Methodological problems evaluating clinical and public health interventions
Although researchers and health educators are motivated to find the best means of
promoting smoking cessation and healthy drinking, evaluating the effectiveness of
any intervention is fraught with methodological problems. For smoking cessation these
problems include: Who has become a non-smoker? Someone who has not smoked in the last month/week/
day? Someone who regards themselves as a non-smoker? (Smokers are notorious for
under-reporting their smoking.) Does a puff of a cigarette count as smoking? Do
cigars count as smoking? These questions need to be answered to assess success rates.
Who is still counted as a smoker? Someone who has attended all clinic sessions and still
smokes? Someone who dropped out of the sessions half-way through and has not
been seen since? Someone who was asked to attend but never turned up? These
questions need to be answered to derive a baseline number for the success rate.
Should the non-smokers be believed when they say they have not smoked? Methods other
than self-report exist to assess smoking behaviour, such as carbon monoxide in the
breath, cotinine in the saliva. These are more accurate, but time-consuming and
expensive.
How should smokers be assigned to different interventions? In order for success rates to
be calculated, comparisons need to be made between different types of intervention
(e.g. aversion therapy versus cue exposure). These groups should obviously be
matched for age, gender, ethnicity and smoking behaviour. What about stage of
change (contemplation versus precontemplation versus preparation)? What about
other health beliefs such as self-efficacy, costs and benefits of smoking? The list could
be endless.For interventions aimed at changing drinking behaviour, these problems include: What is the desired outcome of any intervention? Being totally abstinent (for the last
month/week)? Drinking a normal amount? (What is normal?) Coping with life?
(What constitutes acceptable coping?) Drinking that is not detrimental to work?
(Should work be a priority?) Drinking that is no longer detrimental to family life?
(Should family life be a priority?) In his autobiography, John Healy (1991) describes
his transition from an alcoholic living on the ‘Grass Arena’ in London to becoming
addicted to chess. Is this success? Should the experts impose their view of success on
a drinker, or should success be determined by them?
How should drinking behaviour be measured? Should intrusive measures such as blood
taking be used? Should self-reports be relied on?STAGE 4: RELAPSE IN SMOKING AND DRINKING
Although, many people are successful at initially stopping smoking and changing their
drinking behaviour, relapse rates are high. Interestingly, the pattern for relapse is124 HEALTH PSYCHOLOGY