The use of confidence intervals for individual utilities: limits to formal decision analysis for treatment choice.

This paper discusses the use of confidence intervals for utility measurements. Classic test theory is applied to estimate confidence intervals for utilities. The theory is enhanced to calculate confidence areas for combined utilities and confidence bands for the threshold line. As an example it is shown that, if confidence intervals are taken into account, the implied preferred treatment of T3-larynx carcinoma patients is uncertain for a wide range of utilities, considering the mediocre reliability of most methods of utility assessment. This implies that although utility measurement and formal decision analysis can be a useful way to look at the decision problem, ambiguity, which must be resolved by other means, will often remain.

Delirium after cardiac surgery: a critical review.

Numerous articles have been published investigating the incidence of and risk factors for delirium after cardiac surgery. Smith and Dimsdale reviewed the literature on postcardiotomy delirium in 1987 using a meta-analysis of 44 research studies. However, doubts about their methods and results caused the authors to re-examine the literature using these 44 references as well as computerized literature searches to gather research and review papers from medical journals. Delirium after cardiac surgery appeared to be ill-defined in most of these studies. The methods and instruments used to assess delirium proved to be very different, and the patient samples were rather heterogeneous. Therefore, in most cases, the results are not comparable. Only a small number of the studies that were examined fit the criteria for statistical meta-analysis. On the basis of our analysis, a tentative conclusion may be drawn that the incidence of postcardiotomy delirium has declined slightly and that no strong risk factors have yet been identified.

Dimensionality and hierarchical structure of disability measurement.

Since the D-code of the International Classification of Impairments, Disabilities, and Handicaps (ICIDH) in its full form has proven to be impractical, an instrument based on a selection of 28 items is used to measure disability in Dutch patients undergoing rehabilitation. The items are categorized into 5 domains of physical, activities of daily living (ADL), social, psychological, and communicative activity. Measurement is made on a 4-point scale ranging from 0 (not disabled) to 3 (severely disabled). As a result of the ordinal character of the rating, statistical and mathematical manipulations of the scores are complicated. The aim of this study was to obtain more insight in the dimensionality and hierarchical structure of the items, to overcome problems in comparing disability between items, between patients, and within patients between different moments in time. Mokken scale analysis of the disability scores from 1,967 rehabilitation inpatients showed that the 28 items constitute hierarchical scales. However, categorization of the items into the 5 original domains was not replicated. Five other scales or dimensions were investigated, measuring the level of extended ADL, extended psychological, fine motoric, work/leisure, and hearing/seeing activity, respectively. The number of items per dimension ranges from 14 in the extended ADL dimension to 2 each in the work/leisure and hearing/seeing dimensions. Although each disability item may be of importance in clinical case management, a reduced set of extended ADL items suffices to describe the disability level in this dimension for epidemiological research purposes. The other dimensions need further specification to provide reliable and sensitive measuring of disability.

Patient participation in clinical decision-making for treatment of T3 laryngeal cancer: a comparison of state and process utilities.

PURPOSE: To study the use of two different approaches, and feasibility of four commonly used utility assessment methods to assess preferences for treatment of T3-laryngeal cancer by surgery or radiation therapy (RT). METHODS: Utility assessment methods, namely, time trade-off (TTO), standard reference gamble (SRG), rating scale (RS), and direct comparison (DC), were used to assess utilities in two groups of former cancer patients (n = 10 for both), a group of clinicians (n = 9), and a group from the general population (n = 10). For the treatment modalities, ie, surgery and RT, two types of scenarios were developed and used: the state scenario, which describes a stable health state after treatment, and the process scenario, which describes a dynamic process. First, utilities were assessed based on state scenarios. Next, respondents were thoroughly informed and educated with respect to the relevant aspects of both treatment modalities. Subsequently, utilities were again assessed, but now based on the process scenarios. The outcome of each approach was calculated and expressed in a quality-adjusted life-expectancy (QALE) score for each treatment modality, and the treatment with the highest outcome was said to be the preferred treatment modality. RESULTS: In general, a higher QALE score for each treatment modality was found for clinicians and for the general population as compared with the former-cancer-patient groups. When the outcome of both approaches was compared on an individual level dependent on the utility assessment method, 32% to 43% of respondents showed an inconsistent treatment preference. CONCLUSION: The approach to assess utilities and the extent to which respondents are informed about treatment modalities have a major effect on individual treatment preferences.

Characteristics of patients from hospital-based in-patient rehabilitation in The Netherlands.

The University Hospital Dijkzigt in Rotterdam is one of the first hospitals in The Netherlands to have a rehabilitation department providing multidisciplinary, hospital-based-in-patient rehabilitation (HBIR). The main goals of HBIR are to estimate the final optimum level of patient's functioning, to investigate the most appropriate treatment setting to achieve this and to fulfil requirements for discharge from hospital as soon as possible. One of the central issues in rehabilitation in the Netherlands is to obtain more insight into the current daily practice of HBIR. The aim of this study was to describe the characteristics of the patients treated in HBIR in Dijkzigt Hospital. Registered data from 1967 HBIR patients treated between 1988 and 1990 were analysed retrospectively. The patients were on average 53.1 ( +/- 19.7) years old; the men (63%) were significantly younger than the women (mean age of 50.8 and 57.0 years, respectively). The main diagnostic groups were stroke (27%), progressive and regressive neurological conditions (17%), hand injuries (9%), spinal cord lesion (9%), orthopaedic injuries (8%) and amputation of lower extremity (8%). The median duration of HBIR and hospitalization were 13 days and 33 day, respectively. Most patients were discharged home (63%), or to a nursing home (14%) or a rehabilitation centre (8%). The results of this study can be used in policy discussions on HBIR and provides a description of the population which needs to be studied longitudinally in future research on the outcomes of HBIR.