The Written Summary of the EM:RAP Monthly Audio
Transcription
The Written Summary of the EM:RAP Monthly Audio
The Written Summary of the EM:RAP Monthly Audio Program January 2014: Volume 14, Issue 1 Editor-in-Chief: Mel Herbert Executive Editor: Paul Jhun Associate Editor: Marlowe Majoewsky, MD www.emrap.org Accidental Hypothermia Part 1 The duration of CPR does not predict outcome, and a long transport time should not dissuade you from transferring. Mel Herbert MD and Doug Brown MD u Brown DJ et al. Accidental hypothermia. N Engl J Med. 2012 Nov 15;367(20):1930-8 PMID: 23150960. Case #1 Case presentation. A 29 year old orthopedic surgery resident was skiing with friends in Norway. She fell into a creek and became trapped upside down. Water flowed all around her but she had an air pocket in front of her face. Her friends were unable to extricate her so they watched her while they waited for help. After 40 minutes, she lost consciousness. After 80 minutes, rescuers arrived and extricated her. She was lifeless and in asystole. They started CPR and transported her via helicopter to a tertiary care center. At the time of her arrival, she had been down for 130 minutes and received 90 minutes of CPR. Her potassium was 4.3 mEq/L and her pH was 6.6. She was placed on AV bypass. She had the lowest recorded core temperature of 13.7 degrees Celsius (56.7 degrees Fahrenheit). 10 minutes after bypass was initiated, she went into ventricular fibrillation. At 15 minutes, she had a pulsatile rhythm. After 180 minutes, she was rewarmed. 24 hours later, she went into florid pulmonary edema and remained on extracorporeal membrane oxygenation (ECMO) for 5 days. She spent a month in the ICU with the usual complications: renal failure and ischemic colitis. u What is hypothermia? A core temp of less than 35 degrees Celsius (95 degrees Fahrenheit). u The approach to patients with accidental hypothermia can be broken down into two pathways: patients with and without vital signs. Hypothermic patients with no vital signs. This is a 3 step process. #1. Is CPR needed? Do they have obvious signs of irreversible death: decapitated or death before cooling? If there is a possibility that they were cooled (less than 28 to 32 degrees Celsius [82.4 to 89.6 degrees Fahrenheit]) before their heart stopped, they have a chance at recovery and should be transferred to a center capable of performing ECMO or cardiopulmonary bypass (CPB). These patients can survive spectacular durations of CPR and still have a good outcome. #2. Do you need to transfer to an ECMO center? #3. Supportive care while transporting. u Who is cold and salvageable and who is cold and dead? There are 5 criteria you can use to determine if they are dead. 1) A clear history of cardiac arrest prior to cooling. 2) If you have a core temperature greater than 32 degrees Celsius (89.6 degrees Fahrenheit) and they are in asystole, hypothermia is not the cause. 3) If they are frozen solid and the chest is not compressible. 4) They have potassium greater than 12 mEq/L. 5) Special circumstances: trauma, drowning and avalanche burial. q Blunt traumatic arrest has a less than 1% survival and hypothermia will decrease survival even further. q The history of drowning is important. Drowning can be divided into submersion versus immersion. • Submersion is when the patient goes under the water immediately, has a hypoxic cardiac arrest and then cools down. These patients have a dismal prognosis. One caveat is the pediatric patient outlier: there is a report of a 2.5 year old child who fell into a 5 degrees Celsius (41 degrees Fahrenheit) creek and was submerged for 66 minutes. He cooled extremely rapidly and had simultaneous cardiac arrest. He had a core temperature of 19 degrees Celsius (66.2 degrees Fahrenheit) and asystole. He received 2 hours of CPR, was transferred to ECMO and made a full recovery. • Immersion is rare: patients are in the water but the head is above the water and they continue to breathe while they cool down before they eventually arrest. The prognosis here can be good and these patients may survive even after prolonged CPR time. q Selection of patients for ECMO should be judicious; using ECMO in patients that are unlikely to benefit will dilute the resource with cases that don’t do well. EM:RAP Written Summary January 2014: Volume 14, Issue 1 1 The Written Summary of the q u The unconscious, hypothermic patient with vital signs. When should you not worry and when should you transfer to a CPB center? q Things that are not worrisome in the unconscious hypothermic patient with vital signs: bradycardia, mild hypotension and atrial fibrillation. These are normal and to be expected with accidental hypothermia. However, if you see cardiac instability, ventricular arrhythmias, hypotension out of proportion to the degree of cooling (this is difficult to determine sometimes) or very low core temperature; less than 28 degrees Celsius (82.4 degrees Fahrenheit) and especially below 24 degrees Celsius (75.2 degrees Fahrenheit): these patients are high risk for experiencing cardiac arrest. You come across a pulseless, hypothermic patient. q Should you start CPR immediately? Yes. This was controversial in the past due to a concern that the patient might have a pulse that is not easily palpated and CPR might result in asystole. However, the perfusion from CPR is better than whatever limited perfusion they might still have. q Should you warm them, cool them or do nothing during transport? The literature doesn’t give good answers. The position taken in the paper is that you should attempt minimally invasive rewarming. Insulate them. Use warm water bottles or a Bair Hugger if you have it. This won’t rewarm them, but it will prevent further heat loss. q q 2 Most avalanche victims who die, do so as a result of one of three things: severe trauma, the airway is filled with snow and they have a hypoxic arrest, or they have an air pocket and are able to breathe before cooling down and arresting. If the patient has not been buried long enough to get cold, you know hypothermia is not the cause of the cardiac arrest. The magic number is 35 minutes. Patients buried in an avalanche cool at a rate of 10 degrees Celsius per hour. If they are buried greater than 35 minutes, their airway is not packed with snow, and you don’t think that they died from massive trauma immediately: this is the patient to consider for prolonged CPR and transfer to a quaternary care center. Monthly Audio Program If you need to manage the airway, do it. If you need to give fluids, give warm saline at 38 degrees Celsius (100.4 degrees Fahrenheit). In the emergency department, focus on high quality CPR. Advocate for transfer to an ECMO/CPB center. If you are within 6 hours of transfer to an ECMO center and have a patient with a good history, the patient should be transferred, even if you are able to achieve ROSC in your department. There is a good chance the patient will progress to florid pulmonary edema and require those services anyway. q A trial of epinephrine and electricity is reasonable. But if it isn’t working, don’t persist. Give 1-2 doses, wait until the patient is warmed 5-10 degrees Celsius and then try again. q What about chest tube or peritoneal lavage? There is no role for it unless you are not able to transfer the patient to an ECMO/CPB center within 6 hours. q Is there any evidence to support thoracotomy with rewarming of the heart? There is no literature on this. There is literature supporting chest tube lavage with a method of either one or two chest tubes. However, this is a distant second best; the literature is not as robust as that supporting transfer to an ECMO/CPB center. q Hypothermic patients who experience cardiac arrest have a survival rate of 50% when treated in an ECMO center but only 10% in a non-ECMO center. If you have a patient with cardiac instability due to hypothermia, you should package them up and transfer to an ECMO center; they will do much better. Paper Chase 1: A New Chest Pain Decision Rule Sanjay Arora MD and Michael Menchine MD u Hess EP et al. The chest pain choice decision aid: a randomized trial. Circ Cardiovasc Qual Outcomes. 2012 May;5(3):251-9 PMID: 22496116. u Chest pain is a huge problem. We admit patients to obs units and keep them there for 12-24 hours for serial troponins, stress testing or CT angiogram. The yield is low, often in the single percentage rates. It is expensive in terms of time lost, actual cost and likelihood of indeterminate results leading to additional testing. u Hess et al looked at whether providing patients with information about their level of risk changes the disposition. This assumes that we don’t really want to admit these patients for work-up but do so out of a fear of a rare bad outcome. It also assumes that if the patients knew how low their risk was, they would not want admission either. q They randomized 208 low-risk chest pain patients and showed them an interactive card that provides an assessment of risk. They asked the patients what disposition they preferred: admission for urgent stress test, follow-up in 2-3 days for stress testing, follow-up with PCP or the decision be made by the ED physician. q The control group was asked the same series of questions but did not receive a risk assessment. 77% opted for admission to the hospital. In the group receiving the intervention, only 58% opted for admission. This is a decrease of approximately 20%. EM:RAP Written Summary | www.emrap.org The Written Summary of the q u There were no negative outcomes. What this study tells us. q Over half of patients still want to be admitted even knowing they are low risk. q There is a substantial subset of patients who are willing to forgo inpatient risk stratification and follow-up as an outpatient. Patients were more satisfied with the intervention. q Shared decision making can reduce resource utilization. u u The card given to the patients presented their risk in several different ways. “Of every 100 people with factors like yours who came to the emergency department with chest pain, 2 had a heart attack or a pre-heart attack diagnosis within 45 days of the emergency department visit, 98 did not”. It also showed a diagram representation of 100 figures, 98 are colored blue and the 2 patients with a heart attack are identified by red. If the patient is above 32 degrees Celsius (89.6 degrees Fahrenheit), still in asystole and your clinical measures are not working, hypothermia is no longer the cause. u Can you place a central line in these patients? What if you tickle the heart with it? Do the things you need to do to provide the necessary care for the patient. However, if you are going to do a central line, try to place it shallowly. u The hypothermic, unconscious patient who has vital signs can be divided into those with cardiac stability and those with cardiac instability. q Any patient with cardiac instability should be transported to an ECMO/CPB center. This is defined as a systolic blood pressure less than 90, ventricular arrhythmias or a core temperature less than 28 degrees Celsius (82.4 degrees Fahrenheit). There is some literature (mainly case series) indicating that hypothermic patients with cardiac arrest may have improved survival when treated at an ECMO center but it is much less clear for patients who have a pulse but cardiac instability. However, these patients are at high risk for arrest, and once they do, they are better off at an ECMO center. For this reason, transport of patients with cardiac instability to an ECMO/CPB center is recommended. q For all of your unconscious hypothermic patients, you want to handle them carefully. Do minimally invasive rewarming. Expect that they may arrest. q Will rewarming patients cause cardiac arrest? There is a concern that rewarming the patients may cause cold blood to re-enter the circulation and lead to cardiac arrest and death. This is controversial. Review of case reports in the literature divides this into rescue collapse versus after-drop. Rescue collapse is a real phenomenon. Once patients are cold - under 32 degrees Celsius (89.6 degrees Fahrenheit) and especially under 28 degrees (82.4 degrees Fahrenheit) - they are at high risk for a malignant arrhythmia. They could go into a cardiac arrest at any point during your resuscitation. It is not caused by anything that you have done. It is the natural history of accidental hypothermia. There are excellent physiology studies that demonstrate after-drop; it is a real thing but it is not clinically significant. If you take healthy patients, cool them in ice water and then rewarm them, the core temperature will drop between ¼ and ¾ of a degree Celsius. This is not a clinically significant entity. You should not worry about rewarming your patients. We are still early in understanding how best to communicate this to patients. This article is a great first step. It won’t change practice (only 17% of chest pain patients were eligible for it), but it does show that if you can communicate this to patients, they can understand some of it and make decisions. Accidental Hypothermia Part 2 Mel Herbert MD interviews Doug Brown MD u Monthly Audio Program You have a hypothermic patient in a bad, death-defining rhythm. Is there a temperature you need to warm them to before you can consider giving medications? Can you give ACLS drugs? qThe European Resuscitation Council recommends a modified approach to support consisting of up to 3 defibrillations and withholding of epinephrine until the core temperature is higher than 30 degrees Celsius (86 degrees Fahrenheit). If you use epinephrine, the time between doses should be doubled until the temperature reaches 35 degrees Celsius (95 degrees Fahrenheit). q This is in opposition to the American Heart Association recommendations, which say it might be reasonable to consider the administration of vasopressors in cardiac arrest. q Why are these recommendations different? They are based on two different studies, one which shows epinephrine works early on and one which shows it doesn’t work. q A practical approach. When beginning your ACLS algorithm, give them one dose of epinephrine and one trial of electricity (if you see a shockable rhythm). If it doesn’t work, don’t persist with repeated doses but wait until they warm up 5 to 10 degrees Celsius and then try one more dose. Then wait until they are above 32 degrees Celsius (89.6 degrees Fahrenheit). u If you have a patient in hypothermic cardiac arrest with no chance of transfer or ECMO, what should you do? Perform high quality CPR. Prevent further heat loss by insulating the patient. Place a Bair Hugger and warm up the room. Do a trial of epinephrine and electricity (if indicated). Then insert bladder lavage. If you have the capability to insert peritoneal lavage and thoracic lavage, do those as well. Note: this is only for the January 2014: Volume 14, Issue 1 | www.emrap.org 3 The Written Summary of the patient that you can’t get to ECMO. Rewarm the patient with continued focus on high quality CPR. Watch their core temperature; with every 5 degrees Celsius of increase, recheck for a shockable rhythm and consider another dose of epinephrine. q u You have a hypothermic patient who is still alive. What are the triggers for aggressive rewarming? For unconscious patients with hypothermia, there is no indication for invasive rewarming. Place the patient in a Bair Hugger. Handle them carefully to try to avoid triggering a malignant arrhythmia. Bladder lavage has essentially no morbidity and is reasonable. If you need to give them fluids, give them warm saline at 38 degrees Celsius (100.4 degrees Fahrenheit). Use normal supportive care. q u If the patient has cardiac instability (ventricular arrhythmia, significant hypotension), there is evidence that the patient may benefit from invasive rewarming. Ideally this would occur in an ECMO/CPB center. There is no evidence saying whether or not to perform chest tube or peritoneal lavage. The patient has a pulse but is very hypotensive. Should you use vasopressors? Hypothermia causes some of the most profound vasoconstriction and myocardial irritability of any medical condition. Early in your resuscitation, there is no role for vasopressors, as patients are as tight as they can get. You will just increase their myocardial irritability. However, after you rewarm the patient and the patient continues to experience hypotension with an element of vasodilation, the risk/benefit may change and you may consider using vasopressors. Remember that hypothermic patients are very irritable and vasopressors have the potential to increase irritability, causing a perfusing rhythm to progress to cardiac arrest. You need to be especially careful if the patient has frostbite. Vasopressors may worsen frostbite. u Shivering is the best way for patients to rewarm themselves. Current research is examining strategies for pain medication in trauma patients that does not inhibit the shivering mechanism. u Hypothermic patients will become very volume depleted due to vasoconstriction and cold diuresis. They will require significant volume during rewarming to replace volume loss and keep up with vasodilation. Expect 2-5 liters. Use of pressors during rewarming should have the caveat that you have filled the tank or need a temporizing measure while you catch up with the fluids. u 4 Eventually you will reach 32 degrees Celsius (89.6 degrees Fahrenheit). Once you are above 32 degrees Celsius (89.6 degrees Fahrenheit) give it one more try. If they are dead, then stop. If the patient is rewarmed, rehydrated and in sinus rhythm, but still requiring pressors, this is a red flag that something else is going on such as sepsis or tox. Monthly Audio Program u For more information on this, go to: drdougbrown.ca. The review paper is available online and has an algorithm detailing management of hypothermia. u The International Hypothermia Registry: https://www. hypothermia-registry.org The NNT: Aspirin for STEMI and IV PPIs for Upper GI Bleed David Newman MD u What is the number of patients needed to treat with aspirin to save a life in patients with STEMI? There is one major randomized controlled trial. u Randomised trial of intravenous streptokinase, oral aspirin, both or neither among 17,187 cases of suspected acute myocardial infarction: ISIS-2. Lancet, 1988 Aug 13;2(8607):349-60 PMID: 2899772. q This study was published in 1998. 17,187 patients with MI were randomized to receive either streptokinase, aspirin, streptokinase and aspirin, or two placebos. q This was a high quality study. It found that the aspirin groups, compared to the placebo groups, had mortality of 9.4% compared to 11.8%. This is a 2.4% difference. What is the number need to treat? 42. We need to treat 42 patients with MI with aspirin for one person to survive because of the aspirin. 97.6% of patients were either going to die or live and aspirin wasn’t the swing vote; however, for 2.4%, the aspirin made the difference. This is a clinically significant benefit. If you take a million MIs over several years in the United States, 24,000 or 2.4% of them survive because of the aspirin. This is major. q This is worth doing especially if you compare it to the number needed to harm (167 patients). For every 167 patients you treat with aspirin, you will cause one minor bleeding event. Aspirin did not increase intracranial hemorrhages or transfusion-requiring hemorrhages. u What is the evidence supporting proton pump inhibitors (PPIs) for GI bleeds? Consultants seem to love PPIs for GI bleeds; “Can you just give them a bolus before they go upstairs?” Why do they love it? It is supposed to increase the pH and make it more likely you can control the bleeding and less likely they will experience rebleeds and death. u There are two Cochrane reviews. One is relevant to emergency medicine and the other isn’t. u Dorward S et al. Proton pump inhibitor treatment initiated prior to endoscopic diagnosis in upper gastrointestinal bleeding. Cochrane Database Syst Rev. 2006 Oct 18;(4):CD005415 PMID: 17054257. This is relevant to emergency medicine as it looks at PPIs initiated before the patient is taken to endoscopy. There are 4 randomized trials including 1500 patients. The three effects evaluated were EM:RAP Written Summary | www.emrap.org The Written Summary of the Monthly Audio Program mortality, rebleeding and going to the operating room (OR). For those three effects, there was no benefit with PPI treatment prior to endoscopy. No patients benefited from PPI therapy. q q Patients with GI bleeds can be sick patients; they often have multiple lines and are receiving transfusions, fluids, platelets, antibiotics, sedatives for intubation. If you use a line for a drug that isn’t proven to help a patient, you are likely doing more harm than good. only is there no benefit to giving PPIs if you are in Western industrialized country but there may be some harm. However, this will take some time to become well known by our consultants. u Check out the NNT website at www.thennt.com for links to the reviews and additional info. u There are guidelines that support this: National Institute of Health and Clinical Excellence in the United Kingdom 2012 guidelines for upper GI bleed say do not give PPIs before endoscopy in the acute non-variceal upper GI bleed. They do recommend PPIs if an ulcer is seen. u If you are considering stopping the use of PPIs for upper GI bleeds in the emergency department, read the Cochrane review first. The findings were not all negative; the authors did find some benefit with the use of PPIs such as a significant reduction in the need for intervention with endoscopic hemostatic therapy at the time of initial endoscopy. This may reduce costs. There was no obvious harm to the PPIs. second study wasn’t directly relevant to the emergency department but still interesting. Leontiadis GI et al. Proton pump inhibitor treatment for acute peptic ulcer bleeding. Cochrane Database Syst Rev. 2006 Jan 25;(1):CD002094 PMID: 15266462. uThe q q q If you wait to give PPIs until the patient goes to the endoscopy suite and has confirmed peptic ulcer disease, is there a benefit? The review included 24 randomized controlled trials with over 4000 patients. They didn’t find a decrease in mortality or decrease need for transfusion. They did find that it seemed to decrease rebleeding by about 6.6% and need for the OR by about 3%. About 1 in every 15 patients avoided a rebleed and 1 in every 30 people avoided going to the OR because they had gotten PPIs compared to a placebo or no treatment. This was in patients who had gone to the endoscopy suite and had confirmed peptic ulcer disease. The data suggests a benefit for patients with peptic ulcer disease. If there is no harm for patients without peptic ulcer disease, should we give PPIs (because approximately 40-50% of patients with GI bleed will have peptic ulcer disease)? If you look at the Cochrane review, they did a number of different analyses of the data and one was geographical. Why did they do a geographical analysis? A number of studies were done in Japan/China and others were done in Western countries. These had very different results. When averaged together, it seemed to show benefit for reducing rebleeding and going to the OR. If you were in China and suffering from a peptic ulcer bleed, there was a benefit in mortality as well as a substantial benefit in rebleeding and avoiding the OR. What happened in the other group? Was there harm from PPIs? If you were in a study from a Western country and received a PPI for acute peptic ulcer bleeding, you were more likely to die. There was not a single benefit for any outcome. q Why is there a geographical difference? Nobody knows. q This tells us in the Emergency Department that not Paper Chase 2: Head Trauma in the Anticoagulated Patient Sanjay Arora MD and Michael Menchine MD u The patient with blunt head injury and on warfarin who now feels great is a very troubling clinical scenario. Should you scan them? Should you get two CT scans? Admit them for observation? Keep them in the ED for observation? We know that there is the possibility of delayed bleed. u Menditto VG et al. Management of minor head injury in patients receiving oral anticoagulant therapy: a prospective study of a 24-hour observation protocol. Ann Emerg Med. 2012 Jun;59(6):451-5 PMID: 22244878. q This study only included patients on warfarin and did not study other anticoagulation agents. q In 2002, a neurosurgical guideline recommended that patients on warfarin with a head bonk should be observed for 24 hours. This study looked at this protocol. q This is not a randomized controlled trial. It is a prospective case series. q They included 97 patients with minor head injury who had been on warfarin for at least a week. They excluded patients with a positive initial head CT. 16% of patients had a positive head CT. They included patients with a GCS of 14 or 15. q The patients received a CT scan initially and then were placed in an observation unit where their neurologic status was monitored. The head CT was repeated in 24 hours. 10 patients refused repeat CT head. q They found 5 (or 6%) of these 87 patients, who had January 2014: Volume 14, Issue 1 | www.emrap.org 5 The Written Summary of the an initially negative head CT and a GCS of 14 or 15, had findings on repeat head CT. Only 1 of these had any neurologic changes. What happened to these five patients? 2 were discharged after additional observation. 3 were admitted and 1 had a craniotomy. Of the patients who went home after two negative head CTs and a period of observation, two patients later returned with symptomatic subdural hematomas. One returned 2 days later and one returned 8 days later. q u Nishijima DK et al. Immediate and delayed traumatic intracranial hemorrhage in patients with head trauma and preinjury warfarin or clopidogrel use. Ann Emerg Med. 2012 Jun;59(6):460-8 PMID: 22626015. q This was a prospective observational study conducted in 6 different hospitals. They looked at clopidogrel as well as warfarin. q The study included adult patients on warfarin or clopidogrel. The patients had to be taking these medications within the last 7 days. They looked at the incidence of acute and delayed bleeds. The patients were followed out to two weeks via either medical records, phone interviews or if they were unable to contact the patient, they looked at the Social Security Death Index. q 1,064 patients were included in this study. About 70% were on warfarin and 30% were on clopidogrel. About 10% of the patients did not get an initial CT scan. This suggests that they included patients who were so low risk that they didn’t even scan some of them. q What was the prevalence of a bleed on the initial study? About 12% in the clopidogrel group and only 5% in the warfarin group. These numbers are much lower than traditional estimates. q They found the incidence of delayed bleed was 0% in the clopidogrel group and 0.6% in the warfarin group. Of the patients with delayed bleeds on warfarin, all had a GCS of 15 and 50% died. u 6 What can we conclude from this study? Delayed bleed is a real entity. What is the frequency and what do we do with these patients? They found a delayed bleed frequency of 6% and another 2% returned with bleeds even after the 24 hour observation period and two negative head CTs. Observing people who are feeling fine for 24 hours and then repeating a head CT is impractical for most of us. Patients with an INR greater than 3 and those who were elderly were at higher risk of delayed bleed. Take home points. Immediate intracranial hemorrhage in anticoagulated patients is real and happens 10-15% of the time. Any patient who is coming in with head trauma and is taking warfarin or clopidogrel should get an initial CT scan. Treat a patient on clopidogrel just as you would a patient on warfarin. The highest risk patients are those with high INRs and the elderly. Monthly Audio Program q Do we observe the patient or repeat the CT scan? It is still up in the air. Observation, a repeat CT head or sending them home with good return precautions are all reasonable options. What you do will depend on what the patient in front of you looks like and what their support system is. At the very minimum, the patient should be going home with someone reliable who can watch them and bring them back for a delayed scan. However, an observation period of 24 hours for every patient with a GCS of 15 in our overcrowded system is an impractical solution. You are going to have to read the literature and come up with your own conclusion of what to do. If the patient has risk factors such as a high INR or elderly, it may be worth admitting them for observation and a repeat CT. q Regardless of what strategy you use, remember that there is still the possibility of a delayed bleed several days later. You have to tell the family to watch for subtle changes. The Lin Sessions: PO vs IV Clindamycin, the Dirty Epi Drip and Propofol for Migraines Michelle Lin MD and Zlatan Coralic PharmD Thoughts on vancomycin in the ED. u There has been a trend where healthy patients with cellulitis are given 1 gram of vancomycin in the ED and then are sent home on oral antibiotics. Is this actually efficacious? There is no literature showing benefit to giving one dose of vancomycin and then switching to oral antibiotics. The IDSA guidelines do not recommend this approach. q u What are some downsides to this approach? You are committing the patient to staying in the ED for another 1-1.5 hours or longer if there is an adverse reaction. It increases the cost of the patient’s stay. It does not make sense from a pharmacokinetic standpoint. Recent evidence shows that ability of vancomycin to kill MRSA depends on the concentration of the drug in the blood compared to the minimum inhibitory concentration (MIC) of the bacteria; the magic ratio is 400. It is almost impossible to achieve this level with just one dose of vancomycin. The average patient would need to receive one gram of vancomycin every 8 to 12 hours to achieve these levels. You will be increasing the potential for antibiotic resistance by using just one dose. There is the tendency to give a loading dose of something via IV in the emergency department and then send the patient home on oral antibiotics. If you look at the antibiotics used to treat skin and soft tissue infections, they are almost 100% absorbed from the GI tract. Clindamycin is about 90%. Sulfamethoxazole-trimethoprim (Bactrim) is 100% absorbed. Doxycycline is 100% absorbed. EM:RAP Written Summary | www.emrap.org The Written Summary of the Linezolid (which is not used as a first line agent) is also 100% absorbed. There is no need to load someone on IV antibiotics; this has never been proven to be efficacious. Giving them an oral dose and discharging them home is likely the best practice. u There are a variety of presentations of soft tissue infections in patients. There is no benefit to IV antibiotics in the patient who is healthy and you know as soon as you see them that they will be going home with oral antibiotics. If you know you are going to admit the patient (they are septic, toxic appearing, meet SIRS criteria, etc.), these patients need IV antibiotics. q Deciding whether or not to give IV antibiotics to patients in the gray zone is more challenging. For example, the patient who is kind of sick looking but you may be able to turn them around and send them home with supportive care. What should you do for these patients? Start the IV. Give them fluids and supportive care. Give them the first dose of oral antibiotics. If they improve within an hour, you can send them home on PO antibiotics. If they aren’t improving, you can start them on IV antibiotics and admit them. q What about patients who are admitted to the observation units? Should these patients receive IV antibiotics or PO antibiotics? Your intent is to admit these patients initially so it is not unreasonable to start IV antibiotics in these patients and then discharge them home if they are improving in 24 hours. Monthly Audio Program u The “dirty epi drip” is a quick and dirty way to make an epinephrine drip bedside for anaphylactic shock. u How do you do this? Take your epinephrine that you have available. It does not matter if it is 1:1,000 or 1:10,000 as long as it is 1mg and inject it into a liter bag of saline. This can be the code cart epinephrine which is 1mg in 10cc or the IM epinephrine which is 1mg in 1cc. This gives you a concentration of epinephrine in the bag of 1mcg/mL. Hook this up to the patient and run it wide open until the hemodynamics stabilize. Others advocate starting at the 1mcg/min dose and titrating up. This is not perfect, but it will buy you some time until your team is able to set up the pump and follow hospital protocols. This drug has been tested at such a dilute concentration and found to be stable up to 24 hours. u The maximal rate of infusion will vary with a few things: how high the IV bag is, the length of the tubing, and the size of the catheter. With an open wide 18 gauge, the patient will receive between 20 and 50 mcg/min (or between 20 and 50 mL/min). This is not something you walk away from; you need to stay bedside with the patient with your hand on the roller clamp. If your patient starts responding, you can decrease the dirty epi drip by clamping down the line. If they aren’t doing well you can squeeze the bag and give them more. Case #2 Case #2. You are working at 3am in a low resource area. You receive a patient who is an obese female of 47 years. She presents to your triage area with shortness of breath and difficulty speaking. She has an allergy to peanuts and ate a candy bar prior to arrival. She has a history of hypertension and intubation for a prior episode. She becomes more obtunded with a heart rate of 130, blood pressure of 65/nothing and an oxygen saturation of 89% on O2. Anaphylactic shock is on the top of your differential. The nurse confirms administration of two separate shots of intramuscular epinephrine. The patient responds slightly but remains hypotensive and wheezy. u u Should you continue to give IM epinephrine or switch to an IV drip? If you are giving the second IM dose of epinephrine, you should also be ordering an IV epinephrine drip. The guidelines say that IM epinephrine should be given to stable patients exhibiting signs of anaphylaxis. The guidelines recommend IV epinephrine in patients where the anaphylaxis appears to be severe with immediate life-threatening manifestations. What is the starting dose? Between 1 and 4 mcg/min. You can start at 1mcg/min and titrate up. Should you give the patient push dose epinephrine? The guidelines suggest an IV push dose of 0.1mg of 1:10,000 over 5 minutes. This is a high risk drug and errors occur frequently. There are multiple case reports of errors in IV administration leading to death. There are no standardized guidelines for how to mix this drug or which units to use. Once you dilute the epinephrine, the drug is not stable for very long. It is unlikely that most EDs will have premixed epinephrine drips ready to use. Propofol for migraines. u This is very controversial. This practice started in 2000 with an open label trial where propofol was given to patient with migraines at a migraine clinic. They found that propofol, when given to patients at procedural sedation doses, terminates migraines for most of the patients. u Krusz JC et al. Intravenous propofol: unique effectiveness in treating intractable migraine. Headache. 2000 Mar;40(3):224-30 PMID: 10759925. q This trial included 77 patients with a pain score greater than 7 out of 10 using a visual analog score. This was an open label trial. They gave them 20-30mg IV push doses every 5 minutes for up to an hour. The average reduction in severity was 95.4% at 30 minutes which is the most efficacious treatment so far for intractable migraines. 3 of the 77 patients returned with subsequent headache but most had a sustained response. The average dose of propofol was 110mg similar to procedure sedation. They did not report any adverse events. January 2014: Volume 14, Issue 1 | www.emrap.org 7 The Written Summary of the Monthly Audio Program u In 1995, the NINDS trial came out and was important in getting tPA approved for stroke. It was only approved in patients with symptoms with onset of less than 3 hours prior. Most of the benefit seemed to occur in patients who had their stroke within 90 minutes prior to receiving drug. However, most patients do not present within this window. u In 2003, the ECASS 3 trial showed that tPA was effective up to 4.5 hours after presentation. This is now a fairly widely adopted recommendation. u This was a huge study. It included 3,035 patients in 156 hospitals in 12 countries. This study included people who were able to get tPA versus nothing up to 6 hours after symptom onset. How did this work? If the patient showed up within 3 hours and was a slam dunk for tPA, they gave tPA and didn’t enroll the patient in the trial. If the patient wasn’t a great candidate but not a terrible candidate and the doctor thought the patient might benefit from tPA (for example, the 85 year old presenting within 2 hours after symptom onset or the healthy 47 year old presenting within 5 hours after symptom onset), they were enrolled. This is exactly the type of patient that presents a clinical conundrum in the ED. u What did this mean for patient selection in the trial? They were old. 50% were older than 80 years old. They tended to have bigger strokes or were outside of the 3 hours window. u This was open label; patients were given tPA or nothing. There was no placebo. Patients and their families were aware that they were given tPA. u The primary outcome was alive and well at 6 months. This was assessed using the Oxford Handicap Score which basically asked if they had significant disability (bad outcome) or if they were ok and able to attend to their basic activities of daily living (good outcome). u The Paper Chase 3: TPA up to 6 hours? Not! What did they find? At 6 months, 37% of the patients in the tPA branch had a good outcome versus 35% of patients in the control group. This was insignificant both statistically and clinically. There were bad outcomes overall; approximately 2/3 did poorly. u What did they conclude? “Despite the early hazards, thrombolysis within 6 h improved functional outcome.” u IST-3 collaborative group et al. The benefits and harms of intravenous thrombolysis with recombinant tissue plasminogen activator within 6 h of acute ischaemic stroke (the third international stroke trial [IST-3]): a randomised controlled trial. Lancet. 2012 Jun 23;379(9834):2352-63 PMID: 22632908. u They stated their primary outcome and the primary outcome was clearly negative. Then they did some fancy math… They did ordinal rank logistical regression to say that some more people moved from severely handicapped to less handicapped due to tPA use. This is difficult to understand and determine a number needed to treat. u This is an important article to understand. The conclusions were misleading and if you rely on the abstract, you may be led astray. u There was early harm with tPA. 11% in the tPA group died compared with only 7% in the control group. Intracranial hemorrhage within 7 days occurred in 7% of the tPA group and 1% of the control group. More patients died initially in u Soleimanpour H et al. Effectiveness of intravenous dexamethasone versus propofol for pain relief in the migraine headache: a prospective double blind randomized clinical trial. BMC Neurol. 2012 Sep 29;12:114 PMID: 23020264. q They included 90 patients who presented to the ED with migraines. The patients were divided into two groups: the first group received propofol every 5 to 10 minutes with a maximum of 80mg and the control group received dexamethasone at 0.15 mg/kg. They found similar results to the previous study. q u Most patients had resolution of their migraine symptoms. 2 patients desaturated down to 89% but this resolved after administration of oxygen. Sheridan DC et al. Low-dose propofol for the abortive treatment of pediatric migraine in the emergency department. Pediatr Emerg Care. 2012 Dec;28(12):1293-6 PMID: 23187986. q This was a case control trial with 7 pediatric patients with migraines who received low doses of propofol and compared these to patients who received the usual abortive therapy which was prochlorperazine (Compazine), nonsteroidal anti-inflammatory (Toradol), and diphenhydramine (Benadryl). Propofol was significantly better at decreasing the pain: from 61% to 80%. They had no adverse events. u This is a pretty amazing thing to have in the ED. The standard treatments such as triptans, prochlorperazine or opiates have side effects. Now we have an intervention that can completely obliterate a migraine in 20-30 minutes. However, how many resources does using propofol require? Do you have to do this as procedural sedation with end-tidal CO2 monitoring? This is difficult to do in a single coverage ED. Will this make patients present repeatedly for treatment with propofol? We need more research before this is ready for primetime but it is something to consider in the patient with refractory migraine. Sanjay Arora MD and Michael Menchine MD 8 EM:RAP Written Summary | www.emrap.org The Written Summary of the Monthly Audio Program founded fear as we have always been told that hypothermia equals bleeding. Subarachnoid hemorrhage is not an uncommon cause of cardiac arrest. The literature suggests that not only have you not made things worse, you have probably made things better. the tPA group but by six months it seemed to have evened out and was equivalent to the control group. u There is no evidence that tPA is better than placebo out to 6 hours. Although they claim improved functional benefit, the data does not support it. u If you have a neurologist available, does this mean you now have to call them for patients presenting within a 6 hour window? No. The neurologists seem to like to push this drug. However, the data continue to show that that the benefit is at best modest and when used outside of the window, you kill people and you kill them early. The best data available for tPA outside of 4.5 hours says that there is no difference in functional outcome at 6 months. u However, we do things a little differently if we discover a subarachnoid hemorrhage. You probably have a goal of 32-34 degrees Celsius (89.6 to 93.2 degrees Fahrenheit) in the average post arrest patient. If the neurosurgeon wants to keep the patient at that temperature, that’s fine. Most would let the patient’s temperature drift up to 35 degrees Celsius (95 degrees Fahrenheit). You do not want to rewarm these patients quickly; aim for an increase of ¼ degree per hour: an increase from 33 to 35 degrees Celsius (91.4 to 95 degees Fahrenheit) over 8 hours. At 35 degrees Celsius (95 degrees Fahrenheit), they probably won’t experience an effect on their clotting factors but they will get the benefits from mild hypothermia, such as decreased cerebral metabolism and increased intracranial pressure. u Is the evidence strong for this? There is evidence that induced hypothermia; between 32-35 degrees Celsius (89.6 - 95 degrees Fahrenheit) is good treatment for intractable increased ICP. There is no data showing patient outcome effects from the hypothermia but there is evidence showing that it isn’t harmful. It is similar to giving mannitol or hypertonic; the numbers improve but there is no evidence saying that it will benefit mortality or neurologic outcome. Critical Care Mailbag Scott Weingart MD and Rob Orman MD Who to get warm, who to get cold and who is still a mystery? Hypothermia in Trauma u We cool down most of our post arrest patients to save the brain. The hypothermic trauma patient often is intubated with head injuries and some sort of thoracoabdominal trauma. Is there any evidence to say that mild hypothermia in these patients is a good or a bad thing? u #1. Bleeding. Once you start getting below 35 degrees Celsius (95 degrees Fahrenheit), you begin to experience some mild platelet effects. Between 32 and 35 degrees Celsius (89.6 and 95 degrees Fahrenheit), there is an anticoagulant effect and clotting factors stop working as well. Things become horrible for trauma patients at 32 degrees Celsius (89.6 degrees Fahrenheit). Your goal should be a temperature above 35 degrees Celsius (95 degrees Fahrenheit) the entire time. You should be very afraid if the patient is at 32 degrees Celsius (89.6 degrees Fahrenheit). u #2. Is it good for the brain? Maybe. We don’t know if it is good for the body yet. This is still up for debate in the post arrest hypothermia patients. You should not be cooling down your multi-trauma patients because they will bleed. If they are cold, warm them up. How to induce hypothermia in a remote setting u Strategies for cooling patients in EDs with limited resources. Take 8 bags of normal saline and stick them in one of your medication fridges (about 40F or 4C). Talk to the nurse manager about checking the expiration dates every so often. Now you have the means to cool a patient ready to go. When a post arrest patient comes in, give them 2 liters of the cold saline. Place ice packs (you can fill lab bags with ice from the ice machine) in their groin and axilla. Is this perfect? No, but it will do the job until you are able to transfer them. u Continue the resuscitation as far as you can go. If you can put an arterial line in, great. If not, don’t worry. Try to keep their MAP above 65 and aim for 80. It is ok if you use a pressor to get there. Keep their oxygen status in the middle ground; hypoxia is bad and so is hyperoxia. Check their labs. Try to figure out the cause of cardiac arrest. If it is something you can intervene on, do it. If not, you will be transferring the patient regardless. u Small hospitals should be cooling these patients but not keeping them. If these things seem daunting in your ED, your ICU probably will be unable to provide the same level of care. Cooling is a small component of post cardiac arrest care. It is important and necessary but the patients should be transferred to a cardiac arrest center. Hypothermia in isolated head injury u The patient has isolated traumatic brain injury. Are these patients better off at 35 degrees Celsius (95 degrees Fahrenheit)? We don’t know for sure, but their intracranial pressures (ICPs) might be less. Hypothermia in subarachnoid hemorrhage u You cool down a post cardiac arrest patient and then discover that they have a nontraumatic subarachnoid hemorrhage. Have you done the patient a disservice? No. Cooling the patient is the right thing to do. This is a well- January 2014: Volume 14, Issue 1 | www.emrap.org 9 The Written Summary of the u u You are unable to transfer the patient for 4 to 5 hours. You should have some means for core temperature monitoring. The best way to do this is an esophageal/rectal probe. If you can place an NG tube, you can place an esophageal probe. Try to position it several inches above the xiphoid process. This hooks into regular monitors using a cartridge (both the cartridge and the probes are cheap so it is worthwhile to have them on hand). If you don’t have this in the ED, your anesthesiologist should have it. If you have nothing, do a rectal probe intermittently to check the temperature. Can you overshoot? If you are going too low, take off the ice packs. Chances are it will be more difficult to achieve the desired temperature than it will be to overshoot. You don’t want them too far below 32 degrees Celsius (89.6 degrees Fahrenheit). u Shivering is your enemy. Make sure they are adequately sedated. If they are still shivering, just paralyze them. u Why is there an emphasis on placing arterial lines? It is not because it provides much additional benefit in terms of the resuscitation but because it is a marker of the ability to do post-arrest care at an ideal level. Bag valve mask versus using the ventilator in cardiac arrest u For an intubated patient who arrests, should you disconnect the ventilator and bag manually when doing chest compressions? The safest way is to let respiratory therapy or a colleague bag the patient if they are calm and know what they are doing but there is the risk that the person may bag the patient rapidly (for example, 60 beats per minute). This is really bad for the patient’s outcome. However, the ventilator is not a panacea either. Most ventilators have a set upper pressure limit (usually between 40 and 60) so normal patients won’t get an injurious level of ventilation. The chest compressions will be read by the ventilator as increased pressure in the chest and the ventilator will not deliver a breath. The ventilator doesn’t know if the increased pressure is in the alveoli or external to them. If the patient is receiving chest compressions at 100 beats per minute, there is no time for the ventilator to deliver a breath between those chest compressions. The ideal breath happens at the downstroke of the chest compression. you want to use the ventilator, if you understand the settings, and if you are able to increase the peak pressure limit to about 100, the ventilator will now deliver breaths regardless of whether or not the patient is getting a chest compression. Nothing bad will happen to the alveoli as they are not actually experiencing that pressure. However, if you don’t know how to adjust the settings, the ventilator can be incredibly dangerous for these patients because they won’t get any ventilation. Monthly Audio Program to the tube. Rate of 12 breaths a minute with a FiO2 of 100%. PEEP of zero. If you know what it means, put them on a flow rate of 30L/minute. This will deliver the 500cc over 1 second. Increase the peak pressure to 100. Ideal tidal volume for patients intubated in the Emergency Department u q The days of using 12-15 cc/kg are over. This is a bad idea. You should be starting at 8ml/kg of predicted body weight. If they have any signs that this is too much, go down to 6ml/kg. What is predicted body weight? If you have a patient that weighs 600lbs and is 5 foot 1 inch, they get the tidal volume you predict based on the lung size for someone who is 5 foot 1 inch. There are charts online for this: ardsnet.org/node/77460 q This is one of the few trials that showed mortality benefit. More recent data shows that even healthy patients will be less likely to develop acute lung injury if low tidal volumes are used. The data is not definitive but there is no good reason to use volumes of greater than 8ml/kg, even in your healthy patients. Paper Chase 4: DKA and Cerebral Edema Sanjay Arora MD and Michael Menchine MD u Cerebral edema in kids with DKA is real, but rare. The youngest children are at highest risk. Clinically apparent cases occur in 0.5-1%. However, there may be more cases that go unrecognized. u For those with clinically apparent cerebral edema, the outcomes are devastating; up to 50% of these patients will die. Of those that survive, 1/3 will be in a vegetative state. There has been an association between severity of DKA and intensity of treatment with development of cerebral edema. u Are fluids the culprit? Nobody knows. Early trials showed a relationship with IV fluids but there were no controls and multiple confounders. These were very sick kids who were more likely to get the most fluid. There have been more than 20 retrospective studies on this but no prospective trials until now. Both animal and human studies have shown cerebral edema on MRI prior to treatment. u The largest retrospective trial was published in 2001. Glaser N et al. Risk factors for cerebral edema in children with diabetic ketoacidosis. The Pediatric Emergency Medicine Collaborative Research Committee of the qIf Ideal ventilator settings in cardiac arrest u 10 Put the ventilator on assist control at 550cc tidal volume. This will deliver 500cc to the patient and 50cc is lost Ventilation with lower tidal volumes as compared with traditional tidal volumes for acute lung injury and the acute respiratory distress syndrome. The Acute Respiratory Distress Syndrome Network. N Engl J Med. 2000 May 4;342(18):1301-8 PMID: 10793162. EM:RAP Written Summary | www.emrap.org The Written Summary of the American Academy of Pediatrics. N Engl J Med. 2001 Jan 25;344(4):264-9 PMID: 11172153. q u This is a huge study. They included 61 children with cerebral edema with 6 controls per patient. They found no relationship between the amount or rate of IV fluids and the development of cerebral edema. Despite this data, some important people (like the AAP) say that if you are going to give children with DKA more than 10-20 cc/kg of IV fluid, you need to involve a pediatric endocrinologist. u What is the pathophysiology? No one knows. Some believe that osmotic changes result in rapid influx of fluid into brain cells. However, others believe that it is caused by cerebral hypoperfusion so you should get the fluids in fast. u Glaser NS et al. Subclinical cerebral edema in children with diabetic ketoacidosis randomized to 2 different rehydration protocols. Pediatrics. 2013 Jan;131(1):e73-80 PMID: 23230065. q This was a prospective trial including children between the ages of 8 and 18 with DKA. They did not include the children with the highest risk of DKA (less than 5 years). q The patients were randomized to receive Protocol A, a fluid bolus of 20cc/kg initially and then rapid replacement of their estimated fluid deficit (2/3 over the first 24 hours and 1/3 over the next 24 hours) or Protocol B, a fluid bolus of 10cc/kg followed by slow, even replacement of their fluid deficit over the next two days. q Both groups were given insulin at a rate of 0.1units/kg/h with no bolus. They all got frequent neuro checks and labs. Each patient got 3 MRIs: one MRI at 3-6 hours after initiating treatment, one MRI at 9-12 hours and one MRI after treatment was completed at approximately 72 hours. q The trial included 18 patients and two patients experienced DKA twice so there were 20 total episodes. 2 of the cases had to be withdrawn after randomization; one was too claustrophobic for repeat MRI and the other was too sick to have been enrolled in the first place. All of the kids did fine at the end with no neuro deficits and no clinical cerebral edema. q Remember that the likelihood of clinically apparent cerebral edema is only 0.5 to 1% of cases and this was a small number of patients. However, they were looking for subclinical cerebral edema. q Note: One kid in the rapid rehydration group (Protocol A) had a fast deterioration in his mental status. He was treated for cerebral edema with mannitol with no improvement but then spontaneously improved after a few hours. They eventually determined that this wasn’t cerebral edema, but it is something to keep in mind. Monthly Audio Program q They looked at the ADC (apparent diffusion coefficient) and found no difference between the two groups. The ADC reflects the diffusion speed of water molecules and changes are a harbinger of cerebral edema. q However, based on the graphs, there was no real difference in time to resolution of DKA. The kids with rapid fluid administration did not do better. This was not discussed much in the article. u What does this article tell us? In a small sample of kids, there was no difference in terms of cerebral edema. So if you have patient who is sick, shocky, septic and obviously dehydrated or hypotensive, give them fluids. You are not a bad doctor if you do so. However, be careful in what you consider sick, shocky and septic in kids who are under 5 years old because these are more likely to get cerebral edema. u Document the patient’s neurologic exam and mental status prior to initiating treatment. If they are already obtunded or lethargic before getting treatment, this can save you if the kid develops cerebral edema. There is literature to say that cerebral edema is present prior to treatment. u This same group (PECARN) is currently conducting a huge prospective randomized controlled trial: the FLUID trial (Fluid therapies Under Investigation in DKA). This is a 13 center, nationwide study that includes 4 groups with fast and slow rehydration with normal saline and half normal saline. The protocol was published in Pediatric Diabetes earlier this year. The primary outcome will be GCS of less than 14. They will include children less than 8 years old. Mythbusting Part 1: Corneal Abrasions and Topical Anesthetics Anand Swaminathan MD and Mel Herbert MD u Most of what we thought about the management of corneal abrasions is wrong. Patch the eye? No. Antibiotics? We still do them but the evidence isn’t great. Don’t send your patients home with topical Proparacaine? u Corneal abrasions are common. We see them multiple times a day. You put the drops in. They feel better and ask if they can take the drops home. We say no. We know that systemic analgesics are inadequate for this. Our textbooks all say that it is dangerous to give topical anesthetics because they will go on to develop corneal ulceration and impaired healing. u What are the supposed dangers? #1. Physiologic. Topical anesthetics supposedly inhibit mitosis and cellular migration and impair the ability of the corneal epithelium to oxidize glucose and pyruvate leading to decreased healing. #2. It decreases corneal sensation to touch and the abrasion may worsen without the patient knowing it. #3. There is direct toxicity. The agents increase corneal thickness and cause opacification, stromal infiltration or epithelial defects. January 2014: Volume 14, Issue 1 | www.emrap.org 11 The Written Summary of the u u What do our textbooks say? Both say that topical anesthetics are contraindicated in the management of corneal abrasions and the healing will be different due to sensation issues. This is a single line of text with no supporting references. Monthly Audio Program send patients home with. Of interest, they included 4 physicians. u Where does this come from? q Most of the literature comes from animal studies and two main sources: enucleated rat and rabbit eyes or cell preparations. There are number of different studies. Some showed no difference in healing or corneal permeability, others showed worse outcomes. Is an enucleated rat or rabbit eye really equivalent to a living human eye? q The data that exists for humans is a series of case reports. q Epstein DL and Paton D. Keratitis from misuse of corneal anesthetics. N Engl J Med. 1968 Aug 22;279(8):396-9 PMID: 4232838. Included 5 case reports of patients who applied topical anesthetics for prolonged periods of time and developed ulcerations, thickening of the corneal epithelium and hypopyons or other infections. What were these five cases? The first case used 5% tetracaine. Most preparations in the emergency department are only 0.5%. The patient used it every hour for three weeks and developed an ulcer. The second case used it every 20 minutes for two months. The third case was a veterinarian who used veterinary tetracaine in an unknown concentration, every 1-2 hours for two weeks. The fourth case was a pediatrician who used it every 16 minutes for 5 days. The last case was a patient who used it 10 times a day for three months. In these cases the anesthetics were used much more frequently than we are considering (at most 2-3 days). q q q 12 Willis WE and Laibson PR. Corneal complications of topical anesthetic abuse. Can J Ophthalmol. 1970 Jul;5(3):239-43 PMID: 5472832 . Described 9 cases including a doctor who used 0.5% every 15 minutes for seven days, a woman who used 8 bottles of Proparacaine over 6 weeks and a 50 year old who used it every 30 minutes for 9 days. All of the cases are very similar to this. There was one patient who used Proparacaine 0.5% every 2 hours for three days and returned with ulceration. Even though he used it sort of appropriately, he didn’t have a good evaluation of his abrasion prior to initiating anesthesia (he was a factory worker who was seen by the factory doctor in 1969 and started on the regimen prior to evaluation by an ophthalmologist). Duffin RM and Olson RJ. Tetracaine toxicity. Ann Ophthalmol. 1984 Sep; 16(9):836, 838 PMID: 6508100. Included one patient who used 0.5% tetracaine every 30 minutes for more than 2 months. Are there any safety studies? Initial studies were in the ophthalmology literature with patients who received photorefractive keratectomy and were given topical anesthetics agents for between 24 and 72 hours. The healing was evaluated initially and at 72 hours. All three studies agreed there was no difference in corneal healing at 72 hours. They induced a laceration to the cornea rather than abrasion and there was no difference in wound healing. q Verma S et al. A comparative study of the duration and efficacy of tetracaine 1% and bupivacaine 0.75% in controlling pain following photorefractive keratectomy (PRK). Eur J Ophthalmol. 1997 Oct-Dec; 7(4):327-33 PMID: 9457454. q Brilakis HS and Deutsch TA. Topical tetracaine with bandage soft contact lens pain control after photorefractive keratectomy. J Refract Surg. 2000 Jul-Aug;16(4):444-7 PMID: 10939724. q Shahinian L Jr et al. Dilute topical proparacaine for pain relief after photorefractive keratectomy. Ophthalmology. 1997 Aug;104(8):1327-32 PMID: 9261322. u There is one study in the emergency medicine literature. Ball IM et al. Dilute proparacaine for the management of acute corneal injuries in the emergency department. CJEM. 2010 Sep;12(5):389-96 PMID: 20880433. q This was a small study with only 33 patients but it was randomized. 15 patients were given 0.05% proparacaine as a topical anesthetic to take home. This is a 10-fold dilution of what is normally used in the emergency department. The patients were followed up at 72 hours with no adverse events or change in wound healing. The group with topical anesthetics needed less oral medication. u What are the take-home points? Most of the evidence against topical anesthetics for corneal abrasion is from animal studies with enucleated rabbit or rat eyes and case reports of inappropriate use published years ago. There is some ophthalmology literature saying that topical medications are safe after PRK. The best literature in the ED says that there is no increased risk of side effects but this is small. We need a larger study looking at safety but the urban legend saying that corneal anesthetics are unsafe appears to be based on an unsound research base. We give patients medication that causes bleeding and death when they trip and fall; this doesn’t seem so bad. u Both Herbert and Swaminathan have had corneal abrasions. What did they do? Used proparacaine for several days… Of the 15 case reports, 14 patients used it inappropriately. All cases used a solution that was higher than we would EM:RAP Written Summary | www.emrap.org The Written Summary of the Paper Chase 5: Patient Satisfaction and the Telephone Monthly Audio Program u 1 gram of calcium chloride has three times the amount of calcium as calcium gluconate. This has to do with the size of the molecule. There is very little literature on which one we should be using. It has been said for years that calcium gluconate needs to go through the liver for metabolism but there is little data supporting this and no randomized controlled trials. u Brian Hayes Pharm.D discussed the data on Academic Life in EM. u Martin TJ et al. Ionization and hemodynamic effects of calcium chloride and calcium gluconate in the absence of hepatic function. Anesthesiology. 1990 Jul;73(1):62-5 PMID: 2360741. Sanjay Arora MD and Michael Menchine MD u u u Patient satisfaction is important. It is how you get paid. It is how you stay satisfied with your job. It may help the patient avoid clinical deterioration and avoid a follow-up visit. It may be the only way to measure quality in a healthy patient. It might help patients follow your recommendations. It might help keep you out of legal trouble and it is the right thing to do. It is hard to influence patient satisfaction as there are multiple variables beyond your control such as the cleanliness of the ED, the waiting time, the rudeness of the clerk up front, etc. If you can do something easy to improve patient satisfaction, do it. Patel PB et al. Physician e-mail and telephone contact after emergency department visit improves patient satisfaction: a crossover trial. Ann Emerg Med. 2013 Jun;61(6):631-7 PMID: 23465304. q This was a crossover trial done by Kaiser in Northern California. The physicians contacted all of the patients they saw. The physicians spent 2 minutes on average to make the contact. Even if the contact was a phone call, it was still usually only about 3 minutes. patients were then surveyed about their satisfaction. The patients who were contacted were satisfied with their interaction with the physician 88% of the time compared to 79% for the controls. Satisfaction was assessed for the physician’s skills and abilities, care and communication. q This study was performed in patients who were going to the operating room for liver transplant. There was no way they would be able to turn calcium gluconate into ionized calcium if the liver was necessary. The patients received citrate as part of the pre-transplant algorithm. This can cause hypocalcemia. q The patients became hypocalcemic and had serum calcium levels measured. They were given either calcium chloride at 10mg/kg or calcium gluconate at 30mg/kg. q They measured ionized calcium after administration and found that the ionized calcium in both groups increased a similar amount after 10 minutes. There was no difference between the two. They concluded that you don’t need the liver in order to turn calcium gluconate into ionized calcium and it is safe to give either calcium gluconate of chloride. q The q There were some problems with this study; it was at Kaiser, all of the patients had to have contact information, the satisfaction scores were focused on the physician and not the overall experience. Mythbusting Part 2: Calcium Chloride or Gluconate q u u u A recent discussion recommended the use of calcium chloride over calcium gluconate in the treatment of hyperkalemia. Theoretically, it has quicker onset, has more calcium in the same volume and doesn’t require first past metabolism unlike calcium gluconate. However, most use calcium gluconate as they are worried about extravasation and sclerosis of the veins with calcium chloride. A response by Chris Cresswell MD. This is BS and you need to stop saying it. There is no difference in calcium levels regardless of what is given. Cote’ CJ et al. Calcium chloride versus calcium gluconate: comparison of ionization and cardiovascular effects in children and dogs. Anesthesiology. 1987 Apr;66(4):465-70 PMID: 3565811. q Anand Swaminathan MD and Mel Herbert MD This isn’t really applicable to emergency medicine, but it does highlight that you don’t need a liver to activate calcium gluconate. Included 10 pediatric patients with significant burns going to the operating room. These patients had hypocalcemia and were given either calcium chloride or calcium gluconate with the same outcome. The ionized calcium increased in the serum in about the same amount of time regardless of which formulation they were given. This paper was a little weird as they also included 10 dogs. Heining MP et al. Choice of calcium salt. A comparison of the effects of calcium chloride and gluconate on plasma ionized calcium. Anaesthesia. 1984 Nov;39(11):1079-82 PMID: 6507824. u These aren’t the greatest studies but they are probably the best we will get. They both say there is no difference between administering calcium chloride or calcium gluconate. January 2014: Volume 14, Issue 1 | www.emrap.org 13 The Written Summary of the u Which is more effective to give for hyperkalemia? There isn’t much literature on this. Calcium chloride has not been compared directly to calcium gluconate for hyperkalemia. Even the amount of calcium necessary is unknown. Calcium gluconate is definitely superior to calcium chloride in terms of safety through a peripheral line; it is less likely to cause tissue necrosis if it extravasates. If you have central line access in a patient with hyperkalemia, calcium chloride is a reasonable choice at a dose of 1 gram. However, it is reasonable to use calcium gluconate in everyone since we know it works just as quickly with a better safety profile. You can use a dose of 2 grams. BONUS SECTION SUMMARIES Monthly Audio Program q u Child abuse cases can be dichotomized into two separate areas. q 1. Thinking about child abuse in the first place. If you see a child with vomiting and without diarrhea, it is difficult to think that it may be due to a brain injury. This is the hardest thing to do: seeing a kid where no one is telling you they are concerned for abuse, the child doesn’t have an obvious injury and you have to be the one to say, “Well, maybe this is from abuse.” q 2. When the flag has been raised, deciding whether or not abuse has occurred. Involve others. Don’t feel like you have to do everything by yourself. If you are in an academic center that has a pediatrician specializing in child abuse, call them. They will know the literature and can assist you in what tests to order. They will also be able to go to court and talk with other experts. u How extensive should your work-up be? If you are suspecting abuse, the work-up should include more studies than you think. The more injuries you identify, the easier it will be to protect the child in the face of opposition in court. u What injuries should heighten your suspicion of abuse? q Every injury in a child should be considered in the context of the age of the child and the reported mechanism of injury. For example: retinal hemorrhages, rib fractures, subdural hematomas and skull fractures may be consistent with a history of ejection in a motor vehicle accident and are not always due to abuse. It can be difficult to verify a fall. The most concerning time is age less than 3 years and especially less than 6 months old. Neurologic exams are difficult in this age group and they are unable to provide history. Once the child is able to talk to you and tell you what happened, it is easier. q Spiral fractures have historically been considered pathognomonic for abuse, but there is not a lot of literature to support this. One of the most common accidental fractures in children, the Toddler’s fracture, is usually a spiral fracture. q It was previously thought that bruising of different ages is concerning for abuse and the age of the bruise could be determined based upon the color. However, a review of the literature shows that we are often unable to agree on the color of the bruise and there is no good evidence associating the color of the bruise with its age. This theory was based on studies performed decades ago that were poorly done and not performed in children. An Approach to Child Abuse Ilene Claudius MD and Sol Behar MD with Dan Lindberg MD u How do you approach child abuse? It requires a different way of thinking than usual. Why? q q 14 Unlike every other illness you may face in the pediatric patient, the parents aren’t your partner when abuse is involved, even when they are not responsible. Often they are unaware the child has been abused. For example, even though a father may have abused a child, the mother may just think the child is limping. If they are responsible for the abuse, they may give you a story that is very misleading or incomplete. Our normal clinical gestalt is to look for clinically important injury. If you miss a rib fracture in a patient who fell down skiing and it heals by itself, it doesn’t really make much difference. However, if you miss a rib fracture in a child that is due to abuse and allow the child to return home, it dramatically increases the risk of future abuse. Evidence shows that if they are abused once, they are very likely to be abused again if it remains unrecognized. The abuse may be escalating. One of the reasons the cases of child abuse we see are so depressing is because we are often too late for intervention. We are left trying to put someone in jail, which is not a satisfying outcome. The satisfying outcome is when we are able to recognize a mild abusive injury and protect the child. EM:RAP Written Summary | www.emrap.org The Written Summary of the Monthly Audio Program arm. You reported that the patient had fallen off of the bed. This injury is more severe than we usually expect from a fall of this type, so it is important for us to be sure that we are not missing something else. It is important that we don’t miss something medical that makes your child’s bones fragile or some other kind of trauma that you are unaware of. Unfortunately, sometimes we find that someone may have been hurting your child. We are going take pictures of your child’s other bones to make sure that there are no other injuries and do a CT or MRI to make sure that their brain is ok. We will send some blood tests to make sure that there are no signs of injury to their belly.” q Location of bruises is important. Active children are expected to have bruises on their shins, forearms and forehead. Bruising on the abdomen, neck, ears and under the chin is more concerning. Pattern bruising is important. Bruises in children under the age of 6 months or in children who are not cruising are extremely rare in normal children. In children under 6 months with bruising and no other apparent injuries, the rate of fracture, brain injury or abdominal injury is approximately 50%. q Rib fractures are concerning for abuse. This is true in all age groups. u q Fractures in different stages of healing. q q The classic metaphyseal or bucket-handle fracture. These are fractures around the growth plate. Most of these are in infants. q Consider abuse in any case of subdural or subarachnoid hemorrhage outside of major trauma such as a car accident. We need to weigh the risks and benefits of these studies. The ALARA principle of radiation exposure (As Low As is Reasonably Achievable) and new literature describing which children do not need CT scans are important. However, this is different for cases of suspected abuse. Why? We are missing cases of abuse. This is not because we are lazy or not trying; it can be difficult to detect. q The risk from missed abuse is very high. Children with abusive head trauma have a mortality of about 20% and of those who survive, about 60-80% will experience long term neurologic problems. The lifetime disability and risk of death is as high as any other disease we see. We are much more likely to have seen a child die from abuse than meningitis. Even in a newborn, the risk of cancer due to the radiation from a head CT is only 1/500-1/1000 according to the most conservative studies. The risk from abuse is much higher. q The AAP guidelines say if you are concerned about physical abuse in a child less than 2 years old, a skeletal survey is mandatory. This recommendation has the most supporting evidence. There are studies showing the rates of positive findings on skeletal survey in this population are about 20%. If you work at a center that does not do skeletal surveys, you should consider transferring the child to a center that does. q Your clinical exam is unlikely to be sufficient to detect abusive abdominal injuries. You may screen with AST and ALT in every child with concern for physical abuse up to the age of 5. If these are elevated (>80 IU/L), the likelihood of a positive abdominal CT is about 1 in 5. This will not detect everything. If you have a child who reports that they were punched in the abdomen or with bruising or tenderness, it may be reasonable to go straight to CT. q Neuroimaging with CT head or MRI. There was a study published with children under the age of 2 who were hospitalized for suspected abuse with no signs of brain injury. Patients who were deemed high risk (<6 months, facial bruising, rib fractures, multiple fractures) received neuroimaging. The rate of positive CTs was 37%. q Oropharyngeal injuries in a non-mobile infant. A tear in the frenulum under the tongue is concerning for abuse. q Any long-bone fracture in an infant outside of a major car accident is concerning for abuse. The rate of abuse associated with these fractures in infants is 20% or more. If you are not going to screen for abuse in these patients, you need a good reason (e.g. multiple people unrelated to the family saw the accident happen the way the family described, it was on videotape, it was a very significant mechanism, etc.). The risk of screening is much lower than the risk of missing abuse. u We do a bad job of identifying cases of abuse. Approximately 30% of cases of abusive head trauma are missed initially by physicians. About 1/5 of kids with abusive fractures are seen by physicians and sent home without a diagnosis of abuse. u Burns are complex. We are able to determine approximately how long it would take a child to get burned based on the temperature of the water heater at the house. The location of the burn can indicate the position of the child at the time they were burned. Consider consultation with a burn specialist or a child abuse specialist in any burn case that is not obviously due to the reported mechanism. If you do not have access to an expert, take pictures of the burn. u How do we discuss performing a work-up for abuse with the family in a way that is not punitive or stigmatizing? Your decision to do a work-up should be uniform for a particular injury and not based on whether or not you like the family or their demographics. For example, a patient presents with a fracture of both bones of the forearm after a reported fall off a bed. What can you say to the family? “We found a fracture of both bones in the What tests should you order? January 2014: Volume 14, Issue 1 | www.emrap.org 15 The Written Summary of the Rubin DM et al. Occult head injury in high-risk abused children. Pediatrics. 2003 Jun;111(6 Pt 1):1382-6 PMID: 12777556. q Neuroimaging should be obtained in any child with signs of head trauma, altered mental status, bulging fontanelle or if they are young and high risk (rib fractures, multiple fractures, etc.). If it is a stable infant, you can start with MRI. If they are unstable or older, a head CT is more accessible. If you do not have access to MRI, a CT is reasonable. q Every child needs a thorough physical exam. Get them naked. Look in their ears. Feel their fontanelle. Look at their genitalia. Look inside their mouth for chipped teeth, torn frenulum and mucosal injuries. u What can you do to increase your odds of detecting abuse? Make a list of injuries where you will decide to screen every time. Make a list of screening studies, that when you think about it, you do every single time. Batch order it once you are thinking about abuse. Once this is triggered, your rate of positive findings will be fairly high. Think less and screen more. u Kids that should trigger an abuse work-up. The story is inconsistent with the injury. Children under 6 months with bruising. Children of any age with bruising to unusual parts of the body like the abdomen, ear, neck or pattern bruising. Rib fractures at any age. Multiple fractures in different stages of healing. Classic metaphyseal fractures. Subdural or subarachnoid hemorrhage. Oropharyngeal injuries in non-mobile injuries. Long bone fractures in non-mobile infants. u Some instructive cases. q Children with frenulum tears or oral injuries in non-mobile infants. These injuries are so minor and often heal without intervention. It hasn’t been well publicized that these injuries are concerning for abuse. A child may be crying or not eating food and someone may shove a pacifier, finger, bottle or spoon in the mouth or strike the child in the mouth. q Children presenting with vomiting without diarrhea at a young age are challenging. Most of these children will have gastroenteritis and not abuse but upon review of missed cases of abuse, many of the children had a presenting complaint of vomiting. Consider how long it has been going on. If it has been more than 24 or 72 hours, it is less likely to be gastroenteritis. Do the full physical exam and carefully examine the fontanelle. u 16 Monthly Audio Program for retinal hemorrhages. When we see retinal hemorrhage we are probably correct but we are not good at ruling them out. If the head CT is negative it is unlikely that there are retinal hemorrhages. This is supported by the literature. u Consider trauma in reported medical cardiac arrests in small children. During resuscitation of a toddler brought in as a medical arrest, a bedside FAST showed positive free fluid consistent with abdominal trauma. Another case was found to have a rectal laceration due to sexual and physical abuse. If you don’t think of it and do a thorough physical exam, you won’t find it. Retinal hemorrhages. Retinal hemorrhages obtained after birth trauma usually go away after a month. They are in a different pattern then that seen in abuse. Whenever you are concerned about abusive head trauma with a positive CT or MRI, get a pediatric ophthalmologist involved to look EM:RAP Written Summary | www.emrap.org