Healthy Skepticism International News

December 2002

Does teaching evaluation of drug promotion improve the quality and economy of prescribing?

Aims

To investigate whether teaching GPs how to evaluate drug promotion leads to improved prescribing.

Background

There are large differences between GPs in prescribing rates per consultation for key drug classes.[1],[2] These variations are associated with avoidable adverse health outcomes[3] and opportunity costs.  For example, the main cause of recent growth in PBS expenditure is rapid uptake of new drugs, especially COX2 inhibitors.[4]  COX2s have little or no overall advantage[5] and may have more total adverse effects[6] at much higher cost.  One of the causes of rapid uptake of COX2s is misleading promotion.[7]

Wide variations in the quality and cost of prescribing often correlate with wide variations in use of, and attitudes towards, drug promotion.[8],[9],[10]  Analysis of 1,279 responses to a National Prescribing Service case study found 4 independent predictors of believing in the appropriateness of new drugs despite no good evidence: 1) more recent graduation, 2) stronger belief that “My prescribing is improved by information from drug companies”, 3) belief that a lower % of drug advertisements contain potentially misleading claims and 4) more frequent attendance at drug company sponsored meetings.[11]

Pharmaceutical promotion is a major influence on prescribing[12],[13] often via psychological techniques[14] including use of triggers for “heuristics” (inductive decision making short cuts) eg “newer is better” [15].  Use of heuristics is common when people lack skills or time for evaluating all the evidence.[16]  Using heuristics can be an efficient path to the correct conclusion but can make people vulnerable to being misled.[17]  Promotional triggers are often fallacies[18],[19] that have been shown to work with a range of methodologies since the 1920s, including controlled trials with 25,000 subjects.[20]

Most interventions for improving prescribing involve persuading doctors to accept additional evidence-based information,[21],[22] but doctors are suffering information overload.[23]  It may help to enable GPs to sort justified vs misleading information.  Teaching about promotional fallacies also leads to changes in attitudes towards, and plans for use of, drug promotion[24] and that may then lead to improved prescribing.[8],[9],[10],[11]

Teaching about promotional fallacies and psychological techniques is effective for improving skills and attitudes.[24],[25],[26],[27]  RCTs of educational interventions for GP registrars aimed at improving prescribing are feasible.[28]  Interventions for improving prescribing are more effective if a combination of techniques, such as individualised feedback and providing information about drug costs,[29] are used synergistically.[21],[22]

Research Project
1)  Cochrane systematic review:
Title: Teaching evaluation of pharmaceutical promotion to medical students and doctors.  The Cochrane Effective Practice and Organisation of Care Group has indicated that this topic is likely to be accepted and supported by them.

2a)  Survey of Australian GPs self-reported use of and attitudes towards information sources compared with their responses to clinical vignettes.

2b)  Survey of Australian GP registrars’ self-reported use of and attitudes towards information sources compared with their responses to clinical vignettes and their Health Insurance Commission (HIC) prescribing data.

These surveys will use similar methods to a 1982 Dutch study[8] except data collection will be from questionnaires and the HIC, the focus will be on new rather than old drugs and advances in the study of attitudes[30] will be utilised.

3)  Controlled trial:
Does teaching Australian GP registrars how to evaluate drug promotion improve the quality and economy of prescribing?

Participants:
All final year GP registrars in 2004, whose training consortia agree to include the trial in their curriculum, will be invited to participate.  The reason for choosing registrars rather than established GPs is that concordance with an educational intervention is more likely with registrars if it is part of their course than with GPs unless there were some other incentives.  (NB. It may not be appropriate to generalise the findings to all GPs.  It is hoped to do a separate trial for all GPs later.)

Intervention:
Teaching GP registrars evaluation of the fallacies and psychological techniques commonly used in pharmaceutical promotion integrated with providing: personalised feedback, information about drug costs, skills for evidence-based prescribing decisions and evidence based recommendations for appropriate use of, or for replacing, target drugs.  The target drugs will be COX2 inhibitors plus antibiotics, calcium channel blockers, A2 antagonists, HRT, new diabetes and new asthma drugs.  The list of target drugs may need to be changed according to current prescribing patterns detected with the survey 2b) above.  The material will be based on experience gained over 20 years with MaLAM letters (http://www.healthyskepticism.org/publications/publicat.htm), the Healthy Scepticism NZ newsletter (http://web.archive.org/web/20050211211731/http://www.healthyskepticism.org/NZ/healthy.htm), Teaching Material for Critical Appraisal of Pharmaceutical Promotion commissioned by WHO EDM (Essential Drugs and other Medicines policy department), a pilot RCT with GP registrars[24] and (in preparation) a module about drug promotion for the National Prescribing Service’s National Prescribing Curriculum (http://nps.unisa.edu.au).  The teaching medium will be similar to the later ie multiple brief visits to a password controlled, interactive website using case vignettes[31] and individualised responses depending on the registrars answers to questions.  This will enable automated data collection and national access including for registrars in rural areas.  The intervention will be written during 2003 then updated and fine tuned during the intervention period in 2004.

Comparison group:
Teaching evaluation of diagnostic tests for indications not covered in the prescribing intervention.  The target tests will be the screening tests for cervical, breast, colon and prostate cancer, and coronary artery calcium.  (Teaching about tests may lead to some effects on prescribing.  However we estimate that the effect size will be small.)

Primary Outcome:
Count of initiation (not continuation[32]) prescriptions dispensed for all target drugs recorded by the HIC per consultation.  HIC data has limitations when used for comparing individuals[2] but if the allocation process produces two groups of registrars seeing near equal numbers of concession cardholders overall then those limitations will not hamper comparison of the groups.

Secondary Outcomes:
a) Count of all initiation prescriptions dispensed per consultation
b) average cost to HIC of all prescriptions dispensed.
c) total prescription costs to HIC (NB c=a×b.) The study will not be powered to detect small but important changes in a, b and c but these will be measured in case of large changes.
d) Knowledge: Preferred drugs in responses to vignettes.
e) Attitudes to the teaching and to different information sources.
f) Skills: what sources of information are being used for prescribing decisions and basic evidence based medicine critical appraisal skills.
g) Behaviours: how many contacts with industry and non-industry information sources.

Duration:
Intervention over 1 year then 5 years of HIC data.  Secondary endpoints will be available of analysis and publication at the end of the intervention year.  HIC data at the end of the following year will be analysed for an interim publication.

Design:
Adequately concealed minimisation[33] by consortia, and by factors known to influence prescribing: sex, medical school and rural vs urban.  Some contamination from registrars sharing what they have learned with those in the other group is expected.  However the contamination effect size is probably not be large enough to justify cluster randomisation by consortia.[34]  There will be a planned subgroup analyses of registrars who become full time (Medicare income >$60,000[35]) vs part time GPs because part timers may worsen the noise to signal ratio.[36]

Significance

It has been estimated that improved prescribing for uncomplicated hypertension controlled by mono-therapy could have saved around $50 million in 1998[37],[38]  Had GPs had the skills to evaluate the promotion of COX2 inhibitors that Healthy Skepticism demonstrated in March 20007 then much of the $266 million wasted on COX2 inhibitors in 2001[4] could have been saved.  Improving prescribing may involve increased prescribing and thus increased costs (eg aspirin for patients with atrial fibrillation) but in most cases under-prescribed drugs are off patent and thus relatively cheap (which is why they are not promoted) or they lead to decreased total health care costs (eg aspirin prevents or delays cardiovascular hospital admissions).

If successful the intervention may reduce total prescribing costs by an amount of the order of $ 2,000-20,000 per GP per year, on average.[22]  The intervention could be made available for all GPs across Australia including rural areas at no marginal cost other than payment for the 20 hours involved eg $2,000 each.

Improvements in the quality of prescribing may lead to further savings via prevention or delay of the need for other health care eg less thombolic events and breast cancer from not prescribing HRT. Both the direct and indirect savings would then be available for other areas of health care.  Consequently improving prescribing may both directly and indirectly increase the length and improve the quality of life for large numbers of people in Australia and possibly other countries as well.

View/Hide References

HS Int News index

Page views since 15 March 2010: 5265

Comments

Our members can see and make comments on this page.