Subscribe
×

Preference: General Medicine

312

First do no harm: a real need to deprescribe in older patients

Deliberate yet judicious deprescribing has considerable potential to relieve unnecessary medication-related suffering and disability in vulnerable older populations. Polypharmacy — variously defined as more than five or up to 10 or more medications taken regularly per individual — is common among older people (defined here as those aged 65 years or over). Depending on the circumstances, including why and how drugs are being administered, polypharmacy can be appropriate (potential benefits outweigh potential harms) or inappropriate (potential harms outweigh potential benefits).

While many older people benefit immeasurably from multiple drugs, many others suffer adverse drug events (ADEs). In Australia, one in four community-living older people are hospitalised for medication-related problems over a 5-year period1 and 15% of older patients attending general practice report an ADE over the previous 6 months.2 At least a quarter of these ADEs are potentially preventable.2 Up to 30% of hospital admissions for patients over 75 years of age are medication related, and up to three-quarters are potentially preventable.3 Up to 40% of people living in either residential care4 or the community5 are prescribed potentially inappropriate medications. In both hospital and primary care settings, about one in five prescriptions issued for older adults are deemed inappropriate.6

Polypharmacy in older people is associated with decreased physical and social functioning; increased risk of falls, delirium and other geriatric syndromes, hospital admissions, and death; and reduced adherence by patients to essential medicines. The financial costs are substantial, with ADEs accounting for more than 10% of all direct health care costs among affected individuals.7

The single most important predictor of inappropriate prescribing and risk of ADEs is the number of medications a person is taking; one report estimated the risk as 13% for two drugs, 38% for four drugs, and 82% for seven drugs or more.8 One in five older Australians are receiving more than 10 prescription or over-the-counter medicines.9 People in residential aged care facilities are prescribed, on average, seven drugs,10 while hospitalised patients average eight.11 While hospitalisation might provide an opportunity to review medication use, on average, while two to three drugs are ceased, three to four are added.12

In responding to polypharmacy-related harm, a new term has entered the medical lexicon: deprescribing. This is the process of tapering or stopping drugs using a systematic approach (Box).13 Inculcating a culture of deprescribing will be challenging for several reasons.

Barriers to deprescribing

Underappreciation of the scale of polypharmacy-related harm

Clinicians easily recognise clinically evident ADEs with clear-cut causality. In contrast, ADEs mimicking syndromes prevalent in older patients — falls, delirium, lethargy and depression — account for 20% of ADEs in hospitalised older patients but go unrecognised by clinical teams.14

Increasing intensity of medical care

The drivers of polypharmacy are multiple: disease-specific clinical guideline recommendations, quality indicators and performance incentives;15 a focus on pharmacological versus non-pharmacological options; and “prescribing cascades”, when more drugs are added in response to onset of new illnesses, including ADEs misinterpreted as new disease.

Narrow focus on lists of potentially inappropriate medications

Lists of potentially inappropriate medications (such as the Beers, McLeod and STOPP [Screening Tool of Older Persons’ Potentially Inappropriate Prescriptions] criteria)16 comprise drugs whose benefits are outweighed by harm in most circumstances. Examples include potent opioids used non-palliatively, non-steroidal anti-inflammatory agents, anticholinergic drugs and benzodiazepines. However, such drugs account for relatively few ADEs in Australian practice.17 In contrast, commonly prescribed drugs with proven benefits in many older people, such as cardiovascular drugs, anticoagulants, hypoglycaemic agents, steroids and antibiotics, are more frequently implicated as a result of misuse.18

Reluctance of prescribers and ambivalence of patients

Many prescribers feel uneasy about ceasing drugs for several reasons:19,20 going against opinion of other doctors (particularly specialists) who initiated a specific drug and which patients appear to tolerate; fear of precipitating disease relapse or drug withdrawal syndromes or patient approbation at having their care “downgraded”; paucity of precise benefit–harm data for specific medications to guide selection for discontinuation; apprehension around discussing life expectancy versus quality of life in rationalising drug regimens; limited time and remuneration for reappraising chronic medications, discussing pros and cons with patients, and coordinating views and actions of multiple prescribers.

While agreeing in principle to minimising polypharmacy, patients simultaneously express gratitude that their drugs may prolong life, abate symptoms and improve function. Fear of relapsing disease lessens patients’ desire to discontinue a drug, especially if combined with negative past experiences. Inability to see the personal relevance of harm–benefit trade-offs or lack of trust in their doctors’ advice about drug inappropriateness and the safety of discontinuation plans also act against deprescribing.21

Possible solutions

Ultimately, deprescribing is a highly individualised and time-consuming process, but several facilitative strategies are offered.

Reframe the issue

Both the patient and clinician should regard deprescribing as an attempt to alleviate symptoms (of drug toxicity), improve quality of life (from drug-induced disability) and lessen the risk of morbid events (ADEs) in the future. Its objective is not simply to reduce the burden, inconvenience or costs of complex drug regimens, although these too are potential benefits. Deprescribing should not be seen by patients as an act of abandonment but one of affirmation for highest quality care and shared decision making. Compelling evidence that identifies circumstances in which medications can be safely withdrawn while reducing the risk of ADEs and death needs to be emphasised.22–25

Discuss benefit–harm trade-offs and assess willingness

In consenting to a trial of withdrawal, patients need to know the benefit–harm trade-offs specific to them of continuing or discontinuing a particular drug. Patient education initiatives providing such personalised information can substantially alter perceptions of risk and change attitudes towards discontinuation.26

Target patients according to highest ADE risk

In addition to medication count, other ADE predictors include past history of ADEs, presence of major comorbidities, marked frailty, residential care settings and multiple prescribers.

Target drugs more likely to be non-beneficial

Such drugs, as initial targets for deprescribing, consist of five types.

  • Drugs to avoid if at all possible (comprising those previously cited in “drugs to avoid” lists16).
  • Drugs lacking therapeutic effect for substantiated indications despite optimal adherence over a reasonable period of time.
  • Drugs that patients are unwilling to take for various reasons, including occult drug toxicity, difficulty handling medications, costs and little faith in drug efficacy.
  • Drugs lacking a substantiated indication, either because the original disease diagnosis was incorrect — as often occurs with regard to heart failure, Parkinson disease and depression — or because the stated diagnosis-specific indication is invalid owing to evidence of harm or no benefit. Importantly, indications or treatment targets derived from studies involving younger patients or highly selected, older individuals with a single disease may be inappropriate for older, frail patients with multimorbidity.
  • Drugs that are unlikely to prevent disease events within the patient’s remaining life span. The question here is not how much benefit drugs aimed at preventing future disease events (such as statins and bisphosphonates) may confer, but when. If the time until benefit exceeds the patient’s estimated life span, no benefit will result, while the ADE risk is constant and immediate. Undertaking such reconciliations is facilitated by accurate estimation of longevity using easy-to-use web-based mortality indices (eg, see http://eprognosis.ucsf.edu) and drug-specific time until benefit using trial-based time-to-event data.27

Access and apply specific discontinuation regimens

Detailed descriptions of how clinicians can safely withdraw specific drugs and monitor for adverse effects can be found within various articles and websites.28,29

Foster shared education and training

Interactive, case-based, interdisciplinary meetings involving multiple prescribers, combined with deprescribing advice from clinical pharmacologists or appropriately trained pharmacists have been shown to be beneficial with regard to doctors’ deprescribing practice.30

Extend the time frame with the same clinician

Several patient encounters with the same overseeing generalist clinician, focused on one drug at a time, can provide repeated opportunities to discuss and assuage a patient’s fear of discontinuing a drug and to adjust the deprescribing plan according to changes in clinical circumstances and revised treatment goals. Deprescribing can be remunerated under extended care or medication review items, with patients’ full awareness that such consultations are expressly for reviewing medications.

Conclusion

Inappropriate polypharmacy and its associated harm is a growing threat among older patients that requires deliberate yet judicious deprescribing using a systematic approach. Widespread adoption of this strategy has its challenges, but also considerable potential to relieve unnecessary suffering and disability, as embodied in that basic Hippocratic dictum — first do no harm.

Core themes in deprescribing13

  • Ascertain all current medications (medication reconciliation)
  • Identify patients at high risk of adverse drug events
    • number and type of drugs
    • patient characteristics (physical, mental, social impairments)
    • multiple prescribers
  • Estimate life expectancy in high-risk patients
    • with focus on patients with prognosis equal to or less than 12 months
  • Define overall care goals and patient values and preferences in the context of life expectancy
    • relativity of symptom control and quality of life v cure or prevention
    • patient preference for aggressive v conservative care; willingness to consider deprescribing
  • Define and validate current indications for ongoing treatment
    • medication–diagnosis reconciliation
    • drugs lacking valid indications are candidates for discontinuation
  • Determine time until benefit for preventive or disease-modifying medications
    • drugs whose time until benefit exceeds expected life span are candidates for discontinuation
  • Estimate magnitude of benefit v harm in absolute terms for individual patient
    • drugs with little likelihood of benefit and/or significant risk of harm are candidates for discontinuation
  • Implement and monitor a drug deprescribing plan
    • ongoing reappraisal for illness relapse or deterioration or withdrawal syndromes
    • communication of plan and shared responsibility among all prescribers
    • supervision of drug discontinuation by single generalist clinician

Are high coronary risk patients missing out on lipid-lowering drugs in Australia?

Lipid-lowering drugs (LLD), especially statins, are of proven benefit in preventing future coronary heart disease (CHD), both recurrent events and first events in those at high coronary risk.1–3 The Organisation for Economic Co-operation and Development (OECD) noted that consumption of LLD in Australia in 2011 was the highest of 23 countries reported.4 The rate was 50% higher than the OECD average and had risen more than 300% since 2000.4

Given this high level of use, it is important to know whether LLD are being prescribed for the correct mix of patients. The Australian ACACIA registry reported that statins were prescribed for 75%–89% of patients with acute coronary syndrome in 2005–2007, with the rate varying depending on the clinical presentation.5 Similarly, a large European survey of patients with CHD reported that 89% had been prescribed statins in 2006–2007,6 while a companion survey in the general practice setting found that 47% of “high-risk” patients with hypercholesterolaemia had been prescribed statins.7

The AusHEART study, an Australian general practice survey of risk factor perception and management in 2008, found that 50% of patients with established cardiovascular disease were prescribed a combination of statin, antihypertensive and antiplatelet therapy.8 Only a third of patients without established cardiovascular disease but at high risk of a first event were prescribed statin and antihypertensive medication.8 The AusDiab Study in 2011–2012 reported that 60% of people with diabetes were using statins.9 None of these outcomes take into account the poor long-term persistence in patients prescribed these and other cardiovascular drugs.10–12

In previous studies, we have used the Pharmaceutical Benefits Scheme (PBS) database to explore patient behaviour in those prescribed various medications.10–12 The database also contains information on co-prescriptions, which can be used as a surrogate for accompanying medical conditions. Here, we used the PBS database to explore whether patients arbitrarily defined as being at high coronary risk (those with prior CHD, diabetes or hypertension) are receiving LLD as they should, according to contemporary prevention guidelines.2,3

Methods

Data source

Dispensing is only recorded in the PBS database for patients classified as concession card holders, who are nevertheless estimated to represent 65% of all patients receiving statins.13 We analysed PBS pharmacy payment claim records for a 10% random sample of the included population. The data were drawn from de-identified records held by Medicare Australia, via the Department of Human Services, for the period January 2006 through May 2013, inclusive. Various statin drugs were priced below the general patient copayment threshold for some or all of this period, and no record of their prescription would have been sent to the PBS. Hence, our study was limited to patients who had received > 90% of their PBS prescriptions on a concessional basis during the study period, indicating that they were long-term concession card holders. Prescriptions for LLD included statins, fibrates and ezetimibe, although statin drugs were predominant.

Definition of surrogates

Groups of patients at high risk of future CHD were arbitrarily defined by the following co-prescriptions:

  • CHD: antiplatelet drugs (clopidogrel, prasugrel, not solo aspirin) and anti-anginal drugs (nitrates, nicorandil, perhexiline)
  • Diabetes: all standard drugs
  • Hypertension: all standard drugs (not solo diuretics).

To be classified in a high-risk group, patients needed to have three prescriptions for the specified drugs within 6 months at any time during the study period. Patients could belong to multiple groups.

Ethics approval

Patient identities remained anonymous during this investigation. Ethics and publication approval was obtained from the External Request Evaluation Committee of Medicare Australia.

Results

We extracted information from the PBS claims database on 853 836 concessional patients who had received PBS drugs during the study period (representative of 8 538 360 patients nationally). Of these, we classified 276 212 (32%) as being at high coronary risk. A comparison of the Australian population distribution with the concessional and high coronary risk populations is shown in Box 1. Compared with the Australian population, the distribution of concessional patients was shifted towards older age groups. High coronary risk patients were older than the overall concessional group (mean age, 66.1 [SD, 14.8] years, 44% male v 47.9 [SD, 28.0] years, 45% male).

Of the total concessional group, 657 454 patients (77%) were not prescribed any LLD during the study period. Of the total high coronary risk group, 115 477 (42%) were not prescribed any LLD.

Among the clinical groups of high-risk patients not receiving LLD, there were minor variations by age and sex (Box 2). For patients in a single risk-factor group, the proportion not receiving LLD was lowest in the CHD group (40%). For patients in multiple risk-factor groups, the proportions not receiving LLD were lower, down to 8% in the CHD + diabetes + hypertension group. Among all CHD groups combined, 19% of patients were not receiving LLD.

The proportions of patients in high-risk groups not receiving LLD are shown by age groups in Box 3. Across all risk groups, the proportions not receiving LLD were generally higher in the youngest and oldest age groups and lowest in those aged 51–70 years. This U-shaped relationship in the proportions not receiving LLD by age is clearly apparent for the largest multiple risk-factor groups in Box 4.

Discussion

Australia’s national guidelines for the primary or secondary prevention of CHD provide a general framework for appropriate management of all risk factors.2,3 However, such therapy needs to be individualised according to background risk, prognosis, comorbidities, drug tolerance, lifestyle and living circumstances, and personal choices.2,3 Hence, there can be no simple threshold for the proportion of people not treated that would signify undertreatment. Using conservative definitions for high coronary risk, we identified that the proportions of patients with multiple risk factors who were not receiving LLD are generally low, particularly in the CHD + diabetes + hypertension risk group.

Our findings are broadly consistent with Australian data for patients with CHD5 or diabetes,9 as well as data from Europe.6,7 The AusHEART study reported much higher proportions of patients with cardiovascular disease or at high risk of disease who were not receiving statins, antihypertensive therapy or antiplatelet drugs,8 but the findings were derived from general practice surveys and are not strictly comparable with administrative data from the PBS database.

We found that the concessional patient population was older than the Australian population and, not surprisingly, that the high coronary risk population was older still. The proportions of patients aged 71–80 and ≥ 81 years in the high-risk population were about fivefold those in the Australian population, reflecting a greater presence or risk of disease. The finding that greater proportions of those aged ≥ 81 years, and to a lesser extent those under 41 years, were not receiving LLD across all clinical groups is intriguing. This may represent a form of age discrimination and may not be consistent with national guidelines.2,3 The PBS database contains no further clinical information that might explain the behaviour of prescribers, but if there is a genuine bias against lipid-lowering therapy in these age groups, further research and education will be required.

There are limitations in our analysis. It was restricted to concessional patients; however, we have previously noted that concession card holders account for two-thirds of statin use.13 As the PBS database is administrative, with no relevant clinical information, we had no information on how well risk factors are controlled in these patients. The clinical definitions of high coronary risk we employed were conservative and would have good ability to identify high-risk patients, but would be less reliable in identifying those not at high risk. Very few patients using the nominated drugs would be misclassified as having CHD. There is minimal use of metformin in patients without diabetes (much less so in older patients), while some antihypertensive drugs are used for other indications, notably CHD. The major strength of our analysis is the use of a large, nationwide database.

Despite our study’s limitations, we also examined the reciprocal question of whether too many “low-risk” patients are receiving LLD. After subtracting data for the high-risk group from the total dataset of 853 836 concessional patients, we calculated that only 7% of those not identified as being at high coronary risk were receiving LLD. At limited face value, this is an encouraging statistic, given the high use of statin drugs in Australia.4

Our study suggests a large proportion of patients at high coronary risk, especially those with CHD and multiple risk factors, are being appropriately prescribed LLD in Australia. However, long-term persistence of therapy remains problematic10 and we have no information on how well risk factors are controlled. While use of LLD in Australia may be very high, it appears that middle-aged concession card holders at high coronary risk are being well managed.

1 Comparison of Australian population distribution with concessional and high coronary risk populations, by age*

 

Age group (years)

Population

< 41

41–50

51–60

61–70

71–80

≥ 81

Australian population (n = 22 710 352)

12 546 208 (55%)

3 143 550 (14%)

2 827 714 (12%)

2 165 462 (10%)

1 255 465 (6%)

771 953 (3%)

Concessional patients (n = 853 836)

367 141 (43%)

73 628 (9%)

63 695 (7%)

90 591 (11%)

135 129 (16%)

123 618 (14%)

High coronary risk patients (n = 276 212)

11 429 (4%)

17 122 (6%)

35 793 (13%)

88 210 (32%)

82 816 (30%)

40 842 (15%)

* Estimated Australian population at 30 June 2012.14 The other two groups are based on a 10% random sample of the Pharmaceutical Benefits Scheme database. Percentages refer to proportion of the population group (eg, 55% of the Australian population were aged < 41 years).

2 Mean age, sex and proportions of high coronary risk patients not receiving lipid-lowering drugs (LLD), by clinical group

Clinical group

No. of patients

Mean age (SD), years

Male

Not receiving LLD

CHD alone

6 696

74 (13)

3 375 (50%)

2 676 (40%)

Diabetes alone

12 713

68 (13)

6 191 (49%)

7 007 (55%)

Hypertension alone

166 094

53 (20)

66 105 (40%)

88 942 (54%)

Diabetes + hypertension

39 772

65 (12)

18 971 (48%)

8 543 (21%)

CHD + diabetes

853

74 (11)

465 (55%)

215 (25%)

CHD + hypertension

35 827

71 (13)

17 161 (48%)

6 983 (19%)

CHD + diabetes + hypertension

14 257

70 (10)

7 884 (55%)

1 111 (8%)

All CHD groups

57 633

66 (15)

28 817 (50%)

10 985 (19%)

CHD = coronary heart disease.

3 Numbers and proportions of high coronary risk patients not receiving lipid-lowering drugs, by clinical and age groups*

 

Age group (years)

Clinical group

< 41

41–50

51–60

61–70

71–80

≥ 81

CHD alone

            

No. (%)

56/104 (54%)

88/239 (37%)

141/512 (28%)

312/1480 (21%)

627/2023 (31%)

1452/2338 (62%)

P

0.000

0.000

0.003

reference

0.000

0.000

Diabetes alone

          

No. (%)

2680/3272 (82%)

981/1894 (52%)

871/2118 (41%)

1131/2822 (40%)

795/1800 (44%)

549/807 (68%)

P

0.000

0.000

ns

reference

0.006

0.000

Hypertension alone

          

No. (%)

4981/6173 (81%)

7537/10 579 (71%)

12 297/21 806 (56%)

27 210/56 274 (48%)

22 310/48 471 (46%)

14 607/22 791 (64%)

P

0.000

0.000

0.000

ns

reference

0.000

Diabetes + hypertension

          

No. (%)

546/1451 (38%)

794/3049 (26%)

1429/7147 (20%)

2394/14 120 (17%)

2207/10 901 (20%)

1173/3104 (38%)

P

0.000

0.000

0.000

reference

0.000

0.000

CHD + diabetes

          

No. (%)

4/17 (24%)

7/41 (17%)

13/109 (12%)

27/201 (13%)

65/275 (24%)

99/210 (47%)

P

ns

ns

reference

ns

0.010

0.000

CHD + hypertension

          

No. (%)

45/279 (16%)

97/823 (12%)

221/2418 (9%)

864/8789 (10%)

2098/13 918 (15%)

3658/9600 (38%)

P

0.000

0.028

reference

ns

0.000

0.000

CHD + diabetes + hypertension

        

No. (%)

13/133 (10%)

23/497 (5%)

64/1683 (4%)

180/4524 (4%)

402/5428 (7%)

428/1992 (21%)

P

0.001

ns

reference

ns

0.000

0.000

All CHD combined

          

No. (%)

118/533 (22%)

215/1600 (13%)

439/4722 (9%)

1384/14 994 (9%)

3192/21 644 (15%)

5637/14 140 (40%)

P

0.000

0.000

reference

ns

0.000

0.000

CHD = coronary heart disease. ns = not significant (P > 0.05). * Percentages refer to the proportion of the clinical group. There was a pairwise comparison in each clinical group with the reference age category. The age category with the smallest percentage was used as the reference, employing a two-tailed Z test.

4 Proportions of high coronary risk patients in selected multiple risk-factor groups not receiving lipid-lowering drugs, by age group

CHD = coronary heart disease.

Excessive occupational sitting is not a “safe system of work”: time for doctors to get chatting with patients

Employers and doctors need to recognise and respond
to the health hazards of too much sedentary time

Being able to work usually has a positive impact on health. However, changes in the physical demands of work and increased use of computers have led to many workers now being employed in sedentary jobs.1 While these have traditionally been thought of as safe work environments, recent evidence (including meta-analyses2) suggests this mode of work — often involving long uninterrupted periods of sitting — may be hazardous, contributing substantially to the growing chronic disease burden associated with obesity, diabetes, cardiovascular disease and cancer.3,4 Importantly, being sedentary (ie, too much sitting) is not the same as being physically inactive. Insufficient physical activity is defined in the public health context as not meeting the guidelines to accumulate at least 2.5 to 5 hours of moderate-intensity physical activity per week.5 Both physical inactivity and sedentary time have an impact on health: physical inactivity is estimated to account for 5.5% of all-cause premature mortality, and excessive sitting time, after adjusting for physical activity, accounts for 5.9%.2 Notably, even if workers meet physical activity guidelines (ie, are physically active), they can still have high exposure to sedentary time.

Work health and safety laws in Australia and other jurisdictions require employers to provide a “safe system of work”. For example, section 19 of the Work Health and Safety Act 2011 (Cwlth) states that the “primary duty of care” is to “ensure, so far as is reasonably practicable, the health and safety of workers” by, among other things, “provision and maintenance of safe systems of work”. We contend that: a) the systems of work commonly observed in contemporary offices demonstrate a high likelihood of excessive sitting hazard; b) the degree of harm associated with this hazard is likely to be substantial; c) the evidence for this is now widely known; d) there are available ways to minimise the risk; and e) the cost of these strategies is proportionate to the risk.

Exposure to sedentary time can be substantial for office workers. Objective measurement has found that, on average, over 75% of the office workday is spent sitting, with much of this accumulated in unbroken bouts of at least 30 minutes,3,6,7 demonstrating the hazard is much more common than previously thought.

There is now also evidence that both overall sedentary time and the pattern of sedentary exposure are associated with substantial harm. For example, excessive total sitting time is associated with premature mortality, obesity, cancer (ovarian, endometrial and colon), type 2 diabetes and cardiovascular disease.2–4 Conversely, regular interruption of sedentary time is beneficially associated with biomarkers for chronic conditions.8,9 Occupational exposure can account for around half of total weekly sedentary time,3,7 and thus fractional risk attribution implies occupational responsibility for around half of the harm.

The evidence for the harm associated with excessive sitting is now being widely promoted. The recently updated Australian Government guidelines recommend minimising the time spent in prolonged sitting.5 Further, media stories reported in recent years have highlighted the potential dangers of excessive sitting.10 Advanced organisations are implementing risk control strategies using, for example, a sedentary work practices toolkit.11

Changes to work systems can reduce sedentary time. Alterations to the individual physical environment (eg, sit–stand workstations or active workstations)12,13 and combined approaches (including individual, environmental and organisational changes)6 have achieved substantial reductions in total occupational sitting time and prolonged unbroken sitting time. The extent to which such changes can minimise chronic disease is now being investigated.

Some risk reduction strategies, such as introducing standing meetings, are costless. While other strategies have a cost, the cost does not seem disproportionate given the potential for significant harm (eg, reduced life expectancy and cardiovascular disease).

On the available evidence, it appears that contemporary offices may be failing to provide a safe system of work. Doctors should be prescribing behaviour to reduce occupational sedentary exposure where this may exacerbate, or be exacerbated by, an existing medical condition. A doctor who is aware that a patient has a prolapsed disc in the spine would require the patient to refrain from lifting heavy objects at work. In the same way, a doctor who is aware that a patient’s cardiovascular condition necessitates remaining active and avoiding excessive sedentary exposure should inform the patient and employer of the need for the patient to regularly move to maintain wellbeing. This could be achieved through the redesign of work tasks to enable postural variability or through regular breaks that involve activity such as walking. In addition to raising this matter in routine patient consultations, doctors could be supporting organisations attempting to provide systems of work in which workers “sit less, move more, and move more often”.14

A systematic approach to chronic heart failure care: a consensus statement

Chronic heart failure (CHF) remains a major public health problem. CHF is not a static syndrome; individuals with CHF are at high risk of progressive cardiac dysfunction resulting in either sudden cardiac death or acute hospitalisation. Despite significant advances in CHF management, clinical outcomes are poor and associated with escalating health care costs.1 Worldwide, there are an estimated 23 million people living with CHF and 5.7 million new cases each year.2 With limited Australian data available, the Australian Institute of Health and Welfare has used overseas rates to suggest that 30 000 patients are diagnosed with incident heart failure annually and 300 000 people are living with CHF in Australia.3 The prevalence of CHF continues to rise as the population ages and survival from cardiovascular disease continues to improve.

Between 2006 and 2011, deaths from CHF in Australia rose by 20%.4 CHF prognosis remains poorer than that for common forms of cancer (in terms of individual survival and population life-years lost) for men and women.5,6 Hospital separations for CHF increased by 24% between 2002–03 and 2011–12.7 In the 2007–08 financial year, CHF was a primary diagnosis in 45 212 hospitalisations and a contributory diagnosis in 94 599 hospitalisations.8

The annual cost of CHF in Australia has been estimated at over $1 billion per year, with hospital care being the largest expenditure.9 A significant proportion of this cost is associated with preventable CHF readmissions. Readmissions within 30 days of discharge can be as high as 20%–27%.10,11 Overall, reported rates for readmission with CHF within 3–12 months of initial discharge are between 29% and 49%.12,13

Given the high rate of readmissions, there is profound potential to improve CHF-related outcomes, at both individual and societal levels, through improved quality of care and system change.14

Practice gaps

Recent studies have highlighted significant variations in access to evidence-based care for patients with CHF.15–17 Many individuals are not diagnosed in a timely manner and subsequent management is suboptimal. Initial diagnostic delay is often due to under-recognition of early heart failure symptoms. This is compounded by limited availability of public sector services, particularly in rural and remote areas. Additionally, patient data are not shared across health services, largely due to lack of integrated information systems and care coordination.

These problems are amplified among marginalised populations. CHF is 1.7 times more common, and occurs at a younger age, among Aboriginal and Torres Strait Islander peoples than among other Australians.18 Aboriginal and Torres Strait Islander peoples are also more likely to die from CHF, and their rate of preventable CHF-related hospitalisations is three times higher than for non-Indigenous Australians.18 Such health disparities frequently occur due to poor access to evidence-based care. Availability of culturally appropriate services that provide earlier prevention, detection and management of CHF needs to be improved.

Among people hospitalised with CHF, those who receive evidence-based, multidisciplinary care have better health outcomes than those who do not.19,20 Current Australian guidelines articulate the evidence-based practices necessary to improve care delivery.21 However, the management of CHF remains a pressing concern, with many apparent indicators of poor case detection, non-guideline-based management, poor coordination and communication, and recurrent hospital admission.15–17,22

In this consensus statement, our aim is to guide the policy and associated system changes required to support delivery of evidence-based care. This is not intended as a prescriptive guideline, rather a set of principles to assist health departments, health network administrators, clinicians and consumers in improving care systems for people living with CHF. Our intended audience is policymakers, health system managers, consumers and health professionals in acute and primary care, including cardiologists, general practitioners, nurses, dietitians and other members of the multidisciplinary team. The consensus development process is outlined in Box 1.

The expert panel identified four themes and five principles to inform the consensus statement (Box 2). The recommendations (Box 3) based on these themes have the potential to reduce the likelihood of emergency presentations, hospitalisations and premature death among patients with CHF.

Chronic heart failure care model

Best-practice management of CHF involves evidence-based, multidisciplinary, patient-centred care, which leads to better health outcomes.20,23 Patient-centred care is respectful of and responsive to preferences, needs and values of patients and consumers and should include dimensions of respect, emotional support, physical comfort, information and communication, continuity and transition, care coordination, involvement of family and carers, and access to care.23 Greater care coordination is needed because fragmentation across health care, long-term care and other social support systems effectively impedes a patient-centred focus.24 Research suggests that providing incentive payments through primary care payment schemes may improve care continuity and transition, as would streamlining funding for delivery of different levels of care.24

Multidisciplinary CHF care is distinguishable from generic chronic disease management programs by the special needs of these patients, which necessitate specialised evidence-based treatment strategies associated with optimal outcomes.25,26 Considerations include management of severely ill CHF patients, symptom monitoring, implementation of a range of self-management strategies and titration of medications.

Against a background of recent national health reform linking improved person-centred care with performance and funding arrangements, there is increasing interest in how to realign care systems accordingly. Research suggests that individuals value easy access to services, coordinated care, and information and honesty about their prognosis.27,28 Patient or consumer charters and informed consent policies have been introduced, but there is limited evidence that consumer engagement influences change in care delivery.29 Consumer engagement has been strengthened by the recent introduction of the National Safety and Quality Health Service Standards, which include a component on consumer partnership.30 The Standards deliver a framework that health organisations can use to actively engage and partner with consumers to strengthen health service delivery.30

Research suggests care coordination problems are greatest at the interfaces between health care sectors and between providers.24 Multidisciplinary care can overcome some of these barriers, as can pooling resources between sectors for care coordinators.24 Multidisciplinary program delivery needs to be appropriate to local needs, resources, patient preferences and disease trajectory phase, as well as across a range of delivery models, including home-based, clinic-based and telephone-based approaches, or a hybrid of these.17,31,32

CHF management plans that include a multidisciplinary approach are vital to educate and empower individuals and their carers to manage this challenging condition.11,33 Given limited resources, a risk assessment tool that stratifies patients at higher risk of readmission could be used to ensure those most likely to benefit from a management program are targeted.34 These plans should be clear about responsibilities among health care providers.

There are cardiac clinical networks in most Australian states and territories (Queensland, New South Wales, Victoria, South Australia, Western Australia, Northern Territory) that have championed access to evidence-based care for CHF patients. These networks have significant influence in improving care systems and outcomes, and can evaluate variance in care quality within and across jurisdictions, with the authority to develop funding models, including care packages. They can facilitate improvements in CHF care by fostering awareness, communication, partnerships and links; by engaging leaders across sectors; and by providing advice and advocacy for policy, planning and funding. The expert panel identified the clinical networks’ essential role in ensuring systematic delivery of a multidisciplinary care model, and concluded this role should be strengthened and further developed. Although multidisciplinary CHF management programs exist across Australia, ensuring access for all patients who would benefit remains a challenge. This is particularly so for Aboriginal and Torres Strait Islander peoples, those from non-metropolitan areas and lower socioeconomic backgrounds, and culturally and linguistically diverse populations.16,35

Implementation of patient-centred care approaches can also have clinical and operational benefits, through less frequent readmissions and improved clinician and patient satisfaction.36–38 Other benefits include reduced emergency department re-presentations, fewer medication errors, higher functional status and improved evidence-based clinical care.38,39 There continues to be a large degree of heterogeneity between CHF programs,40 with some delivering high-quality complex care and others a simplistic program with minimal interventions. Research has shown that this has an impact on patient outcomes,41 and national guidelines have been developed to reduce this heterogeneity in Australia.21 Minimum accreditation standards are important for assessing multidisciplinary care services and reporting on best practice.16,42

Access to meaningful data for management and benchmarking

There is a paucity of Australian data on CHF, resulting in reliance on extrapolation of overseas research. Lack of identification of people with symptomatic CHF prevents efficient patient monitoring. Expansion of cardiac registries to include patients with CHF could improve identification. Recall between health care providers to ensure appropriate assessments and treatments are completed at pre-agreed intervals is also often uncoordinated. An electronic health record potentially offers the ideal tool to track, document and supply CHF patients or their carers and health care providers with the appropriate health care information, on demand, to optimise care.

Further, we do not have standardised outcomes to measure and evaluate care effectiveness and enable international and national benchmarking activity. The definition of a quality indicator must be specific, complete, clearly worded and verified across different user groups.43 Another barrier to measuring standardised outcomes is poor data system compatibility across and within health services, which prevents efficient transfer of data and results in duplication of patient data collection. These problems could be reduced through the use of better process measures.

Increasingly, hospital readmission is becoming an important indicator of health care outcomes, as it can be used to identify potentially preventable admissions. However, as a sole indicator, it can be problematic due to difficulties with interpretation, utility in the clinical environment, and problems such as poor attendance at outpatient clinics, which does not necessarily reflect poor hospital care. An operational definition for readmission needs to clearly identify the diagnosis-related group or major disease classification associated with the index admission.

Over a longer period, as readmissions occur due to the chronic nature of the disease, event-free survival provides a measure of quality.44 An event is defined as an emergency presentation, hospitalisation or premature death within any 12-month period.45

Workforce planning

Workforce needs are likely to be driven by the ageing population and associated disability rates, as well as changing technology, increased burden of disease and community expectations.46

One of the main challenges to workforce planning is providing access to services outside large cities. In 2006, Australian capital cities hosted 93% of CHF management programs, despite 40% of the known population with CHF living outside these cities.47 Policies that guide specialisation or multiskilling in the health workforce will become increasingly important.48

GPs should be empowered to lead care for patients with CHF. This may be through the introduction of funding incentives or provision of nurse practitioners and practice nurses in primary care. Any incentive arrangements should favour provision of care for Aboriginal and Torres Strait Islander peoples, those from lower socioeconomic backgrounds and rural areas, and culturally and linguistically diverse populations.49

Research

Future research activity needs to build in processes to ensure the dissemination and translation into practice of valuable knowledge; the creation of ethical and evidence-based research policies; and the promotion, monitoring and implementation of high-quality health research evidence.

Research and quality improvement activity priorities arising from this consensus statement are those relating to CHF care models (including development of readmission risk assessment models), access to meaningful data for management and benchmarking, and workforce planning. Focused investigative teams, such as clinical CHF research networks, could lead this work. In addition, more work needs to be undertaken among populations for whom frequent access to mainstream services is limited; namely Aboriginal and Torres Strait Islander peoples, those from non-metropolitan areas and lower socioeconomic backgrounds, and culturally and linguistically diverse populations.

Individuals with CHF have specific management needs. Future research should consider roles of specialty care teams (eg, cardiology, general medical) and the role of telehealth.

Conclusion

The current and future burden of CHF compels us to strive for equitable outcomes for all Australians. A national policy framework, with agreement between states, territories and the federal government, needs to be developed and implemented to tackle the increasing burden of CHF. Governments at national and state levels, together with cardiac clinical networks, need to ensure that evidence-based care models for people with CHF are standardised, with equitable access.

The core principles and recommendations described in this consensus statement should be incorporated into the various CHF systems of care operating across states and territories. Implementing these recommendations has the potential to improve the quality of care provided to individuals with CHF, reducing associated costs for both the individual and the community. Improvements could be seen not only in the care experienced by patients and their families, but also in clinical and operational benefits. Implementing multidisciplinary, patient-centred care approaches can shorten lengths of stay in hospital, reducing health care costs and improving clinician and patient satisfaction.22,44,45,50 In the longer term, other benefits of patient-centred care include reduced emergency department re-presentations, fewer medication errors, higher functional status and improved clinical care.40–42

These recommendations can empower health care providers and organisations, peak and government organisations, care regulators, education providers and consumers to improve health outcomes for patients with CHF and to reduce harm. This work needs to be underpinned by nationally recognised standards for outcome measurement that are universally recognised and easily applied in practice. Data systems need to support evidence-based decision making, while providing feedback relating to standardised performance measures. Our health care workforce needs to be equipped to deal with the increasing burden of disease associated with CHF, with training, education and research around the delivery of multidisciplinary care in an increasingly complex environment.

These recommendations, if adopted, have the potential to facilitate and promote optimal and equitable health outcomes for all Australians diagnosed with CHF.

1 Consensus development process

The National Heart Foundation of Australia convened an expert panel to provide guidance on policy and system changes to improve the quality of care for people with chronic heart failure. A relevant literature search was performed, limited to evidence from human studies published in English between 2003 and 2013. This was complemented with hand searching of reference lists from reviews and personal collections of the expert panel, and additional peer-review. As there is limited evidence around the system changes required, these consensus recommendations are based on expert opinion. They are not exhaustive, and many other changes and actions can be implemented by both individuals and organisations to improve care outcomes. The recommendations are generally broad, rather than prescriptive, and many can be implemented with minimal resourcing.

2 Themes and principles to reduce emergency presentations, hospitalisations and premature death among patients with chronic heart failure (CHF)

Theme

Principle

CHF care model

Current evidence clearly identifies that accessible, multidisciplinary, guideline-based CHF care improves outcomes.

Access to meaningful data for management and benchmarking

Collecting outcome data is the only accurate way of determining the effectiveness and cost of individual treatments; practice standards can then be based on up-to-date comparative effectiveness research.

Adequate patient information is a prerequisite for reducing unnecessary hospital admissions and medical errors.

Workforce planning

An appropriately trained workforce with access to specialist cardiology support can deliver evidence-based care.

Research

Research is essential to ensure an evidence base.

3 Recommendations to achieve a systematic approach to chronic heart failure (CHF) care*

CHF care model

1. Through state and territory cardiac clinical networks, support health departments to continue leading the development of integrated local care systems and future national quality improvement strategies within and across health services.†

2. Identify and implement mechanisms to champion the uptake of clinical practice guidelines and delivery of integrated CHF services, according to local population need, within and across health services.†

3. Develop minimum standards for CHF multidisciplinary care, which can be used to accredit health services and recognise best-practice health services or networks.

4. Develop robust funding models and examine the role of funded care packages in CHF care.

5. Establish system protocols and pathways to ensure effective clinical handover and service coordination across care transitions, and to activate appropriate services according to clinical need for an exacerbation, emergency presentation, hospitalisation or palliation.

6. Streamline care processes to facilitate early diagnosis, self-management and multidisciplinary care planning, including primary care involvement and appropriate access to palliative services.

7. Embed mechanisms to promote the rights of individuals and their carers to facilitate their active engagement with health professionals and care systems.

Access to meaningful data for management and benchmarking

8. Develop national data definitions for CHF.

9. Expand current cardiac registries to include patients with CHF.

10. Develop mechanisms to promote data linkage across care transitions.

11. Trial an electronic health record for people diagnosed with CHF, so all current and future health care providers could, with the individual’s consent, have access to the same information where and when they need it.

12. Use 12-month event-free survival as an outcome measure nationally to evaluate effectiveness of care systems. Events would include emergency presentations, hospitalisations and premature death.

13. Establish a national mechanism for monitoring and reporting CHF care outcomes against a nationally recognised set of goals and standards.

14. Develop a national set of indicators and standards to evaluate, inform and improve systems of care.

Workforce planning

15. Develop the workforce capacity across hospital and community services to deliver evidence-based care, appropriate to the local population, as identified in Guidelines for the prevention, detection and management of chronic heart failure in Australia31 and Multidisciplinary care for people with chronic heart failure: principles and recommendations for best practice.21

16. Develop robust funding models for the delivery of these services.

17. Examine mechanisms to empower general practitioners and other health care professionals in primary care to deliver evidence-based care for people with CHF.

Research

18. Create investigative teams, such as clinical CHF research networks, with active consumer collaboration.

19. Investigate approaches to optimise care delivery for Aboriginal and Torres Strait Islander peoples, those from non-metropolitan areas and lower socioeconomic backgrounds, and culturally and linguistically diverse populations.

* Level of evidence: expert panel consensus judgement. † Health services include area health services, local hospital networks, primary care, Aboriginal community controlled health organisations, aged care services and other appropriate agencies.

Should general practitioners order troponin tests?

Cardiac troponin I and T are the preferred biomarkers for assessing myocardial injury. Understanding the pathobiology of troponin and the timing of troponin testing is fundamental to the clinical utility of these biomarkers, as troponin and its kinetics are central to the universal definition of acute myocardial infarction (AMI).1 Troponin levels become elevated in serum within a few hours of an AMI, and they remain elevated for up to 7–10 days.2 However, numerous other conditions may elevate troponin levels, so it remains essential that the results of troponin tests be interpreted with clinical findings and electrocardiography results.3 The dynamics of troponin levels (rise and/or fall over time) help distinguish AMI from non-AMI conditions, thus serial troponin testing is the standard approach recommended for assessing patients with suspected acute coronary syndrome (ACS).4 In this article, we explore troponin testing in general practice, including problems faced by laboratories that offer testing in this context.

The absolutist stance

One stance on this topic is that general practitioners should never order a troponin test. The basis of this argument is that the only widely accepted clinical indication for measurement of troponin levels is suspected ACS, which should prompt referral to hospital based on clinical and electrocardiography findings without recourse to troponin testing. Supporting this argument is that serial troponin testing is unrealistic in most general practice settings, and opens the question of how a patient should be monitored while the results are awaited.

One argument against this absolutist stance is that it is an oversimplification. Further investigation and management depends on the degree of suspicion for ACS, and timing of the presentation may obviate the need for serial testing. Chest pain is a challenging symptom and the prevalence of unstable angina or AMI in general practice is low, in the order of fewer than 5% of patients with chest pain.5 Atypical presentations of AMI, such as in young people,6 people with diabetes and older people, are a perennial concern. GPs have been shown to be fairly accurate in assessing chest pain clinically as due to coronary artery disease, but not accurate enough to safely exclude it.7 A system of estimating pretest probability of ACS, or risk of short-term complications, is an attractive approach. Clinical decision-making rules and pretest probability tables have been developed to assist with this process in general practice8,9 and, while some risk stratification tools may be more relevant to doctors in emergency departments, they are potentially useful to GPs.10,11 How troponin testing might fit into risk stratification in general practice is not entirely clear.

How well do GPs understand troponin tests?

Audits from New Zealand suggest that GPs have a generally sufficient understanding of the use of troponin tests in primary care.12,13 Knowledge of false-negative results (eg, due to sampling too soon after symptoms) appears to be better than knowledge of false-positive results (eg, due to non-AMI causes of raised troponin levels). Most GPs would refer high-risk patients without troponin testing, but a small proportion of GPs would defer hospitalisation while waiting for the troponin result (mostly for patients with an intermediate probability of AMI).

Why do GPs request troponin tests?

Our experience suggests that GPs mainly request troponin tests to rule out AMI in one of two situations. The first situation is patients who had symptoms in the preceding days but for whom symptoms have resolved (also the experience of others13,14). One expert has suggested that this may be a justifiable use of troponin testing in primary care15 and troponin testing is suggested in National Institute for Health and Care Excellence (NICE) primary care guidelines in pain-free patients who had chest pain more than 72 hours earlier.9 The second situation is patients who have atypical symptoms and/or a low likelihood of ACS, in whom troponin testing appears to cover the residual clinical uncertainty. Unexpectedly positive troponin results occasionally occur in such situations, which may otherwise not be detected.

How do GPs currently request troponin tests?

Most requests for troponin testing from general practice are requests for a single test, not serial testing.16 This begs the question of whether ordering a single troponin test is an appropriate strategy. Given our understanding of troponin kinetics, a single negative troponin test result a certain time after symptom onset could be clinically useful in ruling out AMI (ie, in “late presenters”). The suggested time frame varies between publications, but is usually in the order of 6–9 hours4,17 with the caveat that the time of symptom onset can be unreliable. Local experts have suggested that a single troponin test 12 hours after resolution of suggestive symptoms (with a normal electrocardiogram and no high-risk features) is useful for this purpose.11 With the so-called high-sensitivity troponin assays, this window may decrease: in an emergency department setting, an undetectable (ie, not merely negative) troponin value obtained from a high-sensitivity troponin assay at presentation has been shown to have a very high negative predictive value for a subsequent diagnosis of AMI,18 but this strategy is experimental. The safest rule of thumb is that a single negative test result for troponin does not exclude AMI in a patient with current or very recent symptoms, nor does it exclude clinically significant coronary artery disease.

Conditions associated with chronic troponin elevation

As most GP requests for troponin testing are for a single test, conditions associated with chronic, non-AMI elevation of troponin levels present a problem. Examples include chronic cardiac failure and chronic kidney disease (CKD). A positive result from a single troponin test could be misleading because it might reflect the underlying chronic disease and not AMI. The prevalence of positive troponin test results (defined as above the 99th percentile of the general population) in CKD depends on the stage of the CKD (positive results are more likely during more advanced stages) and on the troponin assay used. This is exemplified by a recent study of asymptomatic patients who had CKD but were not on dialysis. The prevalence of a positive troponin result (for the whole cohort) was 68% when a high-sensitivity troponin T assay was used, 38% when a high-sensitivity troponin I assay was used, and 16% and 8% for troponin T and I, respectively, when contemporary (“less sensitive”) assays were used.19 Despite the high rates of positive troponin results in this study, a negative troponin result from a sample taken at an appropriate time is useful for ruling out AMI in patients who have CKD, but at the considerable disadvantage of reduced positive predictive value, with the attendant risk of unnecessary hospitalisation. Clinical assessment of the acute event in such patients becomes all the more important if this is to be avoided.

Logistics of troponin testing for outpatients

Offering troponin testing in the community is logistically complex and there is a lack of formal guidance for laboratories in this area. Guidelines on management of ACS recommend that a troponin test result should be available within 60 minutes of blood being drawn and, if not, that point-of-care testing should be available.4 This is aimed at hospital-based laboratories and is not a realistic target for large private pathology networks that may test hundreds of community samples per day at variable geographical distances from large networks of collection centres and general practices. So what is the solution? Accept the longer turnaround times and promote judicious use of troponin tests by GPs? Longer turnaround times may be acceptable if testing is largely confined to patients who have a low pretest probability, or low risk, of AMI. If so, what is a reasonable turnaround time for community samples — three hours? Six? At the other extreme is rigorous pursuit of fast turnaround times to meet the apparent clinical need in the community, probably with the help of point-of-care testing, although there are questions about the performance of point-of-care troponin assays.20 The solution is probably a compromise between the two. The only guidelines that provide advice on this are the NICE guidelines, which state that troponin testing can be undertaken in general practice “providing timely results can be obtained” but do not elaborate on what “timely” means.9

After-hours elevated troponin levels can be problematic for all concerned. For example, when samples are taken late in the afternoon, results might not be available until after clinic hours. A common policy is to treat positive troponin test results as “critical results” and to notify the requesting doctor or a representative (such as a locum GP service, if nominated). In the event that a doctor cannot be found to take the result, which is not uncommon in our experience, laboratory staff (usually pathologists) phone patients directly and advise that hospitalisation is the safest course of action. But when a patient cannot be contacted, laboratory staff face a dilemma: can the result wait until office hours, or should emergency services be arranged? We are aware of anecdotal cases in which after-hours notifications of high troponin levels to patients at home have probably contributed to their early survival — but this raises the question of whether such patients are better served by referral to hospital in the first instance. A published coroner’s case touches on these important issues for both GPs and pathologists.21

Conclusion

We suggest that GPs should have a high threshold for requesting troponin testing and carefully assess risk before ordering troponin tests. Positive troponin test results usually change the course of management, but the time frame in which the result becomes available must be balanced against the risk of delay in diagnosis and therapy. A troponin test should not be requested unless a GP is certain that a robust process is in place by which they can be contacted, day or night, if the result is positive. There is an obvious need for further education, research and inclusion of this topic in future clinical guidelines. Our suggestions for using, or not using, troponin tests in general practice are summarised in the Box.

Suggestions for using, or not using, troponin tests in general practice

The default position

  • The default position for patients who have symptoms suggestive of acute coronary syndrome is hospitalisation without prior troponin testing.

Using a single troponin test

  • It is reasonable to use a single troponin test in general practice to exclude the possibility of acute myocardial infarction (AMI) in asymptomatic patients whose symptoms (typical or otherwise) resolved at least 12 hours prior, so long as they have no high-risk features and a normal electrocardiogram.11
  • A single troponin test may also be useful to investigate an otherwise unexplained creatine kinase elevation.

Using serial troponin tests

  • In patients presenting to general practice within 12 hours of symptom onset who are at low risk of AMI and/or have atypical symptoms, and for whom troponin testing is being considered, serial testing is advised.
  • In patients with conditions that are associated with a high prevalence of positive troponin test results, such as chronic kidney disease, a single test can be misleading. Serial testing may be required to resolve clinical uncertainty.
  • Serial testing is most appropriately performed in hospital. The safety of serial testing in outpatient settings has not been established.

A systematic review of the challenges to implementation of the patient-centred medical home: lessons for Australia

Australia’s first National Primary Health Care Strategy1 and resulting National Primary Health Care Strategic Framework2 initiated growing interest and development in our primary care sector, particularly general practice. Clinicians, governments and organisations are now actively searching for new approaches, models of care and business levers to support the primary care quality, efficiency and access gains sought. In December 2012, then Minister for Health Tanya Plibersek announced a focus on the patient-centred medical home (PCMH) as a model of interest.3 The Royal Australian College of General Practitioners (RACGP) has also been a consistent champion of the model, urging adoption of its elements as part of current reforms and calling for the federal government to fund and implement key elements in its 2013–14 Budget submission.4

The PCMH concept of care was introduced by the American Academy of Pediatrics in 1967, and was adopted in 2002 by the family medicine specialty. Four major primary care physician associations in the United States, along with other stakeholders, formed the Patient-Centered Primary Care Collaborative (PCPCC), and in 2007 endorsed the Joint principles of the patient-centered medical home.5 These include: access to a personal physician; physician-directed medical practice; whole-person orientation; care coordination and/or integration; quality and safety benchmarking through evidence-based medicine and clinical decision support tools; enhanced care availability after hours and via e-health; and practice payment reform. We used this definition of PCMH in our review because it concords strongly with the RACGP’s statement, A quality general practice of the future,6 endorsed by all general practice organisations nationally in 2012.

There is evidence that adoption of the PCMH model can improve: access to care;7–10 clinical parameters and outcomes;11–15 management of chronic and complex disease care;7–9,11,12,14–22 preventive care services (eg, cholesterol tests, influenza vaccinations, prostate examinations);9,10,12,13,17,18,20,23–26 and provide improved condition-specific quality of care14,15,18,19,22,27 and palliative care services.8 Data also indicate that the PCMH model can decrease the use of inappropriate medications,8,22,23 and significantly reduce avoidable hospital admissions and readmissions, emergency department use and overall care costs.8,14,22,28–31

While the PCMH model shows promise in transforming the primary care system into a more integrated and comprehensive model, studies report challenges and barriers to the implementation and adoption of this model. Before its potential can be achieved, more robust information is needed on the actual change process, challenges and barriers associated with implementation of this model.32,33

We undertook a systematic review to identify the major challenges and barriers to implementation and adoption of the PCMH model. The findings from this review will provide lessons for Australian primary health care reform and future PCMH initiatives in Australia.

Methods

A complete description of the methods is provided in Appendix 1.

In December 2012, we searched the PubMed and Embase databases for studies published between January 2007 and December 2012 using the search terms patient centered medical home, patient centred medical home, medical home, or PCMH. Appendix 2 provides details of the search strategy. A snowballing strategy was used to identify other related citations through the reference list of all reviewed articles.

Abstracts were included if they met the following inclusion criteria: 1) published between 2007 and 2012; 2) in English; 3) reported information or data related to the review objective; 4) defined PCMH using the PCPCC Joint Principles, or at least mentioned some of its components.5 There were no restrictions on study design or country of study.

Articles included during the initial screening by either reviewer underwent full-text screening. One reviewer with expertise in the area reviewed the full text of each article and indicated a decision to include or exclude the article for data abstraction. We applied 10 quality criteria that were common to sets of criteria proposed by research groups for qualitative research (Box 1).34–36 Two reviewers independently assessed the quality of each study, and discrepancies were resolved through discussion.

Data extraction and synthesis

A data extraction form was created by the investigative team to assist in systematically extracting information on the study design (type of study, methodology and setting) and key findings related to the review objective. One researcher with content knowledge in the area abstracted the data, while a second researcher reviewed the abstracted data alongside the original article to check for accuracy and completeness.

Thematic synthesis was used in three stages: the free line-by-line coding of data; the organisation of these “free codes” into related areas to construct descriptive themes; and the development of analytical themes.37 Data were configured at a study level using a top-down approach, which allowed individual findings from broad study types to be organised and arranged into a coherent theoretical rendering.38 Synthesis matrices allowed data to be recorded, synthesised and compared.

Results

The search strategy identified 2690 citations, of which 28 studies met the inclusion criteria (Box 2). All studies were from the US. This was not surprising, as the PCMH model is a North American model and the PCPCC Joint Principles of the PCMH definition we used as part of the inclusion criteria is from the US. Of the 28 articles, there were nine exploratory studies, 13 descriptive studies, and six experimental or quasi-experimental studies. All studies met five or more of the 10 quality criteria, and nine of the 28 studies met all 10 quality criteria. Descriptions of included studies (including type of study, method, setting and quality rating) are provided in Appendix 3.

This systematic review identified six key overlapping challenges and barriers to implementation and adoption of the PCMH model. These are presented below, and Appendix 4 includes a summary table of themes identified in each study.

Challenges with transformation and change management in adopting a PCMH model

Eleven studies discussed varying challenges and barriers to transforming to a PCMH model. Transformation calls for significant changes in the routine operations of practices, and these are difficult to achieve and require more than a series of incremental changes.16,27,39–44 Key requirements are: long-term commitment,17,39,43,45 local variation,17,39,45 a focus on patient-centredness,39,45,46 and support through reform of the larger delivery system to integrate primary care within it.17,27,40,47 Even with external payment reform, practices need extensive assistance coaching from external facilitators and expert consultants to transform to a PCMH.16,27,39,43

There were reported challenges41,43,44,48 relating to a shift in paradigm for individuals and practices, which required them to move away from a physician-centred approach towards a team approach shared among other practice staff.17,39,43,44,49 Transformation efforts were slowed or ceased by ineffective change management processes;39,50 lack of leadership,51,52 readiness for change, communication and trust;17,44,50–53 and culture.39,43,52,53 Misinformation or lack of understanding about the PCMH could lead to misunderstandings about what was being asked of practices and staff,41,45,53 causing resistance to change.39,43 Furthermore, practices without capacity for organisational learning and development, or what is called “adaptive reserve” (such as a healthy relationship infrastructure, an aligned management model and facilitated leadership), were more likely to experience “change fatigue”,17,41,43,44,50,54 and less likely to successfully implement the PCMH model.17,43,44

Difficulties with electronic health records

Implementing an electronic health record (EHR) with a clear, meaningful use, and which administers the principles of PCMH, has been a difficult task for primary care practices in transition.17,41,46,55 Implementation and use of an integrated EHR has proved to be more difficult than originally envisioned,27,39,41,43,52,53 requiring significant investment of time, effort, resources (eg, new equipment, training material) and money.39,41,44,52,54 Reported challenges related to setting up EHRs at practices, and providing ongoing technical support and resources to service.

There were also difficulties with functionality (eg, EHR could not provide data for population management; a disease registry was absent or extremely awkward to activate; and e-visits such as telephone, email or video consultations presented challenges), and use of EHRs (eg, accessing electronic records in a timely, easily digestible manner, and accuracy and reliability of information in the EHR).39,42,48,50,51 Furthermore, single-practice EHRs were reported as insufficient and a barrier to effective coordinated care,47 and the lack of interoperability of EHRs hindered collaboration between providers, crucial to the PCMH model.17,46,47,52,54

Challenges with funding and payment models

Sixteen studies reported challenges with the current funding models for PCMH. Most stated that current available funding and reimbursements were likely to be inadequate for the transitional costs and sustainability of the PCMH,39–42,45,49,50,54–59 and the essential functions of the PCMH are not supported by traditional fee structures.41,47,49,55,56 Many studies recommended that new payment structures and incentives for practices and providers be developed to support implementation and sustainability of the PCMH model.39,40,43,45,49,50,52,54–59

Insufficient practice resources and infrastructure

Eighteen studies reported barriers related to insufficient resources within practices to implement the PCMH model. These included lack of resources (eg, equipment, human resources, training material), structural capabilities, time and financial capacity to develop the necessary building blocks to transform their practice into a PCMH.17,45,51–53,59 Substantial support (including non-monetary support) and resources were required to implement change at the practice level.16,27,41,49,50,57,58,60 Smaller practices typically could not employ the same resources as larger facilities due to budget and resources constraints. Therefore, implementation at small practices was challenging due to lack of internal capabilities.21,41,42,44,61

Inadequate measures of performance and inconsistent accreditation and standards

There were several reported challenges relating to variations in PCMH standards, inadequate accreditation and measures of performance.16,17,39–42,45,47,56 Most tools developed to measure achievement of the medical home did not directly correspond to the seven Joint Principles that define the PCMH, and many of these principles were difficult to measure.45 Furthermore, accreditation does not yet capture all the key aspects required for a fully functioning medical home,16 and the criteria for evaluating PCMH were inconsistent.56 Establishing standards, measures and targets proved difficult.16,17,40,42

Discussion

In our systematic review, we found evidence of challenges and barriers to implementation of a PCMH model, including difficulties with transformation to a new system, change management issues, adopting EHRs and adapting payment models. Other challenges were inadequate resources, performance measurement and accreditation.

Our findings have significant importance for current Australian reform initiatives. The RACGP, as part of its 2013–14 Budget submission, called for the federal government to fund and implement key elements of the PCMH, as it “encapsulates the very definition of [future] general practice in Australia”.4 Evidence-based assessment of the barriers and enablers to such transition presented in this article is an essential step to effective implementation.

As in the US, primary care practices in this country are challenged by growing complexity of care, accreditation pressures, and perverse funding and reward systems. Clinicians and organisations are often on the receiving end of policy implementation that is top-down rather than bottom-up, and, as small businesses, struggle to adapt in the defined time frames.62,63 Our review also notes the importance of reform across the larger delivery system to integrate primary care change within it. It demonstrates the importance of a long-term and tangible commitment to change adoption at the practice level (strong “adaptive reserves”), with a focus on teamwork, leadership, high-quality communication, staff development and ongoing support for a culture of change. Appropriate practice resourcing for infrastructure and system support over the “transformation” period is essential, as identified in our National Primary Health Care Strategy.1,2

The literature also highlights the importance of practice and practitioner funding that promotes patient centredness, preventive health, and a focus on complex chronic disease support, case management and hospital avoidance. This is timely in the Australian context, as is the focus on EHRs that promote care coordination, quality and safety benchmarking, and clinical decision support.54

Finally, our findings suggest that reform initiatives should involve accreditation review, such that these frameworks reflect measures of performance and standards that match the key benchmarks of importance, with minimal administrative barriers. Such initiatives are in early development, with the RACGP and Australian Commission on Quality and Safety in Health Care partnering in a review of accreditation process and outcome.64

Our review had some limitations. The search strategy did not include grey literature, and unpublished evaluation studies or reports may have been missed. There could also be other challenges or barriers not reported in the reviewed publications. The review was limited to studies that used the Joint Principles,5 because this definition concords strongly with the RACGP’s A quality general practice of the future,6 but may have missed literature published outside this definition. Data abstraction from qualitative studies can be complicated by the varied reporting styles.65 Relevant study “data” were often not presented in the results section, but integrated into the discussion or recommendations. Hence, a second researcher reviewed abstracted data alongside the original article to check for accuracy and completeness. Furthermore, the synthesis of qualitative data is problematic and dependent on the judgement and insight of the researchers (interpretation bias).37,66,67 To limit this bias, two independent researchers were used in the synthesis process.

Our systematic review indicates that implementation of significant primary care change should be cognisant of several considerations, mostly at the practice–practitioner interface. It comes at an important juncture for Australian health care reform, with reviews into the personally controlled electronic health record and Medicare Locals, and recent ministerial statements regarding funding reform for chronic disease management likely to have a major impact on the sector. For policymakers, they underline the approach and resourcing required to effectively influence service delivery. For clinicians, they highlight the teamwork, commitment and practice infrastructure critical to success. Australian health care reforms demand “a stronger, more robust primary health care system”.2 Addressing documented barriers to change adoption relevant in the Australian context will be a critical evidence-into-policy initiative.

1 Criteria for assessing quality of studies34–36

  • Aims and objectives clearly stated
  • An explicit theoretical framework, study design and/or literature review
  • A clear description of context
  • A clear description of the sample and how it was recruited
  • A clear description of methods used to collect and analyse data
  • Attempts made to establish the reliability or validity of data analysis
  • Inclusion of sufficient original or synthesised data to mediate between evidence and interpretation
  • Use of verification procedure(s) to establish credibility
  • Conclusions supported by results
  • Relevance

2 Flow diagram outlining selection process of studies for analysis

Who’s responsible for the care of women during and after a pregnancy affected by gestational diabetes?

Gestational diabetes mellitus (GDM) is the strongest single population predictor of type 2 diabetes,1 and current Australian prevalence is 10%–13%, depending on the criteria used.2 Poor health outcomes extend to children of mothers who had GDM, due to increased risk of obesity and abnormal glucose metabolism during childhood, adolescence and adulthood.3

Antenatal lifestyle intervention is shown to improve short- and long-term maternal and infant health outcomes.3 In addition, it can effectively prevent type 2 diabetes among women who have had GDM.1 However, although some centres of excellence exist, in many cases, antenatal care is not delivered systematically.4

After their babies are born, women who have had GDM can be described as falling into a health care “chasm”.5 When these women leave hospital, their obstetricians and endocrinologists feel that their work is done. Lack of coordination between the hospital and primary care sectors can mean that no one assumes responsibility for the care of these women.

The opportunity to prevent or delay type 2 diabetes in this high-risk population through primary care was noted more than a decade ago.6 However, defined care pathways and coordination remain elusive; implementation of evidence has not occurred. In many cases, general practitioners may not be aware that the woman has had GDM, and may not have a clear pathway directing responsibility for follow-up care.

There is an urgent need to implement a widespread and coordinated approach to prevent progression to type 2 diabetes in this population. Rectifying this situation requires cooperation and collaboration between all care providers.

Antenatal care: navigating the new gestational diabetes landscape

The health care sector operates under guidelines with conflicting content and differing levels of comprehensiveness and professional endorsement (Box). The Australasian Diabetes in Pregnancy Society (ADIPS) recently released revised consensus guidelines for testing and diagnosing GDM in Australia and New Zealand.7

Women with GDM are managed in hospitals because they are identified as having pregnancies at higher risk of adverse outcomes. The ADIPS guidelines recommend an oral glucose tolerance test (OGTT) for all women (unless already diagnosed with GDM in early pregnancy) at 24–28 weeks’ gestation.7 These guidelines were informed by several studies, including the Hyperglycemia and Adverse Pregnancy Outcome (HAPO) Study, which indicated a strong continuous association of maternal glucose levels with increased diabetic fetopathy.14

A change to testing protocols will be introduced in July 2014 and diagnostic criteria on 1 January 2015 (Aidan McElduff, Clinical Associate Professor of Medicine, University of Sydney and ADIPS President; personal communication). Concerns exist about their potential workload implications and evidence base.

Health service and pathology database analyses have resulted in equivocal projections of the potential workload increases; it is most likely that many will see a doubling of cases.2,15 Workload projections can be difficult, as true prevalence is not known, but it has been suggested that the increasing rate reflects the prevalence of abnormal glucose metabolism in the general population.16

In considering potential workload costs and changes, we need to consider the results from two well executed randomised controlled trials, which demonstrated that treatment of GDM can prevent adverse outcomes.17,18 For most women (80%–90%), GDM could be managed through dietary counselling delivered by a dietitian. In some centres, this proportion will be lower, depending on population characteristics. Medical nutrition therapy is a cornerstone intervention for women with GDM,19 and its appropriate delivery results in reduced insulin requirements and improved blood glucose control.19 However, systematic, evidence-based dietetic care of women with GDM does not occur in many centres in Australia.4 Australian health services require clinician leadership and commitment to partnership and change in (re)allocation of resources to support a multidisciplinary team in providing evidence-based care for improved maternal and infant outcomes.

Some clinicians raise concerns about diagnostic criteria changes based on observational study outcomes, but the previous diagnostic criteria were the product of an ad-hoc working party and lacked the strong evidence base that underpins the current criteria.20

Postnatal follow-up: who’s taking responsibility?

Australian guidelines recommend that all women who had GDM should undertake a 75 g OGTT between 6 and 12 weeks after delivery.7 International guidelines also highlight the importance of lifestyle modification, breastfeeding, birth control and risk counselling to improve health outcomes for these women and their children.12,13

The extent to which these recommendations are integrated into postnatal GP visits is not known, but some studies suggest diabetes testing is suboptimal.21 Self-report surveys of women with prior GDM indicated that about half of participants returned for OGTTs, but only a quarter in the appropriate period.21,22 The potential use of glycated haemoglobin testing instead of the OGTT appeals to many, but the approach may not change until it is approved on the Medicare Benefits Schedule.

Appropriate strategies to engage women in screening are paramount, as the motivation to manage a GDM diagnosis transforms to apathy once GDM resolves.23 Barriers to ongoing screening include a lack of awareness of the need for screening, difficulty attending screening with an infant, dislike of the OGTT process, being a mobile population, and inconsistent advice from health care providers about testing, lifestyle modification and risk.21–23 Findings from the United Kingdom suggest that health care professionals need to balance between reassurance of likely resolution of GDM and adequate information about potential progression to type 2 diabetes.23 Perception of risk is an important motivator; a lack of perceived risk of developing type 2 diabetes is common and can be related to timing, content and tone of messages.23,24

Prevention of diabetes in primary care

Which guidelines?

Three Australian guidelines exist for the follow-up of women who are at risk of type 2 diabetes (Box).7–9 Their core messages are similar, but they vary in several areas, diluting GP awareness and implementation. Beyond the timing of testing regimens, recommendations regarding lifestyle interventions to prevent type 2 diabetes progression are absent from the ADIPS guidelines, but the Diabetes Australia/Royal Australian College of General Practitioners (RACGP) Diabetes management in general practice 2014–20158 and Guidelines for preventive activities in general practice (the “red book”; also distributed to GPs in Australia)25 outline diabetes management and dietary advice for diagnosed cases in general practice and for diabetes prevention.

Many similarities exist between the diet for GDM and diabetes prevention (ie, focus on low glycaemic index, low saturated fat, high fibre content). However, during a pregnancy complicated by GDM, there is a major focus on tightly controlled blood glucose levels, although appropriate diet quality for pregnancy requirements and gestational weight gain is also paramount. By contrast, diabetes prevention diets have a greater focus on weight reduction. Currently, there is no effort to explain to women who have had GDM the difference in approach.

A missed opportunity?

Although GPs view follow-up care as their role within the broader context of general health screening and promotion, this is often opportunistic.26 Advice from GPs is a powerful motivator for women to adopt lifestyle modification.27 However, GPs report not being well versed in guidelines for GDM follow-up care, potentially reflecting the lack of clarity in the literature and their varying knowledge and confidence in provision of lifestyle advice and interventions.28 GPs generally give appropriate exercise advice, but can be less clear about dietary or weight loss goals.26

These practices are reinforced by systems and process barriers of prioritisation of issues during a consultation, a lack of integration of recall tools and intervention resources in daily workflow, and uncertainty about responsibility for screening, as well as poor communication between secondary and primary care sector and fragmentation of pre- and postnatal care services.28

Right information, right people, right time

Clinical trials have demonstrated that lifestyle modifications with weight loss and moderate exercise can reduce the incidence of type 2 diabetes by up to 58% for people at high risk, with an impact still evident 8 years from the intervention onset and 4 years after the active intervention ceased.29 Real-world implementation in the Australian health care system has achieved 40% reduction in the risk of progression to diabetes.30

Agreement between and willingness to work in partnership with key stakeholders — such as ADIPS, Diabetes Australia, the RACGP and the Royal Australian and New Zealand College of Obstetricians — is required for a collective approach to delivering diabetes prevention to this high-risk population.

However, despite convincing evidence about effective programs in Australia, postnatal support after a pregnancy with GDM is lacking and is without coordination. Interventions using technologies such as telephone, SMS and the internet have been trialled for diabetes care and may be useful in prevention. These must be underpinned by behaviour change theories and address barriers to making changes regarding future risk.23 Women have been identified as being receptive to messages several months after birth, which may align with “transition times” (eg, introduction of solids).23 Further efforts are urgently needed to develop lifestyle strategies that meet the specific needs of this group of women.

Diabetes Australia’s National Gestational Diabetes Register (NGDR), part of the National Diabetes Services Scheme, was launched in 2011 as a free service to women with a Medicare card to help those who have had GDM to manage their health and prevent progression to type 2 diabetes. One function of the NGDR is to send regular reminder letters to women and their GPs regarding diabetes checks (at registration, 12 weeks after birth, and annually thereafter). These reminder letters also include general information for the women and their families to help them continue a healthy lifestyle.

Although the NGDR outlines what testing to undertake, its potential to allow implementation and dissemination of a comprehensive, consolidated set of guidelines is perhaps underused. It could facilitate effective connection of women with a history of GDM with specific, effective, evidence-based lifestyle advice as well as clinical guidance for their GPs.

A call to action: the need for a collaborative approach

A clear pathway, developed between all stakeholders, with delineated roles and responsibilities to ensure that best-practice care is delivered along the continuum of antenatal, postnatal, interconception and longer-term care is required. Delivery of coordinated, effective programs is essential for this group of women. Without such clarity, and in the absence of a systems approach to care, we are failing to seize an opportunity to reduce the incidence of type 2 diabetes and promote the wellbeing of these women and their children. After a diagnosis of GDM, women view their GP as the most appropriate source of follow-up care,24 so it is imperative that GPs are given the right guidelines and education to advise these women about preventing or delaying progression to type 2 diabetes.

A comparison of current gestational diabetes mellitus diagnosis (GDM), treatment and follow-up guidelines

 

Details of guidelines

Guideline/society (country)

ADIPS (Australia and New Zealand)7

Diabetes Australia and RACGP (Australia)8

Therapeutic guidelines: endocrinology (Australia)9

ACOG (US)10

ADA (US)11

NICE (UK)12

CDA (Canada)13

Antenatal testing protocol

Universal OGTT at 24–28 weeks; earlier if clinically indicated

Universal screening at 26–28 weeks. Two-step approach recommended (GCT then OGTT).

Universal GCT or OGTT at 26 weeks. Early screening if high risk

—

Universal OGTT at 24–28 weeks in women not previously diagnosed with overt diabetes

At 24–28 weeks if the woman has any risk factors or earlier if GDM in a previous pregnancy

Universal screening 24–28 weeks. Two-step approach preferred (GCT then OGTT)

Timing of first postpartum follow-up visit

6–12 weeks

6–12 weeks

6–12 weeks

6–12 weeks

6–12 weeks

6 weeks

6 weeks – 6 months

Which test(s) for postpartum screening

75 g OGTT

75 g OGTT

75 g OGTT

FPG or 75 g OGTT

75 g OGTT; not HbA1c

FPG

75 g OGTT

Who with?

—

GP

—

—

—

—

—

Frequency of follow-up and recommended test

Dependent on future pregnancy plans and perceived risk of type 2 diabetes, yearly OGTT if planning pregnancy. 1–2 yearly FPG (low risk); OGTT/HbA1c (higher risk)

3-yearly; with FPG

If postnatal test normal: annual fasting or random blood glucose or OGTT every 2 years and before subsequent planned pregnancies

3-yearly, as above

Minimum 3-yearly; with OGTT. If IFG or IGT, yearly

Yearly; no blood test specified

At least 3-yearly and before each pregnancy; not specified

Other postnatal advice included

No recommendations

Increase physical activity, weight loss/healthy diet. Refer to dietitian and/or physical activity program. Preconception advice.

Risk counselling for future type 2 diabetes. Lifestyle advice: diet/physical activity. Subsequent pregnancy: early screening 12–16 weeks repeated at 26 weeks.

Weight loss and physical activity counselling as needed

Women with a history of gestational diabetes found to have prediabetes should receive lifestyle interventions or metformin to prevent diabetes.

Lifestyle advice: weight control, diet and exercise

Lifestyle advice to prevent diabetes and cardiovascular disease should begin in pregnancy and continue postpartum. Encourage breastfeeding for at least 3 months postpartum. Provide risk and preconception counselling.

ACOG = American College of Obstetricians and Gynecologists. ADA = American Diabetes Association. ADIPS = Australasian Diabetes in Pregnancy Society. CDA = Canadian Diabetic Association. FPG = fasting plasma glucose. GCT = glucose challenge test. HbA1c = glycated haemoglobin. IFG = impaired fasting glucose. IGT = impaired glucose tolerance. NICE = National Institute for Health and Clinical Excellence. OGTT = oral glucose tolerance test. RACGP = Royal Australian College of General Practitioners. UK = United Kingdom. US = United States.

Copayments for general practice visits

In reply: Arnold regrets that the debate about copayments is restricted to fee-for-service methods of paying general practitioners.

This is true. The Minister for Health has intimated an overhaul of the whole general practice system, but this was not addressed in the recent federal Budget. This is a pity. It would be good to have a debate about different systems, ranging from our current blended model of fee-for-service with grants to practices for achieving quality indicators, through to capitation systems (eg, the United Kingdom’s National Health Service) and salaries (common in public hospitals and in primary care in some countries).

Perhaps this is also a good moment to try to bridge the great divide between state health (mostly hospitals) and federal health (mostly general practice and private health care), which obstructs much of integrated care.

Tinkering at the edges with copayments seems too trivial.

Shared decision making: what do clinicians need to know and why should they bother?

Lucille is a 2.5-year-old who has had a cold for 3 days. Last night, it became worse — Lucille was restless and had a fever. Her mother was up with her for much of the night, and she settled eventually with paracetamol. The mother and Lucille come to see you today and both look exhausted. The only positive findings on clinical examination are a congested nose and bulging red left ear drum. You diagnose acute otitis media. “OK”, you say, “Lucille has a middle ear infection”. Her mother asks, “What can be done to help her?”

Shared decision making is a consultation process where a clinician and patient jointly participate in making a health decision, having discussed the options and their benefits and harms, and having considered the patient’s values, preferences and circumstances. Shared decision making is not a single step to be added into a consultation, but can provide a framework for communicating with patients about health care choices to help improve conversation quality. It is a process that can be used to guide decisions about screening, tests and treatments. It can also be thought of as a mechanism for applying evidence with an individual patient through personalising the clinical decision. Although, to date, most research about shared decision making has focused on medical practitioners, it is relevant to clinicians of all disciplines, including nursing and allied health.1

Shared decision making can be viewed as a continuum,2 along which the extent to which a patient or a clinician takes responsibility for the decision-making processes varies. At the extremes are clinician-led decisions and patient-led decisions, with many other possible approaches in between.2,3 The extent of involvement will vary between individuals and between consultations, and according to the patient’s preferences and the context in which the decision is occurring. Regardless of whether the patient or clinician takes the lead in the decision-making process, joint discussion should occur.

How can shared decision making help?

Internationally, shared decision making is seen as a hallmark of good clinical practice, an ethical imperative,4 and as a way of enhancing patient engagement and activation. Increasingly, it is being advocated for in clinical guidelines and health care policies.5–7 While shared decision making is applicable to most situations, it is especially important in certain circumstances; for example, where the evidence does not strongly support a single clearly superior option (most clinical decisions) or where a preference-sensitive decision is involved. That is, when there is uncertainty as to which option is superior, each option has different inherent benefits and harms, or the decision is likely to be strongly influenced by patients’ preferences and values.8,9

The relationship between shared decision making and evidence-based practice is becoming increasingly recognised. Shared decision making provides a process for bringing evidence into the consultation and incorporating it into discussions with the patient, along with discussion about the patient’s values and preferences. In other words, it is an important, if under-recognised, component to evidence translation, a route by which evidence is incorporated into clinical practice. Shared decision making may also help reduce the unwarranted variation in care10 that may partially arise from clinicians’ opinions dominating decision making, with insufficient consideration of both empirical evidence and patients’ preferences.11

Patients and clinicians typically overestimate the benefits of interventions and underestimate their harms.12–14 Shared decision making can provide the opportunity for resolving this mismatch between clinician and patient expectations and the demonstrated benefits and harms of screening, tests and interventions. Consequently, shared decision making may reduce the inappropriate use of tests and treatments,15 such as those that are not beneficial for the majority or are associated with substantial risks or harms. As such, it can play a role in reducing the problem of overdiagnosis and overtreatment. Patients tend to choose more conservative options than their clinicians when fully informed about the benefits and harms.15

Most evidence about the effectiveness of shared decision making comes from trials of decision aids, where most research has been conducted to date. Decision aids have demonstrated effectiveness for increasing knowledge and risk-perception accuracy; improving patient–clinician communication; and reducing decisional conflict, feeling uninformed, passivity in decision making, and indecision about the choices made.15

The process of shared decision making

A common misperception is that shared decision making is synonymous with the use of decision aids. However, the core of shared decision making is a process, which might additionally use decision support tools but is not dependent on them. This process varies according to numerous factors related to the patient, clinician and other circumstances. One set of questions that can be used to guide the process is shown in Box 1. Originally phrased for use by patients,16 we have amended the wording to the clinicians’ perspective.

As many patients are unfamiliar with being invited to share in decision making, it may help to briefly explain the process. Outlining that they have some choices that you would like to go through with them before deciding together about the next step may reassure patients who might otherwise feel overwhelmed, uncertain or even abandoned. If the problem or diagnosis is clear, and a decision about the next step is necessary, the next step involves describing the nature of the problem or condition (question 1, Box 1). This should usually include providing information about the natural history of the condition — that is, what is likely to happen without any intervention (“watch and wait”). Eliciting the patient’s expectations about management of the condition, including previously tried approaches and experiences, along with fears and concerns, is important and allows for discussion and correction of misperceptions where necessary (either at this stage or later in the process). The second question triggers a discussion of the options and identification of those that the patient would like to hear more about (Box 1). The third question enables discussion about the benefits and harms of each option, including their likely probability or size (Box 1 and Box 2). The fourth question provides patients the opportunity to weigh up the benefits and harms of the options, and consider them in the context of their preferences, values and circumstances (Box 1). Finally, the fifth question explores whether the patient is ready to make a decision or whether further information, time or involvement of other people is needed before a choice can be made (Box 1). If a high-quality and appropriate decision support tool is available for the decision under consideration, it can be incorporated before, during or after the consultation.

There are other approaches to shared decision making, in addition to the approach shown in Box 1. One alternative model breaks the decision-making part of the consultation into choice talk (helping patients to know that options exist), option talk (discussing the options and their benefits and harms) and decision talk (helping patients explore options and make decisions).19

The role of decision support tools in shared decision making

Evidence summaries can be useful in supporting decision making, but current formats such as clinical practice guidelines and systematic reviews typically do not map well onto decision points in the consultation, nor do they promote patient interaction and discussion. This is particularly so where the evidence is uncertain or where benefits and harms need to be weighed up with patient preferences and clinical contexts to individualise decisions.

Specifically developed decision support tools can help clinicians and patients to draw on available evidence when making clinical decisions.20 The tools take a number of formats (Box 3). Some are explicitly designed to facilitate shared decision making (eg, decision aids). Others provide some of the information needed for some components of the shared decision-making process (eg, risk calculators, evidence summaries), or provide ways of initiating and structuring conversations about health decisions (eg, communication frameworks, question prompt lists).

When appropriate tools do exist, clinicians can use them in various ways: during the consultation (Box 1); as homework tools (where the patient is invited to use them either in the waiting room or at home, before or after the consultation); and either with or without assistance.

Decision support tool use is not always straightforward. They may not fit the patient’s clinical or personal circumstances; clinicians and patients need to be willing to use them; they require clinicians to have some skills in how to use them; and there may need to be support for their use and delivery. Decision support tools alone are not the answer, and providing them does not guarantee that shared decision making will occur. Knowledge exchange alone is not sufficient — shared decision making needs to occur in a context where patients are enabled and supported to have a more active role.21,22

Misconceptions about shared decision making

Shared decision making is making headway in many countries.23 However, a number of misconceptions are hampering its implementation. Box 4 lists the barriers to shared decision making, as perceived by clinicians,24 along with comments and, where possible, research findings about each perceived barrier.

Challenges to shared decision making

There are several key challenges to the widespread use of shared decision making within the Australian health care system. First, training in shared decision-making skills is essential for its uptake, yet unlike various international initiatives,40 limited training options exist in Australia for both student clinicians and practicing clinicians. Second, shared decision making is dependent on clinicians having access to up-to-date high-quality evidence, preferably already synthesised. Decision support tools can assist with this, but they exist for only a minority of health care decisions, are of varying quality, are scattered across multiple sources and are difficult to find, and ones developed internationally may not be readily applicable to the Australian context or vulnerable populations. Third, shared decision making is hampered for clinical questions where quality evidence does not exist or has not yet been synthesised. Finally, shared decision making in the area of Indigenous health and vulnerable populations (such as patients with poor health literacy) is important but presents additional challenges, and there is less research to guide implementation in these areas.36

Australia is drastically lagging behind many other countries in all aspects of shared decision making — policies, lobbying, advocacy, research funding, training, resources and implementation.41,42 In the absence of a coordinated national effort, we encourage individual clinicians to begin incorporating shared decision making into their consultations with patients when a health decision is needed. Australia’s health training and delivery organisations need urgently to begin prioritising and planning to make shared decision making a reality in Australia.

1 An example of one approach to shared decision making

Five questions that clinicians can use to guide shared decision making16

Example phrases (for the opening clinical scenario of a child with acute otitis media)

Comments

1. What will happen if we wait and watch?

“In children, most middle ear infections get better by themselves, usually within a week. The best options to control pain and fever are paracetamol or ibuprofen.”

Quantitative information can be provided where possible, either at this step or in step 3 where each option is described. When this is not possible, descriptive information can be provided (eg, “most people find that the symptoms go away by …”). Eliciting patient’s expectations about management of the condition (eg, “what have you heard or do know about …?”), including previously tested approaches and experiences, along with fears and concerns, is important and may occur here or later in the process.

2. What are your test or treatment options?

“Waiting for it to get better by itself is one option. Another option is to take antibiotics. Do you want to discuss that option?”

For some decisions (such as in this example), the options may be familiar to patients and need little elaboration at this step. In others, a more detailed explanation of each option and its practicalities, including options which are time-urgent, will be required.

3. What are the benefits and harms of these options?

“We know from good research that of 100 children with middle ear infection who do not take antibiotics, 82 will feel better and have no pain after 2–3 days. Out of 100 children who do take antibiotics, 87 will feel better after about 3 days of taking them. So, about five more will get better a little faster. We can’t know whether your child will be one of the five children who benefit or not.” (A graphic representation of these numbers can also be shown at this point, and again after the harms information is discussed [Box 2].) “There are some downsides to antibiotics though. Out of 100 children who do take antibiotics, 20 will experience vomiting, diarrhoea or rash, compared with 15 who do not take them. That means about five children out of 100 will have side effects from antibiotics. But again, we can’t know whether your child will have any of these problems. The other possible downside is antibiotic resistance — would you like to hear more about it?” (The option of delayed prescribing could also be discussed here. Numbers used in this scenario are from Venekamp et al.17)

In addition to descriptively discussing the benefits and harms of each option, the probability of each occurring, where this is known, should be provided. For dichotomous outcomes (eg, having a myocardial infarction), this should be in the form of natural frequencies (ie, the number out of 100 or 1000 people who experience the event) rather than relative risk. For continuous outcomes (eg, number of days of pain, and level of anxiety as reported on an anxiety measure), this may be expressed by the estimated size of the effect (eg, the average reduction in 20 points on anxiety on a tool that measured it as a score from 0 to 100). Decision support tools, if available, can be useful at this stage. Simple visual graphics can be particularly useful in helping to communicate the numbers. Principles of effectively communicating statistical information to patients should be followed, such as using natural frequencies (ie, x out of 100), being aware of framing effects, and using multiple formats.18 The discussion of harms should extend beyond the risk of side effects and include other impacts that the option could have on the patient, such as cost, inconvenience and interference with daily roles, and reduced quality of life.

4. How do the benefits and harms weigh up for you?

“With all I’ve said, which option do you feel most comfortable with?”

This step includes eliciting patients’ preferences and working with them to clarify how each option may fit with their values, preferences, beliefs and goals. Some decision aids include formal value clarification exercises that may be used to supplement the conversation and/or enable the patient to reflect further following the consultation. Clarifying the patient’s understanding of what has been discussed so far, using the teach-back method, can help to identify if any information needs to be repeated or explained in another way.

5. Do you have enough information to make a choice?

“Is there any more you want to know? Do you feel you have enough information to make a choice?”

This provides another opportunity to ask if the patient has additional questions. Patients may feel ready to make a decision at this stage or it may be jointly decided to defer the decision and plan when it should be revisited. The patient may wish to seek further information before deciding, discuss with family, or take time to process and reflect on the information received.

2 A simple graphic that can be shown to parents when discussing the benefits and harms of antibiotics for acute otitis media

3 Examples of types of decision support tools to facilitate shared decision making

Type of tool

Brief description

Examples

Condition-specific

Decision aids

Describe the options, and the benefits and harms of each option, along with a values clarification exercise and sometimes a guide to decision making. The International Patient Decision Aid Standards Collaboration (http://www.ipdas.ohri.ca) provides information about assessing the quality of decision aids

Ottawa Hospital Research Institute A to Z Inventory of Decision Aids: : http://decisionaid.ohri.ca/AZinvent.php

Decision or fact boxes

A short summary of the benefits and harms of an intervention, often presented in two columns

Dartmouth Medicine drug facts box: http://dartmed.dartmouth.edu/spring08/html/disc_drugs_we.php;
University of Laval decision box:

Option grids

A one-page summary of the evidence for the possible options, addressing patient-centred outcomes, and questions and concerns frequently raised by patients; can be useful within the consultation for a patient to highlight what is important to them

Option Grid Collaborative:

Question prompt lists

A predefined list of condition-specific questions for patients to consider using in a consultation

Cancer Institute NSW:

Evidence summaries

Clinical practice guidelines and other summaries of the body of evidence

Clinical Evidence: http://www.clinicalevidence.bmj.com; UpToDate:

Generic

Communication frameworks

A generic set of questions or scripts and a structure for clinicians and patients to use during decision making

Ask Share Know: http://www.askshareknow.com.au; Ottawa Personal Decision Guide:

4 Misconceptions about shared decision making and key research findings that refute them

Misconception

Research findings

The duration of the consultation will be lengthened

This concern is the most frequently reported barrier to shared decision making. Indeed, time constraint is the most frequently reported barrier to any clinical change.25

Three systematic reviews (on interventions to implement shared decision making, and impact of decision aids on processes of care and patient outcomes) indicate that there is no systematic increase in consultation duration when shared decision making is implemented or decision aids are used.15,26,27

Patients will be unsupported when making health care decisions

There is a fear that shared decision making will make patients feel abandoned during difficult decisions.

This is a misinterpretation of the intent of shared decision making. The definition of shared decision making explicitly describes patients and their clinicians sharing the decision together.3 Shared decision making is not about insisting that every patient make the decision (not all patients wish to); rather it helps to ensure that patients are informed about their options and are offered the opportunity to participate in the process.

Not every patient wants to share in the decision-making process with their clinician

Critics of shared decision making argue that not every patient wants to be involved in making decisions with their clinician.

An Australian survey reported that > 90% of women preferred a shared role with their doctor in making decisions about screening and diagnostic tests.28 A European survey of > 8000 people reported a high desire for shared health decision making (> 70% of the sample).29 A systematic review of 14 studies that examined the match between patient preferences about information and decision making with clinician–patient communication found that a substantial number of patients (26%–95%; median, 52%) were dissatisfied with the information given and would have preferred a more active role in decisions concerning their health, especially when they understood the expectations attached to this role.30 A time trend in desired involvement in decision making has also been reported. Of the studies conducted in 2000 and later, 71% reported that most respondents preferred a role in sharing decisions, compared to 50% of studies that were conducted before 2000.31

Most people are not able to participate in shared decision making

Critics of shared decision making question its complexity, believing most people will not be able to manage it.

Shared decision making is comprised of a set of behaviours on the part of the clinician and the patient that can be learnt.32,33 An increasing number of studies have demonstrated that shared decision making can be implemented successfully in clinical practice.27,34

Shared decision making cannot be used with vulnerable people

Shared decision making requires a special set of skills that may be too complex for all patients to acquire, and vulnerable people may not ever be able to share decisions with their clinicians.

Most surveys of patients’ willingness to engage in shared decision making show that the most vulnerable people are less willing to participate.30 Therefore, we need to be careful not to increase health inequities by offering shared decision making solely to the most privileged patients. Individuals with low health literacy want to be involved in health decisions but often lack the knowledge, skills and confidence to communicate with clinicians, navigate the health system and engage in shared decision making.35 They receive less information, ask fewer questions and are less satisfied with health care provider communication.36 More vulnerable patients may be less likely to engage in shared decision making because of lower self-efficacy — a modifiable factor to increase their willingness to do so.34 Indeed, in order to decrease health inequities, more needs to be done to engage the most vulnerable patients to make informed decisions.36,37

“I already do this”

  

Most clinicians feel that they already successfully engage their patients in shared decision making — a belief that may arise from not really understanding what it is and how to do it.

A systematic review of 33 studies that assessed shared decision making using the OPTION (Observing Patient Involvement in Decision Making) scale found low levels of patient-involving behaviours. In studies where no intervention was used to facilitate shared decision making, the mean OPTION score was 23  14 (0–100 scale, where higher scores indicate higher levels of patient involvement in decision making).21 Lack of familiarity with shared decision making by clinicians has been found to be a barrier to shared decision making in a number of studies.38

Engaging patients in shared decision making will raise their anxiety level

Some clinicians are afraid that shared decision making will raise patients’ anxiety levels as they become aware of the inherent uncertainty of evidence.

A Cochrane review of decision aids refutes this misconception, finding no effect on anxiety.15 Anxiety should not be confused with decisional conflict, an intrapsychological construct that reflects the difficulty that individuals can experience when comparing the pros and cons of different options.39

Out of the hands of milkmaids

From cowpox virus to global disease eradication, a winning strategy we must fight to protect

In 1914, the year this Journal was founded, there were 59 324 cases of diphtheria and 5863 deaths due to this disease in England and Wales. Mass vaccination was introduced in 1942, and by 1957 there were only 37 cases and four deaths.1 The situation was relatively similar in Australia. By May 2011, only one fatal case was reported in Australia, in an unvaccinated individual who contracted pharyngeal diphtheria from a friend who acquired the disease overseas.2

If I had known of the perils of being a child when I first set foot on this earth, I might have asked for a raincheck — such were the perils from infectious diseases.

In my first years, I dodged the diphtheria bullet and then I was vaccinated. During my childhood there was only triple antigen vaccine, covering diphtheria, pertussis and tetanus. Thus, inevitably, I contracted measles, mumps and chickenpox.

I remember being very sick with measles. So were my two sons, who were born in the 1960s. However, it was not until I was confronted with the situation of a close friend of mine, whose younger daughter contracted subacute sclerosing panencephalitis, that I really saw the impact as the family struggled to cope, watching the terrible decline in an intelligent, ebullient child. In fact, the mother, Gay Davidson, became a public advocate for vaccination when the Minister for Health and Ageing, Michael Wooldridge, committed his energy and funding to a vaccination campaign, even while Australia was in the throes of the 1996 federal Budget cuts.

At primary school, I remember the poliomyelitis epidemic. We still went to school, but interschool sport was curtailed. We were told not to eat ice-cream. Yet one day, a boy was at school; the next day he was in hospital, and within a week he was dead of poliomyelitis. Morbidity was high, with children in splints or confined to an iron lung.

In the United States, the March of Dimes (a non-profit organisation originally founded as the National Foundation for Infantile Paralysis by President Franklin D Roosevelt, who himself was afflicted by polio) provided substantial research funding. The Salk vaccine was the eventual product of this funding.3 I was in the first cohort to receive the Salk vaccine by injection in my last year at school. Then the oral Sabin polio vaccine appeared and made it far easier to convince the community to take its medicine. Despite the opposition of the March of Dimes bosses who favoured the Salk vaccine, use of Sabin vaccine ultimately prevailed.

My journey through the world of vaccine would not be complete without mention of the oldest vaccination (injecting cowpox virus to protect against smallpox), which gave the whole process its name (from the Latin vaccinus, meaning “from cows”). I needed a smallpox vaccination to go overseas to “the Orient”, as it was colourfully then called, on the SS Taiping in 1956. I also needed to have injections to protect against typhoid and cholera. There was no argument — you had to have them because in the world out there, diseases lurked where there was no clean water let alone sewerage.

Smallpox vaccine needed to be administered by a scarification technique, which was later repeated as a requirement for my medical course. Now smallpox is no more. Thanks to people like distinguished virologist Frank Fenner — the centenary of his birth is this year — this disease has been eradicated, and this is one vaccine consigned to history.

Back in the 1950s, vaccinations for overseas travel were compulsory; now, the only compulsory vaccination is that against yellow fever, and only for entry to certain countries if you have recently been in a designated yellow fever area.4 The typhoid and cholera vaccinations have been vastly improved, are administered orally and recommended for use only where the chances of contracting the disease are high.5 The importance of vaccination against yellow fever was highlighted some years ago by the controversy involving a federal politician when he returned from Venezuela to Australia, having not had a yellow fever vaccination before entering that country.6

What of other diseases where vaccination was unavailable? Unfortunately, I missed contracting rubella, but on at least two occasions when my wife was in early pregnancy I was exposed to rubella and suffered the painful injections of gamma globulin into the arms and both buttocks.

Tetanus toxoid has always presented an interesting situation because it is somewhat different. Originally released in 1914, the prime culprit for tetanus is the toxin produced by Clostridium bacteria. Tetanus is the one vaccination that always seems to be given when someone presents with traumatic injury that was contaminated by soil. Again, tetanus has almost been obliterated, however it was interesting to read recently about the New Zealand couple who objected to their children being vaccinated. One of the children contracted tetanus and almost died. This couple’s attitude towards vaccination changed overnight when they realised they had underestimated the diseases and totally overestimated the adverse reactions.7

I had grown up at the edge of a housing commission area where my father was the local doctor. He appeared one morning with bright yellow sclera and a tawny tinge to his skin; he had contracted infectious hepatitis, which required a time in Fairfield Infectious Diseases Hospital.

Now, as with many other diseases, the whole vaccination schedule has expanded greatly, much to the benefit of the community. The community has nearly covered the microbial waterfront.

Paediatrics has changed markedly over the 100 years in which the Journal has been published. Yet, in this community, we have large numbers of confused and deluded parents who resist having their children vaccinated, which not only puts those children at risk, but, if the immunisation rate falls sufficiently, puts entire communities at risk.

There are legitimate scientific reasons for not vaccinating a very small proportion of people. Such people should be protected by other therapeutic means.

However, there is no place for “conscientious objection” to vaccination. I received my first vaccinations when the world was at war and conscientious objection meant not wanting to bear arms. Now the conscientious objection under discussion is metaphorically not wanting to bare arms in a war against disease and disability.

Wilful objection to vaccination on the basis of spurious science should neither be encouraged nor rewarded, particularly by government. It is time for the whole question of conscientious objection to vaccination to be aired in Parliament if for no other reason than to find out which of our politicians are against vaccination, for they are as dangerous to continued Australian wellbeing as anyone who would challenge the biosecurity of our country.