How much FIO2 do you get from that non-rebreather (NRB) mask? (aka There’s a hole in breather, dear Liza, a hole)


A non-rebreather (NRB) mask ONLY delivers high FIO2 when the flow meter valve is opened fully and even then only gets to about 90%.


I find great joy in reading old literature. I’m pretty sure I would be camped out in a medical library if it weren’t for the interwebs, thank goodness I can do all this from home…For this installment of the basics we look at how much oxygen that non-rebreather (NRB) mask actually delivering.

First, we need to talk about what a non-rebreather does. It’s mostly in the name but it isn’t always intuitive as I see it often used incorrectly. A non-rebreather should not allow you to re-breathe; yes I know how silly that sounds. However, when the NRB is working correctly the reservoir should not deflate with inhalation otherwise you are rebreathing. The NRB has two one-way valves. One is between the reservoir and the mask to allow oxygen to flow into the mask when you breathe in but not when you breath out. The other is at the side of the mask and allows the exhaled air to escape to the world. Ideally, the 100% FIO2 (Fraction of Inspired oxygen) in the bag is inhaled and thus the patient gets 100% oxygen delivery. Unfortunately in the real world this does not happen. One important reason is most NRB’s only have the one one-way valve so the patient doesn’t suffocate if the wall oxygen were to fail. Also the mask may not fit as tightly so oxygen will escape from the mask further decreasing the FIO2. Finally, we use oxygen on sick patients who may have increased respiratory rates and abnormal tidal volumes. Thus the normal minute volume will be abnormal in a sick patient. The goal of preoxygenation is to get the nitrogen part of normal air washed out and replace it with oxygen so our blood will be “super-saturated” and we will have longer times of apnea without the oxygen saturation dropping while a patient is being intubated. Thus we would like as high as possible FIO2 going into the patient. Also we really cant tell what the PaO2 is from the pulse ox monitor because no correlation can be made once SaO2 is >98%.

Thus we NEED to know the FIO2 of the non-rebreather going in. To figure this out we go back in time to 1991 while this might sound scary to go to a time period without iphones; we only have to be there long enough to see the article by Farias, Delivery of High Inspired Oxygen by Face Mask in the Journal of Critical Care. The recruited 5 healthy male volunteers and changed respiratory rates, tidal volumes and oxygen flow rates to see the effect on the FIO2 of a non-rebreather. Now here is the impressive part of the study “FIO, was measured from a catheter positioned through the nose so that its tip was in the pharynx”. The “catheter” was a 6Fr Foley that was inflated once in the oropharynx! These people volunteered for this? You sure they didn’t get paid something???? Yikes! This is why the old literature is entertaining! Anyway…. All wall oxygen comes through a flow valve with a little metal ball that measures the flow out. Normally, with a NRB the flow is set to 15L on the meter and sadly this is the highest the meter reads. So to test various flow rates they had to use their own flow meter and tested rates of 15, 20,30,40 and 60 L/min. They note, “by opening the valve fully on the flow meter, it was possible to attain an 02 flow of 60 L/min. “. At each flow rate the volunteers produced respiratory rates (RR) of 20, 30, and 40 breaths/min (brpm) and also varying tidal volumes (VT) of 500, 750, and 1000 mL. Normal tidal volume is around 500 ml or 7 ml/kg. For a normal RR of 20 and a “normal” VT of 500ml, when the flow was set to 15L the best FIO2 obtained was just shy of 60%. When the RR was 40 brpm and the VT was 1000 ml, this fell to 40% . The study notes, “When the O2 flow rate was increased to 60 L/min, the FIO2, remained at approximately [90%] despite increasing respiratory frequency and tidal volume.”

Hence, this is the reason I say crank that little knobby thing on the oxygen until it stops. Because only then are you getting 60 L/min of oxygen and only then, regardless of RR and VT, does one approximate FIO2 of 100%…well ok 90% but I’ll take it! The study also looks at ways to improve the masks by doing stuff to them but since I’m just using the NRB before I intubate I’ll be sufficiently happy with a 90% FIO2 to preoxygenate my patients besides I don’t need to look any crazier to the staff by making an art project out of the masks before I intubate someone…


Farias, E. et al. Delivery of high inspired oxygen by face mask Journal of Critical Care. Volume 6, Issue 3, September 1991, Pages 119-124.


Screen Shot 2019-03-21 at 12.49.51 AM

Recent Recommendations in Neonatal Resuscitation 2019 UPDATE

Sometimes we need a reminder and update on the basics… Your Welcome..

Recent Recommendations and Emerging Science in Neonatal Resuscitation. Pediatr Clin N Am 66 (2019) 309–320.

– The 2017 NRP guidelines recommend a 30- to 60-second delay in clamping in all term and preterm infants not requiring resuscitation.

– If the placental circulation is disrupted (e.g., placental abruption), the cord should be clamped immediately. Investigators found that delayed clamping reduced mortality before discharge.

– Another option is to clamp and cut a long segment of the cord immediately after birth and then milk the cord. The major advantage of either method of cord milking is that the newborn can be passed to the awaiting resuscitation team without delay while preserving the receipt of the placental transfusion.
– Delivery room temperatures should be at 23 C to 25 C (74 F–77 F).

– Suctioning of newborns should be performed only if the airway is obstructed or if PPV is needed.
– In nonvigorous infants born through meconium-stained amniotic fluid (MSAF), current recommendations include only intubation and endotracheal suctioning for those who need it for ventilation or airway obstruction.

– It is recommended that room air (21% O2 at sea level) be used at the initiation of resuscitation in infants born at 35 weeks gestation.

– Failing to reach SpO2 of 80% at 5 minutes was associated with adverse outcomes including intraventricular hemorrhage, and risk of death was significantly increased with time to reach SpO2 80%.
– Endotracheal intubation is indicated for ineffective or prolonged positive pressure ventilation (PPV) or for special circumstances such as an abnormal airway anatomy.

– Chest compressions are indicated when heart rate remains less than 60 beats/min after at least 30 seconds of PPV. When chest compressions begin, supplemental O2 may be increased until the heart rate recovers and weaned rapidly afterward.
– Epinephrine is indicated when the heart rate is <60 beats/min after 60 seconds of chest compressions coordinated with PPV using 100% O2

Cardiac Arrest in Pregnancy: An 2019 UPDATE

Sometimes we need a reminder and update on the basics… Your Welcome..

Cardiac arrest in pregnancy. SEMINARS IN PERINATOLOGY 42(2018)33–38.

– The maternal heart rate increases by 20–30% or 15–20 beats per minute

– Cardiac output increases by 30–50% or 1.8 L per minute with the uterus receiving approximately 17% of maternal cardiac output in the third trimester.

– The diaphragm elevates by up to 4 cm in the third trimester causing decreased chest compliance

– Functional residual capacity decreases by up to 25% in the supine position at term.

– Pregnant patients experience mild respiratory alkalosis.


– Aortocaval compression (ACC) needs to be addressed during cardiac arrest management.

– The American Heart Association recommends manual left uterine displacement (LUD) throughout resuscitative efforts and during perimortem cesarean section until delivery of the infant.

– In the past, ACC was addressed by placing the patient in a tilt; however, this is no longer recommended.

– Numerous studies have shown that maternal tilt decreases efficacy of chest compressions, which hinders resuscitative efforts.

– Successful manual LUD can be performed from the patient’s right and left side.

– From the right side, the uterus is pushed upward and leftward to relieve pressure from the maternal vessels.

– Care should be ensured that the uterus is not inadvertently pushed down.


– The most common cause of maternal mortality is venous thromboembolism followed by preeclampsia and eclampsia.

– In an analysis of cardiac arrest in pregnancy in the United States the most common causes of arrest were hemorrhage, heart failure, amniotic fluid embolism, and sepsis.


– Chest compressions are performed in the same manner as for non-pregnant patients with a rate of 100–120 compressions per minute and a depth of at least 2 in with minimal interruptions.

–  The most recent guidance state that hand placement for chest compressions should be in the center of the chest on the lower portion of the sternum in the same manner as for non-pregnant patients.

Bag-mask ventilation with 100% oxygen with a rate of at least 15 L/min should be initiated immediately with a compression–ventilation ratio of 30:2.

– Early defibrillation should be provided when appropriate, and modifications in shock energy are not indicated. Studies have shown that transthoracic impedance is unchanged in pregnant patients. Providers should not delay or withhold defibrillation due to concerns for fetal safety. During defibrillation, a minimal amount of energy is transferred to the fetus, and it is safe to defibrillate a patient at any stage of pregnancy.

– Since airway management is more challenging, intubation should be attempted by the most experienced provider available with the use of a smaller endotracheal tube with a 6.0–7.0mm inner diameter to increase the likelihood of successful intubation.

– Medical therapy for cardiac arrest in pregnancy is no different than for non-pregnant patients.

– Medications do not require dose alterations, and no medication should be withheld due to concerns for fetal teratogenicity.

– During active CPR, the AHA guidelines recommend against fetal assessment, and all fetal monitors should be removed from the patient. The goal of CPR is to restore circulation in the pregnant patient. Evaluating the fetal heart rate is not helpful at this time and can interfere with maternal resuscitative efforts.
– PMCD should be initiated after 4 min of failed resuscitative efforts with the goal of delivery within 5 min of initiation of resuscitative efforts.

– As a caveat, if the mother has clearly non-survivable injuries, it is not necessary to wait to begin the PMCD. Transfer to the operating room is not recommended.

– With regards to technique for cesarean section, both vertical and Pfannenstiel incisions are acceptable and are at the discretion of the obstetrician. If the underlying arrest is secondary to trauma, a vertical incision is preferred given that it provides better visualization of the abdomen.
– If restoration of spontaneous circulation (ROSC) has been achieved without undergoing a PMCD, the patient should immediately be placed in the full left lateral decubitus position.

– The 2015 AHA guidelines now state that pregnancy is not an absolute contraindication, and therapeutic hypothermia can be considered on an individual basis.


Sometimes things aren’t as they seem.  As an attending I often hear from a resident “Hey got an easy one for ya, 50 y/o diabetic with bilateral cellulitis vanc given, I’ll get some admit labs” followed by a precipitous mic drop.

I was told once that a good attending “makes a difficult case simple and a simple case difficult” I believe that more and more these days. So is a simple case of cellulitis really that simple? Well in a study by Weng in 2016, 30% of 259 pts were misdiagnosed with cellulitis. A similar study by Li also 30 percent of 116 patients were incorrectly diagnosed as cellulitis” in a study by Li in 116 patients. Thirty percent is a large number, but does it surprise me? Not really.  I have heard so many times a patient is getting admitted for “bilateral cellulitis”.  So you are telling me there are two skin infections starting at the feet and racing to the groin like two trains approaching each other??? If you think about it kind of doesn’t make sense that this would exist much. Maybe in a very immune-compromised patient it can occur but otherwise the patient should be really sick. If I really think about it in IVDU patients with skin abscess all over even they don’t have bilateral cellulitis! In fact every case report of bilateral cellulitis I found was actually on how “bilateral cellulitis” was the WRONG diagnosis.

So if its not cellulitis then what is it (we will get back to the bilateral part in a bit)?


I asked an intern what is his approach to a patient differential diagnosis. “Worst first” he said. I chuckled a bit since I appreciate alliteration. But in an overall way he is on the right track so on that note lets start off with the worst: Necrotizing fasciitis (NF). However, this will not be a lengthy discussion because there are so many good resources on NF. Instead I want to focus on some pearls for this diagnosis. To start one thing that makes me feel better is this quote in Clinical infections disease by Anaya “Establishing the diagnosis of necrotizing soft tissue infections is not easy”. Thanks for that, what no clinical correlation recommended? The typical findings of blisters/bullae, crepitus, gas, fever, tachycardia, hypotension, and shock have a low sensitivity of only 10-40%.  Well is there anything that can help us? When all else fails examine the patient. We need to see if the area is necrotic. So, one helpful trick is the “finger test”in the almost brilliantly named BMJ article “Necrotizing Fasciitis: Always used the finger” (which I would have said give them the finger but the Brits always have to be proper). Here a test incision is made in the suspected area of approximately 2 cm. A positive test would be characterized by the absence of normal blood flow, dirty’ dishwater’ colored fluid and discoloration of the fat. Then a rapid finger sweep at the level of the fascia can be carried out. If the tissues dissect with minimal resistance this again favors the diagnosis.



The simplest way to think about this would be to say there is cellulitis and then there is everything else. So what are some pearls for the diagnosis of these tricksters?

  1. Cellulitis is rarely bilateral! Fact!
    1. Corollary: Vascular dermatitis is usually bilateral
  2. Elevate the leg when you examine it! Dependent redness is often mistaken for cellulitis, but erythema promptly disappears after elevating the leg at the bedside. Dependent redness is often asymptomatic, but can be associated with rest pain from arterial insufficiency.
  3. Stasis dermatitis (AKA venous eczema) is the most common mimic of cellulitis.
  4. Superficial thrombophlebitis can appear similar to lymphangitis however DVT of the lower limb rarely causes cutaneous erythema EXCEPT in the proximal thigh, where the femoral vein lies just below the skin surface
  5. Cellulitis is not typically itchy, its painful
    1. Corollary: IF the area is insensate worry about Nec Fasc!
  6. Malignancy affecting the lymphatics of the lower extremities can closely mimic cellulitis!
  7. Failure of  antibiotics should make you think twice about the diagnosis of cellulitis
  8. Cellulitis should have inguinal lymphadenopathy
  9. Confluent Erythema Nodosum can have a similar appearance but will have a different feel since it is a panniculitis
    1. Corollary: Cellulitis should be smooth and not be made up of firm nodules!
  10. Erythema Multiform can have a similar appearance and mistaken for allergic reaction or cellulitis.


Stasis Dermatitis (aka Venous Eczema):

  • Ill-defined, bilateral, pitting edema of the lower extremities, typically with erythema, hyperpigmentation, serous drainage, and su- perficial desquamation
  • Chronic venous insufficiency causes micro- vascular changes and microthrombi leading to acute cutaneous inflammation
  • ill-defined erythematous plaque with overlying pigment changes and super- ficial desquamation, as well as nonpitting edema


  • Necrosis, and fibrosis of the subcutaneous fat, especially in women
  • It usually develops much more slowly than cellulitis—over weeks to month
  • A sclerosing panniculitis classically described as an “inverted champagne bottle” or “inverted bowling pin” appearance of the leg, ie, the diameter of the leg is sharply narrowed directly below the calf
  • There is an acute and a chronic phase. The acute phase is characterized by inflammation and erythema, and the chronic phase is characterized by fibrosis
  • The acute phase can be difficult to differentiate from cellulitis. venous insufficiency, cutaneous changes of stasis dermatitis, and the absence of systemic symptoms all point to lipodermatosclerosis.

Contact Dermatitis

  • The lesion is a painful, nonpruritic, well-demarcated, erythematous, weeping plaque with scattered vesicles at the periphery, as well as superficial desquamation and scaling.
  • Ask about recent changes in medications, soaps, and laundry detergents, new hobbies, or recent surgeries. The involved site is often confined to the area where the allergen contacted the skin


  • Localized edema of an affected extremity, with induration, erythema, and secondary cutaneous changes such as hyperkeratosis, dyspigmentation, and wart-like architecture
  • Has the patient undergone lymph node dissection? Has the patient had an injury in the affected leg? Lymphedema is overwhelmingly unilateral and nonpitting, and is often seen in overweight people

Eosinophilic cellulitis, or Wells syndrome

  • It is a recurrent hypersensitivity reaction to a drug, to a vaccine, or to an insect bite, or to a viral or fungal infection that pre- sents on the extremities as localized erythema, edema, and induration with sharp borders and a green or gray hue
  • Patients tend to report itching and burning that precedes the
  • onset of plaques the complete blood count typically shows a transient hypereosinophilia


  • is a rare disorder that causes episodes of heat, redness, and burning discomfort provoked by heat and dependency and relieved by elevation and cooling of the extremity
  • Primary occurs more common in women than men
  • Can be a marker of systemic disease associated with: myeloproliferative neoplasms of polycythemia vera, essential thrombocythemia, and chronic myelogenous leukemia.
  • Age of onset is typically in the forties and fifties.
  • The symptoms are nearly always intermittent, with episodes usually lasting from a few minutes to a few hours, but occasionally lasting for several days
  • They are typically provoked or worsened by limb dependency, exercise, and heat and alleviated by the opposite of these


  • The onset of an attack is abrupt, causing severe pain, tenderness, erythema, swelling, and warmth over the affected joint.
  • Evolves rapidly, reaching its maximal intensity within 6 to 12 hours.
  • Commonly expands a considerable distance beyond the joint itself, producing extensive cutaneous inflammation that may strongly resemble cellulitis.
  • Fever occasionally occurs,
  • Careful examination, , usually indicates that the origin of the inflammation is clearly in the synovial space, rather than the soft tissues
  • Movement of the affected joint and pressure over it produce exquisite pain that is less intense with compressing adjacent inflamed, non- articular tissue
  • Don’t order a serum uric acid! It doesn’t help!


  1. Weng QY, Raff AB, Cohen JM, Gunasekera N, et al. Costs and consequences associated with misdiagnosed lower extremity cellulitis. JAMA Dermatol.
  2. Li et al. Outcomes of Early Dermatology Consultation for Inpatients Diagnosed With Cellulitis. JAMA Dermatol. 2018 May 1;154(5):537-543.
  3. Anaya et al. Necrotizing Soft-Tissue Infection: Diagnosis and Management. Clinical Infectious Diseases, Vol. 44, No. 5 (Mar. 1, 2007), pp. 705-710
  4. Necrotizing Fasciitis: Always use the finger. BMJ 2005;330:830
  5. Lower limb cellulitis and its mimics Part II. Conditions that simulate lower limb cellulitis. J Am Acad Dermatol 2012;67:177.e1-9.
  6. Keller EC. cellulitis mimics. Cleveland Clinic Journal Of Medicine. 79: 8 2012 547- 552

Aspirin for All Comer Chest Pain: Is it Naughty or Nice

Screen Shot 2018-12-04 at 2.35.59 AM.png

Like it or not,  we do a fair amount of primary care medicine in the emergency department. Therefore, we need to know some of the basics about primary care.  One of the most fundamental concepts is the use of aspirin for prevention of MI either before the first one (primary) OR after (secondary) your patient has had a heart attack or stroke. Spoiler alert… 2018 was not a good year for aspirin. It would seem our beloved miracle drug has not been able to escape the rigors of time and medical reversal; much as we have seen with so many other treatments in medicine. Don’t panic, using our beloved aspirin during and after an acute MI hasn’t changed (see you don’t even need a towel!). But whom in the ED, do we need to discharge on a daily aspirin? How good is aspirin in the PRIMARY prevention of cardiovascular disease? Well the holiday season of 2018 has gifted us with 3 trials to help us answer these burning questions.


     First a little back story. I was born at a very young age…oh sorry not my back-story, the aspirin in prevention back-story. To understand the controversy we need to define a few terms. The first is MACE, which stands for major adverse cardiac events defined as MI (aka myocardial infraction, aka heart attack), stroke, and vascular death (and sometimes arrhythmias are thrown in to the definition as well). MACE is what separates the terms Primary prevention from Secondary Prevention. The people without MACE are the people we are PRIMARILY trying to prevent bad things from happening. If they HAVE had MACE then we are trying to SECONDARILY prevent a repeat of bad things. As I said above the role of aspirin has been well established in SECONDARY prevention (i.e. prior MI). This was established by the famous Antithrombotic Trialists’ Collaboration in 2009 in over 200,000 patients. The use of aspirin in the PRIMARY prevention however has been somewhat controversial despite 30 years of studies! The harm from aspirin is mostly due bleeding and over the years we have had better and better medications to treat vascular pathology (specifically statins, neuropathy and myopathy aside).

     OK so now that we are again confident that aspirin is great SECONDARILY (once you have a heart attack), we need to figure out if it is useful for PRIMARILY preventing heart attacks (BEFORE you have one). Below we discuss the THREE trials that help us figure this out.

     The first trial we will discuss is the ARRIVE trial. ARRIVE was done in seven countries, including the US mostly in primary care offices. Eligible patients were either males aged 55 years and older with >1 risk factors or females aged 60 years and older with >2 risk factors. The study also checks all the usual boxes for doing a study the right way including randomized, double blind, placebo-controlled, multicenter, and intention to treat analysis. Patients had a moderate vascular risk but NO MACE, so all primary prevention and were randomized to get aspirin or placebo. The primary outcome was a time to event analysis for the composite of who did get MACE. GI bleeding was the major safety outcomes. They enrolled about 12,500 pts and randomized about 6300 to each group. For the primary outcome of MACE the aspirin group had 4.3% vs. 4.5% MACE (HR of 0.8-1.1; p=0.6). For the GI bleeding outcome aspirin had 0.97% vs. 0.46% in the placebo group (HR 2.11, 1.36–3.28; p=0·0007). So, aspirin had no difference in MACE but doubled the bleeding.

     In order to see if we can find a subgroup of people the MIGHT benefit with primary prevention of aspirin we turn to the ASCEND trial. This jolly good UK trial looked specifically at diabetics for primary prevention. They enrolled men and women at least 40 yo who had a diagnosis of diabetes and followed them for 7 years. The primary outcome, which was modified DURING recruitment (naughty, naughty!) was MACE and the primary safety outcome was any major bleeding (ICH, GI, etc.). The trial was a randomized, placebo controlled and intention to treat study.  The trial included 15000 pts. and randomized 7740 to each group (I know that math doesn’t add up but work with me here). For the primary outcome of MACE, the study found a 1% benefit to aspirin overall, with 8.5% for aspirin and 9.6% for placebo (RR of 0.88 with p=0.01). However, that 1% difference occurred only in the first 3 years (2.6% vs. 3.5%) but after 3 years there was no further difference (2.7% vs. 2.7%). The “any” major bleeding risk however was 4.1% in the aspirin group and 3.2 in the placebo group (RR 1.29 with p=0.003). So even if we don’t look at these groups by year (and the subgroups were NOT pre-specified so we shouldn’t – More naughtiness!); the benefit to aspirin for primary prevention was 1% and the harm was 1% making it a wash. This looks like a job for shared decision making in the first 3 years…

     The last study we will discuss looks at another at risk group, the elderly. The ASPREE trial involved men and women from Australia and the United States who were 70 years of age or older (or ≥65 years of age among minorities in the United States). Again this was a randomized controlled trial of aspirin or placebo of approximately 19,000 (only 2500 from the US) “relatively healthy” elderly people who were followed for about 5 years. The primary end point was disability-free survival, which was defined as survival free from dementia or persistent physical disability. The primary composite end point was derived from the first end-point events of death, dementia, and persistent physical disability. Of note the trial was stopped early because “it was extremely unlikely that continuation of the trial intervention would reveal a benefit with regard to the primary end point”. Unfortunately, this trial makes the naughty list for listing two primary outcomes “The primary end point was disability-free survival… The primary composite end point was derived from the first end-point events of death, dementia, and persistent physical disability”. Then they go on to mistakenly list not survival but mortality. About 9500 were allocated to each group. They list “any cause of death” as 5.9% with aspirin and 5.2 with placebo, so no difference. They list CV disease including stroke as 1.0% for aspirin vs 1.2% for placebo. They list major hemorrhage as 0.3% for aspirin and 0.3% for placebo. They also go on to say in the abstract that there were more cancer related deaths in the aspirin group but that isn’t really an appropriate conclusion since we don’t even know if this is an a priori secondary outcome or just statistical jujitsu after the fact.

So to summarize these three large trials: ARRIVE had no change in MACE but increased bleeding; ASCEND had a decrease in MACE but with an equal and opposite increase in bleeding; and ASPREE had no change in mortality, no change in MACE and no change in bleeding. So does the patient with chest pain but without MACE get an aspirin forthright?

I would say no to that…and to all a good night!


1. Gaziano, J. Use of aspirin to reduce risk of initial vascular events in patients at moderate risk of cardiovascular disease (ARRIVE): a randomised, double-blind, placebo-controlled trial. Lancet 2018; 392: 1036–46.

2. Bowman, L. Effects of Aspirin for Primary Prevention in Persons with Diabetes Mellitus. N Engl J Med 2018;379: 1529-39.
 ASCEND Study Collaborative Group.

3. McNeil
J. Effect of Aspirin on All-Cause Mortality in the Healthy Elderly. N Engl J Med 2018;379: 1519-28.
 ASPREE Investigator Group

The Basics: How to interpret a chart review

I can’t stress enough how important it is to understand how to read the literature. It makes a world of difference in interpreting the results. When I read a study I always read the methods and the results section only. Then I make my interpretations and see if they match the authors conclusion. But to make conclusions you need to know a little something about methods. In this review, I want to go over what methods are important for the interpretation of chart reviews. Remember that chart reviews are not only the weakest type of study, they are unable to provide cause and effect. They are observational studies and therefore can only show ASSOCIATION not CAUSATION. There are lots of different checklists but unlike RCT’s there is no universally agreed upon methods for chart reviews (unlike the PRISMA for systematic reviews). Fortunately, we have two great papers in our own journal to base appropriate chart reviews on. The papers are by Goodman and Lowenstein (1996) and by Kaji and Schriger (2014). Definitely a must read for methodologists. In this review I want to just summarize them and formulate my own simple checklist (See Table 1) for validating chart reviews. The table below gives a summary of the 5 sources of biases that should be accounted for when looking at a chart review. Without rigorous methods studies severely lack the ability to answer their stated question. For purposes of statistics here when I refer to the word bias I mean the tendency of a measured sample (study) to INCORRECTLY estimate the population (true effect).

Is there INVESTIGATOR bias?

Why did the investigators do the study? There could be lots of reasons and so looking for financial bias or even intellectual bias (maybe they published in that area in the past) is important. Check for financial disclosures or honorariums given. Is a chart review the appropriate outlet to answer the proposed question? Or even more importantly ensure a primary outcome was proposed.

Is there CHART bias?

What method was utilized to find the charts. Were ICD codes or chief complaints used. Was there a sufficient sample of charts used? Were the charts chosen in a convenience sample or consecutively. The best way to look for charts is to perform random sampling of the charts. This is often not possible but the method of how and what charts were chosen should be discussed. How many charts are needed to arrive at the intended answer? A power calculation should be supplied by the paper. The included and excluded charts should be delineated. Also, why were charts excluded or included. Inclusion and exclusion criteria for picking charts should be determined ahead of time (a priori).  The paper should have a table of the clinical/baseline characteristics of the charts or patients being sampled. Finally, there should be a flow diagram showing how the charts were picked.

Is there DATA bias?

Once the charts are picked the data that is being extracted needs to be determined. This should be done ahead of time. Data collections tables (DAT) should be made and then trialed in a small pilot study to ensure they are adequate to get the intended data. That same DAT should be provided in the paper or in a supplement. What if the some of the data are missing? How this should be handled should also be determined a priori. A sensitivity analysis should be done to see how much missing data will negatively impact the results of the study. What if there is conflicting data in the same the chart? How will that be handled and what impact will it have?

Is there ABSTRACTOR bias?

The abstractors should not be the study investigators. Often times they aren’t even medical providers and therefore they should be trained appropriately in how to collect the data and how to use a data collection tool (DAT). The abstractors should also be blinded to the study hypothesis so there is no bias. Finally, the abstractors should be monitored and the monitoring process should be described in the paper especially if it’s a long trial. Maybe even repeat training should be done. How well do the abstractors agree with each other? Will different abstractors be able to get the same results? This is inter-rater reliability. A measure of this reliability should be given in the study. Both the percent agreement and a kappa score should be reported, at minimum. Lastly, the ability of the abstractor to agree with himself should be measured, that is the Intra-rater reliability.

Is there RELIABILITY bias?

The data should be reproducible and reliable. Kappa values and percent agreement should be reported for the data as well. How many variables should be looked at for reliability? Ideally all of them however at minimum the ones important for the study hypothesis should be checked. Also, how many of the data points should be checked for reliability? The consensus seems to be at minimum 10% of the data should be checked for reliability. Is the kappa level chosen appropriate or should it be higher or lower?

Table 1.

The Checklist:
1.     Investigator bias
        a.     Question appropriate for chart review?
        b.     Financial/Intellectual disclosures supplied?
2.     Charts bias
        a.     Methods of Chart Identification (CC vs ICD-10)
        b.     Sufficiently sampled?
        c.     Was a power calculation supplied?
        d.     A priori inclusion criteria?
        e.     A priori exclusion criteria?
        f.      Table of clinical characteristics?
        g.     Flow diagram delineating how the study population was derived?
3.     Data bias
        a.     Defined A priori?
        b.     Coding guide for abstractors?
        c.     Coding guide provided?
        d.     Standardized Data Collection tool (DAT)?
        e.     Was the DAT pilot tested?
        f.      Was the DAT provided?
        g.     Is there missing or conflicting data?
        h.     How is missing data handled (sensitivity analysis)?
4.     Abstractor bias
        a.     Blinded to study hypothesis?
        b.     Trained appropriately?
        c.     Monitored?
        d.     Inter-rater Reliability?
        e.     Intra-rater reliability?
5.     Reliability bias
        a.     Kappa and percent agreement calculated for the data?
        b.     What level of reliability and why was that level chosen?
        c.     Which of the collected variables were checked for reliability?
        d.     What percent of the data was checked for reliability?


Gilbert EH, Lowenstein SR, KozioI-McLain J, Barta DC, Steiner J. Chart reviews in emergency medicine research: Where are the methods? Ann Emerg Med March 1996; 27: 305-308.

Kaji AH, Schriger D, Green S. Looking Through the Retrospectoscope: Reducing Bias in Emergency Medicine Chart Review Studies. Ann Emerg Med. 2014; 64: 292-298.

Anti-Platelets in PCI: Which one and Why?

Pharmacology is a favorite subject of mine and so I think it super interesting to discuss. Unfortunately, this post may be one of my most boring write ups. Luckily, the content makes up for it! (Hopefully). I bolded the key points to help. This one is a quickie.

  1. Clopridogrel (Plavix)

Loading Dose: 300mg or 600mg depending on preference of cardiologist

Clopidogrel is a thienopyridine which irreversibly blocks the P2Y12 platelet receptor. The CURE trial (n = 12,562) helped show the efficacy and safety when combined with aspirin in patients with NSTE. The combination reduced the composite primary outcome of cardiovascular death, non-fatal MI, or stroke by 2% absolute risk (9.3% vs. 11.4%) vs aspirin alone. A 1% absolute increase in major bleeding events was noted, but there was no significant increase in life threatening or fatal bleeds. Typical loading dose is 300 mg however, for some patients with PCI it is common practice to use a higher loading dose of clopidogrel (600 mg) in order to more rapidly achieve full antiplatelet effect. Clopidogrel is a prodrug. Genetic polymorphisms of CYP3A4 and CYP2C19 cause variation in the effectiveness of clopidogrel. Due to the irreversible binding it has a slow in offset (over several days) which can be clinically important when urgent CABG is required.

2. Prasugrel (Effient)

Loading Dose: 75mg

Prasugrel also requires metabolism to an active form since it is a prodrug. Metabolism is rapid due to esterases, and a single CYP450 so there is minimal change in metabolism. In TRITON-TIMI there was a reduction in recurrent cardiovascular events in the prasugrel treated group vs clopidogrel (9.3% vs. 11.2%), driven by a reduction in non-fatal MI (7.1% vs. 9.2%). A significantly lower rate of stent thrombosis was also seen in the trial with prasugrel (1.1% vs. 2.4%). More bleeding, however, was observed in the prasugrel arm (2.4% vs. 1.8%) and with an increase in fatal bleeding (0.4% vs. 0.1%). Patients with a previous TIA or CVA experienced net harm with prasugrel thus the drug is contraindicated in these patients.

  1. Ticagrelor (Brillinta)

Loading Dose: 180mg

Ticagrelor the first non-thienopyridine ADP blocker and is a cyclopentylazopyrimidine. It reversibly binds to the P2Y12 platelet receptor. In addition to reversible inhibition of the P2Y12 receptor, ticagrelor also inhibits adenosine reuptake in cells. This can cause additional complaints of dyspnea (5% of patients) and bradycardia in addition to bleeding side effects. The PLATO study compared ticagrelor (180 mg loading dose, followed by 90 mg twice daily) with clopidogrel (300–600 mg loading dose, followed by 75 mg daily) in 18,624 patients who presented with either STEMI or NSTEMI.  Ticagrelor was associated with a 2% absolute risk reduction in the primary endpoint of vascular death, MI, or stroke compared with clopidogrel. All-cause mortality (4.5% vs. 5.9%) and stent thrombosis (1.4% vs. 1.9%) were lower with ticagrelor. There was no difference in major bleeding or blood transfusions. The ATLANTIC trial (n = 1875) assessed whether pre-hospital ticagrelor loading was superior to delayed in-hospital ticagrelor. The co-primary outcomes were related to arterial patency at angiography with no different between the two groups. Misdiagnosis of STE was identified in 10% of patients. The difference in timing of loading dose was approximately 30 min. Furthermore, it was also noted that opiate administration impaired oral absorption of ticagrelor. Further, it was also noted that opiate administration impaired oral absorption of ticagrelor. This may be a potential limitation of ALL oral P2Y12 receptors in the STEMI setting.

  1. Cangrelor

Cangrelor is an IV form of the cyclopentylazopyrimidine ADP blocker. It has a faster onset and offset through reversible inhibition of the P2Y12 receptor. It has a very short half-life (<10 min), thus platelet function can be restored within a couple of hours of discontinuation of the infusion. Cangrelor has been studied in three large trials, 70% of whom had presented with an ACS. A meta-analysis of these trials demonstrated a reduction in periprocedural death, MI, ischemia-driven revascularization and stent thrombosis for cangrelor vs clopidogrel (3.8% vs.  4.7%).



Brown. Continuing Cardiology Education, 2017; 3(1) pp11-21


How to interpret a Forest Plot

Let’s take a short trip into the woods and discuss Forest Plots (They are called that because if you lay the graph on its side ups and downs look like tree line…or maybe because everyone gets lost in them…). Let’s look at the one below from a Cochrane review 1.

Screen Shot 2018-06-27 at 10.13.10 AMIt looks daunting but really once we break it down its not bad:

  1. The main graph looks like an upside down T. The bottom horizontal line is a binary decision that tells you if one side or the other is favored. The far left favors control and and the far right favors the experiment. In this case spacer (left) or nebulizer (right). The vertical line is smack dab in the middle of the two and is the point of “no effect”. If anything touches that line then it means that whatever you are looking at has no effect (effect of control=effect of treatment).

Screen Shot 2018-06-27 at 10.36.04 AM

  1. Starting from the top down on the graph we see a box with whiskers. The box is the average result of the study and the whiskers are the 95% confidence interval. Remember that when we ask a question there is no ONE answer (average). The answer lies somewhere within an interval that we can be 95% sure contains that answer.

Screen Shot 2018-06-27 at 10.37.58 AM

  1. Another thing to note about the box and whisker is that a study with a large sample size will have short whiskers and a small sample size will have long whiskers (larger studies by their nature are more precise)

Screen Shot 2018-06-27 at 11.07.45 AM

  1. Included data: The plot will also contain columns to the right of the graph. In this case we have: Study, Mean, Standard Error (SE), Weight, and the 95% confidence interval. I want to point out the weight column. Because some studies have more patients than others they contribute differently to the overall forest plot. The weight tells you how much that single study contributed. Larger studies have higher weights and will contribute more to the overall effect. This is important because a meta-analysis might have its entire results based on ONE study that overpowers all the others

Screen Shot 2018-06-27 at 11.15.47 AM

  1. The Diamond: The middle of the diamond is the mean of the result of all the above trials. The ends of the diamond are the 95% confidence interval. So if the diamond touches the “line of no effect”, there is no difference between control and experiment.

Screen Shot 2018-06-27 at 11.02.15 AM

  1. Heterogeneity: Lastly is the measure of how similar the studies are. Ideally a meta analysis combines a number of randomized control trials that are all done on the SAME type of patient with the same outcome. But remember studies aren’t usually done like that (which sounds crazy, I know). They are done the way the researcher has the ability to get them done. So, for example in this study although we compare spacers to nebs did the researchers really get only COPD patients? Did they look at the same primary outcome (e.g. FEV or Dyspnea score)? This could make the trials fundamentally different. Therefore, we need to have some way to say yes all these studies are very similar (homogenous) or no these studies are very different (heterogenous). So for once we have a stat that makes sense! The “heterogeneity score”! Also referred to as I2, a high heterogeneity score means the studies are very different (and maybe can’t be compared together) and a low heterogeneity score means. A heterogeneity score >50% is not great. The example study is 47% (not great).

Screen Shot 2018-06-27 at 10.58.28 AM


  1. van Geffen, W. H., Douma, W. R., Slebos, D. J. & Kerstjens, H. A. Bronchodilators delivered by nebuliser versus pMDI with spacer or DPI for exacerbations of COPD. Cochrane Database Syst Rev CD011826 (2016).