Sepsis is a leading cause of death in hospital. It can be difficult to identify, especially in pediatric, older, and chronically ill patients because its early symptoms and signs can be subtle and may mimic other conditions as well as normal post-operative recovery.1,2 Timely recognition and treatment of sepsis is important to prevent serious harm to patients and reduce medical-legal risk for physicians.
Inadequate patient assessment was a factor in 75% of the reviewed CMPA cases. Recurring themes in these cases were not recognizing the signs of sepsis, not appreciating the severity or worsening of a patient’s condition, and not including sepsis in the differential diagnosis. The cases described below illustrate these diagnostic issues.
Case 1: Multiple visits to the ED and diagnostic anchoring delay recognition of sepsis in boy with chickenpox
A 9-year-old boy is brought to the emergency department (ED) with fever, nausea, and vomiting, and pruritic vesicles. He had been seen 3 days earlier in the same ED for an upper respiratory infection, and a throat swab from that visit was negative. The patient is diagnosed with varicella and discharged with parental instructions to administer antipyretics as required and to return to the ED if the boy’s condition worsens.
That evening, the patient is brought back to the ED for a third time with continued fever, vomiting, shoulder pain, and a fine red rash covering parts of his body. Without reviewing the patient’s medical record, another ED physician examines the boy and notes a high fever, tachycardia, and slight tachypnea. He orders throat swabs, and discharges the patient with parental instructions to continue his medications.
The next morning, the boy, now listless, is brought back to the ED a fourth time with abdominal and back pain and a new hemorrhagic vesicle on his foot. He has grunting rapid respirations, tachycardia, and hypotension, but he has no fever and O2 saturations are normal. A new ED physician examines the patient, notes clear chest sounds, and interprets the chest X-ray as normal. The physician lances the vesicle, diagnoses local cellulitis and discharges the boy with a prescription for an oral antibiotic and analgesic for his chickenpox.
Five hours later, the patient is admitted to the ICU in profound septic shock. Despite extensive resuscitation efforts, he dies a few hours later. Cause of death is determined to be septicemia due to Group A Streptococcus complicating a varicella virus infection.
A legal action follows. Peer experts opine that the two physicians who saw the patient on the third and fourth visits had sufficient clinical information available to appreciate that the patient was septic and potentially had severe sepsis or septic shock. The experts stated that:
- when the child presented with a red rash, the physician should have suspected that the course was not typical for varicella and identified toxic shock syndrome as a possibility, and ordered appropriate diagnostic investigations (e.g. blood cultures, complete blood count, coagulation profiles, and a serum lactate), reviewed the medical records, and admitted him.
- when the child presented with grunting respirations and abdominal pain, the physician should have appreciated the clinical deterioration; thoroughly examined the boy; ordered appropriate investigations including blood cultures, as well as intravenous (IV) fluid resuscitation and antibiotics; consulted a pediatrician; and admitted the boy to the ICU.
Case 2: Surgeon misses evolving sepsis in elderly post-operative patient
Eight hours after undergoing a laparoscopic cholecystectomy, an 84-year-old man develops a low-grade fever and tachycardia. Overnight, he complains of increased abdominal pain and nausea with vomiting. His oxygen saturation and blood pressure are low. The on-call physician orders IV fluids. The next morning, the surgeon notes that the patient is afebrile, but still tachycardic. She finds the patient’s abdomen slightly distended with no guarding or rigidity and diminished bowel sounds. Her differential diagnosis includes post-operative ileus and dehydration.
That afternoon, the patient becomes short of breath, and his urinary output is low. The surgeon orders bloodwork and a chest X-ray, and IV fluids and antibiotics for possible pneumonia. The serum creatinine levels have increased. The patient’s condition continues to deteriorate. In the early evening, he is transferred to the ICU when he shows more advanced signs of sepsis: hypotension, tachycardia, metabolic acidosis, and early renal failure. On examination, he has significant abdominal tenderness, bilateral guarding, and a silent abdomen. The intensivist diagnoses peritonitis, suspects bowel perforation or bile leak, and refers the patient for exploratory surgery. A bowel injury is found and repaired. The patient does poorly post-operatively with septic shock unresponsive to treatment, renal failure, and extensive cerebral infarction, and dies a few days later.
The patient’s family launches a legal action. With respect to the warning signs of sepsis in a post-operative patient, peer experts are critical of the surgeon’s failure to do the following:
- Formulate a differential diagnosis including intro-abdominal sepsis/severe sepsis when the patient’s symptoms persisted and worsened.
- Consider the significance of the patient’s tachycardia.
- Recognize that the patient was under-resuscitated and going into renal failure.
Case 3: Physician fails to recognize worsening infection in patient with multiple co-morbidities
Five days after taking a 7-day course of nonsteroidal anti-inflammatories (NSAIDs) for recurring gout, a 65-year-old woman returns to see her family physician (FP) for continued inflammation of the joint, and a new large ulcer over the area. She is a smoker and has multiple medical conditions, including diabetes, coronary artery disease with an implantable cardioverter defibrillator (ICD), recent vascular bypass graft surgery, and treated rectal cancer. The FP continues the NSAID and orders twice daily dressing changes with an antibacterial topical gel.
At the next appointment the following week, the gout appears to have improved. However, there is an odour from the ulcer which suggests infection, and the patient reports intermittent fever and chills. No cultures or other investigations are obtained. The FP prescribes a 10-day course of oral antibiotics for the infected ulcer and instructs the patient to return for follow-up in five days. Three days later the patient becomes confused and agitated, and she is hospitalized. She is diagnosed with a staphylococcal sepsis, but despite aggressive treatment, which includes a partial foot amputation, the patient dies from complications of infective endocarditis and sepsis.
A regulatory authority (College) complaint follows. The College concludes the FP should have considered the patient’s co-morbidities that could have increased her risk of sepsis and severe sepsis, and treated the patient’s infection with IV antibiotics.
Peer expert opinions emphasize the importance of early recognition of sepsis. They also uphold treatment bundles that include initial resuscitation, prompt administration of IV antibiotics and early source control, which are consistent with clinical practice guidelines.4,5