OUP user menu


Robert Patton, Catriona Hilton, Michael J. Crawford, Robin Touquet
DOI: http://dx.doi.org/10.1093/alcalc/agh049 266-268 First published online: 13 April 2004


The Paddington Alcohol Test, designed to screen for alcohol related problems amongst those attending Accident and Emergency Departments, is presented in a slightly modified form. It concords fairly well with the Alcohol Use Disorders Identification Test (AUDIT), but can be administered in about one fifth of the time taken to administer AUDIT. Its scoring of units is rapid and specific to the UK. PAT is recommended for use in UK Accident and Emergency Departments.

(Received 30 July 2003; first review notified 2 September 2003; in revised form 30 January 2004; accepted 1 February 2004)

Recent reports on the prevalence of alcohol misuse among patients attending British Accident and Emergency Departments (AEDs) estimate that up to 40% present with alcohol related problems, and between the hours of midnight and 5.00 am this proportion may rise to over 70% (Strategy Unit, 2003). The Alcohol Use Disorders Identification Test (AUDIT) (Babor et al., 1989) may be considered to be too time-consuming to use in a busy medical setting such as the AED; thus shorter screening instruments such as the Paddington Alcohol Test (Smith et al., 1996) have been specifically developed for use in this setting. Within the confines of an AED it is not practicable to screen every patient, therefore the PAT is targeted at those patients who present with conditions associated with alcohol misuse (Huntley et al., 2001) to counter clinical inertia; failure of health care providers to initiate or intensify therapy when indicated (Phillips et al., 2001).

Since it was first published in 1996 the PAT has been revised to increase its reliability and validity. Huntley et al. (2001) amended the PAT by refining the list of 21 presenting conditions likely to be associated with hazardous drinking to develop the ‘top ten’ conditions that accounted for 77% of hazardous drinkers. An initial screening question addressing whether or not a patient ever consumes alcohol (Patton and Touquet, 2002) has also been added. For those patients identified as hazardous drinkers the PAT pro forma now encourages the clinician to provide health consequences feedback, specifically ‘We gently advise you this drinking is harming your health’. A retrospective analysis of the impact of this feedback demonstrated a 23% increase in the proportion of patients willing to accept brief intervention (Patton et al., 2003).

The latest version of the PAT (Fig. 1) amends item 3 (frequency of heavy drinking episode) to include a monthly cut off. In a recent study (unpublished) 468 patients who presented to our local AED completed the PAT and the AUDIT. Levels of concordance are presented in Table 1. The introduction of the monthly cut-off represents considerable improvement to figures reported by Hodgson et al. (2003), which were based on the 1996 version of the instrument.

Fig. 1.

The Paddington alcohol test (2003)

View this table:
Table 1.

Concordance of PAT (2003) compared to AUDIT

N (%)PositiveNegativeTotal
    Positive143 (97.3)38 (11.8)181 (38.7)
    Negative4 (2.7)283 (88.2)287 (61.3)
Total147 (100.0)321 (100.0)468 (100.0)

The time taken to complete the PAT and the AUDIT was recorded for a sub-sample of 47 participants. It was found to be 20 seconds for the PAT (SD = 9.53) and 1 minute 13 seconds for the AUDIT (SD = 27.6). Based on our experiences of a randomized controlled trial of screening and brief intervention based in our local AED (in preparation), we estimate that 5750 patients may be screened over one year. A trainee doctor (Senior House Officer) costs £23 per hour, so screening with PAT would add £735 to annual department costs, compared to £2682 for AUDIT.

The PAT features a table of commonly encountered beverages coded in British units. This offers an advantage over other questionnaires such as the AUDIT that require the participant to convert traditional measures into standard drinks (US coding); any misinterpretation at this point could lead to a significant underestimation of alcohol consumed. The PAT also allows for the different relative strengths of certain products, thus differentiating between a patient who may consume 2 pints (i.e. 4 ‘drinks’) of normal strength beer (4 units) and the same amount of ‘strong’ lager (10 units). By providing the best indication of total units, the PAT is well placed for UK users to provide an indication of hazardous drinking patterns.

Whilst we acknowledge that the AUDIT provides a reliable assessment of hazardous drinking, we feel that the time saved using the PAT may be vital when encouraging staff to screen patients within busy medical settings such as the AED.

PAT has been recommended by SIGN for use in AEDs (SIGN, 2003).

Lengthening the timeframe over which alcohol consumption is assessed to one month allows the detection of ‘binge’ drinkers who may not otherwise be identified. The increased sensitivity, the minimal time taken to complete the questions and the potential impact of drawing the patients' attention to the relationship between their alcohol consumption and attendance to the AED, all commend the use of the PAT as the instrument of choice for detecting hazardous drinkers within busy medical settings.


We are grateful for financial support that we received from the Alcohol Education and Research Council and for the support of patients and staff at St Mary's AED.


View Abstract