Subscribe
×

Preference: Administration and Health Services

2

The opportunities of the new Australian Parliament

The composition of the new Federal Parliament provides excellent opportunities for the development of a health policy for Australia. 

This is because the many points of view that need to be reconciled to achieve a comprehensive and inclusive policy will be represented in the new parliamentary configuration. While this may at first sight seem clumsy and inefficient, it is a process critical to developing a policy that will guide health service provision and financing on behalf of the community.  As the Canadian commentator John Ralston Saul has written, the price of democracy is inefficient conversations – lots of them – that allow for all voices to be heard.

There is a further reason to be pleased that in the new arrangements a more cautious, inclusive and conversational approach will be applied to the development of a national health policy. It is that all too often, insufficient consideration is given to the unintended side effects of what appear to be bright new policy initiatives. 

Amartya Sen, a Nobel Prize winning Indian philosopher and economist, refers to this omission as one of the more common cardinal errors of social policy makers.  If asked, all the players may be able to provide more insights than one thinking alone. 

Simply having many players at the policy table does not, of course, guarantee freedom from this error.  As the Chilcot Report on the United Kingdom’s involvement in the recent Iraq war points out, a ‘coalition of the willing’ failed to question in depth what the consequences of war would be in the longer term. It was as though the policy stopped halfway.

The unintended side effects of the (good) policy to fund clinical psychologists to assist with the management of patients with mental health problems via general practice – workforce redistribution and budget over-runs – are examples of side effects that may have  been anticipated if more ‘thought experiments’ – thinking through what might follow – had been conducted prior to implementation.

By what process might this policy be developed? 

First, parliamentary leadership is required. A policy development oversight group that is genuinely multi-partisan should be established. This is not a matter of setting up yet another expert committee or commission of advice. The politicians need to lead. How the group wishes to proceed is, of course, entirely up to them.

Second, it is critical that high on the group’s agenda be a discussion about what Australia may reasonably expect from its health services, private and public, hospital and community, curative and preventive. There must be limits: what are they? How far do we wish to go in ensuring equity of access? How far in privatising the costs of health care? This is a special problem for patients who have serious and continuing complex problems, as my colleagues and I and many others have documented. Chronic illness is a fast track to poverty at present.

There are many topics to be discussed – which underlines my argument in favour of an inclusive conversation, auspiced by the Parliament, to begin. Attitudes vary in relation to prevention, and in the last Parliament a national agency for prevention was abolished. Is that what we want to do, or the best we can do?

And what are our expectations of research as a society? We know what experts and academics expect, but there are other voices as well that need to be heard, including those of some who have values espoused by science (and some who don’t).

Finally, there are ways of doing policy development well. From my personal perspective, I place a high premium on the contribution that solid data can make to the process. But my experience with policy development leaves me in no doubt that the ‘voice’ of data is but one voice. For a policy to work, data elegance is not enough. There must be buy-in from those whose lives and livelihoods are affected by it.

The new parliamentary structure requires a more humble and inclusive approach to policy formation.  Nowhere is this of more value than in working out where we as a nation are going with health and health care.

Rural doctors want support

In the first part of 2016, the AMA has been involved in three studies concerning rural health care. Briefly, they are:

1. AMA Plan for Better Health Care for Regional, Rural and Remote Australia.

This plan is focused on four clear areas needing our attention: rebuilding and cherishing the ‘good ol’ country hospital’; recruitment and retention of ‘fit-for-purpose’ rural doctors; fostering a new workforce of bright-eyed, bushy tailed young doctors eager to remain in the outback; and support of rural practices – this includes CPD, infrastructure and workforce distribution. 

2. Building a Sustainable Future for Rural Practice – the Rural Rescue Package.

A Joint AMA/RDAA Policy Statement. This is an easy-to-read, short paper with a rather brilliant suggestion for a two-tiered incentive system. The first tier is incentive payments and stipends to entice a happy workforce to stay and thrive in rural regions. The second tier of incentives is to encourage procedural and special skills in the rural workforce. Someone recently said that plumbers are so well compensated in the Outback that to find rural plumbing jobs is difficult. Can you imagine a future where rural medical positions were so sought after that doctors will be competing for that position? 

3.The 2016 Rural Health Issues Survey. 

It is this third study which is the most personal of the three. It is the voice of the rural doctors themselves speaking by way of a survey sent to them. The questions were posed by asking the doctors to rate the importance of 31 different proposals.

I was humbled when I read and re-read the results of the survey.

The way I read the results, the doctors clearly want support more than hard cash. None of the top 10 proposals would increase the income of the rural doctor. Instead, they want more of their kind with them, helping the roster, adding specialised services, locum relief, allied and ancillary staff, trainees to mentor and, of course, they want their family with them.

They also want to see support by way of “stuff”. Such amenities as walls, equipment, facilities. More evanescent, but definitely under the heading of amenities, is the need for broadband. These are doctors 500 kilometres from the nearest restaurant or movie theatre. They may be holed up on the other side of a flooded road in the Wet; they have no privacy – everyone in the community knows “the Doc”. These overworked souls need the worldwide web for sanity, for education, to communicate. Handy that it also helps with work.

Finally, they want to feel confident and competent. Every day there is a chance of going out of one’s comfort zone. They want opportunities to upskill, and support for skills that are not always needed but could be readily used. The support should come from colleges, that camera on the ceiling above the resus bed, the specialists on the phone encouraging them, and a medical board that is slow to criticise.

I tip my hat to you who answered the survey, you are fine people. You ask for things for your fellow doctors before you ask it for yourself. You are looking at the future where there is optimism, self-sufficiency and pride in your work. The choice of a rural location is not just a choice but a calling. You see yourself as part of the Outback, a key and respected member of that rural community. I see that you are trying to bring more doctors to the fold. As you watch another brilliant sunset, with a black cockatoo chiding at you and 10 roos ignoring you, be proud of your career path. You are the best. Thank you for being there.

 

Focus on rheumatic heart disease

 As the new President of the AMA I will, like my predecessors, chair the AMA’s Taskforce on Indigenous health. This recognises and emphasises the importance of the AMA’s efforts to improve the health and wellbeing of Aboriginal and Torres Strait Islander people, and our desire to keep ‘Closing the Gap’ initiatives at the top of our agenda.

The Taskforce, which was established in 2000, is comprised of representatives of the AMA Federal Council, AMA members and Indigenous health groups, including the National Aboriginal Community Controlled Health Organisation (NACCHO) and the Australian Indigenous Doctors’ Association (AIDA).

The Taskforce is a robust and dedicated entity which identifies, develops and recommends Indigenous health policy and strategies for the AMA and oversees the AMA’s annual Report Card on Indigenous Health.

This year, the 2016 Report Card on Indigenous Health will focus on rheumatic heart disease (RHD) – a major preventive health issue that significantly affects Indigenous people, particularly those in remote areas. As AMA WA President, I supported legislative change to improve reporting and reduce the burden of disease in Aboriginal communities in Western Australia.

RHD is a classic example of the many preventable chronic diseases that are largely responsible for the health gap between Indigenous and non-Indigenous people, with its burden largely extinguished in other parts of the Australian community. We can no longer allow the prevalence of chronic diseases like RHD to remain unaddressed.

The 2016 Federal election provided an opportunity for the AMA to present all political parties and candidates with the issues that the AMA sees as vital in meeting the challenge of closing the health gap.

While we have seen some recent improvements in Indigenous health, particularly in reducing infant mortality and smoking rates, the AMA wants to see the Commonwealth commit to improving resourcing for culturally appropriate primary health care for Aboriginal and Torres Strait Islander people.

The AMA has repeatedly said that it is not credible that Australia, one of the world’s wealthiest nations, cannot address the health and social justice issues that affect 3 per cent of its citizens. We say this again. The fact that it is our nation’s first people makes it an even greater moral imperative.

With the re-election of the Turnbull Government, the AMA will continue its call for long-term funding and commitments to Indigenous health. We will work closely with key ministers, government departments and other key stakeholders to ensure that appropriate action is taken.

As outlined in its Key Election Issues statement, the AMA urges the Federal Government to:

  • correct the under-funding of Aboriginal and Torres Strait Islander health services;
  • establish new, or strengthen existing, programs to address preventable health conditions that are known to have a significant impact on the health of Aboriginal and Torres Strait Islander people, such as cardiovascular disease (including rheumatic fever and rheumatic heart disease), diabetes, kidney disease, and blindness;
  • increase investment in Aboriginal and Torres Strait Islander community-controlled health organisations. Such investment must support services to build their capacity and be sustainable over the long term;
  • develop systemic linkages between Aboriginal and Torres Strait Islander community-controlled health organisations and mainstream health services to ensure high quality and culturally safe continuity of care;
  • identify areas of poor health and inadequate services for Aboriginal and Torres Strait Islander people and direct funding according to need;
  • institute funded, national training programs to support more Aboriginal and Torres Strait Islander people become health professionals to address the shortfall of Indigenous people in the health workforce;
  • implement measures to increase Aboriginal and Torres Strait Islander people’s access to primary health care and medical specialist services;
  • adopt a justice reinvestment approach to health by funding services to divert Aboriginal and Torres Strait Islander people from prison, given the strong link between health and incarceration;
  • increase funding for family violence and frontline legal services for Aboriginal and Torres Strait Islander people;
  • appropriately resource the National Aboriginal and Torres Strait Islander Health Plan to ensure that actions are met within specified timeframes; and
  • support the establishment of a Central Australia Advanced Health Research and Translation Centre. Central Australia faces many unique and complex health issues that require specific research, training and clinical practice to properly manage and treat and this type of collaborative medical and academic research, along with project delivery and working in remote communities, is desperately needed.

 Closing the gap in health and life expectancy between Indigenous and non-Indigenous people is an achievable task – it is also an agreed national priority.

The Federal Government must build on existing platforms and ramp up its ambitions to achieve health equality for Aboriginal and Torres Strait Islander people. Without commitment and action from our national leaders, the gap will remain wide and intractable. The Taskforce will inform the AMA’s advocacy in ensuring that this does not remain the case.

 

HIP4Hips (High Intensity Physiotherapy for Hip fractures in the acute hospital setting): a randomised controlled trial

The known Early mobilisation and daily physiotherapy after hip fracture fixation are recommended by guidelines, but there is no evidence that guides the intensity of acute hospital physiotherapy for this patient population.

The new Intensive acute hospital physiotherapy following an isolated hip fracture reduced hospital length of stay by more than 10 days without increasing complication or re-admission rates.

The implications We have provided evidence-based support for intensive physiotherapy programs in the acute hospital setting after hip fracture. Our findings may have significant practical implications, given the large number of inpatient beds occupied by this patient group.

Rising rates of hip fracture in our ageing population and the consequently increasing costs of care are significant problems for the hospital system and the community.1 In Australia, 17 000 hip fractures in people aged 65 years or more incur direct hospital costs of $579 million each year,2 and it is projected that the annual number of hip fractures will rise to 60 000 by 2051.1 Costs continue to accrue after patients leave hospital, with about 25% requiring full-time nursing home care3,4 and 50% of previously independent patients needing a gait aid or long term help with routine activities.3

Many strategies have been tried and guidelines developed for optimising the management of this population, including reducing the time to surgery and improving analgesia.5 Several rehabilitation strategies have also been investigated, including early assisted ambulation (within 48 hours of surgery)6 and multidisciplinary programs.7

There are no recommendations about the intensity of physiotherapy during the acute post-operative phase following hip fracture fixation. One cohort study found that increased intensity of acute hospital physiotherapy was associated with a greater likelihood of discharge home.8 Intensive physiotherapy for trauma patients was recently found to be safe, leading to more rapid improvements in mobility.9 Whether these findings can be extrapolated to patients with fractured hips or if early, intensive intervention has longer term benefits are, however, unknown.

The aim of our study was to investigate the effect of providing an intensive physiotherapy program for patients aged 65 years or more with isolated hip fractures.

Methods

Design and ethics approval

This was a single-centre, prospective randomised controlled trial at The Alfred, a level 1 trauma centre in Melbourne. Approval was granted by the Alfred Health Human Research and Ethics Committee (reference, 32/14). The study was registered with the Clinical Trials Registry in March 2014 (https://clinicaltrials.gov/ct2/show/NCT02088437).

Subjects

Between March 2014 and January 2015, all patients admitted to The Alfred with isolated hip fractures were screened for study inclusion. If patients were unable to provide consent, it was requested from the person responsible for their medical decisions. Patients were included if they were at least 65 years old, had been admitted with an isolated subcapital or intertrochanteric hip fracture, and were treated by internal fixation or hemiarthroplasty. Patients were excluded if the fracture was subtrochanteric or pathological, and if post-operative orders required the patient to be non-weight-bearing on the operated hip. Patients who were unable to move independently or who needed a gait aid prior to admission, and those admitted from nursing homes were also excluded.

Participants were randomly assigned by computer program to one of two groups: usual care (the control group), in which they received one daily treatment session of 30 minutes; or intensive physiotherapy (the intervention group), in which they received two additional daily treatment sessions of about 30 minutes each. Allocation was concealed by using opaque envelopes.

Physiotherapy

The usual care physiotherapist was blinded to the allocation, and treated all patients in the morning. The intervention group physiotherapist and allied health assistant provided physiotherapy during the afternoon; this was not documented in the patient’s medical record, maintaining the blinding of the usual care therapist and treating team.

Usual physiotherapy care

Patients in both groups received daily physiotherapy according to usual practice in the trauma centre, 7 days per week. Treatment was individualised and involved bed-based limb exercises (eg, strength exercises, such as knee extension, and active hip exercises) and gait re-training. The goal was early, independent transfer and mobility, with the objective of discharge directly home or to fast stream rehabilitation.

Intervention group

Patients in the intervention group received two additional daily sessions, 7 days per week. One session, delivered by an allied health assistant, practised the achievements of the morning session. An additional session, delivered by a physiotherapist, aimed to improve the functional advances achieved during the earlier physiotherapy session; eg, increased independence, progression of gait aid (eg, from frame to crutches), and increasing the distance walked.

Discharge criteria

A physician blinded to group allocation reviewed all patients, and determined medical stability and the need for inpatient rehabilitation. Participants were discharged home once they were medically stable, were deemed physically ready to return home by the blinded physiotherapist, and had been cleared by the multidisciplinary team. Physical readiness was defined as independence in bed and chair transfers, walking with a gait aid, and negotiation of any stairs required by the patient to safely enter or leave their home.9,10 If patients were unable to achieve these criteria before medical stability was achieved, they were transferred to an inpatient rehabilitation facility, per hospital protocol.

Data collection

Demographic data collected included sex, age, mechanism of injury, pre-morbid mobility level, social circumstances (residence, support at home) and home set-up (distance and steps to house entrance). Cognitive function was assessed by the physician with the Mini-Mental Status Examination, with scores ranging from 0 to 30 points: patients scoring 26–30 have no functional cognitive impairments, those scoring 20–25 display mild cognitive impairments, and those scoring less than 20 points having moderate to severe cognitive impairments and usually cannot live independently.11 Operative data, including fracture type (subcapital, intertrochanteric), operative time, post-operative weight-bearing status, and American Society of Anesthesiologists (ASA) score,12 were also collected. The ASA score provides a six-category physical status classification system, ranging from the normal healthy person (1) to brain-dead (6).

Outcome data were collected on post-operative Day 5, at discharge, and at 6 months. The primary outcome, the modified Iowa Level of Assistance (mILOA) score, was measured by a blinded assessor on post-operative Day 5 (or on the day of discharge, if this occurred earlier). The ILOA scale was originally designed for assessing patients with lower limb arthroplasty,13 and was later modified for patients who had fractured a hip.6 The mILOA is a functional score with six domains, including bed and chair transfer, ambulation, ascending/descending one step, and gait aid used. The score ranges from 0 (independent in all activities without a gait aid) to 36 (unable to attempt any of the activities). The scale has been validated in an acute hospital population14 and was found to be responsive in an earlier physiotherapy trial.9 Secondary outcome measures included a Timed Up and Go (TUG) test,15 administered by a blinded assessor to patients who were mobilising without assistance at Day 5 (a prerequisite for this test). Data on acute hospital length of stay (LOS), inpatient rehabilitation LOS and combined hospital LOS, inpatient complications, and re-admissions were also collected. Time to physical readiness for discharge, based on the criteria for discharge home applied by an elective joint replacement early discharge program,10 was also recorded. Other data included discharge destination from the acute hospital (home; slow or fast stream rehabilitation) and pain scores before and after physiotherapy. The amount of pain relief medication taken during the first 72 hours after surgery was recorded as the opioid equivalence score.16 Six-month outcome information was derived from routinely collected data in the Victorian Orthopaedic Trauma Outcomes Registry (VOTOR), including discharge destination, and Short Form 12 Health Survey (SF-12), Glasgow Outcome Scale (extended) (GOS-E), and EuroQol five dimensions health state questionnaire (EQ-5D) scores.17 When the patient was unable to provide a response to the EQ-5D, proxy responses were substituted.18

Statistical analysis

Ninety-two patients were required to detect a genuine difference of seven units on the mILOA scale at Day 5 (80% power, α = 0.05 [two-sided]),13 assuming that the standard deviation of the response variable was 11 (based on data reported for this outcome in a previous study19). This sample size calculation allowed for an anticipated drop-out rate of 15%.

Analyses were performed in Stata 12.0 (StataCorp). Data are presented as means and standard deviations, or as medians and interquartile ranges if the data were not normally distributed. Differences between groups at baseline were assessed with independent samples t tests (continuous data) or χ2 tests (categorical data). Data for the outcomes LOS, time to walk, time to sit out of bed, sum of hours, occasions of service, and physical readiness for discharge were not normally distributed, so these variables were natural log-transformed. Differences in outcome variables between treatment groups were compared by univariate analysis of variance (continuous variables) or with χ2 tests (categorical variables). The GOS-E data were dichotomised into scores of ≥ 7 (“good recovery”) or < 7 (“less than good recovery”)20 and analysed in χ2 tests. mILOA scores and combined LOS were assessed by linear regression; all potential confounders related to these outcomes in the univariate analysis (P < 0.2) were included in this analysis.

The effect of group allocation on time to discharge (as measured by combined LOS) was evaluated by Cox proportional hazard regression. Corresponding Kaplan–Meier curves were constructed and compared in log-rank tests.

Results

Between March 2014 and January 2015, 170 patients were screened; 68 were excluded, and 92 patients were recruited (46 allocated to each group). One control group patient was unable to participate in physiotherapy during the first 7 days because of a post-operative complication (dislocation prior to the initial physiotherapy intervention). No patients withdrew from the study, and Day 5 outcome measures were collected for all patients. The trial complied fully with CONSORT guidelines (http://www.consort-statement.org/); the CONSORT flow diagram for participation is included in the Appendix.

Demographic data for the two groups are summarised in Box 1. The groups were well matched for age, fracture type, and length of surgery, as well as for pre-operative mILOA and ASA scores. There was a significant difference between the groups for anaesthetic type, with patients in the intervention group more likely to have had a general anaesthetic. The patients in the intervention group were also less likely to have support at home, and there was a trend towards a greater proportion of women in this group.

There was no difference between the groups in the primary outcome measure, mILOA score on post-operative Day 5, with a mean difference of 2.7 points (mean scores: control group, 19.2 [SD, 8.4]; intervention group, 16.5 [SD, 9.4]; P = 0.10) (Box 2). However, after controlling for potential confounding factors (sex, anaesthetic type, carer at home, stairs at home), the Day 5 mILOA score was lower (better) in the intervention group (P = 0.04, Box 3). Hospital LOS (acute and inpatient rehabilitation) was 10.6 days shorter in the intervention group (median LOS: control, 35.0 days [IQR, 19.0–49.8]; intervention, 24.4 days [IQR, 16.4–31.6]; P = 0.01). This difference persisted in a linear regression that controlled for possible confounding factors (Box 4).

A Cox proportional hazards model that controlled for sex, anaesthetic type, presence of a carer at home, and presence of stairs at home found that the intervention conferred a benefit in terms of earlier discharge from hospital (hazard ratio, 2.30; 95% CI, 1.40–3.78; P < 0.001) (Box 5).

There was no difference between the two trial groups in mean pain or opioid equivalent scores over the first 3 post-operative days (Box 6), nor were there differences in the rates of complications or re-admission, or in longer term patient-reported outcomes (Box 2).

Discussion

We found that an intensive physiotherapy program beginning on the first day after surgery for an isolated hip fracture in patients over 65 years of age is safe and effective. There was no statistically significant difference between control and intervention groups in the primary outcome of functional mobility on post-operative Day 5, but hospital LOS was significantly lower for the intensive physiotherapy group. There were also no differences between groups with respect to pain levels or opioid pain relief requirements. These findings further support the current guidelines, which recommend physiotherapy after a hip fracture to facilitate early assisted ambulation.

After controlling for relevant confounders, our primary outcome, mILOA score, was significantly better in the intervention group at Day 5 (Box 3). Functional mobility at discharge is a major determinant of mortality,21 and whether this improved mobility affects longer term outcomes should be investigated. We powered our study to detect a 7-point difference, as this has been reported as the smallest difference in mILOA score of clinical importance.13 The between-group difference in favour of the intervention group did not reach this threshold, so the difference we found may not be clinically significant. Unfortunately, we only collected data on the primary outcome measure at one post-operative time point; Day 5 may be too early to detect a substantial difference in functional mobility in this patient group, and assessing function at longer time intervals would be desirable.

The economic burden of hip fractures is great;3,22 they account for 14% of all osteoporotic fractures, but for 72% of costs, which are largely associated with inpatient care.22 We found that hospital LOS (a secondary outcome measure) was shorter in the intervention group, with a 10-day reduction across acute and subacute care, reflected by a similar improvement in the time to physical readiness for discharge. This was achieved without an increase in rates of in-hospital complications or re-admissions, and without affecting 6-month outcomes. The difference in LOS is still evident well into the rehabilitation period (Box 5). This may represent an important cost saving, given the large numbers of patients requiring management in hospital systems worldwide, and the high costs associated with their care.

The strengths of our study were its robust design, the randomisation protocols, and the blinding of both the usual care physiotherapist and the outcome assessor to trial group allocation. No patient was lost to follow-up during the in-hospital phase of data collection, and 6-month outcome scores could be collected for more than 85% of participants.

The limitations of our study include the relatively small numbers in each trial group (the result of the single institution trial design) and the short duration of the intervention (limited to acute hospital stay). Although our cohort was randomised, there were minor differences between the groups at baseline; more patients in the intervention group had received a general anaesthetic, and fewer had support at home. These differences highlight the importance of our findings, as previous studies have shown that patients who receive a general anaesthetic have an increased rate of post-operative complications,23 and those without support at home are more likely to require assistance after discharge and to be discharged to a higher level of care.24 Further, most patients were at home prior to admission, limiting the generalisability of our findings. We have not provided a robust cost analysis of the project or detailed information regarding resource allocation, although the financial benefit of reducing hospital LOS is clear.

Conclusion

We found that an intensive physiotherapy program for patients in the acute care setting after an isolated hip fracture can reduce hospital LOS by more than 10 days without increasing the rates of complications or re-admissions. Our study provides support for providing intensive physiotherapy programs for those with hip fractures, and this should be considered in future care guidelines.

Box 1 –
Demographic data for the usual care (control) and intensive physiotherapy (intervention) groups

Characteristic

Usual care

Intensive physiotherapy

P

Number of patients

46

46

Age (years), mean (SD)

81.3 (9.0)

81.3 (7.5)

1.00

Sex

Men

21 (46%)

12 (26%)

0.05

Women

25 (54%)

34 (74%)

Fracture type

Subcapital

30 (65%)

25 (54%)

0.29

Intertrochanteric

16 (35%)

21 (46%)

Pre-operative modified Iowa Level of Assistance score, mean (SD)

1.74 (2.80)

1.39 (2.56)

0.54

Pre-operative American Society of Anesthesiologists score

1

4 (9%)

1 (2%)

0.36

2

10 (22%)

15 (33%)

≥ 3

32 (70%)

30 (65%)

Mini-Mental Status Examination score*

26–30

5 (14%)

6 (17%)

0.15

20–25

16 (44%)

8 (23%)

< 20

15 (43%)

21 (60%)

Residence

Home

44 (96%)

42 (91%)

0.24

Retirement village

1 (2%)

0

Low level care

1 (2%)

4 (9%)

Carer at home*

Yes

24 (53%)

14 (31%)

0.03

No

21 (47%)

31 (69%)

Stairs at home

Yes

36 (78%)

29 (63%)

0.11

No

10 (22%)

17 (37%)

Health care funding*

Private

8 (21%)

13 (31%)

0.36

Medicare

30 (79%)

28 (67%)

Transport Accident Commission

0

1 (2%)

Anaesthetic type

General

34 (74%)

42 (91%)

0.03

Spinal

12 (26%)

4 (9%)

Length of surgery (minutes), median (IQR)

82 (62–101)

74.5 (55–101)

0.96

* Missing data mean that numbers do not add to column total.

Box 2 –
Outcomes for the usual care (control) and intensive physiotherapy (intervention) groups

Outcome

Usual care

Intensive physiotherapy

P

Number of patients

46

46

Modified Iowa Level of Assistance (mILOA) score, mean (SD)

19.2 (8.4)

16.5 (9.4)

0.15

Timed Up and Go conducted

16 (35%)

24 (52%)

0.09

Timed Up and Go (seconds), median (IQR)

69 (51.5–96.5)

47 (31–89)

0.08

Acute hospital length of stay (days), median (IQR)

7.4 (5.4–8.8)

6.0 (4.6–7.2)

0.08

Inpatient rehabilitation length of stay (days), median (IQR)

29.9 (15.9–42.9)

21.0 (13.8–26.8)

0.06

Combined hospital length of stay (days), median (IQR)

35.0 (19.0–49.8)

24.4 (16.4–31.6)

0.01

Time to physical readiness for discharge (days), median (IQR)

27.7 (14.0–37.6)

16.3 (5.8–25.6)

0.02

Physiotherapy occasions of service, median (IQR)

5 (3–6)

10.5 (7–14)

< 0.001

Time spent in physiotherapy (hours), median (IQR)

115 (70–160)

248.5 (145–290)

< 0.001

Time to sit on bed (days), median (IQR)

1.02 (0.9–1.8)

1.0 (0.9–1.5)

0.43

Time to walk 3 metres (days), median (IQR)

1.95 (1.0–4.6)

1.93 (1.1–3.0)

0.19

Discharge destination from acute hospital

0.31

Home

5 (11%)

10 (22%)

Fast stream rehabilitation

11 (24%)

12 (26%)

Slow stream rehabilitation

30 (65%)

24 (52%)

Final discharge destination

0.86

Home

33 (72%)

36 (78%)

Low level care

7 (15%)

6 (13%)

High level care

6 (13%)

4 (9%)

In-hospital complications

Hip dislocation

1 (2%)

0

0.32

Troponin levels > 26 ng/L

3 (7%)

5 (11%)

0.46

Anaemia requiring transfusion

9 (20%)

11 (24%)

0.61

Acute kidney injury

8 (17%)

4 (9%)

0.22

Re-admission within 6 months

0.25

Hip-related

4 (9%)

2 (4%)

Not hip-related

12 (26%)

3 (7%)

6-month Glasgow Outcome Scale (extended) (GOS-E) score*

0.10

< 7

26

17

≥ 7

15

21

6-month EuroQol (5 dimensions) (EQ-5D), mean score (SD)†

63.1 (22.3)

70 (16.7)

0.15

6-month SF-12, mental components, mean score (SD)‡

59.3 (3.7)

56.8 (7.0)

0.27

6-month SF-12, physical components, mean score (SD)‡

44.3 (11.8)

41.7 (11.9)

0.56

SF-12 = Short Form 12 health survey. * n = 41 (control group), n = 38 (intervention group). † n = 36 (control group), n = 36 (intervention group). ‡ n = 12 (control group), n = 18 (intervention group).

Box 3 –
Linear regression analysis of the relationship between modified Iowa Level of Assistance (mILOA) score and group allocation, after adjusting for relevant confounders (sex, anaesthetic type and home setting)

Unstandardised coefficient for mILOA score (standard error)

P

Group allocation (intervention v control)

–4.12 (1.99)

0.04

Sex (men v women)

–0.11 (1.96)

0.95

Anaesthetic type (general v spinal)

–6.59 (2.64)

0.02

Carer at home (v no carer at home)

1.25 (1.95)

0.52

Stairs at home (v no stairs at home)

–1.53 (2.13)

0.48

Box 4 –
Linear regression analysis of the relationship between hospital length of stay and group allocation, after adjusting for relevant confounders (sex, anaesthetic type and home setting)

Unstandardised coefficient for length of stay (standard error)

P

Group allocation (intervention v control)

–16.80 (4.90)

0.001

Sex (men v women)

7.64 (4.83)

0.12

Anaesthetic type (general v spinal)

–15.60 (6.52)

0.02

Carer at home (v no carer at home)

–10.06 (4.82)

0.04

Stairs at home (v no stairs at home)

4.01 (5.26)

0.45

Box 5 –
Adjusted Kaplan–Meier analysis of the probability of discharge, after adjusting for sex, anaesthetic type, carer at home, and stairs at home*

* The probability of discharge in the intervention group was greater than for control patients at all time points (P < 0.001).

Box 6 –
Pain and pain management scores for the usual care (control) and intensive physiotherapy (intervention) groups during the first 3 post-operative days

Outcome

Usual care

Intensive physiotherapy

P

Number of patients

46

46

Post-physiotherapy pain scores, mean (SD) (maximum score: 10)

Day 1

5.3 (0.5)

4.3 (0.4)

0.16

Day 2

4.2 (0.5)

4.7 (0.5)

0.44

Day 3

4.0 (0.5)

4.2 (0.5)

0.79

Opioid equivalence score, mean (SD)*

0–24 hours post-operative

20.0 (14.0)

20.9 (15.8)

0.81

24–48 hours post-operative

26.3 (16.6)

34.1 (26.7)

0.12

48–72 hours post-operative

27.0 (19.8)

29.2 (22.9)

0.67

* Reference analgesic is oral morphine (mg).

No strong evidence bicycle helmet legislation deters cycling

A focus on helmet legislation detracts from concerns about cycling infrastructure and safety

Opponents of helmet legislation often argue that mandatory bicycle helmet legislation (MHL) is the primary impediment to an increase in cycling.1 The public debate regarding MHL recently flared up with the Leyonhjelm Senate inquiry2 and the Australian Capital Territory proposing a relaxation of their MHL in low speed areas.3 As there are numerous health and social benefits to cycling, such arguments need to be evaluated with rigour against the highest quality evidence available.

Victoria was the first jurisdiction in the world to adopt MHL in July 1990. Other Australian states and territories adopted similar legislation by July 1992. Around this time, the governments of New South Wales, Queensland, South Australia, Victoria and Western Australia commissioned research to assess the impact of their state-specific laws on helmet wearing.4–8 In SA and WA, stratified random sampling surveys were employed to estimate cycling frequencies before and after the introduction of MHL.6,8 The results of these surveys suggest that there was no real impact on cycling frequencies following MHL (Box 1). Cycling frequencies were not estimated for NSW, Queensland or Victoria.

Recent evidence suggests that active travel modes (ie, walking, cycling and public transportation) steadily declined following World War II because more Australians used cars as their primary transportation mode.9 In a 2011 survey regarding barriers to cycling, over 50% of the responses from current cyclists and non-cyclists related to lack of cycling infrastructure and concerns about safety (Box 2).10 Dislike of helmets constituted only 6% of the responses for both groups and was the tenth and 13th most cited response among cyclists and non-cyclists respectively. Yet, MHL is sometimes proposed as a major barrier to cycling.11

Complex study designs, such as the stratified random samples discussed above, are essential for estimating cycling frequencies to ensure that the sample is representative of the population. Stratified random sampling surveys have long been used in countries with an entrenched cycling culture, such as the Netherlands,12 and also recently in Australia since 2011 through the National Cycling Participation Survey conducted by Australian Bicycle Council (http://www.bicyclecouncil.com.au/publication/national-cycling-participation-survey-2015). There is a lack of international evidence using such methods to support the case that MHL acts as a deterrent to cycling. A Canadian study, in which participants were randomly identified from three sampling frames with elements of stratification and clustering, found no evidence that MHL deters cycling.13

Although complex survey designs are important for obtaining reliable data, they also require many resources and are not always feasible. On the other hand, convenience sampling often requires fewer resources, making it attractive for some studies. However, the cost of convenience is a biased sample. In NSW and Victoria, convenience sampling was used to collect data at various roadside locations in each state over one time period before the introduction of MHL and multiple time periods after its introduction. In addition to bias from convenience sampling, these studies were designed to estimate helmet wearing not cycling frequency. Nonetheless, some authors have used the cyclist counts from these surveys to demonstrate a reduction in the number of cyclists following the introduction of helmet legislation. There is an argument that convenience samples taken over time are representative of population trends; however, such an argument involves many assumptions that are almost always violated.14

In summary, there are reports indicating a decline in cycling based on convenience sampling data following the introduction of helmet legislation in Australia. However, there is also evidence based on better quality data which shows no significant impact on cycling participation. When faced with conflicting evidence, it is important to consider differences in study design and data quality. When these parameters are taken into account, the best evidence suggests that MHL has never been a major barrier to cycling in Australia. In addition, the focus on helmet legislation detracts from more important discussions around the uptake of cycling. These include concerns for personal safety, which can be addressed by the construction of dedicated cycling infrastructure,15 education of all road users, and supportive legislation to protect cyclists, such as minimum passing distances.16

Box 1 –
Frequency of cycling before and after the introduction of helmet legislation in South Australia and Western Australia, according to stratified random sampling surveys6,8

Frequency

1990

1993

1989

1993

South Australia

At least weekly

21.8%

21.0%

At least monthly

5.2%

6.0%

At least every 3 months

3.9%

4.4%

Less often or never

69.1%

68.6%

Western Australia

At least weekly

26.6%

27.7%

At least every 3 months

11.1%

11.6%

At least once per year

10.3%

11.5%

Never

52.0%

49.2%

Box 2 –
Percentage of responses for reasons for not riding a bike for transport more frequently (current cyclists) or for transport (non-cyclists), 2011*

Response category

Current cyclists (n = 158/386)

Non-cyclists (n = 515/1289)

Lack of infrastructure†

30.6%

25.0%

Concerns about safety‡

27.7%

25.8%

Uncontrollable issues§

26.9%

24.4%

Do not like wearing a helmet

6.5%

6.3%

Miscellaneous

8.3%

18.6%

* Adapted with permission from the National Heart Foundation of Australia (numbers of respondents/responses).10 † Speed/volume of traffic, lack of bicycle lanes/trails, no place to park/store bicycle, no place to change/shower, nowhere to store clothes. ‡ Unsafe road conditions, do not feel safe riding, do not feel confident riding. § Weather conditions, destinations too far away, too hilly, not enough time, health problems. ¶ Not fit enough, none, unsure of best route, other, do not own a bicycle, need to transport other people (eg, children), do not know how to ride a bike.

News briefs

Unnecessary EOL treatment widespread

University of New South Wales reviewers, reporting in the International Journal for Quality in Health Care, have found that more than a third of elderly patients hospitalised at the end of their life received “invasive and potentially harmful medical treatments”. The analysis of 38 studies over 2 decades, based on data from 1.2 million patients, bereaved relatives and clinicians in 10 countries including Australia, found that the practice of doctors initiating excessive medical or surgical treatment on elderly patients in the last 6 months of their life continues in hospitals worldwide. Dr Magnolia Cardona-Morrell, who led the research at UNSW’s Simpson Centre for Health Services Research, said rapid advances in medical technology have fuelled unrealistic community expectations of the healing power of hospital doctors and their ability to ensure patients’ survival. “It is not unusual for family members to refuse to accept the fact that their loved one is naturally dying of old age and its associated complications and so they pressure doctors to attempt heroic interventions,” Dr Cardona-Morrell said. “Doctors also struggle with the uncertainty of the duration of the dying trajectory and are torn by the ethical dilemma of delivering what they were trained to do, save lives, versus respecting the patient’s right to die with dignity.” The study revealed 33% of elderly patients with advanced, irreversible chronic conditions were given non-beneficial interventions such as admission to intensive care or chemotherapy in the last two weeks of life while others who had not-for-resuscitation orders were still given CPR. The researchers also found evidence of invasive procedures, unnecessary imaging and blood tests, intensive cardiac monitoring and concurrent treatment of other multiple acute conditions with complex medications that made little or no difference to the outcome, but which could prevent a comfortable death for patients.

Breast cancer treatment impacts independent living

Researchers from the US have found one in five women undergoing breast cancer treatment for a year became “incapable of performing some of the basic tasks required for independent living”. Published in Cancer, the study also found that a simple survey can help identify which women are at risk of such functional decline. Cynthia Owusu from Case Western Reserve University in Cleveland, Ohio, and her colleagues studied a group of 184 women aged 65 years and older who had been recently diagnosed with stage I to III breast cancer. The researchers used the Vulnerable Elders Survey, a 13-item self-administered tool that has been validated in community-dwelling elders to predict functional decline or death within 12 months. Patients completed the survey just prior to breast cancer treatment. Within 12 months, 34 of the 184 patients developed functional decline and seven died. The risk of functional decline or death rose with increasing survey scores. Women without an education beyond high school were disproportionately affected. “Our findings are important because the study validates the Vulnerable Elders Survey as a useful tool for identifying older women with breast cancer who may be at increased risk for functional decline within a year of treatment initiation,” she said. “This instrument offers the opportunity for early identification and will inform the development of interventions to prevent and address functional decline for those particularly at risk, such as women with low socioeconomic status.”

[Correspondence] “The Bedouin predicament”

I was moved when reading Richard Horton’s Offline: The Bedouin predicament (April 9, p 1498) about the semi-Nomadic Bedouins and the poverty and illiteracy that challenge provision of health services in the Negev desert.1 The UK also has semi-Nomadic communities: Gypsy and Traveller families. Of all the many ethnic groups recognised in the Census, “white Gypsy or Irish Traveller” communities have the worst health—for example, in terms of limiting long-term illness.2 These populations experience multiple inequalities3 likely to undermine health in every generation.

Use of medicines by older people with type 2 diabetes

This report describes dispensing patterns of glucose lowering medicines and medicines for other conditions associated with diabetes in a concessional population cohort of Australians aged 65 and over diagnosed with type 2 diabetes. It uses linked data from the National Diabetes Services Scheme and the Pharmaceutical Benefits Scheme to explore medicine supply patterns in 2012 by age and time since diabetes diagnosis. It shows that, in general, the longer the time since diagnosis, the more likely it is that an individual would be supplied with all medicine types and the more intense their glucose lowering treatment regimens would be. This report highlights the complexity of pharmacological management in older people with type 2 diabetes and the diversity of medicine supply patterns in relation to age and time since diabetes diagnosis.

Prioritising general practice research

Cuts to federal funding put us in grave danger of wasting the investment made to achieve current gains in research capacity

General practice is critical to the provision of primary health care (PHC) for Australians. About 85% of the Australian population claim at least one general practice service from Medicare per year.1 Over 137 million such consultations were delivered by 33 279 general practitioners in 2014–15.1 In 2011–12, PHC spending was $50.6 billion (36.1% of total health expenditure), with $28.6 billion spent on predominantly general practice-based medical services and medications.2 Multiple studies have shown that a strong PHC system is associated with greater efficiency, lower rates of hospitalisation, fewer health inequalities and better health outcomes, including lower mortality.3 Thus, ensuring that the cornerstone of PHC delivery, general practice, has a robust evidence base is of paramount importance. Despite this, there are major gaps in the evidence supporting clinical practice and health service delivery in general practice.

From the perspective of general practice, health and medical research appears poorly targeted. There is a mismatch between the burden of diseases commonly managed in general practice and the number of randomised controlled trials exploring their effective management,4,5 and between the frequency with which conditions are encountered in general practice and publication rates of research and clinical guidelines.6 There are unmet needs for evidence specific to general practice in the recognition and management of early stage disease; assessment and management of disease risk for prevention; and patient care in complex situations, including multimoribidity.7

Recommendations for PHC made in disease-specific clinical guidelines are often based on evidence from studies performed outside general practice. For example, in 22 National Institute for Health and Care Excellence guidelines, only 38% of the publications that were cited to support primary care-relevant recommendations had studied patients typical of primary care.8 Yet, patients seen in tertiary referral centres with morbidities commonly seen in general practice differ markedly from the average general practice patient. This affects judgements on the risk–benefit ratio of treatments and the performance of diagnostic tests. Treatment benefits are usually higher in people at higher risk of adverse outcomes from their disease (eg, those attending tertiary referral centres) than in lower-risk patients seen in general practice, so testing an intervention in a hospital setting and then applying that expected benefit to general practice patients may result in overtreatment or less cost-effective treatment.

Further, exclusion criteria of clinical trials frequently eliminate patients with complex conditions, leading to potential underestimation of the risks of harm from treatment for general practice patients, in whom multimorbidity is common. In one example of this, at best 21% and at worst only 3% of “ordinary patients” would have been eligible to participate in a range of trials of osteoporosis treatments.9 Most were excluded due to comorbidities and concurrent medications such as glucocorticoids, but these patients still require treatment in the real world of general practice! The dangers of applying diagnostic testing in general practice, where diseases typically have a lower prevalence or are early stage, are also known. In an Australian study, just over half of general practice patients with fatigue had pathology testing performed, but only 3% had a significant clinical diagnosis based on an abnormal pathology test result.10 The low pre-test probability of illness in general practice patients11 reduces the post-test probability of them having a condition, given a positive test result. As post-test probability drives clinical decision making, applying an overestimated post-test probability observed in a high risk population to general practice patients can lead to inappropriate management.

None of this is news. Recognition of the importance of general practice research to the profession, government and community has been increasing since the 1990s, coinciding with a period of substantial commitment to building general practice and PHC research and evaluation capacity. As early as 1990, the General Practice Evaluation Program (GPEP) was established to develop the evaluation skills of researchers in general practice.12 The GPEP was part of the federal government’s General Practice Strategy, the review of which, in the late 1990s, increased focus on the need for a strong general practice research culture and workforce, with part of its vision into the 21st century being that GPs would be “actively involved in research, evaluation and teaching and be appropriately remunerated for these activities”.12

In 2000, the Primary Health Care Research, Evaluation and Development (PHCRED) program commenced, with four key components:

  • the Australian Primary Health Care Research Institute (APHCRI), with a focus on providing leadership in PHC research;

  • the Research Capacity Building Initiative (RCBI), which funded university departments of general practice and rural health to provide training and support in PHC research, particularly for GPs;

  • PHC research grants, administered through the National Health and Medical Research Council (NHMRC); and

  • the Primary Health Care Research and Information Service (PHCRIS), established to support dissemination and exchange of knowledge.

These efforts have resulted in many high quality research projects being conducted, and their results published. A small sample of such research conducted by Australian GPs includes a decade of research altering post-excision wound care in general practice13,14 and publications in the world’s top clinical medical journals on other important matters for GPs and their patients, such as counselling for women disclosing intimate partner violence15 and treatment for chronic knee pain.16

So why do we have to justify the importance of prioritising general practice research yet again? Put simply, the plug is being pulled on federal funding when the job is only partly done! The RCBI was defunded in 2011. More recently, funding to APHCRI has ceased; PHCRIS has just received a 6-month reprieve of withdrawal of funding; and the federal government contribution to the Bettering the Evaluation and Care of Health (BEACH) program has ceased. BEACH has been the major source of data on general practice activity in Australia. Government provided only about 18% of the total funding for this longstanding program (Associate Professor Helena Britt, University of Sydney, personal communication) but, combined with other difficulties in securing funding, BEACH now has to close.17 This incredibly cost-effective source of data for government and the profession will not readily be replaced.

Two ways to measure general practice research capacity and output are publication rates and numbers of successful doctoral candidates. General practice publication rates increased threefold from 1990–1999 to 2000–2007,18 which sounds impressive until one examines the absolute figures — this still amounts to only three publications per 1000 GPs per year, comparing poorly with rates for physicians (160/1000/year) and surgeons (68/1000/year).18 From 2005 to 2014, 76 GPs were awarded a PhD from an Australian university.19,20 There are no corresponding data for other disciplines but, in the same period, at least 200 physicians completed a doctorate under NHMRC scholarships.

Access to research funding for general practice research remains very challenging. The General Practice Strategy review noted that, “in Australia there is a considerable imbalance between the amount of funding devoted to research in secondary care compared with what is spent on primary and community-based research (including general practice)”.12 This imbalance remains. From 2000 to 2008, only about 1.9% of NHMRC-administered grants were PHC-related, and over a quarter of these were actually funded through the PHCRED strategy.21 This suggest two things. First, the moiety of research funds directed at PHC research is completely out of proportion when one considers the health dollars spent on PHC and its importance to the community. Second, PHC researchers may not yet be sufficiently developed to be competitive in what is arguably the most important health and medical research funding stream to which they have access in Australia.

This latter point is critically important. Underpinning the decisions to withdraw funding from capacity-building initiatives appears to be a failure to recognise that progressing high-performing general practice researchers from PhDs to independent research leaders is a long term challenge. Specific people support for the career development of general practice researchers has been sporadic and inconsistent. From 2009 to 2014, NHMRC grant outcome data show that eight general practice or PHC Early Career Fellowships (available to researchers within 2 years of achieving their PhD) were awarded, but none subsequent to this.22 In 2011, four NHMRC/PHCRED Career Development Fellowships (aimed at people 2 to 12 years after achieving their PhD) were awarded.23 Data before 2013 specifying if other levels of people support for GP researchers were awarded are not available on the NHMRC website at this time, but from our knowledge of the general practice research environment, this support has been minimal.

General practice research workforce issues are substantial. The challenge of supporting general practice research is now left to the university departments of general practice and rural health, which are hamstrung by lack of resourcing for capacity-building since the RCBI funding was withdrawn. As they are currently structured, general practice academic departments will find it hard to increase their activities. From examining university websites, we estimate that in June 2015 there were 72 GPs with Level D or E (Associate Professor or Professor) appointments employed at Australian universities, only 52 of whom had a doctorate. Only two universities had a sufficient concentration of these appointments at the one site to make an effective potential research team. A quarter of these GP academics were aged 60 years or older. The median time since primary medical qualification for successful GP PhD candidates was 23 years, giving them a potential median research career of about 17 years.20 Replacing the ageing GP research workforce will be difficult if this low rate of young doctoral students continues.

There are other problems affecting general practice research capacity, including:

  • lack of funding for the infrastructure needs of general practices to participate fully in data collection and research;

  • lack of support for practice-based research networks (PBRNs);

  • lack of clearly defined clinical research career pathways; and

  • the ongoing need for strengthening research culture within general practice.

PBRNs provide the equivalent of the biomedical laboratory for primary care. Currently, there are 17 university-linked PBRNs in Australia, whose research efforts are coordinated by the Australian Primary Care Research Network (APCReN). These networks have contributed to quality Australian research, including some mentioned above and the internationally important ASPREE randomised controlled trial of low dose aspirin for primary prevention in healthy older people,24 which has recruited over 16 000 general practice patients. Overseas, public funding has proved essential for the development and governance of PBRNs,25 but federal government funding of APCReN has been limited and may not be ongoing.

Prioritising general practice research was a key part of the Royal Australian College of General Practitioners’ (RACGP’s) pre-Budget submission to the federal government this year.26 The RACGP’s recommendations (Box), if implemented, would go a long way towards overcoming the obstacles blocking the continued development of sustainable, long term general practice research capacity for Australia. This is doable, through a more balanced distribution of current research funds to prioritise general practice research or through the Medical Research Future Fund, given the Australian Government’s wish “to support the sustainability of the health system and drive medical innovation through transforming how health and medical research is conducted in Australia”.27 In the meantime, we are in grave danger of wasting the investment made to achieve current gains in capacity, leaving our profession and the Australian population to make do with a severely restricted evidence base to support PHC in this country.

Box –
Recommendations of the Royal Australian College of General Practitioners (RACGP) for prioritising general practice research*

The RACGP recommends that the federal government prioritises primary health care research and:

  • commits $27 million over nine years to establish a general practice research fellowship program, offering eight 4–5-year fellowships to develop general practitioner (GP) research leaders
  • allocates 10% of the National Health and Medical Research Council’s (NHMRC’s) project grants budget to general practice-specific research projects (ie, projects with direct relevance to general practice and which involve one or more GPs as chief investigators)
  • invests $2.5 million to establish an NHMRC Centre for Research Excellence in General Practice/Primary Care
  • invests $200 000 per annum to support the maintenance of practice-based research networks, specifically the Australian Primary Care Research Network (APCReN)
  • provides $2 million per annum across university departments of general practice and rural health to facilitate practice-based research networks
  • implements a practice incentive payment to enable practices to facilitate and implement research

* Reproduced with permission from the Royal Australian College of General Practitioners. 2016–17 pre-budget submission. East Melbourne, Vic: RACGP, 2016. Available at: http://www.racgp.org.au/yourracgp/news/reports/20160205prebudget