TITLE: EMBOLIC PHENOMENA DURING SHOULDER SURGERY IN THE SITTING POSITION

AUTHORS: Paul B. Audu, MD, Joseph L. Seltzer, MD
AFFILIATION: Thomas Jefferson University Hospital, Philadelphia, PA.

INTRODUCTION: Negative intrathoracic pressures as occurs during spontaneous ventilation and elevation of the surgical field above the level of the heart both facilitate venous air embolism (VAE)¹ . A high incidence of this phenomenon would therefore be expected in spontaneously ventilating patients undergoing open shoulder surgery in the sitting position. We report preliminary results of a study to determine the incidence of Embolic phenomena (EP) in this setting.
METHOD: Institutional approval and patients' informed consent were obtained. After applying standard monitors, patients received either an interscalene brachial plexus block or a balanced general anesthetic with mechanical ventilation. They were positioned sitting (beachchair). A prechordial Doppler was placed over the left hemithorax and adjusted until a change in Doppler heart tones (DHT) was obtained following intravenous injection of 10 ml of agitated saline. Auscultation was continued for the duration of the surgery. Variables recorded included the right atrium (RA)-shoulder height; type of surgery; mode of ventilation; and DHT changes. results were analyzed using chi- squared.
RESULTS:

Two of 17 patients (11.7%) had intraoperative EP. Neither the RA-shoulder height nor the mode of ventilation was significantly different between those with and without EP. Two of six (33%) hemiarthroplasty (HA) patients versus none of 11 (0%) acromioplasty patients (P<0.05) had EP.
DISCUSSION: EP were neither related to mode of ventilation nor to RA-shoulder height; EP were related to type of surgery, occurring only during HA as the prosthesis was being inserted into the humeral shaft. Previous authors have noted EP (air, fat and/or marrow) occurring during insertion of the prosthetic stem in hip arthroplasty ^{2, 3}. We postulate a similar mechanism to account for EP during shoulder HA. Our incidence of 33% compares with 30% found during hip arthroplasty ³. Events were followed by minor hemodynamic changes not warranting special therapeutic interventions.
REFERENCES:

1) Neurosurgery 3(3):380-384, 1978;

2) Anesthesiology 40(4):405-7, 1974;

3) Anaesthesiology 35:858-862, 1980.