Continuing a now 12-year tradition, the International Committee For Prevention of Anesthesia Mortality And Morbidity (ICPAMM) met in Sydney, Australia, as a satellite meeting of the 11th World Congress of Anesthesiologists in April. The eclectic agenda provided reports of ongoing anesthesia incident and outcome studies, safety activities, and some provocative discussion of research methods and challenges to conventional wisdom.
Jeffrey B. Cooper, Ph.D., organizer of ICPAMM, described to new participants how the organization came to be as a continuation of the International Symposium For Prevention of Anesthesia Mortality and Morbidity in Boston in 1984. Several of the roots of the APSF and worldwide initiatives in standards and safety can also be traced to that gathering. ICPAMM, he explained, is more like a “club” than a “committee;” it provides a place for anyone in the world of anesthesia who wishes to learn from colleagues about processes for studying issues related to anesthesia safety and to learn of successes and failures of programs and practices in other countries. The goal of ICPAMM parallels that of the APSF: “to ensure that no patient is harmed by anesthesia.” An occasional newsletter advises of publications and safety-related events of interest.
Several themes emerged from this meeting. Critical incident studies appear to be fairly common around the world and provide a way for a country to get started in collecting data on a local level. Generally, there is now more acceptance of the idea of reporting anesthesia problems. Several studies indicate a trend toward involvement of epidemiologists and social scientists in outcome and safety research studies. Developing countries, where resources are scarce, are still able to focus on basic safety and endeavor to apply the most critical elements of what has been learned in countries where resources are more readily available. The economic pressures on medicine and on anesthesia, specifically in the U.S., are being felt in other countries as well (although one wouldnÕt necessarily conclude that from the overwhelming presence of new electronic anesthesia systems, infusion devices, data management systems and other technologies in the exhibit section of the World Congress meeting). Summaries of specific presentations follow.
AIMS
The meeting began with reports on various aspects of the host countryÕs pacesetting national study, the Austral-Asian Incident Monitoring System (AIMS). Drs. Lynn Currie, William Runciman and John Russell described how the system is functioning, some key findings and consequent publications, and how results have led to recommendations and system changes. Since the start of AIMS in 1988, data collected by the Australian Patient Safety Foundation (“APSF,” but not to be confused with the parent organization of this Newsletter) have so far been only from Australia, yet there has been some expansion to other countries and now the first applications of corrective strategies. The effort is now funded by the Australian Department of Health, which may represent a cultural change. Protection from discovery by federal legislation and confidentiality are responsible for practitioners’ cooperation with reporting. More recently, pilot studies have been done in six other medical specialties and a dozen other countries have undertaken incident monitoring in some form based on AIMS.
Among the findings from AIMS are that problems in use of drugs appear in 8% of incidents, that endobronchial intubation is still the most frequently occurring problem leading to hypoxia, and that the most common cause of intraoperative hypertension is some type of drug administration problem. Stress, most often from self-generated haste, was seen as a contributing factor in 14% of incidents. A full third of critical incidents reported in anesthesia are related to equipment human-interface problems (only 8% were pure equipment failures, which has led recently to an industry/AIMS liaison allowing limited access to data by manufacturers).
About 20% to 30% of crisis situations would have been diagnosed sooner or managed better had a specific protocol been followed. This has led to the ABCD COVER algorithm, a plan for dealing step-by-step with untoward developments during an anesthetic. A new sub-algorithm, SCARE, has been developed: Scan every five minutes, Check on the unexpected, Alert/Ready if suspicion of a problem and turn to Emergency mode in a rapidly deteriorating situation.
The plan for the future is to include more specialties, hospitals and countries. Analysis of patient factors such as obesity are underway as is examination of preoperative assessment and preparation, the most commonly reported contributing factors in deaths in Australia. There is now also a study of clinical pathways underway.
Dr. Runciman listed several kinds of health care outcome and quality studies ongoing in Australia: incident reporting, M&M committee processes, and analysis of medical legal cases. Starting with six pilot studies funded by the federal government, there is now significant funding for incident monitoring in intensive care, general practice and obstetrics. The federal government provides money to state governments with certain requirements. There is federal and state protection from access to data. Dr. Runciman said that all the systems are in place for AIMS: collection, analysis and feedback of reports. The net for capturing anesthesia incident information is very broad and includes any event that could have caused harm to someone or any complaint. Cases are sent to the APSF via a relay station in the National Bureau of Statistics, which removes identifiers. The database is accessible only on a local area network, with no access to the outside world. The data entry system allows for “parallel coding,” i.e., very complex problems can be characterized. The classification is by “natural” categories, i.e., ones that describe the data as they are described by the reporter.
They have created various guidelines, protocols, checklists and manuals for feedback of findings. Dr. Runciman has been impressed with the power of this qualitative method (this issue was raised again later in the day). Some elements of AIMS that he believes have been important to its success are that it is confidential, provides rapid feedback, is non-threatening and is inexpensive relative to case review. He estimates that the cost is about $30,000/year to operate the system in one hospital.
Via AIMS, problems heretofore unappreciated have been recognized, for instance, that infections related to failure to replace IV cannula collectively cause more morbidity than all anesthesia events combined. From other findings it has been computed that, to be cost-effective with respect to preventing overnight admissions for ambulatory day-of-surgery patient (assuming a 1/250 rule), a completely effective treatment for prevention of postoperative nausea and vomiting would have to cost less than $2.00 per anesthetic.
Separate from AIMS, the Department of Health funded an analysis of 14,000 randomly collected case notes. Sixteen percent of admissions were found to be associated with adverse events, of which 80% were deemed preventable. This extrapolates to 250,000 adverse events a year and 14,000 preventable deaths, thought to account for 10% of acute-care hospital costs. While 50% of events were in association with surgery, less than 2% were associated with anesthesia.
During the discussion, it was asked why it isn’t possible to have access to individual cases since this precludes the possibility of a deeper qualitative analysis. Dr. Runciman believes that absolute anonymity is so important that it is an over-riding consideration mitigating against allowing for follow-up. Dr. Grobee questioned if there could not be ways to allow for follow-up with reporters and still allow for anonymity.
Dr. Frederick Orkin postulated that much is lost by not having more data about the patient, e.g., risk factors. Dr. Runciman believes that qualitative analysis of these data can still be productive in the absence of having complete demographics of patients, i.e., the elements of detail about what happened are sufficient to point toward designing solutions.
Dr. Russell reported on three examples of how AIMS data have been applied to change practice: To eliminate problems arising from soda lime dusting, they have worked with a manufacturer to create new procedures for loading absorbent, commonly supplied in bulk in Australia. The simple solution was to provide a larger container so the granules can be scooped instead of poured. They have been developing and testing this solution and are now publicizing findings.
Because AIMS identified that the greatest source of drug swaps and of drug-error related morbidity have involved relaxants and reversal agents, the U.S. color-coding standard was adopted. They have also introduced a special syringe with a colored plunger for these drugs. Introduction into practice began in 1994 and preliminary data suggest a sharp reduction in reported syringe swaps. But, he also believes that publicity from a symposium issue of the journal of the Australian Society of Anesthesiologists describing results of AIMS analyses, including drug swaps, may also have had an impact.
The analyses of endobronchial intubations, also reported in the symposium issue, do not appear to have been followed by a change in the rate of AIMS reports. In Australia, a mark is now placed on endotracheal tubes to indicate correct placement. But, power analysis suggests that about six years of new reports will be needed to study effects of this alteration in the tubes. This may be impractical unless there is an increasing rate of reporting.
Postoperative Cognitive Dysfunction
Dr. Jakob Trier Moller from Denmark provided a preliminary report on the large, prospective, multi-national study of postoperative cognitive dysfunction. The aim is to identify incidence and risk factors of postoperative cognitive dysfunction. Patients 60 years or older undergo a battery of cognitive tests before and at several times following surgery for up to three months and six months for those patients who had some residual dysfunction at three months. He illustrated several of the tests, which are quite challenging and distinguishing of various cognitive alterations. A control group of non-hospital subjects is also tested to identify the normal variation in dysfunction and in the learning effects of repeated testing. Preliminary results suggest that 15% of anesthetized patients vs. 3% of control patients can be identified as having some dysfunction. There was controversy about the value of control groups and the way they were used, but this was not resolved.
Reporting Increase and Outcome Study in Germany
Dr. Wolfgang Dick reported on a study of mortality in Mainz, Germany, which is based on the NCEPOD method used in the U.K. They have had great difficulty securing funding, but found a small sum to finance a small study. The forms used in NCEPOD were translated into German. From 1989-1992, 350 reports were filed. There was no further financial support so analysis was delayed and conducted by the University of Mainz Anesthesia Department. Dr. Dick reviewed the characteristics of 200 of the cases. About half of the procedures had actually been reviewed by the Chief of Surgery, but some had been reviewed only by non- specialists. There was no preoperative assessment in 15% of cases. In cases of aspiration, only 20% had been premedicated appropriately. Not surprisingly, there were differences in how anesthetists and surgeons rated the urgency of the surgery.
Although only 14% of cases had an autopsy and 13% were reported in a mortality conference, surgeons were present at most autopsies and conferences. Anesthetists were rarely present, which led Dr. Dick to ponder why anesthetists appear to be less interested in what had happened.
Although these results only give a very superficial view of anesthesia mortality in Germany, they have pointed to the need for more information. The German Society of Anesthesia has established a committee on perioperative outcomes. A new database has been established and 75,000 cases have been collected to date, but no analysis has yet been performed. This is indicative of a growing change in attitude in Germany toward reporting outcomes. By law, it is not possible to retrieve information on specific cases from a hospital.
Incident Reporting on the Internet in Switzerland
Dr. Jan Davies from Canada, reporting for Dr. Sven Staender, described the Critical Incident Reporting System (CIRS) that now can be accessed via a World Wide Web page at the University of Basil, Switzerland (http://www.medana.unibas.ch or directly: http://www.medana.unibas.ch/ ENG/CIRS/Cirs.htm). The home page gives instructions for completing the form, which is comprised of pull-down menus for entries of incident characteristics based on the AIMS forms and hypertext links to instruction and definitions. The system is believed to be very secure, although there was concern expressed from the audience that it is not completely possible to guarantee protection from discovery despite the removal at receipt by the server of information describing the source.
Some concern also was raised about the possibility of false incident entry by pranksters or troublemakers. It is suggested that protection issues be clarified. On the other hand, not providing access or feedback to reporters may limit interest. John Russell noted that electronic reporting of AIMS has produced higher quality, more detailed information, which he suspects is related to the spontaneity associated with the character of e-mail.
Perioperative Mortality and Morbidity in the Netherlands
Dr. Sesmu Arbous reported on progress of the case controlled study of severe mortality and morbidity in anesthesia in the Netherlands. Data collection started in 1995. The main objective is to study severe morbidity and mortality associated with anesthesia, assess frequency of events and formulate preventive measures. A 24-hour postoperative period is included. Three of the 12 Dutch provinces are involved, which included 55% of the total population who have an estimated 400,000 procedures annually. The two-year study will prospectively gather patient, provider and related data on all of the included outcomes and on controls matched only for age and sex. Dr. Arbous explained that matching has been done for practical reasons; matching is not required for validity and can actually introduce bias. From this point of view, the only requirement is that cases and controls are sampled independent of the determinants (anesthesia risk factors) under study. The analysis will be controlled for the influence of confounding factors and risk modifiers, e.g., type of surgery, which is possible given the size of the sample. There was disagreement about this approach.
The Ministry of Health has provided protection of anonymity of the data source. Besides a quantitative approach, data will be analyzed qualitatively: Two reviewers will reach consensus on appropriateness of care and contribution of anesthesia; a third reviewer will be used in cases of disagreement. Dr. Ingram noted in the discussion that the NCEPOD is now using the same approach to overcome the bias of a single reviewer in placing blame given knowledge of the outcome.
Of 64 eligible hospitals, 61 agreed to participate. Eighty-eight percent of those have submitted at least one case since the start of data collection; the total of cases in 1995 was 376. There has been some difficulty in reporting of controls; anesthesiologists needed stimulation to submit these cases. So far, it seems that anesthesiologists are contributing to the study with enthusiasm and data collection will continue.
Assessment of Standards of Anesthesia Care
Dr. Amr Montasser from Egypt described methodology that has been developed to assess standards of care in hospitals in developing countries. Three levels of hospital have been defined: low, medium and high levels of resources. A ten-part spreadsheet presents elements of care such as preoperative care, equipment and supplies, availability of medications, qualifications of personnel, availability of recovery facilities. A sample of 12 hospitals representing the different levels was used to assess how well the spreadsheet could identify the variations in levels of care and resources. His data indicated the great differences in Egypt between the levels ranging from the high level, which approach levels of the American Society of Anesthesiologists to the low level in which sterile technique is ignored and there are few medications or basic supplies.
Dr. Montasser illustrated how this simple spreadsheet tool has been useful to define the basic needs of developing countries. He commented about the potential utility of international anesthesia minimum standards such as those adopted by the WFSA at the 1992 World Congress.
Update on NCEPOD
Dr. Stuart Ingram gave an update on the ongoing studies under the National Enquiry into Perioperative Deaths in the U.K. The 1992-93 report was published last year. NCEPOD is a joint venture between anaesthetists and surgeons. During this report period, there were a total of 19,000 reports of mortality. Eighty-two percent of surgeons and 86% of anaesthetists returned at least one report. The cases represent a relatively older patient population. The most likely reason for not returning a report form for a documented death is unavailability of medical records. In classification, there has been difficulty in adherence to definitions of the study, for instance, the term emergency is not used as intended, i.e., to indicate a procedure during which resuscitation and operation are being performed simultaneously. There has recently been an emphasis to promote the use of more than one anaesthetist in the most difficult cases and this seems to be having some effect as seen by the reports. There do seem to be too many instances of junior anaesthetists working without supervision especially with severely ill patients, but there are now few locums personnel involved. There are still instances of emergency procedures being done at times when a recovery facility is not staffed. The data also suggest a continuing lack of high dependency (intensive care) units, for which efforts are presently being made to correct.
Dr. Ingram noted the gulf that exists between the developed and developing countries considering the deficiencies noted in preceding presentation and in NCEPOD. He also commented on how the press has generally concentrated on the negative aspects of the NCEPOD findings.
The next report will be out in November with a focus on one case from each surgeon to identify the characteristics of non-responders. This has produced a large number of fractured neck of femur cases.
Reporting of Adverse Events in New Jersey
Since 1989, there have legal requirements in the state of New Jersey for many aspects of anesthesia care. Dr. Ervin Moss has been the leader in this effort and has previously described the rules and the process by which they have been made and implemented (APSF Newsletter, June, 1989; see also the related article on Page 34). Dr. Cooper, reporting information provided by Dr. Moss, concentrated on issues related to legal protection to maintain confidentiality. Indeed, New Jersey is the only state in the U.S. to have such protection, which withstood a challenge by a consumer group to have data made public. He felt this to be a significant event in the U.S. and hoped it would lead other states to follow suit. There have been about 5,000 reports to date that are believed to have been useful to some of the reporting hospitals for identifying problems that were remediated. Beyond that, there has not been a comprehensive analysis of the data and recent cutbacks in state funding have halted further work for the time being.
Patient Safety Moves Forward in Japan
Dr. Yosuhiro Shimada described the continuing progress in introducing anesthesia patient safety to Japan. The JSA began an annual survey of mortality and morbidity in Japan in 1992. Hospitals are not required to report and only 30% are doing so. Still, more than 200,000 cases are now being reported annually. From these data, they estimate that the rate of anesthesia-related fatalities to be two to three per 100,000 cases (0.04/10,000 intraoperatively; 0.29/10,000 within seven days of operation). Guidelines of minimal monitoring have been established. There is a great shortage of anesthesiologists in Japan: only 3,000 are board certified, but there are an additional 10,000 “registered” anesthesiologists who have a different primary specialty and varying degrees of anesthesia training. Thus, about three times the current number of appropriately trained anesthesiologists are needed.
Dr. Shimada has recently established “The Society for Safety in Medical Practice,” which has 300 members and is not limited to anesthesia interests. It is unique in being open to anyone interested in safety and includes patients and their families as well as manufacturers, nurses, doctors, lawyers and researchers in medical and human sciences. Particular stress is being put on the opinions of patients in the development of guidelines for safety.
During the discussion of Dr. Shimada’s report, it became clear that there still exists a problem in differentiating anesthesia vs. surgical outcomes. Dr. Gisvold noted that in their outcome reporting system in Norway, they do not try to make that separation but together with the surgeons examine all adverse outcomes.
Critical Incident Methods and Qualitative Research
In a provocative presentation, Dr. Cooper made a case for the expanded use of qualitative methods in the study of the process of outcomes. He explained the path of his conversion, only relatively recently appreciating the power of the methods that were used by his group in the original anesthesia critical incident studies in the 1970s. The concepts underlying qualitative research and the methods by which it is conducted have since become well defined and accepted in the social sciences. Although qualitative research is still looked upon with disdain by most who use purely a positivistic approach and quantitative methods, Dr. Cooper explained how he came to appreciate the power of developing and testing theories via the approach of grounded theory. His colleagues, Drs. Stephen Small and John Biglow, have been using this and other methods to study the effectiveness of training in Anesthesia Crisis Management and learn more about the process of critical events. A brief description was given of some of the features of qualitative research and grounded theory that differentiate it from quantitative methods. Dr. Cooper encouraged the audience to learn more about research methods in the social sciences, best done by collaborating with someone such as a sociologist. A list of references was provided (a few key references are listed below). Dr. Grobee, an epidemiologist working with Dr. Arbous, commented that there perhaps was not so much distance between the two approaches.
Patient Safety Initiatives in Greece
Dr. Helen Askitopolou described the many activities toward improving anesthesia patient safety in Greece. Rather than beginning with a study of outcomes, they began first by documenting the structure of anesthesia services and comparing these to other countries in Europe. As a result of this study, in 1995, the Ministry of Health formed a Committee on Anesthesia Safety in Greece.
In Thessaloniki, an outcome study from 25,000 cases at one hospital has been undertaken; the data are now being analyzed. In 1995, the forms for the AIMS study were translated to Greek and reports were solicited after announcement at a national meeting. Reports have not been forthcoming so plans are beginning for a more intense campaign to encourage participation. Planning is underway to form a Greek Foundation for Patient Safety in Acute Medicine. The broader scope is needed, she said later, to attract the support that will be required to establish such an organization in Greece.
Dr. Askitopolou described some details and results of the study of anesthesia services: a structured questionnaire was sent to anesthesia departments in 109 state hospitals, representing 70% of all anesthetics administered in Greece. The anesthesia directors of 43% of the hospitals were interviewed. The broadest spectrum of information was gathered about clinician training, case coverage, practices, equipment, etc. The survey covered about 300,000 cases in more than 500 operating rooms. Some of the findings so far were disturbing: PACU availability is limited, and those PACUs that exist have personnel with limited training. Adherence to the international standards for monitoring is limited, e.g., in 38% of departments, the presence of an anesthesiologist is not continuous, and only 46% of operating rooms have a pulse oximeter.
These findings led the newly-formed Safety Committee to propose to the Ministry of Health standards for availability of equipment, of staffing of anesthesia services, organization of PACU, accreditation of departments, of training departments, etc. The actions of the Ministry are awaited with some trepidation because of the cost implications of the recommendations. Dr. Gisvold noted that, in Norway, the publication of such standards alone by the Anesthesia Society made an impact even before the adoption by the government.
Dr. Askitopolou invited the audience to attend the European Society of Anesthesiology in Greece in September 1998, at which time she expects to report on progress in these many efforts.
Expanding the Scope of Anesthesia Outcome
Dr. Frederick Orkin again encouraged the ICPAMM group to take a wider perspective on mortality and morbidity in anesthesia. He stressed that safety should continue to be considered of great importance. Yet, he also believes that the continuing reduction in anesthesia deaths, which he traced from Beecher-Todd to recent times, argues for expanding concerns beyond the most serious adverse outcomes. The airline industry, which in the U.S. experience averages about one or two fatal accidents annually, has expanded its interests beyond safety, without diminishing the concern for safety. Similarly, he urges more interest and focus on patient satisfaction as an “outcome” to be measured and improved upon. He used an example from a study in the 1980s reporting that 84% of patients felt that they had been insufficiently treated for pain postoperatively. In the Winnipeg studies of Canadian anesthesia outcomes, it was found that minor outcomes, e.g., nausea and vomiting, were greatly under-reported. Further, patients having those outcomes were likely to report that the anesthesia was unsatisfactory. From a study that he first described at the last ICPAMM meeting and is just now submitting for publication, patients report that they are willing to pay to prevent symptoms such as PONV. Dr. Orkin urged that the research attention be turned to looking more at how the patient views the quality of the anesthetic rather than purely from the view of the provider.
Dr. Mckay felt that the view of the patient was being considered in that the wish of most importance to a patient is to wake up from the anesthetic. Dr. Orkin countered that the same is true for an airline flight, but the airlines still now pay much attention to issues of satisfaction. Others argued that such issues were not of critical importance compared to the more serious outcomes. The debate continued in this vein, with arguments about the relative importance of “safety” and “quality” and the relationship between the two. An argument was also made that safety issues would again become of greater interest as economic pressures remove the redundancies in hospital systems that have, in inexplicable ways, maintained safety despite complexity. Dr. Orkin countered that managed care had the potential to reduce adverse outcomes merely by reducing unnecessary surgery.
This ICPAMM meeting was sponsored by the Anesthesia Patient Safety Foundation (U.S.), and Dr. John Warden was responsible for all local arrangements and technical assistance. The next meeting of ICPAMM will be at the European Congress in Frankfurt in 1998.
Dr. Cooper, ICPAMM organizer, is Director of Biomedical Engineering at Partners Healthcare System, Inc., Boston, MA, Associate Professor of Anesthesia at Harvard Medical School, and a founding member of the U.S. APSF and a member of its Executive Committee.