Too much of a good thing

We all understand the role of clinical trials in testing new medicines but many people would  be surprised to find out that there are different regulatory processes in place for the introduction of new medical devices and surgical techniques and procedures.

This can mean that the evidence-base for these components of health care is less than those for medicines and therefore there is potential for them to cause unnecessary harm.

Below Professor Ian Harris, Professor of Orthopaedic Surgery at UNSW, discusses the problems with the current regulatory environment and suggests changes to improve the safety and quality use of devices and surgery.

This article is published as part of the TOO MUCH of a Good Thing series, which is investigating how to reduce overdiagnosis and overtreatment in Australia and globally, and is published as a collaboration between Wiser Healthcare and Croakey.

To follow the series, bookmark this link, and follow #WiserHealthcare on Twitter.

---

Ian Harris writes:

Most people have some idea of the rigorous testing that is required for any drug to be accepted for clinical use and government reimbursement: the drug must be compared to other drugs or a placebo in a rigorous scientific trial in terms of potential benefits and harms – usually a randomised trial and often with patients ‘blinded’ to what treatment they are receiving.

And only when these tests have been reviewed in detail by expert panels will they be approved for any kind of government reimbursement. It’s a high bar and it needs to be, otherwise public money may be wasted and/or people may be unnecessarily harmed.

Regulation of medical devices

How high is the ‘bar’ for surgery to clear before acceptance? Well, it depends. For a start, there are two different pathways: one for devices (things that are implanted in the body, like joint replacements, stents and heart valves) and one for procedures. Implantable devices are not required to have high quality (randomised trial) evidence comparing them to previously approved devices or placebo. In fact, they just need to show that they are “similar” (structurally) to existing (already approved) devices and that they are safe.

This ‘low bar’ for devices has led to problems like the worldwide recall of some hip replacement devices after being used in thousands of people. The mechanical properties and structural design of these hip replacement devices were very similar to already-approved devices and they passed the approval process based on that similarity. Unfortunately, however, the subtle differences in design in these new hip replacement devices were enough to cause catastrophic failure in a large number of cases – something that would have been easily detected if clinical evidence of actual effectiveness in real people had been required, like it is for drugs.

Some years ago, partly in response to such cases as described above, the government made “clinical evidence” for new devices a requirement for acceptance for clinical use, yet the type or standard of clinical evidence has never been made clear. Simply showing that the device has been implanted in some people without terrible early failures seems to be enough – nothing about comparative studies where the device is tested against a successful device.

Regulation of techniques and procedures

Separate to devices is the ‘bar’ that techniques and procedures are required to clear. Here, there is almost no bar if the procedures falls within an existing description. As an example, orthopaedic surgeons recently started doing hip replacements using a new technique: the anterior approach, which involves literally coming at (“approaching”) the hip from the front rather than the back or the side.

There was a bit of a learning curve involved, as there were more complications associated with using this new technique (for example, the thigh bone would break more often when trying to insert the hip). Eventually, things worked out and this new technique is commonly used and probably not any worse than other techniques. For a while however, it looked like a potential disaster, with no regulatory or ethical oversight involved, just individual surgeons trying a new technique.

In fact, the regulatory environment that does exist around new techniques is not only inadequate, it is actually counter-productive. For example, for a surgeon to invent and perform a new technique (perhaps yet another way to perform hip replacements), there is no regulatory oversight, either from the government or from institutional ethics committees: it is not necessary to get ethics committee approval to trial new procedures or techniques.

However, if the surgeon wanted to gather information on the effectiveness of a new technique by performing a comparative study against the old technique or contacting patients to accurately measure the results of surgery, that surgeon would not be permitted to do so without approval from an ethics committee. So it is OK to try new things without measuring the outcome or studying it scientifically, but if you want to find out if it works or is not harmful, you are faced with an onerous process of getting ethics approval to contact the patients and publish the findings. Many would argue the exact opposite, that ethical approval should be required if a surgeon did not measure and report the results of a new technique.

Changing the paradigm

This problem reflects our underlying tendency to assume that something is effective when it has not been tested: we end up with wasteful, ineffective and often harmful procedures and devices being approved and used. At best, we end up with procedures with unknown effectiveness.

There is no good reason to have such a low bar for the introduction of new devices and techniques. The reason we have such a low bar is tradition, which is almost never a good reason.

The reason for this low bar is that many (probably most) of the currently performed surgical procedures listed in the Medical Benefits Schedule (MBS) were “grandfathered” in when the scheme began in the 1980s. Many have never been subjected to rigorous testing against non-surgical alternatives or no treatment at all. The only time rigorous evidence (an appropriate bar) is required for a new surgical procedure or device is if it does not lie within descriptions currently listed in the MBS. In these cases, applications for listing new surgeries are frequently rejected for failing to have sufficient high-quality evidence.

The system needs overhauling, but the inertia inherent in the current systems is such that this may never occur. It may be a massive task to test (already approved) surgical procedures that lack good evidence for effectiveness, but that doesn’t mean we should never start. It wouldn’t be too hard to begin with a few procedures and build up the evidence over time.

For example, instead of funding spine fusions for back pain indefinitely, why not put the money into trials of effectiveness? This could potentially save billions of dollars and patient harms. Any resulting savings could be put into testing the next procedure on the list. This could be done with one simple decision from the government: to make the bar for procedures currently on the MBS (that were grandfathered in in the 1980s) the same as the bar for newprocedures to be listed on the MBS. Otherwise, it must justify having such a paradoxical system of approval.

Ian Harris is Professor of Orthopaedic Surgery at UNSW and the author of Surgery, The Ultimate Placebo

---

This article is part of an ongoing series that is published as a collaboration between Wiser Healthcare and Croakey.org.

The series investigates how to reduce overdiagnosis and overtreatment in Australia and globally. The articles are also available for republication by public interest organisations, upon request.

Bookmark this link and follow #WiserHealthcare on Twitter.

Many GPs who in 2018 received Health Department letters regarding their opioid prescribing criticised the strategy to stem over-prescribing, saying it unfairly targeted doctors who had legitimate reasons to prescribe opioids, such as those working in palliative care.

The appropriateness of these letters is now being looked at by the Australian National Audit Office (ANAO) as part of a wider enquiry into health provider compliance measures.

Jason Soon, a PhD candidate at the Menzies Centre for Health Policy, understands the concerns but says it would be a shame if such poor targeting led to the end of an approach which has shown promise on over- and under-prescribing in a number of settings.

His article is published as part of the TOO MUCH of a Good Thing series, which is investigating how to reduce overdiagnosis and overtreatment in Australia and globally, and is published as a collaboration between Wiser Healthcare and Croakey.

To follow the series, bookmark this link, and follow #WiserHealthcare on Twitter.

---

Jason Soon writes:

Audit and feedback is widely used by healthcare service providers to improve professional practice. It involves measuring providers’ performance on clinical and other indicators and then feeding summaries of this performance information back to them.

While it has applications beyond the healthcare sector, audit and feedback is typically undertaken with the aim of inducing clinical behaviour change among healthcare professionals to ensure that they adhere to clinical guidelines and evidence-based practice.

For instance, healthcare professionals may have their performance bench-marked against some measures of wider sector or peer adherence to a specific guideline or evidence-based practice.

This means that audit and feedback can be used to reduce overdiagnosis and overtreatment (e.g. over-prescription of opioids) as well as under-diagnosis and under-treatment (e.g. encouraging the use of spirometry for diagnosis of respiratory conditions).

The value from audit and feedback is not simply because it reminds healthcare professionals about clinical guidelines with which they should already be familiar. Audit and feedback is also aimed at inviting professionals to reflect upon their performance relative to their peers.

The most recent comprehensive review undertaken of the effectiveness of audit and feedback (based on the results of 140 studies) found, among other things, that audit and feedback is more effective when the professionals being targeted are not performing well to start with (so that there is a lot of room for improvement) and when the source of the feedback is a trusted supervisor or colleague.

The same review concluded that the effectiveness of audit and feedback is highly variable and can range from ‘little or no effect to a substantial effect’, with a majority of studies finding an effect clustering around a 4 per cent increase in compliance with desired practice.

Rating ‘performers’

More recent studies, which have relied on the use of the audit and feedback to enforce ‘peer pressure’ to improve specifically on poor performers rather than the entire clinician group, have produced less variable results.

For instance, one US trial customised its audit and feedback letters, writing to doctors with the lowest inappropriate prescribing rates for antibiotics (for respiratory infections) via monthly email to tell them they were ‘Top Performers’ while telling the remainder that they were ‘Not a Top Performer’.

These emails also compared the share of inappropriate antibiotic prescriptions written by doctors who were not top performers with the share written by those who were.

High prescribing doctors subjected to this audit and feedback achieved statistically significant reductions of around 16 per cent in inappropriate antibiotic prescribing compared to a control group.

A follow up study found that 12 months after the monthly letters were discontinued, the inappropriate prescribing rate had increased by only 1.5% , meaning that durable improvements were achieved by this approach.

Antibiotics vs opioid prescribing

The Australian Department of Health has also recently targeted audit and feedback initiatives at GPs to reduce rates of inappropriate prescribing of antibiotics.

In 2017, the Department sent letters to GPs whose prescribing rates for antibiotics put them in the top 30 per cent for their practice region. These letters were signed off by the Chief Medical Officer and compared the GP’s prescribing rates with the regional average.

The Department found that the peer comparison letter, when combined with an attention-grabbing graphical representation of the peer comparison, achieved the greatest impact, reducing antibiotic prescription rates by 12.3 per cent over a six-month period.

It even outperformed the same letter when combined with educational materials (but without the graphics).

However, a 2018 trial based on the same peer comparison approach but targeting GPs with high rates of opioid prescribing faced a significant shortcoming.

The Department did not identify (and exclude from targeting) those GPs working in specialist pain settings or with oncology and palliative care patients.

Many GPs who were targeted felt that the letter did not sufficiently account for their patient needs, which required higher rates of opioid prescription.

This promoted a backlash  in the medical community which may have reduced the credibility of these comparisons among their target audience.

Similar campaigns based on peer comparison letters are currently underway to target over-prescription of antipsychotic medication and over-use of imaging (paywall). The imaging campaign has already faced similar criticisms  (paywall) as the one for opioid prescribing (e.g. not being able to account for special circumstances where the appropriate level of prescription or imaging is higher than average).

In response to these and related complaints, the Australian National Audit Office (ANAO) is now looking at the appropriateness of these letters though as part of a wider enquiry into Department of Health provider compliance measures.

The ANAO report is due out in May 2020 and its findings will be of interest to health policy analysts. While the apparently poor targeting of more recent peer comparison letters highlights one of the implementation challenges of audit and feedback, it would be a shame if it led to the abandonment of an approach which has also demonstrated promise in a number of different settings.

---

This article is part of an ongoing series that is published as a collaboration between Wiser Healthcare and Croakey.org.

The series investigates how to reduce overdiagnosis and overtreatment in Australia and globally. The articles are also available for republication by public interest organisations, upon request.

Bookmark this link and follow #WiserHealthcare on Twitter.

We hear increasingly about the importance of tackling overtreatment, but less often about how overtreatment plays out in the lives of patients and their families.

In the article below, Associate Professor Magnolia Cardona introduces us to “Walter”, a fictitious character whose story is typical of many older patients, and explains how tackling overtreatment could improve his care.

This article is published as part of the TOO MUCH of a Good Thing series, which is investigating how to reduce overdiagnosis and overtreatment in Australia and globally, and is published as a collaboration between Wiser Healthcare and Croakey.

To follow the series, bookmark this link, and follow #WiserHealthcare on Twitter.

---

Magnolia Cardona writes:

“Walter” is a 92 year-old man with an abnormal heart beat that needs daily anticoagulants. He has had progressive and irreversible chronic lung disease for over a decade from being a heavy smoker in his younger age. He hardly recognises his family now due to dementia, and has lost weight without trying.

In the past year he has been hospitalised several times with flare ups of his respiratory condition and bleeding complications from his medicine. Every time he is discharged, he is more frail and less able to return to his normal daily activities at home.

But he and his family may be unaware that he will not recover fully, because dying is not just a point in time but a journey that can take over a year.

Clinicians may not have communicated this trajectory or prognosis openly to Walter for fear of taking his hope away, or because of the relentless sense of duty clinicians have to save lives and ‘do everything possible’.

Importantly, they may not have asked Walter if he considers all these treatments acceptable.

Drawing the line

When a treatment will not make a difference to the patient’s prognosis and is more likely to cause harm and prolong dying or impair the quality of their end of life, this is called overtreatment.

Examples of non-beneficial overtreatment are:

• attempting cardiopulmonary resuscitation (CPR) on older patients who already have a not-for-resuscitation order;
• starting chemotherapy or radiotherapy or dialysis in the last month of life;
• continuing cholesterol-lowering tablets in patients who will not survive long enough to benefit from the long-term effect of those medications;
• performing invasive procedures in ICU on a person who has expressed their desire to die peacefully at home.

In Walter’s case, the repeat hospital admissions will likely be associated with additional medications, tube feeding, invasive ventilation or transfer to intensive care following aspiration pneumonia.

Why would clinicians administer overtreatments?

In non-cancer conditions like Walter’s, clinicians tend to be cautious in conversations about what to do when the inevitable time comes. This is often due to uncertainty about the patient’s time to death and the variable trajectories illnesses may take.

It is not unusual for older patients with cognitive impairment like Walter to have a fall and associated hip fracture, or to experience further confusion or delirium while in hospital due to immobilisation, sleep deprivation, dehydration, deconditioning, or emotional distress.

Decisions that may need to be made at this point include whether to start further psychoactive medications, parenteral nutrition, whether to proceed with surgery, and if he survives, whether to transfer him to an aged care facility after discharge.

In some cases, these treatments are administered contrary to patient preferences because families are anxious and pressure clinicians, or because neither clinicians nor relatives have discussed personal values with the patient before decisions are made.

Clinicians, in their well-meaning sense of duty, tend to adopt a surrogate decision-maker position using transfer of information to patients and relatives, rather than engaging in consultation with them.

They are trained to save lives and cure disease, so sometimes they overtreat because they are not aware that death is not the only outcome that matters to patients.

In cases of cognitive impairment, this may also happen if people like Walter had not previously expressed to their families their views on what is unacceptable treatment or where they prefer to be managed.

Consequences of overtreatment

In the absence of discussions on less aggressive end-of-life management options, patients and families are unaware they can opt to limit or withdraw from treatments. Hence patients present repeatedly to emergency services or are hospitalised for treatments that are sometimes administered too late to be of benefit or are too aggressive, such as intensive care unit (ICU) procedures.

ICU treatments in frail old people often disadvantage patients, who usually die anyway after experiencing days or weeks of poor quality of life, whether during the hospitalisation or in the months following discharge.

This is known to cause family dissatisfaction, long term regret among relatives who consented to the aggressive treatments, and long lasting grief among surrogate decision-makers.

Overtreatment can burden families and make health system budgets unsustainable.

Possible solutions

Fortunately, there are several options for society and the health system to reduce overtreatment near the end of life:

1. Evidence-based prognostic tools have become available to supplement clinical judgment. They are checklists with clinical risk factors that indicate level of risk of death, so doctors and nurses can more confidently identify near end of life status.

As these tools are relatively inexpensive and some do not take long to be administered, they could be one of the solutions to prevent or reduce overtreatment. Yet, uptake of these objective prognostic tools is limited. Some clinicians are sceptical about prediction using numbers, others prefer to wait until validation reassures them of near perfect accuracy, others retain the practices they have always had, and others have not read the medical literature to familiarise themselves with these checklists.

2. Disclosing prognosis if there is readiness for the truth. Public attitudes on prognosis later in life are changing, with growing desire for more information on life expectancy and interest in involvement in their treatment decisions near the end. This is an opportunity to fulfil ethical obligations of explaining the anticipated future symptoms and risk of death to the extent that patients and families choose to know.

Contrary to commonly held belief, prognostic disclosure is not always distressing nor detrimental to their emotional wellbeing, and for some people it is liberating. We recommend a gradual approach, depending on the level of patient readiness for prognostic detail, respecting their right to not find out.

Empowering patients and families with the knowledge of a forthcoming death over the ensuing months gives them an opportunity to be part of decisions for their future critical illness, express their views on what they consider unacceptable management, have their affairs in order and say their goodbyes.

3. Public education on advance care planning for future critical illness and on reducing demand for non-beneficial treatments can further assist in containing overtreatment. Involving patients and relatives in informed decision-making can facilitate agreement on management that is not potentially harmful to patients.

Relatives with prior knowledge of the patient’s wishes are less likely to experience guilt about their treatment choices or decisions. Ultimately this can translate into curbing of unnecessary use of healthcare near the end of life and higher satisfaction with healthcare.

4. Reallocating resources into community services. Many older patients dying of progressive chronic illness (natural causes) prefer to be at home for as long as possible, not die in hospital, and have a say on their management to keep their dignity and control of the process.

The health system needs to adapt to cater for services in the community by training community nurses and allied health professionals in supportive and palliative care, to turn these preferences into reality.

Patients like Walter

Clinicians caring for patients like Walter could anticipate his high risk of death using objective clinical parameters and discuss prognosis with his family.

Information on future possible complications and preferences to withhold interventions or withdraw non-beneficial medications in case of future critical episodes could be disclosed to patients and families and documented in an advance care directive.

With the increasing life expectancy and forecast doubling of the world’s proportions of 65+ year-olds by 2050, these conversations on end-of-life treatment preferences need to be held routinely in general practice and aged care facilities before a health crisis arises.

Further suffering for Walter could be prevented by a timely, honest conversation about the inevitable declining trajectory and potential harms of further hospitalisation or aggressive treatments when a natural death approaches.

He could be referred to palliative care assessment earlier; or, if sufficient funding for home nursing and community support existed, his last days could be managed at home, or in another non-acute care facility of his choice.

These resources would ensure he receives comfort care and symptom control that ensure he can achieve a good death.

• Magnolia Cardona is Associate Professor of Health Systems Research and Translation at the Institute for Evidence-Based Healthcare, Faculty of Health Sciences and Medicine, Bond University, and conjoint academic at Gold Coast University Hospital.

---

This article is part of an ongoing series that is published as a collaboration between Wiser Healthcare and Croakey.org.

The series investigates how to reduce overdiagnosis and overtreatment in Australia and globally. The articles are also available for republication by public interest organisations, upon request.

Bookmark this link and follow #WiserHealthcare on Twitter.