Prevalence of pulmonary tuberculosis among patients presenting with cough of any duration in Addis Ababa, Ethiopia

Background The current practice in Ethiopia to diagnose tuberculosis is screening patients with cough for at least two weeks. A health facility based study was conducted to estimate the prevalence of smear and culture positive pulmonary TB among patients presenting with cough ≥2 weeks and <2 weeks in Addis Ababa, Ethiopia. Methods A cross-sectional study design was used to recruit patients with cough of any duration from four selected health centers in Addis Ababa, between August and December 2016. Sputum samples were collected from patients reporting productive cough of any duration and screened for Pulmonary Tuberculosis (PTB) using smear microscopy and culture methods. Mycobacterium tuberculosis isolates obtained from culture positive samples were characterized using RD9 deletion typing. Results Majority (39.7%) of the 725 study participants was in the age range of 20-30 years, and 5.0% were smear positive using smear microscopy. The prevalence of smear positive PTB among patients presented with cough duration of ≥2 weeks was significantly higher compared to those patients presented with cough duration of <2 weeks (10.9% versus 0.7%; χ2=38.98; p=0.001). Using culture method, a total of 86 (11.9%) participants were positive for mycobacteria, and the prevalence (14.6%) of PTB among patients presented with cough duration of ≥2 weeks was not significantly higher compared to prevalence (9.9%) in those patients presented with cough duration of <2 weeks (χ2=3.63; p=0.057). Molecular characterization of 86 culture positive mycobacterial isolates showed that 41 were infected with Mtb; 19(46.3%) from those who had cough duration of <2 weeks and 22(53.7%) from those who had cough duration of ≥2 weeks. Conclusion Screening of PTB using smear microscopy alone and cough duration of at least two weeks would negatively affect early diagnosis and treatment initiation in a considerable number of PTB patients who reports cough duration of <2 weeks with the potential of contributing to the spread of TB. Therefore, screening of patients with cough of any duration using both smear microscopy and culture methods is likely to contribute to the success of any effort towards the control of TB.


Background
Tuberculosis (TB) is responsible for ill health among millions of people each year in low income countries [1]. An estimated 10.4 million new TB cases, 1.3 million TB deaths and an additional 0.37 million deaths as a result of TB disease among HIV-infected people were reported globally in 2016 [1]. Most of the estimated number of incident cases in 2016 occurred in the WHO South-East Asia Region (45%), the WHO African Region (25%) and the WHO Western Pacific Region (17%) [1]. However, TB mortality rate is falling at about 3% per year and TB incidence is falling at about 2% per year [2].
In Ethiopia various efforts have been made to control TB since 1950s, when TB was recognized as a major public health problem [3]. Despite all these national and international efforts, TB still remains one of the major public health problems in the country [1].
The recommended strategy to control TB in low income countries including Ethiopia, where 95% of the TB cases occur, is to detect and promptly treat smear-positive cases [4]. It is known that delayed diagnosis results in more extensive disease, more complications and leads to a higher mortality [5]. It also leads to an increased period of infectivity in the community [6].
To prevent further spread of infection among families and communities, early detection of PTB cases is very important in the TB control [6]. Currently, screening for TB in Ethiopia is limited to patients presenting with cough lasting at least for 2 weeks. On the other hand, a long delay in the diagnosis of TB is a well-documented problem in the management of TB [7][8][9]. Previous studies also indicated that excluding patients who reported cough lasting less than 2 weeks from screening for TB leads to continued transmission to others and delayed diagnosis [10,11]. Therefore, this study was conducted to provide additional evidence on the prevalence of PTB among patients presented with cough duration of <2 weeks and among patients presented with cough duration of ≥ 2 weeks using smear microscopy and culture methods. Study participants were recruited among patients attending four health centers of Addis Ababa, Ethiopia.

Study area and participants
This cross sectional survey was conducted from August to December 2016 in four selected health centers of Akaki kality and Nefas silk Lafto Sub-cities in Addis Ababa. Addis Ababa is the capital and chartered city of Ethiopia. Administratively, Addis Ababa has 10 sub-cities which are further divided into 116 districts. Akaki Kality and Nifas Silk-Lafto are two of the 10 subcities of Addis Ababa in which the current study health centers are located.
According to the 2017 population projection of Ethiopia, the total population of Akaki Kality subcity is 227,182 and that of Nefassilk Lafto sub-city is 396,486 [12]. The study participants consisted of individuals whose age is above 12 years presented to the four health centers and considered to be part of the study.

Sample size estimation
The sample size was estimated using the following assumptions: a 21.3% prevalence of smear positive PTB among patients who will present with cough duration of ≥2 weeks [13] and 1.9% prevalence of smear positive PTB among patients who will present with cough duration of <2 weeks [11], 4% margin of error and 95% confidence level. With these assumptions, 302 individuals with cough lasting for ≥2 weeks and 423 individuals with cough lasting <2 two weeks were required to estimate the prevalence in each of the two groups.

Collection of questionnaire based information and sputum samples
A questionnaire was used to collect relevant data from PTB suspects recruited from the outpatient departments of the four health centers and consented to be part of the study. Health workers identified from each study sites were given responsibility of registering patients whose age is at least 12 years with productive cough of any duration. Consenting the identified potential study participant and requesting the consenting study participants to submit two sputum samples (spot-spot schedule) as per the national guidelines [14]. Socio-demographic characteristics of the study participants including age, sex, marital status, occupation, education, duration of cough, history of TB, history of close contact with TB patients, history of khat chewing, alcohol consumption and cigarette smoking were documented on the questionnaire. PTB suspects were also interviewed about various symptoms such as fever, chest pain, weight loss and loss of appetite.

Processing, transportation and laboratory analysis of sputum sample
A portion of the sputum samples collected from each individual was processed using Ziel-Neelsen staining (ZN) technique for smear microscopy and examined on the same day to identify acid-fast bacilli (AFB) [15]. The remaining portion of sputum sample was transported to TB laboratory of Aklilu Lemma Institute of Pathobiology (ALIPB) using a cold box and stored at -80 o C until processed for culture as previously described [16].
About 0.5 ml of neutralized sputa was inoculated onto test tube containing Lowenstein-Jensen culture medium with 12.5ml glycerol or 0.4% w/v pyruvate and incubated for 8 weeks. Cultures were considered negative when no colonies were seen after 8 weeks incubation period [17]. In case of culture with visible colonies [18], ZN staining was performed to confirm the presence of AFB using microscopy. AFB positive isolates were heat-killed at 80 o C for 1 hour using water bath, and stored at -20 o C until molecular characterization was performed. Mtb isolates were identified using PCR-based genotyping for RD9 deletion as previously described [19]. Internal R: TTGCTTCCCCGGTTCGTCTG. The mixture was heated in a thermal Cycler (Gene AMP 9700) for 15 minutes at 95°C and then subjected to 35 cycles of one-minute duration at 95°C, one minute at 55°C, one minute at 72°C and 10 minutes at 72°C. A molecular weight of 396bp was considered as Mtb, while a molecular weight 575bp was considered as M. bovis [20]. was based on bands of different sizes, as previously described [20].

Data management and processing
Data was computerized using Epi Data version 3.1 and analysis was performed using Stata version 12. The proportion of patients with smear positive PTB was calculated for the two study groups (i.e. according to their cough duration). A possible association between positivity for PTB and patients' background characteristics as well as duration of cough was investigated using bivariate and multivariable logistic regression. Statistically significant association was reported whenever p-value was less than 5%.

Associations of sociodemographic characteristics of the study participants with prevalence of PTB confirmed by smear and culture methods
Socio-demographic characteristics of 725 study participants and their associations with prevalence of PTB are summarized in

Discussion
In this study, a health facility based prevalence of PTB is reported using smear microscopy and culture methods among patients presenting to the facility with cough of any duration. Based on the smear microscopy examination the overall prevalence of smear-positive PTB was 5.0%, and the prevalence was higher among patients who reported cough for ≥2 weeks (10.9%) compared to those patients who reported cough for <2 weeks (0.7%). The smaller number of positive cases detected among patients with cough less than two weeks should not be considered insignificant since these individuals would contribute to the transmission of the disease among the families and the community if they were not screened and treated [9,21]. According to the manual of Tuberculosis, Leprosy and TB/HIV Prevention and Control Program in Ethiopia, any person with cough duration of two weeks or more is usually suspected for PTB and screened as per the guideline [14]. This practice excludes patients who complained cough for less than two weeks causing delay in the diagnosis of potential cases thereby increasing the risk of transmission to others. Previous studies also identified smear positive PTB cases among patients who reported cough for duration of <2 weeks [10,11,22].
Based on culture result, the overall prevalence of PTB was 11.9% which varies depending on reported duration of cough. The observed prevalence among patients with cough for <2 weeks was comparable with patients who had cough for ≥2 weeks. Majority of culture positives were from smear negative cases which is similar to the findings observed in Malawi among patients with short duration of cough [22].
Bacteriologically confirmed PTB among study participants with short duration was 10.2% which is in line with a study in China in which acute cough was independent predictors of bacteriologically positive TB diagnosis [23]. The prevalence in the current study is higher than the previous report in Eastern Ethiopia [24] and elsewhere [22]. However, the prevalence of bacteriologically confirmed PTB in the current study among patients with longer duration of cough is lower than the previous findings from different parts of Ethiopia [11,21,25] and elsewhere [22]. The observed difference might be the variation of the study participants in which this study screened all patients who had cough of any duration in our study.
The findings of the current study supports the previous findings that showed suspects who were reported as smear positive turned out to be culture negative [21,22,26,27], this might be partly be explained by reduced recovery of Mtb from cultures result based on specimen stored a very long time [28,29] or due to a rather harsh decontamination process [29], or could be false smear positive report due to technical problems.
In the current study duration of cough and night sweating were identified as potential risk factors for being PTB positive and currently employed was identified as a significant protective factor.
In previous study unemployment was independently and significantly associated with being TB positives [23]. Employed individuals perhaps can get information from co-workers which helps them to make informed decision and visit health institutions on a timely manner when they feel having symptoms and diagnosed TB. In Tanzania [10] and in Malawi [28] smear positive PTB was similar among study participants who had coughed for less than two weeks and among those who had cough for at least two weeks.
The proportion of culture positive samples that gave positive signal to RD9 deletion typing in the study is significantly lower than the proportion reported in the molecular characterization of Mycobacterium tuberculosis complex study in Gambella Regional state of Ethiopia [32] and elsewhere [39].
Not being able to further characterize 45 isolates for nontuberculous mycobacterial (NTM) infection because of shortage of reagents is one of the limitations of the current study. The second limitation is potential recall bias linked to duration of cough since patients remember quite often the recent most cough duration. Culture was not done for each sample within three days as recommended by World Health Organization and this could potentially affect the prevalence of culture positivity.

Conclusion
The results of this study indicate that screening of PTB using smear microscopy alone and use of cough duration for at least two weeks have a potential negative effect on early diagnosis and

Declarations Ethics approval and consent to participate
The study protocol was approved by the Institutional Review Board of ALIPB, Addis Ababa University and Ethics Committee of Addis Ababa Health Bureau before starting data collection.
Written informed consent was obtained from each participant (guardians for children) after describing the objective of the study to each participant/guardian before enrolment into the study.
As part of the routine clinical care, all patients with smear positive PTB were referred to the TB clinic for treatment and smear negative patients with suggestive diagnosis of TB were treated according to the national guideline. Furthermore, the health centers physicians/health officers/nurses were informed about those patients who were found positive by the culture method since OPD health workers have the mandate of referring TB patients to TB clinic for appropriate management and patients were contacted.

Consent to publication
Consent was obtained from each participant after describing the objective of the study to publication.

Availability of data and materials
The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.