Tuberculosis case detection in Nigeria, the unfinished agenda

Size: px

Start display at page:

Download "Tuberculosis case detection in Nigeria, the unfinished agenda"

Gerald Hensley
6 years ago
Views:

1 Tropical Medicine and International Health doi: /tmi volume 20 no 10 pp october 2015 Tuberculosis case detection in Nigeria, the unfinished agenda Joshua Obasanya 1, Saddiq T. Abdurrahman 2, Olanrewaju Oladimeji 3,4, Lovett Lawson 3, Russell Dacombe 4, Nkem Chukwueme 1, Tubi Abiola 1, Gidado Mustapha 5, Christophe Sola 6, Jose Dominguez 7 and Luis E. Cuevas 4 1 National Tuberculosis and Leprosy Control Programme of Nigeria, Abuja, Nigeria 2 Federal Capital Territory Abuja Tuberculosis And Leprosy Control Programme, Abuja, Nigeria 3 Zankli Medical Centre, Abuja, Nigeria 4 Liverpool School of Tropical Medicine, UK 5 TB Care, KNCV / TB CARE I, Abuja, Nigeria 6 Microbiology Department, Universite Paris-Sud, Orsay, France 7 Servei de Microbiologia, Hospital Universitari Germans Trias i Pujol, Badalona, Spain Abstract objective Underdetection of TB is a major problem in sub-saharan Africa. WHO recommends countries should have at least 1 laboratory per population. However, this recommendation is not evidence based. methods We analysed surveillance data of the Nigerian National TB Control Programme ( ) to describe TB case detection rates, their geographical distribution and their association with the density of diagnostic laboratories and HIV prevalence. results The median CDR was 17.7 (range %) in 2008, increasing to 28.6% (range %) in 2012 (P < 0.01). The CDR 2012 was associated with the 2008 baseline; however, states with CDR 2008 < 30% had larger increases than states with CDR 2008 > 30. There were 990 laboratories in 2008 and 1453 in 2012 (46.7% increase, range by state 3% to +118). The state CDR 2012 could be predicted by the laboratory density (P < 0.001), but was not associated with HIV prevalence or the proportion of smear-positive cases. CDR 2012 and laboratory density were correlated among states having < and > than 1 laboratory per population. conclusion There are large variations in laboratory density and CDR across the Nigerian states. The CDR is associated with the laboratory density. A much larger number of diagnostic centres are needed. It is likely that a laboratory density above the recommended WHO guideline would result in even higher case detection, and this ratio should be considered a minimum threshold. keywords tuberculosis, surveillance, diagnostic centres, case detection, HIV Introduction Tuberculosis (TB) is a major public health problem, and the poor detection of cases by TB control programmes remains one of the main barriers for control. Up to one-third of the cases estimated to occur each year are never reported to the national TB control programmes (NTPs) worldwide. In Nigeria, despite the National TB and Leprosy Control Programme (NTBLCP) reporting cases in 2012, this number only represents 51% of the cases estimated to have occurred in the country for that year[1]. Better case detection in the most populous country in Africa is therefore urgently needed[2]. Most NTPs rely on sputum smear microscopy for TB diagnosis and decentralise the diagnostic centres as much as possible to primary healthcare centres to improve service accessibility. WHO recommends this decentralisation, and it is recommended that NTPs provide at least one TB diagnostic centre per population[3]. This ratio is based on the maximum workload that a laboratory could handle to provide good quality smear microscopy and is based on expert opinion assuming that the ideal number of smears examined per microscopist per day should range between 2 and 20 per day[4]. Although this recommendation has stand for many years, the premise for these calculations is not ideal, as service provision should ideally reflect the number of centres that are needed to ensure service accessibility. For example, in Nigeria, very few states achieve this ratio and patients have to travel to distant centres for diagnosis[5]. Nigeria also has one of the largest HIV-infected populations[6], with urban areas having higher prevalence John Wiley & Sons Ltd

2 than rural populations[7]. As HIV increases the risk of TB progression, it is likely to reduce the proportion of cases confirmed by smear microscopy. The association between the number and density of TB diagnostic centres and the proportion of estimated TB cases detected by the states has not been reported in Nigeria. We therefore analysed the geographical distribution of TB case detection in Nigeria and its association with the prevalence of HIV and with the number of TB diagnostic centres by state during a 5-year period of standardised data collection. Methods Setting Nigeria had an estimated population of 154 million in 2009, with approximately 54% living in poverty[8]. The country is divided into 36 states plus the Federal Capital Territory (FCT). States are subdivided into 774 local government areas (LGAs) and geographically clustered into six geopolitical zones. The health system is structured along primary, secondary and tertiary levels roughly corresponding to the LGA, state and federal levels. The Table 1 Demographic and TB surveillance data for 2008 and 2012 State Population Smear positive (%) All TB CDR % Change Zamfara Abia Jigawa Sokoto Kebbi Osun Anambra Akwa Ibom Enugu Rivers Bayelsa Ekiti Imo Katsina Ondo Oyo Delta Edo Ogun Kwara Kogi Cross River Borno Taraba Yobe Niger Plateau Kano Ebonyi Kaduna Gombe Bauchi Benue Lagos Adamawa FCT Nasarawa John Wiley & Sons Ltd 1397

NTBLCP was established in 1989 and operates within the government health system. LGA health facilities provide TB treatment services; 65% of them have diagnostic laboratories[9].

3 NTBLCP was established in 1989 and operates within the government health system. LGA health facilities provide TB treatment services; 65% of them have diagnostic laboratories[9]. Study design This is an analysis of the NTBLCP routine surveillance database comprising data collected from 2008 to Data in the NTBLCP database were collected from all government centres providing TB treatment and diagnostic services in the country and from all private facilities that have established agreements with the NTBLCP to participate in the programme activities. All reporting centres use dedicated TB registers and report using a purpose-built standardised electronic data entry forms with error-trapping routines. Reports prepared by the centres are sent to the LGA TB offices, where an LGA report is compiled by TB supervisors. LGAs submit their databases to the state offices, where they are further aggregated and sent to the geopolitical zone. The geopolitical zone report is submitted quarterly to the national headquarters to prepare the national report. The national database is updated at pre-defined data submission deadlines and reports to WHO annually. All definitions used by the NTBLCP follow the WHO definitions for TB. The complete database for the years inclusive was included in this analysis. The TB case detection rate (CDR) was defined as the number of annual TB cases reported divided by the expected number of all-forms TB cases by population. The states population for 2008 and 2012 was obtained from the National Bureau of Statistics (2012). The expected number of cases for the states was calculated using the estimated 2012 national TB prevalence of 171 cases per population[1, 3], assuming that the prevalence was the same across the states. Data were grouped into quintiles and mapped to describe geographical distributions. The association of the CDR and independent variables was explored using correlation Map 1.Proportion of smear-positive cases reported (2012) Q1 (40 to <48%) Q2 (48 to <56%) Q3 (56 to <63%) Q4 (63 to 71%) Q5 (71 to 78%) Map 2. Percentage of the population with HIV infection (2010) Q1 (1 to <3.4%) Q2 (3.4 to <5.8%) Q3 (5.8 to <8.1%) Q4 (8.1 to <10.5%) Q5 (10.5 to 12.7%) Q1 (0.47 to <0.75%) Q1 (11 to <23%) Q2 (0.75 to <1.0%) Q2 (23 to <36%) Q3 (1.0 to <1.25%) Q3 (36 to <48%) Q4 (1.25 to <1.5%) Q4 (48 to <60%) Q5 (1.5 to 1.77%) Q5 (60 to 72%) Figure 1 TB case detection rate (a) and number of diagnostic centres (b) by population, proportion of smear positive (c) and percentage of the population with HIV (d) by state John Wiley & Sons Ltd

4 coefficients with 95% confidence intervals (95% CI) and multiple regressions. Residuals were plotted to further describe their distribution. As this was a correlational analysis of a large database of secondary data, it was deemed unnecessary to calculate sample size for the study. Results Totals of and cases were reported in 2008 and 2012 (Table 1), showing a 38% increase over 5 years. Changes in the number of cases were uneven, ranging from a 28.6% decrease to 353% increase from the 2008 baseline. The proportion of smear-positive cases ranged from 34.8% to 86.2% in 2008 and from 40.3% to 78.1% in Western states had a higher proportion of smear-positive cases than eastern states, as shown in Map 1 (Figure 1). HIV prevalence ranged from 1% to 12.7% in 2010 (Figure 1, map 2). States in the south-east and south south geopolitical zones had higher HIV prevalence than northern and south-east states. There were 990 laboratories in 2008 and 1453 in 2012, showing a 46.7% increase (Table 2). Changes in the number of laboratories by state ranged from a 3% decrease to a 118% increase. The geographical distribution of laboratories was heterogeneous, with a median of 0.9 laboratories per population. Only 15 (39%) states reached the recommended 1 laboratory per population ratio in 2012 (Figure 1, map 3). States in the south-east had higher laboratory ratios than in the north-east and central east states. The states with the higher laboratory ratios in 2008 also had the highest ratios in 2012 (Figure 2). The expected number of cases and the CDR by state had a heterogeneous distribution (Table 1). The CDR ranged from 4.7% to 75.8% (median 17.7%) in 2008 and from 10.6% to 72.4% (median 28.6%) in 2012 (P < 0.01). The north-east, south-east and south south states had lower CDR than the north central and southwest states (Figure 1, map 4). The 2008 and 2012 CDRs were correlated (Figure 2). CDR was correlated with the number of diagnostic centres (correlation coefficient = 0.36 for 2012) (Figure 3). This correlation did not seem to vary among states that had < or > 1 laboratories per population. Multiple linear regression to explore whether the states CDR could be predicted by the HIV prevalence, the proportion of smear-positive cases or the density of diagnostic laboratories indicated that only the latter had a statistically significant association (P < 0.001). Table 2 Number and density of TB diagnostic laboratories by state in 2008 and 2012 State Discussion Diagnostic centres 2008 Density 2012 Density Zamfara Abia Jigawa Sokoto Kebbi Osun Anambra Akwa Ibom Enugu Rivers Bayelsa Ekiti Imo Katsina Ondo Oyo Delta Edo Ogun Kwara Kogi Cross River Borno Taraba Yobe Niger Plateau Kano Ebonyi Kaduna Gombe Bauchi Benue Lagos Adamawa FCT Nasarawa The data presented here confirm significant differences in the number of cases reported by the states and that underreporting of TB is a major problem in the country, with only half of the estimated cases being reported to the NTBLCP. Some DOTS centres have incomplete or poor quality reporting practices. However, it is unlikely that these failures account for the large discrepancy between the numbers of cases estimated and reported. It is thus likely that a high proportion of cases do not reach the health system, that patients are misdiagnosed and treated 2015 John Wiley & Sons Ltd 1399

5 Diagnostic centres Diagnostic centres 2008 Case detection rate Case detection rate 2008 Figure 2 Number of diagnostic centres and case detection rates in 2008 and 2012 by state. CDR Diagnostic centres per Figure 3 Number of TB laboratory centres and TB case detection rate by state in for other diagnoses or that they attend private health facilities not participating in the NTBLCP. The latter group is likely to receive non-standardised and suboptimal TB treatments with poor outcome. Furthermore, the lack of association between the case detection rate and HIV was unexpected and may reflect the low proportion of TB cases tested for HIV. Due to the diversity of Nigeria, the national prevalence rate does not give an accurate picture of the large geographical variation in TB case detection rates, with western and northern states having lower detection rates than east central states. Several factors could determine these differences. DOTS treatment centres were introduced in southern Nigeria earlier than in the north with a consequential increase in case detection[10] and better reporting. Despite their earlier start, in 2009, the CDR was only 40% in these regions[11], indicating that still a large proportion of the population had not accessed the services. Availability of laboratory capacity and the proximity of the premises are likely to be essential, as states with poor CDR had the lowest density of diagnostic laboratories. The quality of the laboratory services also varies across facilities, adds barriers to patient access and generates inaccurate results. Laboratories at DOTS centres often have limited facilities, mostly providing smear microscopy diagnosis using the standard Ziehl Neelsen staining. Centres located outside the district hospitals do not have access to molecular diagnostics, such as GeneXpert MTB/RIF, and culture facilities are restricted to selected reference laboratories. The lack of laboratories therefore is a marker of a poor primary healthcare system. Paradoxically, the proportion of smear-positive cases was high in some states with low case detection, suggesting that patients who attend facilities in these areas may be screened more stringently or that patients attend with more advanced disease stages because of low awareness of the symptoms of TB and poor service accessibility. Furthermore, it is likely that there are differences in diagnostic routines, with some facilities focusing on reporting patients with positive smear microscopy, while others rely more frequently on X-rays and clinical diagnosis. States with the highest ratio of diagnostic centres and CDR in 2008 had the highest ratio and CDR in 2012, which may suggest an overall better programme performance. Performance is, however, dependent on a wide range of factors, including the type and availability of a functioning primary healthcare system, local resources, the presence of social unrest and conflict, frequent industrial action by healthcare providers, staff work ethics and the characteristics of the population. The association of the CDR with the number of centres therefore may be a mar John Wiley & Sons Ltd

6 ker of a better functioning system and not a causal relationship. The TB epidemic is also believed to be concentrated in densely populated rural areas[12], and therefore, the location of facilities is skewed to urban areas, with larger cities and higher population density having more, better functioning and accessible facilities and more resources than states with largely rural populations. For more than 15 years, national programmes were guided by the WHO and the United Nations Millennium Development Goals (MDGs)[13] benchmark of detecting > 70% of estimated TB cases[14]. This target was based on Styblo s & Bumgarner s postulation that removing at least 70% of new cases from the community could reduce TB incidence by 5 10% per year[15]. Although most recently the case detection rate was discontinued due to a perceived uncertainty in the calculation of TB incidence[16] and a shift to measuring progress in terms of incidence, prevalence and mortality, the proportion of cases detected is still considered an indicator of the quality of the NTPs. Although the target CDR was discontinued, the proportion of cases detected will still need to be surpassed substantially to meet post-millennium targets, which are more ambitious and demanding. Improving case detection will require increasing the number and upgrading diagnostic facilities in areas where access to services is not universal, especially in the many states in Nigeria which still have less than half of the WHO recommended laboratory density. The origin of WHO guidance on the proportion of one laboratory per population is unclear. Its basis was to provide an ideal workload and enough smearpositive cases to assure proficiency. The average setting was expected to have one new smear-positive case per population per year with a positivity rate of 10% and was not related to patient access or increasing case detection. Indeed, issues of access are much more likely to be influenced by costs to access services (food and transport)[17], community awareness of the disease (e.g. symptoms awareness and self-treatment) and poor engagement with non-government sectors[18]. Only one of the four best performing states in Nigeria had more than 1 diagnostic centre per population, and the states with the highest laboratory density had CDR around 50%. Uganda and Kenya have the highest density of microscopy facilities per population in sub-saharan Africa (3.2 and 4.2 per , respectively) and much higher CDRs (69% and 72%) than Nigeria[1]. This suggests that the current standard of 1 laboratory per is unlikely to be sufficient and that Nigeria should aim to expand its diagnostic laboratories and decentralise diagnostic services much more ambitiously. Furthermore, this expansion needs to be combined with other initiatives to make a significant impact on case detection. In Nasarawa and the FCT, which have the highest CDR but low laboratory density, many faithbased NGOs are engaged in TB control and the local NTP has initiated initiatives for active case detection in the community, suggesting that a multipronged approach is likely to be more synergistic. In conclusion, the case detection rate is associated with the density of diagnostic laboratories in the country. Although this might not be a causal relationship, increasing the number of diagnostic centres in Nigeria should be given a high priority. The location of new diagnostic centres should focus in areas with low CDR and low laboratory density. However, this development needs to be part of a package of interventions to increase case detection to fully realise the benefit of this investment. Acknowledgement We acknowledge all the NTBLCP staff, tuberculosis state control officers, the LGA TB supervisors, and the laboratory officers for their commitment. We are also grateful to all partners for their unrelenting support to TB projects in Nigeria and extend our appreciation to Mrs Funmilayo Omosebi, Mr Segun Adeshina, Drs Oluwatoyin Kuye and Amos Omoniyi of Monitoring and Evaluation unit in NTBLCP for their immense contribution on the data validation. The study was funded by a Strategic Award grant from the European and Developing Countries Clinical Trial Partnership (EDCTP). The funders had no role in study design, data collection and analysis, decision to publish or preparation of the manuscript. References 1. World Health Organization. Global Tuberculosis Report, World Health Organization, Geneva, Switzerland, Abebe M, Doherty M, Wassie L et al. TB case detection: can we remain passive while the process is active? Pan Afr Med J 2012: 11: World Health Organization Tuberculosis Global report, World Health Organization, Geneva, Switzerland, WHO.Laboratory Services in TB Control 1998; WHO: Geneva, Switzerland, p Parsons LM, Somoskovi A, Gutierrez C et al. Laboratory diagnosis of tuberculosis in resource-poor countries: challenges and opportunities. Clin Microbiol Rev 2011: 24: Chaisson RE, Martinson NA. Tuberculosis in Africa combating an HIV-driven crisis. N Engl J Med 2008: 358: National AIDS/STI Programme. National HIV sero-prevalence sentinel survey, John Wiley & Sons Ltd 1401

7 8. TB Care 1. Nigeria (Available from: [15 March 2015]. 9. Federal Ministry of Health Nigeria, Department of Public Health, National Tuberculosis and Leprosy Control Workers Manual (NTBLCP) - (5th edn, Revised Edition.) [15 March 2015] 10. Awe AKM, Odusote T, Daniel O, Omoniyi A. Improved programme performance from tripartite partnership (USAID, WHO, NTBCP): The Nigerian experience th International Union Against Tuberculosis and Lung Disease (The Union) - African region; March Abuja, Nigeria, World Health Organisation Global Tuberculosis Control, World Health Organisation: Geneva, Switzerland Federal Ministry of Health N. TB and Leprosy Control Programme, Annual Report, Revised Edition. Federal Ministry of Health Nigeria, Department of Public Health, (NTBLCP) Workers Manual: available on request from the National Tuberculosis and Leprosy Control Programme of Nigeria 13. Dye C, Maher D, Weil D, Espinal M, Raviglione M. Targets for global tuberculosis control. Int J Tuberc Lung Dis 2006: 10: World Health Organisation. The Stop TB Strategy. World Health Organisation: Geneva, Switzerland, Styblo K. The impact of HIV infection on the global epidemiology of tuberculosis. Bull Int Union Tuberc Lung Dis 1991: 66: World Health Organisation. The Global Plan to Stop TB World Health Organisation: Geneva, Switzerland, Kemp JR, Mann G, Simwaka BN, Salaniponi FM, Squire SB. Can Malawi s poor afford free tuberculosis services? Patient and household costs associated with a tuberculosis diagnosis in Lilongwe. Bull World Health Organ 2007: 85: Storla DG, Yimer S, Bjune GA. A systematic review of delay in the diagnosis and treatment of tuberculosis. BMC Public Health 2008: 8: 15. Corresponding author Luis E. Cuevas, Liverpool School of Tropical Medicine Pembroke Place, Liverpool, L3 5QA, UK. Tel.: , luis.cuevas@lstmed.ac.uk John Wiley & Sons Ltd

NIGERIA ROUTINE IMMUNIZATION IN NIGERIA. 31% of children who received the 1st dose of pentavalent vaccine did not get all 3 doses

NIGERIA ROUTINE IMMUNIZATION IN NIGERIA. 31% of children who received the 1st dose of pentavalent vaccine did not get all 3 doses MICS/NICS 206/7 NIGERIA 206 207 PROGRESS TOWARD GVAP GOALS Penta3 coverage 00% goal 90% goal States with 80% penta3 coverage NATIONAL IMMUNIZATION COVERAGE SURVEY (NICS): National Brief ROUTINE IMMUNIZATION