Background
Tuberculosis is one of the oldest disease that affect human and the leading cause of death globally. It is caused by Mycobacterium tuberculosis mainly affecting the lungs, but can affect other sites as well and it is curable if properly treated. World Health Organisation (WHO) declared tuberculosis as a global public health emergency by the year 1993 and DOTS program has been promoted as a control strategies [
1]. Unlike the availability of an appropriate prevention and control strategies, tuberculosis continues to challenge the world by its incidence and mortality fueled by HIV pandemic and drug resistant TB. The burden of the disease doubles in Africa due to theHIV and a low coverage of HIV testing [
2].
The success of TB treatment is the sum of the patients who are cured and those who have completed treatment under the Directly Observed Therapy Short Course (DOTS) strategy. In 1995 Directly Observed Therapy Short Course (DOTS) was initiated for the first time in some areas of Southern Nations Nationalities and Peoples Region and the program has raised the case detection and treatment success rate by folds. However, tuberculosis (TB) remains the leading cause of morbidity and mortality in the region and it is the third cause of inpatient death in the region. Studies conducted in the region revealed that the mortality is higher in the first 2 months of treatment [
3].
In the southern region of Ethiopia, tuberculosis was the third cause of death in hospitalized patients. Researches revealed that the mortality rate was 2.5% per annum in successfully treated tuberculosis patients in SNNPR [
3].
Although the key target of TB control in DOTs is to detect the disease and treat the cases, delayed presentation for treatment, incomplete treatment or poor compliance as well as treatment interruption or default, relapse, and death are the major challenge that TB programmes face in resource-constrained countries [
4].
Moreover, the federal ministry of health (FMOH) reported in 2011 that tuberculosis was the third leading cause of death in Ethiopia. World Health Organisation (WHO) global report in 2012 estimated that there were 8.7 million new cases and 12 million prevalent cases of tuberculosis globally in 2011 [
5]. The DOTs strategy, developed by the WHO for the prevention and control of TB in the early 1995s, was believed to be the most valuable strategy; and a cost-effective health intervention for reducing the incidence and death of TB in developing countries [
5].
It is estimated that approximately one-third of the world’s population is infected with mycobacterium tuberculosis. Death from active form of tuberculosis is expected to increase to 5 million by the year 2015. The burden the disease will be expected to be much higher in sub-Saharan African countries. The main reasons for the high morbidity and mortality of TB in these regions are dramatic increase in poverty, HIV epidemic, and emergence of drug resistant TB [
6].
According to the 2011 national population survey of Ethiopia, there were an estimated 15,000 deaths due to tuberculosis [
7].
Ethiopia was ranked 7th in the world for TB burden and 3rd in Africa in 2008, with an estimated TB incidence (all forms) of 378 new cases per 100,000 persons and163 new smear positive cases per 100,000 persons [
8].
A study conducted among 6580 registered tuberculosis patients (3147 males and 3433 females) in Addis Ababa health centers showed the following treatment outcomes: 18.1% cured, 64.6% completed treatment, 3.7% died during follow-up, 5.1% defaulted, 0.4% failed the treatment and 8.2% were transferred out to another health institution. From this study, it was found that year of enrollment was significantly associated with the treatment success [
8].
A study conducted in Uganda revealed that a limited information of patients about the disease resulted in high rate of defaulting and Mortality. Among 657 TB patients to assess a long term outcome of smear positive TB. Accordingly, 326 (49.6%) interrupted 1or more times. Of which 95 (29.1%) were in intensive phase, 82 (25.2%) in continuation phase and 149 (45.1%) interrupted their treatment in both phases [
9]. WHO estimates that 1.9% of all new TB cases may be resistant to the first line anti-tuberculosis drugs while about 9.4% TB cases may be resistance to the previously treated drugs. In patients on anti-tuberculosis drug therapy, poor adherence is recognized as a major cause of treatment failure, relapse and drug resistance [
10].
In another study conducted in Ibadan, Nigeria, the following proportion of treatment outcomes were observed: cure (76.6%), failure (8.1%), default (6.6%), transferred out (4.8%), and death (1.9%). The mean age of cured patients was 31.2±3.1 years, which was significantly lower than the mean age of those with poor treatment outcomes. In this study, males had a higher risk of a poor treatment outcome than females. Moreover, patients with a poor knowledge of tuberculosis had a higher risk of having a poor treatment outcome compared to those with a good knowledge about the disease [
11].
According to the study conducted in the central part of Ethiopia, the outcome of smear positive pulmonary TB treatment success rate was higher as compared to the WHO targets and showed 10.8% of unfavorable outcome. The unsuccessful treatment outcome was also higher in age groups more than 40, family size>5, retreatment and unemployed compared to their inverse [
12].
One study conducted on treatment outcome of tuberculosis patients under directly observed treatment and another five years retrospective study conducted on childhood tuberculosis and its treatment outcomes in Addis Ababa at various times showed almost similar increasing proportions of smear positive PTB, smear negative PTB and EPTB [
13,
14].
Treatment outcomes were documented for 95.2% of children of whom 85.5% were successfully treated while rates of mortality and defaulting from treatment was 3.3% and 3.8%, respectively. The proportion of children with TB tested for HIV reached 88.3% during the final year of the study period compared to only 3.9% at the beginning of the study period [
14].
Mortality was significantly higher among under-five children (p < 0.001) and those with HIV co-infection (p <0.001). On multivariate logistic regression, children 5–9 years [AOR = 2.50 (95% CI 1.67-3.74)] and 10–14 years [AOR = 2.70 (95% CI 1.86–3.91)] had a significantly higher successful treatment outcomes. On the other hand, smear positive PTB [AOR = 0.44 (95% CI 0.27–0.73), HIV co-infection (AOR = 0.49 (95% CI 0.30–0.80)] and unknown HIV sero-status [AOR = 0.60 (95% CI 0.42-0.86)] were predictors of poor treatment outcomes [
14].
From the study conducted in Mizan Aman hospital, it was found that out of 2043 TB patients, male patients outweighed (58.00%) female patients (42.00%). The following treatment outcomes were also observed in this study: 79 (3.87%) cured, 4 (0.20%) defaulted, 1575 (76.99%) transferred out to other health facility, and 25 (1.22%) died [
15].
Despite the continued DOTS TB treatment practice in Hadiya zone, treatment outcomes have not been assessed yet in the area. Therefore, the aim of this study was to assess treatment outcomes of TB patients and associated factors in the past five years in Nigist Eleni Mohammed general hospital, Hossana, SNNPR, Ethiopia.
Discussion
In this facility based retrospective study, data was extracted from 768 registered TB patients; and the proportion of male patients (55.3%) dominated female patients. This study was consistent with the finding in Uzbekistan (60%) [
16], Turkey (65%) [
17], and at Mizan-Aman general hospital (57.2%) [
15]. These studies reported an outweighed disease proportion of male patients over female patients. This might be attributed to either males were more likely to develop the disease or more likely to utilize the health services than females. However, one study conducted in Addis Ababa on childhood tuberculosis contradicted this finding.
In this study, out of 768 TB patients, 651 (84.8%) were pulmonary TB patients, and 117 (15.2%) were extra pulmonary TB patients. And among the pulmonary TB patients who were registered during the study period, the majority were affected by smear negative pulmonary TB (64.7%). And this finding was almost similar with the study finding conducted in Uzbekistan. In this study, pulmonary TB (PTB) was present in 77%, of which 43% were smear-positive and 53% were smear-negative [
16], with the study finding conducted at Mizan- Aman general hospital (51.05%) [
15], and This finding was also supported with the study finding conducted in Kwekwe district, Zimbabwe. In this study 42.4% of patients were with pulmonary tuberculosis and 8.2% of patients were with extra pulmonary tuberculosis. And among the pulmonary TB patients who were registered during the study period, the majority were affected by smear negative pulmonary TB (64.7%) [
18].
As to TB HIV co-infection; this study reported 16.4% TB HIV co-infection which was in line with the Ethiopian federal ministry of health report by 2009/10 (15%), and with the finding observed in the study conducted at Mizan-Aman general hospital (16.5%) [
2,
15]. In this study, HIV+ve TB patients were more likely to develop risk of poor treatment outcomes as compared to HIV-ve TB patients (AOR=0.796, 95% CI;0.512,1.236). This finding was almost in line with the study conducted in a cohort of tuberculosis patients in Recife, Pernambuco state, Brazil and showed that HIV+ve TB patients had developed poor treatment outcomes as compared to their counterparts (AOR=3.19, 95% CI; 1.31,7.73), and with the finding reported from the study conducted in Kwekwe district, Zimbabwe (RR=2.07, 95% CI;1.12,3.81). Another [
18,
19].
In this study, the overall proportion of cured patients was 10.9%, which was higher than the proportion observed at the study conducted in Mizan-Aman general hospital (3.87%) [
15]. The difference might be explained due to the high transfer out observed in Mizan-Aman general hospital (76.99%). In this study, the cure rate of TB smear positive was 79.39%, which was higher as compared with the study reported in Addis Ababa (64.8%) and Mizan-Aman general hospital (62.7%) [
14,
15]. This difference might be due to a better access to health care services and information in the latter settings.
The defaulting rate in this study (1.4%) revealed lower finding than the previous studies conducted in Nigist Eleni general hospital (20%), in Uzbekistan (6%), in Turkey (3.9%), in hospitals in Imo state, Nigeria (9.8%), and in Gondar University teaching hospital (36.4%), [
16,
17,
20,
21]. This lower defaulting rate in this study might be due to a better supervision and health education activities than the previous study areas.
The overall proportion of death (2.9%) in this study was lower than the death rate reported in the study conducted in Dessie and Woldiya town health institutions, Northeast Ethiopia (8.1%), in hospitals in Imo state, Nigeria (6.5%), in Addis Ababa (4%), in Felege Hiwot referral hospital, Northwest Ethiopia (5.8%), and in Kwekwe district, Zimbabwe (8.7%) [
14,
18,
20,
22,
23]. The observable higher death rate in the above study areas might be due to the lack of strict follow up and defaulter tracing mechanism. However, the overall death rate observed in this study was higher than the death rate observed in Turkey (2.4%) and in Recife, Pernambuco state, Brazil (2.8%) [
17,
19].
Furthermore, this study reported a treatment failure rate of 0.5%, which was consistent with the rate of treatment failures reported in Enfraz health center (0.5%) and in Felege Hiwot referral hospital, Northwest Ethiopia (0.5%) [
23,
24]. However, the treatment failure reported in this study was slightly higher than the finding in different health centers in Kotabharu, Kelantan, Malaysia (0.2%), and in Addis Ababa (0.4%) [
13,
25]. On the other hand, the treatment failure observed in this study was lower than the study findings observed in: Dessie and Woldiya town health institutions, Northeast Ethiopia (0.8%), Uzbekistan (3%), Turkey (1.1%), Recife, Pernambuco state, Brazil (2.1%), hospitals in Imo state, Nigeria (1.5%), and Kwekwe district, Zimbabwe (0.9%) [
16‐
20,
22].
There was also a high transfer out rate (51.7%) observed in this study among the patients who registered for the TB treatment. But this finding was lower than the study finding in Felege Hiwot referral hospital, Northwest Ethiopia (68.6%) [
23]. This high rate might be due to the overflow of patients from the rural areas (53.5%). However, the transferred out cases might have been cured or successfully completed the treatment. It could also happen due to unfavorable TB treatment outcomes like treatment failure and default unless adequate care and information is considered to the patients. This means, lack of information about transferred out cases therefore, limits the strength of reports about treatment success rate from DOTS clinics.
The overall treatment success of TB was 43.3%, which was high as compared to the previous finding in Felege Hiwot referral hospital, Northwest Ethiopia (26%) [
23], but very low as compared to the findings in: Dessie and Woldiya town health institutions, Northeast Ethiopia (88.1%), Uzbekistan (83%), Turkey (92.6%), Recife, Pernambuco state, Brazil (70.1%), Kotabharu, Kelantan, Malaysia (93%), hospitals in Imo state, Nigeria (81.4%), and Kwekwe district, Zimbabwe (72.4%) [
16‐
20,
22,
25]. Possible elucidations for the observed differences between the findings might be explained due to a high and/or low transfer out rate in each of the study area.
In this study, residence and treatment outcomes were highly associated. This means those patients who came from rural areas were more likely to develop risk of poor treatment outcomes as compared to patients who reside in urban areas (AOR=0.145 95% CI; 0.104,0.201). This finding was parallel with the study findings in: Uzbekistan (AOR=1.3,95% CI;1.2,1.4), Felege Hiwot referral hospital, Northwest Ethiopia (RR =7.0, 95% CI;3.89,12.63), and Kwekwe district, Zimbabwe (AOR=1.91, 95% CI;1.14,3.20) [
16,
18,
23].
This might be due to better health seeking behavior of patients living in the urban areas. Moreover, the health care institutions might be located nearby the patients living in urban areas, which in turn might contribute for a better health seeking behavior and treatment outcomes.
In this study, there was also a significant association between treatment out come and TB category. This means patients who presented pulmonary TB+ve were more likely to develop risk of poor treatment outcomes as compared to the patients with extra pulmonary TB and pulmonary TB-ve (AOR=1.915,95% CI;1.213,3.028). This finding contradicted the finding that says no significant association between unsuccessful treatment out comes and a status of pulmonary TB patient as reported in the study finding conducted in Dessie and Woldiya town health institutions, Northeast Ethiopia (AOR= 0.58, 95% CI; 0.29,1.14), in Enfraz health center, North west part of Ethiopia, and in Felege Hiwot referral hospital, Northwest Ethiopia (AOR= 1.4, 95% CI; 1.13,1.84). In these studies, the risk of developing unsuccessful treatment outcomes was less likely among the pulmonary TB patients than their counterparts [
22‐
24]. This might happen due to the proportion of TB patients considered in each category for the study.
In this facility based retrospective study, data was extracted from 768 registered TB patients; and the proportion of male patients (55.3%) dominated female patients. This study was consistent with the finding at Mizan-Aman general hospital (57.2%), Turkey (65%), and Uzbekistan (60%). [
15‐
17]. These studies reported an outweighed disease proportion of male patients over female patients. This might be attributed to either males were more likely to develop the disease or more likely to utilize the health services than females. However, one study conducted in Addis Ababa on childhood tuberculosis contradicted this finding. The study showed a high number of EPTB proportion among female patients [
14].
In this study, out of 768 TB patients, 651 (84.8%) were pulmonary TB patients, and 117 (15.2%) were extra pulmonary TB patients. This finding was supported with the study finding conducted in Kwekwe district, Zimbabwe. In this study 42.4% of patients were pulmonary tuberculosis and 8.2% of patients were extra pulmonary tuberculosis. Among the pulmonary TB patients who were registered during the study period, the majority were affected by smear negative pulmonary TB (64.7%). And this finding was almost similar with the study finding conducted at Mizan- Aman general hospital (51.05%) [
15], and with the study finding conducted at Uzbekistan. In this study, pulmonary TB (PTB) was present in 77%, of which 43% were smear-positive and 53% were smear-negative [
16].
As to TB HIV co-infection; this study reported 16.4% TB HIV co-infection which was in line with the Ethiopian federal ministry of health report by 2009/10 (15%) [
2], and with the finding observed in the study conducted at Mizan-Aman general hospital (16.5%) [
15]. In this study, HIV + ve TB patients were more likely to develop risk of poor treatment outcomes as compared to HIV-ve TB patients (AOR = 0.796, 95% CI;0.512,1.236). This finding was almost in line with the finding reported from the study conducted in Kwekwe district, Zimbabwe (RR = 2.07, 95% CI; 1.12, 3.81). Another study conducted in a cohort of tuberculosis patients in Recife, Pernambuco state, Brazil showed that HIV + ve TB patients had developed poor treatment outcomes as compared to their counterparts (AOR = 3.19, 95% CI; 1.31,7.73) [
18,
19].
In this study, the overall proportion of cured patients was 10.9%, which was higher than the proportion observed at the study conducted in Mizan-Aman general hospital (3.87%). The difference might be explained due to the high transfer out observed in Mizan-Aman general hospital (76.99%). In this study, the cure rate of TB smear positive was 79.39%, which was higher as compared with the study reported in Addis Ababa (64.8%) and Mizan-Aman general hospital (62.7%) [
14,
15]. This difference might be due to a better access to health care services and information in the latter settings.
The defaulting rate in this study (1.4%) revealed lower finding than the previous studies conducted in Nigist Eleni general hospital (20%), in Gondar University teaching hospital (36.4%) (24), in Turkey (3.9%), in Uzbekistan (6%), and in hospitals in Imo state, Nigeria (9.8%) [
16,
17,
20,
21]. This lower defaulting rate in this study might be due to a better supervision and health education activities than the previous study areas.
The overall proportion of death (2.9%) in this study was lower than the death rate reported in the study conducted in Addis Ababa (4%), in Felege Hiwot referral hospital, Northwest Ethiopia (5.8%), in Dessie and Woldiya town health institutions, Northeast Ethiopia (8.1%), in hospitals in Imo state, Nigeria (6.5%), and in Kwekwe district, Zimbabwe (8.7%) [
12,
14,
18,
20,
22]. The observable higher death rate in the above study areas might be due to the lack of strict follow up and defaulter tracing mechanism. However, the overall death rate observed in this study was higher than the death rate observed in Turkey (2.4%) and in Recife, Pernambuco state, Brazil (2.8%) [
17,
19].
Furthermore, this study reported a treatment failure rate of 0.5%, which was consistent with the rate of treatment failures reported in Enfraz health center (0.5%) and in Felege Hiwot referral hospital, Northwest Ethiopia (0.5%) [
12,
24]. However, the treatment failure reported in this study was slightly higher than the finding in different health centers in Addis Ababa (0.4%), and in Kotabharu, Kelantan, Malaysia (0.2%) [
13,
25]. On the other hand, the treatment failure observed in this study was lower than the study findings observed in: Dessie and Woldiya town health institutions, Northeast Ethiopia (0.8%), Uzbekistan (3%), Turkey (1.1%), Recife, Pernambuco state, Brazil (2.1%), hospitals in Imo state, Nigeria (1.5%), and Kwekwe district, Zimbabwe (0.9%) [
16‐
20,
22].
There was also a high transfer out rate (51.7%) observed in this study among the patients who registered for the TB treatment. But this finding was lower than the study finding in Felege Hiwot referral hospital, Northwest Ethiopia (68.6%) [
12]. This high rate might be due to the overflow of patients from the rural areas (53.5%). However, the transferred out cases might have been cured or successfully completed the treatment. It could also happen due to unfavorable TB treatment outcomes like treatment failure and default unless adequate care and information is considered to the patients. This means, lack of information about transferred out cases therefore, limits the strength of reports about treatment success rate from DOTS clinics.
The overall treatment success of TB was 43.3%, which was high as compared to the previous finding in Felege Hiwot referral hospital, Northwest Ethiopia (26%) [
12], but very low as compared to the findings in: Dessie and Woldiya town health institutions, Northeast Ethiopia (88.1%), Uzbekistan (83%), Turkey (92.6%), Recife, Pernambuco state, Brazil (70.1%), Kwekwe district, Zimbabwe (72.4%), hospitals in Imo state, Nigeria (81.4%), and Kotabharu, Kelantan, Malaysia (93%) [
16‐
20,
22,
25]. Possible elucidations for the observed differences between the findings might be explained due to a high and/or low transfer out rate in each of the study area.
In this study, residence and treatment outcomes were highly associated. This means those patients who came from rural areas were more likely to develop risk of poor treatment outcomes as compared to patients who reside in urban areas (AOR = 0.145 95% CI; 0.104,0.201). This finding was parallel with the study findings in: Felege Hiwot referral hospital, Northwest Ethiopia (RR = 7.0, 95% CI;3.89,12.63), Uzbekistan (AOR = 1.3,95% CI;1.2,1.4), and Kwekwe district, Zimbabwe (AOR = 1.91, 95% CI;1.14,3.20) [
12,
16,
18].
This might be due to better health seeking behavior of patients living in the urban areas. Moreover, the health care institutions might be located nearby the patients living in urban areas, which in turn might contribute for a better health seeking behavior and treatment outcomes.
In this study, there was also a significant association between treatment out come and TB category. This means patients who presented pulmonary TB + ve were more likely to develop risk of poor treatment outcomes as compared to the patients with extra pulmonary TB and pulmonary TB-ve (AOR = 1.915,95% CI;1.213,3.028). This finding was in opposite to the study finding conducted in Dessie and Woldiya town health institutions, Northeast Ethiopia (AOR = 0.58, 95% CI; 0.29,1.14), and in Felege Hiwot referral hospital, Northwest Ethiopia (AOR = 1.4, 95% CI; 1.13,1.84). In these studies, the risk of developing unsuccessful treatment outcomes was less likely among the pulmonary TB patients than the extra pulmonary TB patients [
12,
22]. This might happen due to the proportion of TB patients considered in each category for the study. On the other hand, patients with smear positive (pulmonary TB + ve) were more likely to develop a risk of poor treatment outcomes as compared to smear negative (pulmonary TB-ve) patients (AOR = 1.915 95% CI;1.213,3.028). This result contradicted the finding that says no significant association between unsuccessful treatment out comes and a status of pulmonary TB patient as reported in Enfraz health center in North west part of Ethiopia [
24].