Controversial Ideas in Medicine

Not a particularly controversial post as the evidence is quite clear.

First: here is a review article from the American Journal of Obstetrics & Gynaecology 2009 going through the various studies on flu vaccine safety in pregnancy.

The conclusion is that influenza is pretty dangerous in pregnancy, while the influenza vaccine is pretty safe. Additionally, no long term side effects from the small dose of thimerosal which is a mercury based preservative.

Thimerosal free vaccines are available.

Second: here is an earlier review article from The Lancet Infectious Diseases 2008

Once again, conclusion is that benefit outweighs risk. As this was prior to the large prospective double blind randomised controlled trial in 2008, there was not enough data at the writing of that article regarding the flu vaccine and safety in pregnancy.

Finally here is the New England Journal of Medicine 2008 prospective double blind RCT that seals the deal with enough evidence- 340 patients.

House of God by Samuel Shem is full of many sorry truths of hospital medicine, of General Medicine in particular. One of those is the art of the "buff" and "turf". No-one wants extra work. It is an eternal rule of human nature (unless you are a workaholic such as myself and find work somehow interesting, exciting or, heaven forbid - fun). And it is true that being in hospital for too long is bad for patients. The "buff" is the polishing up of the patient so that they are as healthy as you can get them from your point of view. The "turf" is the act of sending them to another medical team, to rehab or home or to a nursing home. And you want to do this in such a way that they don't "bounce" - otherwise known as a failed discharge. On the whole this can be an effective system. Certainly it is the kind of system that everyone seems to like - administrators, consultants, registrars, residents and the patients themselves. Less work, less costs, less time in hospital- you can see the advantages right there. Sadly though this leads to the very predictable problem wherein no-one wants patients unless somehow the rules state they can't be discharged or turfed. Usually the "buff" is very incomplete at this stage. Because we are all just focussed on the turf. So we have some 75 year old patient with an uncomplicated heart attack being admitted under general medicine rather than cardiology because she has a urinary tract infection as well. Or no-one has bothered to check liver function tests and someone with ascending cholangitis ends up on general surgery instead of gastroenterology. Or neurosurgery takes a patient who "definitely has an acute disc prolapse" because overnight no-one wants to argue with the emergency registrar who wants to get patients out of a full emergency department and they turn out to have septic arthritis. It is well documented that admitting patients under the appropriate speciality unit leads to significantly improved outcomes - in particular coronary care units and acute stroke units are cited as examples. It leads to shorter hospital stay, lower complication rates and marked improvements in morbidity and mortality. This is relevant to both speciality and general units, I feel. Often general medicine is better for complex or geriatric patients because rehabilitation and multiple referrals are streamlined, while speciality units can be very focussed- and can miss multisystem disorders. Perhaps what will happen in the futures is that we will have speciality multisystem units- those dealing with "metabolic syndrome and smoking diseases", those dealing with disorders of immunity and infections and such-like. Either way, appropriate unit protocols can be a way of reducing fighting over rejecting patients. It certainly simplifies the process of admission. Another thing that must be done is reducing bed pressures and simplifying routine task management for junior doctors. It is high (and unnecessary) workloads and often very unfair bed concerns that mean the "buff" is not complete. And so they bounce.

The nuff-nuff is the bane of our existence. That person who comes in complaining of tummy pain or feeling weak or some vague and inconsistent symptomatology completely without medical explanation and almost certainly psychological in nature. They have some tawdry and dreary sort of social background and probably reside in a slumlike place (oh boo-hoo). They have all sorts of medications for somatic relief. The endones and pethidines and maxolons and valiums. Not to mention the psychiatric pharmacopoeia of zolofts and efexors and zyprexas. Some of them drink and smoke too much (but not enough, sadly). We do all the tests (or just enough to satisfy ourselves) and find nothing while they abuse us for "not caring". So we send them home, kicking and screaming while they threaten "if you send me home I will be back here tomorrow" (which we know is true). And sit back exhausted but relieved that they've finally left. They are the bane of our existence, the nuff-nuffs. The bane. But would you know that:

• Schizophrenia and other psychotic illnesses
• Depression, anxiety
• Personality disorders
• Substance abuse

are all associated with greater rates of illness and mortality? Multiple presentations to hospital are a bad sign that there is something going wrong- whether it be physical or psychiatric. Or both for that matter. Even worse, as they say "you can't prove a negative"; in other words there are many for whom we do not find the real answer (even though it exists). Additionally, chronic diseases lead to significant psychiatric burden- people become depressed, anxious, sometimes even epileptics have psychological "pseudo-seizures". And the mentally ill just do not look after themselves properly. So what do you do when faced with the nuff-nuff? There is I suppose only one thing you can do - exclude organic pathology while treating the psychiatric disorder.

• Approach things with an open mind each time. Ask yourself "why" and be prepared for a complex answer. Sometimes people with worsening disease become anxious and depressed and present to hospital even without a need to. Sometimes it is a cry for help. Sometimes despite mostly being psychiatrically unwell or having a pathological personality there is genuinely a medical cause for the symptoms.
• Involve a psychiatric team early. Sometimes the diagnosis is psychiatric and this in itself is the main thing.
• Keep a high index of suspicion for both organic and psychiatric disease- nuff-nuffs get sick too (and sometimes die)
• Have a sympathetic but professional attitude; be caring but consistent and do not get too wound up in the transferred anxiety.
• Do not get angry, violent or abusive. Additionally do not take on their stresses as your own.
• Be as prepared for gratitude as for the potential to be verbally abused for your approach (and don't take it too personally)
• Exclude dangerous things
• Do what your duty of care towards your patient behoves you to do; do not discriminate on the basis of psychiatric illness.

Doctors are angered by these patients for more than one reason. We are educated to believe that only organic disease is "real" or "significant". But even more than that, being (often unconsciously) manipulated and having large amounts of angst and anxiety offloaded onto us - and then finding that the underlying cause is not what it seems- makes people feel betrayed or lied to (even if that is not the conscious aim of the patient). But why should we feel betrayed? A psychiatric diagnosis (even that of a somatoform or personality disorder) is still a medical - and pathological- condition that causes harm. Often the best thing we can do for these people is to acknowledge their underlying issues and refer appropriately. Sometimes that is all they have been hoping for.

Every speciality has its bugbear. Emergency departments have D-dimers (always positive if you spend more than 2 mins in a hospital, seemingly). Gastroenterologists have "melaena" (often just dark brown stool or even frank PR bleeding). Rheumatologists, chronic back pain (the worst thing you can do is admit them to hospital). And cardiologists have troponins. Good ol' troponin. Touted as "the" test for myocardial infarction (heart attacks to the layman), troponin has gone from being the wunderkind of cardiology to a much maligned villain due to its (sometimes) excessive sensitivity. "Why did they order a troponin?" I hear you cry. "Don't bother with it, it's just troponinitis." What is troponinitis? What is this entity which like "acopia" appears to have entered the everyday parlance of the hospitalist? It is, quite simply, a troponin higher than the 3 times the reference range which for whatever reason is not felt to be due to a myocardial infarct. Troponin is an enzyme that only occurs in cardiac muscle, and thus is only released during death or damage to cardiac myocytes. There is obviously a small baseline amount of damage that is normal wear and tear. The main worry with a very high troponin, then, is a large amount of cell death secondary to a clot or stenosis completely obliterating arterial supply to cardiac muscle - an acute myocardial infarction. Rises in troponin can also occur secondary to sepsis, uncontrolled tachycardia, pulmonary embolus, cardiomyopathy and other causes for cardiac strain. Additionally a troponin can be mildly elevated when it isn't been excreted by the kidneys- in renal impairment. So should we worry about a high troponin in the setting of sepsis, tachycardia or other causes? The key issues here are as follows. Is this:

1. A real myocardial infarction, and if so:
   1. Is this caused by thrombus (clot) post plaque rupture (type I MI)?
   2. Is this caused by cardiac failure/hypotension in the setting of fixed coronary artery disease (type II MI)?
2. Is this another condition causing a raised troponin, and if so:
   1. Does the raised troponin mean there is cardiac damage?
   2. If not, does the raised troponin have any significance?
3. Is there coexisting cardiac as well as other pathology?

Obviously the above must be evaluated in every patient. Additionally any test - including a troponin - should be ordered at an appropriate time when you have a high pre-test probability of an acute coronary syndrome. Thus, underlying risk factors (especially diabetes), clinical features- history and examination, an appropriate acute deterioration, etc. Now, for the controversial bits:

• It is possible to have a watershed infarct in the setting of fixed coronary artery disease where hypotension, tachycardia or cardiac failure is present (eg sepsis or arrhythmia)
• It is also possible to have clinically significant cardiac damage in those settings even without true infarction secondary to strain
• Raised troponin has been shown to be associated with a significantly worse prognosis in acutely unwell patients where another pathology is shown to be the cause

Which leads to my next conclusion:

A raised troponin in an acutely unwell patient, especially one with risk factors and evidence of haemodynamic compromise and/or heart failure should never be ignored.

"But what are you going to do about it?!" you say, incredulously. Watershed infarct without a thrombus? "Troponinitis" due to sepsis? What the heck are you going to do about that? Shouldn't we just stop checking the damned thing? Given that these patients have in fact had the possibility of thrombosis, fixed coronary atherosclerosis, severe cardiac strain/hypoxia or just very bad underlying disease the following should be done:

1. Follow up troponins to monitor trend - MI less likely if troponin persistently raised
2. Aspirin (if not contraindicated) and prn anginine
3. Optimisation of fluid status, strict fluid balance and daily urea, electrolytes and creatinine
4. A transthoracic echocardiogram to assess heart function and see if there is in fact evidence of either infarction or heart failure
5. Cardiologist review in all patients
6. Referral for percutaneous coronary intervention/angiogram:
   1. In patients with ST elevation and likely infarction and who are well enough for the procedure and do not have sepsis (risk of septicaemia and endocarditis)
   2. In patients who will benefit and are well enough for the procedure in whom there is high pre-test probability- as an elective procedure when well.
7. Therapeutic anticoagulation for 48-72 hours in patients with high pre-test probability of infarct with no ST elevation and no contraindication to anticoagulation
8. Addition of cardiac risk factor modifying agents and optimisation of cardiac failure medications
9. Treatment of acute illness, underlying problems and optimisation of chronic conditions
10. Discussion with the patient and family regarding diagnosis and prognosis

Of course, unfortunately the one thing that everyone will groan about is the fact that I have listed 10 things that need to be done for sick patients with "troponinitis". It is much easier to ignore it and do nothing- but that is a grave disservice to our patients and very contradictory to the actual evidence on the matter.