Over the last 20 to 30 years there has been a radical change in the way in which we approach people with mental handicaps and associated physical, emotional or social disabilities. This is reflected in the use of new terms such as “learning disabilities”. The new approach is designed to recognise the learning potential of such people, to acknowledge that they, as citizens, have equal rights, and to encourage acceptance of the fact that they are, indeed, individuals. As a result, the various types of institutional care, in particular, that reflected the once prevalent social policies and attitudes towards such people have undergone a marked change. New models of care have emerged that go far beyond simply making asylums more suitable places for human beings to live in. These new models of care rely on the philosophy of normalisation and the process of deinstitutionalisation and rehabilitation; they are community based, and they provide improved survival and quality of life, as well as opportunities for the individuals to develop their potential by means of education and vocational training, which in many cases will make possible their social integration.

A three-year deinstitutionalisation and rehabilitation pilot intervention project was implemented at Leros PIKPA for people with severe learning disabilities. Initial conditions at the asylum were appalling. Residents suffered severe deprivation, extreme institutionalisation, and violation of basic human rights. Intervention involved professionals from different disciplines, and involved residents, their families, care staff, the institution, and the local community. As a result, resident care and adaptive behaviour has started to improve. Communication between residents and families has increased. Owing to training and sensitisation, care staff's poor resident-management practices and negative attitudes toward disabled people have changed. Living and hygienic conditions have been upgraded and building renovation is under way. Asylum administration and the local community have been sensitised to residents' needs. Eleven residents have moved to the project's pilot community home in Athens; two others now live with foster families. The results suggest that deinstitutionalisation and rehabilitation can be successfully initiated even in residential institutions of the severest kind.

The pilot intervention project at Leros PIKPA asylum was resisted at all levels of its implementation. Resistance ranged from implicit and passive to explicit and hostile. It took the form of strong rejective attitudes, defensiveness, and repressed emotions of guilt, shame and fear. It was apparent in efforts to delay, obstruct or reverse the progress of the project and undermine the work of intervention team members. It was evident in the interactions of the intervention team with all parties involved in implementing the project: medical professionals and welfare services, the state and local administrations, the central and local administrations of PIKPA staff, the local community, and the asylum residents' families and relatives. It is argued that this resistance to change is due to the interplay of cultural/attitudinal, psychodynamic and socio-economic factors that are not unique to Leros or Greece. The analysis of the psychosocial processes involved in resistance to changing the Leros PIKPA asylum may offer insights to the deinstitutionalisation of custodial settings in general.

The pilot intervention project at Leros PIKPA asylum sought to change the management of residents by decreasing care staff's tendency to view people with learning disabilities with attitudes of custodial segregation (e.g. “they should all live in institutions”) and categorisation (e.g. “they are all alike”). To that end, staff were trained on the job in resident management and participated in seminars, sensitisation groups, clinical case presentations and visits to model rehabilitation units. To investigate whether staff attitudes towards people with learning disabilities had been influenced, staff members responded to a questionnaire of such attitudes, in March 1991 (initial assessment) and a year later (reassessment). Factor analyses of attitude ratings at initial assessment revealed a strong negative attitudinal construct, “Rejective and custodial segregation – Categorisation”, according to which institutionalisation was the only appropriate type of care for people with learning disabilities, regardless of individual differences. At reassessment, however, this factor had split into two independent ones, “Rejective and custodial segregation” and “Categorisation”, suggesting that the staff had understood that custodial care did not apply to all persons with learning disabilities indiscriminately. Reassessment factors further indicated that staff had comprehended the appropriateness of alternative care forms – community and home care. Analysis of Categorisation and Custodial segregation attitude ratings showed that only staff's tendency to view the disabled as “all alike” (Categorisation) had decreased at reassessment. The latter change and the structural change revealed by the factor analyses imply that staff's attitudes underwent a small but significant modification.

The life expectancy of institutionalised people with learning disabilities is shorter than that of the general population. Data on population dynamics in institutions for such people are vital for planning purposes. Mortality can be considered a crude measure of quality of health care. Mortality data on the 914 admissions to the Leros PIKPA asylum for children and young adults with learning disabilities and associated problems in the years 1961–91 were reviewed. The overall crude mortality rate was 59.2 deaths per 1000 person-years. Twenty-two per cent of the deaths occurred within a year after admission. Age-specific mortality rates were particularly high for those aged one to four years, and declined thereafter. Male residents had lower mortality than female residents in almost all of the age groups. Compared with sex- and age-specific mortality data for the general population of Greece, the observed rates were 20–150 times higher but still comparable to those reported for people with more severe learning disabilities in institutions in other countries. Lifetable analysis by length of stay showed that male residents had a statistically significant higher probability of survival than female residents, which could not be attributed to age-related differential mortality. Culture-specific differential admission criteria may account for the observed sex difference.

The intervention carried out at the Leros PIKPA asylum with the assistance of European Community funds has proved beyond all doubt that financial resources, essential as they may be, are not in themselves sufficient to ensure the success of an effort to reform a custodial care structure. It might seem a unique opportunity for a country to receive financial support to improve its public mental health care system, especially during a period of economic recession, and one might have expected that, if the financial resources were sufficient, reform would be relatively easy. However, our experience has shown that this is not the case and that much time and effort are needed. Emphasis needs also to be placed on a number of factors that we have found to be of decisive importance.