Share this article on:

Head Trauma in Infants: The Isolated Skull FracturePart 3 in a Series

Roberts, James R. MD

doi: 10.1097/01.EEM.0000288908.34786.24

Author Credentials and Financial Disclosure: James R. Roberts, MD, is the Chairman of the Department of Emergency Medicine and the Director of the Division of Toxicology at Mercy Health Systems, and a Professor of Emergency Medicine and Toxicology at the Drexel University College of Medicine, both in Philadelphia. Dr. Roberts has disclosed that he is the recipient of a speaker honorarium from Elan and the recipient of grant/research funding from T-System. Lippincott CME Institute, Inc. has identified and resolved all faculty conflicts of interest regarding this educational activity.

Learning Objectives: After reading this article, the physician should be able to:

  1. Discuss the radiographic approach to linear skull fractures in children.
  2. Describe the clinical presentation of linear skull fractures.
  3. Summarize the disposition of children with such injuries.

Release Date: December 2006

Emergency physicians frequently evaluate patients with seemingly minor head trauma, and the issues are generally straightforward. The history, circumstances of the injury, and the exam and clinical gestalt all have their respective roles. CT scans are used liberally, although minor lumps and bumps rarely require such sophisticated technology. Recently, the issue of excessive radiation to the brain of children has become an issue that may limit scans that previously had little indication except fear of litigation or misguided parental demands. It is possible that the significant radiation delivered by CT scanning to a developing brain may predispose children to cancer many years later. This month's column continues the discussion on evaluating the minor head trauma omnipresent in any ED with a look at isolated skull fractures in children.

Last month's column began a discussion of a rather narrow topic: head trauma in infants. It's very difficult to be scientific on this issue. Recommendations on the nuances of CT scanning are all over the map, and there is clearly no agreement on even basic issues, such as who deserves a CT scan.

In reviewing this complex topic, I concluded a few things: Children and adults with head trauma present vastly different scenarios. While adults may declare their head trauma quite readily, signs and symptoms are of little value in sorting out clinical nuances in young children who sustain head trauma. A scalp hematoma in an adult is usually a garden-variety benign issue. A scalp hematoma in an infant is a significant physical finding that is highly associated with skull fracture, which in turn correlates with intracranial injury, and this is true even in the asymptomatic infant. The current column tackles the issue of isolated linear skull fractures in infants. It seems like an infant's skull cracks quite easily, fortunately with few real clinical consequences.



Infants with Isolated Skull Fracture: What Are Their Clinical Characteristics, and Do They Require Hospitalization?

Greenes D, Schutzman S Ann Emerg Med 1997;30:253

These authors, prolific investigators in pediatric head trauma, attempted to identify the historical factors, physical findings, and clinical consequences of infants who have an isolated linear skull fracture. An isolated skull fracture by definition was associated with the absence of intracranial injury. From Boston Children's Hospital, this study retrospectively reviewed 101 children with radiographically proven isolated skull fractures who were admitted to the hospital. All were under 2, and 90 percent were under 12 months. In this study, a skull fracture diagnosis was confirmed with either plain radiography or head CT. In all cases, the CT scan showed no acute intracranial injury.

Apparently the protocol at Boston Children's Hospital in 1977 was to hospitalize all young children with such injuries. As expected, the most commonly reported mechanism of injury was a fall (89%). Many falls were minor and from low heights, such as a fall from standing or a height of less than five feet. At least one clinical sign that traditionally has been correlated with serious head injury was seen in 71 percent of cases, including loss of consciousness, seizure, vomiting, lethargy, irritability, depressed mental status, or a focal neurologic finding. Most of the time (96%) a skull fracture produced a local traumatic finding on the head, such as a palpable fracture, soft tissue swelling (hematoma), or signs of a basilar or skull fracture.

Despite some clinical fractures previously considered ominous in about 70 percent of the children, not a single patient demonstrated a clinical decline during hospitalization. All were neurologically intact at discharge. Systemic signs and symptoms were imperfect predictors of ultimate injury. The authors note that a diagnosis of isolated skull fracture should be considered in infants with a minor mechanism of head injury, even though they appear clinically well. They also conclude that most infants with an isolated skull fracture will manifest local signs of head injury on physical examination. They believe that if these children have no specific clinical concerns to necessitate hospitalization, the nondisplaced linear skull fracture alone does not mandate admission. Such children can be considered for discharge home.

Comment: An isolated skull fracture is a common reason for hospitalization of children with head injuries, and admission for observation of well appearing children, with normal CT scans in the presence of a skull fracture, is still the norm for many children's hospitals. Few clinicians are able to discharge a child from the ED with a fresh skull fracture, however clinically benign it appears. As in other studies, the mechanism of injury in this study was often trivial or reported to be minor in children who experienced a skull fracture. Even a minor fall put these children at risk. A child's skull bone is, apparently, quite easily cracked, and does not act like the common greenstick fracture of an extremity. Other investigators have commented on the incidence of skull fracture in infants following seemingly minor head trauma. (J Ped Surg 2001;36[1™:119; Lancet 1997;349[9055™:821; J Neurol Neurosurg Psych 2000;68[4™:416.)

It is frustrating to consider that even today there is no consensus regarding the most effective criteria for determining which child with minor head trauma should receive a head CT. (Eur J Ped 2006;165[3™:142.) As a theme in the literature, skull films seem like a reasonable initial approach to be used quite liberally in infants with minor trauma (note, infants only), followed by CT if a fracture is seen. A CT is usually considered mandatory for all children with signs and symptoms suggestive of intracranial injury, but there is little correlation between systemic signs and symptoms with intracranial pathology in most studies involving young children. Vomiting, irritability, and lethargy get the attention of most clinicians and all parents, but when studied carefully, the presence of these scary things usually do not portend pathology. On the other hand, absence of any sign or symptom usually rules out serious intracranial injury, but it does not rule out a linear skull fracture in young children.

To me, the data in this and other studies are crystal clear. Children readily fracture their skulls with even minor trauma, often following a seeming nonserious mechanism of injury. Most will have some external signs of injury, most commonly a scalp hematoma. Note: You can fail to appreciate even a widespread scalp hematoma on first glance, so unless you palpate the scalp, you can miss this finding. An isolated skull fracture rarely causes a problem in itself, but it is associated with intracranial injury in some children. Intracranial injury can only be ferreted out with a CT scan. Likewise, depressed fractures and basilar fractures are also best evaluated with a CT scan.



A plain film skull series is a reasonable screening tool in the asymptomatic and apparently normal infant who fell off the changing table and is brought in for a checkup. Plain skull films are currently akin to clinical dinosaurs, and most young doctors are not adroit at reading them. Small linear fractures are difficult to appreciate, and the multiple sutures in a child's growing skull further complicate the picture. If a skull radiograph suggests a fracture, everyone reflexively orders a CT scan. Basilar skull fractures are rarely seen with plain films; they require a CT scan.

Although a plain radiograph may identify most children with a simple linear skull fracture, it seems to me a waste of time, energy, and radiation if there are signs/symptoms or a scalp hematoma is felt. I go directly to a CT scan if I'm interested in evaluating the skull and brain of a head-injured child. My definition of “interested in evaluating” is signs/symptoms or physical findings of head trauma.

Unlike adults, it is clear that infants can have significant intracranial injury and skull fractures in the absence of symptoms classically associated with head trauma; loss of consciousness, seizures, vomiting, and altered mental status don't have to be present. And don't expect to find a classic focal neurological finding in these cases. Even though they may have subdural or epidural clots, such children usually have a nonfocal neurological examination. By the time you see a dilated pupil or hemiparesis, the child is in big trouble.

On the other hand, it has been my experience that adults with skull fractures are clinically ill, with severe headache, dizziness, and nausea, and they let you know something is wrong. External trauma may be subtle, but most adults let you know their skull is broken.

Not every kid with a boo-boo, however, needs a skull series or CT scan. The older the patient, the less likely will clinical judgment fail you. Minor stroller mishaps and forehead bumps on the coffee table in toddlers are not the same as an 8-month-old falling from the highchair to the linoleum floor. No one would likely eschew a head CT scan in any child who hit his head and was lethargic, had a seizure, had even a hint of loss of consciousness, or who had vomited more than once, even though such findings have very poor correlation with real pathology. For severe head trauma, projectile vomiting is the answer on the board exam, but it's an insensitive sign of increased intracranial pressure. With regard to bile-stained vomiting, think of other serious problems in children: this scenario congers up increased intracranial pressure, bowel obstruction, or sepsis.

In the old days when CT scans were not available, we simply watched these children in the ED. Some looked quite ill, and my nonchalant approach as a resident now scares me in retrospect. The child with blunt trauma often was irritable for a while and often vomited, but in a few hours he was back to his normal annoying self, running around the ED. I am sure some of those I sent home had skull fractures. Only recently have I appreciated the scalp hematoma issue.

Then came the CT scan, identifying all sorts of pathology never before imagined by the neophyte (AKA never sued) emergency physician. Numerous findings on CT often never amounted to much, but those long descriptive radiology reports made us all paranoid. This is reminiscent of the CT scan for chest trauma where now we find lung contusion and small pneumothorax with regularity on the CT but never on the traditional chest x-ray. However, we no longer practice medicine in a society where clinical judgment is superior to a test. We have the technology to identify minute skull fractures and subtle intracranial injury. But we are not smart enough to rule them out with physical examination or historical parameters, so I don't see the issue. The radiation to the brain not withstanding, a CT scan is an everyday occurrence in this country, and often ordered with very minimal indications.

In my hands, a child with a skull fracture is admitted to the hospital. Although they all usually do well, I have not yet reached the state where I can confidently say to a parent, “Your child has a skull fracture, he'll be OK, and you can leave if you just keep an eye on him.” Although some academicians may claim to have the data to back up this caviler approach, I'll bet they don't discharge children with a skull fracture on a regular basis. If my kid has a skull fracture, I want him in the hospital for at least 24 hours. And I probably want him to have a repeat CT scan before he goes home.

Finally, never utter the words that send the message: “I don't care if there is a skull fracture; I only care about brain injury.” Maybe you don't care, but parents certainly do. You can downplay the clinical importance of an isolated linear skull fracture, but don't alienate or confuse the family with complicated medical reasoning that no mother or father will ever understand.

While researching this topic, I came across a relatively obscure condition known as a “growing skull fracture.” I had not previously heard of this condition, but apparently less than one percent of infants with an uncomplicated linear skull fracture will develop a defect in the bone that fails to heal and becomes wider over time. This has led some authorities to suggest a repeat plain film or CT scan one to two months following the identification of an isolated linear skull fracture. I have never seen this condition, but it is yet another reason to consider routine neurosurgical consultation in otherwise asymptomatic children with skull fractures. Growing skull fractures are caused by a tear in the dura under the fracture that subsequently allows herniation of meningeal tissue into the fracture line. Associated injury to the brain almost always accompanies such fractures. The pressure exerted by the herniated tissue is one factor that appears to promote the growth of the skull fractures. Although most scalp hematomas resolve spontaneously, a persistent one might just indicate a growing fracture. Although I have never seen it, if I encountered a child with a scalp hematoma that lasted for more than three or four weeks, I would now consider a growing skull fracture as a potential complication. Surgical repair of the fracture and the defect in the dura are often required for these unusual injuries. By the way, resist the temptation to aspirate a scalp hematoma.

Basilar skull fractures can be difficult to find, almost invisible on plain x-rays. And, of course, a hemotympanum that escapes the detection of the clinician's otoscope is also obvious on a CT scan. The CT scan also can show some indirect signs of a basilar skull fracture. This fracture occasionally produces air in the mastoid cells, termed pneumocephaly. Opacification (with blood) of the mastoid air cells signifies bleeding from a basilar skull fracture. In fact, disruption of any of the paranasal sinuses raises the possibility of a skull fracture. Fractures that involve the sinuses predispose to subsequent intracranial infection.

Back to Top | Article Outline

Why Do Children Vomit After Minor Head Trauma?

Study To determine the factors associated with vomiting after minor head injury in children.

Population 563 children, age 0–13 with minor head trauma. Results 16% vomited after minor head injury.

Associated factors No relationship to mechanism of injury, site of impact, presence of scalp hematoma, skull fracture/brain injury, history of migraine in patient or family.

Positive association (p value less than 0.05) Prior “vomiting episodes” and history of motion sickness.

Conclusion Vomiting after head injury is related to unknown “intrinsic factors” rather than the specifics of the trauma.

Bottom line: Children vomit after minor head trauma, it means very little (although persistent vomiting was not mentioned), and it is not known why this occurs.

Source: Accident Emerg Med 2000;17(4):268.

Back to Top | Article Outline

Reader Feedback

Readers are invited to ask specific questions and offer personal experiences, comments, or observations on InFocus topics. Literature references are appreciated. Pertinent responses will be published in a future issue. Please send comments to Dr. Roberts requests feedback on this month's column, especially personal experiences with successes, failures, and technique.

Back to Top | Article Outline

Dr. Roberts

With some trepidation, I am writing regarding your articles on community-acquired MRSA. (“Community-Acquired MRSA: Still Virulent,” EMN 2006;27[7™:32; “Treating CA-MRSA Soft Tissue infections,” 2006;28[8™:23.) In our population, this nasty bug accounts for roughly 50 percent of soft-tissue infections and abscesses encountered in the ED. Your CA-MRSA antibiogram neglects gentamicin. This bug is universally susceptible to this tried and true antibiotic in our population. Once-daily dosing of gentamicin is accepted therapy at 5–7 mg/kg. Combined with 2 g cefazolin and 1 g probenecid, this makes an effective, cheap, and convenient once-daily therapy for the initial treatment of skin and soft tissue infections. (Tim Wolfe, MD, a colleague and friend, came up with this.) I have been able to treat several patients as outpatients with this regimen until culture results come back. I think all these abscesses should be cultured these days. Although half of infections are due to CA-MRSA, that obviously leaves half that are due to various strep species and good old MSSA. (This point was embarrassingly driven home a while ago when a patient with a classic presentation for CA-MRSA had the temerity to grow out MSSA that was resistant to the TMP/SMX but sensitive to cephalexin, all the worse because I had spent a good deal of time educating an intern about why the former and not the latter should be prescribed. Oops.) Our patients simply cannot afford linezolid so it is not an option. Vancomycin requires BID dosing, and some part of me still feels it should be reserved for proven resistant infections or those with clinically apparent sepsis, shock or risk factors for HA-MRSA. I try to avoid clindamycin due to the auto-inducible resistance phenomenon. I also will add five days of rifampin purely in an attempt to eradicate empirically the carrier state. (I have also tried mupirocin BID under the fingernails and in the nose for five days for the carrier state.) I don't have any evidence to back either of these practices. I am unsure what to do now about CA-pneumonia. Should we start covering for CA-MRSA empirically with gentamicin or vancomycin? I would appreciate any insights you may have into this CA-MRSA Gordian knot. — R. Stephen, MD, Salt Lake City, UT

Back to Top | Article Outline

Dr. Roberts responds

The gentamicin/cefazolin (Ancef) combination is certainly reasonable for a few days in the compliant outpatient. I really like adding probenecid to keep the cephalosporin around longer. I thought I was the only one who used that tactic. You are correct that vancomycin is a BID drug. High-dose gentamicin scares the internists, and the ID consultant uses the culture to guide therapy. In the interim, both seem to like empiric vancomicin and the expensive stuff, like linezolid (Zyvox) and now daptomycin (Cubicin). Once in the hospital, I doubt if anyone would appreciate or continue your regimen. I also tend to culture abscesses these days, but last month's letters regarding that concept did not agree with us. (See Dr. Daniel Mullin's take on this. [EMN 2006;28(10):29.™) Likely, most small abscesses will get better with simple drainage, even if CA-MRSA inhabits the cavity. It's all a guess; you should do that study, Dr. Stephen. With regard to any very sick patient (even pneumonia, and especially pneumonia in a nursing home patient), I now always give vancomycin in the ED just in case, and let the cultures or smarter doctors figure that one out later. I have seen patients die of CA-MRSA infections, even seemingly straightforward CA-pneumonia, when they just should not harbor this bug. Cast a wide antibiotic net in the ED, and let the other anglers identify and fry the specific fish in question.

Back to Top | Article Outline

Dr. Roberts

A few years ago, a colleague cared for an elderly patient who came in comatose at 9:30 a.m. with a huge subdural hematoma. (“Minor Head Trauma in the Anticoagulated Patient,” EMN 2006;28[9™:26.) He had been seen the prior day in the same ED for a ground level fall with a negative CT and an INR of 2.5 from Coumadin. He was discharged home. The initial CT was rechecked and was truly negative. I have heard that being on Coumadin doesn't increase your risk for an acute bleed, but makes a delayed bleed 40 times more likely. The obvious pitfall in this case was that the patient had no one at home to check on him. The questions it raises, however, are many. When should such a patient be admitted for observation? Can he go to telemetry? Is home observation safer if there is a responsible adult? Is there a role for vitamin K prophylactically in such a situation? (It might have helped this man.) What are your thoughts on this man who seems like the subject of the song: “It's raining. It's pouring. The old man is snoring. He bumped his head and went to bed…” You know the rest. — Brady Pregerson, MD, California (P.S. For me a guru [teacher™ is not anyone who agrees with me, but rather, exactly the opposite.)

Back to Top | Article Outline

Dr. Roberts responds

Many of our colleagues would have provided the same initial care, and most would have gotten away with it. There is no standard of care, just varying levels of paranoia. At least someone ordered an INR; many clinicians still don't even do that. But then after you knew he was anticoagulated (degree of external trauma not specified here), discharge home alone was the retrospective error. Elderly anticoagulated patients who bang their heads are clearly problematic; they simply can't be left to fend for themselves. To fend for them, we admit elderly patients for a lot less, so admission or prolonged observation should be used liberally, even if it is in the ED overnight, with a bona fide repeat exam the next day. As for the cause, I have read that it is the atrophied/sagging elderly brain that gets minute hemorrhages that are slowly exacerbated by Coumadin. I always question the accuracy of the history. Old folks fall a lot, don't remember, always want to go home, and often outright prevaricate to do so. A responsible caretaker at home is a plus, but that's putting a lot of medical trust on a layperson. If this man goes home, he needs to see a doctor within 24 hours, but it seems that it would have been too late for a follow-up in this case. I wonder how pristine and well documented the requisite repeat neurological exam was just prior to discharge. I think I would have admitted this patient, withheld vitamin K (fresh-frozen plasma is the key for proven bleeding), and ordered a repeat CT 12 hours later, but now you have me even more scared.

Back to Top | Article Outline

Dr. Roberts

Thank you for your thorough and superb discussion of “head trauma in patients with congenital bleeding disorders.” (“Head Trauma in Patients with Congenital Bleeding Disorders,” EMN 2006;28[10™:27.) For years, hematologists caring for individuals with hemophilia followed the adage, “If in doubt, treat.” They preached a message that it is critical to replace missing clotting factor in instances when bleeding be occurring: treat first, diagnose later. Nothing has changed. This practice remains the standard of care. Emergency physicians must appreciate that providing clotting factor to injured individuals with hemophilia is both time-sensitive and vital. Emergency physicians recognize that with myocardial infarctions, “time is muscle.” With hemophilia emergencies, time may be muscle; it may be brain or it may be airway. Attending hematologists should be consulted after the emergency evaluation of individuals with hemophilia, but decisions to provide clotting factor replacement are typically straightforward and should never await hematology consultation. Clotting factor administration is now safer than ever with both factor VIII and factor IX available in recombinant preparations. The challenge for emergency physicians is to provide these products rapidly. This is not an issue if the patient brings medication with him to the ED or if he has infused to 100% levels before the visit (50 u/kg factor VIII or 120 u/kg recombinant factor IX). Unfortunately, many hospitals do not stock factor VIII or factor IX, and many emergency physicians don't know how to get it. This gives rise to a time-sensitive and scary scenario. I'd encourage physicians to be proactive and to figure out how to address this before a problem arises. At our institution, we store enough recombinant factor VIII and IX to treat an average weight adult to 100% levels. The products are costly, but they have long shelf-lives (years), especially when refrigerated. We've worked out an agreement with our pharmaceutical supplier to avoid waste. If vials approach outdates, the supplier will recirculate them to high-use areas, and we will obtain new vials with a longer expiration date. Each hospital must address the issue of urgent, real-time product procurement. I'd encourage physicians to individually investigate the system they now have in place for securing clotting factor and determine if it meets the needs of your department and the patients you serve. If not, begin a dialogue with your pharmacy director and create a plan. Local chapters of the National Hemophilia Foundation may be helpful in addressing this issue. — Michael Coyne, MD, Pittsfield, MA

Back to Top | Article Outline

Dr. Roberts responds

Well said, Dr. Coyne. Right on.

Back to Top | Article Outline

Isolated Skull Fracture in Infants

  • In infants, skull fractures result from seemingly minor falls, often from low heights.
  • A small percentage of these are child abuse, motor vehicle crashes, and other mechanisms.
  • Any skull fracture should raise the issue of child abuse.
  • Children under age 1 have a higher risk for skull fracture than do older children.
  • Most linear skull fractures (about 70%) involve the parietal bone.
  • Frontal bone fractures are uncommon.
  • Most skull fractures have associated soft tissue swelling or hematoma overlying the fracture.
  • Palpable bony abnormalities are rarely appreciated by a physical exam.
  • Isolated skull fractures frequently occur in the absence of loss of consciousness, vomiting, lethargy, seizures, or irritability. The presence of these does not correlate with the pathology.
  • Basilar skull fractures are easily missed with plain x-rays.
  • Indicators of basilar skull fractures include Battle's sign, raccoon eyes, CSF rhinorhea or otorrhea, blood in a sinus, or pneumocephaly.
  • Basilar skull fractures often have no abnormalities on the examination of the skull or scalp.
  • Linear skull fractures require no specific intervention except for those that involve the frontal sinus. These fractures require surgical repair to prevent intracranial infection.
  • Asymptomatic children with a normal CT scan may be sent home with responsible parents after a period of observation in the ED, according to some authors.
  • Scalp hematomas should not be aspirated; they resolve spontaneously over one to three weeks.
  • Some linear skull fractures develop into a “growing skull fracture,” which can manifest as a scalp hematoma that does not resolve. A follow-up examination one month after injury is often suggested.
© 2006 Lippincott Williams & Wilkins, Inc.