Electronic Letters to:

PAPERS:
Annie McCloud, Ben Barnaby, Nicola Omu, Colin Drummond, and Andy Aboud
Relationship between alcohol use disorders and suicidality in a psychiatric population: In-patient prevalence study
The British Journal of Psychiatry 2004; 184: 439-445 [Abstract] [Full text] [PDF]
*eLetters: Submit a response to this article

Electronic letters published:

[Read eLetter] The AUDIT and Suicidal Risk
Muzaffar Husain   (26 May 2004)

The AUDIT and Suicidal Risk 26 May 2004
Top
Muzaffar Husain,
Senior House Officer (Psychiatry)
None

Send letter to journal:
Re: The AUDIT and Suicidal Risk

muzhusain{at}hotmail.com Muzaffar Husain

I read with interest McCloud et al’s (2004) paper considering the association between alcohol use disorders and suicidality in a psychiatric population(1). Whilst the AUDIT (Alcohol Use Disorders Identification Test) has indeed been established as a valuable screening tool in determining the prevalence of alcohol misuse in psychiatric inpatients(2), the extension of this association to actually predict or to identify the risks of suicidality runs into some trouble.

Firstly, from your data-set it seems that the OR (Odds Ratio) for the association of AUDIT scores of 8 or more with suicidality, albeit high (2.98), has a relatively wide 95% confidence interval (1.66-5.33). This phenomenon is even more so for the OR associating AUDIT scores of 16 or more with suicidality (95% CI: 3.23-20.3). Such a result would generally make a false positive finding in this study very likely.

One might argue that the above quoted figures still point toward a statistical difference between the two groups in both cases. However, if one were to consider that the outcome of suicidality is quite broadly defined in the study, and that in clinical practice one would concentrate on a much smaller number of high risk patients (i.e. patients with previous self-harm or existing plans), due to smaller numbers the results would lose statistical significance.

Secondly, we are not told directly of the association between suicidality and AUDIT scores of less than 8. A high OR in this group of patients could point toward any ‘background’ suicidality that may have confounded the overall results.

Finally, if the AUDIT were to be routinely used in secondary care for psychiatric patients, perhaps the best place to test it for associations with suicidality would be parasuicide assessments in Casualty departments and Medical Admission Units rather than psychiatry wards. This is especially true when decisions have to be made about addressing underlying substance misuse issues during later admissions to psychiatry wards.

The AUDIT may still have a vital role to play in identifying alcohol misuse in parasuicide assessments given the ease with which it can be quickly administered. In this case it would be a possible alternative to the CAGE questionnaire and would help clinicians broaden both diagnostic and therapeutic options. But the use of the same tool in the other direction, i.e. identifying suicidal risk by means of the AUDIT, remains to be proven.

References

1 McCloud, A., Barnaby, B., Omu, N.,Drummond, C., and Aboud, A., (2004) Relationship between alcohol use disorders and suicidality in a psychiatric population. In-patient prevalence study. British Journal of Psychiatry,184, 439-445. 2 Barnaby, B.,Drummond, C., McCloud, A. ,et al (2003) Substance misuse in psychiatric patients: comparison of a screening questionnaire survey with case notes. BMJ,327, 783-784.