Skip Navigation LinksHome > Blogs > M<sup>2</sup>E Too! Mellick's Multimedia EduBlog
M2E Too! Mellick's Multimedia EduBlog
The M2E Too! Blog by Larry Mellick, MD, presents important clinical pearls using multimedia.

By its name, M2E Too! acknowledges that it is one of many emergency medicine blogs, but we hope this will serve as a creative commons for emergency physicians.

Wednesday, September 03, 2014

A sucking chest wound or open pneumothorax is a fairly uncommon event off the battlefield, and civilian guidelines as a result are highly dependent on the experience of the military expert opinion concerning their management. Past recommendations were to place a three-sided occlusive dressing to allow air to egress to prevent a tension pneumothorax, but battlefield guidelines calling for an occlusive dressing closed on three of four sides have not proven to be effective or realistic. Covering the wound improves respiratory mechanics, but the three-sided occlusive dressing on bleeding chest wounds is no longer recommended. Current tactical prehospital guidelines recommend a vented chest seal or closing the wound and observing the patient for development of a tension pneumothorax if one is not available. A needle is placed in the chest to release the air if a tension pneumothorax develops.

 

The Tactical Combat Casualty Care Guidelines recommend a non-vented chest seal if a vented chest seal is not available, and that physicians treat by burping or removing the dressing or by needle decompression if the patient develops increasing hypoxia, respiratory distress, or hypotension, and a tension pneumothorax is suspected. (National Association of Emergency Medical Technicians, Tactical Combat Casualty Care Guidelines, Oct. 28, 2013; http://bit.ly/1pdHwIw.)

 

The recommendations for a vented chest seal is a change from the 2011 guidelines that recommended immediate application and securing of an occlusive material to cover the defect and subsequent monitoring of the casualty for the potential tension pneumothorax. (ItsTactical.com, Tactical Combat Casualty Care Guidelines, Aug. 8, 2011; http://bit.ly/1oe4Cjc.)

 

Five or six brands of chest seals are available. At least one, the SAM chest seal, comes in vented and non-vented forms. Not much research is available on this topic, but two 2013 articles looked at vented chest seals. The first article found that HyFin, SAM, and Sentinel vented chest seals were equally effective in evacuating blood and air in a communicating pneumothorax model. All three also prevented tension pneumothorax formation after penetrating thoracic trauma. (J Trauma Acute Care Surg 2013;75[1]:150.)

 

The second article studying injured pigs found that vented and unvented chest seals provided immediate improvements in breathing and blood oxygenation in their model of penetrating thoracic trauma. The unvented chest seal led to tension pneumothorax, hypoxemia, and possible respiratory arrest in the presence of ongoing intrapleural air accumulation, while the vented chest seal prevented these outcomes. The outcome for the unvented chest seals was not expected because a tension pneumothorax was simulated by injecting approximately 1.4 liters of air into the thoracic cavity of the animal models. (J Emerg Med 2013;45(5):686.) A 2008 study found that the Bolin and Asherman chest seals equally prevented the development of a tension pneumothorax in this open pneumothorax model. (Injury 2008;39[9]:1082.)

 

Not all sucking chest wounds are created equally, however. Some current chest seals with their relatively small footprint would be unable to seal and vent large sucking chest wounds. Several years ago we treated a 4-year-old girl who was attacked by the family’s pit bulls. One of the dogs was not chained up, and grabbed her torso and tore open the chest wall, creating a large sucking chest wound. (Figure 1/video.) Multiple bite injuries were also sustained to her hand and leg. (Figure 2.)

 

 

Click here to watch a video of a 4-year-old with a sucking chest wound.

 

The ED who transferred her to us applied sterile petrolatum gauze and a large abdominal dressing over the wound. Her initial hypoxemia (SaO2 79% on room air) was treated with supplemental oxygen by face mask. The first chest x-ray showed no evidence of a pneumothorax and the petrolatum gauze appeared to be letting air escape, so we did not change the initial management. The patient was subsequently taken to the operating room for further management of the open pneumothorax and washout of her bite wounds. A chest tube was inserted in the operating room, and the large thoracic wound was closed.

 

  

 

Current recommendations for managing a sucking chest wound have changed. The old recommendation to place a piece of plastic wrap or aluminum foil on top of the wound and tape it on three sides should be considered no longer appropriate, even in the austere environment of the wilderness. (Wilderness Medicine. Philadelphia: Elsevier: 2011, p. 456.) You may still use whichever appropriate material you have in the wilderness — plastic wrap, a piece of a plastic bag — but seal the wound on four sides after maximally venting the air from the chest by timing application with full expiration. Then be sure to monitor the patient for increasing hypoxia, respiratory distress, or hypotension, and tension pneumothorax. If signs of decompensation develop, treat by burping or temporarily removing the dressing or, if necessary, by needle decompression.


Wednesday, August 06, 2014

Relative or paradoxical bradycardia in the face of severe blood loss and hypotension fits squarely in the “it-just-ain’t-right” category. It isn’t right because it happens far more frequently than we recognize, and it contradicts what we have been taught about the body’s response to hemorrhage. A definite and significant subset of hypotensive patients will not mount a tachycardia response to hypotension, and a third or more of hypotensive trauma patients will present with bradycardia. (J Trauma 1998;45[3]:534 and 2009;67[5]:1051; J Am Coll Surg 2003;196[5]:679.)

But it just isn’t relative bradycardia in the face of hypotension and hemorrhage where the heart rate’s reliability as a marker for severe blood loss fails. The truth is that even though there is an association between hemorrhage and tachycardia, it is actually unreliable and nothing like the four classes of hypovolemic shock described in that well-known table published by the Advanced Trauma Life Support course. The evidence shows that the ATLS table is more wishful thinking than reality. (Resuscitation 2011;82[5]:556; 2010;81[9]:1142; 2013;84[3]:309.) We still need to respect tachycardia in the face of hypotension, but heart rate is not a reliable indicator in the trauma patient. (J Am Coll Surg 2003;196[5]:679; J Trauma 2007;62[4]:812.)

The sensitivity and specificity of tachycardia is poor. Patients with rapid heart rates frequently do not have serious blood loss, and others who are not tachycardic will be found to be seriously hypovolemic.

Nevertheless, paradoxical bradycardia in the face of massive hemorrhage and hypotension is a fascinating physiologic response. It is very real, and has the potential to trip up the unsuspecting clinician. Several years ago, a 20-year-old woman presented to our emergency department with severe vaginal bleeding. Obstetrics and gynecology was consulted shortly after her arrival in the emergency department, but it was early in the academic year, and delay after delay occurred. Multiple calls to the on-call resident were made. Eventually, the patient demonstrated hypotension that rapidly worsened.

The paradoxical or relative bradycardia associated with her hypotension was truly impressive. Her heart rate was recorded at 58 beats per minute even when her blood pressure was 66/34 mm Hg. Emergent blood transfusion bought us time until the obstetrics gynecology consultants finally arrived to take the patient for an emergency dilation and curettage. Nevertheless, a 4 g drop in hemoglobin was documented over a relatively short period of time. And, of course, the consultants were heard blaming the emergency department for not calling them sooner. Click here to watch a video of this patient.

A number of potential explanations exist for this phenomenon. A vagal or parasympathetic response is often proposed because it is a relative common occurrence in penetrating or blunt abdominal trauma. (J Emerg Med 1989;7[4]:335.) Some experts also say this is one stage of the body’s response to significant blood loss. Heart rate slowing allows the ventricles to fill and prevents the empty heart syndrome. In fact, tachycardia reportedly occurs after fluids and blood are replenished.

We have been taught for years that heart rate is an integral vital sign in the assessment of patients at risk of hemorrhagic shock, but it just isn’t reliable. And it will mockingly try to trick you at the extreme when the heart rate paradoxically slows in the face of hemorrhagic shock. Don’t get caught unaware!


Wednesday, July 02, 2014

Looks can be deceiving. I was not supposed to have a difficult airway, but we found out the hard way that wasn’t true. I was about to undergo an exploratory laparotomy for suspected appendicitis in 1989, and my wife was adamant that an attending anesthesiologist perform my intubation. The anesthesiology resident at a large Midwest teaching hospital made a claim that I suspect many of us did as we advanced through training and began to feel procedurally invincible. He claimed that the attending anesthesiologist rarely did intubations, and that the residents who performed them daily were more prepared. Unaware of any unique issues with my airway and wanting to be a good patient, I didn’t attempt to override the resident.

I woke up from anesthesia with a swollen, bleeding, and extremely painful pharynx, and my right central incisor was abraded and roughened. We learned that the resident made four attempts before the attending took over and successfully placed the airway. The soft tissue injury to the pharynx resolved with time, and my dentist filed the tooth smooth again. Subsequent surgeries, however, included a successful bougie intubation and a not-so-pleasant awake intubation.

This May, my shoulder pain reached a point where I knew it was time for surgery again. My left rotator cuff muscle needed to be repaired, just like the right one had years before. Intubation methods were a necessary topic of conversation with my anesthesiologist again.

My “3-3-2” airway assessment is abnormal on two counts. Instead of three fingers, my mouth will only open to allow two and maybe part of a third finger between the upper and lower incisors. The second “3” is the distance between the mentum and the hyoid bone. This gives an idea of how much space is available for the tongue to be displaced during laryngoscopy. My facial anatomy appears fine in this regard. The “2” refers to the position of the larynx in relation to the base of the tongue. Two or more fingerbreadths between the hyoid bone and the upper anterior edge of the thyroid cartilage are ideal. This signifies that the larynx is located sufficiently beyond the base of the tongue. I barely get one fingerbreadth with this measurement. These two variances add up to a markedly anterior larynx and a difficult-to-visualize airway. I have never been told what my Cormack-Lehane classification is, but my Mallampati score is a Class II.

Visualization of Glottis: Cormack-Lehane Classification
Grade 1:
Full view of glottis and vocal cords
Grade 2: Vocal cords and glottis are partially visible
Grade 3: Only epiglottis seen; glottis is not visible
Grade 4: Glottis and epiglottis are not visible


 

Visualization of Upper Airway: Mallampati Class
Class I:
Faucial pillars, soft palate, and uvula visualized
Class II: Faucial pillars and soft palate visualized; uvula not seen (masked by tongue)
Class III: Only soft palate and base of uvula visualized
Class IV: Faucial pillars, soft palate, and uvula not visualized (only hard palate visualized)

The anesthesiologist who agreed to take my case chose a fiberoptic intubation using nebulized lidocaine, a Parker Flex-Tip endotracheal tube with the unique beaked tip, and the endoscope plastic oral airway guide. In fact, we agreed to make a teaching video of my procedure, even knowing it might not be pretty. It wasn’t. Everything the anesthesiologist did was technically correct. Unfortunately, the lidocaine nebulization failed to anesthetize my airway. That, however, was my fault. Instead of consistently inhaling the 4% lidocaine, I stopped multiple times to greet acquaintances in the operating room and to direct the filming of my video. My reward was vigorous coughing and bucking as the endotracheal tube and fiberoptic scope were inserted.

The intravenous midazolam also caused a short period of apnea, and it was a little unnerving to listen to the pulse oximetry monitor alarm drop in pitch when I reviewed the video later. I couldn’t actually tell how low my oxygen saturation went, but the oxygen was turned up, and the first reported reading was of 93%. Thankfully, I have no memory of these events.

The truth is that events like this happen all the time despite our best-laid plans, and they are the substance of life in the ED that keeps us simultaneously on our guard (paranoid) and humble. I love the video because it shows a few of the realities of difficult airway management that can be instructional for learners. Click here to watch a video of the interscalene brachial plexus block used for my shoulder surgery.

Have you ever had a patient with a difficult airway like mine? How did you manage it? What tips do you have for difficult intubations?


Wednesday, June 04, 2014
Limb exsanguination is no longer the most common cause of preventable death on the battlefield because of emergency tourniquets. Hemorrhage control for wounds in the junction between the trunk and the limbs and in the neck are an obvious care gap, most commonly in the pelvic area, including the buttocks and groin proximal to the inguinal ligament. (US Army Med Dep J 2011 Apr-Jun:38.) Managing hemorrhagic shock also requires support of central aortic pressure to maximize perfusion of the brain and heart and to control bleeding effectively.
 
Noncompressible torso hemorrhage is recognized as the leading cause of death in trauma, but definitive hemorrhage control has been challenging. A review of combat casualties from 2001 to 2011 showed that the most common sites of lethal hemorrhage were truncal (67.3%), junctional (19.2%), and peripheral-extremity (13.5%). (J Trauma Acute Care Surg 2012;73[6 Suppl 5]:S431.)
 
The direct application of pressure over the distal aorta with an abdominal tourniquet and the insertion of a resuscitative endovascular balloon to occlude the aorta (REBOA) has recently come to the forefront as a potential life-saving option. The problem is bleeding from noncompressible vascular injuries. Traumatic or iatrogenic injuries to the common iliac, external iliac, internal iliac, profunda femoral, and superficial femoral can cause bleeding that is extremely difficult to control. (J Cardiovasc Surg (Torino) 2012;53[4]:495.) Exsanguination from one of these vessels can occur in minutes or even seconds.
 
Pulsed-wave Doppler measurements have documented that blood flow to the common femoral artery can be stopped by applying direct pressure over the distal abdominal aorta or proximal iliac artery. (Prehosp Disaster Med 2006;21[6]:379.) A number of reports have now documented the ability to stop blood flow to these major vessels with the Abdominal Aortic and Junctional Tourniquet (AAJT) and the REBOA. (Mil Med 2013;178[11]:1196; J Trauma Acute Care Surg 2013;75[1]:122; J Trauma Acute Care Surg 2013;75[3]:506.) Major trauma center experience with REBOA suggests that it is feasible and effective for proactive aortic control in patients with end-stage shock from blunt and penetrating mechanisms. (J Spec Oper Med 2013;13[3]:1.) Likewise, real-life experience with the AAJT also suggests that it can save lives. (J Spec Oper Med 2013;13[2]:1; Surgery 2011;150[3]:400.)

 
The AAJT and REBOA are relatively new concepts in emergency medicine, and their potential benefits and roles in managing life-threatening bleeding remain untapped to a significant degree. A respectable amount of research has been done for these tools and the FDA has approved several different applications of the AAJT, so it’s time to get broader clinical and research experience with them.
 
The AAJT has other applications besides inguinal and groin hemorrhage. The FDA approved it for life-threatening axillary hemorrhage, and it has been used off-label for life-threatening neck hemorrhage.

 
But what about using the AAJT or REBOA instead of cross-clamping the aorta when an open thoracotomy is performed? One animal study specifically looked at the use of REBOA for that purpose. (Can Med Assoc J 1964;91:128.) The AAJT would obviously occlude the aorta lower than the thoracic diaphragm cross-clamping location typically used during open thoracotomy, but cross-clamping the aorta during an open thoracotomy has been fraught with difficulty and plagued by operator inexperience. Many an esophagus has been inadvertently cross-clamped, and the traumatic injury to the aorta and its branching vessels from dissection and cross-clamping can be significant. The lower-placed AAJT would not occlude blood flow to the kidneys and the superior mesenteric artery, but maybe it’s not a bad idea to preserve blood flow to the kidneys and mesentery in addition to the brain and heart.
 
The biggest drawbacks to the REBOA are the technical skills required for insertion and the actual time it takes to insert the balloon. I want something fast when my patient is hemorrhaging uncontrollably, and that is where the AAJT fills the bill: it can be applied in less than 60 seconds. It’s possible these two tools can be used together effectively, with the AAJT being applied immediately to allow additional time for REBOA placement.
 
REBOA for stopping uncontrolled bleeding in pelvic fractures has been reported as quite effective. Again, it would be great to see some experience with the application of the AAJT for that same indication. Researchers reported in 1964 that compression of the aorta with the heel of the hand just above the bifurcation of the common iliac arteries was an effective way of controlling catastrophic pelvic hemorrhage during laparotomy. (Ann Emerg Med 2013 Oct 23; doi: 10.1016/j.annemergmed.2013.09.026.) More recent reports describe external manual compression for abdominal bleeding associated with penetrating injuries and ruptured abdominal aortic aneurysms. (J Anesth 2002;16[2]:164; Obstet Gynecol 2014;123[4]:804.) Pelvic binding and the AAJT might slow blood loss significantly in pelvic fractures.
 
Postpartum hemorrhage is another life-threatening emergency that can occur after a Caesarean section or vaginal delivery. It is one of the most common obstetric maternal complications, and is responsible for approximately half of severe maternal morbidity. (Transfusion 2014 Mar 12; doi: 10.1111/trf.12550.) Worldwide, postpartum hemorrhage is a common etiology of maternal death, and uterine atony is the major associated cause. (WHO Guidelines; http://bit.ly/1nECNO7.) External manual compression of the aorta is a common treatment for life-threatening post-partum hemorrhage. (Anaesth Intensive Care 1994;22[5]:571.) A commonly referenced case report describing successful management of a hemorrhaging patient with this technique was published in 1994. (Circulation 1993;88[4 Pt 1]:1916.) Again, here is another condition where it seems the AAJT could play a life-saving role in slowing uterine bleeding while other interventions are being started.
 
Applying the abdominal tourniquet during CPR has been discussed as one way to shunt blood flow selectively to the heart and brain, meeting CPR’s goal of maintaining blood flow to the vital organs until the heart can once again function independently. Aortic compression to augment cardiopulmonary resuscitation has been written about for decades. The results were variable, though, and a number of well-done papers demonstrated improved resuscitation parameters and outcomes when the aorta was occluded during CPR. (Am J Emerg Med 2002;20[5]:453; Resuscitation 2001;50[3]:319; Resuscitation 1999;40[3]:171; Resuscitation 2011;82[8]:1087; Resuscitation 2007;75[3]:515; Crit Care Med 1997;25[6]:1003.)
 
Applying the AAJT during CPR is a study that simply must be done. A significant step forward will have occurred in advanced cardiac life support if a distinctly positive outcome similar to the animal studies is found.
 
Both these tools can save lives for a wide spectrum of different life-threatening conditions. Unfortunately, the AAJT and the REBOA are currently relatively unknown and not being used. Emergency medicine and trauma programs need to include training in how to apply the AAJT and how to insert the REBOA. And, of course, we need increased research and assessment of various clinical applications.
 

Additional information about the Abdominal Aortic Junctional Tourniquet can be found at http://www.speeroptech.com/.

Dr. Mellick disclosed that he has no conflicts of interest related to the AAJT, but that the chairman of his department, Richard Schwartz, MD, is one of its inventors.
 

Thursday, May 01, 2014
I don’t have a problem with lumbar punctures in febrile neonates. In fact, my son ended up with three lumbar punctures before aging out of the febrile seizure protocols. I do have a problem, however, with doing unnecessary spinal taps. The emotional stress of a neonatal LP on parents is significant, and the physical stress of the procedure on the infant is also substantial.
 
 
The pain of the needle and the unique restraint required for the procedure are also potentially problematic. The pain of the needle can be minimized by EMLA cream (eutectic mixture of local anesthetics [lidocaine and prilocaine]) and local anesthesia. The risk of patient hypoxia is a well-recognized complication when scrunching a baby into a lateral, recumbent fetal position. No matter your level of experience, a number of traumatic taps will unavoidably occur (10-35%). (Pediatr Emerg Care 2011;27[11]:1057.)
 
Even though the red and white blood cell counts can work with various formulas, the abnormal cell counts and the Gram stain’s crowded field leave a degree of diagnostic uncertainty in the end.
 
Controversy exists on this topic. Some providers feel strongly that lumbar punctures should be performed on every febrile infant under 90 days of age. Others, including many pediatricians, feel confident in assessing individual risks and recommending lumbar punctures on a case-by-case basis. (JAMA 2004 10;291[10]:1203.) Neonates under 30 days old are placed in a separate category by most studies. Blood, urine, and CSF cultures are recommended for neonates, as are hospital admission and antibiotic therapy with a combination of ampicillin and a third-generation cephalosporin or gentamicin. Surprisingly and contrary to my perception, recent research shows that even the guidelines for infants under 28 days old are not followed rigorously. In fact, one in six febrile neonates are discharged from pediatric emergency departments after varying degrees of laboratory evaluation and antibiotic treatment. (Pediatrics 2014;133[2]:187.)
 
Is it possible to identify clearly which febrile infant has a virus and which has a serious bacterial infection? A number of risk assessment protocols have been published and validated. The Boston, Philadelphia, and Rochester protocols are most commonly used, and the Rochester protocol does not mandate lumbar puncture as part of the evaluation. (Pediatrics 1994;94[3]:390.)
 
 
The Rochester criteria also recommend admission for low-risk children whose parents were not mature, who are socioeconomically unstable, who are not within 30 minutes of the hospital, who could not return in 24 hours, or who did not have a car, thermometer, or telephone.
 
A review of the literature suggests that it is possible to avoid mindlessly performing a lumbar puncture on every febrile neonate. Consider the following.
 
Most are viral infections (approximately 90 percent): Viral infections are by far the most common cause of febrile illness in these infants: urinary tract infections, bacteremia, and very rarely meningitis occur in descending order. A 2012 retrospective study by Morley et al. attempted to determine the rates and the etiology of bacterial infection in children under 2 months during the vaccination era. In children younger than 28 days, 2.7 percent of blood, 10.7 percent urine, and zero percent (range 0.0-3.9%) of cerebrospinal fluid cultures were positive. In children 29 to 60 days, 1.5 percent of blood, 8.5 percent of urine, and 1.7 percent of CSF (range 0.0%-5.0%) cultures were positive. (Pediatr Emerg Care 2012;28[2]:125.)
 
Low risk and serious bacterial infections: Infants classified as low risk have low rates of serious bacterial infections. Huppler et al. demonstrated in their extensive review of the literature that the rates of serious bacterial infections in low-risk patients in retrospective and prospective studies using empiric antibiotic treatment were the same (2.7%), and were significantly different from the rate of serious bacterial infections in low-risk patients in prospective studies in which antibiotics were withheld (6/870 patients [0.67%]). (Pediatrics 2010;125[2]:228.)
 
They noted that the low-risk criteria functioned best when used in prospective studies in which low-risk patients underwent observation alone and empiric antibiotic treatment was not an option. Urinary tract infections (false-negative urinalysis) and bacteremia will occasionally be missed, but multiple studies have shown that the risk of meningitis is extremely low. (Pediatrics 2010;125[2]:228.) Low-risk, healthy-appearing infants did well without complications in various studies even when sent home and the infection was detected within 24 hours. The negative predictive value of the Rochester criteria is 98.9% with a 95% confidence interval of 97.2% to 99.6% for any serious bacterial infection, and 99.5% with a 95% confidence interval of 98.2% to 99.9% for bacteremia. (Pediatrics 1994;94[3]:390.)
 
Meningitis: No cases of meningitis occurred in the five prospective studies that used observation alone without antibiotics (0/870 patients), according to the review article by Huppler et al. that looked at low-risk patients. Two cases of meningitis were found in 3114 (0.06%) patients classified as low risk in the prospective studies that treated all patients with antibiotics. The authors hypothesized that the meticulous assessment of study patients was affected when all patients were ultimately treated with antibiotics. (Pediatrics 2010;125[2]:228.) Nevertheless, this information strongly suggests that the risk of meningitis is very small in low-risk infants who have been carefully evaluated.
 
The rate of meningitis is low even in high-risk patients: approximately 1-1.7 percent. (Pediatr Emerg Care 2012;28[2]:125; Pediatrics 2004;113[6]:1662.)
 
Viral testing and serious bacterial infections: Infants who test positive for a viral infection also have a significantly lower risk of having a low rate of serious bacterial infections. Again, however, the associated serious bacterial infections are primarily urinary tract infections with bacteremia; meningitis was almost nonexistent. A 2004 study by Byington et al. compared serious bacterial infections in infants with and without documented viral infections; a serious bacterial infection occurred in 4.2 percent of those with viral infections compared with 12.3 percent of infants without a confirmed viral infection. (Pediatrics 2004;113[6]:1662.)
 
No cases of bacterial meningitis were detected in infants with viral infections and in six (0.67%) patients without an identified viral infection. Bacteremia occurred in five of 491 (1%) compared with 24 of 894 (2.7%) infants without an identified viral infection. Another study of serious bacterial infections found that infants with a positive RSV test had a 5.4 percent rate of urinary tract infections as compared with a 10.1 percent rate of UTIs in the RSV-negative infants. The risk of bacteremia (1.1% vs 2.3%; risk difference: 1.2%; 95% CI: -0.4% to 2.7%) or meningitis (0/251) as compared with RSV-negative infants did not achieve statistical significance. (Pediatrics 2004;113[6]:1728.) Very similar numbers were described in another study that tested 1091 infants for influenza viruses. Again, statistical significance was not met, but none of the influenza-positive infants had positive blood cultures or meningitis, and urinary tract infections predominated as serious bacterial infections. (Pediatrics 2009;124[1]:30.)
 
High risk with a virus: What about infants deemed high risk according to the Rochester criteria who have a documented viral infection? High-risk infants with a documented viral infection (HR+) have a lower risk of serious bacterial infection than high-risk infants without a viral infection (HR-). (Pediatrics 2004;113[6]:1662.) The risk of a serious bacterial infection in HR+ infants (18/323 [5.5%]) was comparable with low-risk infants (9/289 [3.1%]). Importantly, the high-risk infants who were positive for a viral infection (N=324) had no cases of meningitis (0/324), 3.7 percent (15/324) had UTIs, and 0.92 percent (3/324) had bacteremia.
 
Hib and PCV immunizations: Immunizations for the common childhood illnesses don’t begin until 60 days of age. The Hib vaccine protects against Haemophilus influenzae type b, and the PCV vaccine protects against pneumococcus. Interestingly, it appears that these bad players are not a problem in the vast majority of febrile infants. These classic pathogens are essentially nonexistent, thanks to herd immunity. (Pediatr Emerg Care 2012;28[2]:125.)
 
Disposition home: Low-risk infants over 30 days of age (carefully evaluated according to the Rochester criteria) with or without positive viral testing can be sent home without a spinal tap.
 
These suggestions can help safely avoid lumbar punctures in febrile neonates:
 
• Perform a meticulous physical examination and detailed history.
• Don’t rush the assessment. Spend time watching the infant and paying close attention to parental observations and concerns. Remember, a child is not considered low risk if he does not look well, regardless of laboratory findings.
• Assess risk for a serious bacterial infection by using one of several proven protocols. (The Philadelphia and Boston protocols are commonly used, but only the Rochester criteria allows for not doing a lumbar puncture if low-risk criteria are met.)
• If available, perform viral testing for the common viral agents such as respiratory syncytial virus and influenza. Even better, more expansive full respiratory virus panels are now available, and their price has become increasingly reasonable.
• Every infant should have blood and urine cultures sent off to the lab in addition to the complete blood cell count and urinalysis. Some institutions will add a CRP. Stool should be checked for fecal leukocytes, culture, and rotavirus if diarrheal illness is present. Consider a chest radiograph if a patient has respiratory symptoms and for the herpes simplex virus if he has vesicular skin or mouth lesions. Appropriate management is mandatory.
• Be sure your institution has an efficient system for tracking and reporting blood and urine cultures. Tracking cultures during the first 24 hours is critical to preventing missed serious bacterial infections. A culture-proven UTI can occur despite a normal urinalysis in 35-50 percent of these infants. (Pediatrics 1994;94[3]:390.) The time needed to identify a true positive culture is under 24 hours for most cultures. (Ann Emerg Med 2000;36[6]:602.)
 
The final disposition for low-risk infants managed without a lumbar puncture or antibiotics will be admission to the hospital for careful observation or discharge home with clear return precautions and close follow-up the following day. Admission to the hospital is more expensive, but some consider it the most risk-adverse approach. Admission would occur for low-risk infants without a lumbar puncture and without antibiotics. Other practitioners will feel compelled to perform an LP on every febrile neonate under 90 days, and will often start antibiotics as part of the admission process.
 
If hospitalization is chosen, 24 hours or less will be sufficient time for observation of most infants and for most cultures to become positive. (Ann Emerg Med 2000;36[6]:602.) If you send home a low-risk patient without a spinal tap, it is essential to have confidence in the caregivers, provide clearly delineated return precautions, and ensure easy access to medical care with close follow-up the next day. (Pediatrics 1994;94[3]:390.)
 
The bottom line is that a small risk of serious bacterial infection remains for even low-risk patients with a documented viral illness. Know ahead of time what level of risk you are comfortable accepting for your patients. Thankfully, the serious bacterial infection risk for meningitis is very, very low.
 
A spinal tap may no longer be avoided if the child is high risk without a documented viral infection. This video outlines three steps to maximize the success of infant spinal tap and to avoid the dreaded traumatic spinal tap. Antibiotics should not be given before the spinal tap. The rate of meningitis is quite low, so about 99 percent of spinal taps will fail to demonstrate bacterial meningitis, but it is possible that the spinal fluid will be sterilized if antibiotics are given before the spinal tap, and this uncertainty may contribute to an unnecessarily prolonged hospitalization when discharge after 24 hours of observation without antibiotics would have been appropriate.
About the Author

Larry Mellick, MD
Dr. Mellick is a professor of emergency medicine and pediatrics at Georgia Regents University in Augusta, the former chairman of emergency medicine at Georgia Regents Health System, and a professor of emergency medicine and pediatrics at Georgia Regents Medical Center and Children’s Hospital of Georgia.

Blogs Archive