Health professional perspectives on the management of multimorbidity and polypharmacy for older patients in Australia

Abstract

Background

delivering appropriate care for patients with multimorbidity and polypharmacy is increasingly challenging. Challenges for individual healthcare professions are known, but only little is known about overall healthcare team implementation of best practice for these patients.

Objective

to explore current approaches to multimorbidity management, and perceived barriers and enablers to deliver appropriate medications management for community-dwelling patients with multimorbidity and polypharmacy, from a broad range of healthcare professional (HCP) perspectives in Australia.

Methods

this qualitative study used semi-structured interviews to gain in-depth understanding of HCPs’ perspectives on the management of multimorbidity and polypharmacy. The interview guide was based on established principles for the management of multimorbidity in older patients. HCPs in rural and metropolitan Victoria and South Australia were purposefully selected to obtain a maximum variation sample. Twenty-six HCPs, from relevant medical, dentistry, nursing, pharmacy and allied health backgrounds, were interviewed between October 2013 and February 2014. Fourteen were prescribers and 12 practiced in primary care. Interviews were digitally audio-taped, transcribed verbatim and analysed using a constant comparison approach.

Results

most participants did not routinely use structured approaches to incorporate patients’ preferences in clinical decision-making, address conflicting prescriber advice, assess patients’ adherence to treatment plans or seek to optimise care plans. Most HCPs were either unaware of medical decision aids and measurements tools to support these processes or disregarded them as not being user-friendly. Challenges with coordination and continuity of care, pressures of workload and poorly defined individual responsibilities for care, all contributed to participants’ avoiding ownership of multimorbidity management. Potential facilitators of improved care related to improved culture, implementation of electronic health records, greater engagement of pharmacists, nurses and patients, families in care provision, and the use of care coordinators.

Conclusion

extensive shortcomings exist in team-based care for the management of multimorbidity. Delegating coordination and review responsibilities to specified HCPs may support improved overall care.

multimorbidity, polypharmacy, care coordination, primary care, older patients

Background

The majority of older adults have multimorbidity, defined as experiencing two or more long-term health conditions [1]. A rapid increase in multimorbidity research has occurred over the past decade, linked to an increasing prevalence of multimorbidity and increasing challenges in areas such as workforce redesign and appropriate funding mechanisms to deliver appropriate care for such patients [2, 3]. The contiguous phenomenon of increasing polypharmacy—using multiple medications—has received similar attention [4]. Patient-reported challenges to optimising medication management in the context of multimorbidity and polypharmacy include out-of-pocket costs, care coordination, conflicting therapeutic advice from prescribers, low levels of engagement in decision-making and knowledge or skills deficits [5–7].

Challenges with care delivery for multimorbidity arise from factors such as limited consultation time, inadequate service coordination, lack of evidence, inadequate consideration of multimorbidity in clinical guidelines, inadequate medication review and multiple prescribers acting independently [8]. Most older people experience multimorbidity, but guidelines often focus on individual conditions and thereby have uncertain applicability for many patients [9]. Applying multiple individual guidelines can lead to complex medication regimens and treatment conflicts [10]. Emerging evidence suggests that multimorbidity and associated challenges with providing appropriate care may increase treatment burdens, reduce quality of life, diminish medication adherence and cause adverse drug events, unnecessary hospital admissions and unnecessary healthcare expenditure [11, 12].

To support best practice for multimorbidity [13], the American Geriatrics Society (AGS) published Guiding Principles for the Care of Older Adults with Multimorbidity (2012) [14]. It is unclear if these generic principles are being applied in different health systems to support multimorbidity management. Multimorbidity management has largely been explored among a small number of primary care professions, particularly general practitioners (GPs) [8, 15–18]. The application of AGS principles for multimorbidity is also largely unexplored from the perspective of an often diffuse healthcare team. This is despite the increased likelihood of hospital admissions and transition between different care settings for patients with multimorbidity [12], and their greater susceptibility to fragmented care, inappropriate polypharmacy and the adverse consequences of polypharmacy [19]. The objective of this study was to explore current approaches to multimorbidity management, and perceived barriers and enablers to appropriate medications management for community-dwelling patients with multimorbidity and polypharmacy, from the perspectives of Australian HCPs.

Methods

Design and setting

Semi-structured interviews were undertaken with HCPs working in metropolitan and rural areas of Victoria and South Australia. The healthcare structure is similar in each Australian state, with nationally funded primary care services and state-funded hospital care. The primary source of universal healthcare is typically a GP. GPs also act as gatekeepers to other government-funded health services. The interview guide was based on the AGS Guiding Principles for the Care of Older Adults with Multimorbidity [14]. These principles align with recommendations from several Australian policy documents, but there is no single Australian report that comprehensively frames the same issues. The study was approved by the Monash University Human Research Ethics Committee.

Study sample

A maximum variation sample of participants from a diverse range of HCPs in primary, secondary and tertiary care settings was purposively recruited through investigator networks. Potential participants were identified according to their professional background and experience in managing multimorbidity. Because a wide range of individuals influence prescribing and support for patients around medications, the sample included a range of medical and non-medical prescribers and non-prescribers to ensure a broad understanding of perceived roles, responsibilities and challenges experienced with multimorbidity and medication management.

Overview and health professional characteristics

Overall, 26 HCPs were recruited from relevant medical, dentistry, nursing, pharmacy and allied health backgrounds, and almost half worked in primary care settings (Table 1). Participants were aged from 29 to 70 years and more than half (14/26) were prescribers (Table 1).

Table 1.

Healthcare professional characteristics

Social demographics	N = 26	%	Details
Age
20–40	8	30.8
40–60	15	57.7
>60	3	11.5
Area
Rural	9	34.6
Metropolitan	17	65.4
Setting
Community	12	46.2
Hospital	14	53.8
Prescribing status
Prescriber	14	53.8
Non-prescriber	12	46.2
Field of practice
Acute care medical specialitya	6	23.1	Three general medicine consultants, two geriatricians and one clinical pharmacologist
General practicea	5	19.2	Five GPs
Nursing	6	23.1	Two practice nurses, two nurse practitionersa and two hospital nurses (including one transition of care coordinator)
Pharmacy	6	23.1	Four hospital pharmacists and two community pharmacists (including three home medication review accredited pharmacists and one transition of care coordinator)
Dentistry	2	7.7	One dentista and one dental hygienist/centre coordinator
Physiotherapy	1	3.8	One physiotherapist

^aDenotes disciplines/subdisciplines with prescriber authority.

Data collection

The semi-structured interview guide was designed around six domains relating to current multimorbidity management, individual strategies for patient-centeredness, barriers to AGS-defined best practice and how they think the healthcare team should function to improve care. Five of these domains aligned with AGS principles [14]: elicit and incorporate patient preferences; recognise limitations of the clinical evidence base; frame treatment decisions in the context of risk, benefit and prognosis; assess the clinical feasibility of treatment options and use strategies to ensure that treatments optimise benefit, minimise harm and enhance quality of life for older adults with multimorbidity. Because these principles focus on individual care, a sixth domain was added to explore care coordination.

To facilitate consistent interpretation, commonly used definitions were provided before interviews. Polypharmacy was defined as the use of five or more different medications [20]. Multimorbidity was defined as an individual having two or more chronic conditions [13]. It was clarified that the interview focus would be medication management for community-dwelling patients with both multimorbidity and polypharmacy. The interview guide was pilot tested for face validity by conducting interviews with two pharmacists and a physician.

Twenty-four interviews were conducted face-to-face at the participant's practice setting or another mutually convenient location, and two were conducted by telephone. The average interview duration was 45 minutes (range 25–61 minutes). Written informed consent was obtained to audiotape interviews. Interviews were conducted until data saturation occurred, where no new data, themes or coding emerged. To ensure data saturation, independent coding of transcripts by multiple researchers was employed, and interviews were structured to ensure continued exploration of key issues with multiple participants (minor amendments to the interview guide were introduced after pilot interviews to further explore, confirm or contest issues of interest identified from previous interviews).

Data analysis and synthesis

Interviews were digitally audio-taped and transcribed verbatim. After familiarisation with all interviews, two researchers (B.B. and H.A.) carefully read all transcripts line by line and independently coded data using a constant comparison approach. Conflicting codes were discussed until consensus was reached: first by the two coding researchers and then with other authors when required. B.B. and H.A. agreed, after discussion with other authors, on a set of codes to apply to all subsequent transcripts. These codes were then aggregated into themes using AGS Guiding Principles as a framework for analysis [14, 21]; any conflicts were again discussed for consensus.

Results

Incorporation of shared decision-making and patient preferences

All HCPs perceived shared decision-making as important to improve medication adherence and maximise treatment benefits. Apart from geriatricians, most prescribers either recognised that they did not routinely practice shared decision-making or sidestepped the subject, when asked how they incorporated shared decision-making into routine care. Somewhat related to shared decision-making, nursing and pharmacist participants described filling gaps in patient knowledge or addressing patient concerns and preferences after medical decisions had sometimes been made without adequate patient engagement.

I spend a lot of time doing and getting complete medical history and medication reconciliation. Then the next thing is making sure they [patients] really want to be on the medicine. Because if they don't want to be on the medicine there is no use in me prescribing the medicine.

Geriatrician 2

The big thing which I think sometimes the doctors forget is consent…you go in there [to the patient and say], ‘Hi XXXX, doctor XXXX has requested that you do cardiac rehab’, and he said ‘I am not going to cardiac rehab. I did it two years ago, I've got a sore hip, I have got pain [in] the guts, I don't wanna. I don't drive, my wife has to bring me in, once a week, I can't do it and he didn't even ask me.

Registered nurse 1

Prescribing HCPs, particularly GPs, described how the limited time available for care episodes reduced the likelihood of incorporating patients’ preferences into treatment decision-making for multimorbidity. While exploring or confirming patient acceptance of ‘clinical’ goals was commonly mentioned, no participant made reference to exploring goals for daily living as a means of eliciting patient preferences.

Cognitive impairment, hearing problems and poor health literacy were cited as prevalent additional challenges. Several pharmacists and nurses suggested that patients are also often reluctant to ask physicians questions, fearful of consuming physicians’ valuable time.

Evidence base

Prescribing HCPs questioned the applicability of clinical guidelines when making decisions for patients with multimorbidity.

Guidelines that have been developed for individual diseases are based on research that excluded patients with multimorbidity […] we try and follow those [guidelines] where they are applicable but otherwise I guess it is just having the experience of working with these patients and working out what works and what doesn't.

Geriatrician 1

…general guidelines don't work very well for complex poly comorbid patients. In fact, the general guidelines [… h]ave got information that's […] wrong and dangerous and risky if you actually follow them. […] The heart failure, the cardiology guidelines still recommend very restricted diets for patients with heart disease and those guidelines are appropriate for 50 year old overweight patients, but 85 year old patients with multiple chronic diseases they are very at risk of malnutrition so those guidelines are really not appropriate for these common patients particular as they start to age.

Clinical pharmacologist

HCPs more often relied on personal clinical judgement and experience. Lacking this experience, junior doctors reported being dependent on advice from senior colleagues. Nurses reported major problems with conflicting advice from different prescribers, leading to patient uncertainty.

Patient prognosis

All HCPs reported problems with incorporating patients’ prognoses into decisions about therapy appropriateness. Most were unaware of available prognostic tools. The HCPs who were aware of these tools considered them unhelpful because they presented probability of mortality rather than life expectancy. Therefore, likelihood of benefits from treatment is only clear for patients with very high or very low estimated risks of mortality. Conversely, unvalidated mortality predictors such as longevity of patients’ family members were reported for estimating patients’ life expectancy and deciding the appropriateness of therapies. Some HCPs reported ethical concerns around denying treatments and litigation fears if, after withholding preventative medications, a related clinical event occurred (e.g. stroke after statin withdrawal). Prescribing of potentially inappropriate preventative medications was a concern, but no HCP could suggest how to resolve this issue.

If something estimates a risk of death in five years of 20% or 40 […] it doesn't really help me (to decide if I should) prescribe a statin or not. If someone's risk of death is 80% within a year, I wouldn't prescribe that person a statin.

Clinical pharmacologist

You need to have an idea of obviously what their life expectancy is, their quality of life, all those things, their genetics, 'cause if they come from a family where they live a long time that might make you prescribe something that is prophylactic.

GP 2

Clinical feasibility of treatment plans

There appeared to be no clear method for assessing patients’ likely adherence to treatment plans if modified. HCPs also struggled to articulate how they assessed the likely impact of changes to treatment and care plans on patients’ capacity to adopt recommendations. None mentioned using available tools to measure medication management capacity or seeking patient feedback about self-management capacity. Hospital-based HCPs reported that treatment plans are often changed during admissions, but that they do not have sufficient information or time to assess patients’ adherence capability. Post-discharge appointments in primary care were considered by these HCPs as the most appropriate setting to assess patient adherence to treatment plans and self-management.

I might see a patient once, maybe twice or thrice, and then I might never see them again. Cause my, I am there to just give the GP advice, so in the end of the day I have to hope that wherever else they go, people are also giving them the same information, doing the same things, checking and educating.

Geriatrician 1

(There's) not many (steps I can take to assess feasibility) [laughing], uhhhhhh…there is probably not very much we can do other than talk to them and say it (adherence) is important […] I don't tend to see patients on a regular ongoing basis. I see them in you know an acute setting in emergencies […] I am not sure that there is a great deal that we can do that actually improves compliance.

General medicine consultant 2.

Optimising therapies and health management plans

Participants commented most extensively on issues around optimisation of care plans and therapies, and related difficulties with continuity of care. Supplementary quotes for these two issues are provided in Box 1. Most participants admitted, some with embarrassment, that they should, but do not, routinely engage in optimisation of care plans and complicated medication regimens. It was not a priority for time-limited patient encounters.

Oh deprescribing! Yes, I'd like to do that more than I do, yeah we sort of scratch things off and delete things, we do that too. One of the last things you would do when it comes to the medication is to cease certain things.

GP 1

Box 1.

Summary of themes relating to optimisation of therapy and care coordination identified from participants

Theme	Quotes
Optimising therapies and health management plans	GP 2: Oh deprescribing! Yes, I'd like to do that more than I do, yeah we sort of scratch things off and delete things, we do that too. One of the last things you would do when it comes to the medication is to cease certain things Interviewer: in what kind of situations do you do that? GP 2: usually the true answer, you want the truth? The truth is when you remember, when you are not rushed, then you can do that. Because that is the next thing on your list, the last thing you would do. One of the last things you would do when it comes to the medication is to cease certain things. Nurse 1: But sometimes there is such a discrepancy they get to the GP and they might have been discharged on Friday and then come and see the doctor and the appointment was made by the hospital on the Monday or Tuesday. The GP doesn't have a discharge summary, does not have an updated medication list and the patient turns up at the appointment that was made by the hospital and they go ‘why are you here?’ and the patient goes ‘well don't you know?! I've been on the waiting list because I've been in the hospital last week’ and that gets them cross which is fair enough, they are paying good money to follow up for their condition for what they are in hospital for and the GP is not even aware of it. Because I guess there is a time lag between the discharge summary being done and actually the GP receiving it. So here you've got the patient trying to explain to the GP what happened to them in hospital to fill in the picture and tell them something while they are at the appointment. Geriatrician 2: Or there is often a view that I hear from patients is well my doctor prescribed this for me therefore it must be appropriate therefore I must keep taking it, even when it is clearly not appropriate. GP 2: Because I can't usually just ring up and talk to the (hospital) doctor who wrote the prescription, I have to go through the health information service and sometimes they are more and less helpful cause sometimes they won't tell you because they are worried about privacy so then you have to go through a process to get it, to have your signature to say are you happy for me to have that information. But that delays things; it means I can't make a decision now. Clinical pharmacologist 1: I think people do practice very defensively because they fear lots of things, they fear criticism by their colleagues as well as litigation, litigation is actually very rare in terms of being sued. There is other things as well, the fact that we work in a public health system, and it is perceived that you might be ignorant of new treatment or negligent if you don't offer it and someone isn't offered treatment. So it is not just the fear of litigation that drives people to offer treatment. General medicine consultant 3: I think it is quite important. I have to be honest I don't do that myself. I think it's when, uhmm, as people age I think that a lot of doctors forget that you know, because we see them so often we forget that they actually are getting older now. The benefits of certain medications may not be actually effective, say when they are aged 80 years old compared to when they are 40 years old and I think that is a big problem in doctors that don't work in aged care. I think a lot of the doctors forget that you know, maybe we should start weaning of medication.
Coordination of care	GP 2: Because I can't usually just ring up and talk to the doctor who wrote the prescription, I have to go through the health information service and sometimes they are more and less helpful cause sometimes they won't tell you because they are worried about privacy so then you have to go through a process to get it, to have your signature to say are you happy for me to have that information. But that delays things; it means I can't make a decision now. So I have to wait for that information to come back. It might be today, it might be tomorrow or it might be the next week, so it is difficult for those people who are in transit. Community pharmacist 2: One of the issues with just chemotherapy in general is that GPs will prescribe cardiovascular medicines for diabetes or whatever it might be, but they don't want to get involved in oncology they go ‘oh! I don't know anything about it’ so they let that happen. When we speak to our oncologists they go I have not a clue about cardiovascular drugs, ‘I just know about oncology’, so they don't worry about that and that is where you get the complexity because not everyone is across (everything) and they are not willing to step in […] The scope of practice doesn't allow them to so who is looking after the patient across the board. And that's when I think pharmacy comes in I guess is that we've got an opportunity to at least address that and be a bit of a middle man to help pull it together. Nurse practitioner 1: They get some contradictory information for example patients might be told to drink a lot of fluid if they are diabetic and to help improve their renal function but if they've got poor left ventricular function because they've got a cardiomyopathy or just a tricky heart really they will be told to limit their fluid intake to 1.5 litres a day. So then they get in a conflict like ‘my endocrinologist and renal doctors say I've got to drink three litres a day and you're telling me 1.5 and now I don't know what to do’. So I think sometimes there is a conflict in the messages, they may get a conflict and don't quite know how to understand. Or you know which track that they need a pathway down the middle and who do they need to believe? GP 2: Private hospitals have no access to discharge summary system whereas the public hospitals do have a discharge summary. And it gets faxed to me. Usually within 2 or 3 days when the patients have been discharged unless it is a complicated and difficult inpatient admission with lots of changes. Then the discharge summary can often take 2 or 3 weeks […] before we get the information. So for the most important one there is a bigger problem. Dental centre coordinator: We rely on the patients themselves in providing their medical history and their medications and we get given all sorts of small papers with handwritings of medications and so on […] GPs are not easy people to deal with and communicate due to shortage of time, not criticising them and sometimes they are hard to get in touch with. Dentist 1: I have had issues with GPs, they don't want to answer questions about a patient's history. They say if you want to do a procedure, why should I get involved? They misunderstand their roles and responsibilities, that is not the norm, but it has happened when the GP would refuse to cooperate. Practice nurse 1: …there's not a lot of communication between the hospital and us here, and I think a lot of that is to do with time constraints as well. The discharge summaries from the hospital aren't very user friendly but that is a working thing in progress, we are looking at, at the moment, a whole new process with IT and that sort of thing that we will be able to access all of that information as well. So we'll see with the new collaborative programme that's out, we're very involved with that, with patient hand held records and things like that. So that's all in the process and I think in five years’ time it will be wonderful but at the moment it's not.

There was also a reluctance to ‘interfere’ with other HCPs’ prescribing driven by fear of disturbing therapeutic relationships, hesitation to contradict prescribing by other HCPs and among junior doctors, poor confidence to change treatment plans stemming from inadequate knowledge and experience. Some geriatricians and pharmacists mentioned that some patients ‘love taking medicines’, making it difficult to initiate conversations about deprescribing.

Some patients are, if they have been told once, you have to take this medication for the rest of your life, they really hang on to that view, even when it becomes inappropriate for them to continue on that medication.

Geriatrician 1

Prescribing HCPs reported not having a systematic approach to deprescribing or optimisation of care plans. Many of the community-based HCPs were unaware of tools to address potentially inappropriate polypharmacy (e.g. START/STOPP, Beers Criteria). Some of those hospital-based HCPs, who were familiar with these tools, considered them outdated, containing obvious interventions or too lengthy. Most participants reported awareness of potentially inappropriate key medications for older people, such as benzodiazepines and anticholinergic medications.

Coordination of care

Some HCPs, particularly GPs, acknowledged that when multiple prescribers provide care in silos for the same patient with multimorbidity, overly complex medication regimens result that adversely affect patient adherence to treatment plans. Conflicting opinions existed regarding the optimal time to review and optimise medication regimens. Some HCPs believed that the best time is during a hospital admission because all specialists are available in one place for discussions. Hospital-based HCPs were considered more accessible for consultation compared with community-based counterparts because of perceived better collegiality, respect for each other's profession and the opportunities to discuss issues in person.

…hospital is a very supportive environment and is very collegial, why is suppose, you know, you have, you don't feel like you are alone so much. Where I imagine if you work in at the community pharmacy it would be very difficult to have that ..

Hospital pharmacists 1

Other HCPs reported trying not to adjust medication lists in hospital to avoid errors and confusion for both patients and GPs at discharge and post-discharge.

…no bits of the list of responsibilities and roles for management of polypharmacy for elderly people can be done successfully by any other profession other than general practitioners in my opinion

GP 2

In non-medical settings such as dental clinics, HCPs similarly felt that there were considerable difficulties with acquiring the necessary information to make treatment decisions from GPs.

Hospital-based HCPs acknowledged the importance of arranging post-discharge appointments, but many reported not doing enough themselves to ensure appropriate post-discharge care coordination. They cited time constraints and reluctance to assume responsibility as key factors and concluded that primary care is the most appropriate setting to evaluate treatment plans and patient adherence. Community-based HCPs argued that delayed, lost or vague discharge summaries or medication lists make it equally impossible for them to coordinate post-discharge requirements.

Hospital-based HCPs complained of limited abilities to meet care planning responsibilities because of inaccurate and outdated medication lists provided by GPs. Transition of care problems were reportedly far more common among patients with multimorbidity, since updating the more complex medication lists and writing discharge summaries for these patients takes a lot more time and has greater potential for delays and mistakes. Difficulty or inability to contact HCPs in other settings contributed to transitional care challenges. HCPs relied heavily on patients’ self-reported information to clarify ambiguities, which was considered error-prone to error and very time-consuming.

The GP doesn't have a discharge summary, does not have an updated medication list and the patient turns up at the appointment that was made by the hospital and they go ‘why are you here?’ and the patient goes ‘well don't you know?!..’ […] Because I guess there is a time lag between the discharge summary being done and actually the GP receiving it. So here you've got the patient trying to explain to the GP what happened to them in hospital to fill in the picture and tell them something while they are at the appointment.

Registered nurse 1

Overall, participants reported that the problems described above evolve because nobody assumes responsibility for optimising care plans and because of poor coordination of care. After reflection, many participants concluded that their current situation did not allow them to shoulder these responsibilities. Only GPs asserted themselves as being the key care coordinators for patients with multimorbidity. However, because of previously described challenges, they also often reported not meeting this responsibility. Geriatricians felt their role was more to support GPs’ optimisation of care, and that potential input to ongoing care coordination was limited due to the short duration of their interaction with these patients.

GPs and nurses found that home medication reviews undertaken in collaboration with pharmacists were helpful to optimise treatment.

I know through my other role, district nursing out at [rural town] – community pharmacist had just recently been to a home to go through all the medications and the patient was really happy and really pleased and really understood things so much better. So absolutely, the community pharmacist is fantastic.

Practice nurse 2

Both GPs and pharmacists complained about the level of government remuneration provided. Pharmacists criticised the standard remuneration for home medication reviews regardless of workload (AUS$210.93 at 2 July 2015). Patients who are most in need of a home medication review and most complex were considered less likely to receive equally detailed reviews because pharmacists seemed unwilling to substantially extend the review duration without additional remuneration.

Potential enablers for improving the quality of care for patients with multimorbidity are presented in Box 2. Issues around team structure, electronic health and IT, proper resourcing and generation of better evidence were prominent.

Box 2.

Recommendations made by healthcare professionals to improve care for patients with multimorbidity and polypharmacy

Working as a close healthcare team and removing the hierarchical structure that puts specialists above generalists.

To overcome the time constraints of medical consultations, nurses reported their willingness to continue working closely with patients to help them prioritise problems and empowering them to take responsibility for their conditions.

Geriatricians recommended greater family involvement to elicit patients’ problems and preferences and ensure that information is communicated correctly to prescribers and the patient.

Most HCPs expressed the view that pharmacists could play a bigger role in continuously reinforcing information to patients in both the hospital and the community setting.

Community pharmacists mentioned they are ideally situated to reinforce information as the last in line before patients are given medications but needed to take more responsibility to ensure delivery of information with every script.

Practical suggestions were raised to overcome the shortcomings of clinical guidelines. These include: addition of warning sections in every guideline detailing if extra attention should be paid when patients are above a certain age or when certain comorbidities exist; software programmes to send out warnings about, e.g. renal impairment, anticholinergic load or interactions; an extension of the evidence base with numbers needed to treat or risks versus benefits in this specific patient population.

More involvement of the hospital pharmacists was identified as helpful for facilitating faster processing of medication lists at discharge. Another highlighted enabler to overcome the problems in transition of care was the development and improvement of communication systems, including the use of e-health systems. To overcome the lack of optimisation of care plans by prescribers, participating GPs tended to refer patients to pharmacists for a medication review when they use a certain number of medications, when optimisation of medication regimens seems necessary or when patients are confused or in need of education about their medications.

HCPs pointed out that there needs to be one key carer who should be involved in optimising care plans and coordination of care.

In some of the explored practices in rural areas, nurses were involved in identifying patients to relieve the burden for GPs.

Discussion

This study provides a comprehensive account of the challenges for healthcare teams in attempting to optimise medication-related aspects of multimorbidity management. While many previous studies identify shortcomings with care for community-dwelling patients with multimorbidity, our findings provide a new multidisciplinary and system-wide perspective. Some challenges with medication use related specifically to potentially inappropriate polypharmacy (e.g. unnecessary adverse outcomes), while others had general relevance to polypharmacy (e.g. difficulty with medication reconciliation). Widespread personal reluctance to assume the responsibilities of multimorbidity management and poor coordination of care was the key challenge identified from a team perspective. Only GPs asserted their responsibility for ensuring care coordination among patients with multimorbidity but often could not meet the relevant responsibilities due to workload pressure and poor communication from the hospital.

Despite growing consensus about the value of patient-centred care and stepwise management of multimorbidity [14], our findings suggest that considerable system-wide implementation gaps remain. The lack of individuals systematically taking responsibility for key aspects of multimorbidity care and the absence of coordinated care emerged as dominant issues. Addressing these issues is not emphasised in AGS guidelines. The Ariadne principles were published subsequent to our research, with extensive input from experts internationally [22], and build upon AGS principles to make recommendations about multimorbidity management with specific reference to primary care. Care coordination and the effects of referral on burden of treatment are given considerable emphasis in the Ariadne statement, possibly validating our decision to add a domain around care coordination.

Findings such as reluctance to ‘interfere’ with prescribing by other HCPs, resulting in treatment inertia, and incomplete medical history may disproportionately affect patients with multimorbidity because they are more likely to require multiple prescribers. Participants relied on patient recall for medical histories, which may be less reliable for more complex multimorbid cases. Likewise, HCPs appeared to restrict their focus of care to acute presenting issues as a means of avoiding complex multimorbidity challenges related to inadequate documentation of care or ambiguity around management decisions.

Previous research has explored perspectives of GPs and practice nurses regarding multimorbidity [8] and to a lesser extent those of pharmacists and hospital physicians [15, 18]. Our findings add an important system-wide and multidisciplinary perspective. For example, seeking the view of multiple HCPs allowed an understanding of how the absence of coordination or leadership led to general inaction on multimorbidity. These problems are largely not amenable to resolution by individual HCPs. They need macrosystem-led changes for guideline implementation to support clinical teams. Recent initiatives such as the Australian Deprescribing Network may provide necessary leadership.

Primary care capacity to accommodate the needs of patients with multimorbidity and polypharmacy must be increased in light of the growing prevalence of this patient group [23]. Overall, the evidence base for interventions to improve management of multimorbidity is limited [13]. A systematic review posits that multifaceted organisational interventions involving dedicated care coordinators may be beneficial, but these benefits appear more clearly established for the management of specific comorbidities rather than general management of multimorbidity [13]. Our participants recommended the appointment of care coordinators to facilitate better communication between settings, an innovation that seems both feasible and effective in primary care [24] and a preferred option for older people with multimorbidity [7].

Our study suggests that pharmacist-delivered clinical medication review may be acceptable to support patient coordination from a medication perspective, but remuneration systems may require amendment to optimise care of complex cases. Studies exploring health outcomes following medication reviews have conflicting results [25], possibly explained by variation in the interventions and patient groups investigated. Interestingly, our participants did not cite examples of other available services potentially suited to addressing polypharmacy. Options included geriatrician-led comprehensive geriatric assessment [26], GP health assessments for people aged 75 years and older [27] and home nursing programmes.

Patients’ experiences of models for multimorbidity management also remain poorly understood and warrant extensive exploration for health system redesign [6]. In keeping with literature, patient preferences and capacity to implement recommendations did not appear to strongly influence HCPs where clinical uncertainty exists as to the preferred course of action [28]. Our findings that patient difficulties, concerns and conflicting prescriber instructions more commonly become apparent in downstream settings during discussions with other HCPs have been reported previously [29]. Developing mechanisms to ensure that prescribers have the resources, skills and motivation to engage patients at the point of making treatment decisions might therefore deliver more efficient healthcare processes as well as improving patient-centeredness.

Strengths and limitations

Further research is needed to investigate the generalisability of our findings and system-wide strategies needed to promote the uptake of best practice for multimorbidity management. This is particularly relevant in light of widely varying models for care and care coordination, both in Australia and internationally. In line with the qualitative methodology, sampling strategy was designed to achieve a maximum variation of participants rather than being numerically representative. The pharmacy background of the interviewer might have biased the direction of conversations or promoted socially desirable responses regarding pharmacist-delivered services.

In conclusion, participants perceived multimorbidity management including medication use as important but it was typically relegated to a secondary priority in the face of more acute patient needs. Interprofessional communication seemed to be better coordinated within hospitals but less likely to focus on the ongoing patient needs compared with primary care. GPs were concerned by continuity of care, but time pressures and siloed care appeared to inhibit translation into practice. System-wide initiatives to resource care coordination may support leadership and increase capacity in this area.

Key points

• Multimorbidity is acknowledged by healthcare professionals as important but acute issues take priority more often than not.

• Healthcare professionals do not regularly use recommended strategies for optimising management of multimorbidity.

• The inherently complex nature of multimorbidity increases potential for communication problems and inadequately resourced care.

• Poorly defined roles and responsibilities for the management of multimorbidity engender reluctance to assume clinical leadership.

Acknowledgements

K.M.’s salary during this project was supported through a postdoctoral fellowship from the National Heart Foundation of Australia.

Authors’ contributions

All authors made significant contributions to the study concept and design, and reviewed several manuscript drafts. B.B. conducted all interviews. B.B. and H.A. independently coded data, and K.P.M.N. facilitated resolution of coding issues. B.B., K.P.M.N. and H.A. led the initial drafting of the paper.

Conflicts of interest

None declared.