While in operation, the feasibility of conducting the CBR program at Thiruporur was well established . The program addressed most objectives set out by the National Mental Health Program. Strategies implemented included training lay community health workers, clinical services, networking, public awareness campaigns and implementation of simple psychosocial rehabilitation strategies through lay community workers. Training of health care personnel at the primary health care facilities aimed to meet the NMHP objective of integrating mental health care with primary care [8, 9].
When we made our exit from the project area in 1999, we were quite satisfied with the spectrum of activities we had initiated. However, a visit in 2005 proved to be a little unsettling and it was this, which motivated us to take up this research. In Thiruporur, our training over a period of several months did not lead to sustained psychiatric services being available to the community due to a variety of reasons. The primary factor was that physicians at PHCs were often on temporary posting and were transferred within 6 months. Other reasons included the fact that there was only one physician per PHC and he often had to deal with over 100 patients per day, leaving him little time to deal with psychiatric patients. Inclination and willingness to handle mental disorders had also much to be desired . PHC's were also unable to provide psychiatric medication to the patients as they had a limited budget . Budget allocation works on the principle of "greatest good for the greatest number" thereby allowing psychiatric patients to fall through the cracks.
The expectation that the patients would continue treatment in the centers to which they were referred was not realised primarily due to the fact that they found it expensive to commute to the referred center. This was despite the fact that they realized that regular treatment was necessary.
Another factor contributing to the discontinuation of treatment was the lack of awareness among families about the nature of the illness. When the symptoms were controlled and the patient had recovered sufficiently to become functional, he was seen as being "cured" and no longer in need of treatment.
Dropping out of psychiatric care is not an uncommon phenomenon, especially in the case of chronic mental illnesses. Several studies including our own have addressed this issue. While some have reported more drop-outs in improved patients who were satisfied with the level of care , the contradictory finding of persons dropping out because of lack of improvement was seen in other studies . Our own followup study revealed that those patients who had remitted had dropped out .
Alternatively when a patient on treatment relapsed the family lost faith in treatment. Conflicting explanatory models about mental illness  and the availability of religious "healing" centers in the vicinity further compounded this problem. Our own research revealed that people held more than one explanation for a particular condition that led them to seek simultaneously different kinds of interventions. This is best exemplified by the fact that we were able to convince some of the persons with psychoses who " were residing in temples" for cure of their disorder to start taking medicines.
Regular psycho-education programs must be conducted in the community to educate the families about the nature of their relative's illness and the need for sustained medical treatment.
Most support groups for persons with mental illness are located in urban areas and comprise of families of persons with mental illness. The Citizens' group initiative of this programme was quite different, being rural and comprising of persons who had not been directly affected by mental illness. The primary objective of this group was to mobilise resources within the community that would facilitate the implementation of the CBR program. During the period of one year that the group was functional, it acted as a vital resource in the community, assisting in identifying and referring patients, organising camps, disseminating information about camp calendars and helping transportation of patients from distant villages. This group however, stopped functioning as soon as the project co-ordinator left the area. Lack of adequate leadership among the members, limited access to a mental health facility and inability to raise funds accounted for the termination of this group.
Some of these factors apply also to the continuation of the self employment schemes. The family must be co-opted in the treatment and in the self-employment schemes if it is to be sustained as evidenced by the fact that when patients relapsed they had to give up their income generating resource. Schemes that tap the innate skills that the patient and the family already possess will ensure better sustainability.
A revolving fund will provide the seed money to help start self-employment schemes. Patients/families who take a loan from the fund to start a venture will pay back in full within a predetermined timeframe, enabling the cycle to continue. These issues need to be addressed in a comprehensive manner in all future endeavors.