Serum Rhubarb Please… Stat!
We must weigh the benefits and risks of every test we perform – our patients’ lives hang in the balance
Chris Tiplady | | Opinion
Many of you will know what the serum rhubarb level is about – a gentle dig at the newly qualified medic ordering the latest tests to prove their diagnosis is correct (or, more often, to rule out something rare). The serum rhubarb may take weeks to come back, be notoriously hard to interpret, and is only requested by this particular doctor.
We see it in the labs as a sudden increase in the number of requests for serum rhubarb without any other explanation or resources. Pathologists then find themselves in the position of having to help interpret borderline serum rhubarb levels, explain negative results in those who clearly have the disease and, most challenging of all, advise on otherwise fit and well patients who have accidentally had their rhubarb levels checked.
We have seen tests and interventions change over the years and I think it is important to reflect on the historical development of pathways and guidance on how we use them. As an example, CT pulmonary angiography was originally buried in diagnostic algorithms for pulmonary embolism and was not a first-line test for pleuritic chest pain like it is now. The more CT scans you do, the more you realize that you are actually as (or even more) likely to find something other than a pulmonary embolism (the so-called “incidentaloma”) when you do them. This has a direct impact on cost, scanner utilization, and radiology time. It increases patients’ exposure to radiation. It also has a more insidious impact on patient anxiety; up to 15 percent of people will require follow up and some will need interventional procedures (1).
We think only surgery, drugs, or radiation can be harmful. I am arguing that simple blood tests can be harmful.
Serum rhubarb testing can lead to diagnostic delay. Rather than trusting a clinical picture, we feel we must wait for tests to prove it. Tests lead to tests – and, like with every single assay we do, 5 percent of the normal population will have a result outside the normal reference range. Tests need follow-up, repeating, explaining. They cause anxiety, they cost, they require appointments, travel, referrals, and time. Our patients’ time is the most precious commodity they have. We should not be wasting it.
It is for those very reasons of potential harm that very clear guidance must exist about serum rhubarb level measurement. Once laboratory tests are used outside this clear guidance, it is like they have escaped into the wild. But why are these tests escaping?
We face pressure to turn patients around quickly to get them out of the emergency department. This leads to tests being done to rule out conditions out when their original purpose was to rule diagnoses in. The drive is medicolegal, inexperience, influence, a bad experience, the latest case study, or the anxiety of both patient and clinician. We have to do everything now; we can’t wait and see. We no longer consider pre-test probability.
“It might be rhubarbitis; we have to check for it.”
If resources were more tightly managed, would we allow such creep? Should we make the fences higher to keep the tests in our fold? Have we become complacent with the true costs of testing? Have we forgotten that diagnostic pathology comes at a huge healthcare cost to our systems?
Professor Sir Muir Gray illustrated this issue by referencing Donebedian’s models of healthcare resourcing (2,3). The Donebedian model looks at benefit, risk, and cost, exploring the relationship between investment in healthcare resources and the value obtained. “Value,” in this case, is not just financial, but global. Donebedian showed what many call the law of diminishing returns. As we invest more and more into healthcare at a population level, benefits increase rapidly, especially in the initial stages when it is very clear who will benefit from intervention.
An example you may relate to – have you seen the passionate cyclists on the road on a Saturday morning? They all want to go faster. They buy beautiful bikes, carbon wheels, nice components… and they do indeed speed up. But they reach a point at which, however much money they spend on fancy gear, they will not see the same return on their investment as in the initial stages. But the harm they cause to their bank accounts and relationships will increase with every purchase!
As we invest more and perform more interventions on more people, the total “harm” will increase –anxiety, retesting, follow-up, and so on. There is a point at which increased investment reduces “value” – we move past the point of optimality. Testing has escaped into the wild and there is a reduction in net benefit.
For an individual, the perspective on this “value” is important. Everything we do to a patient, whether a test or a procedure, comes with that risk of harm. When we have constraint and guidance, we do those things to people who will clearly benefit. As resources are more generous and less restrained, interventions are offered to those less and less likely to benefit, but equally likely to come to harm. A different point is reached at which the serum rhubarb test is “futile” clinically and economically.
Think of it like the risk you take running across a busy road. If you are running away from a mountain lion, you might briefly look left, then, right and sprint – happy to take the risk for the substantial benefit of the safety on the other side. But, if you are running from a mouse, maybe having spent a few hundred dollars on good running shoes, taking that same risk is foolish.
So what do we do about this and the serial serum rhubarb tester?
Doctors need to consider their approach. We are not just “doing some tests.” Every single thing we do to a patient is an intervention, be that taking a history, drawing blood, or performing surgery. When patients fully understand what is going on, the evidence is that they will choose fewer interventions, choose things different to what their doctor predicts they would choose, and be much happier with the outcome (4).
We need to make sure patients are fully informed of their choices at every single point in their healthcare. This is shared decision-making. This is consent. You do not always know what is best; you can offer assistance, but the patient ultimately needs to make the choice. I recommend reading and using NHS Scotland’s resources on “realistic medicine” (5).
We have reached a point where we are literally putting too much “work” in the workup. Don’t assume anything. Take time to explain everything. You must describe the benefits – and the risks – of everything you do, even that serum rhubarb test.
- O Anjum et al., “Computed tomography for suspected pulmonary embolism results in a large number of non-significant incidental findings and follow-up investigations,” Emerg Radiol, 26, 29 (2019). PMID: 30238172.
- M Gray M et al., “Personalised healthcare and population healthcare,” J R Soc Med, 111, 51 (2018). PMID: 28920755.
- A Donabedian, Explorations in Quality Assessment and Monitoring. Health Administration Press: 1980.
- A Mulley et al., “Patients’ preferences matter: Stop the silent misdiagnosis” (2012). Available at: https://bit.ly/3qUZAi2.
- NHS Scotland, “Realistic Medicine” (2022). Available at: https://bit.ly/3sev8ip.