Addiction (1998) 93(4), 589± 594

RESEARCH REPORT

Problems encountered with opportunisticscreening for alcohol-related problems inpatients attending an Accident and Emergencydepartment

JEAN PETERS, CHARLIE BROOKER, CHRIS McCABE & NEILSHORT1

School of Health and Related Research (ScHARR), University of Shef® eld, Shef® eld and1Chester® eld and North Derbyshire Royal Hospital NHS Trust, Chester® eld, UK.

Abstract

Aim s. To assess the value of opportunistic screening in Acciden t and Emergency (A&E) for patients with

alcohol-related problems and provision of an intervention. Design. Screening of A&E attendees for the

purpose of recruitment to a randomized trial of a counsell ing intervention. Setting. A General Hospital

A&E department. Participants. All 17 000 adult A& E attendees , during a 6-month period and all

nursing staff working within the department. M easurem ents. Patients ’ self-reported alcohol consumption,

responses to the CAGE questionnaire (four question s designed to identify problem drinking) and proportions

offered, and taking up offer of help. Findings. Only 4663 (28%) adult attendee s at A&E were actually

screened and of these 2% declined and 25% were judged unable to answer. Of the rest, 86% drank alcohol,

with 22% drinking in excess of current guidelines or with two or more positive answers to CAGE. Only 41%

(264) of those drinking to excess were offered help and 88% of these declined it. This left 13 patients to be

included in the trial. Conclusion. There is a signi ® cant need for an effective intervention in this area but

considerable barriers exist to testing the ef ® cacy of potential screening strategies and interventions.

Introduction

Excessive alcohol intake can lead to many public

health problems including premature mortality,1

physical and psychological ill-health, marriage

and family breakdown,2 child neglect and abuse3

public disorder and violence, criminal activity,

road traf® c accidents, accidents at work or in the

home, employment problems, ® re and drowning.

Alcohol is also associated with a higher use of

health services; by attendees of Accident and

Emergency (A&E) departments2,4 for acute

admissions to hospital5,6 and in consultations

with general practitioners.2

Drinking clearly leads to many of the problems

that bring patients to an A&E department. In

one study, 40% of 702 patients at a large teach-

ing hospital screened positive in a breath test.4 In

an inner city district general hospital 27% of 104

consecutive, new, unselected, emergency admis-

sions were attributed to alcohol consumption.6

In a third study of 2626 patients attending emer-

gency rooms in four North American hospitals

Correspondence to: Dr Jean Peters, Public Health Section, ScHARR, Regent Court, 30 Regent Street, Shef® eld,S1 4DA, UK. Tel: 0114 2220680 ; fax: 0114 2724095 .

Submitted 19th February 1997 ; initial review completed 4th July 1997; ® nal version accepted 17th October 1997.

0965 ± 2140 /98/0400598 ± 06 $9.50 Ó Society for the Study of Addiction to Alcohol and Other Drugs

Carfax Publishing Limited

590 Jean Peters et al.

the prevalence of problem drinkers was 19%

compared with 11% in a general population.7

Brief interventions have been reported to be

effective in reducing problem drinking8,9

although a recent commentary suggests that the

degree of effect, to some extent, may have been

misinterpreted.10

Twenty per cent (11 million) of the UK popu-

lation attend A&E departments in the United

Kingdom every year for a wide variety of prob-

lems, including acute intoxication and chronic

problems as a result of alcohol. A&E therefore

provides an opportunity to identify people with

alcohol-related problems, to assess their physical

and mental health needs, and offer help.11 How-

ever, in only one study has this been

attempted.12 This paper reports the experience

of attempting to conduct a randomized con-

trolled trial of brief intervention for alcohol prob-

lems which used opportunistic screening in a

District General Hospital A&E department to

identify subjects for recruitment to such a study.

M ethods

All patients, including stretcher patients, aged 18

and above who presented at an A&E department

of a General Hospital in Trent Region for medi-

cal attention between October 1995 and March

1996 were eligible for screening unless they were

judged medically un® t to be screened.

The screening tool was a brief questionnaire.

First it asked if the person drank alcohol or not.

Those answering `yes’ were then asked:

· the number of days each week, on average, on

which they normally had a drink;

· an estimate of the average quantity drunk each

day; and

· the four `yes/no’ questions of the CAGE ques-

tionnaire (to identify problem drinking):13,14

(i) Do you feel you should cut down on your

drinking?

(ii) Does anyone annoy you or get on your

nerves by telling you to cut down your

drinking?

(iii) Do you yourself feel bad or guilty about

your drinking?

(iv) Do you have a drink ® rst thing in the

morning, to steady your nerves or get rid

of a hangover?

All staff in the department were supplied with

ready reckoner cards giving the number of

alcohol units corresponding to named drinks,

supplied as either pub or home measures. The

screening questionnaire was administered by the

triage nurse as part of a routine triage assess-

ment, adding a maximum of 2 minutes to this

procedure.

A positive screen was de® ned as either an

alcohol consumption above governmentÐ rec-

ommended guidelines or two or more positive

answers to the four CAGE questions. Initially

the government guidelines were set at 14 and 21

units for women and men, respectively15, but

changed to 21 and 28 units in December 1996

following publication of the government’ s report

on Sensible Drinking.16 The objective of the

screen was to recruit people with an alcohol

problem to the brief intervention study. It was

felt that individuals who drank more than the old

recommendations but less than the revised rec-

ommendations would be unlikely to accept that

they could bene® t from assistance. Signi® cant

time and effort would thus be expended with

little likelihood of recruiting any one in this

group to this study. Therefore the new guidelines

were used as the screening criteria for recruit-

ment after their publication in December 1996.

Patients who screened negative were informed

immediately of their result. For all those who

screened positive, their questionnaire was

attached to their notes and, during their treat-

ment, in the privacy of the treatment cubicle, the

patient’ s named nurse fed back the results and

offered the patient a specialist outpatient clinic

appointment. All patients who accepted an

appointment were randomized to receive either a

brief intervention (this was a maximum contact

time of 3 hours counsell ing provided in one or more

sessions) with a clinical nurse specialist for alcohol

plus a self-help booklet or the self-help booklet

only.

Prior to commencement of the study all nurs-

ing staff in the A&E department received a half-

day training session covering the following

issues:

· using the questionnaire,

· providing feedback of results to patients,

· recruiting patients to the study, and

· general issues associated with the handling of

alcohol problem patients.

Support for the study was obtained from senior

A& E attenders with alcohol problems 591

Figure 1. Flow diagram of screening outcome of all attendees at an Accident and Emergency department of a general hospital

over a 6-month period, 1995/6.

managers within the NHS Trust and within the

A&E department itself and ethical permission

was obtained from the relevant body. At the end

of the study semi-structured interviews were

conducted with a random selection of one-third

of all nursing staff within the A&E department

(strati® ed by grade). The progress of the screen-

ing and recruitment was also discussed with

departmental managers.

Results

Response rates

Approximately 17 000 adult patients attended

the A&E department during the 6-month

recruitment period but screening questionnaires

were allotted to only 28% (4663). Of these, 1162

(24.9%) were judged `unable’ to be screened and

their questionnaires were not completed

(although only three women and ® ve men were

described as `too drunk’ ). A few patients (103,

2%) refused to answer the screening questions

(Fig. 1).

`Refusers’ (mean 40.9 (SD 18.5) years) and

`unables’ (45.3 (20.0)) were both, on average,

slightly older than non-refusers (38.9 (15.2))

with statistically signi® cant differences between

`unables’ and `non-refusers’ (t 5 9.68, p , 0.001)

(information on age was missing for 237 cases).

Slightly more men refused to answer the ques-

tions, 3% (70) compared with 2% (28) of

women but, conversely, proportionately more

women, 29% (483), were judged unable to

answer compared with 22% of men (620)

( c 2 5 30.52, df 5 2, p , 0.001). Approximately

twoÐ thirds (2851) of screened patients were

men (information on sex and ® rst name was

missing in 144 cases). Breakdown of patients for

whom screening questionnaires were allotted by

age and sex are presented in Table 1.

Drinkers of alcohol

For the 3381 patients with completed question-

naires, 86% (2917) said they drank alcohol, 90%

of men and 80% of women ( c 2 5 68.2, df 5 1,

p , 0.001). Drinkers were younger, mean age

37.8 (SD 14.2) years, compared with non-

drinkers 45.9 (18.9) years (t 5 8.7, p , 0.001).

Twenty-two per cent (645) of all identi® ed

drinkers drank either in excess of the current

guidelines or had two or more `yes’ answers to

the four CAGE questions (Fig. 1) with twice as

many men (29%) as women (14%), ( c 2 5 126.3,

df 5 1, p , 0.001). In the drinkers screening

positive, alcohol was consumed on a similar

number of days by both sexes (t 5 0.72, p 5 0.47)

but with men consuming an average of 44.7

units/week (SD 35.2) and women, 29.2 units/

week (18.4) (t 5 5.23, p , 0.001).

Feedback and recruitment

Only 264 of those who screened positive (41%)

were given feedback on their screened status and

the majority of these (88%, 233) subsequently

refused an appointment for help (Fig. 1). Of the

31 patients who accepted an appointment three

were admitted as inpatients, 15 failed to show,

and 13 patients were randomized to receive

either a brief intervention and self-help booklet

(7) or booklet only (6). This is 2% of all who

screened positive, 5% of all given feedback and

42% of all who had accepted an appointment.

592 Jean Peters et al.

Table 1. Patient screening status (%) by gender and age

Screened positiveusing *current

All cases Drinkers guidelines

Gender (n 5 4519) (n 5 2850) (n 5 627)Men 63 68 86Women 37 32 14

(n 5 4429) (n 5 2812) (n 5 616)Age(years)

. 20 5 5 620± 29 28 31 3630± 39 25 27 2640± 49 17 18 1750± 59 11 11 1060± 69 7 5 470± 79 5 3 1

80 1 3 1 , 1Total 100% 100% 100%

*Current guidelines: these refer to the two sets of guidelines used duringthe recruitment period.

Interviews with nursing staff

Seventeen staff, comprising four male and 13

female, 14 trained nurses and three auxiliaries,

with a range of nursing experience in an A&E

department from less than 5 to 15 years were

interviewed. Two key points emerged. The staff

did not see the link between the study and the

potential to improve patient care in A&E. Sec-

ondly, there was no sense of ownership of the

study among the nursing staff with an over-

whelming feeling that they were being required

to do additional work for someone else’ s study.

Discussion

In spite of the relatively small number of patients

screened, the percentage who screened positive

is similar to that reported elsewhere.7 The idea of

opportunistic screening in a service setting

appeared acceptable to patients in that only a

very small percentage refused to answer the

questions. However, recognition of potential

health problems associated with alcohol intake in

those who screened positive was less acceptable,

with a large percentage refusing help. Thus,

while A&E may provide opportunistic access to a

population containing proportionately more peo-

ple with alcohol-related problems than in the

general population, this is of little use unless

screening is followed by uptake of a health

intervention. One possible bene® t, although

unprovable and unquanti® able in this study, is

that `just asking’ can make a difference and

increase intentions to take action.17 Thus one

barrier to the success of this study was imposed

by the patients although, even with the low

response rates, recruitment targets could ulti-

mately have been achieved, albeit over a longer

time frame, if all patients had been screened as

proposed.

The interviews with nursing staff and discus-

sions with management identi® ed a number of

factors which, combined with staff perception of

the study, might help to explain the general

failure of nursing staff to screen and feed back

results to those patients who had screened posi-

tive. First, the A&E department was undergoing

a major refurbishment which was predicted to be

completed before the study started but contin-

ued throughout the whole of the recruitment

period with unavoidable disruptions to the ser-

vice. In addition, one A&E consultant post was

vacant throughout the entire recruitment period

and a second became vacant during this time.

The Clinical Nurse Manager for the department

left during the recruitment period and there was

a general freeze on any nursing staff replace-

ment. Finally, in the middle of the recruitment

period over the Christmas period, there were

several unexpected fatalities within the A&E

A& E attenders with alcohol problems 593

department. All these factors are likely to have

had signi® cant implications for management

workload and morale within the department and

therefore on the importance that was attached to

the study. Equally, there is little that the study

could have done to prevent these events and

several of these factors were already current

before the study commenced.

It might be argued that the half-day training

given to the nursing staff proved insuf® cient and

that the design of the study was therefore at

fault. However, the original study design had

included a full day of training for all nursing staff

in the A&E department. It was the service con-

straints on these staff which meant that only a

half-day of training could be provided.

Irrespective of the service issues identi® ed

above, the attitudes of the nursing staff to this

study are probably the key factor to its lack of

success. The majority of the nursing staff inter-

viewed, while claiming a holistic approach to

health care, did not equate this with the incorpo-

ration of screening and health promotion activity

into the routine triage examination. The research

was seen in terms of the imposition of someone

else’ s work. Thus, the second, and greater, bar-

rier to the success of this study was the nurses’

perception of the research study.

If the recommendations for nurses to become

more involved with research18,19 are to be imple-

mented successfully, service implications must

be acknowledged and the commitment of both

the managers and the nursing staff involved must

be secured. Given the changes in the funding of

research within the NHS, subsequent to the

Culyer reforms there is likely to be an increase in

the proportion of service staff explicitly involved

in research.20 It is important therefore that the

issue of research is addressed as part of nurse

education and training.

The aims of this study, which were to identify,

and offer provision of immediate help to,

patients with alcohol-related problems on their

attendance in an A&E department had not been

tried before. Green et al.’ s study12 had some

similarities but although those authors reported

greater subject numbers, the opportunistic

screening employed in this study identi® ed a

larger proportion of patients, even though only

13 were treated. In spite of the limitations of this

study we believe that the results con® rm the

extent of alcohol problems within a population

presenting to an A&E department. The paper

also demonstrates that routine screening for

alcohol problems is feasible within such a popu-

lation if nursing staff perceive that this, and the

essential follow-up promotion of help, is part of

their nursing role.

Acknowledgements

The authors wish to acknowledge the help given

by Trust senior management, the nursing staff in

the Accident and Emergency department, with-

out whom the questionnaire could not have been

administered, and the patients for their co-oper-

ation. The study was funded by a research grant

from Trent Regional Health Authority, Research

and Development.

References1. GRONBAEK, M., DEIS, A., SORENSE N, T. I. A.,

BECKER , U., SCHNO HR, P. & JENSEN , G. (1995)Mortality associated with moderate intakes ofwine, beer, or spirits, British Medical Journal, 310,1165± 1169.

2. BUCHAN , I. C., BUCKLEY , E. G., DEACO N, G. L. S.,IRVINE , R. & RYAN , M. P. (1981) Problemdrinkers and their problems, Journal of the Royal

College of General Practice, 31, 151± 153.3. BIJUR, P. E., KURZON , M., OVERPEC K, M. D. &

SCHEIDT , P. C. (1992) Parental alcohol use, prob-lem drinking, and children’ s injuries, Journal of the

American Medica l Association, 267, 166± 171.4. HO LT, S., STEW ART , I. C., D IXON, J. M., ELTO N, R.

A., TAYLOR , T. V. & LITTLE , K. (1980) Alcoholand the emergency service patient, British Medica l

Journal, 281, 638± 640.5. CHICK, J., DUFFY, J. C., LLOYD , G. G. & RITSON ,

B. (1986 ) Medical admissions in men: the riskamong drinkers, Lancet, ii, 1380 ± 1383 .

6. LOCKHART, S. P., CARTER , Y. H., STRAFFEN, A.M., PANG, K. K., M CLOUGHLIN, J. & BARON , J. H.(1986) Detecting alcohol consumption as a causeof emergency general medical admissions, Journal

of the Royal Society of Medicine, 79, 132 ± 136.7. WEISNER , C. & SCHMIDT, L. (1993 ) Alcohol and

drug problems among diverse health and socialservice populations, American Journal of Public

Health , 83, 824± 829.8. NUFFIELD INSTITUTE FOR HEALTH (1993 ) Brief

interventions and alcohol use, Effective Health

Care, 7, 1± 14.9. BIEN, T. H., M ILLER, W. R. & TONIG AN , J. S.

(1993) Brief intervention for alcohol problems: areview, Addiction, 88, 315± 336.

10. HEATHER , N. (1995 ) Interpreting the evidence onbrief interventions for excessive drinkers: the needfor caution, Alcohol and Alcoholism, 30, 287± 296.

11. PETERS, T. J. (1994 ) A&E medicine and the healthof the nation: violence, Journal of the Royal College

of Physicians, 28, 353.12. GREEN , M., SETC HELL , J., HARNES , P., STIFF, G.,

594 Jean Peters et al.

TOUQUET, R. & PREIST, R. (1993 ) Management ofalcohol abusing patients in accident and emer-gency departments, Journal of the Royal Society of

Medicine, 86, 393 ± 395.13. MAYFIELD, D., M CLEO D, G. & HALL, P. (1974)

The CAGE questionnaire: validation of a newalcoholism screening instrument, American Journal

of Psychiatry, 131, 1121 ± 1123.14. EWING , J. A. (1984 ) Detecting alcoholism. The

CAGE questionnaire, Journal of the American

Medica l Association, 252, 1905 ± 1907.15. DEPARTMENT OF HEALTH (1993) The Health of the

Nation. A new strategy for health in England, pp. 1±126 (London, HMSO).

16. DEPARTM ENT O F HEALTH (1995) Sensible drinking.

The report of the Inter-Departmental Working Group,pp. 1± 103 (London, Department of Health).

17. SKINNER , H. A., ALLEN , B. A., MCINTOSH , M. C.& PALM ER, W. H. (1985) Lifestyle assessment: justasking makes a difference, British Medica l Journal,290, 214.

18. BRIGGS, A. (1972) A Report of the Committee on

Nursing, pp. 1± 327 (London, HMSO).19. DEPARTM ENT OF HEALTH , NURSING D IVISION

(1989) A Strategy for Nursing, pp. 1± 44 (London,Department of Health).

20. CULYER , A. (1994) Supporting Research and Devel-

opment in the NHS (London, HMSO).