The European Journal of Public Health Advance Access originally published online on August 26, 2005
The European Journal of Public Health 2005 15(6):621-626; doi:10.1093/eurpub/cki053
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Health Services Research |
Repeat prescriptions: refill adherence in relation to patient and prescriber characteristics, reimbursement level and type of medication
Kristin Andersson1, Arne Melander1,2, Carin Svensson3, Owe Lind4 and J. Lars G. Nilsson1
1 NEPI Foundation, Stockholm, Sweden
2 Department of Community Medicine, Malmö University Hospital, Malmö, Sweden
3 Qulturum Pharmacy, Ryhov Hospital, Jönköping, Sweden
4 Jönköping County Council, Jönköping, Sweden
Correspondence: J. Lars G. Nilsson, PhD, NEPI Foundation, S-1881 Stockholm, Sweden, tel: +46 8 466 1357, fax: +46 8 466 1505, e-mail: lars.nilsson{at}nepi.net
Received May 14, 2004, accepted September 6, 2004
 |
Abstract |
|---|
 Top Abstract MethodsResultsDiscussionConclusionsReferences |
|---|
Background: Repeat prescribing used in long-term pharmacotherapy is often associated with inadequate patient medication, including non-adherence. In this paper we explore patients' drug refill adherence with repeat prescriptions and relate refill data to patient age and gender, type of prescriber, type of prescribed drug, and reimbursement level. Methods: During one week of 2002, copies of 3636 repeat prescriptions filled at 16 large Swedish pharmacies were collected. Satisfactory refill adherence was defined as dispensed refills covering 80–120% of the prescribed treatment time. Under- and oversupplying were defined as <80% and >120% coverage, respectively. Results: The average level of refill adherence was 57%, and the level of under- and oversupplying 21% and 22%, respectively. There was no gender difference. Patients who were exempt from payment had higher oversupplies than others (33% versus 19%), and patients of general practitioners had higher refill adherence than patients of hospital physicians. The highest refill adherence was observed for contraceptives (81%) and the lowest for anti-asthmatics, proton pump inhibitors and non-steroidal anti-inflammatory drugs (30–40%). Conclusions: Refill non-adherence includes both under- and oversupplying and may vary due to different attitudes between prescribers and between patients. Different therapeutic indications and reimbursement systems are other apparent causes. These observations should be considered in programs aiming to assist patients in following medication prescriptions.
Keywords: oversupply, refill adherence, repeat prescriptions, undersupply
Long-term pharmacotherapy is usually carried out using repeat prescriptions. This mode of prescribing has advantages and disadvantages. Drug refills without renewed consultation are convenient for patients and prescribers alike, but several problems may also ensue, as recently reported.1
Repeat prescribing might be associated with inadequate patient medication, including non-adherence.1,2 Medication reviews by pharmacists can reduce the drug-related problems associated with repeat prescriptions.1,3 Adherence to prescribed medication is important for patients' health, and a Cochrane review4 showed that patient outcome may be improved following interventions that promote adherence.
Activities to assist the patients in better following of medication prescriptions should preferably be based on knowledge of the level of non-adherence. However, measuring adherence is complicated and all available methods have disadvantages.5 Estimation of adherence based on refill of repeat prescriptions is a comparatively convenient method.6 Such data not only show how the patients have their drug supply refilled, they may also indicate the total therapeutic adherence, as shown by several authors.6–8 Refill adherence data compare favourably with adherence data obtained using alternative methods.6
In order to be able to design future effective interventions to improve adherence, we wished to estimate adherence among Swedish patients in relation to different influencing factors using refill adherence data. We have therefore explored how patients had their repeat prescriptions filled and have related refill data to patient age and gender, type of prescriber, type of prescribed drug, and reimbursement level.
|
Methods |
|---|
|
TopAbstractMethodsResultsDiscussionConclusionsReferences |
|---|
All Swedish pharmacies belong to one organization, Apoteket AB (the National Corporation of Swedish Pharmacies), presented elsewhere.9 The prescription identifies the drug, its strength, amount and dosage as well as the prescriber, and it is valid for 1 year from the date of issue.10 The prescriber must indicate on the prescription form the number of times the prescription may be dispensed. The same prescription form may thus be used for multiple dispensings without visiting the doctor. At the pharmacy, computer-generated information concerning date of dispensing and amount of drug dispensed is transferred to the prescription form each time the prescription is filled. In the present study the repeat prescriptions included were those that had been dispensed at least twice.
For long-term medication, prescriptions are commonly issued for 1 year of treatment, with 3-month dispensation intervals. Swedish physicians, dentists (included dental hygienists), nurses and midwifes have the right to prescribe. For the latter three groups, prescribing is restricted to certain types of drugs.
When a repeat prescription has been filled the maximum number of times prescribed, the prescription is retained at the pharmacy. The patient pays the full cost of drugs up to 900 SEK (approximately 
100 or US$125) during a 12-month period, and is exempt from further payment above 1800 SEK per year.
Data collection
The study was performed in the southern Swedish County of Jönköping, where drug utilization approximately equals the Swedish average.11 The county has a population of 328000 inhabitants, which is 4% of the Swedish population. It comprises 13 municipalities with 35 health-care centres, three hospitals and 33 pharmacies (information regarding the county of Jönköping is available at http://www.lj.se).
During one week of November 2002, repeat prescriptions that had been dispensed at least twice were collected at the 16 largest pharmacies in the county. Pharmacy staff were instructed to make photocopies of both sides of the prescription form and send the copies to the research group.
Collected data included patient age, gender, exemption from payment, drug name and ATC code, date of issue, type of prescriber and prescribed days of treatment. The dates of the dispensings and the amount of drug dispensed each time were also registered. From these data the actual number of days between each fill could be calculated, to be compared with prescribed days of treatment. The identities of the patients remained confidential.
Definition of refill adherence
Satisfactory refill adherence was defined as dispensed refills covering 80–120% of the prescribed treatment time.12
 |
|
Data analyses
The distribution of refill adherence was analysed in relation to patient age, gender and exemption from payment, and also in relation to type of prescriber and type of drug.
Statistical analyses
Z-tests were used to determine the statistical significance of the differences between the mean values of adherence related to patient and prescriber characteristics, type of drug and reimbursement level.
|
Results |
|---|
|
TopAbstractMethodsResultsDiscussionConclusionsReferences |
|---|
The 16 participating pharmacies dispensed 27 434 prescriptions during the week of data collection. The number of copies of repeat prescriptions included in the study was 3636, i.e. 13% of the total number of dispensed prescriptions of the participating pharmacies. The included prescriptions had generated 10844 pharmacy dispensations during the preceding 3–12 months that they had been used by the patients. The number of intervals between two separate dispensations was 7208. Ten days or more had elapsed between issue date and first fill of one-third (34%) of the prescriptions.
The refill adherence for different patient age groups is shown in table 1, and the distribution of the collected prescriptions in relation to the adherence levels is shown in figure 2. The result is an approximately bell-shaped curve. A total of 2058 (57%) prescriptions had a satisfactory refill adherence level, i.e. 100 ± 20% (table 1), 762 prescriptions <80% and 816 prescriptions >120%. This is an approximate ratio of 1:3:1.
|
|
The number of prescriptions increased with patient age up to 70–79 years and then declined. The highest adherence (66%) was seen among very old patients (
90 years of age). Among 40 to 89-year-old patients, 55–60% had satisfactory refill adherence. Adherence was lower among those aged 30–39 years (49%). The youngest patients (<30 years of age) had too few prescriptions to allow reliable evaluation of refill adherence. There were more prescriptions for female (60%) than for male (40%), but no gender difference in refill adherence were seen (57% versus 55%; P = 0.2). However, men had more oversupplies (25%) than women (21%; P < 0.001), and fewer undersupplies (20% versus 22%; P = 0.04).
More women (448; 56%) than men (358; 44%) were exempt from payment. Patients with full exemption had lower refill adherence (51%) than those without exemption (58%; P < 0.001). This was mostly due to oversupply. There were 33% oversupplies among patients with full exemption and only 19% among those without exemption (P < 0.001). Conversely, undersupply was lower (16%) in the exempt group than in those without exemption (22%; P < 0.001).
Patients with prescriptions from nurses and midwifes had the highest refill adherence (94%), whilst the lowest adherence was seen among those who had prescriptions from company-based physicians (44%) (table 2). Patients with prescriptions from general practitioners (GPs) had higher refill adherence than those with prescriptions from hospital physicians, and it was higher among those with prescriptions from private than from publicly employed GPs. Patients visiting chief physicians had higher refill adherence than those seeing other hospital physicians. Repeat prescriptions from dentists were too few to allow evaluation of refill adherence (n = 17).
|
The highest refill adherence in relation to type of medication was 81% (hormonal contraceptives for systemic use) and the lowest was 20% (anti-dementia drugs) (table 3). Contraceptives predominated among the former and were mostly prescribed by midwifes. Much lower refill adherence was recorded for inhalation anti-asthmatics (adrenergic anti-asthmatics 43%, others 34%), proton pump inhibitors (39%) and COX-2 inhibitors (23%).
|
|
Discussion |
|---|
|
TopAbstractMethodsResultsDiscussionConclusionsReferences |
|---|
Repeat prescriptions are intended for patients on long-term medication. The adherence among these patients has been estimated to 50% or less.13 Complete analysis of a patient's utilization of repeat prescriptions would necessitate comparison of prescriber records with the actual dispensations to the patients. We only had access to the latter. However, British and Swedish studies indicate that 10–20% of repeat prescriptions never reach a pharmacy.14,15 Accordingly, underutilization of repeat prescriptions is likely to be higher than appears from the current data on refill non-adherence, providing that the patients used only one prescription for the same drug at the same time.
The county of Jönköping was selected for the study since the level of prescribing in this county is close to that of Sweden as a whole. The month of November was chosen because the drug sales this month are close to one-twelfth of the sales for the whole year. We therefore consider that our sample of repeat prescriptions is representative for the country. All data were collected prospectively, i.e. we assessed not only repeat prescriptions dispensed the maximum number of times (which are the only repeat prescriptions stored at the pharmacies) but also those still open to more dispensations. This should allow a more appropriate measure of refill adherence than a study using only retrospective collection of dispensing data. Apoteket AB (see above) estimates that 40% of prescriptions dispensed at Swedish pharmacies are repeat prescriptions (at least one refill), whereas our sample comprised 13% of all prescriptions reaching the study pharmacies during the study week. This was due to lack of time at some pharmacies to copy all repeat prescriptions. However, the one-third of expected prescriptions that we obtained should be a large enough sample to permit analysis of refill adherence.
The repeat prescription forms with multiple dispensations noted can be seen as records of the patients' refill adherence behaviour for the last 3–12 months. The fact that the prescription copies were collected at the pharmacies during only one week is therefore of minor importance. The patient identities remained confidential, but it can be assumed that the repeat prescriptions represent separate individuals. Therefore, our data show the number of patients who have a satisfactory refill adherence as well as the number of those who undersupply or oversupply the system. Satisfactory medication adherence is commonly defined as one 80%.16 However, we chose the less common interval of 100 ± 20% as satisfactory adherence. It is much less common to indicate an upper level of refill adherence, probably because data on oversupplies are much more difficult to interpret than those on undersupplies. There are, however, reports12 where 100 ± 20% as been used to indicate of ‘good compliance’.
One-third of the prescriptions were initially dispensed only 10 days or longer after the date of issue. The corresponding patients may have had drugs available from older prescriptions, but it is also likely that some of these patients actually had therapy gaps and should be part of the refill non-adherence group, which would hence be larger than estimated.
As in previous studies,17,18 refill adherence tended to be lower in younger and higher in older patients. The gender ratio of repeat prescriptions (60/40 women/men) was similar to the Swedish average,11 and there was no gender difference in total refill adherence. However, men had higher oversupplies (25%) than women (21%) and women slightly more undersupplies (22%) than men (20%). While small, the latter difference would signify 
20 000 prescriptions in Jönköping county and 500 000 in the whole of Sweden.
Most probably, undersupply signifies gaps in patients' medicine supply, i.e. periods when the patients do not have medicine available and therefore have no possibility to be adherent. For the patients with oversupplies it is possible that they use too much of the prescribed medicines and in that in a sense they are non-adherent. However, the patients may also be stockpiling medicines, particularly if they are exempt from payment. The proportion of patients with oversupplies was almost twice as high among the patients with full exemption as among those who paid for their medicines (33% versus 19%).
Refill adherence was also dependent on the type of prescriber. As might be expected, a very high adherence was found for prescriptions from nurses and midwifes, especially for contraceptives. Prescriptions from private practitioners had a higher refill adherence than prescriptions from publicly employed GPs. Prescription from the former group were associated with higher oversupplies and fewer undersupplies (22% versus 17%) than were the prescriptions from the latter group (20% versus 22%). The seemingly small differences represent large prescription numbers; in the whole of Sweden, oversupplied prescriptions from private GPs would be 688 000 and undersupplies 532 000 in 2002. Hence the difference in undersupply and oversupply between private and publicly employed GPs may be clinically relevant.
Prescriptions from chief physicians had higher refill adherence than those from other hospital physicians. Also, these differences represent large number of prescriptions and seem clinically relevant. Reasonably, most chief physicians have longer clinical experience and see more selected patients than the other hospital physicians. All this could influence the refill adherence. The lowest adherence was found for dentists and company-based physicians, but the numbers of observations were few and hence the data are unreliable.
The largest differences in refill adherence were related to different types of medicines, ranging from 81% for contraceptives to 20% for anti-dementia drugs. Two more groups had refill adherence >70% (cardiac glycosides and thyroid hormones). It should be noted that table 3 only lists medicines that are used for long-term treatment. Several important groups of medicines, e.g. antibiotics, do not appear in the table.
Surprisingly, tricyclic antidepressants had a higher refill adherence than selective serotonin reuptake inhibitors, even though the latter are claimed to have fewer side effects, which should favour higher adherence.19 Another surprising finding was that there was higher refill adherence with selective beta-blockers than with, in turn, ACE inhibitors, loop diuretics, thiazides and AII blockers. Like the antidepressants, the AII blockers have fewer side-effects than other antihypertensives20 and a better adherence would be expected. The adherence may be influenced by the fact that the selective beta-blockers, ACE inhibitors and loop diuretics may also be given for some other, more symptom-prone indications beside hypertension.
Both non-selective non-steroidal anti-inflammatory drugs and coxibs had a low refill adherence. It is possible that this reflects the fact that pain conditions vary in intensity, making patients liable to vary their intake of these drugs. Anti-asthmatics for inhalation, both adrenergics and corticosteroids, also had low refill adherence; the former apparently being oversupplied and the latter undersupplied. This is in agreement with previous analysis.13 Another type of medication with low refill adherence and apparent undersupplies was proton pump inhibitors. This might be due to cessation of intake as soon as the stomach discomfort has been alleviated.
|
Conclusions |
|---|
|
TopAbstractMethodsResultsDiscussionConclusionsReferences |
|---|
Treatments based upon repeat prescriptions may seem practical but confer major risks of therapeutic failures, refill non-adherence being one reason. This includes both oversupply and undersupply, and may vary due to several factors, among them different attitudes between prescribers and between patients, in the latter case partly related to age and gender. Different therapeutic indications and reimbursement systems are other apparent causes. These observations should be considered in programs aiming to assist patients in following medication prescriptions.
Key points
|
|
Acknowledgments |
|---|
We wish to thank the pharmacies that provided us with the prescription copies.
|
References |
|---|
|
TopAbstractMethodsResultsDiscussionConclusionsReferences |
|---|
1 Zermansky AG, Lowe CJ, Petty DR, et al. Clinical medication review by a pharmacist of patients on repeat prescriptions in general practice: a randomized controlled trial. Health Tech Assess 2002;6:1–86.
2 Levy RA. Failure to refill prescriptions. In: Cramer JA, Spiker B, editors. Patient compliance in medical practice and clinical trials. New York: Raven Press, 1991:11–8.
3 Granås AG, Bates I. The effect of pharmaceutical review of repeat prescriptions in general practice. Int J Pharm Pract 1999;7:264–75.
4 Haynes RB, Montague P, Oliver T, et al. Interventions for helping patients to follow prescriptions for medications. (Cochrane Review). The Cochrane Library, Issue 3. Oxford: Update Software, 2001.
5 Nichol MB, Venturini F, Sung JCY. A critical evaluation of the methodology of the literature on medical compliance. Ann Pharmacother 1999;33:531–40.[Abstract]
6 Steiner JF, Prochazka AV. The assessment of refill compliance using pharmacy records: methods, validity and applications. J Clin Epidemiol 1997;50:105–16.[CrossRef][Web of Science][Medline]
7 Wei L, Wang J, Thompson P, et al. Adherence to statin treatment and readmission of patients after myocardial infarction: a six year follow-up study. Heart 2002;88:229–33.
8 Morningstar BA, Sketris IS, Kephart GC, Sclar DA. Variation in pharmacy prescription refill adherence measures by type of oral antihyperglycaemic drug therapy in seniors in Nova Scotia, Canada. J Clin Pharm Ther 2002;27:213–20.[CrossRef][Web of Science][Medline]
9 Westerlund T, Almarsdóttir AB, Melander A. Drug-related problems and pharmacy interventions in community practice. Int J Pharm Pract 1999;7:40–50.
10 Medical Products Agency. Uppsala: Sweden LVFS, 1997;10:37–8.
11 Swedish Statistics of Drugs 2000. Stockholm: Apoteket AB, 2001.
12 Al-Saffar N, Deshmukh A, Eid S, Carter P. Health beliefs and drug compliance of depressed patients in Kuwait. J Soc Adm Pharm 2003;20:142–50.
13 World Health Organization. Adherence to long-term therapies. Evidence for action. Geneva: World Health Organization, 2003.
14 Rashid A. Do patients cash prescriptions? BMJ 1982;284:24–6.
15 Nilsson JLG, Johansson H, Wennberg M. Large differences between prescribed and dispensed medicines could indicate undertreatment. Drug Inf J 1995;29:1243–6.
16 George CF, Peveler RC, Heiliger S, Thompson C. Compliance with tricyclic antidepressants: the value of four different methods of assessment. Br J Clin Pharm 2000;50:166–71.[CrossRef][Web of Science][Medline]
17 Rovelli M, Palmeri D, Vossler E, et al. Noncompliance in organ transplant recipients. Transplant Proc 1989;21:833–4.[Web of Science][Medline]
18 Leventhal EA, Crouch M. Are there differentials in perceptions of illness across the life-span? In: Petrie KJ, Weinman J, editors. Perceptions of health and illness. Amsterdam: Harwood Academic Publishers, 1997:77–102.
19 Swedish National Board of Health and Welfare. Current medications—changes and trends. Quarterly report 2000:1. Item: prescriptions of certain psychotropics for children and teenagers, Socialstyrelsen, Stockholm, 2001.
20 Law MR, Wald NJ, Morris JK, Jordan RE. Value of low dose combination treatment with blood pressure lowering drugs: analysis of 354 randomised trials. BMJ 2003;326:1427–34.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  S. A. Chambers, R. Raine, A. Rahman, and D. Isenberg Why do patients with systemic lupus erythematosus take or fail to take their prescribed medications? A qualitative study in a UK cohort Rheumatology, March 1, 2009; 48(3): 266 - 271. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. T. Westerlund and H T. Bjork Pharmaceutical Care in Community Pharmacies: Practice and Research in Sweden Ann. Pharmacother., June 1, 2006; 40(6): 1162 - 1169. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||