Ty-Bryn Unit, St Cadoc's Hospital, Lodge Road, Caerleon, Gwent NP6 1XQ,
1 Institute of Medical Genetics, University of Wales College of Medicine, Heath Park, Cardiff CF14 4XN and
2 Cardiff and Vale NHS Trust, Community Addiction Unit, House 56, Cardiff Royal Infirmary, Cardiff CF24 0SZ, UK
Received 21 December 1999; in revised form 28 September 2000; accepted 31 October 2000
ABSTRACT
This paper reviews the literature on the use of rating scales within the treatment of the alcohol-withdrawal syndrome. A computer-assisted literature search identified trials of therapy for and rating scales used in alcohol-withdrawal states. Eighteen rating scales were identified. There is a wide variation in symptom items included in these scales. Scales also vary in their length and ease of application. We conclude that it is important to use validated and reliable assessment scales in research if proper comparisons of treatments for the alcohol-withdrawal syndrome are to be made.
INTRODUCTION
One of the major problems for researchers and reviewers of treatment methods for alcohol withdrawal is the lack of a widely used, reliable and validated rating scale (Williams and McBride, 1998). Several different scales have been used within this field of research. Comparison difficulties are further exacerbated by the failure to use strict, comparable inclusion and exclusion criteria for study and control groups. Use of recognized diagnostic criteria, such as those laid out in ICD-10 (World Health Organization, 1992
) or DSM-IV (American Psychiatric Association, 1994
), with standardized ratings of dependency would aid comparison of the study populations.
An ideal rating scale in this area of research should: (i) aid the diagnosis of the withdrawal syndrome; (ii) indicate when drug therapy is required; (iii) alert staff to the development of serious withdrawal symptoms requiring more intensive medical input; (iv) reveal when medication can be discontinued and the patient safely discharged. Such a tool would be useful in research and would facilitate comparisons between studies on existing and newer medications. Study groups could be compared both in terms of symptom presentation and severity. This would then allow treatment response to be accurately and consistently measured. In clinical practice, such a tool would allow clinicians to assess and predict those who require pharmacological treatment on the basis of symptom severity and to titrate the dose required.
The aims of the present work were to identify rating scales used in the assessment of acute alcohol withdrawal described in the literature and then to compare their content and ease of application. Information with regard to reliability and validity was also sought.
METHODS
A computer literature search and reference search of review articles traced papers published in the English language between 1973 and 1999 on pharmacological treatments of alcohol-withdrawal states. The year 1973 was selected as the earliest date because it was the year Gross et al. (1973) published the Total Severity Assessment (TSA) and the shortened version, the Selected Severity Assessment (SSA). Of the 38 papers reviewed, 23 described rating scales in sufficient detail for their content to be analysed. Only those using a standardized system of scoring specific symptoms of withdrawal, and producing an overall measure of severity, were included. Four used a previously published scale so that, in total, 18 different scales were included.
RESULTS
Table 1 shows the symptoms and scoring systems used by all the studies. This reveals the lack of consensus between existing scales as to which symptoms constitute the most significant indicators of the alcohol-withdrawal syndrome. No single symptom was included in all the scales analysed and scales differed in the numbers of items included. A total of 30 symptoms and signs were described.
|
Table 2 illustrates the wide variations in the weightings given to different symptom groups by the rating scales. Figures given are the percentage contribution of each group of symptoms to the total scale scores. The seven groupings were those of affect, gastrointestinal (GI) disturbance, autonomic nervous system (ANS) disturbance, neurological disturbance, sleep, psychotic features and seizures. This comparison emphasizes the disparity in the weighting of items in the scales. Reasons for the particular make-up of the scales concerned were not published in the original descriptions. Some of the scales appear to be biased towards the known actions of the drugs under investigation. For example the scale which gives the highest weighting to autonomic symptoms, such as hypertension and raised pulse rate (Scale 11, Kraus et al., 1985), was used in investigating the beta blocker, atenolol. The scale that gave the highest weighting to seizures was used in investigating carbamazepine (Agricola et al., 1982
).
|
Total Severity Assessment scale
The TSA scale was developed by Gross et al. (1973) in an attempt to improve differentiation of degrees of severity in alcohol-withdrawal states and facilitate the quantification of the withdrawal syndrome. Gross et al. began by reviewing 10 previously published scales utilized in a variety of treatment studies. These varied from those that simply recorded the presence or absence of symptoms, to those that rated severity. Gross et al. (1973) considered them all to be inadequate and produced their own prototype TSA. The scale contains 30 variables which are rated on an 8-point scale. Zero indicates the absence of a symptom, whilst 7 indicates the maximum severity. The 30-item scale was intended as a research tool and a shorter 11-item scale, the SSA, was produced for clinical use.
Reliability of the instruments was assessed by randomly selecting 18 in-patients in acute withdrawal who were then assessed by two nurses, individually, each day. The different nurses visited the patients 21/2 h apart. Correlation coefficients were then calculated. Of the 30 items, eight, including three SSA items, were not evaluated. Six of the 22 items rated (including two SSA) did not show statistically significant correlation between the two assessors' scores. This left only 16 of the 30 items in the TSA (six of 11 SSA) with statistically significant correlations, suggesting that they might be reliable measures of withdrawal severity. Gross et al. (1973) explained the low level of reliability by the inherent fluctuation in symptom severity in what is an acute organic brain syndrome. Subsequent trials revealed the TSA to be valid when compared to global rating scales, but the extensive training of evaluators required to achieve reliability limited widespread use.
Clinical Institute Withdrawal Assessment (CIWA) scale
The CIWA scale for alcohol was developed from the SSA, to enable use at more frequent intervals during the day. This resulted in a 15-item scale, which retained just seven of the 11 SSA items. Inter-rater reliability was demonstrated by comparing assessments made by seven trained nurses on three video cases. Validity was considered by comparison of scores on CIWA rated by nurses, with a 3-point global rating of severity of withdrawal made by a physician at the initial assessment (Shaw, 1981). Further trials have shown a modified CIWA to minimize under- and over-dosing with benzodiazepines in the treatment of alcohol withdrawal. This 15-item scale is known as the CIWA-Ar (Sullivan et al., 1991
). This scale takes only 1 min to complete by assessors with experience in its application, but a further development, an 8-item scale, the CIWA-AD, requires training to ensure adequate reliability (Sellers et al., 1991
). Although the time taken to administer these rating scales may be of limited relevance in a research setting, it is often one factor limiting widespread clinical use.
Metcalfe et al. (1995) produced a further modification of the CIWA scale in the 10-item Windsor Clinic Alcohol Withdrawal Assessment Scale. This scale used fewer objective criteria, in an attempt to improve ease of use. More recently, Wetterling et al. (1997) have derived the Alcohol Withdrawal Scale (AWS) by measuring the internal consistency of items in the CIWA, producing an 11-item scale. The authors claim that the AWS allows rational therapy by predicting those who are at greater risk of complicated withdrawal.
DISCUSSION
With advances in neurophysiological and pharmaceutical knowledge, there are ever more compounds which could be used for treatment of alcohol withdrawal symptoms. Such research will add substantially to current knowledge only if a valid, reliable rating scale is used. Such a rating scale will also enable clinicians to manage alcohol withdrawal more effectively, as they are able to assess severity and progress in a more defined way.
This paper has reviewed published scales which have been used in the management of acute alcohol withdrawal. In most of the articles describing these scales, the primary objective was an assessment of treatment modalities and not the development of a rating scale. As a result, few of the scales have been subjected to a rigorous assessment of validity and reliability and few authors have attempted to justify their reasons for including the individual components of the scale. The weighting of scores often seems to favour symptoms likely to respond significantly to the medication under investigation. Studies have also tended to look at small numbers of subjects, e.g. 24 (Borg and Weinholt, 1980), 23 (Gallimberti et al., 1989
).
Few of the studies sought to establish the clinical usefulness of the rating scales used. An overall estimate of inter-rater reliability was made in only three of the papers (values of 0.94: Shaw, 1981; 0.64: Wetterling et al., 1997; 0.84: Metcalfe et al., 1995). Flygenring et al. (1984) stated that the inter-rater variation was reduced by a series of practice sessions, but they made no attempt to quantify it. Wetterling et al. (1997) developed their scale from the CIWA by choosing only those items with a Cronbach's >0.8. Shaw (1981) also considered internal consistency and quoted a correlation between subscores of 0.92.
Validity was considered for only three of the scales. This has already been discussed with respect to the CIWA and TSA. However, for the CIWA, the gold standard against which scores were compared was simply a global rating of severity, and by a physician at the initial assessment. The paper (Shaw, 1981) did not specify how many physicians performed the assessment, but did imply that there was more than one. There were several reservations about this approach. Although physicians were asked to categorize the withdrawal symptoms as mild, moderate or severe, this was nevertheless a highly subjective judgement. Metcalfe et al. (1995) looked at concurrent validity, by comparing scores obtained using their assessment tool with those obtained using the CIWA. This produced a correlation coefficient of 0.85.
Four papers pointed out the need for training of staff administering the assessments (Gross et al., 1973; Shaw, 1981
; Kraus et al., 1985
; Wetterling et al., 1997
). None of the studies considered the sensitivity or specificity of the rating scales.
The CIWA and its derivations are now probably the most widely used of the withdrawal scales. There may be several reasons for this. First, the CIWA was originally derived from the well-established SSA, so that assessment could be made as frequently as every half an hour. Second, the authors were able to justify why they made changes to the original scale in relation to its relevance in clinical practice. Third, its reliability and validity were determined in the original paper by Shaw (1981). Since its development, the CIWA has been used both in research and in clinical management, increasing the experience of clinicians in its use. It can be easily and quickly administered by trained staff, making it suitable for use by nursing staff in the ward setting and in the community (Bates, 1997).
Most rating scales have been used once and then discarded. If measures are to improve, they will need to be subjected to trials designed solely to assess their reliability and their validity on adequate numbers of subjects. Only if these conditions are met, will it be possible to make proper, scientific comparison of treatment methods, both old and new.
FOOTNOTES
* Author to whom correspondence should be addressed.
REFERENCES
Agricola, R., Mazzarino, M. and Urani, R. (1982) Treatment of alcohol withdrawal syndrome with carbamazepine: a double blind comparison with tiapride. Journal of International Medical Research 10, 160165.[ISI][Medline]
American Psychiatric Association (1994) Diagnostic and Statistical Manual of Mental Disorders, 4th edn. American Psychiatric Association, Washington, DC.
Bates, D. (1997) Another perspective on management of withdrawal. Canadian Family Physician 43, 421424.[ISI][Medline]
Benzer, D. (1990) Quantification of the alcohol withdrawal syndrome in 487 alcoholic patients. Journal of Substance Abuse Treatment 7, 117123.[ISI][Medline]
Bjorkqvist, S. E., Isohanni, M., Makela, R. and Malinen, L. (1976) Ambulant treatment of alcohol withdrawal symptoms with carbamazepine. Acta Psychiatrica Scandinavica 53, 333342.[ISI][Medline]
Borg, V. and Weinholt, T. (1980) A preliminary double blind study of two dopaminergic drugs in the treatment of alcohol withdrawal syndrome. Current Therapeutic Research 27, 170177.[ISI]
Borg, V. and Weinholt, T. (1982) Bromocriptine in the treatment of alcohol withdrawal syndrome. Acta Psychiatrica Scandinavica 65, 101111.[ISI][Medline]
Brunning, J., Mumford, J. P. and Keaney, F. P. (1986) Lofexidine in alcohol withdrawal states. Alcohol and Alcoholism 21, 167170.[ISI][Medline]
Essardas Daryanani, H., Santolaria, F. J., Gonzalez Reimers, E., Jorge, J. A., Batista Lopez, N., Martin Hernandez, F., Martinez Riera, A. and Rodriguez Rodriguez, E. (1994) Alcohol withdrawal syndrome and seizures. Alcohol and Alcoholism 29, 323328.[Abstract]
Flygenring, J., Hansen, J., Holst, B., Petersen, E. and Sorensen, A. (1984) Treatment of alcohol withdrawal in hospitalised patients. Acta Psychiatrica Scandinavica 69, 398408.[ISI][Medline]
Foy, A., March, S. and Drinkwater, R. N. (1988) Use of an objective clinical scale in the assessment and management of alcohol withdrawal in a large general hospital. Alcoholism: Clinical and Experimental Research 12, 360364.[ISI][Medline]
Gallimberti, L., Gentile, N., Cibin, M., Fadda, F., Canton, G., Ferri, M., Ferrara, S. D. and Gessa, G. L. (1989) GHBA for treatment of alcohol withdrawal syndrome. Lancet, September 30, 787789.
Gross, M., Lewis, E. and Nagarajan, M. (1973) An improved quantitative system for assessing the acute alcoholic psychoses and related states (TSA and SSA). In Alcohol Intoxication and Withdrawal: Experimental Studies, Advances in Experimental Medicine and Biology, Gross, M. ed., vol. 35, pp. 356376. Plenum Press, New York.
Kraus, M. L., Gottlieb, L. D., Horwitz, R. I. and Anscher, M. (1985) Randomised clinical trial of patients with alcohol withdrawal. New England Journal of Medicine 313, 905909.[Abstract]
McGrath, S. D. (1975) A controlled study of chlormethiazole and chlordiazepoxide in treatment of acute withdrawal from alcohol. British Journal of Addiction 70(suppl. 1), 8190.[ISI]
Metcalfe, P., Sobers, M. and Dewey, M. (1995) The Windsor Clinic Alcohol Withdrawal Assessment Scale (WCAWAS): Investigation of factors associated with complicated withdrawals. Alcohol and Alcoholism 30, 367372.[Abstract]
Poutanen, P. (1979) Experience with carbamazepine in treatment of alcohol withdrawal. British Journal of Addiction 74, 201 204.[ISI]
Ritola, E. and Malinen, L. (1981) A double blind comparison of carbamazepine and chlormethiazole in treatment of alcohol withdrawal syndrome. Acta Psychiatrica Scandinavica 64, 254259.[ISI][Medline]
Saunders, J. B. (1987) Drug treatment in alcoholism. In Drugs in Psychiatry, Burrow, G. D. and Norman, T. R. eds, vol. 4, pp. 343369. Elsevier, Amsterdam.
Sellers, E. M., Sullivan, J. T., Somer, G. and Sykora, K. (1991) Characterization of DSM III-R criteria for complicated alcohol withdrawal provides an empirical basis for DSM-IV. Archives of General Psychiatry 48, 442447.[Abstract]
Shaw, J. M. (1981) Development of optimal treatment tactics for alcohol withdrawal. I: Assessment and effectiveness of supportive care. Journal of Clinical Psychopharmacology 1, 382387.[ISI][Medline]
Sullivan, J. T., Swift, R. and Lewis, D. C. (1991) Benzodiazepine requirements during alcohol withdrawal syndrome: clinical implications of using a standardized withdrawal scale. Journal of Clinical Psychopharmacology 11, 291295.[ISI][Medline]
Wetterling, T., Kanitz, R., Besters, B., Fischer, D., Zerfass, B., John, U., Spranger, H. and Driessen, M. (1997) A new rating scale for the assessment of the alcohol withdrawal syndrome (AWS Scale). Alcohol and Alcoholism 32, 753760.[Abstract]
Williams, D. and McBride, A. J. (1998) The drug treatment of alcohol withdrawal symptoms: a systematic review. Alcohol and Alcoholism 33, 103115.[Abstract]
Wilson, A. and Vulcano, B. A. (1985) Double blind trial of alprazolam and chlordiazepoxide in management of acute ethanol withdrawal syndrome. Alcoholism: Clinical and Experimental Research 9, 2327.[ISI][Medline]
World Health Organization (1992) International Classification of Diseases, 10th edn. World Health Organization, Geneva.