The present study was a case-control study conducted over period of one year from January 2012 to December 2012. The study was carried out in five major cities i.e. Ludhiana, Jalandhar, Amritsar, Patiala and Bathinda, which inhabit approximately half of the urban population of Punjab and have a large network of private practitioners. For this study the PPs were divided into two groups, referral group and non–referral group. Referral group involved the registered PPs who referred at least one patient with presumptive pulmonary TB per quarter for first two quarters of the study period to the nearest DMCs for diagnosis. Non Referral group included PPs who did not refer any patient to DMC during this period.

Sample size, as calculated with Epi info, was 165 for referral and 165 for non-referral group which came out to be 33 per city for each group. So to address the problem of non-response and incomplete data, 43 PPs for referral group and 43 matched PPs for non-referral group were selected per city.

List of DMCs was obtained from District Programme officer –Tuberculosis (DPO-TB) of the particular district. Laboratory technicians (LTs) of DMCs in the selected cities were sensitized to correctly record the details (name and address) of referring doctors for two quarters (January 2012 to June, 2012). Referral group of practitioners was identified from laboratory records of these two quarters and subsequently 43 PPs per city were selected from this list by population proportionate to sample size method based on qualification. Simultaneously a list of all registered PPs was obtained from the Civil Surgeon’s office and respective professional bodies (e.g. Indian Medical Association, National Integrated Medical Association etc.) to which PPs in a particular area were registered.

Non-referral Group was selected by a process of one to one matching. The criteria of matching was the nearest (by distance in either direction) registered PP in the same Tuberculosis Unit (TU), with the same qualification (allopathic, ayurvedic, homeopathic etc.) as the PP from referral group, and who had not referred even a single patient with presumptive pulmonary TB to DMCs for diagnosis during these two quarters. The distance and accessibility of nearest DMC was comparable in both groups. Tuberculosis Unit (TU) is an administrative unit under RNTCP which has five DMCs under it with each DMC catering to a population of one lakh.

Five field investigators (FIs) belonging to respective cities with graduate medical qualification (BSc, BSc (Nursing), GNM etc.) were recruited and trained by the principal investigators for the study. The training included sensitization regarding TB, RNTCP, study protocol and filling the proforma and questionnaire. To minimize inter-observer bias, the comprehension of FIs about the study and proforma was evaluated during these sessions before sending them in field. The interviews of the PPs, collection and analysis of data were conducted during last two quarters (July 2012 to December 2012) of the study period. Lists and mapping of non-referring PPs in the locality were prepared by FIs with the help of Senior Treatment Supervisor (STS) and Senior Tuberculosis Laboratory Supervisor (STLS) of the respective areas. The non-referral group was selected from this list as per definition. The data collection was done using a pre-tested, structured proforma containing information about age, sex, professional qualification, experience etc. of the PP and questionnaire regarding their knowledge of TB and RNTCP (Annexure 1 & 2). The selected PPs (43 referral and 43 non-referral) were visited by the FIs and the proforma administered and various possible reasons for the referral/non-referral of the pulmonary TB suspects to DMCs were investigated. Field investigators collected the information and reported to the respective District Program Officer- Tuberculosis (DPO-TB) who supervised the whole process of data collection and transportation. The quality of the data collection was ensured by random visits of Principal investigator/ Co-investigators to 10-20% of the PPs visited by the FI. The project initiation was done after approval by institutional ethics committee. Written informed consent was taken from the participants after informing the scope of the study. The data so generated were entered into Microsoft excel by a data entry operator and further analyzed using Statistical Package for Social Sciences (SPSS package), version 20 (IBM SPSS statistics 20.0.0, 2011). Chi-square test and Odd’s ratio was estimated for each variable.

A total of 430 private practitioners (215 each in referral and non-referral groups) were interviewed. Demographic characteristics (Age, gender) of two groups were comparable. The participants (PPs) comprised of 220 allopathic doctors, 170 ayurvedic doctors, 26 homeopathic doctors and 14 other Registered Medical Practitioners (RMPs) who did not have formal medical qualification but were licensed with state medical council to practice medicine. The two study groups were also comparable with respect to qualification and experience in years [Table/Fig-1]. Knowledge regarding TB and RNTCP was assessed on the basis of questionnaire. Only 46.5% of practitioners from referral group and 37.7% from the non-referral group had adequate knowledge regarding TB. Although knowledge was higher in the referral group but the difference between the two groups was not significant statistically. A 38.1% PPs from referral group and 25.6% PPs from non-referral group had adequate knowledge regarding RNTCP and this difference was highly significant [Table/Fig-2]. The difference in knowledge regarding TB and RNTCP with respect to age and gender distribution in the study groups was not statistically significant. Allopathic doctors had significantly higher knowledge regarding TB as compared to ayurvedic, homeopathy and RMP doctors whereas the difference in knowledge regarding RNTCP was not statistically significant in these four streams of doctors [Table/Fig-3].

In the referral group, both the knowledge of PPs regarding nearest DMC as well as perception about its accessibility for the patients were found to be significantly higher than the non-referral group. 98.6% of PPs in the referral group knew about the nearest DMC whereas 82.8% of PPs from non-referral group had knowledge regarding nearest DMC. Most of practitioners (92.7%) from both groups were not fearful of losing patients if the patient was referred to DMC for diagnosis. Previous bad experience on referral was not the reason for non-referral of patients to the nearest DMC in 86% of PPs from the non-referral group. Only 15.3% of PPs in non-referral group had ever been contacted by RNTCP staff. Comparison of two groups with respect to faith in Government Health Sector was not significant statistically [Table/Fig-4].

Most of the patients with suspected pulmonary TB contact a private practitioner in the first place. Despite the role that the private sector plays as a major health care provider, it continues to be neglected by the public health system in the delivery of care [9]. Sputum smear examination is the most reliable way to diagnose patients with pulmonary TB, as it has a higher specificity though lower sensitivity. The positive predictive value of sputum smear result is more than 90%, while that of chest X-ray is around 66% [10]. There were about 13,000 DMCs in the country during 2012 and sputum microscopy forms the backbone of RNTCP [11].

Majority of the PPs do not refer their patients to the DMCs for sputum examination leading to lesser case detection [12–14]. In a study conducted in Andhra Pradesh (India), compliance to internationally recommended standards of TB (ISTC) among private physicians was quite poor as only 34%, meaning only few practitioners rightly suspected TB and prescribed sputum microscopy for diagnosis of TB as their first choice [12]. Studies show that most of the PPs generally rely on other investigations like X-ray, ESR and serological tests for diagnosis of TB [13]. In another study conducted in Gwalior district, more government practitioners relied on sputum examination for diagnosis and follow-up as compared to PPs [14].

Study conducted in Ballabgarh, Haryana (India), documented that among 242 pulmonary TB suspects, 61.6% were not prescribed a sputum AFB test, even after they had presented with complaints suggestive of pulmonary TB [15].

In the present study, overall knowledge regarding TB (37.7%) and RNTCP (25.6%) among PPs was very low in the non-referral group. Allopathic stream of PPs had significantly more knowledge regarding TB. Similarly, a knowledge, attitude and practice study conducted among non-allopathic private practitioners in a rural setting stated that though majority of the non-allopathic practitioners knew the causative agent, only 5% of the 128 practitioners could answer the correct full form of DOTS under RNTCP and only 14% could tell sputum examination to be the first investigation of choice in suspected TB case indicating a low knowledge among non- allopathic PPs. [16]

While evaluating some of the other factors responsible for not referring suspected TB patients to the nearest DMCs in the present study, lack of awareness and accessibility about nearest DMC, lack of contact and sensitization by RNTCP staff emerged as important factors. Fear of losing patients and loss of faith in government sector were not the factors responsible for non-referral of patients in our study. Majority of practitioners from the non-referral group denied being ever contacted by the RNTCP staff. Given the magnitude of the problem of TB and the size of the private sector in India, ignoring the private sector or dealing with it in a superficial manner cannot be an option. So, more awareness generation about RNTCP, training workshops by RNTCP and establishing new DMCs to cater to the ever growing population may be the need of the hour. Parallel mechanisms should be developed within the private sector to support it to diagnose TB patients properly as per RNTCP guidelines.

The main factors responsible for non-referral of pulmonary TB suspects to DMCs for diagnosis in the present study included low knowledge regarding RNTCP, lack of awareness regarding place and accessibility of nearest DMC, and inadequate sensitization by the RNTCP staff. Orientation of PPs to the policies and practices of national programmes must be carried out through continuing medical education and accreditation programmes in collaboration with their professional associations in order to build capacity of the private sector to effectively contribute to national disease control efforts.

The study was conducted with the funding from State Health Society-RNTCP, Punjab as an operational research project.