The problem in this case was not the selection of hydromorphone as the initial analgesic agent. Hydromorphone is frequently used safely in the ED to treat severe pain associated with conditions such as sickle cell vaso-occlusive pain crisis, renal colic, and long-bone fracture. Issues arise when hydromorphone is combined with a benzodiazepine (in this case, diazepam), which by itself causes sedation and anxiolysis. Central nervous system (CNS) depression may be additive and occur when benzodiazepines are used concomitantly with drugs that also cause CNS depression (eg, opioids).1 This combination can lead to excessive sedation, resulting in partial airway obstruction and hypoxia.1 For example, in an investigation by Bailey et al,2 in human volunteers, neither hypoxemia nor apnea was evident after administration of .05 mg/kg of IV midazolam. In patients who received 2 mcg/kg of IV fentanyl alone, hypoxemia occurred in 50%, but apnea did not occur in any of the patients studied. However, when the same doses of these drugs were administered together, 92% of participants exhibited hypoxemia and 50% became apneic.2
When a combination of an opioid and benzodiazepine are given over frequent intervals, the clinician crosses over from treating pain to performing procedural sedation and analgesia—whether he intended to or not. As such, the patient in this case required proper monitoring, including cardiac monitoring and pulse oximetry; he also should have been placed on supplemental O2. Ideally, the patient would have benefited from end-tidal carbon dioxide (ETCO2), monitoring, if available. This is a noninvasive measurement of the partial pressure of CO2 in exhaled breath. Hypoventilation from respiratory depression results in an increase in ETCO2, and hyperventilation results in a decreased ETCO2. While pulse oximetry is excellent at monitoring O2 saturation, it is ineffective in the early detection of respiratory depression, hypoventilation, and apnea. The hypercarbia precedes the hypoxemia—by as much as 60 seconds (range 5-240 seconds), according to a study by Deitch et al.3
Finally, rather than relying solely on the reports from the nurse, the EP should have personally reassessed the patient at some point. Nursing updates are extremely helpful, but when ordering repeated doses of IV opioids and benzodiazepines, the EP should personally reassess the patient.
Hot Red Knee
64-year-old man presented to the ED with a chief complaint of right knee pain, which he stated began approximately 2 days earlier. He denied any injury or trauma or a recent history of fever, chills, or other joint complaints. He described the pain as constant, worse with weight bearing, and becoming progressively more painful. The patient had a history of gout; however, previous attacks had only affected his great toes and elbows. His medical history was also significant for hypertension, for which he was taking lisinopril and hydrochlorothiazide. He admitted to moderate alcohol consumption but denied tobacco use.
On physical examination, the patient appeared uncomfortable due to the knee pain. All of his vital signs were normal. A focused examination of the affected knee revealed a small effusion, diffuse tenderness to palpation, mild erythema, and slight increased warmth. The patient exhibited pain with flexion and extension of the right knee. The right ankle examination and right dorsalis pedis pulse and posterior tibial pulse were all normal. No laboratory or imaging studies were obtained.
Based on the patient’s history and physical examination, the EP believed the patient’s symptoms were due to an episode of gout. He prescribed oral colchicine, allopurinol, and acetaminophen/hydrocodone; he also advised the patient to apply warm compresses to the affected area and limit his activity. He discharged the patient home with instructions to follow up with his primary care physician.
Two days after discharge, the patient returned to the same ED via EMS. On this presentation, he was febrile, with a temperature of 102.6oF; a HR of 120 beats/minute; and a BP of 92/50 mm Hg. He also had altered mental status. The patient’s right knee appeared more swollen, and he would not flex it due to the severe pain. The EP was concerned for sepsis, and ordered blood cultures, a complete blood count, basic metabolic profile, and lactic acid evaluation. The patient was administered 2 L normal saline IV and broad-spectrum antibiotics. Despite the addition of vasopressors, he continued to deteriorate; he ultimately went into cardiac arrest and died.
The patient’s family sued the EP from the initial ED visit for failure to diagnose the right knee pain and swelling as septic arthritis (SA). The plaintiff’s attorney argued that this failure to diagnosis directly caused the patient’s sepsis and death. The EP argued that the patient’s history and physical examination were consistent with an acute gout attack, that there was no evidence of infection in the right knee, and that this was not the cause of the patient’s death. At trial, the jury returned a verdict in favor of the defense.