INT J TUBERC LUNG DIS 10(4):436 440 2006 The Union Are patients who present spontaneously with PTB symptoms to the health services in Burkina Faso well managed? S. M. Dembele,* H. Z. Ouédraogo, A. I. Combary,* B. Sondo, J. Macq, B. Dujardin * Programme National de Lutte Antituberculeuse, Ouagadougou; Institut de Recherche en Sciences de la Santé (IRSS), Ouagadougou, Burkina Faso; Ecole de Santé Publique, Université Libre de Bruxelles, Brussels, Belgium SUMMARY SETTING: Six health districts selected from a total of 53 in Burkina Faso. OBJECTIVE: To evaluate the performance of the health services in identifying infectious pulmonary tuberculosis (PTB) cases in Burkina Faso. DESIGN: Retrospective review of initial consultation registers in the first level health centres and the laboratory and treatment registers kept at the Centres for TB Diagnosis and Treatment (CDTs) in 2001. RESULTS: The rate of detection of sputum-positive cases of PTB was 11.7 cases per 100 000 population. Cough was the reason for consulting for 10.6% of 248 730 adults; 1.1% had chronic cough. Among patients with chronic cough, 66% had been referred for smear microscopy, 69.7% of whom were registered at the CDT to which they were referred. A positive diagnosis was made in 22.5% of the suspects referred and traced to the CDT. Among those with a positive diagnosis, 87.1% were put on treatment in the same CDT. CONCLUSIONS: The PTB case detection rate in Burkina Faso is low, due to the loss of cases at each of the stages leading to the diagnosis of TB. Case detection depends on the operational effectiveness of the staff working in the health services, as well as the referral of suspect patients to the CDT. KEY WORDS: selection; reference; case finding; tuberculosis; Burkina Faso TWO OF THE KEY ELEMENTS of a good National Tuberculosis Programme (NTP) are early diagnosis and rapid treatment of infectious forms of tuberculosis (TB). 1 The success of a programme therefore largely depends on its ability to find incident cases of pulmonary tuberculosis (PTB). In Burkina Faso, a developing country in West Africa with a population of 10 743 041, 2 the number of patients diagnosed with new smear-positive PTB remains low, at approximately 13 cases per 100 000 population, 3 compared with a World Health Organization (WHO) estimate of 60/100 000. 4 According to routine unpublished data, the incidence in 1999, 2000 and 2001 was 13, 13.5 and 12.8 cases/100 000, respectively. The prevalence of human immunodeficiency virus (HIV) infection among TB cases is 30%. 5 Diagnosis is the result of a process that begins with the selection of TB suspects during consultation. This is performed by nurses at the social health centres (Centres de Santé et de Promotion Sociale CSPS), which are the primary health services, or at the medical and surgical centre (Centre Médical avec Antenne chirurgicale CMA), the district hospital. The CSPS and the CMA together constitute the operational health structure of the district. Consultation with a physician at the CMA is restricted to those patients who are referred by a nurse during the initial health consultation. Patients identified as suspects are referred for smear microscopy in one of the especially equipped laboratories at the diagnostic and treatment centres (Centres de Diagnostic et de Traitement CDT). The TB case-finding rate therefore reflects the quality of the selection of TB suspects based on the criteria employed by the health personnel at the initial consultation, the quality of the referral of patients suspected of PTB and the quality of the smear microscopy performed among suspects referred to the CDT. The TB case-finding rate also depends on the rate of use of the health services and on the traditional stigmatisation of TB. The data discussed in the present article are the result of a survey conducted in 2002 which aimed to evaluate the identification of cases of infectious PTB in the health services in Burkina Faso. Correspondance to: Hermann Ouédraogo, IRSS, 03 BP 7192 Ouagadougou 03 Burkina Faso. Tel: ( 226) 36 32 15. Fax: ( 226) 36 28 38. e-mail: ouedher@hotmail.com Article submitted 21 June 2004. Final version accepted 13 September 2005. [A version in French of this article is available from the Editorial Office in Paris and from the Union website www.iuatld.org]
Tuberculosis case detection in Burkina Faso 437 MATERIAL AND METHODS The NTP covers all 53 districts of Burkina Faso. The first series of three sputum smear examinations, aimed at confirming the diagnosis, is performed free of charge for patients. A smear-positive patient is defined as a suspect who has at least two positive smear examinations. The standardised 8-month treatment regimen is also free. The mean distance between the CSPS and the CDT is 37 km. This retrospective survey covered the period 1 January to 31 December 2001. The study setting included six health districts (HD): Gorom-Gorom and Dori in the north, Secteur 15 Bobo and Banfora in the west, and Pissy and Secteur 30, Ouagadougou, in the central region (Ouagadougou is the capital of Burkina Faso). In 2001, the populations of these districts were 160 812, 231 982, 340 583, 304 417, 427 267, and 405 546, respectively. The criteria aimed at obtaining a representative selection of HDs in terms of socioeconomic, geographic, demographic and cultural characteristics. The HDs of the central region, situated in the country s administrative capital, are characterised by a very dense, mainly urban population, living from several types of economic activity; the health centres are geographically accessible. The HDs of the northern region are characterised by a sparse, rural, nomadic population that lives mainly from rearing animals. The HDs in the western region experience large movements in its trading population, which comes from different areas. Information obtained from the initial consultation registers were first transcribed onto individual survey forms for each adult ( 15 years) consulting for chronic cough. All consulting patients recorded were included in the study. The individual survey forms contained information on the date of consultation, the identification of the patient (family name, first name, age, sex, district or village of origin) and whether or not the health worker prescribed a smear examination. Patients referred for smear microscopy were then looked for in the laboratory TB registers, and in the TB registers at the health centres of the six districts, based on the name recorded in the initial consultation registers. The time to arrival at the CDT was defined as the time between the first consultation and the first smear examination at the CDT. For patients diagnosed as positive, the time to treatment was the time between the first smear examination and the start of treatment at the CDT. Reasons for consulting recorded in the initial consultation register as cough, pneumonia, pneumopathy, bronchitis, bronchopathy or broncho-pneumopathy were considered as cough. In the same register, all cough mentioned as chronic, persistent, long-term or of more than 2 weeks was considered chronic. Referred patients were those for whom a request for smear microscopy was recorded in the register. Data were entered and analysed using Excel (Microsoft Excel, Palisade Corp, Newfield, NY, USA) and SPSS (SPSS Inc, Chicago, IL, USA). The results are expressed as absolute frequencies, percentages and median delays. The case-finding rate was calculated based on the total number of new smear-positive cases actually diagnosed and recorded in the laboratory TB registers, taking the population of the district as the denominator. The 2 or the Fisher s exact test were used for the comparison of percentages, and the Kruskal Wallis test for the comparison of medians between HDs, with a significance level of 0.05. RESULTS A total of 511 506 patients presented for an initial consultation in the six districts, of whom 248 730 were adults (aged 15 years). Table 1 shows the number of adult patients consulting and of smear-positive cases/100 000 by HD. The number of patients at each stage of the casefinding process is shown in Table 2. Cough represented 26 440 (10.6%) of the reasons for consulting among adults; 300 (1.1%) of the coughers were chronic. Among those with chronic cough, 198 (66%) were referred to the CDT for smear microscopy, while 15% were referred to a consultation with a physician and 19% were sent home without a prescription for smear microscopy. Among the chronic coughers who were referred, 138 (69.7%) were recorded in the CDT laboratory registers. Among patients with chronic cough referred and traced to the CDT, 31 (22.5%) had a positive diagnosis; of these, 27 (87.1%) were put on treatment in the CDT to which they had been referred. If all the 300 chronic coughers were recorded in the CDT laboratory registers we could expect 68 [(31/138) 300] patients with a positive diagnosis. The number of smear-positive patients lost is therefore 37 (68 31) cases. The proportions of patients with chronic cough who reached each stage of the case-finding process are shown in the Figure. Only 138 (46%) of the 300 chronic coughers reached the CDT. Table 1 Distribution of consulting and smear-positive patients by health district Number of individuals consulting Detection of smearpositive patients Cases/ 100 000 P* Pissy 64 546 9.6 0.15 Secteur 30, Ouagadougou 84 178 7.9 0.01 Secteur 15, Bobo Dioulasso 38 282 1.2 0.001 Banfora 29 791 22.3 0.001 Gorom Gorom 14 886 27.4 0.001 Dori 17 047 12.9 0.56 Total 248 730 11.7 * Compared to all other districts.
438 The International Journal of Tuberculosis and Lung Disease Table 2 Distribution of patients with cough, patients with chronic cough, referred patients, patients traced to the laboratory, patients with a positive diagnosis, patients put on treatment and number of smear-positive patients lost, by health district Pissy Secteur 30 Ouagadougou Health districts Secteur 15 Bobo Banfora Gorom Gorom Dori Total Number of individuals consulting 64 546 84 178 38 282 29 791 14 886 17 047 248 730 a) Patients with cough (% a/total of individuals consulting) 6 337 (9.8) 6 138 (7.3) 4 779 (12.5) 4 137 (13.9) 2 104 (14.1) 2 945 (17.3) 26 440 (10.6) b) Patients with chronic cough (% b/a) 64 (1) 49 (0.8) 45 (0.9) 41 (1) 80 (3.8) 21 (0.7) 300 (1.1) c) Patients with chronic cough referred to the CDT* (% c/b) 51 (79.7) 28 (57.1) 18 (40) 27 (65.9) 67 (83.8) 7 (33.3) 198 (66) d) Patients referred and traced to the CDT (% d/c) 38 (74.5) 14 (50) 9 (50) 24 (88.8) 51 (76.1) 2 (28.6) 138 (69.7) e) Patients traced to the CDT with a positive diagnosis (% e/d) 6 (15.8) 2 (14.3) 1 (11.1) 3 (12.5) 19 (37.3)* 0 (0) 31 (22.5) f) Diagnosed positive and put on treatment (% f/e) 5 (83.3) 1 (50) 1 (100) 3 (100) 17 (89.5) 0 (0) 27 (87.1) g) No. of patients who would have been positive if all those with chronic cough had reached the CDT g ( e/d) b 10 7 5 5 30 0 68 h) Number of smear-positive patients lost (g e) 4 5 4 2 11 0 37 * CDT diagnosis and treatment centre. Compared to all other districts (P 0.05). Compared to all other districts (P 0.001). For the 138 patients with chronic cough who were referred and traced to the CDT, the time to arrival varied between 0 and 79 days, with a median of 1 day (Table 3). For the 27 patients diagnosed as positive and put on treatment, the time to start of treatment varied from 0 to 6 days, with a median of 1 day. These delays varied depending on the HD (Table 3). Figure Percentages of: patients with cough among adults attending the health services; patients with chronic cough among those with cough; patients referred among those with chronic cough; patients traced to the laboratory among those referred; patients with a positive diagnosis among those traced; and patients put on treatment among those with a positive diagnosis. DISCUSSION An analysis of the data from the laboratory registers of the six HDs included in the survey indicates that the level of case finding among smear-positive patients, whatever their circuit, was weak, at 11.7 cases/ 100000, giving a case-finding rate of 19.5%, similar to that at national level. 3 This is inadequate compared with the 70% target fixed by the WHO. 6 In Burkina Table 3 Sex, age, time to arrival and time to start of treatment among patients with chronic cough referred and traced to the CDT, by health district Pissy Secteur 30 Ouagadougou Health districts Secteur 15 Bobo Banfora Gorom Gorom Dori P Total Sex 0.07 n 38 14 9 24 51 2 138 % male 57.9 64.3 55.6 50 80.4 100 65.9 Age (years) 0.22 n 36 14 6 24 24 1 105 Med (min max) 34.5 (19 80) 34.5 (24 51) 32 (28 40) 40.5 (16 73) 33 (19 63) 60 35 (16 80) Time to arrival (days) 0.001 n 38 14 9 24 51 2 138 Med (min max) 2 (0 79) 5 (1 36) 2 (1 5) 1 (0 4) 2 (0 79) 9 (4 14) 1 (0 79) Time to start of treatment (days) 0.03 n 5 1 1 3 17 27 Med (min max) 3 (1 5) 1 2 2 (1 2) 1 (0 4) 1 (0 6)
Tuberculosis case detection in Burkina Faso 439 Faso, the estimated number of new smear-positive cases is 60/100 000. 4 The case-finding rate varied from one HD to the next. The Gorom Gorom HD found significantly more cases per population than the other districts (27.4/ 100 000), despite the fact that it is a district where the geographic accessibility of the health centres is low. This suggests that case finding does not depend only on the patient s health seeking behaviour. It could also be due to a more efficient organisation of the health services in this district confronted with the TB problem. Given that 1.1% of all patients with cough had chronic cough, the Gorom Gorom HD identified significantly more suspects (3.8%). It is well known that the accuracy of the selection of suspects depends essentially on the competence and motivation of the health professionals at the first consultation. 7 All patients with chronic cough should be referred as suspects by the nurse to the CDTs. Although referring patients for consultation with the physician may improve the referral rate further, nearly a fifth of patients with chronic cough returned home without a prescription for smear microscopy. The non-referral of patients with chronic cough partly explains the missing 37 smear-positive cases. It was surmised, based on the detection of suspects traced to the laboratory, that the health personnel probably referred those patients who met the clinical criteria for PTB, and that the suspects who respected their referral are those who felt that they had TB. Approximately a third (n 60, 30.3%) of the 198 suspects referred were not traced to a CDT. This could be due to a problem of access to the CDTs; however, those referred suspects who reach the laboratory do so within a reasonable time frame of not more than 2 days. The most plausible explanation is that those who were not found at the CDTs included in the survey went to the centres of other neighbouring districts that were geographically closer or to a private laboratory in the city, or they didn t have the means of travelling to the laboratory, as reported by managers of the CDTs in the HDs surveyed. Among the 138 patients with chronic cough referred at first consultation who were traced to the laboratory, a positive diagnosis was made in 22.5% (31 cases). This relatively high rate could be due to excessive filtering of TB suspects or late consultation. Late consultation can be explained by difficult access to the health centres, such as Gorom Gorom (which has the least satisfactory health coverage), leading to a rate of positive diagnosis that is even higher than in other districts. Of the 31 patients diagnosed as positive, four were not put on anti-tuberculosis treatment in the CDT surveyed. These patients may have been registered in other centres felt to be more accessible. One of these patients, who was traced, claimed that he had changed centres for his treatment so that his sister-in-law, who worked in his local CDT, wouldn t discover that he had TB. The waiting times for treatment are generally short. CONCLUSION The level of case finding for PTB in Burkina Faso is low compared to the WHO s estimated incidence of 60/100 000. It is affected at each of the stages of the diagnosis of TB. Case finding depends on the referred suspects reaching the diagnostic centre, as well as the operational efficiency of the health services and personnel in identifying and referring suspects. Loss of cases occurs between classification of a suspect and referral to the diagnostic centres by the nurse, and between referral and the arrival of the suspect at the diagnostic centre. The poor use of the health services no doubt contributes to the low identification of TB suspects. The need to systematically prescribe a sputum smear examination for all TB suspects has not yet been understood by health workers. It is also important to understand why patients do not respect referral, particularly by examining the circuit of referred suspects who were not traced to the laboratories to which they had been sent. A better understanding of the reasons for this non respect of the referral system should allow us to devise strategies to detect more TB patients in Burkina Faso. References 1 Salaniponi F M, Harries A D, Banda H T, et al. Care seeking behaviour and diagnostic processes in patients with smear-positive pulmonary tuberculosis in Malawi. Int J Tuberc Lung Dis 2000; 4: 327 332. 2 Institut National de la Statistique et de la Démographie. Recensement général de la population et de l habitat au Burkina Faso, 1996. Ouagadougou: INSD 2000: 330p. 3 Dembele S M. Quatre ans de lutte antituberculeuse à l ouest du Burkina Faso: quels résultats? Opinion des acteurs de terrain. Bruxelles, Belgique: ESP/ULB, Mémoire de DEA; 2000: 30p. 4 OMS/Région de l Afrique. Tuberculose et lutte antituberculeuse en Afrique de l Ouest : bilan et perspectives. INFO/TUB 2003; 2: 2 8. 5 Nyarko EA. Aperçu sur la situation de la tuberculose dans la région africaine. OMS-AFRO/Observatoire de la Santé en Afrique 2001; 2: 4 6. 6 WHO/IUATLD/KNCV. Revised international definitions in tuberculosis control. Int J Tuberc Lung Dis 2001; 5: 213 215. 7 Dujardin B. Tuberculose-sida : de nouveaux enjeux pour les services de santé? Développement et Santé 1996; 122: 21 27.
440 The International Journal of Tuberculosis and Lung Disease RÉSUMÉ CADRE : Choix raisonné de six districts sanitaires sur 53 au Burkina Faso. OBJECTIF : Evaluer les performances d identification des cas de tuberculose pulmonaire (TBP) contagieuse dans les services de santé au Burkina Faso. SCHÉMA : Révision rétrospective des registres de consultation des centres de santé de premier niveau et des registres de laboratoire et de traitement des centres de diagnostic et de traitement de la TB (CDT). La période concernée à été l année 2001. RÉSULTATS : Le taux de dépistage des cas de TBP positive est de 11,7 cas per 100 000 habitants. La toux représentait 10,6% des motifs de consultation des 248.730 consultants adultes ; 1,1% des toux étaient chroniques. Parmi les patients tousseurs chroniques, 66% ont été référés pour examen microscopique des frottis d expectoration, et 69,7% des référés ont été enregistrés dans le CDT où ils ont été référés. Les diagnostics positifs représentaient 22,5% des suspects référés et retrouvés au CDT. Parmi les diagnostics positifs, 87,1% ont été mis en traitement dans le même CDT. CONCLUSIONS : Le niveau du dépistage de la TBP au Burkina Faso est faible, du fait de déperditions de cas dans chacune des étapes qui conduisent au diagnostic de la TB. Le dépistage est dépendant de l efficacité opérationnelle des personnels des services de santé, ainsi que du recours au CDT par les patients suspects référés. RESUMEN MARCO DE REFERENCIA : Muestra selectiva de seis de los 53 distritos sanitarios de Burkina Faso. OBJETIVO : Evaluar el rendimiento diagnóstico para los casos de tuberculosis pulmonar (TBP) bacilífera en los servicios de salud. MÉTODO : Estudio retrospectivo de los registros de consulta de los centros sanitarios de atención primaria y de los registros de TB de laboratorio y de los centros de diagnóstico y tratamiento (CDT) en 2001. RESULTADOS : La tasa de detección de casos de TBP con baciloscopia positiva fue del 11,7 casos por 100 000 habitantes. La tos fue el motivo de consulta del 10,6% de los 248 730 adultos y 1,1% presentaron tos crónica. El 66% de los pacientes con tos crónica fueron remitidos para baciloscopia del esputo y 69,7% de ellos se registraron en el CDT al cual habían sido referidos. Se estableció el diagnóstico definitivo en el 22,5% de los pacientes con presunción de TB que acudieron al centro. El 87,1% de los pacientes con diagnóstico positivo inició tratamiento en el mismo CDT. CONCLUSIONES : La tasa de detección de casos de TBP es baja en Burkina Faso, por causa de una pérdida de casos en cada una de las etapas que llevan al diagnóstico definitivo de TB. La detección de casos depende de la eficacia operativa del personal que trabaja en los servicios de salud y de la asistencia al CDT de los pacientes referidos con presunción de TB.