A systems analysis of an area mental health service

Abstract

The mental health services existing in a defined geographic area in New Zealand were studied in an effort to establish priorities for service development. It was clear from a preliminary review of the services that while community based services were operating in the area, they were under-resourced and very poorly monitored. They did not provide care that was comprehensive and the population they were intended to service was not clearly specified. Information about who used the services, where they came from and why, was difficult to obtain. There were no psychiatric beds in the general hospital in the area which served a population of approximately 300,000. Acute inpatient care was provided in a 300 bed psychiatric hospital remote from the main centres of population. The acute area of this hospital provided 50 beds and appeared to be under strain most of the time. No standardised criteria for admission and discharge to and from the service were in place. The aims of the acute service, the boundaries of clinical responsibility between assessment, symptom control, treatment and continuing care were vague. To begin the study, a scale was developed and used to assess the feasibility of evaluating individual psychiatric services. A one day census of patients registered with the services was then carried out. The services included inpatient services, general hospital and court liaison services, community mental health centres, day hospital programmes, domicilliary nursing services, a substance abuse service, a rapid assessment service and a child and family service. The census uncovered a lack of adequate record systems for both patient and management information. As a consequence, a patient case register, based on a minimum data set of patient information, was developed. Meantime, a preliminary review of the acute services in the psychiatric hospital revealed a service under stress. Major sources of this stress were identified as inadequacies in the physical environment and a large and apparently uncontrollable variation in amount of service use by individuals admitted to the acute area. In the last phase of the research, the problem of uncontrollable variance in amount of acute inpatient service use was examined. This focus on acute inpatient care was justified because it was the most restrictive and expensive form of care offered and there was clearly a need to describe and explain how this service worked and whether it was cost effective. A study of five hundred and thirty eight people admitted to the acute inpatient service during 1988 was designed to determine the factors which predicted amount of service use. An instrument was constructed to record information on selected predictor variables. The instrument was built around the registration form developed in the process of establishing the patient information system. Predictor variables were chosen from an extensive literature review of factors determining service use, and from clinical and management opinion on what information was most necessary for service development and clinical management. Choices were also made to include information which was being gathered at the time, albeit unsystematically, by existing clinical history taking and admission and discharge procedures. The research method involved selecting a comprehensive range of predictor variables as potential indicators for amount of acute psychiatric service use, and dividing them into systems variables, which were by definition outside the control of the patient, socio-economic variables and behavioural variables which were consumer characteristics. A range of socioeconomic, systems and behavioural variables was approved for inclusion in the study by the researcher, clinicians and management. These variables were measured and related to the criterion variable, amount of service use (ASU). The criterion (ASU) was defined as number of bed days spent in the acute inpatient service during a twelve month period. Variables categorised as systems variables (environmental and service features over which the consumer had no control), were found to be the strongest predictors of service use, compared to sociodemographic factors or behavioural characteristics of the users. This suggested that the system of service was provider rather than consumer driven. The results also indicated that, in the absence of evidence to support the effectiveness of acute inpatient care, there was a need to develop alternatives. There was an equally urgent need to implement evaluation procedures which will begin to measure the effectiveness of care delivered in all mental health service settings. The methods developed in this thesis (to determine evaluation feasibility, to develop a client case register, to determine predictor variables for amount of service use) resulted in a useful general model which can be used to describe, explain and to some extent predict and control patterns of human service use in a variety of settings. The theoretical underpinnings for the work, which were originally derived from general systems theory, from the discipline of evaluation research and from the community psychology literature, are strengthened and endorsed by the results. The data obtained extend the case register system developed in the early phases of the project to provide a substantial foundation on which to build a comprehensive data base. This can be regularly updated and expanded allowing for ongoing service evaluation. For the first time in the region, clinical and service policy decisions can be based on information which is both objective and reliable.