Abstract
Background
Introduction
The growing evidence for the benefit of treatment of elevated cholesterol levels, both in primary and secondary prevention of coronary heart disease (CHD), was consolidated during the mid 1990s following publication of the results of major intervention trials using statin therapy [1,2] Guidelines, incorporating the results of these trials and outlining comprehensive evidence-based practice have been compiled to help nurse practitioners and General Practitioners (GPs) improve the management of CHD [3]. However, effective assessment and management of CHD depends not only on health care professionals (HCPs) ability to implement the guidelines, but also on patients’ decisions about whether to follow the recommended treatment.
Despite statins being a relatively well tolerated class of drug [4], discontinuation rates are reported to be approximately 50% one year after commencement of therapy [4] decreasing substantially over time [5]. The reasons for poor compliance with therapy remain unclear because studies that have sought to identify the factors influencing statin compliance have produced conflicting results. One study showed that compliance was poorer with increased age, lower socio-economic status and in instances when more than 16 drug products per year were taken [6]. Other studies have suggested that poor compliance is-related to the cost of the drug [7]; patients’ preference for non-pharmacological methods; being unconvinced of the need for drug therapy [5] or recommendations for life-long treatment [4]; concomitant serious illness, and anxiety about potential [8] or experienced adverse events [4,5,7]. To date no work has been undertaken to explore patients’ understanding of hyperlipidaemia or compliance with statin therapy. As health care professionals, we need to acknowledge that compliance with therapy will be influenced by many factors and that some patients may attach less significance to hyperlipidaemia, or the need for statin therapy, than their health care advisor. It is clear that we cannot expect patients to blindly follow treatment advice. If we are to help patients make an informed choice about the extent to which they will act on the advice given, we need to understand and take account of patients’ perspectives, a view supported by the British Royal Pharmaceutical Society (RPSGB) [9].
The purpose of this study was to explore patients’ perspectives on statin therapy and compliance with their prescribed regimen.
Methods
This qualitative exploratory study was conducted using in-depth one to one interviews with patients who had been diagnosed with CHD, or as being at risk of CHD. Those who had commenced therapy less than three months prior to the review date were excluded from the study. The interviews took place between January 2001 and August 2001. The minimum time participants had been prescribed statin therapy at the point of interview was 8 months. The maximum length of time was 7 years and 9 months.
Definition and measurement of compliance
While acknowledging that use of the term ‘compliance’ is no longer considered appropriate by some [10,11], in this article it is used merely to describe the level to which an individual acts, consciously or unconsciously, upon their prescribed drug regimen.
For the purpose of determining level of compliance with statin therapy, an initial assumption was made that the prescription refill data available on electronic patient records (GPASS) was accurate and that, when a prescription was ordered, drugs were taken by the patients for whom they were prescribed. Each patient's level of compliance was determined by adding the number of tablets re-filled over a 12 month period (1st January 2000 through January 2001) and calculating this as a percentage of the number of tablets prescribed over the same period. As prescriptions were often re-filled on a two-monthly basis, data for one month immediately prior to or immediately following the stated review period were included when appropriate. Where patients had been prescribed statin therapy <12 months prior to their review date, compliance level was calculated using data from drug initiation date until review date.
We defined compliance according to the three groups used in the WOSCOPS study [12] (70–100%=good, 41–69%=moderate, and<41%=poor). This enabled us to identify potential informants with different experiences of medication taking. Compliance behaviour, as identified by the repeat prescription data, was ratified at interview by asking respondents directly for this information.
Sample selection
The target population was identified from a total patient population registered with a large primary care practice (n=5,324) in the West of Scotland, an area of high social deprivation as defined by the Carstairs Deprivation Index [13]. Consent for the conduct of the study was confirmed by Greater Glasgow Community/Primary Care Local Research Ethics Committee. A total of 252 patients aged ≥18 years who had been prescribed statin therapy for at least three months were identified. All had been diagnosed with existing CHD or as being at risk of CHD. According to the three compliance categories used in the study [12], 216 (85.7%) were identified as ‘good compliers’, 21 (8.3%) as ‘moderate compliers’, and 15 (6%) as ‘poor’ compliers.
Our aim was to interview patients who were able to provide information on the phenomenon under investigation from a variety of perspectives, that is patients from each of the three compliance groups [12]. As the number of patients identified as ‘moderate’ and ‘poor’ compliers was small (n=36), our initial intention was to recruit all of these patients plus 24 patients who were identified as ‘good’ compliers (a total of 60 patients). Recruitment was by letter and follow up telephone call. As we were unable to contact three of the moderate compliers by telephone, a purposive sample of 57 patients were invited to participate. Response rate in the ‘good complier’ group was higher (n=24; 71%) than that in the ‘moderate’ and (n=8; 44%) ‘poor’ complier group (n=15; 53%). We recruited 33 patients to the study, an adequate number for this in-depth, qualitative study where the characteristics of the sample met the needs of the study [14].
Data collection
Following ethics approval, in-depth one-to-one interviews were conducted by the first author in the respondents’ own homes. Prior to commencement of the interview, written consent for the interview to be taped, and for the research team to access medical records was provided by respondents. Each interview lasted approximately 1–1.5 h. A topic guide (see Table 1) was used to provide focus during the interview. Field notes were recorded immediately post interview to document additional information provided by respondents and to note circumstances that might have influenced the interview process.
Topic Guide
00028-8-table1.png)
Topic Guide
To ensure the rigour of the data collection and analysis process, findings from existing literature and theoretical models of health behaviour were used to aid development of the topic guide. As the interviews proceeded, the topic guide progressed to incorporate issues raised by respondents[15]. Throughout the interview, the interviewer re-phrased respondents’ comments back to them to ensure that their comments had been interpreted correctly. Following the interview, audio-tapes were reviewed, transcribed and checked for accuracy. Because the data was taped and transcribed we were able to examine it repeatedly and avoid the problem of selective recall [15].
On a small number of occasions (n=3) self-report compliance did not match that of prescription record. As the rationale for using a purposive sampling method was to identify different dimensions of compliance behaviour, and this discrepancy prohibited reliable identification of this dimension, we excluded these data. The themes identified were discussed and agreed with the second author. Rich quotes exemplifying these themes are presented in the findings.
Data analysis
The audio-taped interviews were transcribed and then analysed thematically [16] with the assistance of the qualitative analysis package QSR NUD. IST. This involved intensive reading and re-reading of the transcripts. Following on from this, the analysis process involved organisation of the data in line with the broad areas listed in the topic guide (Table 1). The data within and among interviews was then compared and contrasted [15], and common themes identified. As indicated in Table 2, six main themes were identified within two broad categories, i.e. ‘Patient–health care provider communication’ and ‘Health beliefs’.
Themes and sub-themes
00028-8-table2.png)
Themes and sub-themes
The following section, which presents the findings of the interview analysis, is discussed under the two main categories, Patient-health care provider communication, and Health Beliefs, and the six themes (Table 2) that emerged from the analysis. The demographic and compliance characteristics of respondents are presented in Table 3. As illustrated, most respondents were classified in social class six [13] reflecting the relatively high social deprivation within the study areas.
Demographic and compliance levels of respondents (n=33) according to GPASS data
00028-8-table3.png)
Demographic and compliance levels of respondents (n=33) according to GPASS data
Scoring system commonly used in Scotland to quantify levels of relative deprivation or affluence which is derived from participants postcodes. (McLone P. Carstairs scores for Scottish postcode sectors from the 1991 census data. Glasgow: Public health Research Unit, 1997).
Not known.
Not mutually exclusive.
Initiation of therapy
The importance of the circumstances in which statin therapy was initiated appeared to influence patients’ perceptions of hyperlipidaemia and statin therapy. As illustrated by respondents’ comments below, the credence patients’ attached to the prescriber and their perceptions of the primary purpose of the consultation visit at which the drug was initiated seemed important. [P7: female, moderate compliance] But if they (statins) are so important to somebody with angina I think that [hospital physician] would have recommended it… the hospital never …its not even my own doctor prescribed them…
In addition, patients attending a nurse-led clinic which, in their opinion, was for a routine health or diabetic check tended to view ‘cholesterol’ as less important than other health issues discussed during the visit. [P3: female, discontinued therapy] Well they've [medical practice] never asked me to go for that (cholesterol) itself… I've not to go for that particularly…
In the main, respondents’ who complied with their prescribed regimen spoke of how their nurse-practitioner or GP had emphasised the need for, or benefits of, drug therapy in relation to the limitations of dietary modification. In some instances the GP or clinic nurse had specifically pointed out the potential consequences of the disease. [P9: male; good compliance] I mean I've not seen any difference since taking them. I just took the pills because the doctor explained there's no symptoms with it… It's just the explanation the doctor gave… It's taught me how dangerous it was …you could wake up one morning with a stroke… And it's just you'd rather take these things (pills) just now rather than going into hospital later with it happening – some kind of stroke.
Such emphasis was less evident (but not always absent) in the accounts of patients who did not comply with their prescribed regimen. The information given to patients by HCPs, or patients’ perceptions of the usefulness of that information, varied. Some patients felt they had been given useful and adequate information. Others reported that they had acquired most of their information from the media. [P3: female; discontinued medication] Not really that much information… you more or less listen to it (on television). [P21: male; good compliance] I read in the paper that after a certain age your cholesterol levels off.
Opportunities to obtain information by asking HCPs directly for information were not always available to or utilised by patients. As illustrated below, some patients felt disinclined to ask questions while attending the medical centre, usually due to a concern that they would take up too much of the health care provider's time. [P18: female; poor compliance] Why should I go and annoy these people. I only go if I'm really not well. [P27: male; good compliance] I'm sure the doctors have got far more pressing things to do than to listen to some silly… coming in here saying ‘Why am I taking this and Why am I taking that’.
Others did not appear to know what questions to ask. However, on some occasions patients were merely informed by telephone that the drug had been prescribed and a prescription had been left for them to collect. Such circumstances provided little opportunity for questions.
Although, in the main, information contained in the drug insert was accessible in that it was considered easy to understand, reactions to its content varied. Some respondents appeared unconcerned about the content of the drug inserts, accepting that unavoidably ‘all drugs have side effects’. Others viewed the inserts as overly pessimistic, expressing preference for more positive information. [P15: female; moderate compliance]… it's all the negative things …I want to know about the good side if I've to keep taking these (pills)… [P24: female; good compliance] They just say ‘it [side effect] can happen’, but they don't—they really should emphasise the part where it can and that it doesn't always (happen).
This perceived bias towards negative information within drug inserts may negatively impact on compliance.
Subsequent feedback
Irrespective of medication-taking behaviour, when there was no feedback on individual progress, or there were long spells between appointments, patients rightly or wrongly assumed everything was satisfactory and that ‘they must be doing something right’. In addition, although patients expressed the opinion that they should be kept informed of their blood results and the level for which they were supposedly striving, as illustrated below, being provided with this information was not always helpful. [P12: male; good compliance] I think it was around approximately 5.7. I don't know what 5.7 means… [P7: female; moderate compliance] And I said to [practice nurse] ‘well I don't think six is all that bad’.
Without adequate supplementary information, patients are unable to use the information provided and, at times, use their own frame of reference in deciding the significance of their results. Similarly, unexplained changes to medication whether in relation to dosage, appearance, or packaging increased anxiety or led to patients ‘second-guessing’ as to why the change had occurred. [P7: female, moderate compliance] They keep changing… the packages-shape of… the tablets and because I don't take pills –I'm very wary of taking things that don't look the same as the ones before… [P3: female, discontinued therapy] Oh I was fine. I was getting on fine with that one not feeling sick, not feeling any side effects then all of a sudden I had to change. I said what am I changing for when I'm feeling fine, but I never – I (thought) they will probably be a penny cheaper to buy or something. That's what was in my mind.
Clearly, patients need to be informed about changes to their medication if compliance with therapy is to be encouraged. We cannot expect patients to blindly continue taking drugs for the rest of their lives merely because the drug has been prescribed. Particularly when they are unsure about what has been dispensed to them. While it may be clear to HCPs and pharmacists that characteristics, formulation and packaging of equivalent prescribed drugs may differ, this is not always evident to the recipients.
Although health care systems and processes are apparent to those working within the organisation they may be less obvious to patients. In this study, many patients had no clear idea of whether they were expected to initiate a follow-up appointment after commencement of statin therapy, whether subsequent appointments would be arranged for them, or whether they were necessary at all. Some deferred making a future appointment. [P24: female; good compliance] Probably if a got a letter in from the doctors saying ‘you should come down an get it checked’, I would go. But, because it's up to yourself you say'Och’ – you've got other things to do.
Such lack of definiteness with regard to follow-up may result in patients’ minimising the importance of hyperlipidaemia, particularly when other personal or health demands seem more pressing.
Sources of misconceptions
Patients’ misconceptions about the significance of hyperlipidaemia, potential CHD risk, and recommended treatment, were frequent and occurred at all stages in the patient–health care provider interaction regardless of the purpose of the contact. A number of respondents did not recognise, or understand, the association between hyperlipidaemia and existing symptoms such as angina or claudication. Or the need for long-term treatment. [P1: male, moderate compliance] I thought the cholesterol was just a wee thing… you know it just really happens once and that's it finished with. [P7: female, moderate compliance] …I felt that once you took it for a while your cholesterol would come down. You know like see if you weren't well and you got antibiotics and you took the tablet and you got better so… just – this cholesterol – how do you have to keep taking it (statin) all the time?
As illustrated above, this resulted in misunderstandings about the problem and the need for ongoing therapy. Unless specifically told otherwise, some patients commenced therapy assuming that at some point medication would be discontinued or that it could be taken intermittently. In addition, patients’ perceptions of statin therapy as insignificant (in comparison to other medications) were unwittingly confirmed or established when practitioners were inattentive to what they were being told. Or when comments were taken out of context and/or misconstrued by the listener. [P3: female, discontinued therapy] Oh he never [wrote down] – he said it wasn't important… I said to [GP] ‘I don't take my cholesterol pills’, and he said ‘well that's not really life… threatening or anything’ because he was running out of space (on the benefits form).
Thus, it seems that when HCPs focus on a specific task during their conversation with patients, rather than actively listen to what is being said they may be misinterpreted and subsequently miss opportunities to address non-compliance issues.
Health beliefs
Unconditional acceptance
Respondents who took their medication regularly appeared to have a high regard for their health care practitioner or for advances in diagnosis and treatment of disease. They believed drugs would not be prescribed unnecessarily and took their medication because their doctor ‘told me to take it’. [P8: male; good compliance]… I leave it up to the doctor to decide… I put my life in their hands – I mean, we're in a position we have to rely on our doctor. It's a simple fact ‘cause we're no smart enough. [P26: female; good compliance] There would be no reason for them to give me them if they thought I didn't need them… and that's why we have professional people you know; they've studied all these things, they know actually what they're doing.
The esteem in which these patients held their GP suggests that they were happy for decisions to be taken by the doctor. As such, the treatment recommended was not questioned. Others who took their medication regularly and without question had often been prescribed therapy after experiencing a sudden and unexpected health crisis or had visible symptoms such as xanthelasma (plaque deposits on the surface of the skin, most often found on the inner eyelid). [P4: female, good compliance] This is what the nurse was explaining this is how its came out (xanthelasma). Its sort of popped out you know…
Although these cues to action [17] may have impacted on some respondents’ perceived susceptibility this was not always the case. Despite hospitalisation for CHD and persistent symptoms, some patients still chose not to continue with statin therapy. Therefore, it is clear that factors other than those described above impact on compliance. In this study, compliant patients were more convinced of the benefits of drug therapy as a means to prevent future problems than non-compliers. Some believed that taking statins ‘allowed’ a certain amount of freedom from dietary restrictions. [P6: female; good compliance] [The doctor] reckoned that it was a hereditary thing and he didn't think that anything in my diet was going to help… Well there is nothing much that I can do other than take this pill… [P15: male; good compliance] Well the tablets easiest you know, that's the way the specialist put it – it's easier to pop a pill than sit and worry about whether to eat a bacon sandwich.
For patients who struggled to restrict their diet, or those who believe that restricting their diet made little difference, taking medication was considered an easier option.
Conditional acceptance
Unlike ‘good compliers’, ‘moderate compliers were more likely to question the rationale behind drug therapy. The decision not to take their medicine as prescribed seemed to be based on personal health beliefs and the conclusion they came to regarding the efficacy of the drug compared to that of other strategies. [P5: female; moderate compliance] I'm convinced it's up (cholesterol) ‘cause the doctor tells me its up. But I think myself if I stop eating certain foods that my cholesterol will come down… I feel sometimes people just say ‘Well that's what you've got so there's this tablet’.
When unconvinced about the need for their prescribed statin regimen, patients tended to look for and weigh up evidence, such as family history, symptoms experienced, or the experience of a friend or colleague, which would confirm or refute the diagnosis. However, even when confirmed, increased risk seemed to be acknowledged as possible rather than probable and the drug was assessed by ‘testing’ it for positive (did it make any noticeable difference) and negative (did it cause any problems) effects. [P7: female; moderate compliance] My dad had angina… So I think maybe the cholesterol thing is on my side. I'm thinking myself that the cholesterol is associated with the angina but then when I look at [my friend] (prescribed lipid-modifying drugs) she hasn't got angina… so I mean I don't know. [P5: female; moderate compliance] I always found that well … when my cholesterol was high, my veins were all sticking out …And there's nothing like that right now.
At times, inner conflict resulted when the treatment recommended was contrary to the respondent's own solution. Resolution of the conflict seemed to be sought through partial compliance with the drug being taken only some of the time. [P5:female;moderate compliance] … the doctors trying to help me and I'm laxing with it… well that's the only reason I take it. Otherwise I really wouldn't take it at all… But that guilt there you know…
It was factors such as these rather than experienced side effects, or difficulties in incorporating medication-taking into daily routine, which adversely affected compliance. And although side effects, or planned alcohol consumption were at times proposed by patients as a reason for modifying their prescription, they frequently acknowledged that they probably used these reasons merely to justify their medication taking behaviour. This was substantiated when, as illustrated below, alternative reasons for experienced side effects were offered, and further discussion revealed that, unless forgotten, drugs perceived as necessary for symptom relief were taken regardless of alcohol consumption. [P3: female; discontinued medication] That could have been just nerves that was making me sick (commenting on previously reported side effect) it could have been anything.
Deferment and rejection
The distinction between respondents who took their medication some of the time and those who discontinued it altogether was blurred. Both groups considered drugs as ‘curative’ rather than ‘preventative’. Similarly, both groups seemed intolerant of individuals who they considered to be ‘neurotic’ about cholesterol. [P7: female; moderate compliance] All they [friends] can talk about is this cholesterol, cholesterol. And it just sort of seems to go through me like that… And I say to myself ‘that isn't right, that isn't right at all’.
However, the main issues for respondents who had discontinued their medication altogether was absence of any symptoms that they believed could be attributed to raised cholesterol and perceived ‘wellness’. Consequently, they saw no reason to take the prescribed drug. Their decision to discontinue treatment without seeking medical advice seemed to be based on their perceived ‘healthy’ status. [P19: male, discontinued therapy] …to be quite honest with you I don't even know if I should be taking any of them. I feel all-right …I feel fine.
Patients who had discontinued statin therapy appeared to be more sceptical of medicine in general. Thus, they chose to reject the diagnosis or defer action. In addition, changes in the advice given or treatments on offer resulted in cholesterol being perceived as a ‘new fad’ or doubts about treatment recommendations. [P20: male; discontinued statin] whether it was my blood count or cholesterol [the nurse] actually said to me because of the lowering (recommended cholesterol levels), GPs are being advised to put (more) people on the tablets… And I thought well things change by the week you know, next week it might be something else. [P19: male; discontinued statin]… it's just the style and all that low fat milk, you know what I mean, it's… just going round in a circle. Maybe next year it'll all be back in fashion again…
The illustrations above highlight the views ‘poor compliers’ seem to hold about health recommendations and what they considered to be its inconstant or cyclical nature.
Discussion
Summary of main findings
The findings of this study highlight the variation in patients’ beliefs about cholesterol and the effectiveness of statin therapy. These beliefs are likely to have influenced the way in which health advice was interpreted and the extent to which it was adopted. In this study, compliant patients were more likely to report that the consequences of hyperlipidaemia and the need for drug therapy had been emphasised by a HCP than were patients who did not comply with their prescribed regimen. This supports the results of a previous study [18], which showed that HCPs could influence patients’ perceptions of the significance of a disease and the need for treatment.
Interactions between patients and HCPs may result in a positive outcome, when for example a HCP emphasises the potential benefits of the recommended treatment. However, if HCPs inadvertently misinterpret patients’ questions or attempts to seek information without directly asking questions, more negative outcomes may ensue. Patients’ decisions about acting on treatment advice may then be based on inadequate or erroneous information. In addition, when practitioners focus solely on the issues that they consider relevant at the time of consultation, without taking account of those patients’ perceive as important, patients may assume that the issues not discussed during consultation, are of lesser significance.
Previous studies have shown that some patients have personal and unique beliefs about their medical condition [19] and that some are doubtful about the need for drug therapy [5]. Yet decisions about whether to comply with a prescribed drug regimen are not always expressed. These results are supported by the findings of this study where a number of respondents expressed a doubt about the need for statin medication. Unknown to their health care provider, these patients partially substituted medications for non-pharmacological treatments or rejected them altogether. Furthermore, it should be noted that while the need to emphasise the significance of hyperlipidaemia may be particularly pertinent in the absence of symptoms, it is clear from respondents’ accounts that even when evidence of personal risk was present, some patients actively choose not to follow their prescribed drug regimen. Unless HCPs are alert to patients’ attempts to communicate personal concerns [20], and actively encourage patients’ to discuss their personal beliefs about their diagnosis and advised treatment, factors that could impact on an individuals’ compliance behaviour may not be identified.
Strengths and limitations
Before concluding, it is important to acknowledge the limitations of this small-scale study.
Firstly, although respondents raised a number of issues regarding poor information provision, it is possible that they had forgotten or ‘selectively’ remembered what they were told. Secondly, the number of participants identified in this study as ‘moderate’ or ‘poor’ compliers was less than that of ‘good’ compliers. However, our aim was to recruit patients who were able to provide information on the phenomenon under investigation from a variety of perspectives. As intended, we were successful in identifying patients in the three compliance groups defined in the study methods.
Implications for practice
Despite its limitations, we believe that the findings of this study provide valuable information that could aid General Practitioners, Practice Nurses and Specialist Nurses whose work involves the ongoing care of patients at risk of CHD or with established CHD. Health care practitioners who acknowledge the existence of non-compliance, and accept, without judgement, patients personal beliefs about their advised treatment will be better placed to address the relevant issues and thus help patients make informed treatment decisions. It is clear that patients sometimes misinterpret the information they are given. However, at times, HCPs misread situations and subsequently inadvertently mislead patients with regard to the significance of hyperlipidaemia or need for continued therapy. Practitioners who have a better understanding of patients’ perspectives will be better equipped to ensure that the information they provide to patients is adequately communicated and understood and that appropriate, relevant and acceptable compliance improvement strategies are utilised.
Footnotes
Acknowledgements
The authors will like to extend sincere thanks to the study participants. They would also like to thank the secretarial and practice staff who provided support for the study and assisted with the initial identification of the target population. Source of Funding