Tuesday, November 10, 2015


Today in morning report we discussed the case of a patient who presented with hypothermia. We don't come across this every day in GIM so it is good to review the causes. Hypothermia is defined as a temperature less than 35 degrees Celsius. It is important to consider getting a second reading from a different body site to confirm that this is a true reading.

Environmental exposure is a common cause of hypothermia which can usually be elicited from the history. However, sometimes there is no history of cold exposure. In these cases you should think about the other causes of hypothermia. You can divide these into broad categories [that relate to the reasons your house might be too cold]:

Increased heat loss [the doors/windows are open]
  • Drugs that cause vasodilatation (drugs, alcohol)
  • Iatrogenic causes (cold infusions, CRRT, bypass)
  • Burns or severe skin conditions (psoriasis)

Reduced heat production [the furnace is not working properly]
  • Endocrine causes (hypothyroidism, adrenal insufficiency, hypopituitarism)
  • Hypoglycemia
  • Malnutrition
  • Reduced muscle activity (extreme elderly, inactivity)
Impaired regulation [the thermostat is incorrectly set]
  • Central CNS pathology (Stroke, intracranial hemorrhage, hypothalamic dysfunction, parkinsonism, MS, CNS drugs
  • Peripheral CNS pathology (spinal cord transection)

Miscellaneous causes [your house has an infection? ok, this one doesn’t apply to the house analogy]
  • Sepsis
  • Pancreatitis

Check out the following article for a nice review of hypothermia: http://www.aafp.org/afp/2004/1215/p2325.html  

Sunday, November 1, 2015

Evaluating a patient with headache

We discussed an interesting case of a young man presenting with a headache. This is such a common complaint that it is worth having a solid organized approach. Here are a few key points we discussed in morning report.

Think of the common and serious causes when evaluating a patient with a headache

Headache is a common, non-specific complaint that can be benign or life-threatening. For this reason it is important to have a good approach to headache. One approach is to come up with a differential for the causes that you don’t want to miss and also a list of the most common causes. You can use these lists to direct your history, physical exam and investigations.

Always consider epidemiology when creating your differential. For example, patients who are immunocompromised or those with previous intracranial pathology/surgery may invoke other diagnoses. Here is the differential diagnosis that we came up with at morning report (and one which I generally start with). Keep in mind that there are many other causes of headache that you may have to consider based on the specific clinical situation.

Common causes: Migraine, Tension-type, Cluster, Medication withdrawal, Headache associated with volume depletion or other systemic illness.

Serious causes: Meningitis, Subarachnoid hemorrhage, Temporal arteritis, CNS lesion, Sinus venous thrombosis, artery dissection, hypertensive emergency.

Use a history and neurological exam to rule in (or out) ‘serious’ causes of headache and guide your subsequent workup

The ‘POUNDing’ criteria was described in JAMA as a method of deciding whether a patient with a headache has a migraine or should undergo neuroimaging (JAMA Article Link). The 5 criteria are: Pulsatile, lasts 4-72 hOurs, Unilateral, Nausea/Vomiting and disabling. If a patient has 4/5 of the POUNDing criteria, the positive likelihood ratio for this being a migraine is 24.

It is also important to ask about red flags associated with the ‘serious causes’ listed above. Your history and physical exam will then direct your investigations. For example, you may order neuroimaging (ie. suspecting subarachnoid hemorrhage or CNS lesion), get a lumbar puncture and give antibiotics (ie. suspecting bacterial meningitis), start steroids (ie. suspecting GCA), and so on.

Bacterial meningitis should always be on your differential for headache

Check out this phenomenal one page CMAJ article on bacterial meningitis that gives a really great overview of 5 things you should know about bacterial meningitis (CMAJ Link). If there is one teaching point that I can highlight it is the following: Do not delay antibiotics (and steroids) in patients who you suspect have bacterial meningitis. This means that if you need neuroimaging or think that the LP will be delayed – give the antibiotics right away. Antibiotics might lower your chances of obtaining a positive diagnostic CSF culture but early antibiotics will improve the outcome for your patient.  Plus, you can still use other CSF indicators, like the WBC count/differential, to help support your diagnosis of bacterial meningitis.

If you are unsure about whether the patient has aseptic meningitis or bacterial meningitis you should consider treating empirically with antibiotics

The patient we discussed in morning report had aseptic meningitis. There is a long list of viral causes of aseptic meningitis.  The major families of viruses to think about include herpesviruses (ie. HSV-II), arboviruses (ie. West Nile virus), enteroviruses (ie. coxsackie) and don’t forget about HIV. Other causes to consider in the appropriate patient includes tuberculous meningitis or fungal meningitis. Choose specific CSF testing based on the clinical picture. 

If the diagnosis is not clear and there is still a suspicion for bacterial meningitis, it is often safest to treat empirically with antibiotics and await culture results.  If you suspect HSV as the cause of aseptic meningitis (ie. oral or genital lesions present) consider treating empirically with acyclovir. 

Detsky, Michael E., et al. "Does this patient with headache have a migraine or need neuroimaging?." Jama 296.10 (2006): 1274-1283.
Moayedi, Yasbanoo, and Wayne L. Gold. "Acute bacterial meningitis in adults." Canadian Medical Association Journal 184.9 (2012): 1060-1060.

Wednesday, August 26, 2015

Elevated Troponin and Rheumatic Heart Disease

Today we discussed a case of a woman with chest pain, new atrial fibrillation and an elevated troponin. She likely had rheumatic fever as a child and an echo revealed severe MR with a dilated left atrium. There were some great learning points that came up! Lets discuss.

Troponin is sensitive, not specific! 
When you assess a patient with an elevated troponin do not immediately jump to ACS as the cause! While this will be the most common cause (and certainly one of the more concerning) there are many other etiologies. Always be rigorous in your history and physical exam to decide whether or not ACS is the most likely diagnosis and to rule out others. Some other causes of elevated troponin include:

  • Myocardial Infarction
  • Post-instrumentation (PCI, cardiac surgery)
  • PE
  • End stage renal disease
  • Pericarditis/Myocarditis
  • Aortic Dissection
  • Heart Failure
  • Sepsis
  • Trauma
  • Stroke or intracerebral hemorrhage
Check out this great article on the differential diagnosis of elevated troponin: http://www.ncbi.nlm.nih.gov/pmc/articles/PMC1860726/

Rheumatic fever is less common in developed countries but we still need to know about it.
Acute rheumatic fever occurs a few weeks after group A strep pharyngitis. However, not everyone who gets group A strep pharyngitis will get rheumatic fever! In fact, in developed countries, only about 3% of patients with untreated group A strep pharyngitis develop rheumatic fever. This number is lower than in developing countries - most likely due to improved hygiene (reduced transmission) and higher rates of antibiotic use for 'strep throat'.

Rheumatic fever is an illness characterized by classic signs and symptoms in a patient who had recent group A strep pharyngitis. 
We are all familiar with the 'Duke Criteria' for endocarditis. Well, rheumatic fever has the Jones Criteria to help you diagnose it:
There is a high probability that your patient has acute rheumatic fever if they had a recent group A strep infection and fulfill 2 major criteria or 1 major + 2 minor critera.

Acute rheumatic fever can lead to rheumatic heart disease.
Those who develop acute rheumatic fever are at high risk of having cardiac involvement and developing cardiac complications. Lasting complications of rheumatic heart disease are due to heart valve destruction and include CHF, stroke, endocarditis and death. How does pharyngitis cause heart disease you ask? Well it's not perfectly understood but you can explain it using buzzwords like 'molecular mimicry' (antibodies against strep will attack native cells/tissue, like myocytes). You'll have to check out another source if you want to learn more about that...

So the key is to prevent acute rheumatic fever by promptly diagnosing and treating group A strep pharyngitis in the first place!

Mitral regurgitation is the most common valve pathology of rheumatic heart disease. These patients will frequently develop chronic mitral regurgitation. According to the ACC/AHA guidelines (which I have summarized here), you should consider surgery for severe mitral regurgitation when any of the following indications are met:
  • Acute, symptomatic severe MR
  • Chronic MR which is symptomatic 
  • Chronic MR with LV dysfunction
  • Chronic MR for patients undergoing other cardiac surgery 
  • Atrial fibrillation
  • Pulmonary hypertension
  • Mitral valve prolapse and ventricular arrhythmias
Thanks Team 1 for a great case!

Check out the following resources for more info:
  1. Seckeler MD, Hoke TR. The worldwide epidemiology of acute rheumatic fever and rheumatic heart disease. Clinical Epidemiology. 2011;3:67-84. doi:10.2147/CLEP.S12977.
  2. Korff S, Katus HA, Giannitsis E. Differential diagnosis of elevated troponins.Heart. 2006;92(7):987-993. doi:10.1136/hrt.2005.071282.
  3. Bonow RO, Carabello B, de Leon AC, Jr., et al. ACC/AHA guidelines for the management of patients with valvular heart disease: A report of the American College of Cardiology/American Heart Association Task Force on practice guidelines (Committee on management of patients with valvular heart disease)12. J Am Coll Cardiol. 1998;32(5):1486-1582. doi:10.1016/S0735-1097(98)00454-9.

Tuesday, August 11, 2015


Today we discussed a very interesting case of pancytopenia. This is a great internal medicine topic that is definitely worth reviewing. The causes of pancytopenia are extensive so it can be overwhelming to come up with a comprehensive differential diagnosis. That's why it's great to have an organized approach and stick to it every time!

One approach that we discussed is to divide the causes of pancytopenia into those associated with bone marrow which is either hypocellular or hypercellular:

Hypocellular Marrow:
  • Drugs/Toxins: Methotrexate, chemotherapeutics, various antibiotics, allopurinol...
  • Infection: Parvovirus B19, EBV, HIV, TB, Hepatitis, Sepsis 
  • Autoimmune: Lupus, HLH
  • Paroxysmal Nocturnal Hemoglobinuria (PNH)
Hypercellular Marrow
  • Malignancy: Leukemia, Lymphoma, MDS
  • Nutritional: Deficiencies in B12, folate, copper
  • Marrow Replacement: Myelofibrosis, Granulomatous disease (TB, sarcoid, fungal), solid tumor metastatic disease
A few words about paroxysmal nocturnal hemoglobinuria (PNH):
This is a rare acquired stem cell disorder which can present as unexplained hemolytic anemia, thrombosis, hemoglobinuria and non-specific symptoms. It can certainly present as pancytopenia along with the associated complications. PNH should be considered in Coombs negative hemolytic anemia or other unexplained cases of anemia. The diagnosis is made with flow cytometry using a technique that determines whether certain cell wall proteins (GPI) are present or not - they are absent in PNH. Definitive treatment of PNH is a stem cell transplant.

Another few words about Myelophthisic Anemia:
What a great word! Whoever decided to put a `ph` next to a `th` must have enjoyed watching people struggle with pronouncing things... This entity was brought  up today in morning report. Myelophthisis refers to bone marrow being displaced by some other process. This could include tumor, infection (TB), granulomatous conditions (sarcoid) or fibrosis. There are often clues on the peripheral blood film that may suggest a myelophthisic process. These include teardrop cells and nucleated red blood cells.

Check out this article for a general overview of the causes of pancytopenia: http://www.ncbi.nlm.nih.gov/pubmed/23270895 

Thursday, August 6, 2015


Hyperkalemia can be fatal! Now that I have your attention, time for some learning…

We recently discussed a case of hyperkalemia in morning report. This is a ‘bread and butter’ internal medicine topic for which trainees have been taught the same approach for eons: pre-renal, renal and post-renal. Instead of focusing on the approach to the many causes of hyperkalemia I will review the acute management.

Stop potassium from coming in
  • Low potassium diet
  • Stop potassium containing IV fluids - sounds obvious but makes sure to check!
  • Also, remember to stop nephrotoxic medications - think about NSAIDs, ACE-inhibitors, etc...
 Stabilize the cardiac membrane
  • Calcium gluconate (or chloride) will stabilize the cardiac membrane and (hopefully) prevent a dangerous arrhythmia. But the effect lasts under an hour so give calcium first to buy yourself time to initiate the definitive measures below.
Shift potassium into cells
  • Insulin 10-20 units IV to shift potassium into the cells.
  • Dextrose D50W 1-2 amps to avoid hypoglycemia from your insulin. The sugar load will also stimulate release of endogenous insulin, thereby further shifting potassium into cells.
  • Beta agonists are also effective at transiently lowering serum potassium.
  • Sodium Bicarbonate infusion has been shown to cause a small reduction in serum potassium. However, you may not always need to give it if your other measures are successful.
Remove potassium from the body
  • Renal: Is the patient producing urine? It is important to address any cause of AKI. Healthy kidneys with a reasonable GFR should eliminate any excess potassium. If the patient is volume overloaded and the clinical situation permits, you can give a loop diuretic to promote renal excretion of potassium. If the patient is pre-renal, perhaps he just needs some IV fluids to start excreting potassium. 
  • GI: Resins are often used to treat hyperkalemia (ie. Kayexalate aka sodium polystyrine sulfate) but they are frequently not effective after a single dose and their use has been linked to intestinal necrosis. So if you're going to use a resin, make sure the patient is not post-operative and has no GI problems (ileus/obstruction). Another school of thought is to avoid kayexalate and instead use lactulose to induce a diarrhea which will result in potassium loss. Regardless of your approach, the bottom line is that after you do all (or some) of the above, make sure to follow the serum potassium closely (recheck in 1-2 hours), make sure that one of the methods of potassium excretion is working and address the underlying cause.
  • Hemodialysis: If the hyperkalemia is refractory to your management (above) you should consult your friendly nephrologist for consideration of dialysis.

Check out the following review article from CMAJ on Manegement of Acute Hyperkalemia: http://www.ecmaj.ca/cgi/content/citation/182/15/1631

Thursday, July 16, 2015

Managing a patient with suspected ACS

We had a great discussion this morning about a patient presenting with classic concerning cardiac chest pain who had minimal cardiac risk factors, negative biomarkers and no significant ECG changes. One of the points we discussed was: Should this patient be treated for ACS prior to risk stratification? The answer is the same as for most of our decisions in medicine – it is a balance of risks vs benefits.

I will present a general approach to ACS. But remember – this is just a guide and all patients with chest pain are unique and can have subtle differences that can completely change management!

First, decide if the patient has an Acute Coronary Syndrome (ACS).
 Remember, ACS represents ‘acute’ plaque rupture and can include unstable angina, NSTEMI and STEMI. On your initial assessment, there are 4 things to consider when making this decision.
1)      History
2)      Physical Exam
3)      ECG
4)      Biomarkers
Use these tools to determine whether or not your patient has a high likelihood of ACS. If there is objective evidence of ACS (positive troponin, ECG changes) or if the likelihood of ACS is high then you will treat. If you do not think that there is an ACS or if the likelihood of ACS is low, look for another cause of chest pain. If you can’t find one, consider risk stratification as an outpatient (assuming there are no other concerning features).

And don't forget to include other diagnoses on your differential that may mimic ACS (ie. PE, pericardititis, etc...) and those conditions that are treated entirely differently (ie. aortic dissection).

If you decide that your patient has an NSTEMI or unstable angina, you can use various risk calculators to determine prognosis and decide on a strategy (conservative or early invasive).
 The TIMI risk score is a commonly used scoring system (http://www.timi.org/index.php?page=calculators) that predicts 14 day risk of mortality, new/recurrent MI or severe ischemia requiring urgent revascularization. The TIMI risk score calculator is based on 7 factors: age>65, 3 or more CAD risk factors, known CAD, ASA use in the past 7 days, 2 or more episodes of angina in 24 hours, ST changes >0.5mm, positive cardiac biomarker.

Key Point
: Remember, only use this risk calculator once you have decided your patient has or likely has ACS. It is not appropriate to use this scoring system to determine if your patient has ACS.

In general, use a conservative strategy for low risk (medication and non-invasive risk stratification) and use an early invasive approach for high risk (medication and angiogram within 24-48hrs). Patients diagnosed with ACS should receive ASA and anticoagulation along with the other cocktail of cardiac medications which we won’t go into in this blog.

So, how do you manage a patient with classic chest pain but no ECG changes and normal troponin?
 This patient could be labelled as ‘possible’ ACS. This is a grey zone that may lead to some confusion and maybe even variation in practice among internists.
  • If the ECG is non-ischemic and serial troponins negative and the patient remains free of chest pain BUT there is still a suspicion of ACS, a stress test is generally required to rule out unstable angina. This can be done as an inpatient (if intermediate/high risk features) or as an outpatient within 72 hours (if low risk features and no other concerns). Regarding acute medical management, the AHA guidelines suggest treating (ASA and anticoagulation) patients with definite or likely ACS.
  • According to AHA guidelines (2014), for patients with possible ACS but no objective evidence of myocardial ischemia (non-ischemic ECG and normal troponin):
    • It is reasonable to observe them in a telemetry unit with serial ECGs and troponin at 3-6 hour intervals.
    • It is reasonable to perform non-invasive risk stratification before discharge or within 72-hours of discharge.
    • In low-risk patients who can be discharged with outpatient risk stratification it is reasonable to give them daily ASA, short acting nitro and provide instructions about activity level and follow-up.

So the bottom line is: In general, patients with ‘possible’ ACS without objective ischemia (negative trops, normal ECG) do not require full anticoagulation prior to risk stratification. However, there are exceptions to every rule and these patients should be carefully assessed. The guidelines say that ‘likely’ or definite ACS should be treated. So how do we determine what to do for patients who fall in the grey zone between ‘possible’ ACS and ‘likely’ ACS? This is a case by case decision where you need to weigh the risks and benefits of therapy. This is where clinical experience and judgement are paramount.

Amsterdam, Ezra A., et al. "2014 AHA/ACC guideline for the management of patients with non–ST-elevation acute coronary syndromes: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines." Journal of the American College of Cardiology 64.24 (2014): e139-e228.

Tuesday, July 7, 2015

Liver Cirrhosis

Thank you Team 3 for presenting a case of a patient with cryptogenic liver cirrhosis.  Cirrhosis is a very common diagnosis and we frequently care for patients with this disease. However, it is less common that we see patients with cryptogenic cirrhosis. This combination of a common condition with a less common etiology makes the case a perfect one for Morning Report!

A healthy liver is pink, smooth and filled with happy hepatocytes. The word ‘cirrhosis’ comes from the Greek word kirrhos which means ‘yellowish’ – describing the colour of the diseased organ. In addition to becoming discoloured, cirrhotic livers are firm, nodular and composed of fibrous tissue. This description of abnormal structure gives you a sense of how impaired the liver function can become! Here are six things to know about cirrhosis…
There is a long list of causes of cirrhosis but there is a short list of common causes:
     1. Alcohol
     2. Viral Hepatitis (C & B)
     3. Non-Alcoholic Fatty Liver Disease (NAFLD)
     4. Hemochromatosis
The other less common causes to think about include: Autoimmune, Primary Biliary Cirrhosis, Primary Sclerosing Cholangitis, Veno-occlusive disease, Wilson’s Disease, alpha-1 antitrypsin, medications (ie. methotrexate), right-sided heart failure and even celiac disease!  

Most patients remain asymptomatic until they develop decompensated cirrhosis which is characterized by the development of any of the following complications:
     - Variceal Bleed
     - Ascites
     - Encephalopathy
     - Spontaneous Bacterial Peritonitis
     - Jaundice
In addition to the acute complications of cirrhosis listed above, think about some of the other associated complications:
     - Hepatocellular carcinoma
     - Hepatorenal Syndrome
     - Osteoporosis

Bloodwork can provide clues that a patient has cirrhosis. Imaging and in some cases biopsy are used to confirm the diagnosis. According to the JAMA rational clinical exam, a platelet count of less than 160 increases the likelihood that the patient has cirrhosis. Remember, this is a sensitive finding – not specific! Other abnormal lab values in patients with cirrhosis include elevated bilirubin, elevated INR, reduced albumin and transaminitis. Abdominal ultrasound is the initial imaging modality of choice to assess for cirrhosis. Finally, a liver biopsy may be necessary to help determine an etiology.

Unfortunately, cirrhosis is irreversible so the management should focus on limiting or removing the precipitant and either preventing or dealing with the complications. This will commonly include managing ascites (diuretics, paracentesis, TIPS in select cases), SBP (antibiotics acutely and for prevention in certain cases), varices (non-selective beta-blockers, OGD with band ligation), encephalopathy (lactulose). Patients with decompensated cirrhosis should be considered for referral for liver transplant. There are many variables that go into deciding whether a patient is eligible for liver transplant including Model for End-stage Liver Disease (MELD) score, comorbidities, presence of complications and absence of contraindications.

A word about cryptogenic cirrhosis. This entity is not that uncommon, representing at least 5% of cases of cirrhosis. While the term cryptogenic implies that the cause is unclear, it is thought that many of these cases are due to unrecognized NASH, silent autoimmune hepatitis, non-B/C hepatitis or possibly from prior unreported alcohol use. Diagnosing cryptogenic cirrhosis requires a thorough history and extensive investigations. Biopsy can be used to help identify a cause of cirrhosis – particularly metabolic causes. The management is similar to that of other causes of cirrhosis. 

Finally, when diagnosing and/or managing cirrhosis - do not hesitate to refer to a gastroenterologist. UHN is home to world class hepatologists with interdisciplinary liver clinics that provide high quality care for these complex patients.

1. Schuppan, Detlef, and Nezam H. Afdhal. "Liver cirrhosis." The Lancet 371.9615 (2008): 838-851.
2. Udell, Jacob A., et al. "Does this patient with liver disease have cirrhosis?."JAMA 307.8 (2012): 832-842.3
3. Caldwell, Stephen. "Cryptogenic cirrhosis: what are we missing?." Current gastroenterology reports 12.1 (2010): 40-48.

Thursday, June 11, 2015

SAB - Staphylococcus aureus bacteremia

Thank you to Dr. Panisko and Team 3 for presenting a case of Staphylococcus aureus bacteremia (SAB) at morning report yesterday.

SAB is a common and nteresting clinical problem that is encountered frequently on the general internal medicine ward.  It is very rare to see a general internist who does not enjoy managing SAB!

In Canada and the USA, the incidence of SAB ranges between 20 and 40 cases per 100 000 population.  Not surprisingly, a substantial portion are hospital acquired (40%).  Hospital acquired SAB is more commonly related to intravascular devices, as well as hospital acquired pneumonia. Complicating the entire picture is the rise of MRSA (both community acquired and healthcare associated) that makes the management of SAB somewhat more complicated.

Staph. aureus is one of the most virulent pathogens affecting humans, and the overall case fatality ranges between 20 - 30%.  Not only does it have a high mortality, but also because of it's virulence it has a high associated morbidity.  Staph. aureus can cause widely metastatic infections producing abscesses in the lungs, spine, kidneys as well as heart valves.  It is certainly an organism that one does not take lightly.

A collection of clinical pearls that I've gleaned over the years of working at UHN regarding Staph. aureus bacteremia.

1.) Staph. aureus in the blood cultures should always be taken seriously 

The presence of "gram positive cocci in clusters" should never be ignored or written off.  Given the extremely high mortality, when a clinician receives this result (usually by telephone call), repeat blood cultures and prompt initiation of therapy (typically empiric Vancomycin) should begin.

2.) Search for a source / signs of metastatic spread 

I believe one of the reasons general internists respect this clinical problem is because SAB forces the clinician to perform a thorough history and physical exam.  Looking for peripheral signs of endocarditis (embolic phenomenon), palpating for bony tenderness related to osteomyelitis of the spine, listening for a murmur, and examining the skin for signs of a soft tissue infection should all be performed.  Given the high mortality, a search for a source should ensue and any intravenous or intra-arterial lines should be removed.  Many times patients need imaging to identify occult abscesses.

3.) Consult infectious diseases (ID) - mandatory at UHN / Mt. Sinai

There have been a number of recent studies published that have shown mortality benefit for patients associated with prompt consultation with ID consultants.  Most of the mortality benefit is thought to be related to a number of clinical factors.  When ID is consulted, patients are more likely to have repeat blood culture sampling, more echocardiograms (both TTE and TEE), more effective and appropriate initial therapy, and typically longer courses of antibiotics.  The NNT for consulting ID and preventing death in SAB is 10.

4.) Staph. aureus in the urine should prompt a search for bacteremia 

 Dr. W. Gold has taught me many things about medicine, but this pearl has yielded a diagnosis of SAB on at least two occasions so far!  Occasionally, a urine culture that has been drawn as apart of a work up for "sepsis NYD" or "fever NYD" will yield the growth of Staph. aureus.  This does not typically represent a Staph. aureus UTI as this is a very rare phenomenon that is seen in patients with instrumentation or recent urological procedures.  When a clinician observes staph. aureus in the urine, this should prompt the collection of blood cultures as the cocci are very small and can be shedding from the blood, and filtered through the glomeruli into the urine.

5.) A transthoracic echocardiogram (TTE) is not sensitive enough to rule out endocarditis 

The sensitivity of TTE to detect endocarditis is thought to range between 65 - 80%.  This is not sufficient to rule out a vegetation in a patient with a high pre-test probability for endocarditis.  The presence of persistent bacteremia, non-resolving fevers, or other stigmata of endocarditis should make one ponder the need for a trans-esophageal echocardiogram (TEE).  Luckily, by involving an ID consultant this clinical dilemma can be discussed and appropriately managed.

6.) Antibiotic choice will vary depending on the patient, the presence of MRSA, and the ID consultant!

Some ID consultants (purists?) will recommend that all MSSA (methicillin sensitive staph. aureus) bacteremia be treated with high dose cloxacillin IV for 4-6 weeks, where as other ID consultants are not so dogmatic and will vary between cloxacillin or cefazolin.  There are studies that have been conducted recently by clinicians in the ID department at the University of Toronto to try and provide an answer to this question (Clox or Ancef?).  For some patients, receiving an IV infusion every 4-6 hours is not practical or feasible by their community care services.  Outpatient Vancomycin therapy for patients with MRSA SAB can also be anxiety provoking as the need for drug monitoring and potentially dose adjustments can complicate the matter.  Again, discussion with an ID consultant, community care, and close follow up are always recommended.

Management of Staphylococcus aureus bacteremia and endocarditis: progresses and challenges. Winfried V. Kern. Current Opinion in Infectious Diseases. 2010. 23:346-358.

Friday, May 8, 2015


Thank you to Team 1 (Amina) and Dr. McNeely for discussing a very interesting topic, and a favourite clinical problem of many great general internists (including Dr. Abdullah and former CMR Dr. Jacobs)!
A lonely pink eosinophil

Not only is eosinophilia challenging to pronounce, but it is also a challenging diagnostic dilemma, as the differential diagnosis is quite broad and often involves invasive investigations to characterize the underlying etiology (bone marrow biopsy, tissue biopsies, etc.)

The basic approach is to classify the eosinophilia as either a secondary phenomenon, or a primary eosinophilia. 

Within the differential for secondary causes of eosinophila are the following:

Infectious causes:
Tissue-invasive parasitosis (i.e. Strongyloides, Filariasis)

Noninfectious causes:
Drugs (Sulfa, CBZ)
Atopy / Allergies - asthma, atopic dermatitis
Autoimmune - eGPA, GPA (Churg-Strauss, Wegener's), Sarcoidosis
Malignancy - T-cell lymphoma, metastatic disease
Endocrine - Addison's disease

The differential for primary causes of eosinophilia is a collection of hematologic malignancies and myeloproliferative disorders:

Acute leukemias (AML)
Chronic Myeloid disorders - CML, and other molecular defined chronic myeloid disorders (8p11 Syndrome)
Clinicopathologically defined chronic myeloid disorders (MDS, PRV, mastocytosis)

Taking a thorough history is of critical importance, in terms of obtaining a detailed travel and exposure history that might suggest a parasitic infection.  A history of atopy, or other symptoms of vasculitis (wrist drop, hematuria, fever, arthritis, etc.)  

Tefferi, A. Blood eosinophilia: a new paradigm in disease classification, diagnosis, and treatment. Mayo Clinic Proceedings. 80(1): 75-83, 2005 Jan.

Tuesday, May 5, 2015

Vasculitis - one of the great mimickers

Thank you very much to St. Michael's Hospital and Dr. Yuna Lee's team for inviting me to do morning report (at lunch time!) at their amazing hospital!

We discussed a very interesting case of a patient presenting with a subacute mononeuropathy that ended up being diagnosed with microscopic polyangiitis (MPA).  This brought up a very interesting discussion of vasculitis and the different clinical presentations that patients often present with.  As a general internist, one of the most important methods of diagnosing a vasculitis is having an index of suspicion, and being aware of the different patterns that each vasculitis often presents with.

Vasculitis is a rare, and is a mimicker of many common conditions such as atherosclerotic disease, infections such as endocarditis, and viral infections.  The most helpful way to organize a basic approach to vasculitis is to think about the size of the vessel that the particular disease is affecting, from small vessel all the way up to large vessel.

Small Vessel:

Palpable Purpura

Microscopic Polyangiitis (MPA) - often presents with pulmonary-renal syndrome.  Can present with neuropathies as well.  Many patients with MPA will have positive p-ANCA or anti-MPO antibodies.

GPA (formerly Wegener's) - Again, often presents with pulmonary and renal involvement.  Also involves the nose/sinuses.  These patients can present with diffuse alveolar hemorrhage, and occasionally require ICU admissions.  Patients with GPA often have c-ANCA positivity, or anti-PR3 antibodies.  The classic rash seen in small vessel vasculitis is palpable purpura.

EGPA (formerly Churg-Strauss) - The clinical presentation is similar to GPA, but often there is a history of reactive airway disease or asthma.  These patients also have chronically elevated eosinophils, which is the main differentiation from GPA.

Anti-GBM (formerly Goodpastures) - presents very similar to GPA with pulmonary-renal syndrome.  Does not involve the nasopharynx typically.

Medium Vessel:

Livedo reticularis

In adults, the classic diagnosis is PAN (polyarteritis nodosa).  These patients often present with mesenteric ischemia, testicular pain, and the classic rash is livedo reticularis.  They often get involvement of axonal neurons resulting in wrist drops, and foot drops (mononeuritis multiplex).  PAN also classically has an association with Hepatitis B.

Kawasaki's disease is the other classic medium vessel vasculitis that presents in childhood, and occasionally early adulthood.

Large Vessel:

Takayasu's disease is a large vessel vasculitis that often affects younger women of Asian descent, which is the classic phenotype.  These patients often present with absent pulses (radial, ulnar) and can get symptoms of intermittent claudication (of upper and lower extremities).  They may also have blood pressure differences between the various limbs.

Giant Cell Arteritis - This is classic and relatively common large vessel vasculitis.  This typically presents in individuals over the age of 55 with nonspecific constitutional symptoms.  There is a strong association with PMR.  The classic symptoms patients will complain of include diplopia, headache, and jaw claudication.

Vasculitis is a large and interesting field that is very near and dear to the heart of internists in Toronto.  The UHN/Sinai's own Dr. Simon Carette and Dr. Christian Pagnoux are world experts in the field of vasculitis research!

Other clinical pearls that I've learned from rotating through the rheumatology service at UHN under the watchful eye of Dr. Carette are the following:

The classic CBC pattern that you will see in a patient (it is not-specific) but may be a clue that a vasculitis is occurring, is anemia of chronic disease (Hb in the 90-110 range) along with a mildly elevated leukocytosis (11.5-14 range) and a reactive thrombocytosis (450-650 range).  This pattern is non-specific, but is common in patients with chronic infections (osteomyelitis, and systemic illness likely vasculitis).  This is contrast to a patient presenting with SLE where cytopenia is the classic hematolgic manifestation (thrombocytopenia, anemia, lymphopenia)

Atherosclerosis is common!  atherosclerosis  is by far and away the most common cause of "disease of the blood vessels".  The vast majority of patients with ischemic limbs, mesenteric angina, and stroke have underlying atherosclerotic risk factors (HTN, diabetes) and this can masquerade as a "vasculitis".  As the old saying goes, when you hear hoof steps, its more likely to be a horse than a zebra (unless the case is being presented in morning report!)

Think about vasculitis in patients with multiple organ systems involved, particularly the classic pulmonary-renal syndromes.  Also, many clues to an underlying vasculitis often lie in the skin (palpable purpura, livedo reticularis, etc.) and so a consult with dermatology for skin biopsies can often lead to a diagnosis.


Friday, April 10, 2015

Systemic Lupus Erythematosus

Thank you to Megan Himmel and Dr. Rodrigo Cavalcanti for facilitating a fascinating case of a new diagnosis of systemic lupus erythematosus (SLE).


As Dr. Cavalcanti pointed out, SLE is more common in women of childbearing age, and also in individuals of African-American or African-Caribbean descent.  The overall incidence in the US population was 5.1 cases per 100 000 per year, and the prevalence is 52 per 100 000, which is increasing likely due to improved survival and increased recognition and diagnosis.


With effective treatment, the 10 year survival is reported at 92%.  The major causes of death in the first 5 years were SLE  and infections, and in the remaining 5 years, thrombosis was the most common cause of death.

Diagnostic Criteria:

I like the classic MD SOAP BRAIN mnemonic as a memory tool to remember the ACR diagnostic criteria for SLE.  Although the diagnostic criteria were recently updated in 2012 by the Systemic Lupus International Collaborating Clinics (SLICC)

ACR - Clinical Criteria -  (SLICC criteria in brackets)

M = Malar rash - (acute cutaneous lupus, malar rash, photosensitive lupus rash)
D = Discoid Rash - (chronic cutaneous lupus, hypertrophic lupus, panniculitis)

S = Serositis - (pericarditis, pleuritis, etc.)
O = Oral ulcers - (oral or nasal ulcers)
A = Arthritis - (synovitis involving two or more joints, and >30 min. morning stiffness)
P = Photosensitivity - (included in the M in the SLICC criteria)

B = Blood - hematologic manifestations - (Leukopenia, Thrombocytopenia, Haemolytic Anemia)
R = Renal - (proteinuria, glomerulonephritis)
A = ANA - (Anti-nuclear Antibody) - 95-98% sensitive for SLE 
I = Immunologic Criteria - (Anti-Ds-DNA, Anti-Smith, APLA, Low complement)
N = Neurologic - (seizures, psychosis, mononeuritis multiplex, confusion)

Lab Testing:
Sensitive = (negative - rule out) - ANA - 95-98% Sensitive
Specific = (positive - rule in) - Anti-ds-DNA - 95-98% Specific

Based on the ACR criteria for a diagnosis of definite lupus, a patient required 4 of the 11 criteria.  In the SLICC criteria, there are 17 criteria and a diagnosis of definite lupus is made if 4 of 17 are present.

The treatment of lupus is usually undertaken by specialists from rheumatology and/or nephrology depending on the degree of renal involvement.  The mainstay of treatment is immunomodulating/ steroid sparing agents, steroids, and increasingly biologic therapy.

Hydroxychloroquine (plaquenil) is the usual first line agent for mild lupus.  It is a immunomodulator and is useful for arthritis, skin rashes and systemic symptoms.  It is also the treatment of choice for lupus in the context of pregnancy.

NSAIDs - Are often used in combination with other agents for their analgesic and anti-pyretic properties.

Corticosteroids - Prednisone is often used for moderate to severe SLE flares as it is an extremely potent immunosuppressive.  For moderate flares, you will often see doses in the 0.5 - 1 mg/kg range.  For severe lupus flares, occasionally steroid pulses and high dose regimens are used.

Steroid sparing agents - Mycophenolate mofetil, Azathioprine, Methotrexate, Cyclophosphamide - All of these agents in various mechanisms target DNA synthesis at various steps in the synthetic pathway.  A rheumatologist will choose between the various agents depending on the severity of the flare, the prsence of renal involvement, and other clinical factors.

Biologic Agents - Rituximab has been investigated in a couple of randomized control trials (EXPLORER and LUNAR).  It has shown beneficial effects in certain sub-populations of patients, and will likely have an increasing role in the management of patients with difficult to control lupus.

Final Thoughts:  
SLE is a complex multi-system disease that requires consultation from a number of specialists, including rheumatology, nephrology, dermatology, and often psychiatry for the psychosocial aspects.

 It can mimic many common clinical presentations, and is on the differential diagnosis for almost any clinical presentation (from dyspnea to confusion).  The TWH is a world renowned site for SLE management and research, it is definitely worth an elevator ride down to the first floor of TWH whenever you are managing a patient with SLE!

Lisnevskaia, L., et. al. Systemic Lupus Erythematosus. Lancet. Volume 384. November 22, 2014.

Wednesday, March 18, 2015

Opportunistic Infections in patients with HIV

Today in morning report Dr. HPK discussed the various opportunistic infections (OI) that can infect a patient with HIV/AIDS.  He also provided us with a history lesson on the progression of the AIDS epidemic during the 1980's, and the experience of patients, nurses, and physicians during this challenging time!

The easiest way to approach this clinical problem is to think about the various infections that can infect patient as the CD4 count declines, and their overall risk for infections increases.  It is step-wise, and so obviously if your CD4 count is less than 100 you are at risk for all of the illnesses at the preceding steps (i.e. if CD4 count < 100 you are at risk for Toxoplasmosis, Cryptococcus and PCP, thrush, and TB).
Courtesy of Wikipedia - HIV virus (green) and a lymphocyte
Any CD4 count: Mycobacterium Tuberculosis (TB)
In any patient with HIV, irregardless of the CD4 count, Mycobacterium Tuberculosis (TB) can cause both pulmonary, and extra-pulmonary infections.  The current recommendations from the CDC are that any patient with HIV be screened for latent TB with a TB-skin test.  If a patient screens positive, and they have no evidence of active TB (i.e. pulmonary TB) then they should be treated for latent TB for 9 months (CDC guidelines).

Other infections to consider irregardless of the CD4 count include both bacterial and viral infections:

Streptococcus pneumoniae, influenza A/B, Human Papillomavirus (HPV), VZV, and Hepatitis A/B.

Fortunately, all of these pathogens have appropriate vaccines, and so making sure your patient has up to date vaccinations is a critical step.

1. Streptococcus pneumoniae (13-valent or 23-valent vaccine) - check guidelines for further details
2. Influenza A/B - seasonal vaccine each year
3. VZV - Varivax if never received it / or never exposed
4. HPV - for both male and females (13-26 years of age) - quadrivalent HPV vaccine
5. Hepatitis A/B - Hep A vaccine, Hep B vaccine, and combined Hep A/B vaccine

CD4 count < 250  - Coccidiodomycosis

If you reside in an endemic area for coccidiodomycosis (beautiful Arizona or California) then prophylaxis with fluconazole 400 mg PO daily should be initiated.

CD4 count < 200 - Pneumocystis Pneumonia (PCP) and Thrush (esophageal candidiasis)

Any patient with a CD4 count less than 200 or oral thrush/history of an AIDS defining illness should receive prophylaxis for PCP.  This is usually in the form of Septra (TMP-SMX).  Alternatives include Dapsone and aerosolized pentamidine.

The CDC does not recommend routine prophylaxis against esophageal candidiasis because it is highly treatable with fluconazole, and does not have a high associated mortality.

CD4 count < 150 - Histoplasmosis

Similar to coccidiomycosis, if you reside in an endemic area and the patient is at high risk because of occupational exposure or the location is a hyperendemic area, then prophylaxis with Itraconazole can be considered.

CD4 count < 100 - Toxoplasma Gondii and Cryptococcus

When the CD4 count drops below 100 your patient is at risk for developing both Toxoplasma and Cryptococcus.  Toxoplasma can cause space occupying lesions, seizures and other severe neurological symptoms.  Cryptococcus can present as insidiously and ultimately lead to a meningitis.

Prophylaxis for toxoplasma consists of Septra, and so if your patient has a CD4 count less than 100 they are most likely already going to be receiving Septra for PJP prophylaxis.  No prophylaxis is currently indicated for cryptococcus.

CD4 count < 50 - Mycobacterium Avium Complex (MAC) disease

If the CD4 count has drifted to below 50, then your patient is at risk for all of the above opportunistic infections and MAC.  The prophylaxis for MAC infection is a macrolide antibiotic (Azithromycin 1200 mg PO once weekly) or Clarithromycin.  before initiating prophylaxis, the clinican should first rule out active disease.

Magic - Many people still remember his November 7, 1991 press conference when he announced to the world that he was HIV positive.  Times have changed greatly since then!


The most important management decision when looking after a patient with HIV is the initiation of HAART (Highly Active Antiretroviral Therapy).  HAART has been one of the most incredible advances in medicine over the past few decades.  Many patients are able to maintain a normal CD4 count with undetectable viral loads for many years, and do not develop any of these opportunistic infections.

As Dr. HPK noted, in the 1980's the Toronto Western Hospital had dozens of patients admitted at any one time with pneumocystis pneumonia or cryptococcal meningitis.  Today, it is increasingly rare for internists to manage these problems.

CDC Guidelines on opportunistic infections