Delays to anti-tuberculosis treatment intiation among cases on directly observed treatment short course in districts of southwestern Ethiopia: a cross sectional study

Tuberculosis (TB) has been recognized as a global public health problem since 1993 when an estimated 7–8 million cases and 1.3–1.6 million deaths occured [1] . Since then, different global strategies have been designed and implemented for the control of TB; consequently an estimated 54 million lives were saved between 2000 and 2017 [2]. Despite such achievement, TB remained to be among the major global public health problems. Globally, 10 million incident cases and 1.3milion deaths were estimated to occur in 2017. Of the estimated incident cases, only 6.4 million (64%) were notified to National TB programs (NTPs) [2].

Tuberculosis is among the major public health problems in Ethiopia accounted for third cause of hospital admission and second cause of death [3]. Ethiopia is among the 30 TB High Burden Countries (HBC) where about 172,000 new cases and 25,000 deaths were estimated to occur in 2017. Out of the estimated cases in 2017, only 117,705 (68%) were notified to NTP [2]. The first national prevalence survey in 2011 revealed smear positive pulmonary TB (PTB) prevalence of 108/100,000. The survey further reported 55% of the cases were first identified by the survey and were not on treatment until the survey [4].

Early detection and treatment of TB cases have been a priority in the prevention and control of TB [5]. The delay to TB treatment can be categorized into: 1) patient delay that constitute time elapsed between onset of TB symptoms and first self presentation to formal care, 2) provider delay as time elapsed between first presentation to formal care and anti-TB treatment initiation and 3) total delay as time elapsed between onsets of TB symptoms and anti-TB treatment initiation. When diagnosis of TB is delayed, patients go without treatment for long and transmit the disease. In high prevalence settings, each infectious case of TB can result in up to 20 secondary infections until detection [6] . Moreover, delays to diagnosis and treatment of TB result in more serious illness by the time of diagnosis, increased length of infectiousness and poor treatment outcomes [7‐11] and represent a time span for additional costs [12]. Consequently, delays to diagnosis has been emphasized as a major obstacle to TB control [13].

Systematic reviews of studies across the world reported median patient delay of 30 days and provider delay of 7-28 days [13, 14]. On the other hand, diagnostic delays ranging from a median of 8-191 days at different levels of health care delivery systems [7, 8, 15‐18] were reported. Studies from different parts of Ethiopia also reported longer delays to initiate treatment for TB with median patient delay of 20–90 days [16, 17, 19] and provider delay of 6-34 days [17, 19‐21]. A prevalence survey in northern Ethiopia [22] reported two-thirds of active TB cases were not detected timely. The long patient, provider and total delays are attributed to female gender, older age, low awareness about TB, repeated visits to health facilities, visiting lower level health care facilities and traditional healer [23, 24].

The studies in Ethiopia are limited to smear positive PTB cases [21], focused on cases presenting to either health centers [20] or hospitals [25] but not both. Moreover, the studies were conducted when TB had been treated for 8 months. Studies have shown that shorter treatment regimens could dramatically accelerate the reductions in TB incidence and mortality that can motivate suspects to seek care early and treatment [26]. In Ethiopia, treatment regimen of 6 months was introduced in 2011. Thus, complete picture of the patterns of delay of different categories of TB at different health care setups and in the era of reduced treatment regimen is scanty in the country. As a result the national TB research advisory council has set identification of barriers for the different forms of delays to TB treatment as a national priority research agenda [27]. However, recent evidences are lacking in Ethiopia in general and the study area in particular. Therefore, this study investigated time delays to initiate formal care seeking, diagnosis and treatment along with factors associated with the delays among new TB cases on treatment at health centers and hospitals in districts of southwestern Ethiopia.

Prompt detection and treatment of cases has been a priority in the prevention and control of TB that can be realized upon timely care seeking by patients and diagnosis by the health system. Our study assessed the delay and associated factors with the care seeking, diagnosis and treatment of TB in rural districts of southwestern Ethiopia. We found long patient, provider, and total delays, among TB cases on treatment. TB patients had waited for a median of 25 and 55 days respectively to initiate formal care seeking and anti-TB treatment. Both patients and provider delays contribute nearly equally to the total delay which is consistent with other studies from Ethiopia [20] and elsewhere [14]. The longer time elapsed since onset of illness to treatment initiation implies increased risk of morbidity and mortality among the cases and diseases transmission in the community [10, 33‐35]. All forms of the delays are attributed to patient, disease and health system related factors. This implies the delays are multifaceted that calls for intensified TB case finding through strengthening healthcare facilities and health promotion activities.

Patients had waited for a median of 25 days until seeking help at formal health care providers which is consistent with 30 days in northern part of Ethiopia [17, 31, 36], 28 days in Uganda [14] and 30 days in Angola [13]. The relatively lower delay could be due to better access to HCF where nearly 80% of the cases traveled less than an hour to reach the first HCF. However, only a third (32%) of the cases had made visits to formal health care within 15 days of recommended timeline for TB suspects to visit HCF [29].Besides, more than half (57%) of the cases did not perceive their care seeking was delayed which implies the symptoms are taken as common and less severe to urge consultation of healthcare provider.

Before visits to formal health care facility, patients had taken variety of actions those influence timing of care seeking. Consistent with other studies, patients who took prior self-treatment were more likely to delay seeking formal care compared to those first consulted health care providers [37, 38]. This could be due to use of some home remedies or over the counter antibiotics or analgesics those might lessen the manifestation of the illness for the time being [39]. Moreover, those patients first visited traditional healer and holy water are more likely to have higher total delay. This could be due to the beliefs attached to the traditional care and holy water those might inhibit timely presentation and diagnosis [24].

Patients co infected with HIV are more likely to delay care seeking compared to those non- infected. This could be due to the alteration of classical clinical manifestations and signs of TB among HIV co infected patients [40]. On the other hand, patients suspected or tested HIV positive delay to present themselves to HCF due to fear of stigma attached to the co-occurrence of TB and HIV [9, 41, 42]. As a result, the high mortality among HIV co infected TB patients is partly explained by the delays to TB treatment [13, 43]. Those extra pulmonary cases patients are more likely to delay seeking care compared to the pulmonary cases. A similar finding was also reported form a study in northwest Ethiopia [25]. This could be due to the fact that extra pulmonary cases manifest with less severe and non-specific symptoms that urge patients to perceive the illness to be non-serious [44].

Patient’s perception and knowledge towards the TB disease and control activities do have impact on the care seeking practices. The study revealed that patients with good knowledge towards the TB disease and program are less likely to delay seeking care. Similarly, other studies have reported lack of knowledge and mistrust of the TB program as a reason for delayed care seeking [23, 45, 46]. In contrast, awareness and belief about TB’s curability is associated with longer patients delay [7, 14]. These imply need for awareness creation towards the TB illness and its control program to clear the paradox between the belief and longer delay.

The first consultation at HCF was made primarily (69.9%) at lower level public health care units (health posts and health centers) and private clinics. This is consistent with a study in Mediterranean countries that reported two thirds of patients first visited private sectors [46]. Nonetheless, diagnosis of TB was made primarily at hospital (61%) despite only few had made first visit to the hospitals. As a result, majority of the cases had made more than one visit to different HCFs until diagnosis. Similarly, studies from Uganda [14] and China [47] reported significant number of patients had visited more than one HCF until diagnosis. Subsequent to the missed opportunities during the repeated visits, diagnosis of TB had been made after a median of 22 days from the first visit to HCF. This is consistent with 21 days in a study from Amhara Region in Ethiopia [17] but higher than 6 in Addis Ababa [20] and 9 days in Tigray, northern Ethiopia [19]. The discrepancies could be due to the differences in accessibility to well equipped HCFs and skilled providers usually situated in cities like Addis Ababa. The longer provider delays due to the repeated visits portray high level of missed opportunities of early diagnosis that could have increased infectious period of the cases and costs incurred by the patients and households.

Longer provider delays had been reported among patients who first visited HCFs not providing DOTS services including health posts and private clinics. This is consistent with a study from Uzbekistan [38] and India [44].The higher provider delay at lower care units and private clinic could be due to lack of supplies, guidelines and skilled providers that enhance adherence to the national TB control program. In the current study, diagnosis of TB was mainly made at hospitals that depict unnecessary drugs and treatments provision at series of HCFs incurring extra cost. Besides the cost incurred, medications at the repeated visits themselves lead to delays to diagnosis and serious outcomes including drug resistance [48].On the other hand, the lower health care units and private clinics might lack proper logistics and skilled providers to timely diagnose the cases. Hence, expansions of the DOTS package to the lower and private clinic is required to curb the prevailing long delays.

Patients delayed to seek care are also more likely to have delayed diagnosis and treatment after they initiated care seeking. In contrast a study in Georgia [49] reported those patients with increased patient delay are less likely to have prolonged diagnostic delay. The discrepancy could be due to differences in measurement of the two delays. The higher odds of provider delay among those delayed to seek care can be explained by patients use of different forms of self-treatment and homemade remedies those might alter the manifestations of the TB illness that pose difficulties in timely diagnosis [39, 49]. Those patients delayed to seek care might have been ill for such long duration at which time productivity is compromised. Hence, the patients might be unable to cope with the costs required to have timely diagnosis even after initiating the care seeking. In addition, the long ill days are also associated with more severe disease at presentation [10] which hinders timely diagnosis [21].

Our study has few limitations. First, the study was carried out on those cases ultimately sought care and on treatment at the time of the study. Thus, measurement of delays relied on patient self-report that is liable to recall bias. We minimized this bias through interviewing patients soon after diagnosis and helping them to recall using local events. Second, the assessment of patients’ knowledge towards the TB diseases and its control program might be influenced by the information provided during treatment initiation. Therefore, this could have brought egg and chicken dilemma as the knowledge or care seeking preceded one to other. Third, we studied only new and adult cases so that the findings cannot be generalized to all forms of TB cases among all age groups. Fourth, we were not able to interview patients died during the intensive phase of treatment those could underestimated the true total delay attributed to the severe illness. Lastly our study lacked qualitative assessment of patients and health system those would have supplemented the quantitative data. On the other hand, consecutive enrollment of cases that minimized selection bias and triangulation of data sources from patient interview and chart reviews could be mentioned as strength of the study. Finally, our study is valid and generalizable to new adult cases in similar settings.

Tuberculosis patients in the study area had passed through too long journey to initiate anti-TB treatment. The long journey is attributed to patient and provider delays those are positively correlated and contributed nearly equally to the total delay. The long delays at different levels of healthcare facilities portray high level of missed opportunities of early diagnosis of TB. The patient, provider, and total delays are attributed to the patient, disease and health system related attributes reflecting need for multifaceted intervention at all levels. Therefore, improving community awareness, involving informal providers, health extension workers and TB treatment supporters can reduce the patient delay. Similarly, cough screening and improving diagnostic efficiencies of healthcare facilities should be in place to reduce the provider delays.