Rainer W. Alexandrowicz

The criterion validity of the Geriatric Depression Scale: a systematic review

Objective:  The objective was to provide a systematic review of the screening accuracy of both versions of the Geriatric Depressions Scale (GDS‐30, GDS‐15).

The Consequences of Non-Cognitive Symptoms of Dementia in Medical Hospital Departments

Objective: To our knowledge, there are no studies investigating the non-cognitive symptoms of patients with dementia such as depression, agitation, or delusions among general hospital inpatients. The aim of this study was to investigate the frequency of such non-cognitive symptoms among medical inpatients and to analyze their impact on the length of hospital stay and on admission to nursing homes. Method: The sample consisted of 372 elderly inpatients admitted to four internal medical departments (i.e., not including psychiatric wards) in Austria. Patients were investigated by research psychiatrists using the Clinical Interview Schedule. For the analyses of the non-cognitive symptomatology, only marked and severe symptoms were included. To identify predictors for the length of hospital stay and for nursing home placement, multivariate regression procedures were used. Results: Of all inpatients, 27.4% met criteria for dementia according to DSM-III-R. Of those with dementia, 27.8% had marked or severe non-cognitive symptoms. A diagnosis of dementia markedly increased the risk for nursing home referral and prolonged the duration of inpatient treatment. Among the demented, both, cognitive and non-cognitive symptoms turned out to be significant predictors for nursing home placement and for prolonged duration of acute hospital stay, even when controlling for other independent variables. Conclusions: Non-cognitive symptoms occur frequently among medical inpatients with dementia and considerably increase both the duration of inpatient treatment and the risk of nursing home placement. Since such non-cognitive symptoms are treatable, they should receive attention from the hospital staff.

The "carers' needs assessment for Schizophrenia": an instrument to assess the needs of relatives caring for schizophrenia patients.

ObjectiveFor the purpose of service planning, an instrument was developed for the systematic assessment of interventions needed by the caregivers of schizophrenia patients.MethodThe development of this instrument was based on in-depth interviews and focus groups. It consists of 18 areas describing common problems of schizophrenia caregivers. For each of these problem areas several possible interventions are offered. Concurrent validity, inter-rater and retest reliability were tested among 50 caregivers.ResultsThe kappa values for the inter-rater reliability are predominantly excellent (kappa > 0.75). The values for the retest reliability show a wide range between excellent (kappa > 0.75) and fair agreement (kappa 0.40–0.60). The significant correlations found between summary scores of this new instrument and several sub-scales of the Family Problem Questionnaire support the assumption that the concurrent validity is satisfactory.ConclusionThese results suggest that this instrument is both valid and reliable.

The criterion validity of different versions of the General Health Questionnaire among non-psychiatric inpatients

PurposeWhile the General Health Questionnaire (GHQ) has an excellent screening performance among outpatients and in the community, its accuracy to detect mental disorders among non-psychiatric inpatients was reported to be lower. The aim of the present study was to compare the criterion validity of different scoring methods, i.e. the Bimodal scoring, Likert scoring, Modified Likert scoring and Chronic scoring, of the 30-, 20- and 12-item version of the GHQ.MethodsThe GHQ was available from 511 inpatients recruited from surgical, medical, gynaecological and physical rehabilitation wards. The Clinical Interview Schedule was performed for psychiatric case-identification and was used as external care criterion.ResultsSensitivities of all versions were between 0.612 and 0.701, and specificities between 0.601 and 0.759. The Overall Misclassification Rate (OMR) varied for the Bimodal and the Modified Likert scoring method between 0.257 and 0.281, for the Likert and the Chronic scoring method between 0.325 and 0.386. Overall, the Bimodal and the Modified Likert scoring method showed significantly better specificity and OMR than the Likert and the Chronic scoring method, while we could not find any differences for sensitivity.ConclusionsOverall, the Bimodal and the Modified Likert scoring method seem to be more accurate than the Likert and the Chronic scoring method. Nevertheless, due to the high misclassification, none of these versions can be recommended for routine screening among non-psychiatric inpatients.

Need and utilization of psychiatric consultation services among general hospital inpatients

ObjectiveThe aim of the present study was to estimate the need for psychiatric consultation services and psychiatric inpatient referral among inpatients of non-psychiatric hospital departments, and to assess the actual utilization of these services.MethodsThe study was carried out among 728 inpatients of physical rehabilitation wards, medical, surgical, and gynecological wards in Austrian hospitals. Psychiatric case identification was performed by research psychiatrists using the Clinical Interview Schedule (CIS). Diagnoses were given according to DSM-III-R. The assessment of need for consultation and inpatient referral was based on the clinical judgments of research psychiatrists.ResultsOverall, 34.2% of the inpatients were cases according to the CIS-criteria, 51.8% of them needing either psychiatric consultation or inpatient referral according to research psychiatrists. In 66.7% of those for whom research psychiatrists had stated a need, this need was not met (“unmet need”), while only 33.3% of them had their need met. In contrast, a psychiatric consultation was performed among 5% of those patients not needing psychiatric services according to the research psychiatrists (“overprovision”). Variables of the health care system (i.e. department type and catchment area of the hospital) were among the predictors for “met needs”.ConclusionThe rate of actual psychiatric consultations and admissions to psychiatric wards was markedly lower than the need according to research psychiatrists’ judgment.

Sample Size Determination Within the Scope of Conditional Maximum Likelihood Estimation with Special Focus on Testing the Rasch Model

This paper refers to the exponential family of probability distributions and the conditional maximum likelihood (CML) theory. It is concerned with the determination of the sample size for three groups of tests of linear hypotheses, known as the fundamental trinity of Wald, score, and likelihood ratio tests. The main practical purpose refers to the special case of tests of the class of Rasch models. The theoretical background is discussed and the formal framework for sample size calculations is provided, given a predetermined deviation from the model to be tested and the probabilities of the errors of the first and second kinds.

Zur Validität eines zweistufigen Screenings am Beispiel des Depressionsscreening

OBJECTIVE Several studies reported that general practitioners often do not recognize depressive disorders among their patients. Routine screening is a time-consuming procedure which might be improved by means of a two-step procedure. Using a very brief screening instrument in a first step reduces the number of patients needing a longer conventional screening in a second step. Only screening positives after this second step will be examined by the general practitioner (GP). METHODS Based upon validity scores from published studies, the expected validity of such a two-step procedure is being investigated. RESULTS This two-step screening procedure decreases the overall number of screening positives who have to be examined for the presence of depression by the GP. While the combined sensitivity slightly decreases, the combined specificity increases. CONCLUSIONS The proposed method allows for a practicable routine screening for depression by the GP. The procedure might further be improved through the development of instruments optimized for the first and the second step.

“GANZ RASCH”: A Free Software for Categorical Data Analysis

This article presents a new and freely available tool for performing analyses according to the Rasch model (RM) and the latent class analysis (LCA). The software allows for the estimation of the model parameters and offers several measures of model fit. A graphical user interface (GUI) provides access to numerous options regarding data, models, and output. For educational purposes, an optional annotate feature allows to augment the output with brief explanations and citations regarding the procedures. Based on published data, the features of GANZ RASCH are briefly illustrated in two worked examples. The program intends to combine ease of use while allowing for performing a full-fledged analysis, thus targeting a wide range of users.

The importance of statistical modelling in clinical research : Comparing multidimensional Rasch-, structural equation and linear regression models for analyzing the depression of relatives of psychiatric patients.

SummaryBackgroundVarious studies have shown that caregiving relatives of schizophrenic patients are at risk of suffering from depression. These studies differ with respect to the applied statistical methods, which could influence the findings. Therefore, the present study analyzes to which extent different methods may cause differing results.MethodsThe present study contrasts by means of one data set the results of three different modelling approaches, Rasch Modelling (RM), Structural Equation Modelling (SEM), and Linear Regression Modelling (LRM).ResultsThe results of the three models varied considerably, reflecting the different assumptions of the respective models.ConclusionsLatent trait models (i. e., RM and SEM) generally provide more convincing results by correcting for measurement error and the RM specifically proves superior for it treats ordered categorical data most adequately.ZusammenfassungHintergrundIn zahlreichen Studien konnte gezeigt werden, dass die Betreuung schizophrener Patienten durch Angehörige eine große Belastung für die Angehörigen darstellen und zu Depressionen der Angehörigen führen kann. Allerdings unterscheiden sich diese Studien in ihrer statistischen Herangehensweise, was einen Einfluss auf die Ergebnisse haben kann.MethodikDie vorliegende Studie vergleicht daher drei verschiedene statistische Modelle, das Rasch-Modell (RM), das Strukturgleichungsmodell (SEM) und das lineare Regressionsmodell an einem Datensatz.ErgebnisseEs zeigten sich zum Teil deutliche Unterschiede zwischen den drei untersuchten Modellen, die sich durch deren spezifischen Annahmen erklären lassen.SchlussfolgerungenLatent-trait-Modelle (d. h. das RM und das SEM) führen im Allgemeinen zu überzeugenderen Ergebnissen, da sie den Messfehler korrigieren. Das RM erweist sich darüber hinaus als überlegen gegenüber den anderen Modellen, da es ordinalskalierte Daten am angemessensten verarbeiten kann.

Using Rasch-models to compare the 30-, 20-, and 12-items version of the general health questionnaire taking four recoding schemes into account

SummaryBackgroundThe present study compares the 30-, 20-, and 12-items versions of the General Health Questionnaire (GHQ) in the original coding and four different recoding schemes (Bimodal, Chronic, Modified Likert and a newly proposed Modified Chronic) with respect to their psychometric qualities.MethodsThe dichotomized versions (i.e. Bimodal, Chronic and Modified Chronic) were evaluated with the Rasch-Model and the polytomous original version and the Modified Likert version were evaluated with the Partial Credit Model.ResultsIn general, the versions under consideration showed agreement with the model assumption. However, the recoded versions exhibited some deficits with respect to the Outfit index.ConclusionsBecause of the item deficits and for theoretical reasons we argue in favor of using the any of the three length versions with the original four-categorical coding scheme. Nevertheless, any of the versions appears apt for clinical use from a psychometric perspective.ZusammenfassungHintergrundDie vorliegende Studie hatte zum Ziel, die 30-, 20- und 12-Item-Versionen des General Health Questionnaire (GHQ) und vier verschiedene Rekodierungen (Bimodal, Chronic, Modified Likert und eine erstmals vorgeschlagene Modified Chronic) bezüglich psychometrischer Ansprüche zu vergleichen.MethodikDie dichtomisierten Versionen (Bimodal, Chronic and Modified Chronic) wurden mittels Rasch- Modell und die polytome Originalversion sowie die Modified Likert Version mittels Partial Credit Modell untersucht.ErgebnisseInsgesamt zeigten die untersuchten Versionen weitgehende Übereinstimmungen mit den Modellannahmen. Nichtsdestotrotz wiesen die rekodierten Versionen Defizite bezüglich des Outfit Index auf.SchlussfolgerungenAufgrund der Item-Defizite und aus theoretischen Überlegungen argumentieren wir für die Verwendung der originalen 4-kategoriellen Kodierung für alle drei Versionen des GHQ, jedoch ist jede der Versionen aus psychometrischer Sicht für den klinischen Gebrauch geeignet.