Géraud Dautzenberg

Summary and general discussion 7 157 As far as unmet needs are concerned, a different picture can be seen in the research. Here, the patients score a much higher number in the relative sense; 0.81 vs 0.49. Although this is a small difference in the absolute sense, it is of great significance. The practitioners underestimated (compared to the patients) the mean number of unmet needs, whereby one out of five reported needs according to the patients had not been met for all domains combined. Therefore, there is a large discrepancy in reporting unmet needs between patients and practitioners, especially in the psychological and social domains, resulting in nearly one out of the three needs not being met in the latter according to the patients. The other domains were scored equally between patients and practitioners. This was also reflected in the Cohen’s Kappa coefficients, a statistical measure of agreement between the staff and the patients, in which all high agreement ratings (good and very good agreement expressed in kappa values of 0.61-0.80 and higher) occurred in the environmental and physical domains, but in which almost all items belonging to the social or psychological domains had a low agreement (κ < 0.40) between the ratings of patients and practitioners. At the item level, the lowest per cent agreement between patients and practitioners was on the items of company, medication, and physical health. As for memory, the staff scores more needs than the patients for met needs, but less for unmet needs. These results seem to be in line with the expectations that patients tend to report fewer memory complaints. In relation to patient characteristics and correlations with the number of needs, our study found that most of the clinical variables, measured: age, MMSE GAFp, CES-D, and YMRS, show a correlation only with the total needs and not with unmet needs scored by patients or their staff. However, there are exceptions to this ‘rule’. The staff rating of total needs only correlated with depressive symptoms (CES-D) and not with mania symptoms (YMRS). Age showed only a correlation with total needs as scored by the staff, and not when rated by patients. As for the correlation with unmet needs, all, besides GAF and patient rating, show no significant correlation. Therefore, if we consider the above findings’ clinical characteristics, we see a tendency that most of these characteristics are recognised by staff and patients to influence the number of total needs, but they are accounted for as there is no correlation with unmet needs. In contrast, the network size, social participation, and quality of life (MANSA), summarised as social variables, tend to have a significant negative correlation with total needs as well as the number of unmet needs. Network size is the exception for unmet needs, according to the staff. This could be explained by the need for these variables not having been accounted for.

RkJQdWJsaXNoZXIy MTk4NDMw