Volume 3, No. 3 • Fall 1988

ICPAMM Conference Reviews Safety Research

Jeffrey B. Cooper, Ph.D.

A meeting of the International Committee for Prevention of Anesthesia Mortality and Morbidity was held on May 22 in Washington, D.C., just prior to the World Congress. There were 61 invited participants, representing 19 countries. ICPAMM is an informal communication network among individuals involved in studying anesthesia mortality and morbidity and related issues or who are involved in creating or implementing policies aimed at anesthesia patient safety.

The primary objective of this meeting was to review recent studies with emphasis on experimental methods. In particular, presenters were asked to share experiences about the relatively successful and not so successful aspects of their investigations.

Six studies we-re described during the morning session: Confidential Inquiry on Perioperative Deaths (United Kingdom) (1), Randomized Trial of Four Anesthetic Regimens (U.S./Canada), Analysis of Malpractice Closed Claims (U.S.)(2), Prospective Critical Incident Survey (Australia), Manitoba Study (Canada)(3) and Recovery-Room-Impact Events Study (U.S.)(4). It became clear in the course of discussions that the study of anesthetic outcomes has become of intense interest in the anesthesia communities of many countries. Further, patient safety is similarly high on the agenda of research and clinical interests.

Morbidity, Mortality

The presentations of M&M investigations illustrated the spectrum of anesthesia outcomes and the ways that they can be studied. Dr. John Lunn described how the CEPOD reviewed reports of deaths within 30 days of a half million operations in three regions of the U.K. Reviewers analyzed these and attributed cause to surgical disease, intercurrent disease, surgery or anesthesia, or a combination of those. The study of malpractice closed claims, reported by Dr. Fredrick Cheney, examined records of 17 insurance companies. Here, only anesthesia-related claims were reviewed and the reviewers assessed appropriateness of care, involvement of error, and preventability. Because there is no denominator of the total number of operations represented, no incidence can be computed. The outcomes include not only death but other injuries (excluding dental).

Dr. Forrest described how the randomized trial among ten hospitals recorded outcomes in 17,201 patients who were randomized across four anesthetic drugs (halothane, enflurane, isoflurane and balanced). Nearly 86% of patients had some form of “outcome” as defined by this study. The vast majority of these, of course, resulted in no untoward longterm sequalae. The study represents the kind of methodology that is required to test hypotheses, in this case, “Do differences exist among the different anesthetic drugs?”

Large Studies

Dr. Marsha Cohen described the Manitoba Study in which adverse outcomes were documented for approximately 1 12,000 patients in one large teaching hospital. Here, all patients were carefully screened by a smaller group of anesthetists, assuring better validity of the outcome data. The overall rate of outcomes was substantially different than that in the randomized trial, owing to the differences in definitions of “outcomes:’ This demonstrates how careful one must be in interpreting and comparing studies.

The study on “recovery-room-impact-events,” described by Dr. Jeffrey Cooper, illustrates yet another outcome measure, similar to that defined in the Manitoba Study. Potentially harmful, anesthesia related events, requiring intervention and continuing care in the recovery room were documented. The overall rate of the-se was not too dissimilar from that in the Manitoba Study. This study illustrated how such outcome measures can be used to test the effects of clinical interventions. But, as Dr. Cohen described in her presentation, none is as powerful or elegant as the randomized trial technique.

Each presentation offered insight into the realities of anesthetic risk, interpretation of research or news about new activities. CEPOD demonstrated that cooperation between anesthesiologists and surgeons in outcome studies is possible. The study provides some comfort in revealing that very few people die solely as a result of causes attributable to anesthesia (11185,000). But concern still exists over the 14% of patients to whose deaths anesthesia at least partly contributed. A positive effect of the study was that surgeons and anesthetists are now meeting to jointly discuss their practices and that ongoing reporting of deaths has now been established. The identification of inadequate supervision of junior anesthetists as a frequent mortality related finding is prompting closer examination of this issue. Dr. Lunn explained that the anonymity and confidentiality and legal protection provided were at least partially responsible for the success of CEPOD. He also noted that intense personal contact with participants and their hospitals is required to obtain meaningful cooperation.

Dr. Forrest indicated how difficult it is to organize a massive randomized trial. A particular problem is obtaining consensus among a large pool of investigators (22 in this case). Intense attention to detail is required to assure that the participants and their institutions understand the study and adhere to the protocol. Verification of data similarly requires a tremendous commitment of resources. In summarizing the study, Dr. Forrest reported his assessment that, while some differences among the four anesthetic regimens were identified, these were relatively minor and probably don’t contribute in a major way to anesthetic risk. He believes that the condition of the patient and the person administering the anesthetic are probably more important than which anesthetic drugs are used.

Dr. Robert Webb described the foundation of a new Australian Anesthesia Patient Safety Foundation. It is non-profit and is privately financed. An anesthesia incident monitoring study has also been established in Australia and is planned to include 50 hospitals and 500 anesthesiologists. Its purpose is to collect critical incidents and analyze that information and share it among the Australian anesthesia community. A vital fink in the study will be the person-on-the-spot (POS) located in each participating institution and responsible for motivating cooperation and ensuring credibility of data.

ASA Closed Claim Study

In describing the closed-claims study, Dr. Cheney reported that analysis of inter-rater reliability is an important new component of their work. Such assessments are vital in the design of outcome investigations where reviewers are used. It was comforting to find that among 42 raters of 48 case summaries there was a high degree of agreement on the appropriateness of care and on the involvement of human error. Agreement on assessing prevention by better monitoring was somewhat less but still quite reasonable(Specific data will be reported in a manuscript in preparation.) The value of the closed-claims review methodology is demonstrated in the way that it identified an interesting subset of deaths and injuries following sudden cardiac arrest from spinal anesthesia and in providing information to assess the potential of monitoring technologies to prevent untoward outcomes(5).

The Manitoba Study offers many lessons. Again, careful study design and attention to detail are required for credibility. All of the centers involved must use similar instruments. Code books must be prepared to define all data items; all participants and data coders must be trained and indoctrinated into the objectives of the study. An appropriately designed anesthesia record can be an instrument for collecting outcome data. But, the design of that re-cord must recognize potential discomfort of documenting complications. Combining the same document for anesthesia billing with collection of outcome data is a good vehicle for improving compliance. The importance of a dedicated, well trained program coordinator, with appropriate clinical background and the respect of other clinicians is vital to the success of such efforts.

In describing the study of RRIES, Dr. Cooper noted how carefully planned experimental designs can be altered by uncontrollable events. To test for the effects of feedback of information to participants, a six month control period and a six month experimental period were planned. This was based on estimates of the required sample size for sufficient statistical power to test the hypothesis. Plans to introduce pulse oximetry sooner than had otherwise been anticipated forced a change in the sampling periods. The smaller sample size impeded testing of the hypothesis although reasonable statistical power was still obtained. Still, this illustrates the extreme difficulty in controlling all of the potentially confounding variables in longitudinal clinical experiments.

During discussions at the ICPAMM Meeting, other participants briefly mentioned ongoing studies in Norway (Dr. Gisvold), in South Africa (Dr. Gaisford Harrison) and in Denmark (Drs. Pederson and Trier Moeller). Dr. Arman Bunatien briefly reported on data collected recently in Moscow. Dr. Harrison summarized results of studies that reflect the decreasing rate of deaths attributable to anesthesia over the past several decades (0.41 1 000 to 0.0711000).

Dr. J.S. Gravenstein described the APSF sponsored survey of fatigue among anesthesiologists. (By this time, analysis of the survey should be underway). His report prompted interest from two other countries in which it is hoped that the same survey will be distributed. The discussion following his presentation indicates that fatigue and work hours are of interest in other countries. In Brazil, the maximum number of working hours for house staff is limited to 60 hours by law. Dr. John Zelcer reported on a study that is underway in Australia to assess the effects of fatigue on performance. And, it is noted that regulations limiting length of continuous work hours also exist in various states of Australia and Canada.

Dr. Helmut Schwilden presented results of a survey recently conducted in several European countries with the objective of assessing what anesthetists monitor during anesthesia. His presentation documented that there are indeed wide variations in anesthesia practices among European states. A large majority of respondents said that they had experienced a situation where a monitoring device detected a life-threatening complication before they would have themselves based on clinical signs alone. There is apparently a need for education among anesthesiologists about the function of various forms of monitoring. Since the majority of the sampled anesthesiologists were residents, the implication is for intensified educational efforts in training programs.

Trainee Selection

Dr. Davies reviewed the literature on the personality of anesthetists and the question of whether personality was related to accidents and could be used in the process of selecting trainees. Although studies have evaluated anesthetists’ personalities, it is not clear if the subspecialty constitutes a distinct “species.” And, no profile of which anesthetist might be prone to accidents is known. A four-year, prospective study of anesthetic residents in Canada, Southern Ireland, and France is planned to better establish the personality characteristics and attitudes of anesthesia practitioners.

Dr. Zairo Viero spoke about patient safety in Latin America, where facilities for anesthesia are very different than the industrialized countries. Electronic monitoring equipment is rare and even where it exists, its useful fife is short due to the unavailability of spare parts. Human resources must be the basis for improving safety. More training programs are generally needed. These must recruit anesthesiologists of the target country. Foreign training often results merely in the loss of talent.

The ICPAMM discussions prompted many useful connections for future studies. The next meeting is scheduled for Warsaw in 1990.


1. Buck, N., Devlin, H.&, Lunn, J.N.: The Report of a Confidential Enquiry into PERIOPERATIVE Deaths. The nUFFIELD Provincial Hospitals Trust, 3 Prince Albert Road, London, 1987.

2. Cheney FW: Anesthesia: Potential risks and causes of incidents in Safety and Cost Containment in Anesthesia. Gravenstein IS and Holzer IF, eds. Butterworths, 1988.

3. Cohen, M.M., Duncan, P.G., Pope, W.D. and Wolkenstein, C.: A Survey of II 2,000 Anesthetics at One Teaching Hospital (1975-83). Can Anaesth Soc 33:22-31, 1986.

4. Cooper, I.&, Cullen, D.J., Nemeskal, R., Hoaglin, D.C., Gevirtz, C.C., Csete, M., Venable, C.: Effects of information feedback and pulse oximetry on the incidence of anesthesia complications, Anesthesiology 67:686-694, 1987.

5. Caplan RA, Ward RJ, Posner K and Cheney FW: Unexpected cardiac arrest during spinal anesthesia: A closed claims analysis of predisposing factors. Anesthesiology 68:5-11, 1988.

Dr. Cooper, Massachusetts General Hospital, is the organizer of ICPAMM and a member of the APSF Executive Committee.