Intramuscular air in the setting of infection can be an indicator of a rapidly escalating and potentially life-threatening infection, such as gas gangrene or necrotizing fasciitis. For a patient with diffuse pain in an extremity and air in the tissue planes as seen on imaging studies, surgical exploration has been strongly recommended1-6. Iatrogenic causes of air in fascial planes are underreported.
We present the case of a child with septic arthritis of the hip that was treated with open irrigation and débridement by the on-call service. There was temporary use of pulse lavage during the operation, resulting in intramuscular and subfascial air. The following day, the medical team that accepted the patient from the on-call service had no knowledge of the use of pulse lavage and, when confronted with a confusing clinical picture and dramatic radiographs, proceeded with urgent surgical exploration. This case demonstrates an iatrogenic cause of intramuscular air and illustrates the importance of good communication between on-call and accepting services. In accordance with local Human Resources and Research Center (institutional review board) policies, the parents of the patient were informed that data concerning the case would be submitted for publication, and they consented.
A ten-year-old boy presented with a history of decreasing ability to walk over a period of five days prior to presentation and fever of as much as 102° F (38.9° C). He was described as a previously healthy young man except for mild asthma, which was well-controlled with albuterol and Flovent (fluticasone propionate) inhalers. There was no history of trauma or injury. The examination was notable for an inability to bear weight on the right lower extremity, and the elicitation of pain with passive range of motion of the hip. There was no pain or swelling in the thigh, knee, or leg. He had a white blood-cell count of 8.9 × 109/L, an erythrocyte sedimentation rate of 112 mm/hr, a C-reactive protein level of 17.8 mg/L, and a hip joint effusion that was noted on ultrasound investigation.
Under the care of the on-call orthopaedic service, the patient underwent open irrigation and débridement of the right hip joint on the night of presentation. A standard anterior approach to the hip joint was performed. There was cloudy fluid within the joint, which was sent for Gram stain and culture. Pulse-lavage irrigation was attempted. The end of the apparatus was placed in the wound, and the trigger was activated. The surgeon noticed that, after five seconds, it was obvious that no saline solution was being pulsed into the wound. When the pulse-lavage apparatus was checked, it was discovered that it had not been hooked up to a container of saline solution. At this point, the on-call team decided to use bulb irrigation instead of pulse lavage. After copious bulb irrigation, a Hemovac drain (Zimmer, Warsaw, Indiana) was placed in the hip joint and the wound was loosely closed with nylon suture. The patient was treated empirically with intravenous clindamycin and vancomycin.
The following day, the accepting pediatric orthopaedic service was called to the bedside by the nursing staff to evaluate the increasing limb pain. They reported that the patient's mental status was alternating between somnolent and inconsolable; he had not fully regained a verbal, communicative mental status since emerging from anesthesia the night prior. He had, however, remained afebrile since surgery and his pulse of 92 and blood pressure of 101/72 mm Hg were stable. On physical examination, the patient was lethargic but reacted to movement and palpation of the limb. There was subcutaneous crepitance throughout the thigh and calf. Neither the knee nor the ankle appeared to have an effusion. There was erythema throughout the lower extremity, but this was interpreted cautiously because the patient's skin was darkly pigmented. The leg and thigh were swollen but not tense. Passive motion of the ankle or knee and palpation of the soft tissues of the thigh and calf produced a subdued pain reaction on par with that produced during hip motion. Plain radiographs had been ordered prior to seeing the patient (Figs. 1-A and 1-B). These radiographs demonstrated intramuscular air in the thigh and calf and extending down to the ankle. The air appeared to follow fascial planes, generating a concern for the diagnosis of gas myositis or necrotizing fasciitis.
The decision was made to take the patient to the operating room for emergent fasciotomies of the thigh and leg. The possibility of an iatrogenic etiology (i.e., the air from the procedure the night before) was considered but believed to be unlikely to explain air extending below the knee. The fact that pulse lavage had been used was not known, and no contact with the operating surgeon from the previous evening was established.
The operative findings did not support the diagnosis of soft-tissue infection or compartment syndrome; there were no signs of muscle necrosis or inflammation. The fasciotomy wounds were loosely closed under no tension. No change in antibiotic coverage was made. Of note, the laboratory results from the blood tests performed before the operation were returned after the procedure, and the pertinent values included a C-reactive protein level of 21 mg/L and a creatine kinase level of 521 U/L.
Postoperatively, the patient improved rapidly with respect to the infected right hip. The fasciotomy wounds healed without complication. The drain in the right hip was removed on the second postoperative day. The patient underwent physical therapy and slowly regained mobility. The cultures from the initial hip surgery grew methicillin-sensitive Staphylococcus aureus. He was transitioned from intravenous clindamycin and vancomycin to intravenous Ancef (cefazolin) and clindamycin. The vancomycin was discontinued. Cultures taken from the thigh and leg fasciotomy sites were negative for any organisms.
On the fifth hospital day, he began having frequent episodes of epistaxis and, on the seventh hospital day, one episode of hemoptysis. The international normalized ratio was found to be 8.57. It was determined that he had a consumptive coagulopathy with depletion of multiple vitamin-K-dependent factors. The patient was given a transfusion of fresh-frozen plasma and was also started on vitamin-K supplementation.
On the seventh postoperative day, he became oliguric and was diagnosed with acute renal failure secondary to cephalosporin toxicity. The serum creatinine level reached a peak of 5.29 mg/dL (467.6 µmol/L) on the eleventh hospital day. The Ancef was discontinued, intravenous nafcillin was started, and the patient was continued on intravenous clindamycin; the renal function and creatinine level normalized thereafter.
At the twelve-week follow-up visit, the patient was doing well. He was walking with no limp and had no pain in the right hip, thigh, or leg. His wounds were all healed. Radiographs showed no evidence of osteonecrosis of the femoral head or other osseous abnormalities, and the patient had no residual bleeding problems or renal dysfunction.
The failure to recognize the iatrogenic etiology of the subfascial air and the incorrect diagnosis of gas myositis in this patient resulted in an unnecessary surgical procedure. We were unaware that the pulse lavage had been used by the on-call team during the first operation. This case illustrates potential complications that can result from poor hand-off of patient care from one medical team to another.
It is unclear, however, if the information about the use of pulse lavage would have altered the decision-making. The risks of delayed surgical débridement of an aggressive intramuscular infection are substantial, and prior to this experience, we might not have been comfortable attributing the extensive amount of air seen on these radiographs to a pulse-lavage "misfire."
The decision to use pulse lavage in this case should also be questioned. The pressure or force exerted from pulse lavage varies among systems. Some systems rely on compressed air for operation, whereas others are battery operated. Morgan et al.7 tested four different pulse-lavage systems and found that the peak force ranged from 0.16 to 1.7 N and that the peak pressure ranged from 0.03 to 0.35 N/mm2. The pulse-lavage system used during our patient's procedure was compressed-air driven, and we do not know the setting on the compressed air tank at the time of surgery, which makes it impossible to calculate the pressure or the amount of air expelled over a period of five seconds.
A review of the literature did not yield any similar reports or studies with similar radiographic findings secondary to pulse lavage. There are, however, reports of compartment syndrome resulting from the use of pulse lavage8-10. Lauber et al.8 described a case of a patient with postoperative compartment syndrome after a two-stage local débridement of proximal tibial osteomyelitis with use of pulse lavage. Interestingly, the compartment syndrome did not develop in their patient until after the second débridement. They calculated that the pressure of the pulse lavage was 760 mm Hg during both surgical procedures, and they concluded that the compartment syndrome was directly related to the pulse lavage. This implies that substantial pressure can develop in vivo with use of lavage systems.
This case illustrates a previously unreported iatrogenic cause of intramuscular air. The use of pulse lavage within a small, deep wound is not recommended because there is a danger of rising pressure within the wound, which can cause fluid or air to be pushed into surrounding compartments, resulting in compartment syndrome or intramuscular air. Improved communication would have improved our ability to recognize the iatrogenic source of the intramuscular air in our patient. Surgeon-to-surgeon communication is vital in cases involving further surgical intervention for the treatment of an acute condition. This case has led to an increased emphasis on direct contact with the attending surgeon during the hand-off of patients after surgical intervention at our institution. 