, Volume 41, Issue 3, pp 737–739

Clinical and laboratory features of tuberculosis within a hospital population in Libreville, Gabon


  • D. U. Kombila
    • Centre Hospitalier de Libreville
  • J.-B. Moussavou-Kombila
    • Centre Hospitalier de Libreville
    • Medical Research UnitAlbert Schweitzer Hospital
    • Institute of Tropical MedicineUniversity of Tübingen
    • Center for Tropical Medicine and Travel Medicine, Department of Infectious Diseases, Division of Internal MedicineAcademic Medical Center, University of Amsterdam
  • B. Lell
    • Medical Research UnitAlbert Schweitzer Hospital
    • Institute of Tropical MedicineUniversity of Tübingen

DOI: 10.1007/s15010-012-0383-x

Cite this article as:
Kombila, D.U., Moussavou-Kombila, J., Grobusch, M.P. et al. Infection (2013) 41: 737. doi:10.1007/s15010-012-0383-x

Tuberculosis constitutes a substantial health problem in Africa [1], particularly in the face of the human immunodeficiency virus (HIV) co-pandemic and the increasing problem of drug-resistant Mycobacterium tuberculosis strains [2]. While epidemiological data of adequate quality and quantity on the disease is available from some regions of sub-Saharan Africa, there is, so far, little data of high quality and visibility at hand for Central Africa [3]. For example, published epidemiological studies from Gabon are mainly available exclusively in French, and only from special subgroups, such as hospitalised children [4, 5] and HIV-positive adults [6], or from special clinical manifestations, such as tuberculous meningitis [7].

The capital Libreville is the largest city of Gabon and has about 450,000 inhabitants. The majority of tuberculosis patients are referred to one of two institutions. Some data from Nkembo Hospital has been published previously [6]. However, in recent years, the “Ward A” of the Internal Medicine Department of the General Hospital (Centre Hospitalier de Libreville, CHL) has started to receive a large proportion of the adult tuberculosis patients of the city.

The objective of this work is to describe the clinical characteristics of tuberculosis in this service over a period of 6 years from January 2004 to December 2009.

In this ward, all subjects with a suspicion of tuberculosis are hospitalised for extended clinical and laboratory examination and the initiation of treatment. The catchment area is greater Libreville and, to a certain extent, the whole country. The ward receives patients above the age of 12 years.

Demographic, clinical and laboratory information was gathered from the patient files of all patients hospitalised with suspected tuberculosis. Data from return visits of patients already captured earlier in the database were not used. Table 1 provides a synopsis of the key findings.
Table 1

Characteristics of hospitalised tuberculosis patients at Centre Hospitalier de Libreville (CHL) over a period of 5 years








Hospitalised cases







Tuberculosis casesa

30 (16 %)

58 (23 %)

93 (26 %)

83 (21 %)

58 (19 %)

322 (21 %)

Age in yearsb

33 (12)

34 (12)

37 (16)

32 (12)

33 (14)

34 (14)

Female sexc

20 (67 %)

29 (50 %)

35 (38 %)

47 (57 %)

34 (59 %)

165 (51 %)

Pulmonary tuberculosisc

25 (83 %)

41 (71 %)

63 (68 %)

54 (65 %)

37 (64 %)

220 (68 %)


19 (67 %)

27 (43 %)

32 (37 %)

30 (34 %)

25 (43 %)

133 (47 %)


17 (59 %)

18 (31 %)

22 (26 %)

23 (31 %)

17 (31 %)

97 (32 %)

aNumber and percentage among all hospitalised

bMean (standard deviation, SD)

cNumber and percentage among all tuberculosis patients; results for Ziehl–Neelsen and HIV were not available for all subjects

Direct diagnosis was by Ziehl–Neelsen-stained sputum specimen, pleura exsudate or tissue biopsies. Distinguishing tuberculous from non-tuberculous mycobacteria (such as M. africanum) was not possible in the present analysis. At the time this study was conducted, the hospital did not have the possibility to perform cultures, and it still has not, nor have molecular diagnostic methods also serving as drug resistance sentinels such as the World Health Organization (WHO)-propagated GeneXpert® technique [8] been installed.

Clinically, the cases were classified as either pulmonary, extra-pulmonary or a combination of both. No case of disseminated tuberculosis was recorded. The diagnostic certainty was graded as either definitive (microscopical detection of acid-fast bacilli) or probable (clinical, radiological and/or histological signs suggestive of tuberculosis without microbiological confirmation; and the attending clinician’s decision to initiate tuberculosis treatment).

An HIV test was performed in all subjects using Determine HIV-1/2 (Abbott Laboratories, Abbott Park, IL, USA) and confirmed with ImmunoComb II HIV 1&2 (PBS Orgenics, Courbevoie, France). A total of 1,508 subjects were hospitalised in the ward during the observation period, of which 322 patients (21 %) were treated for tuberculosis. Of those, 133 cases were considered to be definitive and 189 probable.

Pulmonary tuberculosis was the most prevalent manifestation, with 220 cases (68 %). There were 155 cases (48 %) of extrapulmonary tuberculosis, consisting of 84 cases of pleural tuberculosis, 44 cases of peritoneal tuberculosis, 34 cases of tuberculous lymphadenitis, ten cases of genitourinary tuberculosis and four cases of osteoarticular tuberculosis. Seventy patients (22 %) had tuberculosis in more than one site, with the majority of these having concomitant pulmonary and pleural tuberculosis, which was present in 22 cases (7 %).

An HIV test was performed in 302/322 (94 %) patients, of which 97 (32 %) were positive. CD4+ counts were available for 64 (66 %) of the 97 HIV-positive patients. Of these counts, 56 (88 %), 46 (72 %) and 23 (36 %) had counts below 350, 200 and 50 cells/μL, respectively.

The number of cases over time is shown in the table. During these 5 years of observation, the proportion of tuberculosis cases remained fairly stable, representing between 16 and 26 % of all hospitalised patients, without apparent trend over time. The fairly high variation in the number of hospitalisations between calendar years may be explained by disruptions of activity due to a number of strikes during the observation period.

Follow-up data containing duration of treatment were available for 213 patients. Eleven subjects (5 %) died within 6 months. Among the remaining subjects, only 79 (39 %) finished the treatment course of 6 months.

The cases described probably represent a majority of all tuberculosis patients diagnosed in Libreville. With the exception of the second largest town, Port-Gentil, the rest of the country has rural or semi-urban character. Given the lack of epidemiological data from rural areas, it is unclear as to what extent the characteristics of the population in the present analysis are representative of the whole country. Epidemiological data collection for Gabon is coordinated by the National Anti-tuberculosis Programme (Programme National de Lutte Contre La Tuberculose). However, the data quality for Gabon is considered to be poor [9] and tools to combat tuberculosis on all levels, from implementation of infection control measures over laboratory facilities including resistance testing and a functioning public health system to follow-up of outpatients, are limited.

Initiatives from academia and other non-governmental institutions are needed in order to obtain a complete picture of the tuberculosis situation within the country and the larger region as a basis to fully understand the magnitude of the problem, and to tackle it appropriately. This is especially urgent in light of the high proportion of treatment drop-outs found in our study, as well as the emergence of multi- and extended drug resistance against tuberculosis in Libreville, as anecdotally reported recently [10], and from other parts of the country, as currently assessed, for example, in the Moyen-Ogoué region around the town of Lambaréné (U. Kombila et al., unpublished data). Routine drug sensitivity testing and the feedback of the resistance profile to the treating physician within a reasonable amount of time must be considered as a high-priority target within tuberculosis control goals across Africa.

Conflict of interest

None of the authors has any kind of conflict of interest, including financial, to disclose.

Copyright information

© Springer-Verlag Berlin Heidelberg 2012