Diversity, scholarship and clinical relevance characterized the scientific presentations in the Patient Safety, Epidemiology and Education sessions at the 1989 annual meeting of the American Society of Anesthesiologists in New Orleans. There were over 100 presentations in six sessions, a remarkable increase in interest in this relatively new area, during the meeting. Salient aspects of selected presentations are provided here.
Drs. P.M. Kempen (University of Pittsburgh) and D.W. Learned (University of Michigan) collaborated to study whether anesthesia practice patterns might be vectors of infection. Underlying their interest is recent documentation that 3.3% of surgical patients suffer unexplained postoperative hepatitis not related to transfusion. The majority of 139 questionnaire respondents routinely or frequently used one syringe to administer drugs in the intravenous infusion of more than one patient; 98% reuse multiple-dose vials opened by unknown per sons, and 75% did not discard such vials after refilling syringes used in the care of multiple patients. Similarly, a majority of respondents do not clean laryngoscope blades in accordance with guidelines issued by the federal Centers for Disease Control. The investigators speculate that such practices may contribute to the spread of nosocomial viral infection such as hepatitis.
Dr. A.D. Rosenberg and colleagues (Hospital for joint Diseases, New York) surveyed attendees at a major anesthesia meeting for similar infection related perceptions and practices: 60% of respondents reused syringes in the care of several patients, even though 80% believe that a needle inserted into an intravenous infusion port can become contaminated. Although aware of the AIDS and hepatitis hazards, a majority of anesthesiologists had not obtained hepatitis B vaccination and failed to wear gloves and eye protection. The authors concluded that, even though anesthesiologists are aware of infection hazards, many do not implement protective measures for themselves or their patients.
Dr. S. Roth (University of Chicago) reported on adverse cardiopulmonary sequelae to the administration of OKT3, a monoclonal antibody used to prevent or manage renal transplant rejection. Although seizures, fever, tachycardia and hypotension have been reported in awake patients, there have been no reports of side effects in anesthetized patients. In 25 patients, they noted adverse responses in four: two developed severe bradycardia or hypotension; two other patients developed desaturation up to an hour after administration, requiring low levels of PEEP. High dosage of OKT3, concomitant administration of drugs reducing myocardial contractility, and hypovolemia may be risk factors for development of severe responses to OKT3.
Dr. U. Adourian (The Johns Hopkins University) reported on a study of the immunoreactivity of four commercial preparations of prolamine, prompted by an interest in determining whether they differed in their propensity to cause actual allergic reactions. Using the sera of patients who were known to have experienced life-threatening reactions to prolamine, no differences were found. Thus, there is no apparent advantage to using any particular preparation when reversing heparin anticoagulation.
Dr. S. Aubas (Hospital Lapeyronie, France) reviewed 186 cardiac arrests occurring among 102,468 anesthetics in a university hospital. Of these, 1 5 7 cardiac arrests were deemed unrelated to anesthetic management, with 144 deaths due largely to hemorrhage, ventricular failure after cardiopulmonary bypass and multiple organ failure. Of the 29 cardiac arrests related to anesthetic management, 14 occurred during induction (2 deaths), 14 during maintenance (8 deaths), and one during immediate postoperative recovery (death). Causes of these cardiac arrests included absolute or relative anesthetic overdosage, some with associated hypovolemia, hypoxemia and multiple etiologies. The 11 deaths related to anesthetic management involved multiple etiologies (5 cases), myocardial ischemia (3), dysrhythmia (1), hypoxemia (1), and relative overdosage (1). Cardiac arrest during regional anesthesia had a higher incidence but better outcome than during general anesthesia
ASA Closed Claims
Dr. H.S. Chadwick (University of Washington) presented the obstetric subset from the ASA Closed Claim study, a review of 1455 (thus far) claims of medical malpractice relating to anesthesia care157 (11%) of the claims related to Cesarean section or vaginal delivery, with two-thirds of the cases involving the former. Despite the use of regional anesthesia in 66% (77% of vaginal delivery, 61 % of Cesarean section), respiratory problems predominated: pulmonary aspiration in 6% (10% of vaginal delivery, 5% of Cesarean section) compared to only 2% in the non-obstetric claims. Injuries covered a wide range, including headache, pain during anesthesia, seizure, newborn brain damage and maternal death. Payments were made in 5 1 % of claims, with a median payment of $225,000, compared to 58% of non-obstetric claims with a median payment of $80,000.
Dr. D.A. Kroll and others from the University of Washington examined the incidence of nerve injury among the cases in the ASA Closed Claims Study. Ulnar nerve injuries during general anesthesia were, by far, the most common problem seen. Two thirds of the-se were associated with malpositioning of the arm. Nearly even at second and third on the list were femoral/sciatic injuries associated with regional anesthesia and brachial plexus injuries (appearing in about a 2:1 ratio of general to regional anesthetics). The median and radial nerves were the other identified sites of injury. The authors commented that often the standard of care had been met and the injury was a poor outcome not involving negligence. Also, there was a tendency for late reporting so that many injuries were not identified until more than one week after the anesthetic.
Dr. M. Fischler and colleagues from France discussed the safety of using intramuscular midazolam in a relatively high dose (0. I mg/kg) as a premedication prior to retrobulbar block for ophthalmic surgery. There was minor but statistically significant hemodynamic depression and frequent but brief oxygen desaturation (never below 85% but 8 of 69 between 85 and 90%). Routine pulse oximetry in these patients was recommended.
Gastric acid aspiration prophylaxis received attention. Dr. B.R. Jacobs et.al. from Yale suggested administrating controlled-release metoclopromide was a valuable therapy. The H-2 blocker famotidine was shown effective in reducing acid by Dr. B. P. Capehart from Texas Southwestern Medical Center and Dr. E.A.M. Faure from the University of Chicago suggested that, in an emergency, the sodium bicarbonate solution intended for IV injection can be given PO as an effective short-term acid neutralizing measure.
Confirmation of the value of inspection of the hypopharynx as a means to predict difficult intubations was presented by Dr. S.M. Cohen (Michael Reese Medical Center). Preoperatively, patients were given one of three classifications according to the ability to see tonsillar pillars and uvula. These classes correlated directly with the degree of difficulty of laryngoscopy and the frequency of difficult intubation. It was suggested that this simple test is the stronger predictor of potential problems.
Dr. T. Pederman of the University of Copenhagen reported on the risk of death associated with anesthesia and surgery in a study of 7306 consecutive patients undergoing non-cardiac surgery. The overall in-hospital mortality was 1.2%, but was as high as 76% for emergency gastrointestinal surgery in elderly patients with chronic heart failure and renal failure. Only one in twenty deaths occurred during anesthesia and the remainder were due to progression of the underlying illness.
Dr. M. Pine (University of Chicago) presented a risk model that effectively separates patients into risk categories. The risk model uses a limited number of risk factors including age, emergency status, ASA classification, and NYHA class. The risk category can be used to estimate prognosis and monitor the outcomes of care within an institution.
Dr. L.M. Borland described the Quality Assessment system implemented at the Children’s Hospital of Pittsburgh. Preoperative, intraoperative and follow-up information are entered into a microcomputer for analysis with commercially available database management software. This quality assessment system will delineate types and frequency of anesthesia-related problems in children. Post-op vomiting, hypoxia, laryngospasm and post-intubation croup were the most common events identified. It is suggested this system could be adopted by other institutions for the evaluation of pediatric anesthesia care.
Dr. J.N. Cashman from Guy’s Hospital in London reported his study of the effect of sleep deprivation on mood and cognitive performance. Residents evaluated after having less than four hours of continuous sleep on a call night reported a subjective feeling of fatigue and clumsiness, and did have some impairment of memory, concentration, dexterity and reaction time demonstrated by objective testing, but the differences were not statistically significant.
Dr. J.L. Lichtor (University of Chicago) also presented a study of the effect of sleep deprivation, reporting impairment of auditory, visual, and divided reaction times and coordination, although only auditory reaction time was significantly impaired. Interestingly, the magnitude of change on all tests was identical to that seen in a group of volunteers with a blood alcohol level of 0.05%.
There were two abstracts presented concerning the use of anesthesia simulators. The first was by Dr. H.A. Schwid (University of Washington). This simulator models cardiac and respiratory physiology and a variety of drugs used in anesthesia on a computer. The study asked two questions: is this system a useful teaching tool and is the system user friendly? Based on interviews with subjects, the authors concluded the simulator was a valid teaching tool, predictive of responses in the operating room. Because the simulator required usually less than 20 minutes to learn, it was judged user friendly.
The second abstract was presented by Mr. A. DeAnda (a medical student working with Dr. D.M. Gaba of Stanford University). It examined the role of human error as a source of anesthetic mishaps, using an anesthesia simulator. 88% of the unplanned mishaps (mishaps occurring during the simulation, but not part of the original simulation plan) resulted from human error and that most were either due to incorrect monitor usage or inappropriate allocation of the anesthesiologist’s attention. (During which the anesthesiologist’s attention was diverted from the main problem by a secondary distraction, usually one of the monitors.)
Dr. F. Rhoton (Case Western Reserve) investigated the distribution of critical incidents among anesthesia residents during their first year of anesthesia training at five anesthesia training programs. These investigators demonstrated that the frequency of critical incidents peaked during the months of October and November. They also demonstrated that a small fraction of the residents, 15.5%, were responsible for 58.6% of all critical incidents that occurred. This is an ongoing project with the ultimate goal being to identify patterns of performance that lead to problems and effective means of modifying these patterns.
Dr. I. Moyers (University of Iowa) studied the development of clinical signs of hypoxia during normocapnia in a swine model. During slow incremental decreases in the FiO2 heart rate and blood pressure did not change until PAO2<20 mmHg. The findings reinforce the need to monitor FiO2 and SaO2 during anesthesia. Dr. J.T. Moller (University of Copenhagen) dealt with the incidence of hypoxia in the recovery room. 55% of patients developed at least one hypoxemic episode, defined as a SaO2 of less than 90% for greater than 20 seconds, while in the recovery room and that 22% of patients were hypoxemic at time of discharge from the recovery room. The authors are currently investigating the association between perioperative hypoxemia and morbidity in a multi-institutional study. Dr. I.B. Gross (University of Pennsylvania) concerned the efficacy of supplemental oxygen in preventing hypoxemia during sedation with midazolam and meperidine. Application of oxygen by nasal cannula significantly decreased the incidence and duration of hypoxia in the patients studied.
Dr. E.D. Kharasch (University of Washington) presented data indicating that there is a stress-related significant decrease in serum potassium concentrations in the immediate pre-induction period. These investigators demonstrated an average decrease in the immediate preinduction potassium concentration of 0.8 mEq./L. compared to potassium concentrations measured one-to-two days preoperatively. Almost 50% of the patients previously normokalemic would have been considered hypokalemic by traditional criteria at the time of induction. Accordingly, the authors suggest observation of such values does not warrant either potassium administration or postponement of surgery.
Dr. R. Christopherson (Johns Hopkins) and Dr. G. McCloskey (Yale) both presented papers relating to cardiovascular management and morbidity/mortality in the perioperative period. Dr. Christopherson looked at the incidence of intra and postoperative tachycardia in patients undergoing lower limb major vascular surgery who were randomized to either epidural or general anesthesia. This study was motivate-d by the premise that tachycardia is a major determinant of myocardial oxygen supply and may significantly and adversely alter patient outcome. A rigorous protocol was utilized to quantitate duration of tachycardia in each group. Moreover, a uniform approach to treating tachycardia was utilized. The authors demonstrated that tachycardia occurred more frequently postoperatively than intraoperatively in both the epidural and general anesthesia patient groups, and that the frequency of tachycardia was similar in both groups.
In a retrospective study, Dr. McCloskey looked at the incidence of unscheduled postoperative ICU admissions for evaluation of potential perioperative myocardial infarction. During this study, 53 patients out of 15,030 were evaluated to rule out myocardial infarction. Four patients sustained myocardial infarction and two of these died within 48 hours. Only the presence of postoperative angina was a significant predictor of myocardial infarction. Seventy percent of the cardiovascular events that resulted in admission to the ICU occurred in the PACU. This is consistent with the previously cited observation. Further, the authors concluded that transfer to the ICU too rule out an MI due to mild hemodynamic instability is not cost effective. They suggest that patients developing an MI can be readily distinguished based on the magnitude of the hemodynamic compromise and the presence of greater than 1 mm ST segment depression.
Influence of Technique
Two separate papers examined the influence of regional vs. general anesthesia in two different settings. In a paper by Dr. J.H. McIsaac from Hartford Hospital the influence of the type of anesthesia [epidural vs. general vs. combined epidural-general] on intraoperative blood loss, intra and postoperative blood transfusions, and length of hospital stay were examined in patients who were undergoing radical prostatectomy. intraoperative blood loss and the associated necessity for both intra and post-operative blood transfusions was significantly increased in those patients who had general anesthesia as opposed to those who had either epidural alone or combined epidural with fight general anesthesia. The duration of hospital stay was also significantly longer in patients who received general anesthesia.
In contrast, in a study by Dr. B. Kim from Yale, the type of anesthesia employed [regional vs. general] did not influence outcome in patients who were undergoing hip fracture pinning. This also was a retrospective study but the end points monitored were outcome events rather than physiological parameters. Outcome parameters similar in both groups included significant cardiovascular morbidity (cardiac failure, myocardial infarction and angina) respiratory complications and the development of deep venous thrombosis.
Dr. G. Bashein from the University of Washington examined the influence of aspirin ingestion within seven days of surgery on the incidence of reoperation for bleeding after coronary artery bypass grafting. In this retrospective case control study, the investigators demonstrated that aspirin ingestion was associated with significantly increased risk of reoperation with an estimated odds ratio of 1.82. The aspirin patients had the anticipated significantly longer ICU stays and longer post-operative hospital course than did the controls. It was suggested that in an individual patient the presence of apparently normal platelet count and bleeding time does not necessarily prevent the increased risk associated with recent aspirin ingestion.
Drs. Dull (University of Iowa), Eichhorn (Harvard), Nyhan (Johns Hopkins), Orkin (University of California, San Francisco), and Schwid (University of Washington) moderated scientific sessions on patient safety and epidemiology at the ASA.