System-level change in mental health services in North Wales: An observational study using systems thinking.

PURPOSE: To describe changes to mental health services using systems thinking. METHOD: Structured standardized quality of life assessment (Manchester Short Quality of Life Assessment: MANSA) was used to establish service user priorities for changes to service provision (part of a process known as check in systems thinking). Current service performance in these priority areas was identified, and changes to service arrangements were planned, implemented and monitored by task and finish (T&F) groups (making use of a process known as flow in systems thinking). RESULTS: 81 MANSA assessments were completed at the check stage (by NM). Work finances and leisure activities emerged as service user priority areas for change, and T&F groups were established with representation of all sectors and service users. Ways to make improvements were observed, planned and implemented by T&F groups (the flow stage). CONCLUSION: The systems approach reveals how services and quality of life have been changed for patients in Wrexham. Further generalizable research is needed into the potential benefits of using systems thinking in mental health service evaluation.

The Lancashire Quality of Life Profile: modification and psychometric evaluation.

BACKGROUND: Although widely used in Europe, the conceptual and psychometric qualities of the Lancashire Quality of Life Profile (LQoLP) have not been thoroughly examined. Four issues need attention: coverage, scale construction, systematic missing data, and psychometric properties. METHOD: Concept mapping was used to examine the coverage, and exploratory factor analysis to examine the empirical scale structure of the LQoLP. Data of 518 long-term patients from ten different mental health care settings were used. Modifications to the LQoLP were proposed on the basis of these findings, and its psychometric properties were tested. Thirty-one respondents participated in a test-retest reliability study (T1-T2: 2 weeks). RESULTS: The modified LQoLP covers the quality of life-concept in a more comprehensive manner. Internal consistency, test-retest reliability and validity are good. CONCLUSIONS: The modified version of the LQoLP now covers ten domains, paying specific attention to patients' definition of quality of life (autonomy, coping, self-worth). Domains are now based on factor analysis. The problem of systematic missing data is solved. Psychometric properties are good. Because of moderate alphas, two domains need further investigation.

Do different subjective evaluation criteria reflect distinct constructs?

Four subjective evaluation criteria--self-rated symptoms, subjective quality of life, self-rated needs, and patients' assessment of treatment--were examined to determine whether they reflect distinct constructs and to what extent they are correlated. The four criteria were assessed in 90 newly admitted schizophrenia patients, 170 long-term-hospitalized schizophrenia patients, 154 patients with alcoholism admitted for the short term, and 68 patients with alcoholism in long-term rehabilitation, using identical instruments. The four criteria show substantial intercorrelations, except for assessment of treatment in the two acute treatment groups. One general factor explains between 43% and 55% of the variance in each group. Factor scores are associated with observer-rated psychopathology and objective data. The four criteria overlap in different patient groups, except in the assessment of treatment in acute groups. The use of more than one subjective criterion should occur only with specific hypotheses. A better theoretical framework is needed to explain the differences between and interrelationships of subjective evaluation criteria.

Measuring the quality of life of severely mentally ill people using the Lancashire Quality of Life Profile.

Quality of life (QOL) has become an important outcome measure for many disorders, including mental illness. The Lancashire Quality of Life Profile (LQOLP) was developed for use in operational contexts, and has been translated into several languages. It is in use in several European and North American community psychiatric services. The present paper addresses the questions: how easy is it to use?; how reliable is it?; do the results of the LQOLP vary by setting in a meaningful way?; how do the results co-vary with measures of clinical symptoms and social functioning?; how well does it measure change?; is it clinically useful? While most of the answers to these questions are favourable, there is a need for further research and development of the profile, in particular with reference to the consequences of the use of the profile as a routine monitoring instrument and the most appropriate form of statistical analysis in longitudinal data-sets.

Partnership in transsexualism. Part I. Paired and nonpaired groups.

In a group of 72 transsexuals, 26 out of 55 male-to-female transsexuals had partners and 9 out of 17 female-to-male transsexuals had partners. Forty percent of the group of 35 paired transsexuals had been married, but most of these marriages had ended in divorce or separation. Six male patients were still married at the time of the enquiry. The two groups differed significantly in several respects. Those with partners (the paired group) had more fathers or heads of family in social classes I-III, while those without partners (the nonpaired group) were more often without their father in the first decade of life. Significantly more of the paired group received psychiatrists' rating of stable social adjustment. There was a tendency for the paired group to have a more successful employment history since leaving school and to change their National Insurance cards more often in order to obtain employment in their adopted gender. The groups did not differ in the amount of social drift, self-confessed criminal behavior, age at referral, history of prostitution, or incidence of rejecting parents. There were also no significant differences on the following psychological tests: Wechsler Bellevue, Progressive Matrices, Mill Hill Vocabulary Scale, Maudsley Personality Inventory, Slater Selective Vocabulary--except that the nonpaired group knew fewer male words. Any differences between the two groups are perhaps best described in terms of social adjustment rather than any background social factors, personality, or clinical differences.

Partnership in transsexualism. Part II. The nature of the partnership.

Thirty-five partners were identified in a group of 72 transsexuals. Six male-to-female transsexuals continued to live with their marriage partner, but the remainder of the total group cohabited with a member of the same biological sex. Two of the female-to-male transsexuals had undergone a marriage ceremony with their present partner, but the remainder had not. In most cases the social class of the partner was the same as that of the transsexual. In one case two transsexuals were living together, and this case and three others out of a total of six who were interviewed are described in more detail. The partnerships described support the view that transsexualism is best regarded as an overvalued idea which may be accepted or shared by a partner. The extent to which the idea is shared is related to the bond of affection between the partners, the pattern of dominance between them, and the sexual needs of the partner.

The prediction of the course of minor psychiatric disorders.

Recent studies have suggested that psychosocial factors play an important part in the prediction of the course of minor psychiatric disorders. Fifty-nine new psychiatric out-patients suffering from minor disorders were assessed, using standardized clinical and social interviews, and 52 were followed up after one year and the clinical assessment repeated. Social and clinical factors were equally important predictors of the number of months ill in the survey year, but social and constitutional variables were superior in the prediction of percentage change in symptoms over the year. The results of correlation, factor and multiple regression analyses suggest that the course of minor psychiatric disorder is best predicted by three sets of variables which are, in order of importance, the patient's material social circumstances, his clinical symptoms and his 'genetic risk' scores.