How Can We Help?
Iatrogenic illness.
Scope.
The size of this problem in South Africa is poorly quantified. In the US and Canada about 10% of hospitalisations are related to therapeutic and diagnostic misadventures. This means that in a unit admitting between ten and thirty patients per day, one to three admissions each day are iatrogenic.
Aetiology.
There is often more than one cause. Look for the underlying systems disorder that needs correction – mistakes are usually multifactorial and ascribing ‘blame’ to a single individual or event seldom reduces the probability of future events.
Prophylaxis – minimisation of harm
A patient’s similarity to the randomised controlled trial (RCT) population may be distant. Even efficacy data may not be transferable – a blind elderly person with no phone living alone and with a stable borderline blood glucose will not accrue most of the benefits of moderately tight glucose control described in the UKPDS data, since much of this benefit was a reduction in ‘episodes of photocoagulation’[1]. Her risk of serious hypoglycaemia is higher than that of an affluent city dweller with good family support. Giving more medicines increases the risk of a significant side effect or interaction. The more diagnostic tests you do, the higher the chance of a test complication.
Will this test really help this patient (not just satisfy the doctor?)?
Is it reasonable to assume that the efficacy of this therapy will be the same (or higher) in this patient than in trial patients?
Are the expected side effects a justifiable risk in terms of the expected benefits?
Does the patient understand and agree with these value judgements?
Real world probabilities in patient decision making
Both doctors and patients struggle to understand numerical probabilities. Most medical information is presented statistically, and few except diehard mathematicians make day-to-day decisions in our own lives using formal number theory.
Medical procedures themselves are also associated with risk:
The annual risk in the UK of radiation induced cancer has been estimated at about 1 in 5000 from coronary or cerebral angiography, 1 in 15 000 from a CT, half this again (1 in 30 000) from an abdominal X-ray, and one quarter of this (1 in 125 000) from a mammogram. A plain chest X-ray confers a 1 in a million risk[2]. Another way of looking at this is to remember that a modern abdominal CT exposes a patient to the same amount of radiation as 500 chest X-rays, and has a 1 in 2000 risk of inducing cancer[3].
A similar exercise has been undertaken looking at some medications[4], with estimated fatality rates per 100 000 person years of 76 for rofecoxib taken for arthritis, 35 for clozapine without FBC monitoring, and 2.8 for first-generation anti-histamines taken for four months each year (non-MVA injuries). This was compared with 0.15 for commercial air travel, 1.3 for talking on your cellphone while driving, 45 for being a US truck driver, and 450 for being a motorcyclist.
Perspective: information overload
There is a very real danger of garbling one’s message by providing too much information. Trying to say everything is difficult to assimilate and arguably impracticable in time-constrained environments, and multi-page information leaflets are probably no better. However, as soon as one decides to limit information, one has to decide what to leave out, and one needs to recognise that the leaving out process exposes one’s own biases. For instance, any one of the following figures could be used to provide an estimate of the risk of radiographic contrast material reactions[5], (which can be reduced by corticosteroids, but with a NNT of 200[6]):
Overall frequency (all reported reactions) 5% (1 in 20)
Moderate reactions (e.g. vomiting, urticaria, angioedema) 1% (1 in 100)
Life-threatening reactions <0.1% (< 1 in 1000)
Informed consent
Informed consent is the provision of understandable information on both benefits and risk of diagnostic and therapeutic options in order to allow patients to make rational choices about their health care. It is a map plus a friendly explanation.
Components of the informed consent process.
Provision of information – dialogue or document
Understandable – tailored to the individual’s world view and educational abilities.
Quantify risks and benefits in meaningful terms – being too ‘mathematical’ is often confusing, and analogies with real life probabilities can be helpful.
Implications of the various options should be spelt out in sufficient detail to allow a reasoned decision.
Facilitation of decision-making. Patients often ask for help or appear to opt out of decision-making. If you then say what you think is a reasonable choice, still give the patient space to re-think the suggestion.
Often a documentary trail to the process is worthwhile and the medicolegal implications of not providing this should be weighed against the time involved for you and the ‘finality’/scariness of signing a paper for the patient.
Consent quality: ‘Cosmetic’ consent is a reality in many hospitals – there is a signed consent form in the folder, but it is evident that the patient didn’t understand what he or she was signing. There is little medicolegal value in a piece of paper that the patient admits to signing but didn’t understand.
Frequently voiced reservations about informed consent:
Doctors are often sceptical of the value of the process. Apart from its obvious human rights value, it has immense value in getting patient participation and ‘buy in’ for what may turn into a lengthy and arduous therapeutic course
Can too much information impair decision-making ability? Both patients and doctors may simplify decision-making by ignoring what is judged unnecessary detail. In the archetypal paternalistic model the doctor abrogates to him or herself all the details of this simplification process. The art of effective counselling is finding the best compromise for an individual patient.
Perspective – can we do better?
In a study on HIV transmission in Haiti[7], a trained counsellor had three consecutive meetings with a group of prospective study participants, who then wrote an examination on the content of the consent form, and 80% ‘passed’. A clinician giving the same information to a group in a single session led to a pass rate of only 20%. There is clearly room for improvement, and the interventions were not that difficult – repeating the message using trained staff.
Risks and benefits in perspective
The table shown below attempts to put into perspective some or the risks to which populations (the general South African population) and patients are exposed.
Perspective – Contrast-induced nephropathy (CIN)
Renal failure due to the contrast given for angiography or CT appears iatrogenics; there are multiple risk-scoring schemes[8], but few have been evaluated prospectively. A consensus statement[9] advocates focusing on calculated GFR and diabetes, although hypotension, heart failure, anaemia, old age and volume of contrast are considerations. The table combines information in the statement document, with that in a review article on cardiac angiography[10].
CIN may be defined in terms of absolute rises in creatinine or an increase > 25% above baseline. The probability of CIN also predicts the probability of needing dialysis: when it is 7.5%, less than 0.04% of patients need dialysis, at 15% the risk is 0.1% (but more than 0.7% if after PCI), at 25% it is 1% (4% after PCI) and when the risk of CIN is over 50%, the probability of needing dialysis rises to over 10%.
There are numerous underpowered and poorly executed trials, touting any number of medications for CIN, ranging from theophylline to N-acetylcysteine. No agent is so clearly beneficial that it is worth advocating[11], but in general it seems worthwhile to ensure adequate hydration with intravenous saline, and avoid concurrent nephrotoxins. (A meta-analysis[12] showed that acetylcysteine treated patients had a relative risk of CIN of 0.62 (95% CI 0.44 to 0.88, ARR of 22.8%).
-
McCormack J Greenhalgh T. Seeing what you want to see in randomised controlled trials: versions and perversions of UKPDS data. BMJ. 2000;320:1720-3. ↑
-
De Gonzalez A, Darby S. Risk of cancer from diagnostic X-rays: estimates for the UK and 14 other countries. Lancet. 2004;363:345-51. ↑
-
Kmietowicz Z. Better safe than sorry. BMJ. 2007:335:1182-4. ↑
-
Cohen JT, Neumann PJ. What’s more dangerous, your aspirin or your car? Thinking rationally about drug risks (and benefits.) Health Affairs. 2007;26:636-46. ↑
-
Summary statement JCAAI. The diagnosis and management of anaphylaxis-anaphylactoid reactions to radiographic contrast material. J Allergy Clin Immunol. 1998;101:S465-528. ↑
-
Tramer MR, von Elm E, Loubeyre P, et al. Pharmacological prevention of serious anaphylactic reactions to iodinated contrast media: systematic review. BMJ. 2006;333:675. ↑
-
Fitzgerald DW, Marotte C, Verdier RI, et al. Comprehension during informed consent in a less-developed country. Lancet. 2002;360:1301-2. ↑
-
Mehran R, Aymong ED, Nikolsky E, et al. A simple risk score for prediction of contrast-induced nephropathy after percutaneous coronary intervention: development and initial validation. J Am Coll Cardiol. 2004;44:1393–9. ↑
-
McCullough PA, Adam A, Becker CR, et al. Risk prediction of contrast-induced nephropathy. Am J Cardiol 2006;98:27K-36K. ↑
-
McCullough PA, Choi JP, Feghali GA et al. Contrast-induced acute kidney injury. J Am Coll Cardiol. 2016;68. http://dx.doi.org/10.1016/j.jacc.2016.05.099 ↑
-
Barrett BJ, Parfey PS. Preventing nephropathy induced by contrast medium. N Engl J Med. 2006;354:379-86. ↑
-
Kelly Am, Dwamena B, Cronin P, et al. Meta-analysis: effectiveness of drugs for preventing contrast-induced nephropathy. Ann Intern Med. 2008;148:284-94. ↑