Tuberculosis (TB) continues to be a major health problem in Zambia, despite a long running National Tuberculosis and Leprosy Programme (NTLP). In 2007, the World Health Organization (WHO) estimated the TB burden in Zambia to be at 60,337 cases (all forms of TB) [1]. The TB control efforts have been hampered by the high level of human immunodeficiency virus (HIV) infection, especially in urban settings where prevalence is estimated to be 19.7% [2]. As a result the number of TB and HIV cases threatens to overwhelm the capacity of the general health systems. HIV-TB co-infection rates in Zambia have been estimated at 70% [1].

Zambia adopted the WHO recommended Directly Observed Treatment Short course (DOTS) strategy as its primary approach in TB control in 1993 and has officially reported 100% DOTS coverage in all nine provinces since 2003 [3]. A good functioning primary health care system is crucial in the implementation of DOTS. In Zambia, the NTLP activities have been integrated into the primary health care services. The decentralisation of TB treatment services has provided for more responsibility at the lower levels of the health care system and in the face of an overwhelming TB case-load, this move has proved to be beneficial to the practical implementation of the programme. Despite the human resource challenges, the use of treatment supporters and community volunteers in the implementation of DOTS has contributed to the improvement in cure rates over the past decade from 67% in 2000 to the global target of 85% by 2006 [3]. The goal of the Zambian NTLP is to prevent and control TB through the provision of quality diagnostic and treatment services for TB and TB/HIV- infected individuals at all levels of the health care delivery system [4].

Assessing access to quality of healthcare service delivery is complex and multidimensional and will depend on several aspects that are both patient/community-related and/or health systems/service related. Several questions could be considered in this vein, for example, are patients seeking help when they are sick, and when they do seek healthcare, are they getting the appropriate care they require when they need it and ultimately, is this care effective when they get it? Understanding the factors that affect or influence care actions in different settings, will ultimately result in an improvement in healthcare delivery.

Although there are several reports about health seeking behaviour of TB patients and factors related to their delay in seeking health care, compliance to treatment and the role of these factors in treatment outcome, only a few studies describe patient experience in accessing TB care throughout treatment. This study describes and assesses the care received by pulmonary TB patients from government health care centres, and the association with patient adherence to TB treatment based on previous TB patients' reports. The study also alludes to patient's attitude to seek health care for TB.

The success of a national TB program is multi-faceted and complex. Community awareness; patients' adherence to treatment; patient access to quality of care through competent healthcare staff who are able to provide quality of care through prompt diagnosis and referral, prescription of correct treatment regimens and treatment follow-up; and accessible TB services, are important components of a successful TB program.

Consequently, it is important that people in communities are aware and able to suspect TB in persons who show signs and symptoms suggestive of the disease, such as prolonged cough, persistent fevers, and weight loss. Maybe not surprising, as previous TB patients our respondents showed a good level of knowledge on the symptoms and modes of transmission of TB, attributable to caregiver education during treatment. However, our study revealed vast differences in knowledge regarding the importance of treatment completion compared to knowledge of the importance of follow-up sputum submission; whereas, nearly 90% knew the importance of treatment completion, only 57% knew the importance of the latter, reflective of the low importance given to the relevance of education on this issue. Similarly, other studies have shown that most TB patients know the importance of treatment completion [7–9]. According to our conceptual framework, overall knowledge of the disease was low, mainly due to the low knowledge gap in the role of sputum microscopy in TB treatment by the respondents.

Despite the high knowledge levels of TB symptoms shown in our study, most respondents not only, reported not to have suspected they had TB, but also reported that they delayed seeking care (even when they suspected they had TB). Whereas it is possible that respondents were truly unaware of TB symptoms prior to TB treatment, several other studies have shown that there are various reasons why patients delay seeking care at a health centres. Loss of income, health centre systems or staff attitudes, stigma of the HIV association, severity of disease, lifestyle, for example, alcohol abuse, are among the many explanations [9–13]. The most common reasons in our study, 'I was thinking the symptoms will go away' or 'I did not think it was serious' also appear to be common in different settings [8, 12]. This may be reflective of the commonly practiced self-treatment, which may ameliorate initial symptoms thus temporarily masking the severity of disease and consequently 'buy them time' to continue with their daily income generating endeavours. Only 17% of our study population were in formal employment suggesting that for most respondents an income was dependent on their daily efforts and therefore may not afford the time at the health centre. Further, the period of the study, were the early days of scaling up of free antiretroviral therapy in Zambia and so people may still have been feeling helpless against HIV infection.

Our results showed that only 47% of respondents reported to have submitted follow-up sputum at least twice post diagnosis and that 67% reported submitting follow-up sputum at the end of treatment. These results may be cause for concern because sputum re-examination at the end of the patient's treatment is a much stronger indicator of treatment success than 'treatment completion. Further, data in one of our studies in this population, has shown that among subjects who experienced another episode of TB within one year of completing treatment, there were more who harboured the same M. tuberculosis strain as that of the previous episode (relapses/treatment failures) than those that had a different strain (re-infection) (unpublished data). Furthermore, our study showed a high proportion of respondents taking of drugs for the complete period of treatment (89%) with a notable proportion (22%) reporting stopping medication at some point during treatment. Over half (55%) cited that they stopped because they were feeling better, similar to many other studies [14, 8, 15].

The role of the health worker on patient compliance has been described many times [16–18]. Patient counselling and good communication [19, 20] can improve patient compliance. Our study showed high levels of patient satisfaction when it came to health provider explanation regarding medication. However, we did not see the same positive response with regards to health provider explanation on the role of follow-up sputum submission. Only about half of the respondents reported that they were informed about the requirement (53%) and importance (54%) of submitting follow-up sputum. In fact, these two parameters were shown to be significantly associated with respondent adherence (p < 0.0001 for both). A study in Egypt demonstrated that adherence to recommended sputum smear microscopy schedule was significantly associated with treatment success [21]. Our study also showed that respondent adherence to treatment was significantly associated with respondent's knowledge about the disease and its treatment (p < 0.0001) in contrast to other studies [22, 8].

Moreover, caregivers' communication skills fell short on account of dialogue, giving the patient a chance to ask questions, an important aspect in patient management that ensures patient understanding of disease and treatment. The effects of non-dialogue counselling were demonstrated in a study in Madagascar where reported lack of opportunity to ask questions by patient was significantly associated with non-adherence [16].

Other features of the health system, like distance, convenience of TB services (microscopy, antiretroviral treatment services), how long it takes to see the clinician, prompt diagnosis and referral of TB patients presenting with TB-related symptoms at primary health care facilities, may have an effect on patient access to healthcare. Distance to the health centre for this population was not an issue. Delays in the commencement of treatment have been documented in some settings [23], our study, however, showed that all the respondents were given medication within one week of diagnosis, with 84% commencing treatment within two days post laboratory diagnosis. The NTLP in Zambia has given full responsibility of sputum transportation plus obtaining and communicating results for each patient, to the treatment centres. This not only reduces on the number of patients, who remain undiagnosed following initial health centre visit, but also removes the inconvenience and added travel costs from the patients. The majority of our respondents reported that they used the same treatment centre for sputum submission. Our results indicate that facility-service related factors may not be the main issue in patients' access to TB care in Ndola, unlike the study from KwaZulu-Natal where systems failure was reported as contributing to the ineffectiveness of the National Tuberculosis Program [24].

Admittedly, because this study asked questions about past events, participants' recall may have biased our results. In addition, since the interview was anonymous to ensure complete confidentiality, we were not able to go back to the patient's data files to verify the self-reported data. Nevertheless, the implied cure rate for this sample population is comparable to the average cure rate data for the same period from Ndola. Another limitation for this study is that we did not establish from the respondents how long it took for laboratory results to be available for diagnosis, a factor that could well contribute to delay in TB patient care. However, enquiries from TB focal persons indicated a turnaround time for lab results ranging from the same day to a week. Further, our study did not include all components of TB treatment and care in the National Guidelines and consequently, other components that contribute to this package have not been discussed. Lastly, it is well known that respondents usually consider the interviewer to represent authority or the healthcare system and therefore tend to bias their answers in the way they expect they should to please the interviewer. Consequently, although the study made efforts to use researchers from outside the respondents' healthcare system, it is difficult to completely remove this perception in communities.

In conclusion, TB treatment systems appear to be well in place in NDHMT. However, taken together, these results suggest that closer monitoring systems on guidelines adherence at health centres may need strengthening and more patient counselling on treatment of disease and importance of sputum submission may improve cure rates.