Family Practice Advance Access originally published online on November 30, 2006
Family Practice 2007 24(1):77-83; doi:10.1093/fampra/cml057
The effect of patient self-completion agenda forms on prescribing and adherence in general practice: a randomized controlled trial
a CAPER research practices, Halford Wing Dean Clarke House, Southernhay, Exeter EX1 1PQ, UK
b Academic Unit of Primary Health Care, Department of Community Based Medicine, University of Bristol The Grange, 1 Woodland Road, Bristol BS8 1AU, UK
c Institute of Health and Social Care Research, Peninsula Medical School Exeter EX1 2LU, UK
d Institute of Health and Community Studies, Bournemouth University Royal London House, Christchurch Road, Bournemouth BH1 3LT, UK
Correspondence to William Hamilton, general practitioner, Cotham House, Bristol, BS6 6JL, UK. Email: w.hamilton{at}bristol.ac.uk
Received 14 June 2006; Revised 2 October 2006; Accepted 17 October 2006.
 |
Abstract |
|---|
 Top Abstract IntroductionMethodResultsDiscussionDeclarationAppendix A. The SCAFReferences |
|---|
Background. What the patient wants from a general practice consultation and what the doctor believes they want are not always the same thing. This mismatch may lead to unwanted and unnecessary prescribing.
Aim. To study the effect of a one-page form completed by patients before their consultation and given to the doctor at the start of the consultation, in terms of a reduction in prescribing, satisfaction with the consultation and adherence with prescribed medication.
Design. Randomizedcontrolled trial.
Setting. Ten general practices in Devon and Dorset, UK.
Methods. Unselected patients attending general practice appointments were randomised to receive (or not) a self-completed agenda form (SCAF) asking five questions, including whether the patient considered they should receive a prescription.
Results. Approximately 4125 patients were offered entry; 3124 (76%) agreed to randomisation. In 1783 (57%) of these prescribing or satisfaction outcomes were identified. 457 of 811 (56.4%) of SCAF patients received a prescription, at a median (IQR) cost of £5.60 (inter-quartile range £2.12–£16.10), compared with 418 of 799 (52.3%) of controls, at a median cost of £5.94 (£2.46–£18.90); both results non-significant (P = 0.10 for prescribing and 0.30 for cost). Satisfaction was also similar in both groups: mean satisfaction score in SCAFs 5.37 and in controls 5.40 (P = 0.64), as was adherence: at 12 weeks, adherence of 75% or greater was reported by 92 of 136 SCAF patients (68%) and 105 of 145 controls (72%) (P = 0.31).
Conclusion. The negative result could have two explanations. Either the intervention did not achieve the intention of communicating the patients' agendas, or if it did, patients may have been persuaded by doctors' explanations in the consultation.
Keywords. Adherence, prescribing, primary health care, randomised-controlled trial.
|
Introduction |
|---|
|
TopAbstractIntroductionMethodResultsDiscussionDeclarationAppendix A. The SCAFReferences |
|---|
General practice prescribing accounts for half of the cost of family health services and one-tenth of NHS expenditure.1 In 2005, 720 million prescription items were dispensed in England, with on average 14.3 items per head of population.2 Prescriptions are written in two-thirds of general practice consultations, yet in a fifth of these the prescription is not indicated on medical grounds alone.3 In some consultations, there is a mismatch between the patient's expectation of receiving a prescription, and the doctor's estimate of the patient's wish for one, with GPs probably overestimating their patient's expectations of a prescription.4–8 Approximately 5% of prescriptions are not redeemed, perhaps reflecting such mismatching.9
If prescriptions are dispensed, adherence to prescribed medication also depends upon the patient's understanding of the need for their medication, and presumably also upon their desire to have received medication in the first place. Many interventions aimed at improving adherence have addressed the patient's knowledge of their medication, but with varying results.10
It is possible that improved doctor–patient communication could reduce unwanted prescriptions and improve adherence in those that were issued. Despite the promotion of patient participation, and some evidence of benefit, patients' agendas (such as their ideas about their illness, and their concerns and expectations) are often not voiced or elicited in the consultation.11,12 This deficiency could be addressed by using a pre-consultation questionnaire. Two trials have addressed this issue. The first was of a general leaflet asking patients to list issues for discussion and was associated with improved patient satisfaction but an increased number of investigations.13 There was no change in prescribing in that study, but it was designed with insufficient power to identify a modest, albeit important, change. The second study, published after this study was undertaken, was a trial of a single-page form asking patients to expand on their concerns before the consultation, and sought to identify additional agenda items.14 Although the form asked the patient if they would like a prescription, no changes in prescribing outcome were reported. Against this background, we performed this trial of a self-completed agenda form (SCAF), measuring its effects on prescribing, adherence and patient satisfaction. The specific hypotheses tested were: that the use of a SCAF would reduce both the overall number of prescriptions written and the cost of prescribed medication, and that it would increase both patient satisfaction and adherence to treatment.
|
Method |
|---|
|
TopAbstractIntroductionMethodResultsDiscussionDeclarationAppendix A. The SCAFReferences |
|---|
This was a single-blind, randomised controlled trial, with the intervention at the level of the patient. Ten general practices in Devon (9) and Dorset (1), UK, were recruited. These practices were selected to provide a mixture of characteristics, with rural and urban, deprived and wealthy, or young and old, populations.
Recruitment of participants
The study ran from June 2004 to June 2005, with practices participating for 4–6 weeks. All patients attending during normal working hours were eligible. On arrival at the surgery, patients were offered an envelope with a brief description of the study. If patients wished to proceed, they opened the envelope. This contained a covering letter, a short form to write down the patient's contact details, a pen, and in approximately half, a SCAF. The SCAF had been piloted in two surgeries and 
300 patients previously. Randomisation used a computerised random number sequence (WH). A research assistant was available throughout this process to answer queries, and to assist disabled patients. Carers were allowed to complete the SCAF if appropriate.
The SCAF was a one-sided sheet with five questions (see Appendix A) aimed at eliciting the patient's views on their condition, and what they expected from the consultation. The last question asked whether they thought they should receive a prescription. Patients allocated to a SCAF were asked to complete it whilst waiting for their appointment and to give it to the doctor on entry to the consulting room. The doctor was allowed to use the SCAF in any way they deemed appropriate for that consultation. Each doctor was allowed a maximum of eight study patients per surgery, though it was usually much less. The SCAFs were not retained or returned to the study team, as the consultation occurred before full written consent to access to the medical records was requested (see below).
Assessment of outcomes
A letter was sent to the patient within 24 hours of their consultation. This contained two questionnaires: the Medical Interview Satisfaction Scale,15 and the Satisfaction with Decision Questionnaire,16 and requested consent for the researchers to examine the GP records for prescriptions issued at the consultation, or at any re-attendance within the next week. Prescribing, re-attendance and adherence data were only collected on patients who had given consent for access to the medical records, or on the few who withheld access, but who had given prescribing information when returning their questionnaires. The researcher collecting the outcome data was blinded to the intervention status.
Prescription details and re-attendances were identified from the practices' computer systems. The cost of the medication was calculated at the full NHS price, taken from the contemporary issue of the British National Formulary. Prescribed items were categorised into short term (a medication taken for 14 days or less), or long term (for 15 days or more). To allow measurement of adherence, medications with a specific regimen were defined as drugs in which a regular pattern of use had been written on the prescription. A list of such drugs with a specified regimen was assembled independently by two of the study team (DR, WH) and agreed by consensus. All analgesics, ‘as-required’ medications and skin preparations were excluded because for these adherence was not relevant or was impossible to measure.
Adherence was measured by structured telephone interviews by a researcher blinded to the intervention status (CS), at two weeks (both long and short term) and at 12 weeks (long term only). Up to five telephone calls were made.
The 55 GPs participating in the study were offered a semi-structured telephone interview after participation. This was conducted by a researcher in Medicines Partnership (one of the study's funders) to allow any criticisms of either the SCAF or the study to be more easily aired. The interview focussed on whether the doctors believed the SCAF affected the consultation and their prescribing, including whether any such change in consultation style could have also occurred in control patients.
Analysis
The results were analysed by intention to treat. Adherence was calculated as the number of tablets taken, expressed as a percentage of the number prescribed to be taken. This was then grouped into three categories: below 25% adherence—to reflect those who either do not begin the medication or abandon it rapidly; adherence 25–74.9%—reflecting those who take a less than effective dose; and adherence of 75% or more—in which it is likely that a therapeutic concentration of the drug is maintained.17 Significance testing depended on the particular outcome measure and its distribution. Differences in the proportion of patients receiving a prescription, and of adherence groupings were tested by a chi-square test; in the cost of prescriptions by Mann–Whitney U-tests; and on satisfaction scores by t-tests. The data analysis used Stata, version 8.18
The sample size calculation was based on the primary outcome of the proportion of patients receiving a prescription. This was estimated to be 67% in controls, with a worthwhile reduction being to 60%. For 90% power, with a two-sided 5% alpha, 837 patients were required in each group to identify such a difference. The two-sided significance testing was chosen, lest our prescribing hypothesis—of reduced prescribing—be wrong, with the SCAF actually increasing prescribing.
|
Results |
|---|
|
TopAbstractIntroductionMethodResultsDiscussionDeclarationAppendix A. The SCAFReferences |
|---|
Entry of patients into the trial is summarised in Figure 1. Approximately 4125 patients were offered participation and 3124 (
76%) agreed to do so. As many as 1783 (57%) of these patients replied to the letters, though not all repliers completed both questionnaires and also gave permission to access the records. Those who either gave permission to access the records, or who described whether or not they received a prescription are shown in Figure 1 in the category, ‘prescribing outcomes known.’ These totalled 1610 patients. There were no statistically significant differences between the SCAF and control groups in terms of age or sex, at either randomisation or in those whose prescribing outcomes were known (data not shown).
|
Demographic details of the participants and the characteristics of the practices are shown in Table 1 and prescriptions with their costs in Table 2. Thirtythree participants who were randomised to complete a SCAF did not do so, usually because their doctor was available to be seen before the SCAF could be completed. Prescribing at the index appointment was non-significantly more frequent in those who had had a SCAF. Additionally, 43 participants received a prescription at a re-attendance within a week of the first consultation (SCAF 20, control 23), although 36 of these patients had also received a prescription at the first consultation (SCAF 17, control 19). The mean total costs of the prescription items were: SCAF £14.32 and control £15.35. Satisfaction scores are shown in Table 3. Fewer patients completed the ‘Satisfaction with Decision’ questionnaire, generally because they deemed it inapplicable to what had occurred at the consultation. No differences in satisfaction scores were observed between the SCAF and control groups.
|
|
|
Adherence to medication in those who had received a drug with a specified regimen is shown in Table 4. The overall mean adherence for short-term medication was SCAF 89% and control 85%; for long-term medication at 2 weeks it was SCAF 93% and control 95%; for long-term medication it was SCAF 74% and control 79%. No significant differences were observed between the two groups.
|
Only 29 of the 53 participating doctors completed the telephone interview. Eight doctors (28%) considered that the SCAF had affected their prescribing decision on at least one patient, and nine (31%) believed it had an effect on their consultation style, although any such effect was considered ‘slight’ and only related to patients who had actually received a SCAF.
|
Discussion |
|---|
|
TopAbstractIntroductionMethodResultsDiscussionDeclarationAppendix A. The SCAFReferences |
|---|
Main findings
The results of this study do not support the hypotheses being tested: in terms of the outcome measures, no differences were observed between those receiving and those not receiving the SCAF. The level of prescribing in the study consultations was, if anything, lower,19 and the level of satisfaction higher, than that observed in previous studies.13
Strengths and limitations
The strengths of this study include the successful recruitment of over 1600 patients from 10 general practices, making it one of the largest in this field. The randomisation was successful, and the outcomes were identified for similar numbers in both the intervention and control arms. However, we do not know what patients did with their SCAF and whether it achieved the intention of communicating the whole of the patient's agenda to the doctor. As we did not retain the SCAFs we cannot analyse what patients wrote on them or measure the proportion of patients explicitly stating that they wanted a prescription.
There was considerable attrition in the number of participants between initial entry to the study and assessment of the main outcome. This was inevitable with a study design based on initial verbal consent, followed by postal written consent. The pragmatic nature of this design was based on the future use of any positive result: we wished to mirror what could easily be implemented by practices, were SCAFs found to be of value. It is unlikely that attrition led to the negative result, as the age and sex characteristics of both the initial and the full consent groups were very similar for SCAF and control patients.
This was a trial randomised at the level of the individual patient. An alternative was to use cluster randomisation with the randomisation at the level of the practice. The main disadvantage of this alternative would have been the greatly increased sample size required. The advantage of cluster randomisation would have been to reduce or eliminate contamination. Contamination occurs when control patients unintentionally receive the intervention, either in part or in full. This trial could have experienced contamination at the levels of either the patient or the doctor. Patient contamination would have occurred if control patients saw intervention patients completing a SCAF, and decided to formulate a list of concerns themselves. We did not observe this behaviour, and it is very unlikely that these control patients would have extended their thinking to state explicitly at the consultation whether they wanted a prescription. Doctor contamination would have occurred if seeing a small number of patients with SCAFs would have changed the doctors' style so that they were more likely to enquire about desires for medication in control patients. This is why there was a restriction of the number of trial patients in any one surgery, although the maximum of eight (only half of whom would have completed a SCAF) was rarely reached, one or two being the norm. The doctors who completed the telephone interviews considered that any change to their style was minor, and restricted to those patients who had completed a SCAF. This latter point also fits with many studies showing that it is very difficult to change general practitioner behaviour, even with much more comprehensive interventions than merely seeing an occasional patient with a SCAF.20 Overall, we consider it unlikely that our negative finding can be explained by the absence of cluster randomisation.
Interpretation and comparison with previous literature
Various other interpretations are possible of these negative findings: those based on the assumption that the SCAF did not achieve its intention, and those based on the assumption that it did. Taking the first category, although previous studies have identified a wide range of agenda items during home interviews,21 patients may not expect doctors to be able to address all their concerns in a single consultation. They may not write them all down on the SCAF even when explicitly invited to do so. Patients may also be reluctant to write down items which could seem critical of doctors, such as having concerns about prescribed medication or wanting to discuss their own theories about the diagnosis.21 Other patients may prefer to leave decision-making to the doctor. Perhaps the doctors did not need the SCAF to judge whether patients wanted a prescription or not, but continued to use their usual methods.
If on the other hand, the SCAF did achieve its intention, it is possible that some patients who wanted a prescription before the consultation were persuaded if the doctor said they did not need one, and vice versa. In this situation, measuring patients' expectations before the consultation may not be a good way of identifying ‘unwanted’ prescriptions, although no research has yet explored if measuring patients' preferences afterwards provides a more accurate measure. Alternatively, decreased prescribing in patients who did not want a prescription may have been balanced by increased prescribing for those who did, leaving no overall effect. If the SCAF did provide patients with an opportunity to communicate their agenda to the doctor, it is surprising that this had no impact on patient satisfaction. It is possible that there is a ceiling effect for patient satisfaction and that the background level of satisfaction was so high that it could not be improved.
A comparable study found that leaflets encouraging patients to list issues they wanted to raise with the doctor increased patient satisfaction, although there was no change in prescribing.13 Unlike the leaflet in that study, the SCAF included a specific question about expectations for prescriptions. It would seem, therefore, that even though patients' expectations have been shown to influence GPs' prescribing,3,22 the invitation to make these expectations explicit does not change the level or costs of prescribing. In the study reported here, satisfaction scores were similar in both groups and similar to that seen in the intervention group for the leaflet study.13 An older trial of a SCAF showed that more items were raised during the consultation when a SCAF was used: no increase in satisfaction was seen, but prescribing outcomes were unreported.14
One unexpected finding was the very high level of self-reported adherence. A more usual figure for adherence is that approximately half of prescribed medication is taken.10 Our method of assessment was by telephone interview: it is possible that patients gave higher figures for adherence than had actually happened (either to please the researcher or to avoid possible researcher disapproval). Erroneous reporting would be hard to detect by telephone interview. However, over-reporting of adherence is very unlikely to have occurred predominately in either the SCAF or the control group, so it is unlikely to have affected the overall result. Furthermore, these were triply compliant patients. They had entered the study in the first place, had returned their questionnaires, and had agreed to telephone interviews. Such characteristics may well be associated with high levels of adherence to medication.
Nonetheless, the clear conclusion from this study is that use of a SCAF does not affect prescribing, satisfaction or adherence, either for better or for worse. It may be that these issues are too complex to respond to such a simple intervention.
|
Declaration |
|---|
|
TopAbstractIntroductionMethodResultsDiscussionDeclarationAppendix A. The SCAFReferences |
|---|
Funding. Project grants from the Medicines Partnership, and East Devon and Exeter Primary Care Trusts. Three NHS research practices hosted the study (Barnfield Hill and Mount Pleasant, Exeter; Cornwall Road, Dorchester). WH received funding from a NHS Researcher Development Award. The funding sources had no input into design, analysis or interpretation of the study.
Trial Registration. National Research Register number N0620137250.
Competing interests. None.
Ethical approval. North & East Devon Research Ethics Committee.
|
Appendix A. The SCAF |
|---|
|
TopAbstractIntroductionMethodResultsDiscussionDeclarationAppendix A. The SCAFReferences |
|---|
Self Completion Agenda Form.
- (1) What made you decide to come to see the doctor? Please describe the problem you have e.g. symptoms or current illness.
- (2) Your ideas about your illness: What do you think is wrong with you?
- (3) Your concerns: Have you any particular worries about your illness?
- (4) Your expectations: How do you think your problem should be treated? What do you hope the doctor will do?
- (5) Medication: Do you think you should receive a prescription for your problem?
- (2) Your ideas about your illness: What do you think is wrong with you?
|
Notes |
|---|
Hamilton W, Russell D, Stabb C, Seamark D, Campion-Smith C and Britten N. The effect of patient self completion agenda forms on prescribing and adherence in general practice: a randomized controlled trial. Family Practice 2007; 24 77–83.
|
References |
|---|
|
TopAbstractIntroductionMethodResultsDiscussionDeclarationAppendix A. The SCAFReferences |
|---|
1 Audit Commission. (1994) A prescription for improvement: towards more rational prescribing in general practice(HMSO, London).
2 Department of Health. (2006) Prescriptions dispensed, 2005(The information centre, London).
3 Britten N and Ukoumunne O. (1997) The influence of patients' hopes of receiving a prescription on doctors' perceptions and the decision to prescribe: a questionnaire survey. BMJ 315:1506–1510.
4 Virji A and Britten N. (1991) A study of the relationship between patients' attitudes and doctors' prescribing. Fam Pract 8:314–319.
5 Webb S and Lloyd M. (1994) Prescribing and referral in general practice: a study of patients' expectations and doctors' actions. Br J Gen Pract 44:165–169.[Web of Science][Medline]
6 Little P, Dorward M, Warner G, Stephens K, Senior J, Moore M. (2004) Importance of patient pressure and perceived pressure and perceived medical need for investigations, referral, and prescribing in primary care: nested observational study. BMJ 328:444–445.
7 Bradley CP. (1992) Factors which influence the decision whether or not to prescribe: the dilemma facing general practitioners. Br J Gen Pract 42:454–458.[Web of Science][Medline]
8 Stevenson FA, Greenfield SM, Jones M, Nayak A, Bradley CP. (1999) GPs' perceptions of patient influence on prescribing. Fam Pract 16:255–261.
9 Beardon PH, McGilchrist MM, McKendrick AD, McDevitt DG, MacDonald TM. (1993) Primary non-compliance with prescribed medication in primary care. BMJ 307:846–848.
10 Haynes R, Yao X, Degani A, Kripalani S, Garg A, McDonald H. (2005) Interventions to enhance medication adherence. Cochrane Database Syst Rev 19:CD000011.
11 Little P, Everitt H, Williamson I, et al. (2001) Observational study of effect of patient centredness and positive approach on outcomes of general practice consultations. BMJ 323:908–911.
12 Barry CA, Bradley CP, Britten N, Stevenson FA, Barber N. (2000) Patients' unvoiced agendas in general practice consultations: qualitative study. BMJ 320:1246–1250.
13 Little P, Dorward M, Warner G, et al. (2004) Randomised controlled trial of effect of leaflets to empower patients in consultations in primary care. BMJ 328:441–444.
14 Middleton JF, McKinley RK, Gillies CL. (2006) Effect of patient completed agenda forms and doctors' education about the agenda on the outcome of consultations: randomised controlled trial. BMJ 332:1238–1242.
15 Meakin R and Weinman J. (2002) The ‘Medical Interview Satisfaction Scale’ (MISS-21) adapted for British general practice. Fam Pract 19:257–263.
16 Holmes-Rovner M, Kroll J, Schmitt N, et al. (1996) Patient satisfaction with health care decisions: the satisfaction with decision scale. Med Decis Making 16:58–64.
17 Cook CL, Wade WE, Martin BC, Perri M III. (2005) Concordance among three self-reported measures of medication adherence and pharmacy refill records. J Am Pharm Assoc 45:151–159.
18 StataCorp. (2001) Stata Statistical Software: Release 8.0(Stata Corporation, College Station, TX).
19 Britten N, Jenkins L, Barber N, Bradley C, Stevenson F. (2003) Developing a measure for the appropriateness of prescribing in general practice. Qual Saf Health Care 12:246–250.
20 Bero LA, Grilli R, Grimshaw JM, Harvey E, Oxman AD, Thomson MA. (1998) Getting research findings into practice: closing the gap between research and practice: an overview of systematic reviews of interventions to promote the implementation of research findings. BMJ 317:465–468.
21 Barry CA, Bradley CP, Britten N, Stevenson FA, Barber N. (2000) Patients' unvoiced agendas in general practice consultations: qualitative study. BMJ 320:1246–1250.
22 Cockburn J and Pit S. (1997) Prescribing behaviour in clinical practice: patients' expectations and doctors' perceptions of patients' expectations—a questionnaire study. BMJ 315:520–523.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||