SEARCH

SEARCH BY CITATION

Keywords:

  • Ethiopia;
  • tuberculosis;
  • pastoralists;
  • diagnostic yield;
  • operational research

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Objective

To assess the number of sputum specimens necessary for a reliable diagnosis of pulmonary tuberculosis (PTB) in a pastoralist population in Ethiopia.

Method

Using routine data from Ethiopia, where three sputum specimens are currently recommended for the diagnosis of PTB, we documented, (i) the proportion of persons with suspected, PTB who submitted a first, second and third sputum specimen for smear examination and (ii) the incremental smear-positive yield from the first, to the second and third specimens.

Results

Of 505 persons with suspected PTB, 107 (22%) failed to submit three samples. Of 60 patients who submitted three sputum samples with at least one smear-positive sample, the first sputum sample was smear positive in 56 (93%) cases; the second sputum sample was the first to be positive in 3 (5%) cases and in only one case was the third sample the first to be smear positive (additional yield 2%).

Conclusion

In a pastoralist setting, a reliable diagnosis of PTB can be achieved with two sputum specimens and PTB diagnosis may be adequate with just one sputum specimen. However, if this more radical approach was adopted, ways of increasing diagnostic sensitivity should be explored.

Objectif

Evaluer le nombre d’échantillons de crachats nécessaires pour un diagnostic fiable de la tuberculose pulmonaire (TBP) dans une population pastorale en Ethiopie.

Méthode

A l'aide des données de routine de l’Ethiopie, où trois échantillons de crachats sont actuellement recommandés pour le diagnostic de la TBP, nous avons documenté (i) la proportion de personnes ayant une suspicion de TBP qui ont soumis un premier, un deuxième et un troisième échantillon d'expectoration pour l'examen de frottis et (ii) le rendement additionnel en frottis positif, en allant du premier au deuxième et au troisième spécimens.

Résultats

Sur 505 personnes ayant une suspicion de TBP, 107 (22%) n'ont pas fourni 3 échantillons. Sur 60 patients qui ont fourni 3 échantillons de crachats avec au moins un échantillon à frottis positif, le premier l’échantillon de crachat avait un frottis positif dans 56 (93%) cas, le second échantillon de crachat a été le premier à être positif dans 3 (5%) cas et dans un seul cas, le troisième échantillon a été le premier à frottis positif (rendement additionnel de 2%).

Conclusion

Dans un cadre pastoral, un diagnostic fiable de la TBP peut être réalisé avec deux échantillons d'expectoration et le diagnostic de la TBP peut suffire avec un simple échantillon d'expectoration. Toutefois, si cette approche plus radicale devrait être adoptée, des moyens pour augmenter la sensibilité du diagnostic devraient être explorés.

Objetivo

Evaluar el número de muestras de esputo necesarias para el diagnóstico fiable de la tuberculosis pulmonar (TBP) en una población de pastores trashumantes en Etiopía.

Método

Utilizando datos de rutina de Etiopía, en donde actualmente se recomiendan tres muestras de esputo para el diagnóstico de la TBP, hemos documentado (i) la proporción de personas con sospecha de TBP que entregaron la primera, segunda y tercera muestra de esputo para realizar una baciloscopia y (ii) el incremento en baciloscopias de esputo positivas de la primera, a la segunda y a la tercera muestra.

Resultados

De 505 personas con sospecha de TBP, 107 (22%) no entregaron las 3 muestras. De 60 pacientes que entregaron las 3 muestras de esputo y con al menos una muestra con baciloscopia positiva, la primera muestra de esputo era baciloscopia positiva en un 56 (93%) de los casos; la segunda muestra de esputo era la primera en dar un resultado positivo en 3 (5%) casos y solo en un caso fue la tercera muestra la primera en ser positiva (ganancia adicional 2%).

Conclusión

En un emplazamiento con una población trashumante, el diagnóstico fiable de la TBP puede hacer con dos muestras de esputo e incluso el diagnóstico de TBP puede ser adecuado con solo una muestra. Sin embargo, si se adoptase este enfoque más radical, deberían explorarse maneras de mejorar la sensibilidad diagnóstica.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Ethiopia has one of the highest burdens of tuberculosis (TB) in the world (WHO 2007), and this burden is disproportionately high in the Somali Regional State (SRS) (Barr & Menzies 1994; Agutu 1997; Waldman 2001; Coninx 2007). While weak TB control activities in the region may be to blame (Sheik-Mohamed & Velema 1999; Gele et al. 2009) – leading to many TB cases going undiagnosed and untreated – the situation is made more challenging by the type of population which lives there. Most of the inhabitants are pastoralists, and conventional TB management strategies are not well adapted to this mobile population (Tayler-Smith et al. 2011).

In 2010, Médecins Sans Frontiéres (MSF), in collaboration with the Somali Regional Health Bureau (SRHB), began providing TB care and treatment in Imey district in the SRS. Diagnosis of pulmonary TB (PTB) was based on light microscopy examination of three sputum specimens, each specimen submitted on three separate occasions. There has since been concern that this need for multiple visits, particularly in a pastoralist setting, may lead to high loss to follow-up, compounding delays in TB diagnosis and treatment. This has raised the question: is the current recommendation of three sputum specimens for the diagnosis of PTB appropriate in this setting?

Using routine data from the MSF TB programme in Imey, Ethiopia, we thus report on: i) the proportion of persons with suspected PTB who submitted a first, second and third sputum specimen for smear examination and ii) the incremental smear-positive yield going from the first, to the second and third specimen in those who submitted three sputum specimens.

Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Design

This was a retrospective cross-sectional study using routine programme data.

Setting and population

The study was conducted in Imey, a rural district in the eastern part of the SRS, where MSF, in collaboration with the SRHB, has been providing free comprehensive TB care since 2010. Integrated TB services are provided through a primary healthcare facility situated in Imey town serving a target population of about 65 000 and consisting mainly of pastoralists. This study included all adults (≥15 years) with suspected PTB referred for sputum smear microscopy between December 2010 and March 2012.

TB diagnosis

Diagnosis of PTB was based on clinical examination and microscopic examination of smears made from three sputum specimens submitted on three occasions. PTB was only diagnosed if two or more sputum smears tested positive for acid-fast bacilli (AFB). Internal quality control measures are in place, and external quality control is performed quarterly in an external reference laboratory.

Data collection and statistical analysis

Data were sourced from the TB laboratory registers and double entered, by two independent encoders, into a data entry file created using EpiData Entry software version 3.1 (EpiData Association, Odense, Denmark). The two data files were compared and discordances resolved by cross-checking with the paper registers. The following variables were collected: laboratory serial number, age, sex, number of sputum samples submitted and AFB status (positive or negative) of each sputum sample submitted. To calculate the incremental smear-positive yield, only those who submitted three sputum specimens and were AFB positive on at least one of the specimens were included. Data were analysed using Epi Data analysis software version 2.2 (EpiData Association, Odense, Denmark).

Ethics approval

The study satisfied the criteria for analysis of routine data by the MSF Ethics Review Board, and formal approval was also obtained from the SRHB in Ethiopia.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

There were 505 persons with suspected PTB, of whom 278 (55%) were female. Median age was 38 years (interquartile range, 25–50 years). The number and proportion of persons with suspected PTB who submitted a first, second and third sputum specimen for smear examination are shown in Figure 1. There were losses at each stage, with a total of 107 (21%) patients failing to submit the three recommended sputum specimens. 398 (79%) patients submitted three samples, 69 (14%) submitted 2 and 31 (6%) submitted one sputum specimen. Seven patients submitted none.

image

Figure 1. Submission of sputum specimens by persons with suspected pulmonary tuberculosis (PTB), Imey, Ethiopia.

Download figure to PowerPoint

Altogether, 72 (14%) of the 505 suspected patients with PTB were sputum smear positive. Of 398 patients who submitted three sputum specimens, 60 (15%) were sputum smear positive. The incremental yield of smear-positive results going from the first, to the second and the third sputum specimen is shown in the Table 1. The first specimen identified 56 [93% (95% CI 84–97)] smear-positive patients, the second identified 3 [5% (95% CI 2–14)] and the third identified 1 patient [2% (95% CI 0.3–9)].

Table 1. Incremental yield in smear positivity among patients submitting three sputum specimens with at least one smear positive for acid-fast bacilli (n–60), Imey, Ethiopia
Sputum sampleIncremental yielda
First AFB positive n (%)95% CI
  1. AFB, acid-fast bacilli; CI, confidence interval.

  2. a

    Defined as a positive AFB result on the second or the third specimen when the previous specimens were negative.

First56 (93)85–97
Second3 (5)2–14
Third1 (2)0.3–9
Total60 (100)

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

This is one of the first studies assessing loss to follow-up (LTFU) related to a policy of requesting three sputum specimens for the diagnosis of PTB in a pastoralist setting in Ethiopia. The LTFU rates were unacceptably high. In those who submitted all three recommended sputum specimens, the incremental yield from the second and third smears was very low.

The study's strengths are that (i) the data come from a programme setting and thus are likely to reflect the operational reality on the ground, (ii) laboratory staff are well trained and supervised, (iii) quality control measures are in place and routinely performed and we believe that the results of the sputum microscopy are robust, and (iv) we adhered to the STROBE guidelines on the reporting of observational studies (Von Elm et al. 2007). The main study limitation is the relatively small sample size used to calculate diagnostic yield.

Our findings raise some important issues. First, while the majority of persons presenting with suspected PTB in our setting submit at least two sputum samples, more than one in five are lost to follow-up before submission of a third sample. This is a missed opportunity for diagnosing and treating potential cases of PTB.

Second, our findings suggest that the yield from the third sample is negligible considering that 98% of AFB-positive cases were identified by the first two specimens, which corroborates the findings from other studies (Ipuge et al. 1996; Mase et al. 2007). This finding is also in line with current WHO recommendations for submission of only two sputum specimens for the diagnosis of PTB (WHO 2010). However, the incremental yield gained from the second sputum sample in this setting was small. There are thus two operational options in this setting. The first is to rely on a single sputum smear and start TB treatment on the basis of one positive AFB smear – this recommendation for start of treatment is also in line with current WHO guidelines (WHO 2010). The second is to rely on more rapid, effective and advanced sputum screening techniques that would maximise diagnostic sensitivity with the first sputum submission. Results from a recent study suggest that PTB diagnosis using light emitting diode (LED) fluorescence microscopy (FM) for the examination of two smears prepared from one sputum sample may further improve the efficiency and accuracy of evaluation of patients with presumptive PTB (Cattamanchi et al. 2011). Such an approach would seem worthwhile in our setting, as it would not only help to reduce LTFU rates associated with additional visits for sputum submission, but would also reduce service and patient costs.

In conclusion, in a pastoralist setting in Ethiopia, our findings suggest that PTB diagnosis can be achieved with two sputum specimens and may be adequate with just one specimen. However, if this more radical approach was adopted, ways of increasing diagnostic sensitivity should be explored.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

We are particularly grateful to the MSF staff in Imey district for their hard work, and we would also like to thank the Somali Regional Health Bureau for their support and collaboration. The study was funded by MSF Brussels.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References