  |
|
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| Journal of the New Zealand Medical Association,
28-October-2005, Vol 118 No 1224
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Computerised screening for hazardous drinking in primary
care
Kypros Kypri, Shaun Stephenson, John Langley, Martine
Cashell-Smith, John Saunders, Dan Russell
Hazardous consumption of alcohol is a leading contributor to
the global burden of disease and injury.1 Aggregate consumption in many
countries has stabilised or increased in the last 5 years, after 20 years of
steady decline.2,3 There is evidence of a change (in the pattern of drinking)
toward larger quantities per drinking occasion, particularly among young people
(aged 15–24 years).4 In young people, the intoxicating effects of alcohol
account for a greater burden of disease than the chronic effects, given the high
incidence of injury and other acute outcomes (e.g. sexually transmitted
infections) in this age group.1
Primary care services have been identified as settings where
screening and intervention for alcohol problems might be effective. The World
Health Organization (WHO), therefore, developed the Alcohol Use Disorders
Identification Test5: a 10-item questionnaire with questions on alcohol
consumption, symptoms of dependence, and other alcohol-related problems.
With a cut-off score of 8, the test has a sensitivity of 92%
and specificity of 94% for the identification of individuals with hazardous or
harmful drinking in primary care settings.5 It has consistently been found to
outperform other questionnaires and blood markers in the identification of
individuals with alcohol use disorders.6 The AUDIT is cheap to administer7 and
its use provides practitioners with a suitable opportunity to offer advice to
patients with hazardous drinking.8
A significant advance in the treatment of hazardous alcohol
consumption (and thus the prevention of alcohol-related harm) over the last two
decades has been the development and evaluation of screening and brief
intervention (SBI). SBI typically involves opportunistic administration by a GP
or nurse of a brief screening questionnaire such as the AUDIT and (for those who
screen positive) provision of 5–10 minutes of brief advice or a short
session (<30 minutes) of motivational therapy.9
For people identified with severe problems or an established
alcohol dependence, a referral may be made for further assessment and specialist
treatment.10 More than 40 randomised controlled trials have been published on
SBI, most of which have been in primary care settings.
In 2002, Moyer and colleagues11 published a pivotal
meta-analysis which revealed significant (albeit modest) reductions in hazardous
drinking lasting at least 6–12 months among people who were not
specifically seeking treatment. Among those actually seeking treatment, the
effects of SBI were similar to those of more intensive interventions. On the
basis of an extensive review of the effectiveness of SBI in primary care,12 the
US Preventive Services Task Force13 found that SBI of 15 minutes duration is
helpful, and that multi-contact interventions are effective for patients ranging
from 17 to 70 years of age.
In New Zealand, around one in six persons visiting their
general practitioner meets criteria for hazardous or harmful drinking,14 defined
as a score of 8 or higher on the AUDIT.5 In the only published New Zealand study
in which screening of young people (18–29 years) in primary care has been
examined, McMenamin found that 16% of men and 6% of women met criteria for an
alcohol-use disorder.15 He highlighted the importance of screening, noting that
‘without [it], nearly half of those identified would have been
missed’ (p.128).
McMenamin also found that screening rates were low for young
men (59%) relative to young women (83%).15 He suggested measures to overcome
this barrier included the ‘availability of a patient self-administered
computerised lifestyle assessment not requiring supervision by clinical
staff’ (p.128).
The potential utility of computerised screening or
intervention has been identified in other primary care settings.16 For example,
in a survey of a random sample of university students (n=1,564; response rate
82%), we found that computerised screening and brief intervention was the most
acceptable of a range of brief intervention options, including
practitioner-delivered interventions.17 Four out of five hazardous drinkers said
they would use such a service if they thought they had a drinking problem.17
The aims of this study were to:
MethodsSetting—The
data used for this study were collected during the baseline phase of a
randomised controlled trial of a brief intervention for hazardous drinking.
Participants were students aged 17–29 years who attended the University of
Otago Student Health Service in the period 3–25 March 2003. In 2002, the
service conducted 42,000 consultations with over 10,000 individuals,18 making it
the largest provider of primary care for young people in New Zealand (personal
communication, Dr Jim Jerram, Director of Student Health Service, 2002).
Sampling—True
random sampling (i.e. random selection of individuals from a sampling frame of
some description) was not practicable for this study given that eligible
participants were recruited from patients presenting for care. We opted instead
for a selection protocol which would minimise the risk of systematic biases, and
allow for measurement of the potential bias resulting from self selection. Each
week (Monday–Friday inclusive) of the sampling period was broken into 10
sessions: five morning sessions 9am to 12:30pm and five afternoon sessions
1:30pm to 5pm. Based on the ratio of men to women using the service as measured
in a pilot study,19 and to ensure approximately equal numbers of men and women
in the study, we randomly selected 2 of the 10 sessions in each week for
recruitment of men only.
Illustration 1. The University of Otago Student Health
Service waiting room
 Research assistants were trained in the application of
a study protocol, which stipulated that the assistant should invite the next
patient leaving the reception desk (see Illustration 1) to participate in the
study, go through the informed consent procedure, log the participant into a
computer (see Illustration 2) and return to the reception desk to recruit the
next patient. Instances in which a patient appeared too sick or injured or whose
English was not sufficient to participate were recorded, as were refusals to
participate.
Illustration 2. Computers used for screening
 Consent—A
two-stage recruitment procedure was used, whereby patients were first invited to
complete a computerised survey (stage 1: screening). Patients eligible for the
study on the basis of screening were asked for consent to be contacted for
follow-up surveys (stage 2: assessment and intervention). In accordance with
ethical approval, the study was presented to potential participants as a series
of surveys on alcohol use, not as a randomised trial. Randomisation was effected
by computer upon completion of screening. Participants and researchers were
blind to group assignment. This study reports only on data collected during
stage 1: screening.
Measures—Participants
were asked to indicate their gender, age, and ethnicity, using the questions
from the 2001 census.20 Their drinking risk was assessed with the AUDIT. The
consumption questions were based on standard drinks (10g ethanol) which were
defined and depicted in graphics presented on the relevant pages of the web
questionnaire. The whole questionnaire can be viewed at http://ipru.otago.ac.nz/eSBI2003Demo/Index.html
Participants were also asked to indicate how many standard drinks they had
consumed in their heaviest drinking episode in the preceding four weeks.
ResultsOf 1120 patients invited to complete the screening
questionnaire, 1,010 accepted the invitation (90%). Of these, 35 (4%) failed to
complete screening due to being called for their consultation, leaving 975
individuals (538 women and 437 men) with complete AUDIT data—i.e., a
screening rate of 87%. A summary of the AUDIT data of these patients is
presented in Table 1.
Table 1. Distributions of AUDIT item responses by
gender (N=975)
Two-thirds of patients (66%; 60% of women, and 73% of men)
met criteria for hazardous drinking (an AUDIT score of eight or higher).
Relative to women, men reported significantly higher drinking frequencies (item
1), typical occasion quantities (item 2), and binge drinking frequencies (item
3). For both men and women, the modal pattern was to consume six or more drinks
(>60g ethanol) at least once per week. Men also reported a higher frequency
of being unable to stop drinking once started (item 4). Blackouts (item 8) on at
least a monthly basis were reported by 1 in 5 men and 1 in 10 women.
Alcohol-related injuries (item 9) were reported by 1 in 4 men and 1 in 6
women.
In the 4 weeks preceding their visit to Student Health, 71%
of women and 71% of men had exceeded the Alcohol Advisory Council of New
Zealand’s recommended upper limits of no more than four drinks per
occasion for women and no more than six per occasion for men. Among women, 46%
reported at least one episode of more than 8 drinks (>80g ethanol) while 55%
of men reported at least one episode of more than 12 drinks (>120 g ethanol).
A frequency distribution of AUDIT scores is presented in
Figure 1, with indicators of the standard cut-off score and the more liberal
score of 11 recommended by Fleming et al21 in a study of American college
students.
Of 599 service users (311 women, 288 men) who screened
positive for hazardous drinking, 23 (4%) did not consent to web-based follow-up
assessments as part of the study, leaving 576 individuals (300 women, 276 men)
in the intervention trial.
DiscussionThese results show that between half and two-thirds of this
young population drink at hazardous levels, that most (87%) will complete
computerised screening in a primary care setting, and that only 4% of those who
screen positive for hazardous drinking decline follow-up contact. The study
supports the notion that the primary care setting can facilitate access via
computer to a large number of individuals whose drinking is hazardous.
Limitations include the fact that trained research
assistants issued the invitations to complete screening. Although the
intervention is computerised, when put into routine practice some involvement
from receptionists or other staff dedicated to the task would probably be
required to promote its use to patients. The program assessed in this study has
been delivered as a service (i.e. not as part of a research project) at two
student health services at Victoria University Wellington since April 2005.
Early reports show that even with minimal promotion by receptionists, large
numbers of students have utilised the screening and computerised intervention
program, but acceptance rates under these naturalistic conditions have not been
measured.
The rates of hazardous drinking identified in this study
were remarkably high: 60% of women and 73% of men met the commonly used
criterion of >8 on the AUDIT. These prevalence rates can be compared with
those attained in 2002 from a large probability sample of students aged
17–29 who completed a web survey (n=1,564, response rate 82%). In that
sample, 58% (95% CI: 54%–61%) of women and 70% (66%–73%) of men
scored >8 on the AUDIT.22 While the rates are slightly higher in the present
study, the confidence intervals overlap, suggesting that patients using the
student health service are broadly representative of the student population in
terms of their drinking behaviour.
We did not diagnose students in the present study, but it is
likely, given the AUDIT scores attained, that a far higher prevalence of alcohol
use disorders would be found in this tertiary student population than among
patients presenting to their general practitioners.15 Direct comparisons of
students and non-students of the same age in the general population reveal very
large differences in the prevalence of hazardous drinking.23 The reasons for
this are unclear, but probably relate to the presence in the university
environment of ‘...high concentrations of licensed premises, events that
have a primary focus on drinking, intense advertising, promotion, and aggressive
pricing by the liquor industry, institutional policies that do not adequately
discourage drunkenness, and inadequate enforcement of the intoxication
provisions of liquor legislation’ (pp. 713-714).23
In the last 15 years, New Zealand has drastically altered
its laws with respect to alcohol. There has been a shift away from supply-side
policies, in which the primary mechanism is to restrict the availability of
alcohol to the consumer.24 Examples include the introduction of wine (1989) and
beer (1999) in supermarkets, allowing a wider range of retail outlets to sell
alcohol (1989) and the reduction of the minimum purchase age from 20 to 18 years
(1999), which increased alcohol-related harm among young people.25 This move
toward increased availability at a time when the burden of disease and injury
attributable to alcohol is increasing,1 highlights the need for greater efforts
by public health advocates to influence policy but also for interventions to
reduce heavy consumers’ demand for alcohol. To make an impact at a
population level, such interventions would have to be inexpensive and
deliverable to many. Opportunistic screening for hazardous drinking followed by
brief intervention in the primary care setting meets both of these
criteria.
Despite evidence from at least 36 randomised controlled
trials from several countries,11 screening and brief intervention is not yet a
routine aspect of primary care in any country. One obstacle to its widespread
implementation is the lack of time and remuneration available for preventive
medicine. Another is the view held by general practitioners that their patients
would not accept their advice to drink less if an alcohol-related condition was
not the presenting problem.26
In recognition of these circumstances, and the reported
willingness of students to participate in a computerised intervention,17 we
conducted a pilot randomised controlled trial at the Student Health Service at
the University of Otago.19 Of 112 students who screened positive for hazardous
drinking, 104 agreed to be contacted for follow-up assessments, and were
randomised to a computerised brief intervention delivered in the reception area
(n=51) or a leaflet-only control group (n=53).
Follow-up assessments were conducted 6 weeks and 6 months
post-intervention and were completed by 80% and 90% of participants
respectively. Results showed reductions of episodic heavy drinking and
alcohol-related problems in the intervention group relative to controls of
20–30% over 6 months.19 It was concluded that the results warranted a
larger, more comprehensive trial.
Screening is only of benefit to patients if there is an
viable intervention to offer them. The pilot research described above suggests
that there is a viable intervention and that early signs regarding its efficacy
are positive.
On the basis of the screening described in the present
study, we conducted a four-arm randomised controlled trial in which students
with hazardous drinking were assigned to one of four conditions:
These interventions represent the least
to the most that we judge young people could accept in a primary care setting.
Figure 1. AUDIT scores of Student Health Service users
(N=975)
 Author information:
Kypros Kypri, Senior Lecturer;1,2 Shaun C R Stephenson, Biostatistician;2 John D
Langley, Director/Professor;2 Martine L
Cashell-Smith, Assistant
Research Fellow;2 John B Saunders, Professor;3 Dan Russell, Data Manager2
1School of Medical Practice and Population Health,
University of Newcastle, Adamstown Heights, NSW, Australia
2Injury Prevention Research Unit, Department of Preventive
and Social Medicine, University of Otago, Dunedin
3Centre for Drug and Alcohol Studies, Department of
Psychiatry, School of Medicine, University of Queensland, Brisbane, QLD,
Australia
Acknowledgements:
This research was funded by the Alcohol Advisory Council of New Zealand and the
Health Research Council of New Zealand. We are also grateful to Dr Jim Jerram
and staff of the University of Otago Student Health Service for facilitating the
study.
Correspondence: Dr
Kypros Kypri, School of Medical Practice and Population Health, University of
Newcastle, 2 Edison St, Adamstown Heights, NSW, Australia. Fax: +61 2 49246208;
email: kypri@tpg.com.au
References:
|
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| | Current
issue | Search journal |
Archived issues | Classifieds
| Hotline (free ads) Subscribe | Contribute | Advertise | Contact Us | Copyright | Other Journals |
|