Family Practice Advance Access originally published online on August 1, 2007
Family Practice 2007 24(5):475-480; doi:10.1093/fampra/cmm044
A clustered randomized trial of the effects of feedback using academic detailing compared to postal bulletin on prescribing of preventative cardiovascular therapy
Department of Pharmacology and Therapeutics, Trinity College, Trinity Centres for Health Sciences St James Hospital, Dublin D 8, Ireland
Correspondence to Corina Naughton, Department of Pharmacology and Therapeutics, Trinity College, Trinity Centres for Health Sciences St James Hospital, Dublin D 8, Ireland; Email: naughtc{at}tcd.ie
Received 9 September 2006; Revised 12 April 2007; Accepted 21 June 2007.
 |
Abstract |
|---|
 Top Abstract IntroductionMethodologyResultsDiscussionConclusionDeclarationReferences |
|---|
Background. Interventions to promote prescribing of preventive therapies in patients with cardiovascular disease (CVD) or diabetes have reported variable success.
Objective. (i) To evaluate the effect of prescribing feedback on GP practice using academic detailing compared to postal bulletin on prescribing of CVD preventive therapies in patients with CVD or diabetes at 3 and 6 months post intervention and (ii) to evaluate the intervention from a GP's perspective.
Methods. Volunteer GP practices (n = 98) were randomized to receive individualized prescribing feedback via academic detailing (postal bulletin plus outreach visit) (n = 48) or postal bulletin (n = 50). The proportion of CVD or diabetic patients on statins and antiplatelet agents/warfarin pre- and post-intervention was calculated for each GP practice. Multivariate regression with a random effects model was used to compare differences between the groups adjusting for GP clustering and confounding factors. ß-Coefficients and 95% confidence intervals (CIs) are presented.
Results. There was a 3% increase in statin prescribing in CVD patients at 6 months post-intervention for both randomized groups, but there was no statistical difference between the groups (ß = 0.004; 95% CI = –0.01 to 0.02). Statin and antiplatelet/warfarin prescribing also increased in the diabetic population; there was no significant differences between the groups. GPs participating in the project expressed a high level of satisfaction with both interventions.
Conclusion. Prescribing of preventive therapies increased in both randomized groups over the study period. But academic detailing did not have an additional effect on changing prescribing over the postal bulletin alone.
Keywords. Academic detailing, feedback, prescribing, primary care.
|
Introduction |
|---|
|
TopAbstractIntroductionMethodologyResultsDiscussionConclusionDeclarationReferences |
|---|
Mortality from cardiovascular disease (CVD) continues to decline in the western world.1 This is largely attributed to improved primary and secondary prevention with pharmacological agents and lifestyle changes.2 International guidelines on the management of CVD aim to standardize therapeutic targets for cholesterol and hypertension and promote prescribing of antiplatelet and stain agents as preventive therapies in patients with established CVD or Type II diabetes.3
The uptake of the clinical trial evidence and the guidelines has been slow and variable.4 Ireland is no exception with studies showing age, gender and regional variation in the prescribing of preventive therapies.5,6 At present, there are few interventions aimed at improving primary care physician prescribing of CVD preventive therapies in Ireland. The exception is the Heartwatch programme aimed at secondary prevention of CVD which involves 20% of GPs’ practices in Ireland and has enrolled over 7000 patients. The programme has shown significant improvements in the management of CVD risk factors after only 2 years.7 However, it targets only the highest risk population, those with a history of myocardial infarction, coronary artery bypass surgery or percutaneous transluminal coronary angioplasty and a small number of diabetic patients.
Alternative interventions which complement the Heartwatch programme but have a broader focus on patients with CVD and that can potentially involve a greater number of practices should be evaluated. Interventions involving individualized prescribing feedback alone or in addition to educational outreach visits have been evaluated in other countries.8,9 However, the results from these studies are variable and there is a need for further evaluation of these interventions in different health care systems.
The purpose of this study is to evaluate the effect of randomizing GP practices to prescribing feedback using academic detailing (postal bulletin plus an educational outreach visit) compared to postal bulletin alone on prescribing of cardiovascular preventive therapies in patients with CVD or diabetes. A secondary objective was to evaluate the effectiveness of both interventions from a GP's perspective.
|
Methodology |
|---|
|
TopAbstractIntroductionMethodologyResultsDiscussionConclusionDeclarationReferences |
|---|
Population selection
The sample size calculation was based on demonstrating a 25% improvement in appropriate prescribing, with a power of 80% and statistical significance of 5%, between the randomized groups and allowing for clustering of GPs within practices [cluster size = 2, intraclass correlation coefficient = 0.1]. The number of practices required per arm of the study was 26.10
Following ethical approval, 300 GPs based in the Eastern Regional Health Authority (ERHA) in Ireland were contacted. GPs had to have a minimum of 500 registered patients on the Health Service Executive-primary care reimbursement service (HSE-PCRS) scheme (see below for details). Thirty-seven per cent of GPs (n = 110; n = 98 practices) agreed to participate. These GPs were randomized to receive prescribing feedback using either (i) academic detailing including postal bulletin (n = 48 GP practices) or (ii) postal bulletin alone (n = 50 GP practices) (Fig. 1). The unit of randomization was the GP practice, with all GPs within the same practice being randomized to the same intervention. However, there were only 10 practice (n = 22 GPs) in which there was more than one GP participating in the study (two practices had three GPs, the remaining eight practices had two GPs participating). As the ERHA includes the largest urban centre in Ireland, the randomization was allocated by area of GP practice to ensure a balanced distribution of practitioners working in deprived and non-deprived areas and urban and semi-urban areas.
|
Data collection
The study utilized the HSE-PCRSs pharmacy claims database to provide individualized GP prescribing feedback and evaluate the effect of the interventions. The HSE-PCRS is a means tested health care scheme for all those under 70 years but is free to those aged 70 years or older. The scheme provides free health care including medicines to 30% (n = 1.15 million) of the Irish population. However, it over represents the elderly, the very young and the most socio-economically disadvantaged.
The HSE-PCRS pharmacy database includes a unique patient identification number with demographic information (age and gender) and the registered doctor number. Thus, each prescription can be linked to the patient and their registered GP. Full details of all drugs prescribed are recorded and the drugs are coded using the WHO Anatomical Therapeutic Chemical (ATC) classification system.11 The database does not contain information on patient diagnosis, instead specific drug therapy or combinations of drug therapy are used as surrogate markers of disease. This methodology has been reported by other authors.12–15 The surrogate markers for CVD are antiplatelet therapy (aspirin or clopidogrel, B01AC04 &06) plus or minus a coronary artery vasodilator (nitrates C01DA or nicorandil C01DX16), also coronary artery vasodilators are used to identify ischemic coronary artery disease.16 Diabetes is identified using insulin and oral hypoglycaemic agents (A10A/B). Data on all patients registered to the participating GPs, aged 45 years and older and who received at least one prescription for one of the above CVD or diabetes-related drugs was included in the analysis during the defined study periods.
Intervention: postal bulletin
The postal bulletins contained individualized GP prescribing feedback and educational information based on the 2003 European guidelines on CVD prevention.3 The aim of the bulletin was to increase the level of (i) antiplatelet prescribing in patients with coronary artery disease; (ii) statin prescribing (C10AA) in patients with CVD and (iii) antiplatelet and statin prescribing in patients with diabetes.5 The feedback was displayed using graphs and included the actual number of GP-registered patients not receiving recommended therapy. All bulletins were sent out in June 2005.
Academic detailing
Following distribution of the bulletin, the educational outreach visits took place between June and July 2005. Fifty-four visits were carried out by a single researcher to ensure consistency of the information given, one GP in the academic detailing group declined to participate at this stage. The visits ranged from 15 to 30 minutes and were held in the GP practice. Each visit consisted of a 10-minute powerpoint presentation based on the bulletin data with extra information on CVD risk factor management. The GP was encouraged to ask questions about the data and clarify any points raised. The majority of the visits were one to one with the GP participating in the project, though some practices held group meetings involving other GP partners and the practice nurse.
Evaluation questionnaires
The postal bulletin and outreach visit were evaluated separately. All GPs (n = 110) received a postal questionnaire with their prescribing feedback bulletin evaluating this aspect of the study. In addition, GPs in the academic detailing arm of the study received a separate questionnaire related to the outreach visit. The majority of questions were based on a three-point Likert scale and focused on the content, format and frequency of the bulletin and outreach visit. GPs were also encouraged to make free text comments on the study and its perceived effects on their practice.
Data analysis
GPs with <25 patients with CVD or 10 patients with diabetes were excluded from the analysis to avoid calculations based on small numbers (random error). Thus in the postal bulletin only group, one GP practice was excluded from the CVD and diabetes analysis (n = 49) while in the academic detailing group no GP practices (n = 48) were excluded (Fig. 1). Analysis was on an intention to treat basis.
The GP characteristics, patient population structure and post intervention evaluation questionnaires were compared using chi-square statistic for categorical variables or analysis of variance for continuous variables. The analysis time periods were 3 months (August 2005–October 2005) and 6 months post-intervention (August 2005–January 2006) compared to 3 months (March 2005–May 2005) and 6 months pre-intervention (December 2004–May 2005). All prescribing data and GP patient population structure was obtained from the HSE-PCRS. In each analysis period, the GP registered patient populations with CVD (excludes patients with diabetes), coronary artery disease and diabetes were identified. Following this, the proportion of patients in each disease group receiving statin or anticoagulant therapy (antiplatelet/warfarin therapy; patients on warfarin therapy are considered anticoagulated and may not normally be considered eligible for additional antiplatelet therapy) was calculated. Pre-intervention mean population proportion and the post intervention proportional increase in the population receiving appropriate therapy are presented. The proportion of males and those aged 70 years or older was also calculated for each practice.
The prescribing differences between the randomized groups were examined using multivariate regression with a random effects model for each therapy, e.g. statin therapy in patients with CVD. This allowed for adjustment of GP clustering within practices, baseline prescribing and confounding factors such as practice participation in the Heartwatch programme and patient population structure. ß-Coefficients and 95% confidence intervals are presented. All analyses were performed using SAS (v 9.1 SAS Institute, Cary, NC).
|
Results |
|---|
|
TopAbstractIntroductionMethodologyResultsDiscussionConclusionDeclarationReferences |
|---|
There was no statistical difference between the data at 3 and 6 months. Thus, the later 6-month data are presented because it is more likely to capture an increased number of patients with CVD or diabetes (repeat prescriptions are issued for a 3-month period).
The characteristics of the GPs in both groups were similar with no significant difference seen in GP age, gender, years qualified or participation in the Heartwatch programme (Table 1). Patient panel size and distribution of males were similar between the groups but academic detailing practices had significantly more patients aged 70 years and older compared to the postal bulletin group (P < 0.001).
|
CVD patient analysis
Patients receiving cardiovascular medication represent
30% of the patient population aged 45 years or older in both groups. Pre-intervention just over 50% of patients in both the academic detailing and postal bulletin groups were receiving statin therapy (see Table 1). Post-intervention after adjusting for baseline prescribing and other confounding factors, there was no significant differences seen between the groups. Both groups increased statin prescribing in this population by 3% over a 6-month period. In patients with coronary artery disease (on nitrate therapy), 96% were already receiving anticoagulants. Post-intervention there was a 1% increase in prescribing seen in both groups.
Diabetes patient analysis
Patients receiving diabetic therapy accounted for 7% of patients aged 45 years or over in both randomized groups. Pre-intervention between 64% and 66% of diabetic patients were on statin therapy, both groups increased statin prescribing by 4–5% in the post intervention period but there was no significant difference between the groups. A higher proportion of diabetic patients were on anticoagulant therapy, between 74% and 76%; again post intervention there was a small increase in anticoagulant prescribing with no significant difference found between the groups (see Table 2).
|
Evaluation questionnaire
In the questionnaire evaluating the outreach visit aspect of the study, there was a response rate of 89% (48/54). There was a high level of satisfaction with the content and format of the visit and 81% (39/48) indicated that they would like the academic detailing visits to continue in conjunction with the postal bulletin. GPs found that the visits prompted them to review the postal bulletin and they liked the opportunity to clarify points raised in the bulletin.
There was a response rate of 80% (88/110) in the questionnaire evaluating the postal bulletin aspect of the study. In total, 45 postal bulletin and 43 academic detailing GPs replied with no significant difference between the groups in their replies to the questions. Both groups expressed a high level of satisfaction with the respective interventions they received. All the respondents indicated they would like to continue receiving prescribing feedback as a postal bulletin. The majority 94% (79/84) felt the feedback had some impact on practice. Those who expanded on this point identified an increased awareness of their own performances against recommended practice.
|
Discussion |
|---|
|
TopAbstractIntroductionMethodologyResultsDiscussionConclusionDeclarationReferences |
|---|
There was an increase in the prescribing of preventive therapies seen in both academic detailing and postal bulletin GP practices over the study period. Statin prescribing in patients with CVD and diabetes increased by 3–5%, while anticoagulant therapy in diabetic patients increased by 2–3% in both groups. However, at 3 and 6 months post-intervention, academic detailing did not have a significant additional impact on prescribing of these therapies above postal bulletin alone. Anticoagulant prescribing in patients on nitrate therapy was already at 96% pre-intervention in both groups and it was not anticipated that this could be significantly improved upon as the small numbers of patients not on this therapy are likely to be those who refuse or are unable to tolerate anticoagulants.
In the evaluation of the project from the GP's perspective, 81% of those who received an outreach visit felt it was useful, while all GPs who returned their evaluation questionnaires (80%) would like to see the postal bulletins continue. Both groups acknowledge the limitations of a one-off intervention in having a sustained effect on practice but felt that participation in the project had effected their prescribing in some way and all would like to continue receiving prescribing feedback, with or without the visit.
Academic detailing is described as one of the most promising interventions used to promote changes in physician behaviour9 and has been used widely in the area of CVD. Simon et al.17 reported a significant increase in prescribing of certain anti-hypertensive therapies, while Lobo et al.18 identified significant improvement in preventive measures for CVD but did not include prescribing of preventive therapies. However, negative results from academic detailing have also been reported. Fretheim et al.19 found a single significant change in thiazide use out of nine other quality care indicators in prevention of CVD, while Witt et al.20 reported that academic detailing had no effect on prescribing of asthma medication.
In this study, over 20–40% of high-risk patients were not receiving recommended preventive therapy. The provision of academic detailing in addition to individualized prescribing feedback had a minimal impact on clinical practice though many GPs believed the intervention increased their awareness of this issue. The gap between clinical practice and clinical knowledge is well documented especially in relation to CVD.4 The lack of intensity of the intervention in this study (one bulletin and a single visit with no follow-up) may certainly have contributed to the negative result.21,22 In addition, prescribing feedback with or without academic detailing only provides motivation for GPs to change practice through increasing awareness and reinforcing best practice guidelines. It does not tackle more complex obstacles to effective prescribing in CVD. These factors include patient non-compliance, variation in patient morbidities, lack of computerized patient registers, lack of time, ancillary personnel, space to run dedicated clinics and financial incentive to deal with the extra work load, such obstacles are readily identified in the literature as well as by the GPs participating in this study.23–25
Providing a prescribing feedback service plus or minus prescribing advisers would require a considerable investment in financial resources without guaranteeing substantial improvement in the management of CVD in the community.9 This is particularly likely to be the case in the absence of investment in primary care infrastructure such as computerized patient registers, automated patient follow-up, additional staff to run CVD clinics as in the Heartwatch programme and patient education. Prescribing feedback may be a useful adjunct to multifaceted interventions targeting local barriers to CVD management but in isolation it is not likely to have a substantial effect.
Study limitations
There is a lack of diagnosis information to validate the quality prescribing indicators. This was a voluntary sample of GPs; thus, the acceptability of prescribing feedback to a broader population of GPs is difficult to determine. Not all GPs in multi-partner practices agreed to participate, which will have affected the sample size calculation (based on assuming at least two GPs per practice participated in the study). This also meant that the data analysis was based on the individual GPs’ registered population and not the whole practice patient population. Thus, some patients who may have been prescribed preventative therapy by the participating GP but who were not registered with that GP were excluded from the analysis. In addition, using the registered GP patient number to identify patients with CVD or diabetes does not guarantee that the participant GP in multi-partner practices prescribed the therapies under review. However, patients tend to see their own GP whenever possible especially for management of chronic conditions. Also, for the purpose of data analysis, a stable patient population was required. The pharmacy database covers a particular group of the Irish population and may not be generalizable to the rest of the population; however, the database represents 65% of all prescribing in primary care.
|
Conclusion |
|---|
|
TopAbstractIntroductionMethodologyResultsDiscussionConclusionDeclarationReferences |
|---|
The prescribing of preventative therapies for CVD and diabetes showed a small increase after a randomized intervention to improve prescribing, but academic detailing had no additional impact over postal bulletin alone. There was a high level of satisfaction among participant GPs in both intervention arms of the study. Interventions to further improve management of CVD and diabetes in primary care need to address local barriers to implementing evidence-based medicine.
|
Declaration |
|---|
|
TopAbstractIntroductionMethodologyResultsDiscussionConclusionDeclarationReferences |
|---|
Funding: Health Research Board of Ireland (Grant No. HO1199).
Ethical approval: The Irish College of General Practitioners, Protocol number REC0904-1.
Conflicts of interest: None.
|
Acknowledgments |
|---|
We would like to thank all the GPs in the ERHA who took part in the intervention study and Fergus O'Kelly for his advise and support. Also the HSE-PCRS payments board for the data on which this study is based.
|
Notes |
|---|
Naughton C, Feely J and Bennett K. A clustered randomized trial of the effects of feedback using academic detailing compared to postal bulletin on prescribing of preventative cardiovascular therapy. Family Practice 2007; 24: 475–480.
|
References |
|---|
|
TopAbstractIntroductionMethodologyResultsDiscussionConclusionDeclarationReferences |
|---|
1 Kabir Z, Bennett K, Shelley E, et al. The population mortality benefits of maximizing the number of eligible patients receiving appropriate cardiology treatments in Ireland. QJM (2006) 99:523–530.
2 Yusuf S, Hawken S, Ounpuu S, Tiukinhoy S, Rochester C. Effect of potentially modifiable risk factors associated with myocardial infarction in 52 countries (the INTERHEART study): case-control study. Lancet (2004) 364(9438):937–952.[CrossRef][Web of Science][Medline]
3 de Backer G, Ambrosioni E, Borch-Johnsen K, et al. European guidelines on cardiovascular disease prevention in clinical practice. Third Joint Task Force of European and other Societies on Cardiovascular Disease Prevention in Clinical Practice (constituted by representatives of eight societies and by invited experts). Atherosclerosis (2004) 173:381–391.[CrossRef][Medline]
4 Majumdar SR, McAlister FA, Furberg CD. From knowledge to practice in chronic cardiovascular disease: a long and winding road. J Am Coll Cardiol (2004) 43:1738–1742.
5 Bennett KE, Williams D, Feely J. Inequalities in prescribing of secondary preventative therapies for ischaemic heart disease in Ireland. Ir Med J (2002) 95:169–172.[Medline]
6 Bennett KE, Williams D, Feely J. Under-prescribing of cardiovascular therapies for diabetes in primary care. Eur J Clin Pharmacol (2003) 58:835–841.[Web of Science][Medline]
7 National Hearwatch Programme. Heartwatch Clinical Report, March 2003–December 2005. http://www.icgp.ie (accessed on July 2, 2007).
8 Jamtvedt G, Young JM, Kristoffersen DT, Thomson O'Brien MA, Oxman AD. Audit and feedback: effects on professional practice and health care outcomes. Cochrane Database Syst Rev (2003) 3. CD000259.
9 Thomson O'Brien MA, Oxman AD, Davis DA, Haynes RB, Freemantle N, Harvey EL. Educational outreach visits: effects on professional practice and health care outcomes. Cochrane Database Syst Rev (2000) 2. CD000409.
10 Campbell MK, Thomson S, Ramsay CR, MacLennan GS, Grimshaw JM. Sample size calculator for cluster randomized trials. Comput Biol Med (2004) 34:113–125.[CrossRef][Web of Science][Medline]
11 WHO. Anatomical Therapeutics Chemical (ATC) Classification Index with Defined Daily Doses (DDDs) (2001) WHO, Geneva, Switzerland. http://www.who.int/medicines (accessed on July 8, 2007).
12 Maio V, Yuen E, Rabinowitz C, et al. Using pharmacy data to identify those with chronic conditions in Emilia Romagna, Italy. J Health Serv Res Policy (2005) 10:232–238.
13 Von KM, Wagner EH, Saunders K. A chronic disease score from automated pharmacy data. J Clin Epidemiol (1992) 45:197–203.[CrossRef][Web of Science][Medline]
14 Hoven JL, Haaijer-Ruskamp FM, Vander Stichele RH. Indicators of prescribing quality in drug utilisation research: report of a European meeting (DURQUIM, 13-15 May 2004). Eur J Clin Pharmacol (2005) 60:831–834.[CrossRef][Web of Science][Medline]
15 Kendall H. Why prescribing data are monitored. Pharm J (2004) 272:21–22.
16 Gray J, Majeed A, Kerry S, Rowlands G. Identifying patients with ischaemic heart disease in general practice: cross sectional study of paper and computerised medical records. BMJ (2000) 321:548–550.
17 Simon SR, Majumdar SR, Prosser LA, et al. Group versus individual academic detailing to improve the use of antihypertensive medications in primary care: a cluster-randomized controlled trial. Am J Med (2005) 118:521–528.[CrossRef][Web of Science][Medline]
18 Lobo CM, Frijling BD, Hulscher ME, et al. Improving quality of organizing cardiovascular preventive care in general practice by outreach visitors: a randomized controlled trial. Prev Med (2002) 35:422–429.[CrossRef][Web of Science][Medline]
19 Fretheim A, Oxman AD, Havelsrud K, Treweek S, Kristoffersen DT, Bjorndal A. Rational prescribing in primary care (RaPP): a cluster randomized trial of a tailored intervention. PLoS Med (2006) 3:e134.[CrossRef][Medline]
20 Witt K, Knudsen E, Ditlevsen S, Hollnagel H. Academic detailing has no effect on prescribing of asthma medication in Danish general practice: a 3-year randomized controlled trial with 12-monthly follow-ups. Fam Pract (2004) 21:248–253.
21 Horn FE, Mandryk JA, Mackson JM, Wutzke SE, Weekes LM, Hyndman RJ. Measurement of changes in antihypertensive drug utilisation following primary care educational interventions. Pharmacoepidemiol Drug Saf (2007) 16:297–308.[CrossRef][Web of Science][Medline]
22 Kiefe CI, Allison JJ, Williams OD, Person SD, Weaver MT, Weissman NW. Improving quality improvement using achievable benchmarks for physician feedback: a randomized controlled trial. J Am Med Assoc (2001) 285:2871–2879.
23 Perreault S, Lamarre D, Blais L, et al. Persistence with treatment in newly treated middle-aged patients with essential hypertension. Ann Pharmacother (2005) 39:1401–1408.
24 Grol R, Grimshaw J. From best evidence to best practice: effective implementation of change in patients’ care. Lancet (2003) 362:1225–1230.[CrossRef][Web of Science][Medline]
25 Nazareth I, Freemantle N, Duggan C, Mason J, Haines A. Evaluation of a complex intervention for changing professional behaviour: the Evidence Based Out Reach (EBOR) Trial. J Health Serv Res Policy (2002) 7:230–238.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  R. Ostini, D. Hegney, C. Jackson, M. Williamson, J. M Mackson, K. Gurman, W. Hall, and S. E Tett Systematic Review of Interventions to Improve Prescribing Ann. Pharmacother., March 1, 2009; 43(3): 502 - 513. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||