A. Prescribing factors.

Overall, GPs acknowledged that it is important to consider the quality of life, health status and life expectancy of their patients, prior to initiating a medicine [12].

“Once they go into the rest home, they’re obviously only going to live for another 5 years potentially, I don’t know what the actual figures are. So you look at it in that view point. You know, everything changes. You look at the risks and benefits of every single decision you’re making.” GP3

“Patient factors would be um, their life expectancy. That’s putting it bluntly, isn’t it? Their age, I think you know, if someone’s you know, you might consider treating a 72 year old different from a 92 year old, or a 102 year old, had a few of those.” GP 4

i. Uncertainties regarding applying evidence based medicine: GPs’ knowledge and ability to synergise and apply evidence based practice may be challenged when caring for older people with complex health problems [19]. GPs in this study acknowledged that they experienced a dilemma when applying research evidence to a patient with multimorbidity. This was the case particularly for preventive medicines such as aspirin and statins.

“There’s also a challenge of when do you stop preventive medicines such as statins and aspirin?” GP3

The dilemma of continuing or stopping preventive medicines also concerned GPs who reviewed a patient profile in our recent qualitative study. Our study found that GPs felt more confident about symptom management and adjusting the dose of symptomatic medicines, according to patient response [12]), rather than reviewing the appropriateness of continuing preventive medicines.

Lack of access to user friendly evidence based deprescribing guidelines: Few clinical trials have highlighted that nonadherence to disease-specific clinical guidelines may result in poor quality of patient care [20]. Paradoxically, there is growing evidence that older people with multiple conditions have poorer outcomes when treated according to disease-specific guidelines compared to other patients [10, 21]. In addition, the presence of multiple guidelines discourages physicians from applying these guidelines in clinical practice [22]. One GP in the present study expressed their frustration regarding the complexity of guidelines.

“I have seen these Best Practice Advisory Committee (bpac^nz) articles, but I can’t remember what was in them, you know? Has it changed my prescribing? No it hasn’t, which I suspect applies to a lot of us. We get so many guidelines, there’s so many… that’s the other thing. You have all this stuff (different guidelines) coming at you and everybody’s saying you could do better, and well maybe we could, in this speciality, but actually, when you’re putting it altogether, maybe we’re not as bad as people like to make out because when you’re sitting there with a lot of stuff, I mean we basically see that balancing huge amounts of information.” GP4

“Well like you mentioned earlier, that Best Practice Advisory Committee (bpac^nz) one they’ve written about it, and just making you more aware and giving the time so that I can actually read it, I think I’ve started it, but I haven’t read it. It’s in a pile, so makes you more aware and more confident of making the changes if and when needed.” GP10

Lack of easy-to-use guidelines or decision support is a particular challenge for GPs whilst implementing deprescribing and many physicians are forced to make decisions without much guidance.

“Often research will say if you’re above for example, 80 (years of age) you’re excluded.” GP9

“And that’s the hard thing, whether a little bit of omeprazole, which you’re meant to slowly reduce from 40, to 20, to 10, to stop, and when should that be done? It’s not easy, because I mean guidelines say that you should reduce, but it is often not indicated why people are on say, omeprazole. I find those sort of things hard.” GP10

In a synthesis of ten qualitative studies, physicians highlighted that guidelines focus on health outcomes that may be of little relevance to their patients [22]. This leads to physicians deviating from guideline-directed care. In addition, they mentioned the lack of available tools to help physicians quantify the benefits and harms of prescribed medicines, as this is often a dilemma whilst optimizing medicines in older people. As a result, physicians making a clinical decision, follow an approach that combines prioritizing the main problems most relevant to the patient, stratifying the risk of a disease outcome, modifying guideline-directed goals or interventions according to anticipated adverse effects the patient might experience and then making a clinical decision balancing potential benefits and harms [22, 23].

ii. Fear of consequences: Physicians have a genuine fear of patients experiencing deterioration in their health, shortly after they have stopped or reduced one or more of their medicines. They may also fear others (family or health care staff) having the misperception that coincidental adverse effects the patient may develop, are a direct outcome of the physician stopping medicines.

“Coincidentally, people would die when you stop medicines. Not because you killed them, but they happened to.” GP3

In addition, physicians fear that patients may perceive deprescribing as withdrawal of care before their end of life. These misconceptions prevent physicians from implementing deprescribing. For many GPs, ‘maintaining the status quo’ for multimorbid patients with polypharmacy who are stable, is seen as the most reasonable course of action to undertake.

“If they’re otherwise stable, and life’s going on, it’s the same as prescribing for anybody else really which doesn’t really help you very much. I don’t reduce or stop unless I think it’s needed. If you stop it, are you going to de-stablilise things? So, if the system is working, I have to say, although we look and we wonder, we tend to leave them on.” GP4

“It’s hard to give them the option sometimes, and you feel like you’re writing them off if you’re crossing off all this preventative sort of stuff.”

Even though this fear of consequences exists, some GPs commented that through their personal experience, patients have often improved once their medicines have been stopped.

“I had a patient today, a 90 year old, I reduced her metformin and she said: I’m feeling much better since you did that, you now? So you can often get away with you know, reducing down.” GP6

B. Social influences.

i. Patient and family factors: Several GPs in our study felt the importance of delivering individualised patient-centered care to their patients. This is similar to the views of other GPs who felt the importance of considering the patient’s personal preference and patient empowerment when tailoring care [24]. To achieve this, taking into consideration individuals’ opinions or perceptions about their medicines is important. However, patients may not always be willing to stop or change medicines they have taken for a long time [12], despite the physician’s recommendations.

“Sometimes it is hard because they have been on it for donkey’s years, and they totally believe in it.” GP9

In addition, certain patient characteristics, such as loss of cognition and memory, can further complicate this process and make delivering patient-centered care, even more challenging [22, 25].

ii. Other prescribers: Multimorbid patients tend to visit a greater number of physicians, which could include specialists. This can have several effects on the patients’ care. It has been shown that those patients who visit four or more physicians experience problems such as conflicting medical advice, lack of access to laboratory test results and more commonly, duplication of laboratory tests from multiple prescribers [26].

It was apparent in our study, that GPs’ prescribing is influenced by specialist recommendations. One GP disclosed, they considered which specialist initiated the medicine, before changing the prescription.

“So if they’ve seen a cardiologist and were put on a statin, you feel very nervous about stopping it for example. There’s no doubt that you know a specialist assessment, commenced on a specific drug, you know it does make you reluctant to change things, yeah, it does have that, it influences you.” GP7

In addition, GPs believed that patients have the perception that physicians in hospitals, or specialists in private clinics have more knowledge than GPs practising in primary health care. This could disrupt the trust in patient-GP relationships, leading to compromised shared-decision making between the patient and the GP.

“Yeah, look the big doctor in the white coat in the big house on the hill always knows more than the GP especially the house surgeon who might have a brief amount of experience and does what they’re told and one of the issues with this process is, experienced GP’s still think that the doctor up the road knows more.” GP9

Specialists who operate on a single disease paradigm without an overview of the ‘whole patient’ can lead to fragmented care. Single-disease care is antagonistic to the goals of GPs in primary care. GPs receive poor communication from other care providers in multimorbidity [22]. This uncertainty makes it more difficult for GPs to make prescribing decisions with confidence and impedes GPs from delivering patient-centred care to their patients [22].

iii. Residential care staff: Nurses have a significant role in caring for people living in residential care. Most of the information GPs collect is sourced from nurses instead of a direct conversation with the patient. As a result, biased opinions could be communicated and shared-decision making between the patient and the GP could be compromised.

“We are relying on the nursing staff to feedback information to us, and often that’s not brilliant, hence we’ve lost that kind of one to one with the patient, which makes it much harder to decide how to prescribe, because you’re being biased by other people.” GP3

In addition, nurses could encourage GPs to prescribe certain medicines; in particular those medicines with a sedative effect.

“Uh, some nurses do push for medicines to be prescribed.” GP8

“I do think sometimes you wonder who are we treating? Are we treating the nursing staff who can’t face somebody calling out at night, or are we treating (an) actual patient who may be very well, happen to be calling out once or twice a night at one o’clock, but then fall back to sleep, you know?” GP9

1) Pharmacist medicine reviews in a multidisciplinary integrated approach.

In this study, time constraints were identified as a limitation preventing physicians from reviewing medicine lists for older people. This impedes the process of deprescribing medicines noted to be unnecessary and/or harmful. One way to implement deprescribing is to conduct comprehensive deprescribing reviews. Physicians and pharmacists have worked together using a multidisciplinary approach to reduce polypharmacy and inappropriate medication use (IMU) in older people by carrying out comprehensive medicine reviews [13–16].

The GPs in this study, perceived that treating multimorbid patients, required a collective effort from different health care professional groups. This approach could really help deliver patient-centred care, especially if the patient and/or their family are involved in the process.

“I think if you made medicine reviews compulsory, I think it could be very useful. When people have incentives, for example, if they had to do it for accreditation, they will do it, otherwise it’s not going to happen. It’s simple, you know? People just don’t have, I mean we’re all busy, and don’t have tons of time. I think we need to carry out medication reviews, and not miss people out. Sometimes its good to have somebody else look at it, so working together with a pharmacist is a good idea. Because I think two pairs of eyes looking at the same page, often gets better results than one person looking at a patient.” GP9

“Its little things you know, such as, how about considering a medicine review on this patient as they meet the criteria you know for a medicine review?” GP7

“I’m not against talking with a pharmacist or a geriatrician for half a dozen, ten patients that you look through and even a consultant, and say we’re looking at that for that person. I mean that to me (speaking to a pharmacist and a geriatrician about some patients) would be a reasonable thing, as well (if) it didn’t take too much time and it wasn’t too difficult to prepare. So you could make it more certain with ideas from a pharmacist and a consultant about that. And I would see that as worth thinking about.” GP10

“Well the help of a clinical pharmacist, or certainly a very good nurse manager to help, or the cardiac nurse specialist, because if we’re decreasing medication we’ve got to keep an eye on heart failure, blood pressure, blood sugars (glycated haemoglobin) kind of thing. But having a clinical pharmacist involved would be great. I mean ideally a physician, a clinical pharmacist and the GP, and the nurse would be great. You know, the whole mixture” GP2

Several GPs embraced the idea of pharmacists’ involvement in a multidisciplinary team to review medicines, but had concerns about the delivery of the service. The primary concern was that pharmacists may offer recommendations that may have already been recognised by the GP and these may have not been implemented for compelling clinical reasons. These reasons are often not communicated to or recognised by the pharmacists. GPs were positive about the potential for inter-professional partnerships. In a residential care setting, pharmacists have access to residents’ clinical information and are well placed to make recommendations for GPs to consider. In contrast pharmacists working in the community do not have access to patient information. As a result, GPs fear that the pharmacists’ recommendations might not be relevant in this setting. In addition, lack of resources hinders these multidisciplinary meetings from occurring regularly in practice.

“I think it’s good to know the pharmacists if you can, but it doesn’t have to be a local pharmacist. It’s just that locality means you build relationship with people more easily.” GP6

“The pharmacist down the road are sort of part of the team, but they’re in a different building, and we never see them. They’re a voice on the phone, they do rollover and they lack information; there’s no doubt about that. They don’t know what Mrs. Bloggs has got or why we’re treating them for half the stuff. We would be better if they had the time to, for the pharmacists to sit down with the physician and say you know, it’s possible that you could stop ‘x’ and ‘y’. But how can they do that if they don’t know what the patient’s got? I guess, if you’re part of the treating team, you’re part of the treating team. And that’s the critical thing. And if you’re not, you’re not.” GP4

2) Adequate reimbursement.

GPs in this study felt that the overall structure of residential care prescribing and reimbursement systems are disorganised. Prescribing for older people in residential care is viewed as cumbersome for many, as time constraints and limited resources prevent this process from being carried out efficiently. In addition, they felt that they received insufficient compensation for the amount of work and attention that this area of prescribing requires. These opinions were not exclusive to our study and are similar to the views of GPs in other studies [12, 19, 24].

“The challenges are around time coordination, and um, GPs invariably complain they don’t have enough time. But being cynical, not much here, and long enough in the tooth, that if you pay people adequately, they will make the time. A lot of this is under-funded, and it takes time outside people’s main surgery working hours. GP5

“Your first challenge is; you go to the rest home. You try and find a nurse. You can never find one. You try and find the notes, hard to find. You can’t find the medicine chart, it could be on the rounds somewhere. It’s not computerised, it doesn’t link with our technical notes at the medical practice, so quality just goes down. It shouldn’t be, but at the practice we’ve got the computer, we’ve got light, we don’t have a darkened room in a rest home, and we can actually see what’s going on.” GP3

If I don’t have my rest home rounds by 7:30 in the morning at least 4 days a week, I barely finish up with the rest of my rest home work then occurring jammed in at lunchtime or at 6 o’clock at night, 7 o’clock at night, when you… the nurse may no longer be on duty, the patient’s tired, it all turns to custard. So it is not accorded a high, a very high priority and does take place at these extreme ends of the day, and whereas if it’s adequately funded, for you know, 8–8:30 in the morning, and it’s properly funded, and everybody’s there and ready to go, GP’s will go. But while it’s as ad-hoc, as it is, it’s seen as a chore and it’s not well done.” GP5

“Um, I think having adequate funding for older people so that you have time to really think about what they are having so, and Care Plus (a recent government funding initiative) has been good from that point of view, because it gives some extra funding for people like this who have complex health problems. So that has been a real help having enough you know, that recognition of the time involved in reviewing people adequately in rest homes. So that’s really, so that’s’ fine at the moment. I hope they don’t remove that, because it just puts the cost on patients if you’re going to give time to reviewing their medication. So you know there’s nothing additional that I would do, yeah.” GP6

3) Better communication between physicians at health interfaces.

Clear and transparent communication is essential between GPs in primary care, specialists and hospital physicians. Integrated health care meetings between GPs and specialists, and information technology to clearly outline patient plans in terms of prognosis and care, are some strategies that can improve seamless care for a multimorbid older individual. Bidirectional communication between GPs in primary care and physicians at the hospital is crucial. This would improve the suitability of medicines prescribed to patients and in some cases, may prevent inappropriate medicines being prescribed. Furthermore, better communication would help physicians to ‘speak with one voice’ which would result in greater satisfaction for both physicians and patients, as different stories provoke distrust [12, 22].

“The reverse of hospitals putting them on medications, is that hospitals stop medications sometimes, and a week later they start them on all of the medicines again, so that sort of communication needs to be taken into consideration.” GP1

“If we could stop hospital physicians prescribing single issue medicines for compromised older people, we’d reduce our problems by 50% overnight.” GP5

“I think we are not daring enough, and I think the communication between the hospital and us (GPs in primary care) could be far better.” GP9

4) Deprescribing guidelines, GP education and GP empowerment.

Experiences of practising physicians in this study and other studies, suggest that they struggle with the uncertainties of applying disease-specific guidelines to older people with multimorbidity [23]. One qualitative study, by Smith et al., reported that having a multimorbidity focus in GP training is important as GPs expressed that they lacked confidence and felt that they needed more training and clinical support [25]. Another study by Herzog et al., reported that specific geriatric training for GPs is likely to have a positive effect and may help overcome some of the barriers outlined earlier [19]. In the present study, one GP also highlighted the need for improving GPs’ awareness on such issues.

“Improve GPs’ awareness or education on the issues (involved with prescribing for multimorbid patients), and um, perhaps the medicine review service.” GP7

Improving GPs’ awareness can be augmented by positive testimonies from GPs. This may also help empower GPs, and could encourage them to deprescribe. One GP in this study spoke about the positive outcomes they observed after implementing deprescribing.

“Since I’ve started to look at that more globally, the number of medicines I’m prescribing on average for patients in rest homes is about 50% of what I was prescribing a year ago and they aren’t falling off their perch in greater numbers. Patients like it (being prescribed less medicines). They say oh, a whole big meal of pills, and you know, people are generally better. People wake up, they’re less nauseated, they have fewer falls, all those sorts of things, yeah. GP5

In addition, this study illustrates that GPs feel pressured into continuing the prescription of certain medicines initiated by specialists; even if they question the medicines’ suitability. GPs also find it more difficult to convince their patients of certain changes, as people may perceive physicians in the hospital to have more experience or knowledge.

“Yeah, look the big doctor in the white coat in the big house on the hill always knows more than the GP especially the house surgeon who might have a brief amount of experience and does what they’re told and one of the issues with this process is, experienced GP’s still think that the doctor up the road knows more.” GP9

Therefore, it is important to ensure that current GP training meets the needs of GPs, and is focused around delivering patient-centered care for older people with multiple comorbidities. In addition, further work is required to create and disseminate clinical tools, suitable for use in primary care practice as GPs need different approaches, to help them deliver medical care that fulfills their patients’ priorities [23]. These strategies need to allow GPs to employ flexibility in implementing prescribing guidelines, while responding to the individuals’ needs and preferences [29]. This will in turn empower GPs and provide them with the confidence needed to be experts in prioritising medicines and stopping medicines initiated by specialists who would have dealt with one particular patient issue, without considering all other aspects of the patient’s care.