The number of individuals found in the traditional healers' clinics during data collection period and who responded those traditional healers' clinics as their first choices could indicate the contribution of traditional healers' clinics to the public health system. The number of repeated visits of these clinics by patients and number of individuals that gave information to the patients about traditional healers' clinics that might have previously visited traditional healers' clinics also demonstrated the significance of the traditional healers' clinics for the public health system in Addis Ababa. These showed that a considerable number of the population was treated by the traditional healers' clinics and hence, the contribution of these clinics to public health systems in Addis Ababa.
The majority of patients in this study preferred traditional health care clinics than modern health facilities. Females, individuals with middle-income level and those with education visited traditional healers' clinics more frequently than the rest of informants. This is in agreement with the study done in Trinidad . However, it is different from the studies conducted in California , Israel  and Colombia University  where females, those with higher education and high-income level had statistically significant association with traditional medicine use. In most studies, low income has been mentioned as the reason to visit traditional healers' clinics [9–11] whereas in this study it was not found as a determinant in visiting traditional healers' clinics since other categories were equally important, which was indicated by single sample t test distribution (Table 2).
Reasons indicated by patients that participated in the present study for using traditional medicine as their first choice when they were ill is similar to the study done in Trinidad  where efficacy of traditional medicine was the reason for choosing herbal medicine as the first line of health care option. This high efficacy perception may be because traditional medicine was embedded in the belief and culture of the society [9–11]. On the other hand, the study conducted in Addis Ababa to determine the epidemiology of herbal drug use  showed that the main reasons given for choosing herbal medicine as the first line medication option were dissatisfaction with the services of modern health institutions due to their time-consuming practice, cost and perceived efficacy. Study conducted in Nigeria  also agrees with the present study that high efficacy of traditional medicine and dissatisfaction with modern medicine were the reasons to visit traditional healers' clinics.
The study conducted in the United States  to investigate possible predictors of alternative health care use indicated that those with higher education and poorer health status were associated with alternative medicine use. This is not in agreement with the current study, however level of education had a contribution in visiting traditional healers' clinics.
A majority of patients, in this study, visiting traditional healers' clinics were associated with dermatological cases. Study conducted in Pakistan  showed that 43% of the patients preferred traditional healers for skin disorder treatment indicating that the effectiveness of the remedies given by traditional healers against dermatological diseases.
The finding of this study that majority of patients were satisfied after being treated by traditional healers is corroborated by the study conducted in Zambia  and Tanzania . The study conducted in Nigeria  indicated that 33.4% of the respondents reported that herbal medicines had no adverse effects though lower than the current study. The difference could be due to the variation in the dosage and the type of herbs used.
The source of the healers' knowledge in this study is similar to the study conducted in Tanzania  where for 41.9% of the healers were their families. On the other hand most healers in Tanzania kept patient records containing demographic, diagnosis and treatment data whereas in the current study none of the healers kept patient records. The healers in the current study followed traditional treatment systems. Healers in Tanzania  agree in diagnosis of patients with this study though they also use laboratory test results made in the hospital in addition to history taking, physical diagnosis, and divination to identify diseases.
In the current study, only one healer referred his patients to modern medicine but the study done in Tanzania (27) showed that almost all healers referred their patients to hospitals when they failed with their own treatment. This difference may be because absence of collaboration and lack of training of traditional healers in Addis Ababa.