Acta psychiat. scand. (1977)55, 299-308 Department of Psychiatry (Heads: Prof. N. Kessel and Prof. D. Goldberg), University of Manchester, Manchester, England
A standardized psychiatric assessment scale for rating chronic psychotic patients M. Kruwrecsrl, D. GOLDBERG AND M. VAUGHAN Simple 5-point scales are described together with the method used
to study their reliability, the results of which are shown. The scales are short, easy to administer and sensitive to change, therefore particularly applicable where there is the need for screening chronic psychotic populations. Some uses for the scales are suggested. Key words: 5-point rating scales - screening - reliability psychotic populations.
- chronic
The objective rating of mental state has always been of considerable interest and the assessment and measurement of clinical therapeutic progress has increased the need for a scale suitable for the purpose. It would be necessary for such a scale to be short and yet to allow a reliable assessment of chronic patients in particular and to be sensitive to change in their status. So far no scale fulfils all these criteria. Of the existing scales the Wittenborn (Wittenborn (1955)), the Mental Schedule (Spitzer et al. (1964)), the In-Patient Multi-dimensional Psychiatric Scale (IMPS) (Lorr et al. (1963)), Present State Examination (Wing et al. (1967)) and The Clinical Interview Schedule (Goldberg et 41. (1970)) are lengthy. Only Overall & Graham’s (1962) shortened version of IMPS is more manageable from that point of view, but unfortunately it suffers from comparative insensitivity to change. Although the Present State Examination (Kendell et al. (1968)) yields very reliable ratings for diagnostic purposes it was not designed primarily for measuring change and its length (140 items) makes it difficult to use in this setting. Wing (1961) in an early paper described a simple interview using s-point scales allowing a reliable classification of schizophrenic patients; the main symptom areas were: flatness and incongruity of affect, poverty of speech, incoherence of speech and coherent delusions. The Clinical Interview Schedule of Goldberg et al. (1970)also uses 5-point scales including most of those described by Wing (1961) but enables the research worker to make a total of 22 ratings, thus making it possible to make a diagnosis in community settings. There are obvious advantages in scales which can be used for the assessment of therapeutic change but also allow researchers to classify patients according to Wing’s Scale. The aim of the present study was to produce a short set of rating scales which would provide a reliable clinical assessment of chronic psychotic patients and which would be sensitive to changes in their conditions.
300 METHOD Videotape recordings were made of 10 known psychotic patients being interviewed individually by a psychiatrist. Two psychiatrists acted as interviewers seeing five patients each. It was also possible to produce a video tape in the form of a separate interview explaining the method of scoring the scales with the help of a manual produced for the purpose. This allows new raters to train themselves with the explanatory tape and manual and to compare their ratings on the other 10 tapes with those of the original constructors. This means that it is a relatively simple matter for a new centre to become skilled in the use of the tapes and in that way comparisons can be made between clinical populations. The scale is not intended to be used blindly and the rater is expected to be at the very least familiar with the patient’s notes, and must be aware of the areas in which the patient has previously experienced symptoms. The first four ratings are based on the patient’s replies to questions, the next four on the rater’s clinical observation of abnormal phenomena. If the scale is to be used for the purpose of a drug trial, side-effects can also be rated. The whole scale is shown in Appendix 1. The Manual in the form of “Guide Lines for the use of Five-Point Scales” is attached in Appendix 2. To take an example of depression from the manual: in order to make a rating of depression in the course of the preceding week, the rater must detect a definite persistent abnormality of affect making a contribution to the abnormal mental state. The rater is expected to make any additional probes which will enable him to form an opinion as to the presence of a particular symptom; but his probe questions have to include depression, anxiety, delusions and hallucinations. Other phenomena must be elicited by clinical observation (Appendix 1). Having recorded the 10 video tapes, a reliability study was conducted. The two interviewer psychiatrists independently rated all 10 tapes to provide scores for the comparisons described below. Each tape was viewed and rated before the next was seen. Three other psychiatrists were asked to train themselves using the training tapes and then to rate the 10 eriterion tapes. It was possible not only to determine the reliability achieved for these tape-trained raters but also to compare their ratings with those of the two psychiatrists who were extremely familiar with the scale. RESULTS The overall agreement between the psychiatrists was established using Kendall‘s Coefficient of Concordance, W.This coefficient is a linear function of the mean of the Spearman’s rank correlation coefficients for all possible pairs of raters and therefore reflects the average degree of association between the rater’s scores (Siegel (1966)). Kendall’s W was calculated for each scale using the ranked ratings of all five psychiatrists and then separately those of the three psychiatrists who had been trained only from the manual and training video tape. The results are given in
301 Table 1. Kendall‘s coefficient of concordance with significance levels for groups of five and three psychiatrists
Kendall’s Coefficient of ConcordanceW for five psychiatrists
P
Ratings based on replies to questions Depressed Anxious Delusions Hallucinations
0.87 0.64 0.83 0.78
A standardized psychiatric assessment scale for rating chronic psychotic patients M. Kruwrecsrl, D. GOLDBERG AND M. VAUGHAN Simple 5-point scales are described together with the method used
to study their reliability, the results of which are shown. The scales are short, easy to administer and sensitive to change, therefore particularly applicable where there is the need for screening chronic psychotic populations. Some uses for the scales are suggested. Key words: 5-point rating scales - screening - reliability psychotic populations.
- chronic
The objective rating of mental state has always been of considerable interest and the assessment and measurement of clinical therapeutic progress has increased the need for a scale suitable for the purpose. It would be necessary for such a scale to be short and yet to allow a reliable assessment of chronic patients in particular and to be sensitive to change in their status. So far no scale fulfils all these criteria. Of the existing scales the Wittenborn (Wittenborn (1955)), the Mental Schedule (Spitzer et al. (1964)), the In-Patient Multi-dimensional Psychiatric Scale (IMPS) (Lorr et al. (1963)), Present State Examination (Wing et al. (1967)) and The Clinical Interview Schedule (Goldberg et 41. (1970)) are lengthy. Only Overall & Graham’s (1962) shortened version of IMPS is more manageable from that point of view, but unfortunately it suffers from comparative insensitivity to change. Although the Present State Examination (Kendell et al. (1968)) yields very reliable ratings for diagnostic purposes it was not designed primarily for measuring change and its length (140 items) makes it difficult to use in this setting. Wing (1961) in an early paper described a simple interview using s-point scales allowing a reliable classification of schizophrenic patients; the main symptom areas were: flatness and incongruity of affect, poverty of speech, incoherence of speech and coherent delusions. The Clinical Interview Schedule of Goldberg et al. (1970)also uses 5-point scales including most of those described by Wing (1961) but enables the research worker to make a total of 22 ratings, thus making it possible to make a diagnosis in community settings. There are obvious advantages in scales which can be used for the assessment of therapeutic change but also allow researchers to classify patients according to Wing’s Scale. The aim of the present study was to produce a short set of rating scales which would provide a reliable clinical assessment of chronic psychotic patients and which would be sensitive to changes in their conditions.
300 METHOD Videotape recordings were made of 10 known psychotic patients being interviewed individually by a psychiatrist. Two psychiatrists acted as interviewers seeing five patients each. It was also possible to produce a video tape in the form of a separate interview explaining the method of scoring the scales with the help of a manual produced for the purpose. This allows new raters to train themselves with the explanatory tape and manual and to compare their ratings on the other 10 tapes with those of the original constructors. This means that it is a relatively simple matter for a new centre to become skilled in the use of the tapes and in that way comparisons can be made between clinical populations. The scale is not intended to be used blindly and the rater is expected to be at the very least familiar with the patient’s notes, and must be aware of the areas in which the patient has previously experienced symptoms. The first four ratings are based on the patient’s replies to questions, the next four on the rater’s clinical observation of abnormal phenomena. If the scale is to be used for the purpose of a drug trial, side-effects can also be rated. The whole scale is shown in Appendix 1. The Manual in the form of “Guide Lines for the use of Five-Point Scales” is attached in Appendix 2. To take an example of depression from the manual: in order to make a rating of depression in the course of the preceding week, the rater must detect a definite persistent abnormality of affect making a contribution to the abnormal mental state. The rater is expected to make any additional probes which will enable him to form an opinion as to the presence of a particular symptom; but his probe questions have to include depression, anxiety, delusions and hallucinations. Other phenomena must be elicited by clinical observation (Appendix 1). Having recorded the 10 video tapes, a reliability study was conducted. The two interviewer psychiatrists independently rated all 10 tapes to provide scores for the comparisons described below. Each tape was viewed and rated before the next was seen. Three other psychiatrists were asked to train themselves using the training tapes and then to rate the 10 eriterion tapes. It was possible not only to determine the reliability achieved for these tape-trained raters but also to compare their ratings with those of the two psychiatrists who were extremely familiar with the scale. RESULTS The overall agreement between the psychiatrists was established using Kendall‘s Coefficient of Concordance, W.This coefficient is a linear function of the mean of the Spearman’s rank correlation coefficients for all possible pairs of raters and therefore reflects the average degree of association between the rater’s scores (Siegel (1966)). Kendall’s W was calculated for each scale using the ranked ratings of all five psychiatrists and then separately those of the three psychiatrists who had been trained only from the manual and training video tape. The results are given in
301 Table 1. Kendall‘s coefficient of concordance with significance levels for groups of five and three psychiatrists
Kendall’s Coefficient of ConcordanceW for five psychiatrists
P
Ratings based on replies to questions Depressed Anxious Delusions Hallucinations
0.87 0.64 0.83 0.78
