Brief interventions for alcohol problems
While
for some of us the problem with alcohol is not getting enough, for a
distressing number of people the problem is drinking too much. There are, as
always, gradations, but alcohol is a problem for a significant minority of
people. Some recognise the problem and want to control their intake or stop.
Others do not, or will not, recognise it as a problem.
One
way of trying to deal with alcohol problems is that of a brief intervention,
often by a primary care physician. Those brief interventions have all sorts of
designations, and what is a brief intervention for one is an extended
intervention for another. Then there is the difference between opportunistic
interventions (often in primary care aimed at people with less severe
problems), and those interventions that may be more targeted (often in hospital
in people with more severe problems).
Two
reviews [1, 2] give hope that brief interventions can be beneficial in some
settings. The first [1] also has an excellent review of the issues of the
interventions themselves and the populations in which they are used.
Alcohol problems
Three
sources were used for studies in the first review [1], a specialised database,
literature searches, and reviews and primary report citations. Brief
intervention was defined as an intervention of no more than four sessions,
though the amount of contact time is frequently difficult to estimate from
reports. Studies were classified according to the type of comparison (brief
intervention versus control, or a more extended intervention), and the type of
patient population (treatment-seeking versus non-treatment-seeking).
The
outcomes examined were alcohol consumption, and a composite measure of multiple
different types of outcome, 12 in all, that included alcohol consumption,
abstinence, frequency of intoxication, global ratings, and problems in life
resulting from drinking.
Results
Though
56 investigations were included, there was no definition of types of study
architecture used, nor was there any indication of which (if any) were
randomised trials.
Alcohol consumption
In
34 studies comparing brief interventions to control conditions in
non-treatment-seeking populations, brief interventions had a significant though
small effect on alcohol consumption in trials with a follow up from less than
three months to up to 12 months.
In
20 studies comparing brief interventions to extended treatment conditions in
treatment-seeking populations, there was no difference found.
Composite outcomes
In
34 studies comparing brief interventions to control conditions in
non-treatment-seeking populations, brief interventions had a significant though
small effect on composite outcomes in trials with a follow up from less than
three months to up to 12 months.
In
20 studies comparing brief interventions to extended treatment conditions in
treatment-seeking populations, there was no difference found.
Mortality
Whether
brief interventions for problem drinking had any effect on mortality was
examined in a second review [2]. Here studies included had to be randomised,
compare a brief intervention to a no-intervention control in heavy drinkers,
and have a long follow up with verified death rates. Studies were sought
through consulting previous systematic reviews and meta-analyses, with
additional searches of a number of electronic databases.
Results
Though
there were 32 studies with 7,500 subjects meeting a broader set of inclusion
criteria, only four with 1,540 subjects had verified death rates. Follow up in
these four studies was between one and 10 years.
Death
rates (Figure 1) were lower in the groups with a brief intervention compared
with no intervention. Overall the death rate with brief intervention was 2.1%,
compared with 3.3% with control. In terms of life years, there were 3 deaths
per 1000 life years with brief intervention, compared with seven deaths per
1000 life years with control.
Figure 1: Brief interventions about alcohol, and mortality
The
reduction was statistically significant with a relative risk of 0.5 (95% CI 0.3
to 0.96). For every 80-100 heavy drinkers given a brief intervention to stop or
cut down their drinking, one fewer would die over the succeeding 1-10 years.
Comment
This
is tricky territory for a number of reasons. First there is the intervention.
What is it? A brief intervention defined for these reviews was up to four
sessions of about 15 minutes or so, aimed at helping heavy drinkers stop or cut
down. What is in a brief intervention? A variable feast, this, with a number of
possible components. What is the outcome? Anything from reducing consumption to
death. And whom are we targeting? There's the person who knows he or she
has a problem, and wants help, and those not acknowledging the problem.
The
problem we have with the first review is that we do not know the architecture
or quality of the studies. Any result must therefore be taken with a pinch of
salt, but, with that proviso, the results were fairly consistent. A brief
intervention is better than nothing in those not seeking treatment. A brief
intervention is probably as good as longer, more intensive, interventions in
those who are seeking treatment.
What
buttresses these results is the knowledge that in randomised trials with good
ascertainment of outcomes, death rates were consistently lower with a brief
intervention than with no intervention. Death is a hard outcome, and a
reasonable proxy for other important outcomes.
While
a brief intervention is not very effective, with an NNT of about 100, if an
intervention can be constructed that is easy to do, it can make a difference,
and may be well worth doing. Estimates of problem drinking suggest that
significant portions of the population have alcohol problems.
Reference:
- A Moyer et al. Brief interventions for alcohol problems: a meta-analytic review of controlled investigations in treatment-seeking and non-treatment-seeking populations. Addiction 2002 97: 279-292.
- P Cuijpers et al. The effects on mortality of brief interventions for problem drinking: a meta-analysis. Addiction 2004 99: 839-845.