Physician perspectives on statin continuation and discontinuation in older adults in primary cardiovascular prevention: a qualitative methods study

Abstract

Objectives

In the context of limited evidence on statin use in primary cardiovascular prevention in older adults, we assessed physician perspectives on decision-making about statin continuation or discontinuation in this population.

Design

Qualitative descriptive approach including four focus groups. Inductive and deductive thematic analysis.

Setting and participants

18 physicians including two neurologists, three cardiologists, seven hospital internists and six primary care providers (PCPs) recruited at a hospital and primary care practices in the area of Bern in Switzerland.

Results

Concerning knowledge about statins in older adults, physicians reported defining if a patient is treated for primary or secondary prevention as challenging and that lack of evidence makes the decision to continue or discontinue the statin difficult. In terms of beliefs, fear of a possible rebound effect after statin discontinuation was reported. Regarding decision-making, physicians mentioned that statin discontinuation or continuation should be a shared decision between the patient and the physician. Concerning the professional role, environmental context and resources, the PCP office was identified as the ideal setting to discuss discontinuation, as all necessary information is available and PCPs have a longer relationship with the patients, thus facilitating a shared decision. Discontinuation of a chronic medication was perceived as difficult in general. Furthermore, PCPs noticed a possible negative impact on patient–physician relationship as some patients felt not being worth it, given up or undertreated if the statin was discontinued.

Conclusions

This study highlights the challenges of statin continuation and discontinuation in older patients and the crucial role of PCPs in situations with unclear evidence for a medication, where shared decision-making between physicians and patients is important. More evidence forming the background for a decision aid would be helpful.

STRENGTHS AND LIMITATIONS OF THIS STUDY.

• Inclusion of physicians from different specialties and different settings allowed a more comprehensive overview.

• Physicians both from the outpatient and hospital settings were considered allowing to assess the possible differences arising from this environmental context.

• A notable limitation of our study lies in the small number of participants and further the exclusive recruitment of specialists and internists from only a singular hospital; however, we used an explorative approach and were not aiming to reach theoretical saturation due to the low availability of physicians during COVID-19 pandemic.

• Small number of participants in each focus group seemed to improve the quality and depth of answers by allowing every participant more time to express his views.

Introduction

Statins are one of the most frequently prescribed chronic medications in older adults.^{1 2} While there is strong evidence for their benefit in secondary cardiovascular prevention even for older adults, the evidence in primary cardiovascular prevention in this population is limited.3,6 Still, new statin prescriptions for primary cardiovascular prevention are more frequent in older than younger adults.⁷ Statins are also associated with several potentially serious adverse drug events that can impact the quality of life, such as muscle pain, liver dysfunction, renal failure or gastrointestinal symptoms.^{8 9}

When evidence is unclear for the potential benefits of a prophylactic medication, continuation or discontinuation of this therapy should be discussed in a shared decision-making approach between the patient and the treating physician. Understanding patient and prescriber perspectives can help to better implement this approach in patient consultations. Prior studies assessed some aspects of physician perspectives on this topic, including main reasons for discontinuation and preferred conversational approach for statin discontinuation.10,12 However, a comprehensive evaluation including the comparison of perspectives of different medical specialists and identification of barriers and facilitators is missing.

With this study, we thus aimed to explore the perspectives of hospital internists, cardiologists, neurologists and primary care providers (PCPs) on statin continuation and discontinuation for primary cardiovascular prevention in older adults.

Methods

This article is written according to the COREQ (Consolidated criteria for reporting qualitative research) checklist.¹³

Study design

A qualitative descriptive approach with focus groups (FGs) was chosen to gather extensive information about physician perspectives on statin continuation and discontinuation in older adults in primary cardiovascular prevention.¹⁴ Ethical approval was waived by the Human Research Ethics Committee of the Canton of Bern, Switzerland (Req-2020–03065), as the study did not fall under the Human Research Act according to the Swiss regulation.

Setting, participants and recruitment

We selected physicians who are most involved in statin prescribing, including hospital internists, cardiologists, neurologists and PCPs. Hospital internists, cardiologists and neurologists were recruited at the Bern University Hospital (Inselspital). PCPs were recruited at the Bern Institute of Primary Healthcare. Invitations to participate in the study were sent individually by email by the senior author. Participation was voluntary and participants provided consent for data collection and recording of the FGs. They were informed that their name would not appear anywhere.

As FGs took place during the COVID-19 pandemic and the availability of physicians was limited, we did not aim to reach theoretical saturation but to involve all four kinds of physicians. We thus planned two FGs with PCPs, one with hospital internists and one with cardiologists and neurologists. We expected that this purposeful sampling would give a good overview of the different physician perspectives.¹⁵ We decided to conduct a higher number of FGs with PCPs than with other specialists because of their greater implication in decision-making regarding prescription and discontinuation of (primary) preventive medication. We aimed to include four to seven physicians in each FG to ensure sufficient speaking time for every participant.

Data collection

FGs were conducted by CEA (female, senior physician, MD, MSc), while LB (female, medical resident, MD) took field notes. Both were trained on qualitative data collection and analysis. FGs were conducted virtually, recorded and transcribed verbatim. A duration of 45–60 min was planned. Physicians were compensated financially for their involvement (100 Swiss francs). FGs were conducted following a guide, provided in online supplemental material 1, created by CEA and LB based on the previous literature,^{14 16} and had the following structure: (1) a short presentation of the interviewers and explanation of the study goal, (2) past experiences of physicians with statin discontinuation and (3) information wished by physicians to help them decide to continue or discontinue a statin with a patient. Open-ended questions were followed by prompts to elicit exchange between participants on topics of interest. The FG guide was presented for feedback and improvement to experts in qualitative research and deprescribing guidelines. The FG guide was slightly modified after each FG in an iterative process to explore further insights.¹⁷ To minimise the risk of bias during the interview conduction, the main investigator carefully followed the guide and cared not to speak for or against statin prescription.

Data analysis

Data analysis was performed using a mixed deductive and inductive approach.¹⁸ The deductive approach was based on a systematic review that we have previously published.¹⁶ Thematic analysis was chosen to identify common themes among FGs.¹⁹ With the inductive approach, we hoped to ensure the dynamic detection of possible new, not previously assumed themes, during the analysis. CEA and LB coded FGs independently to increase reliability in coding. Subsequently, CEA and LB discussed and adapted codes until they reached consensus. Codes were grouped in subthemes and themes. Interpretation and presentation of the results were carried out by AM (male, chief resident), SZ (male, medical resident) CEA and LB. Quotes presented in this article were translated into English by AM and LB and reviewed by CEA. Coding was conducted using MAXQDA2020 (Max Weber qualitative data analysis) software for qualitative analysis (VERBI Software, Berlin, Germany).

Patient and public involvement

None.

Results

Of 47 physicians contacted, 26 agreed to participate, and 18 finally participated (eight could not participate because of unavailability at the time of the FGs). Reasons for declining participation were lack of time or interest and lack of answer after three reminders. We conducted a total of four FGs between March and April 2021: FG1 with specialists (ie, cardiologists and neurologists), FG2 with hospital internists, FG3 and FG4 with PCPs. Baseline characteristics of participants are presented in table 1.

Table 1. Sociodemographic characteristics of physicians.

Characteristics	Total (n=18)
Age (years), median (IQR*)	40 (34–46)
Gender
Female, n (%)	9 (50%)
Specialisation, n (%)
Cardiologist	3 (17%)
Neurologist	2 (11%)
Hospital internist	7 (39%)
Primary care provider	6 (33%)
Function, n (%)
Resident	4 (33%)
Chief resident	6 (50%)
Attending physician	2 (17%)

Function only presented for hospitalists (n=12).

^*IQR, interquartile range. Function only presented for hospitalists ().

We identified five themes (1) knowledge, (2) beliefs (3) decision-making, (4) environmental contexts and resources and (5) social/professional role and identity. Each theme included two to three subthemes with corresponding codes. Figure 1 displays the themes and subthemes. We use the following abbreviations to report participant quotes: FG, focus group; S, specialist; I, internist and PCP, primary care provider.

Figure 1. Themes and subthemes.

Knowledge

This theme included physician factual knowledge about cardiovascular prevention and statin medication. Concerning disease-related knowledge, the differentiation between primary and secondary prevention was reported throughout all FGs as challenging (FG4PCP4): “I think it’s difficult to draw the line between primary and secondary prevention… particularly with older patients”. In terms of knowledge about medication, the possible adverse effects of statins seemed well known. Lack of evidence in specific guidelines for both continuation and discontinuation of primary preventive statin therapy in older adults was mentioned as a major issue (FG1S1): “I think we have zero evidence for doing that [discontinue the statin]. In my opinion, we have weak evidence with the study of Giral [referring to²⁰] that this might even be medical malpractice.” (FG2I3): “Because there are certainly no studies that show they still benefit from the statin at that age.” Uncertainty was reported about possible rebound effects after statin discontinuation. Furthermore, a general lack of knowledge about the duration of cholesterol plaque stabilising properties of statins was reported, making it unclear, how long and until what age patients might benefit from statin therapy (FG2I1): “If a statin is used for plaque stabilisation, in both primary and secondary prevention, then I don't know how long that actually makes sense. Maybe there is a high-risk period, and then you can say, ok, the plaques are stabilised.”

Beliefs

This theme presents physician beliefs about the continuation and discontinuation of statins in primary prevention, given the limited evidence in older adults and their corresponding outcome expectancies. Specialists believed there had to have been an indication for the statin at some time point and that age should not influence this indication (FG1S1): “So, there must have been a guideline-based indication at some point. And it does not disappear when the patient is 75.” Specialists also made the assumption that cardiovascular events could have been prevented by the ongoing therapy with a statin (FG1S1): “Maybe, they [the patients] are still in primary prevention because they are taking a statin”.

Some internists thought primary prevention could differ in usefulness depending on biological age and health status (FG2I7): “I think you should not only consider the chronological age, but also the biological age, and also how fit the patient is. A very fit patient would probably benefit more [from a primary preventive statin] than a patient with a lot of comorbidities and dementia”. Discussing the outcome expectancies, positive consequences of discontinuation were highlighted especially by internists and PCPs, and included the reduction of polypharmacy and relief of adverse drug events, improving the quality of life. Some uncertainty about possible negative consequences after discontinuation was reported, which seemed to negatively influence the decision to discontinue primary prevention in older patients. Especially, fear of a possible rebound effect, that is, a cardiovascular event occurring after statin discontinuation, was reported (FG3I5): “…then there is also the fear that if I take that away [the statin], and then a little later he [the patient] has a cardiovascular event”. (FG4PCP1): “I still have a bit of a bad feeling in the back of my mind, have I then served the patient by stopping [the statin]. I see him [the patient] 10 years later and he has vascular dementia… yes that would be an additional argument against discontinuation”. Yet, some internists believed that this fear of a possible rebound was unfounded (FG2I5): “I think the fear of a rebound is rather big, probably unjustified”.

Decision-making

This theme included three subthemes: physician experiences with decision-making, motivation for statin discontinuation and the stopping process itself.

PCPs reported good experiences with discontinuation, especially when they thoroughly explained the reasons to their patients (FG4PCP3): “So, I discussed it with her [the patient], and also explained it, and she was happy that she then had at least one medication less”. Specialists highlighted that they usually did not see patients in primary prevention, and were therefore seldom confronted with decision-making about discontinuation in this situation. Nevertheless, some specialists mentioned still being confronted with decision-making in primary prevention when additional exams infirmed a suspicion of increased cardiovascular risk (eg, no atherosclerosis in the coronary angiography).

Bad experiences with statin discontinuation included making patients feel being not worth it or abandoned when the physician suggested to discontinue (FG4PCP2): “…patients feel that the doctor gives up on them when we stop medications, and I think it is quite important that the patient does not have that feeling, because it would be a barrier”. Furthermore, PCPs mentioned that suggesting statin discontinuation offended some patients who felt undertreated (FG3PCP1): “And now she [the patient] has a disease that I don't treat”. Other aspects mentioned were confusion and insecurity experienced by patients when the discontinuation of a long-term medication like a statin was suggested. Some PCPs experienced negative effects on the therapeutic relationship after discontinuation (FG4PCP1): “When I took over the practice, my predecessor had prescribed statin in primary prevention a lot. When I stopped the medication [the statin], I upset many [patients], and they started to have palpitations or to be afraid of me … When I started it again, everything was better again”.

Regarding motivation, reducing polypharmacy was perceived mainly by PCPs and internists as an important reason to discuss statin discontinuation with an older patient in primary prevention. The occurrence of severe adverse drug events was reported as a reason for statin discontinuation by all groups of physicians (FG4PCP3): “I think it’s a good basis to discuss discontinuation of a statin when patients are experiencing adverse drug events”. Physicians felt that it was common sense to discuss discontinuation when patients are very old and frail, or in case of cancer or short life expectancy (FG1S1): “From a cardiovascular prevention point of view, I see no comprehensive reason for discontinuing a statin apart from the obvious ones: a patient has severe muscle pain, extreme frailty or polymorbidity, short life expectancy and cancer.” A further motivational factor for discontinuation reported by PCPs was to favour quality of life (FG3PCP2): “…they no longer want to live as long as possible, they want to live as well as possible”. On the other hand, PCPs and internists reported they would be more likely to continue primary prevention if an older patient was fit, had few comorbidities and a younger biological age.

The subtheme stopping process presents physicians perspectives on the process of statin discontinuation itself. Discontinuing a chronic medication was perceived as difficult by internists, PCPs and specialists in general. Nevertheless, especially PCPs mentioned reviewing medication as important to identify potentially inappropriate medications or adapt them to patient individual goals of care. PCPs suggested mentioning polypharmacy to facilitate discussion about statin discontinuation (FG3PCP2): “Then I said: “Yes, you have quite a lot of medications. Do you find it difficult?“, and that was a good starting point”. PCPs mentioned exploring lifestyle change possibilities with their patients to support the discontinuation process. Monitoring cholesterol levels after implementing lifestyle changes while pausing the statin was therefore suggested as a possible way to support the decision about permanent discontinuation and make patients feel safer. Patient participation in decision-making was perceived as central by all physician groups (FG2I3): “The key point is that the patients share the decision. Because they also bear the risk”. Thus, the importance of good patient information was highlighted. However, it is crucial that the information is understood, which can be problematic for decision-making as some PCPs reported their patients would not dare to say if they do not understand (FG4PCP3): “Some patients do not dare to say in the consultation that they have not understood it.”

Environmental contexts and resources

Differences in physician perspectives depending on their working environment are presented in this theme. Physicians working in hospitals mentioned they might not consider evaluating statin discontinuation as they lacked full patient history (FG2I3): “We may not have the time to organise the list of diagnoses during a short hospitalisation, and then it is difficult to decide whether it is primary or secondary prevention.” Furthermore, the acute setting was not perceived as ideal to discuss the use of a long-term medication due to time constraints. Conversely, PCPs did not report lack of time to screen medication lists as a relevant barrier (FG3PCP1): “I don't think that [lack of time] is a factor. Because especially older people, you see them… quite regularly.” PCPs mentioned that a decision-aid could be helpful (FG4PCP4): “a visualisation [similar to the existing one for PSA in screening for prostate cancer] showing, how many patients must be treated for one patient to benefit and how many would be harmed. That would really help for shared decision-making.”

Social/professional role and identity

This theme presents both the social interactions of physicians with their patients and their colleagues regarding decision-making about primary preventive statin medication. Trust between physician and patient was seen as important for discussing statin discontinuation because of the uncertainty about the effect of this medication on patient outcomes. PCP role was thus mentioned as central in decision-making, because of the long-term relationship and frequent consultations with their patients (FG4PCP3): “…they [the patients] see their PCP much more often than they see the specialists, which enhances doctor-patient relationship and facilitates the discussion with the patient”.

Regarding interprofessional interactions, hospital internists considered communication with PCPs as an important tool when they needed to make decisions about long-term medications. (FG2I1): “What I do is that I talk to the PCP about the medication. Because it is a shared decision, and the PCP is usually relatively well trusted by the patient.” Conversely, hospital internists and specialists regretted not knowing whether their recommendations were implemented by PCPs. PCPs perceived interprofessional decision-making difficult when patients highly considered recommendations made by specialists. Therefore, PCPs and hospital internists expressed decision-making was more difficult when their opinion differed from that of specialists (FG2I1): “I had trouble discontinuing [the statin] when a specialist recommended it. Then the specialist’s opinion went against my opinion. That is sometimes difficult.”

Discussion

This study reveals important findings about physician perspectives on statin continuation and discontinuation in patients older than 70 years in primary cardiovascular prevention. Physicians with internal medicine, cardiology, neurology or primary care background reported it as challenging to define if a patient is in primary or secondary prevention. Furthermore, the lack of evidence makes the decision about statin discontinuation and continuation difficult. In this unclear situation, neurologists and cardiologists prefer continuing statins, whereas internists and PCPs favour reducing polypharmacy and thereby potentially improving quality of life by avoiding potential adverse drug effects. PCP office was identified as the ideal setting to reach a shared decision on statin continuation or discontinuation, as PCPs know the patients best and see them more regularly than specialists. Discontinuation of a chronic medication was perceived as difficult and having a possible negative impact on patient–physician relationship, because some patients might feel not being worth it, given up or undertreated.

Physician background seems to play an important role in the decision to continue or discontinue a medication when evidence is uncertain. Cardiologists and neurologists reported clear opinion towards statin continuation, justifying this by possible protective effects even in older adults based on weak evidence coming from a retrospective cohort study.²⁰ Furthermore, they tend to continue statins in this situation as a previously established indication does not disappear at a certain chronological age, thus assuming that the statin treatment kept patients in primary prevention by preventing a cardiovascular event. Similar to a previous study,²¹ internists suggested that biological rather than the chronological age and the health status influence the potential benefit of primary prevention. Unlike the specialists, hospital internists and PCPs questioned the utility to continue a primary preventive statin treatment in older adults. They rather favoured reducing polypharmacy and thereby improving quality of life by relieving patients from possible severe side effects. A possible explanation for the diverging attitudes of the specialists and internists could be that hospital internists and PCPs are more often confronted with patients experiencing adverse drug effects, whereas cardiologists and neurologists are rather confronted with patients suffering from severe cardiovascular events.

Compatible with what was published before,^{22 23} physicians through all FG agreed that statins in primary prevention in patients with short life expectancy and high frailty should be discontinued.

PCPs reported a fear of rebound effect after discontinuing a statin. This points out that PCPs are facing a difficult decision. On the one hand, they may want to reduce polypharmacy by discontinuing statins. On the other hand, they fear being confronted in the future by patients suffering from a cardiovascular event after statin discontinuation and feeling responsible for their condition. This fear could be difficult for PCPs particularly, given the frequent long-term relationships with their patients and reduce the likelihood for deprescribing. Conversely, some internists believed that the fear of possible negative outcomes of statin discontinuation in older adults was unfounded. Unlike what was described before,²⁴ our physicians did not mention fear of lawsuits if they do not follow experts’ recommendations. This could be an indication that legal prosecution following undesirable medical events is still relatively seldom in Switzerland. Nevertheless, a decision aid based on data coming from future or ongoing randomised controlled trials (RCTs) such as the STREAM trial (STatin discontinuation as pRevention among the Elderly And Multimorbid, NCT05178420) could support physicians in their decision-making.²⁵

Physicians of all FGs mentioned continuation or discontinuation should be a shared decision between the patient and the physician. Consistent with previous studies,^{26 27} PCP practice was identified as the ideal setting to discuss discontinuation, as all necessary information is available and PCPs have a long-lasting relationship with the patients and see the patients more frequently than specialists, facilitating a shared decision. Mentioning polypharmacy or, as previously described, the occurrence of adverse drug effects¹⁶ was seen as a good starting point to discuss discontinuation. Interestingly, PCPs did not report the lack of time to discuss discontinuation as a relevant factor, whereas patients had reported before that lack of time and support by their PCP was a barrier to drug discontinuation.²⁸ This discrepancy could be explained by different perspectives of patients and physicians on the same situation. Additionally, Swiss PCPs may also have more time available per patient than PCPs in other countries.²⁹

Discontinuation of a chronic medication was perceived as difficult by physicians through all FGs. For example, PCPs experienced a possible negative impact on patient–physician relationship, since some patients might feel not being worth it, given up or undertreated. This shows that good patient information is important to reach a better patient understanding and thus enable shared decision-making. Especially in medical fields of high dynamics as cardiovascular prevention, where guidelines and knowledge about a disease or available medicaments change frequently. By ensuring good patient information, a barrier between physicians and patients could possibly be weakened or even lead to a positive effect on physician–patient relationship. Priorly, PCPs even reported deprescribing as an opportunity to strengthen patient–physician relationship.³⁰

Consistent with previous research, decision-making was reported as difficult in situations with interprofessional divergence of opinion on discontinuation, especially if a PCP was confronted to a clear opinion expressed by a specialist.⁽²⁴⁾ Clear evidence coming from RCTs would minimise the divergence of opinions. However, there will always be situations with unclear evidence or diverging opinions of the treating physicians which might lead to patient confusion. Divergent opinions may also negatively influence patient–physician relationship as we showed above. Therefore, it is crucial that divergent opinions on a treatment plan are subject to an interprofessional discussion before the shared decision discussion with the patient. As direct communication is difficult in the daily clinical work, a suitable communication tool allowing direct contact could help. Additionally, guidelines from general internal medicine considering the patient with all his comorbidities rather than focusing on a disease could help make shared decisions specific to each patient context.

The main strength of our study is the inclusion of physicians from different specialties and different settings, which provides a more comprehensive overview of physician views on decision-making regarding statin continuation and discontinuation. Limitations regard the recruitment at a single hospital in Switzerland and low availability of physicians due to the COVID-19 pandemic. As such, reaching theoretical saturation was not possible and our observations, especially those concerning comparisons between the physician groups, thus have an explorative character and might disagree with the opinions of other physicians. However, since similar themes emerged among the different physician groups despite their different practice contexts, this seems rather unlikely. We have decided to conduct separate FGs for the different groups of physicians included in this study, which can be both a strength and a limitation. Mixing specialists with PCPs or internists would probably have changed our results, most likely due to a more confrontational character of discussion during the FGs, which would have made the analysis more difficult and unclear.

Conclusion

This study highlights physician difficulties to decide whether to continue or discontinue statins in older patients in primary cardiovascular prevention where evidence is unclear about the utility of this medication. We could show the wide range of opinions of physicians of different specialties on this question and that these opinions are often based on physician background and beliefs. In this unclear situation, shared decision-making between physicians and patients was identified as crucial and should happen at the PCP office, given that all necessary information is available and that PCPs have a longer relationship with the patients. More evidence on the use of statins in primary prevention in older adults is needed and might help create a decision aid, which could further facilitate and enhance shared decision-making about statin medication in primary prevention in older adults.

Data availability statement

Data are available upon reasonable request.