Key elements in any TB control programme are early diagnosis and prompt initiation of treatment. Analysis of factors associated with patient and healthcare services delay is an important step to identifying how to improve the quality of TB care and control. This study assessed delays and associated factors in the diagnosis of Tuberculosis in Hohoe Municipality, Ghana.
Contrary to a previous report from Ethiopia  where majority (68.7 %) of their participants reported to a public medical provider within 30 days following the onset of symptoms. This study shows that more than 60 % of patients delayed their first visit to medical provider following symptoms for more than 30 days. Eleven percent of patients consult a healthcare provider within two weeks in this study, compared to 70 % reported by Kiwuwa and colleagues in Kampala Uganda . Two weeks is generally the timeline recommended for a patient with cough to visit a health facility . This long patient delay can reflect poor patients’ knowledge and awareness of TB symptoms and the need for prompt consultation with healthcare services for diagnosis and treatment. The proportion of patients (60.3 %) who delay seeking care for > 30 days is higher in this study than what was found in previous studies in Tanzania and Uganda [15, 17]. However, this study reports a lower proportion of patient delay than what was reported in Nigeria , where patient delay was observed in more than 80 % of patients. In a similar study done in Kumasi, Ghana in 1995 to determine the factors associated with diagnosis delay among pulmonary TB patients, a low proportion of patient delay (46 %) was reported . Also, the median patient delay of 47 days among sputum smear positive patients in this study is longer than what Lawn and colleagues found in their Kumasi study. These differences in patient delay between the two studies could be due to study setting as more urban dwellers living close to health facilities may have shorter patient delay than rural dwellers. Sreeramareddy et al. in their systematic review , found that among low and middle income countries, patient delay varied from 4.9 days in Gambia to 162 days in Tanzania. It is likely that most of these studies may have underestimated patient delays since patients often do not clearly recognize symptoms at the onset of disease.
Patient delay has been reported to be more common among females than males . This study however, finds an insignificant difference between the sexes and increased patient delay, a finding consistent with those reported in Tanzania , Nigeria , Ethiopia , Malaysia , and Norway , where sex was not found to be a significant predictor of prolong patient delay. In contrast to the Kumasi study , males in this study postponed care-seeking longer than females. In a study in India, males delay longer because of fear of cost of diagnosis and treatment . Also, men were more likely to neglect symptoms longer until they are serious before seeking care. Fazlul and colleagues attributed similar differences in case detection to women’s limited decision-making power and failure of health systems to provide accessible and acceptable health care .
This current study shows that patients in employment have 2.9-fold greater median patient delay than the unemployed patients in the univariable analysis. A possible explanation for this variance can be that those employed are busy at their work and are not able to attend a clinic with their symptom.
In this study, patients who first sought care from alternative care providers following symptoms have an increased risk of patient delay, though not statistically significant. One can speculate that this group of patients did not perceive themselves to be at risk of TB and did not take their symptoms seriously enough to seek formal medical care.
Stigma plays an important role in determining the health-seeking behaviour of suspected TB patients . Despite decades of public health efforts, stigma continue to impede progress in diagnosis and treatment of TB . This study, like other studies [25-27] finds that stigma was independently associated with prolonged patient delay. Nonetheless, most of the patients are willing to reveal their illness to others in this study, this is similar to what was reported previously in Ethiopia where 88 % of study participants were willing to disclose their TB status to everyone . Several studies have shown why and how TB has been highly stigmatized throughout history. Most authors recognize the perceived infectiousness of TB as a major cause of stigmatization [26-28]. Whilst the stigma of TB as “a disease of the poor” persists, more recently, HIV/AIDS stigma affects TB patients, particularly in communities where HIV/AIDS is prevalent as shown in studies in Ethiopia . Dordor and Kelly  recognised that fear of stigma can result in infected individuals hiding their disease from their families and others
As found in China  and France , lack of medical insurance was surprisingly found to predict extended patient delay in this study. Though not explored in this study, lack of medical insurance may be related to poverty. Poverty and lack of insurance may discourage people from seeking prompt healthcare. In this view, Ghana National TB programme as in many TB endemic countries had implemented free TB diagnosis and treatment policy to reduce financial burden of patients and hence improve access to TB diagnosis and treatment. In addition to this free policy, the NTP had implemented patients’ “enablers” package to take care of other expenses such as transportation costs, medical consultation fee, cost of folders, chest X-ray, and food supplements for patients who cannot afford.
Factors such as age, educational level, knowledge, which have been identified to be significantly associated with patient delay [13, 19, 29,] could not be established in this study.
This study also failed to find an association between access to healthcare and prolonged patient delay. This may be due to the limited sample size obtained for the study.
This study did not find an increase in patient delay among rural dwellers compared to urban patients, despite the fact that these patients are likely to have poor access to healthcare and that they were found to be less educated compared to patients from urban settings. Yimer et al.  however, found in their recent study among 201 patients attending a referral hospital in Northwest Ethiopia that rural dwellers were 3.4 times more likely to have increased patient delay than urban dwellers, similar to previous report , suggesting that rural patients in this study setting has better access to healthcare than rural dwellers in Ethiopia.
Healthcare services delay
In this study, the time period from patients’ first contact with public health services to diagnosis (45 days) is unacceptably long and contributes significantly to total diagnostic delay. This finding compares favourably with reports from other countries such as Ethiopia , Uganda , and Pakistan  that reported median healthcare system delay ranging between 61 days and 87 days, however, longer than report Nigeria (. The differences can perhaps be due to the variation in the definition of healthcare system delay. Some studies defined it to include the period between TB diagnosis and initiation of treatment. In contrast to the patient delay, the median healthcare services delay in this study is shorter than the previous report from Ghana . Lawn’s study was just a year after establishing a new TB control programme in the country and since then the National Tuberculosis Control Programme has been implementing activities towards improving quality of TB care, which include extensive training of health personnel, decentralization of diagnosis and treatment services, community TB-DOTS, and provision of “enabler” package to health providers and TB patients as well, aiming at providing a quality integrated TB services to people at all levels by means of standardised diagnosis, treatment and community-based care and support. These measures are expected to reduce diagnostic delays. This could explain the shorter median healthcare services delay in the current study. This study dismisses the view held by healthcare providers that the late presentation of patients is mainly “the patients’ doing”. On the contrary, in approximately half of the patients, the healthcare services delay exceeded the patient delay. This reflects the inefficiencies on the part of healthcare services in the diagnosis of Tuberculosis. Proportion of healthcare services delay in this study is higher among urban dwellers than rural patients (78.6 vs. 66.7 %), however, rural dwellers had longer healthcare services delay compared to patients from urban settings (median: 54 vs. 45 days). This is perhaps due to the fact that urban health providers are able to suspect and diagnose TB cases faster than their rural counterparts. The municipal hospital is the only health facility that provide diagnostic and treatment services for TB patients at the time of the study thus patients suspected in the rural facilities must be referred to the hospital for diagnosis. Some of these patients might have delayed at home after referral explaining the prolong time duration between patients first encounter with healthcare services and diagnosis.
In line with a previous study in Kampala, Uganda , this study shows that a significant proportion of TB patients used alternative healthcare as the first choice of care in Hohoe Municipality, with nearly 40 % initially presenting to pharmacy or drug shops, traditional healers, and private clinic. On the contrary, Pronyk et al.  reported that 75 % of patients first visited public health system. This may reflect lack of awareness of free TB diagnosis and treatment services as more than half of patients hoped to receive cheaper services at those places.
This study demonstrates that “multiple healthcare contacts” is the overriding predictor to healthcare services delay. Similar finding was reported in Afghanistan, where making more than one visit to health care providers strongly predicts increased health system delay . This relationship may be associated with poor clinical suspicions of signs and symptoms by healthcare providers especially from primary healthcare facilities and failure to request for proper investigations or refer patients to TB centre for further investigations, which led to patients making multiple contacts because of persistent symptoms. Some authors attribute this delay to provider failure to correctly diagnose TB patient [10,13]. Wrishmeen and colleagues  speculate that multiple health care provider visits influence health system delay when patients receive inappropriate antibiotic that can modify the clinical picture which may cause patients in believing that they will be cured; and in the long run, may choose alternative treatment. Needham et al.  also indicated that centralized public services and a lack of integration between public and private providers prolonged health system delays.
Contrary to the previous report from Ghana , where doctor delay was 3.9 times greater for rural dwellers compared to patients from urban settings, this study found no increase in healthcare services delay among rural patients compared to urban dwellers. This suggest that both rural and urban healthcare providers contribute equally to healthcare services delay in Hohoe Municipality despite the fact that rural dwellers are likely to have poorer access to TB diagnosis services as the only diagnostic facility is located in Hohoe township.
The total median delay is 104 days, similar to what was found in Brazil  during which the patient was transmitting the infection to the close contacts in the community. Approximately 80 % of patients have total diagnostic delay exceeding 2 months, more than what was reported in Ethiopia (47.8 %) . This has a great public health implication since increased transmission in communities may imply that patients may have waited longer in the communities before diagnosis and treatment. Long total delays have also been found in industrialized countries like the United Kingdom (78 days)  and United States (89 days) . The median total delay of 104 days (3.5 months) is shorter than 4 months reported in the Kumasi study. In line with studies in Nigeria  and Nepal , patient delay is a major contributor to the total diagnostic delay in this study, contributing approximately three-fifth of the total delay, finding contrary to previous study in Kumasi where doctor delay was the main contributor to total delay . The proportion of patients with total delay was higher among urban dwellers (89.3 %) than rural patients (77.8 %). Moreover, urban dwellers were more likely to have longer total delay than rural dwellers, which may imply that patients from urban settings delay longer before diagnosis to avoid disrupting perhaps their economic activities. Similar to a report from Brazil , more patients with AFB negative had longer total delay than AFB positive patients.
The most obvious limitation of the study is its small sample size. Therefore firm conclusions about the relationships among variables cannot be drawn. Thus interpretation of the results must be done with caution. Research studies with much larger sample size would therefore be required to ensure appropriate generalization of the findings of the study. Secondly, since only self-report measures were used to estimate delays, response consistency effects may have biased the observed estimates. However, measures were taken to minimize this limitation. For example interviewers used national and local events to estimate date of onset of symptoms. Secondly, the date of first consultation with healthcare provider was self-reported and could not be validated in this study thus leading to potential differential misclassification as a result of recall bias.