos Belén I-3 Yes Yes Rapid test No Bellavista Nanay Health Center Iquitos Punchana I-4 Yes Yes Rapid test No Hospital San Juan de Lurigancho Lima SJL II-2 Yes Restricted b ELISA Under accreditation La Huayrona Health Center Lima SJL I-3 Yes Yes Rapid test No Sagrada Familia Health Post Lima SJL I-2 Only sample collection Yes Rapid test No DOT, directly observed treatment by mouth; ARVT, antiretroviral therapy; TB, tuberculosis; HIV, human immunodeficiency virus; ELISA, enzyme-linked immunosorbent assay. a. Only for patients in the surrounding areas and those who are hospitalized. Diagnosed cases are referred to primary care centers. b. Only for hospitalized patients. Diagnosed cases are referred to primary care centers. consisted of the following sections: iden- Comprehensive care of TB/HIV coin- personally write the referral for the physician tification data, general considerations, fection requires the services aimed at to sign, but the referral is not actually tuberculosis and HIV care and data for managing those pathologies to maintain recorded. This has allowed me to see more constructing programmatic indicators. fluid communication, so that the status people or give more timely treatment [ ... ].” (IQT_02_Ic) The information was analyzed in an Ex- of each patient is known. The principal cel 2010® spreadsheet, using descriptive results of this lack of coordination are The absence of an integrated informa- statistics. that there is no linkage between the ser- tion system hinders monitoring of coin- The study was approved by the Ethics vices, for example, patients who received fected patients and overall data analysis. Committee of the Hipólito Unanue Na- reagent screening in TB services but who Furthermore, cases are registered manu- tional Hospital (Ref. No. 059-2015-CIEI- do not go to the HIV services, or patients ally, so information can be lost. The ab- HNHU) and the PAHO Ethics Review lost to follow-up, in the case of people sence of computer equipment and Committee (Ref. No. PAHO-2015-03- with HIV who were diagnosed with TB properly functioning Internet service 0015). Before collecting information, the but do not seek treatment. contributes to this problem. objective and study procedures were pre- In Peru, TB/HIV coinfection tends to sented to the chief or head of each health be treated at different health facilities, “We only handle files and case histories; there is no system; I would like it to be interconnected, facility. Health workers from the facility’s hampering coordination between the but it is not [ ... ] in the end it is empirical, ev- tuberculosis and HIV strategies were then teams. This is because the TB and HIV erything is done manually, scrambling at the invited to take part in the study. Patients care models differ in several ways. For end of the month.” (IQT_01_Ib) were invited through the health workers. TB, patients are seen in primary care fa- An informed consent form had to be read cilities, close to home, where they receive This barrier is especially apparent at and signed to participate in the study. directly observed treatment (DOT). HIV, the central level, since when data are in turn, is managed in some secondary- consolidated, they are found to be incon- RESULTS or tertiary-level health facilities that have sistent. This, in turn, reflects the lack of trained professionals, where drugs are coordination between the services and Qualitative component delivered every month or two months, the fact that a single patient goes to dif- according to patient needs. Because of ferent health facilities that use different Twenty-two key actors (16 health pro- this fragmentation, health professionals IDs, which creates reporting duplications viders, four patients, and two staff mem- do not know the patient’s status, there is or omissions. bers) were interviewed. Several barriers lack of adherence to treatment, and pa- There are also barriers related to prob- to comprehensive care of patients with tients experience financial problems. lems in the health system, such as insuf- TB/HIV were identified. ficient funding for the strategies and “[ ... ] There is a post near my house; the same The first barrier is little or no coordina- lack of human resources. The workers pneumologist who sees me here [in the hospi- interviewed agree that the real issue is tion of activities, such as HIV screening in tal] sees me there [ ... ] because here it is more not lack of money, but the quality of tuberculosis patients or monitoring com- expensive. I only come here once a month for pliance with treatment. Health workers’ my antiretrovirals. ”(“ H_IV) spending. The funds generally come experience and knowledge influence this from the results-based budget (RBB) relationship. Other factors are staff will- Poor coordination is also seen at the and are not used for the tuberculosis ingness and interest, physical proximity, programmatic level, as there is no policy and HIV strategies, since facility offi- and affinity among those responsible. document standardizing each strategy’s cials allocate these resources to other procedures and responsibilities for man- expenditures. “Well, I do not know what activities our aging TB/HIV coinfection, which creates counterparts [in HIV] do; admittedly we are “[ ... ] because the authorities use the budget uncertainty among health workers. quite removed. I think we would have to pro- for other things. It is a constant struggle to vide the information, because the hospital “There are patients who come on their own try to protect the funds we have; some sup- does not know what patients are in other cen- but I have adopted strategies outside of the port us, others don’t; theirs is more of a wel- ters [ … ]. ”(“ IQT_02_Ic) legal standards or framework [ ... ]. I fare approach [ ... ]” (IQT_01_III)

4 Rev Panam Salud Publica 41, 2017 García-Fernández et al. • Access barriers to comprehensive care for people affected by TB/HIV coinfection in Peru Original research

In Peru, the shortage of health workers FIGURE 3. Coverage of HIV screening in patients with TB in all the health facilities in is due not only to a lack of professionals, the study, 2010–2015 but also to their uneven distribution. 1000 95,2 Health professionals therefore have to 100 85,5 split their time between administrative tasks and patient care. In the case of the 800 80 TB and HIV strategies, the workers are not only in charge of the day-to-day care 57,5 57,7 600 60 of patients, but also of activity program- 49,2 ming, management, and reporting. The issue is compounded by high staff turn- 400 40 Coverage (%) over because of a lack of labor incentives, Number of patients 18,8 which has an impact on all the activities 200 20 of both strategies. “[ ... ] we have to provide care and do admin- istrative work; and we also run the center’s 0 0 micro-network, so it is like we have two re- 2010 2011 2012 2013 2014 2015 sponsibilities.” (SJL_01_IIa) HIV, human immunodeficiency virus; TB, tuberculosis. Personnel shortages are also related to insufficient funding, and the human re- sources lack regular and refresher train- FIGURE 4. Percentage of patients with HIV who began IPT per year, Iquitos, 2010–2015 ing, particularly at the primary care % level. 98 100 91 87 85 90 Quantitative component 75 80 Screening coverage for HIV in tuber- 70 culosis patients steadily increased from 60 18.8% in 2011 to 95.2% in 2015 (Figure 3). No systematized information was found 50 40 37 in HIV registries on ruling out a tubercu- 40 losis diagnosis. 30 Coverage of isoniazid preventive 21 therapy (IPT) at the Regional Hospital 20 11 8 9 stood at over 85% from 2010 to 2013, 10 but fell to 37% in 2015. In Iquitos Hos- 0 pital, ITP coverage dropped from 40% 2010 2011 2012 2013 2014 2015 in 2012 to 11% in 2015 (Figure 4). No information integrated with the tuber- Regional Hospital of Loreto Iquitos Hospital culosis strategy on coinfected patients HIV, human immunodeficiency virus; IPT, isoniazid preventive therapy. who receive ARVT was found, since they tend to be served in different health facilities. DISCUSSION have improved notably. However, iso- Of the 266 patients with TB/HIV coin- niazid coverage in HIV patients is still fection (225 in Iquitos and 41 in Lima) This study identified the program- well below the national target of 100%. from 2010 to 2015, the Iquitos Hospital matic barriers that hinder access to com- The most significant difference in the reported one case of multidrug-resistant prehensive care of patients with TB/ indicators evaluated in the Lima and tuberculosis (MDR-TB) in 2012, while HIV coinfection in seven health facili- Iquitos health facilities was IPT cover- Hospital San Juan de Lurigancho had ties in Lima and Iquitos, Peru. These age. This could be due to current regu- eight cases. Furthermore, two cases of barriers include difficulties coordinat- lations not being implemented or weak monoresitant tuberculosis were reported ing between the tuberculosis and HIV supply of the drugs in the jungle re- in the Regional Hospital of Loreto, and strategies, fragmentation of care in dif- gion, among other possible reasons two at the Sagrada Familia Health Post. ferent health facilities, the absence of that could be explored further in an- No cases of extremely resistant tubercu- policy documents for managing this other study. losis were recorded. condition and the lack of an integrated Collaboration between tuberculosis In total there were 52 deaths among information system. Other barriers are and HIV programs has been analyzed the cases of TB/HIV coinfection (20%), insufficient funding for their work, the in several studies, which found that lack from 2010 to 2015. This percentage was scarcity of health workers and their lack of integration at the different levels is greater in Lima than in Iquitos (24% com- of training. Some indicators, such as a problem common to several contexts pared to 19%). screening tuberculosis patients for HIV, (12, 16–19). Much like in our study, other

Rev Panam Salud Publica 41, 2017 5 Original research García-Fernández et al. • Access barriers to comprehensive care for people affected by TB/HIV coinfection in Peru research has found that this lack of inte- (cure or adherence to treatment) in pa- Management of TB/HIV coinfection gration results in monitoring difficulty, tients with TB/HIV coinfection (28). is fragmented, which hinders compre- loss of cases, fragmentation of care, and This study’s principal limitation is pos- hensive patient care. These findings more missed opportunities (16, 19). sible social desirability bias, especially on point to the need for specific policies These barriers have also been found to the part of patients, since the interviews and regulations to address the problem have effects on indicators, such as IPT were conducted at the health facilities. of TB/HIV coinfection in Peru. As a coverage in patients with HIV and high There could also be participant selection result, a joint technical document on mortality in patients with coinfection, in bias, since those available at the time of TB/HIV was prepared. It establishes addition to the lack of data for construct- the visit were the ones interviewed. How- health interventions and administra- ing indicators, such as ARVT coverage in ever, the information obtained is consis- tive procedures in health services at patients with tuberculosis. tent with the type of key actor interviewed, the national level and aims to reduce Lack of financing and human re- which gives us confidence that the barri- the incidence, morbidity, and mortal- sources are two more major barriers that ers were adequately explored. The limita- ity of people affected by TB and HIV tend to be present in the Peruvian health tions of the quantitative component are in Peru. system (20, 21). As in our study, they related to the data source, since registries limit the integration of joint tuberculosis can underreport cases or activities. This Acknowledgements. The authors and HIV activities (22, 23). Furthermore, bias is expected to be similar across the wish to thank Ariel Bardach, Sebastián lack of staff training is a key factor, since facilities and, therefore, does not affect García, and Fernando Rubinstein, of the it would contribute to better manage- comparisons. In spite of these constraints, Institute for Clinical Effectiveness and ment of TB/HIV coinfection, and of this study contributes information that Health Policy (IECS), for their valuable available mechanisms for streamlining will enable decision-makers to identify contributions during preparation of the the processes for patients. The techni- weaknesses, from a system standpoint, in project; and Miluska Carrasco and Hel- cal document should provide a legal the care of TB/HIV coinfection. len Palma for conducting the fieldwork. framework that prompts changes in Lastly, a barrier identified by this We are also grateful to the authorities ­administrative procedures and, in turn, study is the lack of a technical document and workers at the Regional Health Of- facilitates human resources education for TB/HIV coinfection care. Accord- fice in Loreto (Hermann Silva and Ceci and makes the health interventions ingly, during this research, a document Ríos) and the San Juan de Lurigancho sustainable. was prepared standardizing comprehen- Health Network (Nancy Zerpa, Corina This study did not examine barriers sive care processes for patients with TB/ Vásquez and Flor Domínguez) for facil- from the patient’s standpoint. Other HIV coinfection, addressing the main itating our research. Lastly, the authors studies have found that personal fac- barriers encountered. Furthermore, a would like to thank the study partici- tors, such as lifestyle, socioeconomic meeting was organized between the tu- pants for the valuable information problems, social support, lack of un- berculosis and HIV teams at the central they provided. derstanding of the treatment, stigma level and in some health facilities in Lima and discrimination, and alcohol and to discuss some elements of the technical Funding. This study was financed by drug use affect the treatment of TB/ document. The next steps (validation the Alliance for Health Policy and Sys- HIV coinfection (24–27). Although and subsequent implementation) should, tems Research of the World Health Orga- these factors are important, they are however, involve all the actors; other- nization (WHO). The Pan American difficult to solve from a programmatic wise, it would run the risk of being ig- Health Organization provided technical standpoint, since they require a multi- nored at the operational level (29). cooperation for this project. sectoral approach. Furthermore, better health system performance (measured CONCLUSIONS Declaration. The opinions expressed by the context, integration, support ser- in this paper are the responsibility of the vices, human resources and service There is poor coordination be- author and do not necessarily reflect the continuity and quality) is found to be tween HIV and TB health strategies in opinion or policy of the RPSP/PAJPH related to better treatment outcomes health facilities included in the study. and/or PAHO.

REFERENCES

1. World Health Organization (WHO). Global the Global Burden of Disease Study 2013. for the prevention and treatment of oppor- tuberculosis report. Geneva; WHO; 2014. The Lancet. 2014;384(9947):1005–70. tunistic infections in HIV-infected adults Available at: http://apps.who.int/iris/bit- 3. Doherty M, Ford N, Vitoria M, Weiler G, and adolescents: recommendations from stream/10665/137094/1/9789241564809_ Hirnschall G. The 2013 WHO guidelines the Centers for Disease Control and eng.pdf. Accessed on 5 August 2015. for antiretroviral therapy: evidence-based Prevention, the National Institutes of 2. Murray CJL, Ortblad KF, Guinovart C, Lim recommendations to face new epidemic Health, and the HIV Medicine Association SS, Wolock TM, Roberts DA, et al. Global, realities. Curr Opin HIV AIDS. 2013;8(6):​ of the Infectious Diseases Society of regional, and national incidence and mor- 528–34. America; 2014. Available at: http://aidsinfo. tality for HIV, tuberculosis, and malaria 4. Panel on Opportunistic Infections in HIV- nih.gov/contentfiles/lvguidelines/adult_ during 1990–2013: a systematic analysis for Infected Adults and Adolescents. Guidelines oi.pdf. Accessed on 12 August 2015.

6 Rev Panam Salud Publica 41, 2017 García-Fernández et al. • Access barriers to comprehensive care for people affected by TB/HIV coinfection in Peru Original research

5. Panel on Antiretroviral Guidelines for view&gid=23710&itemid&Itemid=270. choice experiment. PLoS ONE. 2012;7(12):​ Adults and Adolescents. Guidelines for Accessed on 5 August 2015. e50315. the Use of Antiretroviral Agents in 13. Ministerio de Salud del Perú (MINSA). 22. Njozing BN, Edin KE, San Sebastián M, HIV-1-Infected Adults and Adolescents. De- Norma Técnica de Salud N°104/MINSA- Hurtig AK. Voices from the frontline: partment of Health and Human Services; DGSP-V.01 “Norma Técnica de Salud para counsellors’ perspectives on TB/HIV col- 2013. Available at: http://aidsinfo.nih.gov/ la Atención Integral de las Personas laborative activities in the Northwest contentfiles/lvguidelines/Adultand​Ado- Afectadas por Tuberculosis.” RM No. 715- Region, Cameroon. BMC Health Serv Res. lescentGL.pdf 2013/MINSA, 2013. 2011;11:328. 6. United Nation Programme on HIV/AIDS 14. Ministerio de Salud del Perú (MINSA). 23. Nansera D, Bajunirwe F, Kabakyenga J, (UNAIDS). The gap report. Geneva: Norma Técnica de Salud N° 097 -MINSA/ Asiimwe PKJ, Mayanja-Kizza H. UNAIDS; 2014. Available at: http://www. DGSP-V.02 “Norma Técnica de Salud Opportunities and barriers for implemen- unaids.org/en/resources/​campaigns/​ de Atención Integral del Adulto/a tation of integrated TB and HIV care in 2014/2014gapreport/gapreport. Accessed con Infección por el Virus de la lower level health units: experiences from on 5 August 2015. Inmunodeficiencia Humana (VIH).” RM a rural western Ugandan district. Afr 7. Ministerio de Salud del Perú No. 962-2014/MINSA, 2014. Health Sci. 2010;10(4):312–9. (MINSAL). Estrategia Sanitaria Nacional 15. Organización Panamericana de la Salud 24. Daftary A, Padayatchi N. Social con- de Prevención y Control de Tuberculosis. (OPS). Proyecto Demostrativo para inte- straints to TB/HIV healthcare: accounts Informe de monitoreo. Lima: MINSAL; gración de la atención de la Tuberculosis y from coinfected patients in South Africa. 2014. VIH. Organización Panamericana de la AIDS Care. 2012;24(12):1480–6. 8. Ministerio de Salud del Perú (MINSAL). Salud, 2013. Available at: http://www. 25. de Sousa Filho MP, Luna IT, da Silva KL, Dirección General de Epidemiología. paho.org/hon/inde,.php?option=com_ Pinheirod PN da C. Pacientes vivendo Análisis de la situación epidemiológica content&view=article&id=925%3Ase-soci- com HIV/AIDS e coinfecção tuberculose: del VIH/SIDA en el Perú. Lima: MINSAL; aliza-proyecto-demostrativo-integracion-a- dificuldades associadas à adesão ou ao 2013. tencion-tuberculosis-vih&Itemid=229. abandono. Rev Gaucha Enferm. 2012;33(2):​ 9. Ministerio de Salud del Perú (MINSAL). Accessed on 5 August 2015. 139–45. Estrategia Sanitaria de Prevención y 16. Daftary A, Padayatchi N. Integrating pa- 26. Gebremariam MK, Bjune GA, Frich JC. Control de ITS y VIH/SIDA. Informe de tients’ perspectives into integrated tuber- Barriers and facilitators of adherence to TB monitoreo. Lima: MINSAL; 2014. culosis-human immunodeficiency virus treatment in patients on concomitant TB 10. World Health Organization (WHO), health care. Int J Tuberc Lung Dis. and HIV treatment: a qualitative study. United States President’s Emergency Plan 2013;17(4):546–51. BMC Public Health. 2010;10:651. for AIDS Relief (PEPFAR), Joint United 17. Conseil A, Mounier-Jack S, Coker R. 27. Rodrigues ILA, Monteiro LL, Pacheco Nations Programme on HIV/AIDS Integration of health systems and priority RHB, da Silva SED. Abandono do trata- (UNAIDS) and the Global Fund to Fight health interventions: a case study of the mento de tuberculose em co-infectados AIDS, Tuberculosis and Malaria. A guide integration of HIV and TB control pro- TB/HIV. Rev Esc Enferm USP. 2010;44(2):​ to monitoring and evaluation for collabo- grammes into the general health system in 383–7. rative TB/HIV activities. Geneva; WHO- Vietnam. Health Policy Plan. 2010;25(S1):​ 28. Loveday M, Padayatchi N, Wallengren K, PEPFAR-UNAIDS; 2015. Available at: 32–6. Roberts J, Brust JCM, Ngozo J, et al. http://www.who.int/tb/publications/​ 18. Wannheden C, Westling K, Savage C, Association between health systems per- m_and_e_document_page/en/. Accessed Sandahl C, Ellenius J. HIV and tuberculo- formance and treatment outcomes in pa- on 5 August 2015. sis coinfection: a qualitative study of treat- tients co-infected with MDR-TB and HIV 11. Fujiwara P, Dlodlo R, Ferroussier O, ment challenges faced by care providers. in KwaZulu-Natal, South Africa: implica- Nakanwagi-Mukwaya A, Cesari G, Boillot Int J Tuberc Lung Dis. 2013;17(8):1029–35. tions for TB programmes. PLoS ONE. 2014;9​ F. Implementing Collaborative TB-HIV 19. Furin J, Isaakidis P, Reid AJ, Kielmann K. (4):e94016. Activities: A Programmatic Guide. Paris: ‘I’m fed up’: experiences of prior anti-tu- 29. Nabyonga Orem J, Bataringaya International Union Against Tuberculosis berculosis treatment in patients with Wavamunno J, Bakeera SK, Criel B. Do and Lung Disease; 2012. Available at: drug-resistant tuberculosis and HIV. Int J guidelines influence the implementation https://www.theunion.org/what-we-do/​ Tuberc Lung Dis. 2014;18(12):1479–84. of health programs?-Uganda’s experience. publications/.../pub_tb-hivguide_eng_ 20. Huicho L, Canseco FD, Lema C, Miranda Implement Sci. 2012;7:98. web-1.pdf. Accessed on 5 August 2015. JJ, Lescano AG. Incentives to attract and 12. Pan American Health Organization retain the health workforce in rural areas (PAHO). Antiretrovieal Treatment in the of Peru: a qualitative study. Cad Saude Spotlight: A Public Health Analysis in Publica. 2012;28(4):729–39. Latin America and the Caribbean. 21. Huicho L, Miranda JJ, Diez-Canseco F, Washington, DC: PAHO; 2013. Available Lema C, Lescano AG, Lagarde M, et al. Job at: http://www2.paho.org/hq/index. preferences of nurses and midwives for Manuscript received on 6 May 2016. Revised version php?​option=com_docman&task=doc_​ taking up a rural job in Peru: a discrete accepted for publication on 5 October 2016.

Rev Panam Salud Publica 41, 2017 7 Original research García-Fernández et al. • Access barriers to comprehensive care for people affected by TB/HIV coinfection in Peru

RESUMO Objetivo. Identificar as barreiras programáticas que dificultam o acesso à atenção integral de pacientes com coinfecção por tuberculose e vírus da imunodeficiência humana (TB/HIV). Métodos. Estudo de métodos mistos. A pesquisa qualitativa foi realizada mediante Barreiras no acesso à entrevistas aprofundadas com atores-chave, e o componente quantitativo baseou-se na análise descritiva transversal de dados programáticos dos programas de tubercu­ atenção integral de pessoas lose e HIV de estabelecimentos de saúde das cidades de Lima e Iquitos no período de afetadas pela coinfecção por 2010 a 2015. tuberculose e vírus da Resultados. Foram entrevistados 22 atores-chave em sete estabelecimentos. As bar- reiras identificadas foram: pouca ou nenhuma coordenação entre as equipes de tuber- imunodeficiência humana culose e HIV, manejo separado dos casos de tuberculose e HIV em diferentes níveis de no Peru, 2010 a 2015 atenção, financiamento insuficiente, recursos humanos escassos ou pouco capacitados e ausência de um sistema de informação integrado. Constatou-se que o rastreamento de HIV em pacientes com tuberculose aumentou (de 18,8% em 2011 para 95,2% em 2015), a cobertura da profilaxia com isoniazida em pacientes com HIV diminuiu­ (de 62% para 9%) e a proporção média de óbitos entre os casos de coinfecção por TB/HIV foi de 20%. Conclusões. Existe uma má coordenação entre as estratégias de saúde para HIV e tuberculose. O manejo da coinfecção por TB/HIV é fragmentada nos diferentes níveis de atenção, o que prejudica a atenção integral do paciente. Esta pesquisa resul- tou na elaboração de um documento técnico para estabelecer os procedimentos conjuntos que deverão ser implementados para melhorar a atenção integral da coin- fecção por TB/HIV.

Palavras-chave Coinfecção; tuberculose; HIV; Peru.

RESUMEN Objetivo. Identificar las barreras programáticas que dificultan el acceso a la atención inte­gral de pacientes con coinfección por tuberculosis y virus de la inmunodeficiencia humana (TB/VIH). Métodos. Se trata de un estudio de métodos mixtos. La investigación cualitativa se Barreras para el acceso a la realizó mediante entrevistas en profundidad a actores clave y el componente cuantita- tivo a través del análisis descriptivo de corte transversal de datos programáticos del atención integral de las período 2010–2015 sobre los programas de tuberculosis y VIH de establecimientos de personas afectadas por la salud de las ciudades de Lima e Iquitos. coinfección por tuberculosis Resultados. Se entrevistaron a 22 actores clave en siete establecimientos. Las barre- ras identificadas­ fueron: poca o ninguna coordinación entre los equipos de tuberculo- y virus de inmunodeficiencia sis y VIH, manejo por separado de los casos de tuberculosis y de VIH en diferentes humana en Perú, 2010–2015 niveles de atención, financia­miento insuficiente, recursos humanos escasos o poco capacitados y ausencia de un sistema de información integrado. Se evidenció que el tamizaje para VIH en pacientes con tuberculosis se incrementó (de 18,8% en 2011 a 95,2% en 2015), la cobertura de isoniazida en pacientes con VIH disminuyó (de 62% a 9%) y la proporción de fallecidos entre los casos de coinfección por TB/VIH fue de 20% en promedio. Conclusiones. Existe una débil coordinación entre las estrategias sanitarias sobre VIH y sobre tuberculosis. El manejo de la coinfección por TB/VIH es fragmentado en diferentes nive­les de atención, lo que repercute en la atención integral del paciente. Como producto de esta investigación, se elaboró un documento técnico para es­ tablecer los procedimientos conjuntos, el cual deberá ser implementado para una mejora en la atención integral de la coinfección por TB/VIH.

Palabras clave Coinfección; tuberculosis; VIH; Perú.

8 Rev Panam Salud Publica 41, 2017