Health memo to COAG part one: it’s time to tackle waste and inefficiency

With health reform high on the agenda of the forthcoming COAG meeting, it is worth reminding ourselves about why we need it.

It is also worth  being reminded that simply spending more money is no guarantee of better health or even better health care. We currently waste much money on treatments of dubious or uncertain value, often at some significant cost to patients.

Read more about some of the areas that COAG should be examining in this  series of discussions between public health expert Professor Stephen Leeder, Canadian health policy consultant Professor Steven Lewis and Melbourne specialist Dr Richard Barnes.

The discussion began with this article, recently published in the Medical Journal of Australia by Professor Steven Lewis and Professor Stephen Leeder, which argued that:

• Traditional health care is fragmented, marred by quality and safety defects, with a failure to provide evidence-based care, and huge and unjustifiable variations in practice.

• There is abundant evidence that traditional means of delivering health care are obsolete.

• Concerns are deepening about persistent and widening gaps in health status that health care cannot overcome.

• Increased spending on health care has never definitively solved the problems of access, quality, or equity.

• Non-medical determinants of health indicate that the solutions to health problems lie mainly outside health care.

• The current financial crisis may create the urgency and courage to both eliminate the fundamental problems in health care delivery and reduce health disparities.

According to Leeder, the article provoked more response than any other he had written.

This is one of the comments that flew in, from a Melbourne anaesthetist, Dr Richard Barnes. He wrote:

“I believe it is one of the most important articles published in the MJA in a long time; the issues you raise deserve wide, thorough and extensive discussion, not just by doctors and health administrators, but by the general public.

I’d like to comment on one aspect of your article.

You suggest that we, via government, should look particularly to cut  areas of absolute waste – the zero or negative return domains rather than risk the middle-class backlash from cuts to marginally effective services. At face value, your recommendation seems unarguable, yet it may be more complex than is immediately apparent.

For many health interventions, the evidence which would enable us to  assess their effectiveness simply doesn’t exist. Often the research  simply hasn’t been done. In many cases, research has shown a particular treatment to be effective, but the therapy is now applied to groups outside the subject group of the original research. For example,  I often wonder how much benefit octogenarians gain from treatment with statins. (And every second 80-year-old seems to be taking them!)

Earlier in the year, Elshaug et al published “Identifying existing health care services that do not provide value for money” (MJA 2009; 190: 269-273). Disappointingly, but not surprisingly, the table which lists interventions clearly shown to be of no benefit, is most striking for its inability to identify any “really big stuff”.

The authors go on to list some much juicier areas for further research, to which could be  added some other obvious topics, such as the whole area of medical therapy versus Percutaneous Coronary Intervention versus coronary artery graft surgery for ischaemic heart disease. The combination of complexity  and vested interests makes it highly unlikely that definitive answers will be available any time soon.

Many treatments, on the other hand, have been or can be shown to be  of marginal benefit. Prolonged, complex and expensive treatments in the very premature  and very old are obvious examples but it is my impression that a  large amount of what we do in acute tertiary hospitals is futile or  of very marginal benefit.

A few years ago, an anaesthetist colleague made a preop visit to an old man who was scheduled to undergo thoracoabdominal repair of a suprarenal “aortic aneurysm”. How do patients undergoing this operation usually get on? asked the old man. When informed that of the past six done at the relevant hospital, four were dead, one was paraplegic and one was well, the patient discharged himself. In one sense, the proposed procedure offered a marginal benefit; one in six patients had a successful outcome. But the old man had made the entirely correct evaluation, that the benefit did not come without a cost – not just in pain and suffering for all ten, but loss of life, or loss of quality  of life, for the other nine.

The average layman is not unaware of these issues – they are among the main themes of Jodie Picoult’s “My Sister’s Keeper”, for example. Tools such as QALYs (quality-adjusted life year, used to measure disease burden) can be used to add some sophistication  to the evaluation of a particular treatment or therapy.

What of an intervention with virtually no risk, but very little  benefit, involving children: MRI scan under general anaesthetic for assessment of developmental delay. I have anaesthetised hundreds of children for this procedure. An audit I performed of a series of such scans confirmed what the limited  published data also say: the majority will have evidence of hypoxic-ischaemic encephalopathy; the minority will have no MRI abnormality;  in some, the scan will afford some prognostic information; rarely,  the scan will suggest a diagnosis with implications for genetic  counselling; rarely, if ever, will the scan lead to an unexpected diagnosis of a condition for which there is a helpful therapy.The general anaesthetic probably has a 1:100,000 risk of causing death, but the scan costs perhaps $2,000 and every tertiary paediatric hospital in the  developed world has an unmanageable waiting list for paediatric MRIs under GA.

I’m sure all of us have our pet  examples of treatments of dubious value.

What I’m saying is that we should not focus solely on eschewing treatments of proven zero value but we should also lead our colleagues, health administrators and the public (in reverse order) to well-informed, dispassionate discussions and decisions about the costs – including the opportunity costs – versus the benefits of the many medical treatments on or  near the flat of the curve.”

***

Steven Lewis replied:

“Your points are indeed well-taken, and particularly the relative paucity of “blockbuster” no-value-added interventions and services.  However, I do think there are a number that get very close to this, especially if we are talking about NET zero or negative value (i.e., some positive benefit but also harm that equals or exceeds the positive benefit, as in the case you cite).

Here the candidates include:

1.  Off-label use of drugs, e.g., the notorious Bextra for which Pfizer has just paid a $2.3 billion fine for misleading marketing.

2.  Substitution of more expensive drugs for cheaper drugs, e.g., ranitidine vs. cimetidine for acid reflux, where the latter – older and far cheaper – is just as effective in well over 90% of cases and should always have remained the drug of first resort before switching.

3.  Mammography for women under 50 (I know this may be somewhat controversial but there appears to be growing consensus that the sceptics were always right); same for universal or population-based (as opposed to targeted) PSA screening

4.  X-rays for suspected broken ankles (if you follow the Ottawa Ankle rules)

5.  A great deal of precautionary MRI and CT scanning (these are of course of great value when used appropriately and there are gray areas but I think it would not be difficult for expert panels to eliminate a third of all utilization with simple algorithms based on experience to date)

6.  Arthroscopic knee surgery (pretty much all of it) and a good deal of back surgery

7.  The use of BEERS drugs in the elderly (BEERS is a list of medications that may be inappropriate for use in the elderly).

Now, it is very likely true that compiling a comprehensive list might not add up to more than 10% of health care spending, although that is no small change given current expenditure levels (that would be $18 billion in Canada, which could transform both child care and post-secondary education).

But there are other forms of outright waste, as in, avoidable hospitalizations and length of stay, achievable not by denying people what they need, but by giving them what they need when they need it (hence the famous comparison of Kaiser Permanente vs. NHS expenditures a few years ago, which found, remarkably, roughly equal per capita costs, with the Kaiser performance attributable entirely to doing a lot more on the prevention and health maintenance end and using 1/3 of the hospital days per capita).

But as you point out, the implicit issue is whether we should declare something to be “useless” or “valueless” unless it can be proven (not merely suggested) that it is unhelpful or harmful almost all of the time.

Well, we do this quite regularly, mainly in the area of drug formulary decision-making, where the main criterion is usually, it’s not worth it (rather than, it’s entirely useless or adds no value in no cases compared to existing products).  If we apply the standard of absolute sensitivity and specificity to our definitions, there won’t be very much to eliminate categorically.

Hence the real discussion – much more nuanced than we were able to make in our short piece, and in any case I suspect Steve and I would have to have many dinners, some randomly assigned to include a lot of good wine and other potions to test for the effect on our wisdom, before we could agree on exactly where we would draw the line conceptually, let alone empirically, on the issues of probability and risk.

The real challenge is to reduce the “number needed to treat” to achieve positive benefits.  If you prescribe Ritalin to enough kids you will generate a positive effect on ADD, but if we systematically overdiagnose ADD we will cause harm and/or be less efficient.  A vastly oversimplified but nonetheless, I would suggest, reasonable description of contemporary health care is that it is fixated on sensitivity and indifferent to specificity.

In any event, your remarks have done what good remarks should – they have stimulated further thinking.  I can imagine a great 2-day symposium on conceptual and empirical issues in eliminating health care waste, ending in a consensus statement on how to go about (yes I am on occasion an optimist).”