Send email for updates


About updates
Hot topic

Below is one of our selection of Hot Topics, important issues which sometimes generate heated debate over the facts or their interpretation. Click the GO button or the Blue title to trigger a customised search for relevant Findings logo documents. Links to other documents. Hover over for notes. Click to highlight passage referred to. Unfold extra text Unfold supplementary text

Title and link for copying Comment/query to editor Other hot topics

Computerising therapy and advice increases access – but is effectiveness sacrificed?

The prospect that computerised interventions could make substance-related help more available more cheaply to more people, and the increasing feasibility of achieving this as web access expands and becomes more convenient, has driven studies to test whether such approaches would retain effectiveness. Without that, no amount of increased access will be worthwhile. In an emotional and difficult endeavour, the idea that an automated response driven by a computer could help addicted patients overcome often desperate situations seems not just unbelievable, but somehow ‘wrong’ – a denial of the humanity due to them. That reaction is in itself a reason why they might not work, because such therapies would fail to meet a basic criterion for effective psychosocial treatment – that to the patient, it looks like ‘treatment’ – what you do in that culture to get better.

But as web services permeate more of life including social and health-related activities, they too may take their place among culturally accepted routes to overcoming unhealthy substance use. Self-help web sites particularly attract problem drinkers who retain a stake in mainstream society in the form of relationships, jobs, families, and a reputation to lose. These populations have more of the ‘recovery capital’ needed to elevate themselves out of a substance use problem without therapist intervention. Typically the computer-driven programmes are based on cognitive-behavioural principles and techniques.

Stepping down, and particularly in relation to alcohol, are non-dependent but clearly excessive substance users and those potentially facing harm, but not yet. Typically these substance users are offered brief advice after they have been identified by clinical signs or short screening questionnaires. A common core component is giving the user feedback from their screening test or assessment showing how far their drinking departs from the norm. Here a computerised response has a clearer role, because these users are unlikely to seek face-to-face help, their problems need not be so complex and deeply ingrained that only in-person support could be expected to work and would be considered acceptable, and an inexpensive and short intervention would seem in line with the (non-)severity of their problems – more extensive help might in any event be rejected as ‘over the top’.

Dutch studies test computer-aided alcohol interventions across severity range

Prominent among studies testing these assumptions is a set from the Netherlands which in relation to drinking systematically mapped and examined the whole territory. These authors have both reviewed the evidence, and contributed to it by testing a computer-delivered therapy involving ‘text–chat’ conversations with a real therapist for problem and often dependent drinkers, an on-line cognitive-behavioural programme for excessive drinkers, and, at the lowest level of intervention intensity and problem severity, a 10-minute web-based brief intervention for risky drinkers.

Weekly alcohol consumption in gm

For the typically very heavy and multiply problematic drinkers in the first study who volunteered for it in order “to reduce your alcohol intake or quit drinking”, text chats with a real therapist had the greatest and most lasting impacts, but a fully automated process also worked substantially better than being placed on a waiting list and could be made more available with less resources chart. The less severe but still on average heavy drinkers in the second study who engaged with a web-based self-help programme and completed a follow-up survey had more often reverted to non-risky drinking than a control group. The still less heavy but nevertheless risky drinkers in the third study had been identified through screening or responded to ads and been allocated not to a self-help programme but a computerised version of a brief, single-session intervention. Again those who agreed to join the study and completed follow-ups had more often reverted to non-risky drinking than a control group.

Based partly on these results the Dutch team devised a mathematical model which simulates the health gains and costs of incorporating these new technologies in to a health care system for problem drinking. For the Netherlands, the results suggested national health would improve and/or costs be reduced if on-line brief interventions and therapy were added to or replaced conventional alcohol-related health care. However, the weak link in all the studies and therefore in the simulation was that few of the people who might have done joined the studies and completed follow-ups, raising doubts over whether impacts in these presumably atypical minorities would be matched on a larger scale.

Another way to do brief interventions

The concern that routine implementation across a broad population might prove less effective than among the minority who join trials seemed validated in studies of computerised brief interventions in New Zealand and Sweden, seen as the most real-world trials to date of this type of intervention among college students. In both cases the study and/or the intervention appealed to only a minority of students and impacts were at best slight and generally not statistically significant.

In finding at best small effects, these trials were typical of trials of computer-based brief interventions among students. Published in 2015 a study which amalgamated results from trials of (typically brief) computerised interventions found larger and more lasting drinking reductions among general adult samples of risky drinkers than among college students, among whom there was just a 12g per week reduction six months after intervention but no significant reduction after a year. Another review corroborated this disparity, but also argued that in some studies drinking amounts were so skewed that using the average to characterise them was misleading. After accounting for this the findings changed, and among students there was now no statistically significant reduction in drinking due to the interventions.

Stronger findings across non-college adult samples may be because, rather than for incentives or course credits, they join studies and access internet alcohol intervention sites in order to control a level and pattern of drinking which worries them. Drinking on average more heavily than students and having had longer to experience the ill effects, they have more reason and more scope to cut back. In a Canadian study it was only the higher-risk half of risky drinkers among a general adult sample who reduced their consumption and alcohol-related risk levels after being given access to a web-based brief intervention. Drinking less overall, and in a setting where heavy drinking is an accepted rite of passage and may be seen as a passing phase, it seems likely that students have less incentive to act on information and advice which would lead older and heavier drinkers responsible for families and jobs, and facing the possibility of chronic diseases as they age, to cut back.

Nobody is yet suggesting that computers can replace therapists for typical treatment populations, but further down the severity and complexity scale, the evidence is growing that the computer may add substance use reduction to its more familiar competencies among adult heavy drinkers who are drinking enough and concerned enough to respond to ads and join studies. When identified through screening in a college setting, the evidence from the more real-world studies is that effects are probably too little and too transient to make much of a difference to health.

Last revised 02 September 2015. First uploaded 01 May 2012

Comment/query to editor
Give us your feedback on the site (one-minute survey)
Open Effectiveness Bank home page
Add your name to the mailing list to be alerted to new studies and other site updates