A medical error is not an unusual event, although most are minor. The emergency department is one place where medical errors are more likely because of the chaotic nature. Here's a look at common mistakes and some cases where mistakes were made.
Every physician has likely committed a medical error at some time.1 Fortunately, most medical errors are minor, and many do not reach the patient. However, one recent study suggested that medical errors are the third leading cause of death in the United States.2 In a study of internal medicine trainees, 45% reported being involved in a serious medical error.3 In a survey of practicing physicians, 10.5% reported they were involved in a major medical error during the prior 3 months.4 Provider burnout has been linked to medical errors. Medical errors often lead to burnout, and the reverse is also probable.5,6,7
The Emergency Department (ED) is a high-risk area where medical errors are likely. This is, in part, because the ED is often extremely busy with multiple ill or complicated patients. Emergency clinicians often must make rapid decisions in a chaotic environment, so the chance for error is high. (This is not entirely unique to the ED. Pediatricians in office practice may find their office fits this mold on many days!) Many errors in pediatric emergency medicine relate to physician fatigue. Others relate to frequent distractions that are prevalent in the ED setting. In one study,8 attending physicians and nurses were followed in a level 1 trauma center. It was noted that physicians were interrupted 10 times within an hour, and nurses were interrupted 12 times in an hour. These providers performed between1-8 other activities before returning to the original task. They were interrupted by phones, pagers, other staff members. It is easy to see how errors can result from these frequent disruptions. In some cases, ED providers commit cognitive errors, perhaps because of “premature closure.” That is, the clinician fixes on a diagnosis early on and ignores other findings that could lead to a different diagnosis.
In many cases, errors result from poor communication. In the ED it is obviously crucial for staff to work as a cohesive team. Certainly, ED providers must communicate well with nursing staff, and consultants. Numerous errors result when there is a breakdown in this communication. In fact, the Joint Commission notes that communication issues are the third leading cause of sentinel events.9 Likewise, ED providers must communicate well with patients and families to obtain an accurate history and develop a management plan together.
Although this seems obvious, it is not guaranteed. In one interesting study10 involving 2 teaching hospitals in Michigan, patients were asked about 4 domains soon after their ED visit:
1.Do you understand your diagnosis?
2. Do you understand what was done for you in the ED?
3. Do you know what you are supposed to do after your ED visit (discharge instructions).
4. Do you understand what symptoms should prompt you to return to the ED?
Surprisingly, 78% of those surveyed were deficient in recalling one of these domains and 51% were deficient in 2 or more domains. One could interpret this as a failure to communicate! Clearly, there is room for improvement.
In the ED, procedures can also go badly due to clinician error or failure to monitor the patient carefully. Even more common are ED diagnostic errors.
Table 1 summarizes common causes of medical errors in the ED; again, office practices are not immune to these. Whether you practice in a pediatric emergency department or an office setting, the following examples (all real cases) offer good teaching points. It is hoped these cases will help clinicians avoid similar errors in the future.
A 15-year-old boy presented to the ED with the complaint of chest pain for 1 week. He reported that he was hit over his ribs on the right side of his chest when playing football a week ago. His pain, described as achy and stabbing, worsened in the past few days. It was worse with walking and for one day, he had difficulty climbing the stairs. He was short of breath at times and had mild dizziness with standing. He had no fever, no vomiting or diarrhea, no cough or rhinorrhea. Physical examination revealed temperature 100.9° F, heart rate 78 beats per minute (up to 106 beats per minute when standing), respiration 24/minute, blood pressure 128/84 (124/92 standing). He was alert and talkative and in no distress. He had mild tenderness over his sternum. Cardiac exam was normal. Lungs were clear. Abdomen was benign. His exam was otherwise unremarkable.
A chest x-ray showed clear lungs; the heart was noted to be “top normal” in size. (see radiograph Figure 1). An electrocardiogram had ‘non-specific’ findings such as ST elevation in some leads and borderline prolonged QT interval.
The patient received ibuprofen for his pain, and intravenous saline because it was thought he may be dehydrated. His vital signs remained unchanged after the fluids. A cardiologist was consulted by phone. The cardiologist believed the patient had musculoskeletal pain, and he did not need immediate cardiology evaluation. The patient was discharged to home. About 8 hours later, he collapsed and died. Myocarditis was found on autopsy.
Teaching points: In retrospect, the diagnosis of myocarditis is obvious. The patient had classic findings: chest pain for several days, worsening with exertion and leading to shortness of breath. He also had a low-grade fever in the ED, that seems to have been ignored by the cardiologist who was consulted. Myocarditis can be a difficult diagnosis, and in many cases the patient is evaluated multiple times before myocarditis is considered.11
Chest pain associated with fever is concerning. If pneumonia is ruled out, consider myocarditis as the etiology for pain. Chest pain with exertion is concerning for a cardiac etiology rather than musculoskeletal pain, which should improve over time, not worsen. Pay attention to the vital signs! If the patient continues to have tachycardia (or dizziness) with standing, after intravenous fluids, perhaps there is cardiac insufficiency rather than dehydration.
A 5-year-old girl presented to the ED with persistent vomiting. She was seen by her pediatrician who obtained a rapid throat swab that was positive for Group A Streptococcus. She also had an abdominal radiograph that showed constipation. She was referred to the ED because her emesis had a foul odor. It smelled of ‘fecal material.’ In the ED, it was noted that she had been vomiting intermittently for about 3 months, but had 8-9 episodes today. Her emesis was ‘greenish-brown’ and there was no blood in her stool. She had crampy abdominal pain. Physical exam revealed temperature 98.9° F, heart rate 96 beats per minute, respiratory rate 20/minute, blood pressure 106/66. She was alert, well-nourished, with an injected pharynx. Her chest was clear; abdomen was soft and unremarkable.
The patient was given an enema and was able to tolerate oral fluids. She was discharged to home with a diagnosis of constipation, and strep throat. The next day, she had a generalized seizure and was brought back to the hospital. A computed tomography (CT) scan of her brain was promptly obtained and showed a large brain tumor. The child expired the next day.
Teaching points: Vomiting has many etiologies. Viral gastroenteritis is perhaps the most common etiology. However, when a child has vomiting without diarrhea, other conditions must be considered. Vomiting is sometimes associated with strep pharyngitis, and perhaps the clinicians in this case were misled by the rapid test of the pharynx. Constipation is unlikely to cause persistent vomiting-it is wise to look for another etiology before concluding that constipation is the culprit. Diabetic ketoacidosis and pneumonia are other possibilities, but unlikely in the case above. A viral infection, gastritis without diarrhea, is possible. However, clinicians must also consider an intracranial cause for vomiting, such as a brain tumor. In this case, a neurologic exam was not even documented in the medical record. Her history of prolonged vomiting over several months was apparently ignored.
A 2-year-old girl was bitten by a dog. She presented to the ED with extensive facial wounds. A plastic surgeon was consulted, and the face lacerations were meticulously repaired. She was discharged to home. She returned to the ED a few days later because of frequent vomiting and low-grade fever. A head CT scan was obtained and showed a brain abscess. With careful inspection, it was noted there was a defect in the patient’s skull. The dog’s tooth had apparently penetrated the child’s skull, leading to intracranial infection. The skull defect was not noted under the toddler’s extensive hair and thus it was not cleaned or addressed at the initial visit.
Teaching points: When an injured child arrives, clinicians often focus on the most obvious injury, or perhaps the most impressive. It is wise to examine the rest of the child first. No one will miss the dramatic wound, so a careful look elsewhere is advised before devoting attention to the noticeable face laceration. Also, consider that when a big dog attacks a little child, the jaws of the animal can likely encompass a large area of the child. Look for a second wound!
A 2-year-old boy presented to the ED with the complaint of “something in his nose.”12 Despite multiple attempts, the physician’s assistant and the physician were unable to remove a foreign body. In fact, they were not able to clearly visualize the foreign body, believing it was deeply embedded in the nostril. The providers recommended that the child see a specialist in the next few days. When the child was evaluated by an otolaryngologist, a radiograph was obtained. (Figure 2) The rim of the metallic foreign body is evidence that it is a ‘button battery’. This child had a complicated course with perforation of the nasal septum, and he required extensive surgery.
Teaching points: Button batteries in the nose, ears, or esophagus can cause extensive burns and tissue necrosis in a short period of time.13 It is important to visualize a foreign body in these locations. A plastic bead can likely wait for the specialist appointment; a button battery cannot. It is wise to consider a radiograph to identify the presumed foreign body if it cannot be seen on physical examination.
On a warm summer day, a 15-year-old girl fell when playing baseball and hit her head on the turf.14 She was briefly unresponsive with “twitching” movements. She was brought to the ED, was noted to have a scalp laceration, but her exam was otherwise unremarkable. Routine blood tests were obtained, and her scalp laceration was repaired. Her syncope was thought to be related to heat exhaustion. The patient did well until about a year later. She then had a brief syncopal episode when she was doing laundry. She dropped to her knees, was briefly unresponsive and recovered quickly. She was brought to the ED where her exam was unremarkable. Routine blood studies were again unremarkable. She was diagnosed with vasovagal syncope and discharged to home.
One year later, the patient again collapsed during baseball. She was brought to the ED and found to have Long QT syndrome. Unfortunately, she suffered a cardiac arrest and after resuscitation, had a severe brain injury.
Teaching Points: Exercise related syncope is concerning. Consider a cardiac etiology, like Long QT syndrome.15 Routine blood tests are not often helpful with a syncopal episode, but many recommend obtaining an electrocardiogram in most cases of syncope, and for all those related to exercise. Some syncopal episodes can resemble a seizure. This patient had brief ‘twitching’ movements. But unlike a seizure, the patient with cardiac related syncope often recovers quickly, without a post-ictal state. Also, a syncopal episode that results in an injury should raise a red flag. Most children with vasovagal syncope have some warning of the event and fall slowly to the ground. Significant injury is less likely in such cases. With a cardiac arrhythmia, the patient has no warning and may fall to the ground with more force.
A 7-year-old boy developed abdominal pain at school. After being sent home, his parents brought him to his pediatrician for evaluation. The physician noted right lower quadrant pain, but no other concerning findings. The family was instructed to take the patient home and give supportive care. The next day, his parents called the pediatrician to inform him that their son’s pain had worsened. The pediatrician recommended an over-the-counter analgesic and follow up if he did not improve. As the pain continued to worsen, the boy was brought to the ED, where a testicular exam revealed a swollen, tender testicle. He was found to have testicular torsion. The testicle was noted to be necrotic in the operating room and it was surgically removed. After review of the initial evaluation by the pediatrician, it was noted that a testicular exam was not performed.
Teaching Points: This child did not initially present to the ED; however, the case provides an important lesson for ED clinicians and pediatricians in office practice – always perform a testicular exam for males with abdominal pain! Many pathologies can manifest or be interpreted as abdominal pain by a young child. In fact, several studies show that abdominal pain can be the presenting symptom of testicular torsion. One study16 found that 9/76 (12%) patients with testicular torsion presented with acute abdominal pain as their only symptom. Of those 9 patients, 7 did not have a testicular exam initially and the diagnosis of torsion was delayed. Similarly, the child in our case had referred pain to the abdomen from his testicular pathology, therefore, the diagnosis was initially missed, which likely led to his poor outcome. Also keep in mind, even if pain is not referred, a young boy may have difficulty articulating where the pain is originating from or may be embarrassed to discuss this. Therefore, it is very important to consider testicular torsion for any boy with vague complaints, especially abdominal pain.
A 23-day old baby presented to the ED with excessive crying. A brief history was notable for premature birth at 36 weeks and the birth weight was 2778 grams Examination of the baby was documented as grossly normal, except the baby’s weight in the ED was 2268 grams. The infant fed in the ED and his crying improved. He was discharged to home and the parents were instructed to follow-up with his pediatrician if there were any concerns. After arriving home, the baby continued crying. His parents brought him to a different ED where the baby was diagnosed with severe dehydration. When receiving an intravenous fluid bolus, the infant developed cardiovascular collapse. Cardiopulmonary resuscitation was initiated, and he was stabilized. However, the baby suffered severe brain damage and subsequent developmental delay. Close review of the infant’s record revealed multiple factors that likely contributed to his poor outcome, including the mother’s perinatal drug use and child neglect, leading to malnourishment.
Teaching Points: This case reminds us to carefully consider an infant’s social situation before discharge from the ED. In addition, it is essential to always review and interpret the infant’s weight and growth curve. In this case, it is unclear if any of the original ED providers took note that this infant’s documented weight was 20% below his birth weight. If that association were made, perhaps the physician would have recognized a bigger issue, such as severe dehydration due to neglect and poor feeding. This may have prompted more aggressive care, hospital admission, as well as support and education for the family. Taking a few extra moments to review the weight and growth curve can clue the provider into the whole picture and help make more informed decisions.
A 14-year-old male presented to the ED after suffering a left knee injury when he was playing basketball. He was evaluated, radiographs were obtained, and no fracture was identified. He was discharged to home with a knee immobilizer, crutches, and pain medication. The following day, he had significant, worsening pain and he was brought to a different ED. At that time, he was found to have absent distal pulses in left lower extremity. It was discovered that his left popliteal artery was transected, and he required emergent fasciotomy due to compartment syndrome. He subsequently needed multiple surgeries for debridement of necrotic tissue and ultimately underwent left lower extremity amputation. Upon review of the case, it was found that the initial triage notes by nursing staff documented “+3 swelling of left leg and pedal pulses not discernable by palpation.” Based upon documentation, it was apparent that the nurse’s concerning findings were not read by nor discussed with the ED physician.
Teaching Points: This is a very sad, unfortunate case that exemplifies how important communication is in the ED, as well as the importance of a thorough musculoskeletal exam following an extremity injury. Had there been good communication regarding lack of pulses noted by the nurse, one of the cardinal “5-Ps” of compartment syndrome (pain, pulse, pallor, paresthesia, paralysis), the poor outcome could have potentially been prevented. Although there is clinical variation in the diagnosis and management of compartment syndrome, early diagnosis is crucial to avoid morbidity. When clinical examination is concerning (or discrepant in this case), prompt orthopedic consultation for measurement of intramuscular pressure is warranted.17 Admission to the hospital for close monitoring should be considered. Perhaps these measures would have been taken if the nurse’s findings were communicated to the physician. This is a great reminder to read all nursing notes and keep open, strong communication with the medical team.
The above cases highlight just a small sample of errors that can haunt medical providers. Hopefully, these stories shed light on how even the most well intentioned, thoughtful providers can miss something. To prevent errors, it is crucial to maintain good communication, pay attention to detail, listen closely to patients and families, form a broad differential and avoid premature closure before making a diagnosis. It is most important to be in the moment with our patients.