INT J TUBERC LUNG DIS 7(6):550–555 © 2003 IUATLD
Symptoms in patients attending services for diagnosis of pulmonary tuberculosis in Sudan A. I. El-Sony,*† S. A. Mustafa,‡ A. H. Khamis,§¶# S. Sobhi,* D. A. Enarson,† O. Z. Baraka,** G. Bjune†† * National Tuberculosis Programme, Khartoum, Khartoum, Sudan; † International Union Against Tuberculosis and Lung Disease (IUATLD), Paris, France; ‡ Epidemiological Laboratory, National Tuberculosis Programme, Khartoum, Khartoum, § Department of Medicine, Ahfad University for Women, Omdurman, Khartoum, ¶ University of Sudan for Science and Technology, Khartoum, Khartoum, # Bayan College for Science and Technology, Khartoum, Khartoum, ** University of Khartoum, Department of Medicine, Khartoum, Khartoum, Sudan; †† University of Oslo, Oslo, Norway SUMMARY OBJECTIVE:
To describe the variation in clinical features of individuals presenting to a health facility with chest symptoms according to their ultimate diagnosis. M E T H O D S : Of 16 735 patients (52.6% males and 47.4% females) presenting at respiratory centres in seven states in Sudan, 5338 patients were identified with respiratory tract symptoms: 2917 (54.6%) men and 2421 (45.4%) women, with a mean age of 32 years. Those who had cough for more than 3 weeks that was not responsive to a course of antibiotics were screened by microscopy of two or three sputum specimens and chest radiography.
R E S U L T S : A total of 504 (9.44%) were smear-positive, 259 (4.85%) were smear-negative, and 166 (3.11%) had extra-pulmonary tuberculosis, of whom 59 (1.11%) had pleural involvement; the remaining 4409 suspects (82.6%) were non-tuberculous. C O N C L U S I O N : Tuberculosis patients had a constellation of presenting symptoms, with the principal symptom being cough for more than 3 weeks. The accompanying symptoms with greatest predicted significance were weight loss, tiredness and night sweats. K E Y W O R D S : tuberculosis; diagnosis; symptoms; Sudan
THE GLOBAL EPIDEMIC of tuberculosis (TB) is increasing. The breakdown in health services, the spread of the human immunodeficiency virus and the acquired immune deficiency syndrome (HIV/AIDS) and the emergence of multidrug-resistant TB are contributing to the worsening impact of this disease.1–3 One third of the world’s population is infected with tuberculosis, and around 8 million TB cases and some 2 million deaths occur annually. Most cases and deaths occur in developing countries.4,5 Smear examination of sputum specimens obtained from patients presenting with respiratory symptoms suggestive of tuberculosis remains the cornerstone of diagnosis of tuberculosis in most locations with a high burden of tuberculosis. The sensitivity of sputum microscopy to identify all cases of tuberculosis, even in good centres, is only about 60%,6–8 and patients whose smears are negative for acid-fast bacilli (AFB) represent a diagnostic dilemma. Sudan has a low prevalence of HIV among TB patients, of about 7.9%. The HIV prevalence in the country is 1.4% in northern Sudan and 5% in southern Sudan. The incidence of TB is thus not affected by the HIV epidemic.9
The aim of this study was to describe the variation in clinical features of individuals presenting to a health facility with chest symptoms according to their ultimate diagnosis, in order to assist the clinician in diagnosing tuberculosis in settings where resources are limited. A representative sample of patients attending the respiratory disease centres was studied and those suspected of tuberculosis were offered sputum smear microscopy for AFB, radiographic and clinical examination, according to the routine practice of the National Tuberculosis Programme (NTP).
PATIENTS AND METHODS Persons between 15 and 49 years of age with respiratory symptoms seen consecutively in the chest outpatient departments of hospitals and chest clinics at health centres and examined by the health personnel in Gezira, Bahr Al Jebel, Khartoum, Red Sea, Gadarif and Kassala States, including both internally displaced and settled populations, were prospectively enrolled in the study from March 1998 to March 1999. The health services where these patients were seen constitute the first level referral and were specialist services
Correspondence to: Asma I El-Sony, National Tuberculosis Programme, Federal Ministry of Health, PO Box 193, PC-11 111, Khartoum, Khartoum, Sudan. Tel: (249) 11-772 182. Fax: (249) 11-774 412. e-mail:
[email protected] Article submitted 9 July 2002. Final version accepted 3 December 2002. [A version in French of this article is available from the Editorial Office in Paris and from the IUATLD Website www.iuatld.org]
Symptoms in diagnosis of PTB in Sudan
dealing with chest diseases. The age range 15–49 was selected because of the high proportion of tuberculosis cases in this range, because the clinical picture of the disease is different in children and because this is the range in which the differentiation from other diseases is most difficult. Cases with a history of previous treatment for tuberculosis were not included in the study. The 5338 patients identified with respiratory tract symptoms, including those identified as ‘TB suspects’, were referred for routine procedures recommended by the NTP. According to the WHO definition, TB suspects are those with cough for more than 3 weeks.10 These patients were categorised as 1) AFB-positive pulmonary tuberculosis, 2) AFB-negative pulmonary tuberculosis, 3) extra-pulmonary tuberculosis, and 4) patients randomly selected from the consecutive patients who were diagnosed as having conditions other than TB (non-TB). Of those patients whose final diagnosis was other than tuberculosis, 800 (15%) were interviewed for comparison with those with tuberculosis. All suspects were referred to clinical officers (doctors or medical assistants) for screening according to standard procedures. TB suspects were first sent to the laboratory for examination of three sputum smears; those who had two or more positive smears were registered and treated for AFB-positive pulmonary tuberculosis; those who were not positive on at least two smears, even after repeated examination, were referred to the medical officer for further investigation and diagnostic decision. Recommended proce-
Table 1
551
dures for further investigation included radiograph; if this was abnormal and the patient had symptoms that did not respond to standard, broad spectrum antibiotics, the patient was registered and treated for smear-negative pulmonary tuberculosis. Clinical course and treatment outcome were monitored in all patients. Non-TB patients were those who remained without a diagnosis of TB after completion of this procedure. For the purpose of this study, a questionnaire was used to collect information on demography, type and duration of symptoms, and the results of sputum microscopy and chest radiography. Oral consent was obtained from all subjects before inclusion in the study. Sputum smear microscopy for AFB was done using Ziehl-Neelsen staining, according to International Union Against Tuberculosis and Lung Disease (IUATLD) and World Health Organization (WHO) recommended standards.10–12 Statistical methods Data were computerised and analysed using SPSS 9.05 (SPSS, Chicago, IL). Categorical data were compared using 2. Models of symptoms for the different types of TB were obtained by performing a multivariate analysis using logistic regression, from which odds ratios corrected for the covariates, with their corresponding 95% confidence intervals, were obtained. A P value of less than 0.05 was considered statistically significant. The study was approved by an ethics review committee prior to its commencement; all procedures conform to internationally approved ethical principles.
Characteristics of consecutive patients presenting to chest clinics in Sudan according to diagnosis
Characteristic Total Age group (years) 15–24 25–34 35–44 45–49 Sex Male Female Residence Rural Urban Neighbourhood Poor Other Occupation Unemployed Student Housewife Salaried Business
Non-TB n (%)
Pulmonary TB n (%)
Smear-positive TB n (%)
Smear-negative TB n (%)
Extrapulmonary TB n (%)
800 (100)
763 (100)
504 (100)
259 (100)
166 (100)
159 (19.9) 342 (42.8) 235 (29.4) 64 (8.0)
258 (37.4) 245 (32.1) 157 (20.6) 76 (10.0)
187 (37.1) 167 (33.1) 99 (19.6) 51 (10.1)
98 (37.8) 78 (30.1) 58 (22.4) 25 (9.7)
43 (25.9) 61 (36.7) 37 (22.3) 25 (15.1)
0.0001
475 (59.4) 325 (40.6)
466 (61.1) 297 (38.9)
321 (63.7) 183 (36.3)
145 (56.0) 114 (44.0)
87 (52.4) 79 (47.6)
0.037
152 (19.0) 648 (81.0)
129 (16.9) 634 (83.1)
81 (16.1) 423 (83.9)
48 (18.5) 211 (81.5)
34 (20.5) 132 (79.5)
0.480
533 (66.6) 267 (33.4)
665 (87.2) 98 (12.8)
449 (89.1) 55 (10.9)
216 (83.4) 43 (16.6)
143 (86.1) 23 (13.9)
0.0001
62 (7.8) 82 (10.3) 154 (19.3) 149 (18.6) 170 (21.3)
81 (10.6) 99 (13.0) 172 (22.5) 66 (8.7) 144 (18.9)
48 (9.5) 65 (12.9) 102 (20.2) 47 (9.3) 106 (21.0)
33 (12.7) 34 (13.1) 70 (27.0) 19 (7.3) 38 (14.7)
21 (12.7) 20 (12.0) 47 (28.3) 6 (3.6) 34 (20.5)
0.0001
* P value indicates a comparison of the diagnostic groups for each of the characteristics.
P
552
The International Journal of Tuberculosis and Lung Disease
Figure Distribution of longest duration of symptoms in patients presenting to chest clinics in Sudan by diagnostic category.
RESULTS In the period of the study, 16 735 patients (52.6% males and 47.4% females) presented at respiratory centres in seven states in Sudan; 5338 had respiratory tract symptoms, 2917 (54.6%) men and 2421 (45.4%) women, with a mean age of 32 years. Of these 5338, 17.4% (929) were diagnosed as tuberculosis cases: 763 were pulmonary cases, of whom 504 were sputum smear-positive and 259 sputum smear-negative. A further 166 patients had extra-pulmonary tuberculosis: 97 had tuberculous lymphadenitis, 59 tuberculous pleurisy, five skeletal tuberculosis, three genitourinary tuberculosis and two abdominal tuberculosis. A further 800 patients (15%) of those whose final diagnosis was other than tuberculosis were interviewed for comparison with those with tuberculosis. Table 1 presents the personal and social characteristics of the participants according to their diagnostic group. The groups were significantly different in distribution according to age, sex, socio-economic level of residential area and occupation. They were not different in terms of residence in urban or rural areas.
Table 2
The differences were greatest between patients with and without tuberculosis; patients with tuberculosis had closely similar characteristics. Compared with non-tuberculosis patients and other forms of tuberculosis, pulmonary cases were more likely to be young (aged 15–24 years), and to live in poor neighbourhoods. Extra-pulmonary patients were more likely to be older (45–54 years), female, and living in poor neighbourhoods. The longest duration of symptoms according to diagnostic group is shown in the Figure. Non-tuberculosis patients had significantly shorter durations of symptoms than any of the groups with tuberculosis, for whom the duration of symptoms was not significantly different (P 0.05, 2 test). Because sputum smear examination is indicated in ‘suspects’, who are defined as having symptoms for a minimum of 3 weeks, comparison was made among the groups, excluding all cases with a duration of symptoms of less than 3 weeks. The same differences were observed as when the whole group of patients was compared, with the same results of statistical analysis.
Frequency of symptoms in consecutive patients presenting to chest clinics in Sudan according to diagnosis
Symptom Chest symptoms Cough Chest pain Shortness of breath Haemoptysis Systemic symptoms Weight loss Tiredness Fever Night sweats Loss of appetite
Non-TB n (%)
Pulmonary TB n (%)
Smear-positive TB n (%)
Smear-negative TB n (%)
Extrapulmonary TB n (%)
733 (91.6) 284 (35.5) 77 (9.6) 17 (2.1)
721 (94.5) 440 (57.7) 571 (74.8) 151 (19.8)
492 (97.6) 294 (58.3) 387 (76.8) 96 (19.0)
229 (88.4) 146 (56.4) 184 (71.0) 55 (21.2)
70 (42.2) 79 (47.6) 66 (39.8) 11 (6.6)
106 (13.3) 101 (12.6) 114 (14.3) 33 (4.1) 43 (5.4)
699 (91.6) 651 (85.3) 597 (78.2) 479 (62.8) 371 (48.6)
461 (91.5) 431 (85.5) 396 (78.6) 319 (63.3) 254 (50.4)
238 (91.9) 220 (84.9) 201 (77.6) 160 (61.8) 117 (45.2)
129 (77.7) 122 (73.5) 113 (68.1) 94 (56.6) 66 (39.8)
Symptoms in diagnosis of PTB in Sudan
Table 3
553
Significant predictors of pulmonary TB*
Variable Cough Breathlessness Chest pain Haemoptysis Age Sex Constant
B
SE
P
OR
95%CI
0.387 2.290 1.271 2.001 0.043 0.316 0.526
0.240 0.152 0.130 0.290 0.008 0.132 0.389
0.107 0.0001 0.0001 0.0001 0.0001 0.017 0.176
0.679 9.871 3.564 7.444 0.958 0.729
0.424–1.087 7.326–13.301 2.762–4.601 4.229–13.103 0.943–0.972 0.563–0.945
* Pulmonary TB vs. non-TB. B beta coefficients; SE standard error for beta coefficients; OR odds ratio; CI confidence interval.
Table 2 displays the distribution of symptoms according to diagnostic group. Cough was frequent among all chest patients, while other chest symptoms were present in a minority of patients with conditions other than tuberculosis (2.1–35.5%). Among pulmonary tuberculosis patients, the majority complained of shortness of breath (74.8%) and chest pain (57.7%), and a smaller proportion of haemoptysis (19.8%). Systemic symptoms were infrequent among patients with conditions other than tuberculosis (4.1– 14.3%) compared with pulmonary tuberculosis patients, among whom systemic symptoms were very frequent (weight loss 91.6%, tiredness 85.3%, fever 78.2%, night sweats 62.8%, loss of appetite 48.6%). While these symptoms were similarly frequent among smear-positive and smear-negative pulmonary tuberculosis patients, extra-pulmonary patients had substantially lower frequencies of all chest symptoms and somewhat lower rates for systemic symptoms. Table 3 presents the odds ratios (OR), with their respective confidence intervals (95%CI), from logistic regression analysis for pulmonary tuberculosis as compared with patients with other conditions, for the various chest symptoms adjusted for age and sex. Pulmonary tuberculosis patients were significantly less likely to be female (OR 0.73, 95%CI 0.56–0.95) and to be older (OR 0.96, 95%CI 0.94–0.97). Cough was not significantly more frequent in either of the groups. Every other chest symptom was more frequent among pulmonary tuberculosis patients (OR 9.87 for shortness of breath, 7.44 for haemoptysis and 3.56 for chest pain). Table 4 presents the results of a similar analysis for
Table 4
systemic symptoms, adjusted for the presence of chest symptoms other than cough, and for age and sex. The OR for other chest symptoms was substantially reduced (OR 1.65, 95%CI 1.08–2.53) when systemic symptoms were added to the equation. Pulmonary tuberculosis was significantly less frequent with advancing age (OR 0.96, 95%CI 0.94–0.99), but was not different from non-tuberculosis with respect to sex. Each of the systemic symptoms was a significant predictor of pulmonary tuberculosis (OR 13.80 for weight loss, 6.53 for tiredness, 4.22 for night sweats, 2.16 for loss of appetite, 2.15 for fever). Logistic regression analysis was used to estimate OR and 95%CI for symptoms as predictors of smearpositive as compared with smear-negative pulmonary tuberculosis. The two groups were not different in age or sex. Of all the symptoms, only cough was a significant predictor of smear-positive tuberculosis (OR 5.37, 95%CI 2.70–10.68).
DISCUSSION The diagnosis of active TB in developing countries still largely depends on clinical suspicion and radiographic findings where applicable and available. Microscopy of sputum smears for AFB remains the standard diagnostic procedure;2,6,13 in most developing countries culture is not available for routine diagnosis. Reducing the transmission of Mycobacterium tuberculosis depends crucially on prompt diagnosis and treatment, and every effort is necessary to reduce diagnostic delay.14–18 Clinical assessment of patients
Significant predictors of pulmonary TB*
Variable Any chest symptom except cough Weight loss Tiredness Fever Night sweats Loss of appetite Age Sex Constant * Pulmonary TB vs. non-TB.
B
SE
P
OR
95%CI
0.499 2.625 1.876 0.763 1.441 0.770 0.038 0.221 2.184
0.218 0.207 0.206 0.216 0.259 0.252 0.012 0.205 0.503
0.022 0.0001 0.0001 0.0001 0.0001 0.002 0.0001 0.281 0.0001
1.647 13.800 6.528 2.145 4.224 2.160 0.963 0.802
1.075–2.525 9.204–20.692 4.364–9.776 1.403–3.278 2.541–7.020 1.319–3.537 0.941–0.985 0.537–1.198
554
The International Journal of Tuberculosis and Lung Disease
in many locations varies considerably between teaching hospitals and peripheral health units. As the programme in most developing countries depends on paramedics,11,12 it is important to aid the health care worker to identify the patients most likely to be suffering from tuberculosis for the diagnosis to be made as rapidly as possible. This implies a clear understanding of the clinical characteristics that differentiate tuberculosis patients from other cases presenting to the health service, primarily with respiratory symptoms. Few scientific investigations have compared the clinical characteristics of unselected tuberculosis patients from those in the same first level referral chest service without tuberculosis, with the result that, although the ‘tuberculosis suspect’ has been defined for operational purposes, this definition is not based on an extensive body of scientific evidence. Unfortunately, in the setting of this study, it was not possible to confirm that the sputum smear-negative patients were indeed tuberculosis patients, as bacteriological culture facilities were not available. Thus, it is not possible to be certain that these patients were indeed tuberculosis patients, nor that patients designated as ‘non-TB’ did not have tuberculosis. The clinical course of the patients suggest that they were diagnosed correctly: those designated as tuberculous improved on treatment, while the state of those designated ‘non-TB’ did not progress in the absence of specific anti-tuberculosis chemotherapy. Nonetheless, it is not possible to be certain of these diagnoses, although this is the reality of programme conditions in low-income countries. In addition, the results of this study cannot necessarily be applied to communities where HIV infection is frequent. Similar studies must be undertaken in such communities to determine the complex of symptoms among such patients and the ability to distinguish tuberculosis from other diagnoses in such a setting. Diagnosis of patients with pulmonary tuberculosis whose sputum is smear-negative poses a challenge to clinical services and to public health programmes in low-income countries. This study has shown that, in the setting of the chest clinic in a community with a low prevalence of HIV infection, the constellation of symptoms associated with tuberculosis might be termed ‘cough of more than 3 weeks plus . . .’. While cough was very frequent among all patients in the clinic, it was patients who had cough, and in particular systemic symptoms, who were most likely to have tuberculosis. The accompanying symptoms of greatest predictive significance were weight loss, tiredness and night sweats. This was particularly the case among the patients studied if the symptoms were of long duration.
Using this combination of symptoms should assist health practitioners in increasing the efficiency of diagnostic examination and in improving early detection of patients at the first level referral service. Practitioners at the primary care level must maintain a high index of suspicion among all patients who present with prolonged cough, and investigate such patients thoroughly for the presence of tuberculosis. References 1 World Health Organization. Tuberculosis. Fact Sheet 104. Geneva: WHO, 2000. 2 Murray C J L, Styblo K, Rouillon A. Tuberculosis in developing countries: burden, intervention and cost. Bull Int Union Tuberc Lung Dis 1990; 65 (1): 6–24. 3 World Health Organization Global Tuberculosis Programme. A deadly partnership: tuberculosis in the era of HIV. WHO/ TB/96.204. Geneva: WHO, 1996. 4 World Health Organization—Programme of Tropical Diseases Research. Why tuberculosis research? Geneva: WHO, 2000. 5 WHO Report 2002. Global Tuberculosis Control, Surveillance, Planning, Financing. WHO/CDS/TB/2002.295. Geneva: WHO, 2002. 6 Tessema T A, Bjune G, Assefa G, Bjorvatn B. An evaluation of the diagnostic value of clinical and radiological manifestations in patients attending the Addis Ababa tuberculosis centre. Scand J Infect Dis 2001; 33: 355–361. 7 Aung W W, Nyein M M, Ti T, Maung W. Improved method of direct microscopy for detection of acid-fast bacilli in sputum. Southeast Asian J Trop Med Public Health 2001; 32: 390–393. 8 Bruchfeld J, Aderaye G, Palme I B, Bjorvatn B, Kallenius G, Lindquist L. Sputum concentration improves diagnosis of tuberculosis in a setting with a high prevelance of HIV. Trans R Soc Trop Med Hyg 2000; 94: 677–680. 9 Federal Ministry of Health. Sudan National AIDS Programme. HIV/AIDS Surveillance Report. Khartoum, Sudan: Federal Ministry of Health, 2000. 10 World Health Organization, Global Tuberculosis Programme. Managing Tuberculosis at District Level. A training course. Geneva: WHO, 1994. 11 Enarson D A. The International Union Against Tuberculosis and Lung Disease model National Tuberculosis Programmes. Tubercle Lung Dis 1995; 76: 95–99. 12 International Union Against Tuberculosis and Lung Disease. Technical guide. Sputum examination for tuberculosis by direct microscopy in low income countries, Paris: IUATLD, 2000. 13 Samb S, Henzel D, Daley C L, et al. Methods for diagnosing tuberculosis among in-patients in Eastern Africa whose sputum smears are negative. Int J Tuberc Lung Dis 1997; 1: 25–30. 14 Jaramillo E. Pulmonary tuberculosis and health seeking behaviour: how to get a delayed diagnosis in Cali, Colombia. Trop Med Int Health 1998; 3: 138–144. 15 Bayer R, Dupuis L. Tuberculosis, public health, and civil liberties. Ann Rev Public Health 1995; 16: 307–326. 16 Gie R P, Beyers N, Schaaf H S, Donald P R. Missed opportunities in the diagnosis of pulmonary tuberculosis in children. S Afr Med J 1993; 83: 263. 17 Katz I, Rosenthal T, Michaeli, D. Undiagnosed tuberculosis in hospitalised patients. Chest 1985; 87: 770–774. 18 Vaughan J P, Morrow R H. Manual of epidemiology for district health management. Geneva: WHO, 1989.
Symptoms in diagnosis of PTB in Sudan
555
RÉSUMÉ O B J E T : Décrire les modifications des caractéristiques cliniques en fonction du diagnostic final chez les sujets se présentant à un service de santé avec des symptômes thoraciques. M É T H O D E S : Parmi 16.735 patients (52,6% d’hommes et 47,4% de femmes) se présentant dans les centres respiratoires de sept Etats du Soudan, il y a eu 5.338 patients avec des symptômes respiratoires, dont 2.917 hommes (54,6%) et 2.421 femmes (45,4%) avec un âge moyen de 32 ans. Ceux dont la toux durait depuis plus de 3 semaines et qui n’avaient pas répondu à une cure d’antibiotiques ont fait l’objet d’un dépistage par examen microscopique de deux ou trois échantillons d’expectoration et par cliché thoracique.
Chez 504 (9,44%), on a trouvé une TB pulmonaire à bacilloscopie positive, chez 259 (4,85%) une TB pulmonaire à bacilloscopie négative, chez 166 (3,11%) une tuberculose extrapulmonaire parmi lesquels 59 (1,11%) avaient des atteintes pleurales, et finalement 4.409 (82,6%) cas non tuberculeux. C O N C L U S I O N : Les maladies tuberculeux présentaient une constellation de symptômes, dont la principale était la toux depuis plus de 3 semaines. Les autres symptômes avec la plus grande signifiance étaient la perte de poids, la fatigue et les sueurs nocturnes. RÉSULTATS :
RESUMEN O B J E T I V O : Describir las variaciones de las características clínicas de los pacientes que consultan en un establecimiento de salud por síntomas respiratorios, en función del diagnóstico final. M É T O D O S : De 16.735 pacientes (52,6% hombres y 47,4% mujeres) que consultaron en centros respiratorios de siete estados de Sudán, 5.338 habian síntomas respiratorios. De éstos, 2.917 eran hombres (54,6%) y 2.421 eran mujeres (45,4%) y la edad promedio era de 32 años. Aquéllos que presentaban tos que duraba más de 3 semanas sin respuesta a un tratamiento antibiótico fueron sometidos a un examen microscópico de 2 ó 3 muestras de esputo y a radiografía de tórax.
Se encontró una tuberculosis pulmonar con baciloscopia positiva (PTB) en 504 (9,44%) pacientes, en 259 (4,85%) una PTB con baciloscopia negativa y en 166 (3,11%) una TB extrapulmonar, de los cuales 59 (1,11%) tenían un compromiso pleural. Finalmente, no se encontró una TB en 4.409 (82,6%) sujetos. C O N C L U S I O N E S : Los pacientes tuberculosos presentaban una constelación de síntomas, de los cuales el más importante fue el tos que duraba más de 3 semanas. Los otros síntomas de significación más grande fueron pérdida de peso, fatiga y sudores nocturnos. RESULTADOS :
