- Split View
-
Views
-
Cite
Cite
Patrik Midlöv, Åsa Bondesson, Tommy Eriksson, Christina Nerbrand, Peter Höglund, Effects of educational outreach visits on prescribing of benzodiazepines and antipsychotic drugs to elderly patients in primary health care in southern Sweden, Family Practice, Volume 23, Issue 1, February 2006, Pages 60–64, https://doi.org/10.1093/fampra/cmi105
- Share Icon Share
Abstract
Background. Different methods have previously been tested to affect GPs' prescribing habits. Attention has been drawn to benzodiazepines and antipsychotic drugs that are associated with several adverse effects in the elderly.
Objective. To evaluate if educational outreach visits to GP practices can affect the prescribing of benzodiazepines and antipsychotic drugs to the elderly and to evaluate the opinions of the participating GPs on such education.
Methods. In the county of Skåne, Sweden, 41 GP practices were invited to participate in educational outreach visits. Fifteen GP practices accepted the invitation. Practices were randomised to active (8 practices, 23 physicians) and control group (7 practices, 31 physicians). After the educational outreach visits prescribing of benzodiazepines and antipsychotic drugs to patients 65 years or older were measured for 1 year. The control group participated in the education after the study period. The opinions of GPs on educational outreach visits were evaluated.
Results. One year after the educational outreach visits there were significant decreases in the active group compared to control group in the prescribing of medium- and long-acting benzodiazepines and total benzodiazepines but not so for antipsychotic drugs.
Conclusions. Educational outreach visits can be effective in modifying GPs' prescribing habits. We have shown this to be so for prescribing of benzodiazepines to elderly patients in primary health care. Educational outreach visits are also very well appreciated by participating GPs.
Midlöv P, Bondesson A, Eriksson T, Nerbrand C and Höglund P. Effects of educational outreach visits on prescribing of benzodiazepines and antipsychotic drugs to elderly patients in primary health care in southern Sweden. Family Practice 2006; 23: 60–64.
Introduction
In Sweden about 50% of all drugs are prescribed by GPs.1 These GPs obtain most of their information on drug treatment from pharmaceutical companies.2 With rising drug-costs more producer-independent information has been requested. In a Cochrane review on educational outreach visits the authors concluded that these appear to be a promising approach to modifying health professionals' behaviour, especially prescribing.3 In another review of the effect of continuing medical education (CME) strategies where studies from 1974 to 1994 were evaluated the authors concluded that effective CME such as systematic practice-based interventions and outreach visits are seldom used compared with less effective methods such as conferences.4 Of later studies on educational interventions in family practice some have5–7 and some have not8,9 shown improvements in outcome measures. Studies that measured the GPs' knowledge after educational intervention have shown positive effects.6,10
Benzodiazepines (BDZ) are widely used and their use is highest among the elderly.11 In southern Sweden (Skåne) patients 65 years and older use 84.2 defined daily doses (DDD)/1000 inhabitants compared with 27.4 DDD/1000 inhabitants for all age groups.12 BDZ have a wide range of CNS effects such as sedation, drowsiness, memory difficulties and lack of co-ordination.13,14 The elderly experience excessive sedation from BDZ compared with younger individuals.13 In a study on 308 patients with suspected dementia, dementia was attributed in 13 patients to chronic treatment with a single BDZ.15 The prolonged half-life of many BDZ in the elderly increases the risk of accumulation, which could cause the serum concentration to reach a level where a delirium is induced.16 The strongest evidence for BDZ-related impaired cognition in the elderly is from a study that documented the improvement of cognitive deficits upon drug withdrawal.15 According to one study of more than 400 hospitalised patients the relative risk of developing cognitive impairment was 3.5 (95% CI: 1.4–8.8) for those who reported taking BDZ in daily doses equivalent to 5 mg or more of diazepam.17 Reducing BDZ use by elderly patients is important for several reasons. BDZ use by elderly patients is not only associated with cognitive side effects,13,14 but also increases the risk of hip fractures.18 According to an American study treatment with BDZ appears to increase the risk of hip fractures even at modest doses.18 These fractures lead to hospitalisation costs. In a European study it was estimated that costs of accidental injuries related to BDZ use in the EU varied between Euro 1.5 and 2.2 billion each year. More than 90% of these costs were in the elderly with fractures as the major contributor.19
Antipsychotic drugs are often used for the treatment of behavioural symptoms in demented persons.16 These drugs are associated with several adverse effects, such as extrapyramidal and anticholinergic effects including delirium.11,20 Antipsychotic drugs have long been reported to cause delirium. Blockade of α-receptors is a common effect among these drugs, can cause orthostatic reactions, which further increase the risk.16 All traditional antipsychotic drugs have been reported to increase the risk of delirium.21 In one study on patients who were transferred from the psychiatric ward to the medical ward because of delirium, 31% of cases were due to low-potency antipsychotic agents.22
The aim of this study was to evaluate whether educational outreach visits to GP practices can affect the prescribing of BDZ and antipsychotic drugs to the elderly and to evaluate the opinions of the participating GPs towards such education.
Material and methods
Subjects and practices
The committee for research ethics at the University of Lund gave their approval to the study. Each GP gave his or her informed consent to participation. All (41) GP practices in two districts of the county of Skåne in the south of Sweden were offered group-education programmes on “drug treatment that may cause confusion in the elderly” by a physician (PM) and a pharmacist (ÅB). Among the 15 GP practices that accepted eight were randomised to active education and seven to control group (Fig. 1). Characteristics of the practices are described in Table 1. For the active group the group-education was held two times at each centre from October–December 2000. The prescribing of BDZ and antipsychotic drugs was then monitored during the period October 1, 2000–December, 2001 for the active and the control groups. The control group participated in the education after the study period.
Centre number . | Active group. Number of GPs . | Centre number . | Control group. Number of GPs . |
---|---|---|---|
1 | 2 | 9 | 5 |
2 | 2 | 10 | 6 |
3 | 3 | 11 | 5 |
4 | 4 | 12 | 2 |
5 | 4 | 13 | 5 |
6 | 1 | 14 | 5 |
7 | 3 | 15 | 3 |
8 | 4 | ||
Median | 3 | Median | 5 |
Centre number . | Active group. Number of GPs . | Centre number . | Control group. Number of GPs . |
---|---|---|---|
1 | 2 | 9 | 5 |
2 | 2 | 10 | 6 |
3 | 3 | 11 | 5 |
4 | 4 | 12 | 2 |
5 | 4 | 13 | 5 |
6 | 1 | 14 | 5 |
7 | 3 | 15 | 3 |
8 | 4 | ||
Median | 3 | Median | 5 |
Centre number . | Active group. Number of GPs . | Centre number . | Control group. Number of GPs . |
---|---|---|---|
1 | 2 | 9 | 5 |
2 | 2 | 10 | 6 |
3 | 3 | 11 | 5 |
4 | 4 | 12 | 2 |
5 | 4 | 13 | 5 |
6 | 1 | 14 | 5 |
7 | 3 | 15 | 3 |
8 | 4 | ||
Median | 3 | Median | 5 |
Centre number . | Active group. Number of GPs . | Centre number . | Control group. Number of GPs . |
---|---|---|---|
1 | 2 | 9 | 5 |
2 | 2 | 10 | 6 |
3 | 3 | 11 | 5 |
4 | 4 | 12 | 2 |
5 | 4 | 13 | 5 |
6 | 1 | 14 | 5 |
7 | 3 | 15 | 3 |
8 | 4 | ||
Median | 3 | Median | 5 |
Outreach visit
The physician and the pharmacist visited each practice twice with an interval of 2–8 weeks. The first meeting dealt with different causes of confusion in the elderly. These causes included medications but also other reasons e.g. infections and other illnesses. Literature as well as actual cases were discussed. The second meeting dealt with the effects of BDZ and psychotropic drugs in the elderly. Special attention was drawn to the risks of BDZ with medium- or long-acting duration of action. Medium-acting BDZ prescribed in Sweden are alprazolam, nitrazepam and flunitrazepam whereas diazepam is the only long-acting BDZ. The main message was that these drugs only should be used in the elderly after the potential disadvantages have been considered. We did however stress that it is not easy to withdraw long-term therapies. If trying to do so we recommended thorough evaluation of the effects. There were no discussions about what prescribing was going to be measured or when. Actual cases were discussed and commented on. After the second visit the physicians' opinions were evaluated using an anonymous questionnaire. The questionnaire consisted of six questions that could be answered by selecting a number on a scale of 1–10, where 1 mean strongly disagree and 10 mean strongly agree (Table 2). The questionnaire was constructed based on a questionnaire used in a similar study.6 The opinions were evaluated both in the active and the control groups.
. | Q1 (25%) . | Median . | Q3 (75%) . | Range . |
---|---|---|---|---|
(1) Content was relevant to my clinical work | 9 | 10 | 10 | 2–10 |
(2) The material was well presented | 9 | 10 | 10 | 2–10 |
(3) My questions were answered adequately | 9 | 10 | 10 | 2–10 |
(4) This education improved my knowledge | 7 | 8 | 10 | 1–10 |
(5) This education will lead to changes in my prescriptions of drugs | 6 | 8 | 9 | 1–10 |
(6) In the future I would like to participate in this kind of education | 9 | 10 | 10 | 2–10 |
. | Q1 (25%) . | Median . | Q3 (75%) . | Range . |
---|---|---|---|---|
(1) Content was relevant to my clinical work | 9 | 10 | 10 | 2–10 |
(2) The material was well presented | 9 | 10 | 10 | 2–10 |
(3) My questions were answered adequately | 9 | 10 | 10 | 2–10 |
(4) This education improved my knowledge | 7 | 8 | 10 | 1–10 |
(5) This education will lead to changes in my prescriptions of drugs | 6 | 8 | 9 | 1–10 |
(6) In the future I would like to participate in this kind of education | 9 | 10 | 10 | 2–10 |
Values are presented as 25 percentile (Q1), 50 percentile (median) and 75 percentile (Q3).
. | Q1 (25%) . | Median . | Q3 (75%) . | Range . |
---|---|---|---|---|
(1) Content was relevant to my clinical work | 9 | 10 | 10 | 2–10 |
(2) The material was well presented | 9 | 10 | 10 | 2–10 |
(3) My questions were answered adequately | 9 | 10 | 10 | 2–10 |
(4) This education improved my knowledge | 7 | 8 | 10 | 1–10 |
(5) This education will lead to changes in my prescriptions of drugs | 6 | 8 | 9 | 1–10 |
(6) In the future I would like to participate in this kind of education | 9 | 10 | 10 | 2–10 |
. | Q1 (25%) . | Median . | Q3 (75%) . | Range . |
---|---|---|---|---|
(1) Content was relevant to my clinical work | 9 | 10 | 10 | 2–10 |
(2) The material was well presented | 9 | 10 | 10 | 2–10 |
(3) My questions were answered adequately | 9 | 10 | 10 | 2–10 |
(4) This education improved my knowledge | 7 | 8 | 10 | 1–10 |
(5) This education will lead to changes in my prescriptions of drugs | 6 | 8 | 9 | 1–10 |
(6) In the future I would like to participate in this kind of education | 9 | 10 | 10 | 2–10 |
Values are presented as 25 percentile (Q1), 50 percentile (median) and 75 percentile (Q3).
Data collection and statistical analysis
Prescription data were collected from Apoteket AB (National Corporation of Swedish Pharmacies) measured as DDD dispensed during 3-month periods from October 2000 until December 2001. In Sweden prescriptions often last 1 year and many patients have a yearly check-up by their GP. We wanted to investigate if prescriptions were affected 1 year after the educational outreach visit. We intended to do follow-up on the long-term effects of our educational outreach visits. However, after June 2002 there was a change in the way prescribing-statistics are collected by Apoteket AB making it impossible to compare data before and after June 2002. There are no data on prescription for individual GPs but only for all GPs working at a specific practice. We compared logarithmically transformed ratios of mean values, for each practice, of prescriptions (DDD) at each 3-month period after the intervention over the prescriptions during the first 3-month period, October–December 2000. The analyses were performed using a mixed model with the group by period interaction as fixed effects and practices as random effects using SAS proc mixed (ver 8.2, SAS Institute, Cary, NC, USA). Differences of the least square mean estimates between the two groups, and their confidence limits, were calculated for each period. The results are presented as geometric means and confidence intervals.
Results
Characteristics of the practices
The participating practices were situated in villages or small towns except one in the control group that was in a city of 100 000 inhabitants. The centres were all publicly financed. The number of GPs varied from one to six across the practices (Table 1).
Prescription
One year after the educational outreach visits there was a decrease in prescribing of medium- and long-acting benzodiazepines (25.80%) and total BDP (26.63%) in the active group compared with the control group. These decreases were significant, P < 0.05 (Table 3). There were significant decreases after 9 months but not after 3 months or 6 months. For antipsychotic drugs there were no significant differences between active and control group.
. | January–March 2001 . | April–June 2001 . | July–September 2001 . | October–December 2001 . |
---|---|---|---|---|
BDZa | −17.48(−37.96 to 9.75) | −18.26 (−38.54 to 8.70) | −27.87* (−45.77 to −4.08) | −25.80* (−44.20 to −1.32) |
Medium- and long-acting BDZb | −17.31 (−39.31 to 12.66) | −17.18 (−39.21 to 12.84) | −28.27* (−47.36 to −2.29) | −26.63* (−46.15 to −0.03) |
Antipsychotic drugs, active groupa | −20.33 (−52.64 to 34.06) | −14.79 (−49.36 to 43.35) | −21.00 (−53.05 to 32.89) | 1.13 (−39.90 to 70.14) |
. | January–March 2001 . | April–June 2001 . | July–September 2001 . | October–December 2001 . |
---|---|---|---|---|
BDZa | −17.48(−37.96 to 9.75) | −18.26 (−38.54 to 8.70) | −27.87* (−45.77 to −4.08) | −25.80* (−44.20 to −1.32) |
Medium- and long-acting BDZb | −17.31 (−39.31 to 12.66) | −17.18 (−39.21 to 12.84) | −28.27* (−47.36 to −2.29) | −26.63* (−46.15 to −0.03) |
Antipsychotic drugs, active groupa | −20.33 (−52.64 to 34.06) | −14.79 (−49.36 to 43.35) | −21.00 (−53.05 to 32.89) | 1.13 (−39.90 to 70.14) |
Benzodiazepines are drugs belonging to group N05BA + N05CD and antipsychotic drugs are those belonging to group N05A classified by therapeutic group based on the World Health Organisation Nordic Anatomical Therapeutic Chemical classification index (ATC) codes.
Medium- and long-acting benzodiazepines are alprazolam, nitrazepam, flunitrazepam and diazepam.
P < 0.05.
. | January–March 2001 . | April–June 2001 . | July–September 2001 . | October–December 2001 . |
---|---|---|---|---|
BDZa | −17.48(−37.96 to 9.75) | −18.26 (−38.54 to 8.70) | −27.87* (−45.77 to −4.08) | −25.80* (−44.20 to −1.32) |
Medium- and long-acting BDZb | −17.31 (−39.31 to 12.66) | −17.18 (−39.21 to 12.84) | −28.27* (−47.36 to −2.29) | −26.63* (−46.15 to −0.03) |
Antipsychotic drugs, active groupa | −20.33 (−52.64 to 34.06) | −14.79 (−49.36 to 43.35) | −21.00 (−53.05 to 32.89) | 1.13 (−39.90 to 70.14) |
. | January–March 2001 . | April–June 2001 . | July–September 2001 . | October–December 2001 . |
---|---|---|---|---|
BDZa | −17.48(−37.96 to 9.75) | −18.26 (−38.54 to 8.70) | −27.87* (−45.77 to −4.08) | −25.80* (−44.20 to −1.32) |
Medium- and long-acting BDZb | −17.31 (−39.31 to 12.66) | −17.18 (−39.21 to 12.84) | −28.27* (−47.36 to −2.29) | −26.63* (−46.15 to −0.03) |
Antipsychotic drugs, active groupa | −20.33 (−52.64 to 34.06) | −14.79 (−49.36 to 43.35) | −21.00 (−53.05 to 32.89) | 1.13 (−39.90 to 70.14) |
Benzodiazepines are drugs belonging to group N05BA + N05CD and antipsychotic drugs are those belonging to group N05A classified by therapeutic group based on the World Health Organisation Nordic Anatomical Therapeutic Chemical classification index (ATC) codes.
Medium- and long-acting benzodiazepines are alprazolam, nitrazepam, flunitrazepam and diazepam.
P < 0.05.
Opinions of GPs
The opinions of GPs on the educational programmes in this study were overall positive (Table 2). The participating physicians in general agreed with the statements that this education was relevant to their clinical work and that it improved their knowledge.
Discussion
We targeted our intervention to decrease prescribing of BDZ and antipsychotic drugs to elderly patients in general practice since these drugs could cause these patients many problems. For BDZ we succeeded. Since the use of BDZ is a risk factor for cognitive impairment and fractures, the reduction of these drugs among the elderly may have had positive health effects for these elderly patients. Our educational outreach visits did not, however, affect prescribing of antipsychotic drugs. This could be due to different reasons. Antipsychotic drugs are not prescribed by GPs in the same quantities as BDZ,12 therefore it is harder to detect any significant changes due to the educational programme. Since the number of prescriptions of antipsychotic drugs is low our study may not have had enough power to detect a difference in contrast to the prescribing of benzodiazepines. In Sweden patients often receive prescriptions for 1 year. Any effect of an intervention is not registered until a patient uses the new prescription. That means that any changes in the prescribing behaviour will not have an immediate effect on prescribing statistics. This is the reason for our measuring dispensed medications 1 year after the intervention. We intended to measure long-term (2 years) effects of educational outreach visits on GPs' prescribing habits. In Sweden there was a change in the way prescribing statistics were collected from the autumn of 2002. Therefore, we could not measure any long-term effects.
In this study there are rather few participating practices and they were randomised to either active or control group. They are all localized in the same districts and we have no reasons to believe that there are any differences between practices in active and control group that may have affected the results.
We have not examined the cost effectiveness of this education programme but as stated previously the use of BDZ is associated with great hospitalisation costs. The physicians in our study chose to participate in this study. A method like this can of course only succeed among physicians willing to participate in the education programme. Educational outreach visits cannot be forced on anyone.
The GPs were overall positive towards this kind of education. These results are comparable to those in an Australian study on educational programmes.6 As stated previously GPs in Sweden obtained most of their information on drug treatments from pharmaceutical companies. This might explain their positive attitude towards producer-independent education. We also believe that educational outreach visits can be a favoured method since they are convenient for the GPs. In Sweden the authorities and the pharmaceutical companies recently have agreed on stricter regulations on educational programmes offered by the pharmaceutical companies.
In order to attract all interested we offered all participants the same education. The control group in our study also participated in the education although after the study period.
After 2002 when the prescribing statistics (due to new EAN-codes on prescriptions) changed it is only possible to prescribe medium- and long-acting BDZ for 3 months and these drugs can only be prescribed by a physician using a special personal prescription. This is one way that the authorities in Sweden try to influence the prescribing of BDZ in the same direction as we intended with our educational outreach visits. Other educational programmes have been successful in modifying the prescribing of psychoactive drugs in nursing homes without adversely affecting the overall behaviour and level of functioning of the residents.23 Of educational outreach visits aimed at reducing BDZ some have24 succeeded whereas others have not.25 In this study we have shown that educational outreach visits are well appreciated by GPs and could influence their prescribing habits. This could thus be one way to reduce suffering and costs due to the use of BDZ by elderly patients. We do not know if our method is the best intervention. We welcome studies that compare the outcome of different educational methods.
The authors thank pharmacists Christer Luthman and Zoltan Nagy, Apoteket AB, for prescription statistics. We also want to thank the participating GPs for their participation. We thank the department of Primary Care Research and Development in the county of Skåne, Apoteket AB and the Faculty of Medicine, Lund University, for financial support.
References
Ekedahl A, Andersson SI, Hovelius B, Molstad S, Liedholm H, Melander A. Drug prescription attitudes and behaviour of general practitioners. Effects of a problem-oriented educational programme.
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.
Davis DA, Thomson MA, Oxman AD and Haynes RB. Changing physician performance. A systematic review of the effect of continuing medical education strategies.
Lowe CJ, Raynor DK, Purvis J, Farrin A, Hudson J. Effects of a medicine review and education programme for older people in general practice.
Reeve JF, Peterson GM, Rumble RH, Jaffrey R. Programme to improve the use of drugs in older people and involve general practitioners in community education.
Wahlstrom R, Diwan VK, Tomson G, Oké T, Beermann B. An educational experiment in primary care in Sweden.
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.
Borgiel AE, Williams JI, Davis DA et al. Evaluating the effectiveness of 2 educational interventions in family practice.
Jackson C, McGuire T, Dommers E, Nyst P. A GP prescribing educational intervention involving a medication panel.
Flaherty JH. Psychotherapeutic agents in older adults. Commonly prescribed and over- the-counter remedies: causes of confusion.
Lechin F, van der Dijs B, Benaim M. Benzodiazepines: tolerability in elderly patients.
Gray SL, Lai KV, Larson EB. Drug-induced cognition disorders in the elderly: incidence, prevention and management.
Larson EB, Kukull WA, Buchner D, Reifler BV. Adverse drug reactions associated with global cognitive impairment in elderly persons.
Foy A, O'Connell D, Henry D, Kelly J, Cocking S, Halliday J. Benzodiazepine use as a cause of cognitive impairment in elderly hospital inpatients.
Wang PS, Bohn RL, Glynn RJ, Mogun H, Avorn J. Hazardous benzodiazepine regimens in the elderly: effects of half-life, dosage, and duration on risk of hip fracture.
Panneman MJ, Goettsch WG, Kramarz P, Herings RM. The costs of benzodiazepine-associated hospital-treated fall Injuries in the EU: a Pharmo study.
Whitworth AB, Fleischhacker WW. Adverse effects of antipsychotic drugs.
Rosen J, Bohon S, Gershon S. Antipsychotics in the elderly.
Popli AP, Hegarty JD, Siegel AJ, Kando JC, Tohen M. Transfer of psychiatric inpatients to a general hospital due to adverse drug reactions.
Avorn J, Soumerai SB, Everitt DE et al. A randomized trial of a program to reduce the use of psychoactive drugs in nursing homes.
Ray WA, Blazer DGd, Schaffner W, Federspiel CF, Fink R. Reducing long-term diazepam prescribing in office practice. A controlled trial of educational visits.
Author notes
aTåbelund Primary Health Care Centre, Eslöv, bDepartment of Clinical Pharmacology, Lund University Hospital, Lund, cHospital Pharmacy, Lund University Hospital, Lund, and dDepartment of Medicine, Lund University Hospital, Lund, Sweden.