Avoiding Common Pediatric Radiological Errors
Author: Brian Coley, MD, Assistant Director, Department of Radiology, Columbus Children’s Hospital; Clinical Associate Professor of Radiology and Pediatrics, The Ohio State University School of Medicine and Public Health, Columbus, OH
Peer Reviewer: David Kramer, MD, FACEP, FAAEM, Director, Emergency Medicine Residency Program; Vice Chair, Department of Emergency Medicine, York Hospital, York, PA
No one likes to make a mistake. Physicians often believe society expects them to be perfect, and they strive to practice error-free medicine. Medicine is, however, a human endeavor and, thus, has some inevitable inherent error. Education, experience, guidelines, and protocols go far toward promising to minimize mistakes, but human judgment is not infallible. Patient care and treatment all start with an accurate diagnosis. The history, physical examination, and laboratory and radiological information are all integrated with the experience of the clinician, and a diagnosis made that best fits the facts available. All of these factors are, however, subject to errors and inaccuracies.1,2 The emergency department (ED) physician is the critical link between the patient and the diagnostic tests, including radiographs, that are ordered. This article provides information on common radiographic errors to help clinicians improve their accuracy, confidence and subsequent patient care. — The Editor
Introduction
The most important consequence of medical error is, of course, harm to patients. Other consequences include increased healthcare costs, patient complaints, and legal actions against physicians and health care systems.3 In Pennsylvania, missed diagnoses (of which radiological errors are a part) accounted for 38% of all paid claims,4 and accounted for 70% of lawsuits filed against radiologists in Chicago.1 In the United Kingdom, issues related to radiographs are involved in one-half of litigation involving emergency departments (EDs).5 Interestingly, in spite of many quality improvement efforts, the rate of missed diagnoses has not changed appreciably since the work of Garland in 1949, and ranges from 2 -30% for clinically significant errors.1, 2, 6, 7 While changing modalities, interventional procedures, large and complex imaging studies, and ever-increasing physician workloads may contribute to this, Garland’s work emphasizes that human error is a difficult problem to overcome.
However unavoidable, this does not mean that complacency is acceptable. Individual efforts (e.g., education, training, knowledge of prior reports and current history) and system efforts (e.g., communication, documentation, quality improvement processes, working conditions) can minimize error rates and help to provide the best patient care possible. This article reviews common errors in radiology, techniques to avoid errors and specific difficult and error-prone pediatric radiographic diagnoses.
How Big is the Problem?
To say that a radiological error has occurred suggests that a mistake has been made. The supposition that follows is that a correct or better outcome was possible.7 Although some radiographic findings are obvious, there is a significant overlap between normal and abnormal. Presented with the same findings, even experienced radiologists may come up with different interpretations. A review by Robinson and colleagues8 examined interpretations of ED radiographs by three experienced radiologists. In reviewing skeletal, chest, and abdominal films, there was concurrence among all three radiologists in only 74%, 61%, and 51% of cases, respectively. When comparing pairs of observers and whether films were significantly abnormal, there was disagreement in 9-10% of skeletal, 11-19% of chest, and 8-18% of abdominal studies. Differences in image interpretation are not unique to radiology. A study reviewing cardiac catheterization results by cardiologists showed differences of opinion on relevant findings in 20-40% of cases.9
Definition of Error
With such differences of interpretation even among alleged experts, the notion of what constitutes an error is questioned. Some authors have suggested that error rates are “a measure of variance from the perfect result”.6 However, there are some errors made that all would agree are clearly mistakes, while others appropriately may be ascribed to variations of reasonable interpretation. Between the extremes of easy and difficult cases is a zone where accurate interpretations should be made that defines acceptable performance.7 What constitutes acceptable performance is often arbitrary and debatable, decided by individual institutions, professional societies, and judicial bodies.
How much error or variation in interpretation exists among experienced radiologists? The aforementioned study by Robinson and colleagues showed high levels of observer disagreement, and estimated a 3-6% error incidence per observer.8 A review of multiple imaging modalities interpreted by 35 community-based radiologists showed a mean error rate of 4.4%.10 Errors are hardly limited to the community setting. A study of CT scan readings at an academic medical center showed interobserver interpretation differences of 12%, and an overall error rate for individual interpretations of 7.6%.11 Another study also found a miss rate of 12% for findings deemed clinically relevant in abdominal trauma CT scans.12 While error rates will vary depending upon the nature of the examination, rates of 2-20% are consistently found.6 Considering that there are 12.5 million pediatric ED visits per year in the United States13 and that one-half of ED patients will have a radiograph performed,14 the potential number of errors is substantial.
The Nature of Error
As we consider how to reduce radiologic errors, we need to understand how and why those errors are made. Several classification schemes have proven useful.15-17 In any single case where a misinterpretation has occurred, several factors may be involv-ed. However, it is useful to try to isolate the various factors; this facilitates identifying areas for improvement. For purposes of this paper, the following categories will be discussed: technique, perception, knowledge and judgment, and communication.
Technique. Poor performance of the radiographic study can lead to substandard imaging and subsequent errors in patient care. A study of radiology malpractice claims showed that 10-30% of cases were due to faulty examination performance.18 Improper exposure, poor positioning, and inattention to detail all contribute to diminished conspicuity of abnormalities. In the pediatric setting, the ability of technologists to interact appropriately with children under stressful conditions to obtain their cooperation is critical to successful radiography. These issues also apply to performing more complicated studies such as CT imaging and ultrasound. Expensive equipment will produce worthless images if the child is not handled properly.
Errors in technique also include imaging the wrong part of the body, imaging it incompletely, or choosing the wrong imaging method. Ideally, a patient should have a physical examination prior to the ordering of any radiographic study.5 For some conditions there are well-defined clinical guidelines for when to image, and a thorough physical examination will refine which images are needed. An appropriately focused radiographic examination increases the likelihood that an abnormality will be perceptible. In general, two orthogonal views of any affected region should be performed to appropriately visualize and localize pathology. For long bone injuries, the proximal and distal joint also should be included to exclude concomitant joint injury or a second fracture.5
Perception. Once a properly performed study is completed, the abnormality must be observed and recognized to make an accurate diagnosis. A “true” miss occurs when a finding is visible but remains unnoticed by the reviewer, constituting a false-negative result. False-positive results, interpreting a normal radiograph as abnormal, are generally much less common.3,14,16,19 While it would seem that false-positive interpretations may lead to less adverse consequences than false-negative results, they can lead to unnecessary anxiety, and expose the patient to treatments and other diagnostic examinations that have cost and potential patient morbidity.
Causes of misperception are many, but may be grossly divided into issues intrinsic to the interpreter, and issues extrinsic to the interpreter himself that nonetheless affects his performance. Extrinsic factors include the quality of the examination performed, viewing conditions, speed of work, fatigue and stress, and the prevalence of the disease or condition being searched. There is a reason radiologists sit in dark rooms. Lesions with subtle contrast difference from normal structures will be visible only when there is little interfering background illumination. PACS systems are not immune to this problem; monitors can fade and drift over time, adversely affecting perceptual performance.20, 21 Radiologists tend to like quiet environments. It takes some effort and concentration to properly evaluate an image, and focused attention can be difficult in a chaotic setting. Clearly, this may be difficult to achieve in a busy ED, but an effort should be made to give some undivided time to the images when re-viewing them, as free from interruption as possible. Workloads and stress have been well documented to decrease clinical performance, and the same applies to radiological interpretation.11 In our emergency health care system, it is often impossible for physicians to control the amount of work they are expected to accomplish, but reasonable limits of performance should be kept in mind. Lastly, the prevalence of a condition, or the likelihood of a positive radiographic finding, affects the likelihood of missing a diagnosis. If a disease is likely, or the study is likely to show findings, then the chance for missing those findings is decreased, as opposed to studies with an unlikely positive yield.10
Intrinsic factors in misperception, those inherent in the physician, include issues of who should read ED studies, training and expertise, and patterns of search when reviewing images. So who should be reading? Setting aside ego, territory, money, and acrimony that may exist between radiology and ED physicians, ideally, radiologists would read all emergency imaging studies in a timely manner that actually would be useful to the treating physician.5,14 In most studies, radiologists generally perform better than non-radiologists, a fact that one would expect.3, 14, 22 However, a recent study by the American College of Radiology of 97 EDs23 revealed that after routine hours, 8% of EDs had no radiology coverage, and only 38% had the ability to consult with a radiologist about radiographic studies. Most EDs had CT scans read by radiologists at night, although 8% did not. Clearly, most ED physicians are on their own a great deal of the time, when staffing may be lower and workloads may be higher, conditions conducive to missing relevant findings.
Regardless of specialty, level of training plays an important role in proper perception of radiographic findings.3, 5, 7, 14, 19, 24 Attending physicians make fewer perceptual errors than residents (regardless of specialty), and non-radiology attending physicians often perform as well or better than junior radiology residents. As will be discussed below, certain types of injuries are missed more than others, allowing us to direct education toward common errors and to improve patient care.
So what makes an expert observer? One of the first things is a careful and systematic approach to the image.7, 12 Trainees often are taught specific patterns of search to ensure that all parts of an image are reviewed. While this is useful, many experienced radiologists are more flexible in how they view an image, but they still will make sure all parts of an image are examined. Sticking to a pattern helps to avoid “satisfaction of search”, the phenomenon that one tends to overlook additional findings after a significant finding has been made. One study suggests that complete search patterns improve perceptual performance, but a flexible approach based on clinical history and physical examination provides the best results.25
Knowledge and Judgment. The value of a radiological examination lies not just in observing a finding, but also in integrating that finding — whether positive or negative — with other patient data to make appropriate treatment decisions. As with perceptual ability, the ability to appropriately use radiological information improves with experience and training. Radiology studies should not be viewed as isolated information, but as part of the larger process of clinical decision making, of which radiology is one piece. Appropriate conclusions about radiographs come only with knowledge of the patient’s history and physical examination findings. The ability to draw conclusions from past experience and to incorporate new information and knowledge are characteristics of expert physicians.7
Many radiologists have been trained to view radiographs prior to knowing any patient information so as not to bias their interpretation. Although this approach may sound theoretically and academically desirable, it would be odd, indeed, for a clinician to adopt such a practice. The availability of prior radiographic studies and reports also is valuable to the diagnosing physician and can improve confidence in diagnoses. 26 Studies show that knowledge of accurate patient clinical information improves the rate of true positive diagnoses, without a concomitant increase in false-positive readings (e.g., over-calls or biased reports).7, 25-30 Similarly, inaccurate information had a detrimental effect in a study of CT scan interpretation.30 Accurate quality clinical information regarding a patient facilitates the job of the radiologist and clinician to make an accurate diagnosis.25, 31
Communication. A brilliant diagnosis does little good if made too late or if not communicated properly. As previously discussed, many EDs are unsupported by radiologists after hours,23 potentially delaying diagnoses. In such situations, some mechanism for review and comparison of conclusions is necessary, as is a method for reliable communication of discrepant interpretations.14 Radiological reports should not only list findings, but also attempt to draw relevant conclusions and diagnoses when possible. Again, the quality of the diagnostic output partly depends upon the quality of the clinical input. Reports should eschew obfuscation and use plain language with as little ambiguity as possible. Ideally, there is direct and contemporaneous communication between radiologist and emergency physician about important or complex studies, with one author arguing “…the radiological process has not been completed until such a meeting has taken place….”.6
Common Pediatric Radiological Errors
Skeletal Errors. In the ED, most radiological misses are fractures, accounting for 26.9-79.7% of missed cases in a review of a general ED3, 14 and 69% of missed cases in a large dedicated pediatric ED.19 Most pediatric skeletal misses involve buckle fractures, physeal injuries, and avulsion fractures.5, 14, 19, 32 Al-though sometimes subtle, many of these injuries are readily apparent on subsequent review.3 For practitioners who infrequently image children and those with minimal radiologist support, obtaining comparison views of the nonsymptomatic contralateral extremity helps to avoid missing subtle injuries and overcalling normal skeletally immature anatomy as abnormal.33 To understand the radiographic appearance of pediatric fractures, having a concept of the unique fracture types in children is useful.34
Pediatric Fracture Types. Plastic bowing deformity is an injury unique to children due to relative bony plasticity. Most commonly occurring in the ulna and radius, bowing results from microscopic fractures on the convex side of the diaphysis, which do not return to the previous form when the injuring force is ended.34
Buckle or torus fractures commonly occur in the long bone metaphyses from axial loading forces. Buckle fractures may result in mild angular deformity, or may only demonstrate minimal angulation or irregularity of the outer bony cortex. Common sites include the proximal and distal radius, distal ulna, proximal and distal tibia, and phalanges.
Greenstick fractures are the result of continued force beyond that producing bowing deformity. With continued force, the convex side of the diaphysis fails, producing a fracture through only one side of the cortex. The radius and tibia are common sites.
Complete fractures extend completely through the bone, with or without angular deformity or displacement. Fractures may be transverse, oblique, or spiral depending upon the causative force.34
Injuries about the physis are characterized by the Salter-Harris system. The five classic types include: I – injury to the physis only, II – injury involving the physis and metaphysis (most common type, present in 75% of physeal injury), III – injury involving the physis and epiphysis, IV – transphyseal injury involving the epiphysis and metaphysis, and V – crush injury of the physis. Type V and non-displaced type I injuries are difficult to diagnose,35 contralateral views may be useful, but often these injuries are diagnosed on clinical grounds. The other types generally are visible on routine studies, although oblique images often are useful to detect subtle non-displaced injuries.
Specific Injuries
Buckle Fractures. The margins of the bony cortex always follow a smooth arc, no matter what the radiographic view. The cortical margins should be followed along their entire course on all views. Any interruption of this smooth line should be viewed with suspicion,36 particularly if it is associated with soft-tissue swelling and local pain (Figure 1).
 |
Figure 1. Buckle Fracture Lateral view of the wrist shows buckling of the dorsal cortex of the distal radius (arrow), with more subtle cortical irregularity of the dorsal ulna (arrowhead), both indicating buckle fractures. |
Bowing Deformity. These injuries may be rather subtle, but can result in significant deformity and functional impairment if not recognized (Figure 2). Bowing deformities of the ulna are often associated with radial head dislocation, a variant of a Monteggia injury (discussed below).37
 |
Figure 2. Bowing Deformity and Greenstick Fracture Frontal view of the forearm shows a greenstick fracture involving only the lateral cortex of the radius (arrow). There is marked lateral bowing of the ulna (arrowheads) without a demonstrable focal fracture. |
Proximal Radius Injuries. Radial head and neck fractures often occur from a fall on an outstretched hand or are associated with elbow dislocations. Radiographic findings may be subtle, with minimal cortical irregularity present. Dedicated oblique radial head views often are useful (Figure 3).
 |
Figure 3. Radial Neck Fracture Oblique radial head view shows focal angulation of the radial neck (arrow) indicating a fracture. The routine lateral view (not shown) showed an elbow joint effusion, but no fracture. |
Radial head subluxation may occur with nursemaid’s elbow, but radiographs are usually normal as reduction occurs either before imaging or during the process of positioning the patient for radiographs. In Monteggia fractures, there is fracture of the ulna and dislocation of the radius at the radiocapitellar joint. A line drawn through the long axis of the radius should intersect the capitellum on any radiograph (Figure 4). If it does not, radial head dislocation has occurred.
 |
Figure 4. Monteggia Fracture Dislocation Lateral view of the elbow shows a joint effusion (arrowheads) and fracture of the proximal ulna (arrow). The radial head (R) is dislocted from its normal articulation with the capitellum (C). |
Distal Humerus Injuries. Injury to the distal humerus is very common in children after falling on an outstretched arm. Supracondylar fractures account for 60% of elbow fractures in children38 and are classified into three groups: Type I fractures– non-displaced or minimally posteriorly angulated, Type II fractures – posteriorly displaced but with an intact posterior cortex, and Type III fractures – completely displaced fractures. The anterior humeral line normally should bisect the capitellum on a lateral elbow view, allowing evaluation of posterior displacement (Figure 5). Good quality images with a true lateral view are essential to accurate diagnosis and classification,32 and significantly affects treatment decisions.
 |
Figure 5. Type II Supracondylar Fracture Lateral view of the elbow shows a joint effusion (arrowhead) and a fracture of the distal humerus. The capitellum (C) is displaced posteriorly from the anterior humeral line. |
Elbow joint effusions are seen as uplifting of the normal anterior and posterior fat pads from the olecranon fossa. Even if a fracture is not appreciated, an occult supracondylar fracture should be assumed to be present. Although a recent study demonstrated that a minority of these children actually have supracondylar fracture,39 prudence still demands immobilization and appropriate subspecialty follow-up.
Medial epicondylar avulsion is a common injury resulting from contraction of the forearm flexor muscles. Radiographically, there is usually soft-tissue swelling and separation of the epicondyle from the medial humeral metaphysis resulting from avulsion through the cartilaginous apophysis. Occasionally, the avulsed medial epicondyle may be displaced into the joint and mimic the trochlear ossification center (Figure 6). This can produce a relatively normal radiograph, unless one knows that the medial epicondyle normally begins to ossify two years prior to the trochlea. The presence of what looks like a trochlear ossification center with an absent medial epicondyle should alert one to the diagnosis.
 |
Figure 6. Medial Epicondylar Avulsion
with Entrapment Frontal radiograph of the elbow shows what appears to be a trochlear ossification center (arrow), but is actually the medial epicondyle displaced into the joint. There is marked medial soft-tissue swelling and absence of ossification at the expected location of the medial epicondyle (arrowhead). |
Proximal Humerus and Shoulder. Injuries to this region are seldom radiologically problematic. Some confusion may arise out of the appearance of the normal proximal humeral physis, which can mimic a fracture. Additionally, normal irregular ossification centers of the coracoid, acromion, and distal clavicle also can be misinterpreted as fractures.
Cervical Spine. Radiographic evaluation of the pediatric cervical spine can be an anxiety provoking experience, and may result in the failure to recognize normal findings. For the infrequent observer, fortunately, there are a number of useful signs to separate the normal from abnormal pediatric spine. Prior to the age of approximately 8 years, the biomechanics of the pediatric spine differ from older children and adults, and most cervical injuries occur from the occiput to C4. Pseudosubluxation of C2 on C3 is common, but can be differentiated from a true injury through the use of the cervicolaminar line. The cervicolaminar line is drawn from C1 to C3 and should pass within 1 mm of C2 (Figure 7). Another common pediatric normal variant, the C3 vertebral body commonly is flattened along its anterosuperior margin as well, mimicking a compression injury.
 |
Figure 7. Cervical Pseudosubluxation Lateral view of the cervical spine shows anterior subluxation of C2 on C3. However, the posterior cervicolaminar line (white line) from C1 to C3 intersects C2, indicating that no abnormal displacement is present. Also note the normally rounded anteriosuperior aspects of C3 and C4, and the lucency at the base of the odontoid (arrow) representing the normal synchondrosis. |
Odontoid fractures do occur in children, but care should be taken to avoid calling the normal dens synchondrosis a fracture, especially because the prevertebral soft-tissue thickness is a less reliable indicator of injury in children than in adults.
Pelvis and Proximal Femur. Slipped capital femoral epiphysis (SCFE) is the most common adolescent orthopedic hip disorder, particularly in overweight patients. Diagnostic delays of up to 10 months are common, and the initial examiner misses the diagnosis in 30-50% of cases.40 Frontal pelvis radiographic findings include asymmetry of the physes and possible displacement of the epiphysis. The frog leg lateral view is the essential image, which more reliably shows the displaced femoral head (Figure 8). A line drawn along the lateral femoral neck (Klein’s line) should intersect the lateral portion of the epiphysis; if it does not, SCFE should be suspected.
|
Figure 8. Slipped Capital Femoral Epiphysis A: Frontal view of the pelvis in an obese teenage boy (note the pannus) with right hip pain shows widening of the right proximal femoral physis (arrow) relative to the left side. A line drawn along the lateral femoral neck intersects some of the left epiphysis, but not the right, indicating epiphyseal displacement. B: The right frog lateral view shows this much better, clearly demonstrating the posteromedial displacement of the femoral epiphysis (arrow). |
The pelvis serves as an attachment for several large muscle groups. In athletic adolescents, chronic stress or sudden contraction may produce avulsions of these muscular attachments. The most common of these injuries is avulsion of the hamstrings from the ischial tuberosity, followed by avulsion of the sartorius and tensor fascia lata from the anterior superior iliac spine, and the rectus femoris from the anterior inferior iliac spine.41 Acutely, these injuries may show minimal displacement and often are missed (Figure 9).42 Oblique views, magnetic resonance imaging (MRI), or close follow-up to detect healing can be useful.41
 |
Figure 9. Pelvic Avulsion Fracture Frontal view of the right hip in a teenage sprinter with acute pain shows an avulsed anterior inferior iliac spine (arrow). |
Lower Leg and Foot. Toddlers and young children are prone to injuries of the lower leg and foot, which are sometimes difficult to localize clinically. The classic toddler’s fracture is a non-displaced oblique fracture in the distal tibial diaphysis, which may be visible only on oblique views (Figure 10), or may be radiologically occult.43 Follow-up films for continued pain will disclose periosteal new bone formation within 7 to 10 days after injury.
 |
Figure 10. Tibial Toddler’s Fracture Frontal view of the lower leg shows an oblique nondisplaced fracture of the distal tibia (arrowheads). |
Other less appreciated injuries also may occur. Hyperextension of the lower leg can result in subtle anterior buckle fractures in the region of the tibial tubercle.44 Jumping on a plantar flexed foot produces a longitudinal loading force upon the metatarsals, usually proximally.43 This so-called “bunk bed” fracture usually involves the first metatarsal, and the fracture may be visible only on an oblique view (Figure 11).
|
|
Figure 11. First Metatarsal “Bunk-Bed” Fracture Frontal view of the foot of a 3-year-old child with pain after jumping shows a nondisplaced fracture of the base of the first metatarsal (arrow). |
Through a similar mechanism, the cuboid may be compressed and fractured between the fourth and fifth metatarsals and the calcaneous. These injuries are subtle, probably underdiagnosed, and may appear only as a band of sclerosis during healing (Figure 12).43
 |
Figure 12. Cuboid “Nutcracker” Fracture Lateral view of the foot in a child with continued foot pain after a jump from a height shows a band of sclerosis at the proximal pole of the cuboid (arrow) indicating a healing fracture. |
Child Abuse
Fifteen to thirty percent of fractures in children younger than 3 years are due to physical abuse,45 and up to 55% of abused children will have fractures.46 Most abuse-related fractures (e.g. rib fractures and metaphyseal corner fractures) are highly specific for inflicted injury. Multiple fractures, fractures in different stages of healing, and fractures out of proportion to the explanation of injury also are useful clues.
Rib fractures tend to occur at the posterior costovertebral junctions and the lateral chest as a result of the chest being squeezed.48 Acute fractures are difficult to detect if not displaced, and usually are diagnosed as they heal, producing recognizable callus. Remember to look at the bones on all chest radiographs, as victims of abuse may present with vague or misleading histories, and fractures may be visible on otherwise routine chest images (Figure 13).
 |
Figure 13. Child Abuse Frontal chest radiograph in an infant brought to the ED for irritability who was noted to have unusual bruising shows multiple posterior and lateral rib fractures (arrowheads) with healing callus. |
Metaphyseal corner fractures are seen most commonly in the lower extremities and result from twisting of the extremity. These fractures are subphyseal and parallel the physis before exiting at the metaphyseal corners. Depending upon the radiographic projection, these injuries appear as corner fractures or so-called bucket-handle fractures (Figure 14).49
 |
Figure 14. Child Abuse Frontal view of the left knee shows irregularity of the distal femoral metaphyseal corners (arrows) indicating fracture. The slightly angled view of the proximal tibia shows a crescentic ossific fragment (arrowhead), the so-called “bucket-handle” fracture. Note the periostitis along the lateral cortices of the femur and tibia indicative of healing. |
Airway and Chest
After skeletal injuries, Halsted and colleagues19 found misdiagnoses involving the airway and chest were most common. The young child’s pharynx and trachea are quite pliable, which can produce confusing radiographic images. In expiration, the trachea appears buckled and tortuous, and the retropharyngeal soft tissues abnormally thickened. Proper imaging with the head and neck extended, and the child in full inspiration will minimize these effects (Figure 15). If concern still exists, airway fluoroscopy provides a quicker and less expensive alternative to CT imaging, and exposes the child to less ionizing radiation.
 |
Figure 15. Effect of Inspiration on Retropharyngeal Soft Tissues A: Lateral view of the neck in a child with pain and fever shows minimal air in the hypopharynx (H) and thickening of the retropharyngeal soft tissue (*). |
 |
B: Repeat examination with better inflation of the hypopharynx and a normal prevertebral space. |
Pneumonia can be difficult to diagnose, particularly in the middle and lower lobes19 of young children. Normal right middle lobe bronchovascular markings may be overcalled on frontal radiographs (Figure 16), but the lateral view should clarify that the lungs are clear. Airspace disease behind the heart or deep in the costophrenic sulci may be difficult to observe. Children with lower lobe pneumonias may present with abdominal pain, and abdominal radiographs often may reveal the pneumonias better than chest images.
 |
Figure 16. Normal Chest Frontal view of the chest in an infant with fever shows normal right middle lobe bronchovascular markings (arrow); the lateral image was entirely normal. Note the normal wavy borders of the thymus (arrowheads). |
The cardiac and mediastinal contour is normally relatively wide in young children due to thymic tissue, so care must be taken in overcalling cardiomegaly. As the thymus lies in the anterior mediastinum, a lateral view will show a normal anteroposterior dimension of the heart.
Conclusion
Diagnostic radiology is an integral part of the evaluation of many pediatric ED patients. As in any human endeavor, error is possible. However, understanding the factors involved in the radiological process —from ordering the appropriate examination, to performing it well, to proper interpretation and integration with other clinical data, to communication of results— allows us to minimize mistakes. Working in concert with one another, providing mutual education and support, and openly reporting and analyzing errors when they occur, radiologists and emergency physicians have the opportunity to facilitate accurate diagnoses, which ultimately will improve patient care.
References
1. Berlin L, Hendrix R. Perceptual errors and negligence. AJR 1998;170: 863-867.
2. Garland L. On the scientific evaluation of diagnostic procedures. Radiology 1949;52:309-328.
3. Guly H. Diagnostic errors in an accident and emergency department. Emerg Med J 2001;18:263-269.
4. Shively C. Quality management in radiology. Imaging Economics 2003; November:63-70.
5. Touquet R, Driscoll P, Nicholson D. Teaching in accident and emergency medicine: 10 commandments of accident and emergency radiology. BMJ 1995;310:642-645.
6. Goddard P, Leslie A, Jones A, Wakeley C, Kabala J. Error in radiology. Br J Radiol 2001;74:949-951.
7. Robinson P. Radiology Achilles’ heel: error and variation in the interpretation of the Rontgen image. Br J Radiol 1997;70:1085-1098.
8. Robinson P, Wilson D, Coral A, et al. Variation between experienced observers in the interpretation of accident and emergency radiographs. Br J Radiol 1999;72:323-330.
9. Banerjee S, Crook A, Dawson J, et al. Magnitude and consequence of error in coronary angiography and interpretation (the ACRE study). Am J Cardiol 2000;85:309-314.
10. Siegle R, Baram E, Reuter S, et al. Rates of disagreement in imaging interpretation in a group of community hospitals. Acad Radiol 1998;5:148-154.
11. Bechtold R, Chen M, Ott D, et al. Interpretation of abdominal CT: analysis of errors and their causes. J Comput Assist Tomogr 1997; 21:681-685.
12. West O, Anderson J, Lee J, et al. Patterns of diagnostic error in abdominal CT. Emeg Radiol 2002;9:195-200.
13. O’Neill K, Molczan K. Pediatric triage: a 2-tier, 5-level system in the United States. Pediatr Emerg Care 2003;19:285-290.
14. Saab M, Stuart J, Randall P, et al. X-ray reporting in accident and emergency departments - reducing errors. Eur J Emerg Med 1997;4:213-216.
15. Fitzgerald R. Error in radiology. Clin Radiol 2001;56:938-946.
16. Renfrew D, Franken EJ, Berbaum K, et al. Error in radiology: classification and lessons in 182 cases presented at a problem case conference. Radiology 1992;183:145-150.
17. Smith M. Error and Variation in Diagnostic Radiology. Springfield: Charles C. Thomas; 1967.
18. Brenner R, Lucey L, Smith J, et al. Radiology and medical malpractice claims: a report on the practice standards claims survey of the Physicians Insurers Association and the American College of Radiology. AJR 1998; 171:19-22.
19. Halsted M, Kumar H, Paquin J, et al. Diagnostic errors by radiology residents in interpreting pediatric radiographs in an emergency setting. Pediatr Radiol 2004;34:331-336.
20. Ikeda M, Ishigaki T, Shimamoto K, et al. Influence of monitor luminance change on oberver performance for detection of abnormalities depicted on chest radiographs. Invest Radiol 2003;38:57-63.
21. Herron J, Bender T, Campbell W, et al. Effects of luminance and resolution on observer performance with chest radiographs. Radiology 2000;215:169-174.
22. Eng J, Mysko W, Weller G, et al. Interpretation of emergency department radiographs: a comparison of emergency medicine physicians with radiologists, residents with faculty, and film with digital display. AJR 2000;175:1233-1268.
23. Saketkhoo D, Bhargavan M, Sunshine J, Forman H. Emergency department image interpretation services at private community hospitals. Radiology 2004;231:190-197.
24. Hillier J, Tattersall D, Gleeson F. Trainee reporting of computed tomography examinations: do they make mistakes and does it matter? Clin Radiol 2004;59:159-162.
25. Peterson C. Factors associated with success or failure in radiological interpretation: diagnostic thinking approaches. Med Educ 1999; 33:251-259.
26. Aideyan U, et al. Influence of prior radiographic information on the interpretation of radiographic examinations. Acad Radiol 1995;2:205-208.
27. Berbaum K, el-Khoury G, Franken EJ, et al. Impact of clinical history on fracture detection with radiography. Radiology 1988;168:507-511.
28. Berbaum K, Franken EJ, Dorfman D, Lueben K. Influence of clinical history on perception of abnormalities in pediatric radiographs. Acad Radiol 1994;1:217-223.
29. Doubilet P, Herman P. Interpretation of radiographs: effect of history. AJR 1981;137:1055-1058.
30. Leslie A, Jones A, Goddard P. The influence of clinical information on the reporting of CT by radiologists. Br J Radiol 2000;73:1052-1055.
31. Swett H. The power of knowledge in radiologic education and decision making. J Digital Imag 1991;4:199-201.
32. Paterson J. Children’s fractures “not to be missed”. Hosp Med 2002; 63: 426-428.
33. Swischuk L, Hernandez J. Frequently missed fractures in children (value of comparative views). Emerg Radiol 2004;11:22-28.
34. Jones E. Skeletal growth and development as related to trauma. In: Green N, Swiontkowski M, eds. Skeletal Trauma in Children. Philadelphia: W.B. Saunders Company; 1994:1-13.
35. Perron A, Miller M, Brady W. Orthopedic pitfalls in the ED: pediatric growth plate injuries. Am J Emerg Med 2002;20:50-54.
36. Hernandez J, Swischuk L, Yngve D, et al. The angled buckle fracture in pediatrics: a frequently missed fracture. Emerg Radiol 2003;10:71-75.
37. Kemnitz S, De Schrijver F, De Smet L. Radial head dislocation with plastic deformation of the ulna in children: a rare and frequently missed condition. Acta Orthopaedica Belgica 2000;66:359-362.
38. Wu J, Perron A, Miller M, Powell S, et al. Orthopedic pitfalls in the ED: pediatric supracondylar humerus fractures. Am J Emerg Med 2002;20:544-550.
39. Donnelly L, Klostermeier T, Klosterman L. Traumatic elbow effusions in pediatric patients: are occult fractures the rule? AJR 1998: 243-245.
40. Perron A, Miller M, Brady W. Orthopedic pitfalls in the ED: Slipped capital femoral epiphysis. Am J Emerg Med 2002;20:484-487.
41. Pope L. Avulsion and chronic avulsive-type injuries to the skeleton: part I. Contemporary Diagnostic Radiology 2003;26:1-6.
42. Gidwani S, Jagiello J, Bircher M. Avulsion fracture of the ischial tuberosity in adolescents - an easily missed diagnosis. BMJ 2004; 329:99-100.
43. John S, Moorthy C, Swischuk L. Expanding the concept of the toddler’s fracture. RadioGraphics 1997;17:367-376.
44. Swischuk L, John S, Tschoepe E. Upper tibial hyperextension fractures in infants: another occult toddler’s fracture. Pediatr Radiol 1999;29:6-9.
45. Oral R, Blum K, Johnson C. Fractures in young children: are physicians in the emergency department and orthopedic clinics adequately screening for possible abuse? Pediatr Emerg Care 2003;19:148-153.
46. Taitz J, Moran K, O’Meara M. Long bone fractures in children under 3 years of age: is abuse being missed in emergency department presentations? J Paediatr Child Health 2004;40:170-174.
47. Scherl S, Miller L, Lively N, Russinoff S, Sullivan C, Tornetta P. Accidental and nonaccidental femur fractures in children. Clin Orthop 2000;376:96-105.
48. Kleinman P. Bony thoracic trauma. In: Kleinman P, ed. Diagnostic Imaging of Child Abuse. 2nd ed. St. Louis: Mosby; 1998:110-148.
49. Kleinman P. Skeletal trauma: general considerations. In: Kleinman P, ed. Diagnostic Imaging of Child Abuse. 2nd ed. St. Louis: Mosby; 1998:8-25.
50. Schlesinger A. Pitfalls in the interpretation of pediatric chest and airway radiographs. Curr Probl Diag Radiol 1998;27:73-101.
The emergency department physician is the critical link between the patient and the diagnostic tests, including radiographs, that are ordered. This article provides information on common radiographic errors to help clinicians improve their accuracy, confidence and subsequent patient care.
Subscribe Now for Access
You have reached your article limit for the month. We hope you found our articles both enjoyable and insightful. For information on new subscriptions, product trials, alternative billing arrangements or group and site discounts please call 800-688-2421. We look forward to having you as a long-term member of the Relias Media community.