1Department of Anaesthesiology and Intensive Care Medicine, University of Ulm, D-89070 Ulm, Germany. 2Department of Medical Informatics, University of Utah, Salt Lake City, Utah 84132, USA*Corresponding author
Accepted for publication: March 6, 2000
![]() |
Abstract |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Br J Anaesth 2000; 85: 27180
Keywords: anaesthesia, complications; records, anaesthetic
![]() |
Introduction |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
The German Society of Anaesthesiology and Intensive Care Medicine (Deutsche Gesellschaft für Anästhesiologie und Intensivmedizin (DGAI)) launched a study3 following the introduction of quality assurance and cost-containment regulations imposed by German health-care law. The aim of the project was to implement and evaluate a methodology for the standardized, long-term reporting of everyday anaesthesia outcome measures. Routine assessment of anaesthesia-related incidents, events and complications (IEC) was recommended for each patient and every anaesthetic procedure performed. A standard minimal data set is collected for national benchmark comparisons.4 Our department has been one of the pilot centres in this nationwide project.
This article attempts to (i) describe the methodology of the project concerning the immediate perioperative time-frame of anaesthesia care, and (ii) review the implementation and results of a routine data acquisition system that includes data collection, validation and staff education in one academic anaesthesia provider institution.
![]() |
Methods |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Quality measurement dimensions
The recommended outcome measurements of the whole project have three dimensions:
(i) perioperative reporting on IEC occurring during anaesthesia;
(ii) postoperative reporting on IEC occurring after anaesthesia;
(iii) postoperative patient satisfaction inquiry.
Perioperative IEC, according to Donabedians triad,1 represent quality-of-process measures of the immediate anaesthesia care process. Postoperative variables represent the quality of outcome. Methodology and results for quality of outcome have been described in detail elsewhere for a specific study period at our hospital7 and are beyond the scope of this article.
Standard data set
The standard data set was designed to reflect several key issues as to which patient (e.g. demographic and administrative data, risk factors, pre-existing disease, admitting surgical department) experiences which problems during what procedure (e.g. the kind and duration of anaesthesia and operation). The DGAI recommended the following definition of anaesthesia-related IEC:3
(i) IEC occur during the anaesthetists perioperative responsibility, i.e. from the induction of anaesthesia until discharge of the patient to the ward or intensive care unit (ICU), either from the postanaesthesia care unit (PACU) or directly from the operating theatre.
(ii) IEC lead to intervention by the anaesthetist.
(iii) IEC have or could have caused morbidity or mortality if the anaesthetist had not intervened.
All three conditions have to be fulfilled. By definition, it does not matter if the incident occurred by chance, was caused by the patients preoperative condition, or was a consequence of inadequate treatment by the responsible physician. Usually, physicians are able to assess IEC with standard clinical methods. The DGAI has defined 63 kinds of IEC (Table 2) and five ascending grades of severity as follows:
(i) IEC without any impact on postoperative careno additional postoperative care necessary (including near incidents).
(ii) IEC clinically important only for care in the PACUno impact on transfer to the ward.
(iii) IEC clinically relevant for postoperative careprolonged stay in the PACU or particular observation on the ward is clearly necessary.
(iv) IEC clinically important for postoperative careproblem cannot be solved satisfactorily in the PACU, and transfer to intermediate care unit or ICU necessary.
(v) IEC presumably severe, with permanent damage or death.
Acquisition of data
With the advent of our participation in the DGAI project in 1991, we decided to implement a uniform, paper-based, yet computer-readable anaesthesia record system for every monitored anaesthetic procedure. The system was developed according to ergonomic principles8 and was available at any anaesthetic location throughout the hospital. Each record set contained three pages:
(1) Preoperative evaluation with patient details, medical history, laboratory and pathophysiology findings, American Society of Anesthesiologists (ASA) classification of physical status, and surgical specialty.
(2) Intraoperative documentation of anaesthesia care, surgical procedure, monitoring techniques and notes on transfer.
(3) Postoperative course in the PACU, timing of operation and anaesthesia, length of stay in the PACU, type and severity of IEC, and the patients transfer destination.
The documented course of vital signs, the type and dosage of anaesthetics and the administration of other drugs in the operating theatre and PACU were not to be stored in the database. Before storing data in the database, every record set was validated in four ways and was corrected if necessary (Fig. 1).
|
(ii) Documentation clerks checked each record set for technical errors to guarantee proper scanning of the paper.
(iii) Rules implemented in the database system checked each record set for completeness, consistency and plausibility.
(iv) If there were any doubts in the second or third step of the validating system, the anaesthetists in charge of the documentation work group (B.S. or U.B.) were asked for assistance. If necessary, the anaesthetist who had performed the anaesthesia was asked to complete the data.
During a pilot phase from September 1991 to June 1992, clinical and documentation staff were trained to use the documentation system and to classify IEC. Written guidelines and educational meetings supported the staff education efforts. During the pilot phase, we conducted time studies and staff surveys. After implementation of the IEC methodology, our regular documentation activity began in July 1992.
Hospital, patient and anaesthesia characteristics
The data were collected from a tertiary university teaching centre completing about 18 000 anaesthetics per year. The case mix covered patients from an urban and a rural area, as the university surgical clinics function both as city hospitals and tertiary referral centres. Our department has a strict policy for the guideline-driven performance of anaesthesia according to our institutional practice recommendation.9 The current study was approved by the Institutional Review Board (Health Sciences) of the University of Utah.
![]() |
Results |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
|
|
|
|
|
|
![]() |
Discussion |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Definitions of anaesthesia outcome
There has been a consistent lack of agreement on how to appraise adverse events, adverse outcomes and quality in anaesthesia. Various countries and research groups differ widely in their criteria for reporting events: some restrict reporting to events with legal or disciplinary consequences;16 others report mortality or major morbidity only.11 12 Also, there is an inconsistent definition of the perioperative period: some reports cover the time from induction of anaesthesia until discharge from the recovery unit15 and others the time from the beginning of anaesthesia until 24 h postoperatively,11 or an even longer period.12 14 15 Furthermore, the method of data acquisition varies from voluntary individual reporting of critical incidents15 to prospective studies covering all anaesthetics during a certain period for one16 or several14 departments.
Instead of reporting particular findings in particular patient groups, the IEC definitions were designed to document everyday problems and their clinical impact on the entire patient population. Documentation was to continue during that precisely specified episode of care when the anaesthetist is completely and solely responsible for the patient, i.e. the induction and maintenance of anaesthesia in the operating theatre and through recovery and transfer from the PACU.
Accurate definitions of criteria for adverse events are difficult to set. Many problems cannot be measured by technical instrumentation (e.g. intubation problems, difficult punctures). Even if threshold values are measurable, rules about them may vary substantially in a complex clinical context. For example, a decrease in systolic arterial blood pressure from 130 to 80 mm Hg may or may not be classified as hypotension depending on whether it occurs in an old patient with cardiovascular disease or in a young healthy woman. For automated incident reporting, the defining of criteria may never reach the sophistication of the caring clinician.17 Therefore, the severity grading of an IEC is not defined by numerical values but rather by its clinical impact on the immediate and postoperative course.
The DGAI defined no criteria to determine what is causally responsible for an event (e.g. the anaesthesia or the surgical procedure). Most of the time, particularly in critical situations, this is hard to determine because interactions are abundant and complex. Even if it were possible to make them, such distinctions would offer no gain in quality information. In addition, it could be a common reason for doctors to refrain from documenting certain incidents for fear of being blamed.
International incidence comparisons
The overall incidence of IEC in our population, 22.0%, was higher than in other German reports using the same definitions.18 19 An overall incidence of 8% was shown to be low and inaccurate when the records were reviewed afterwards.22 Reported complication rates of 0.05%10 may be explained only if restricted inclusion criteria are considered (e.g. incidents with potential medicolegal consequences). Cooper et al.15 found a rate of adverse events of 18%. These events were called recovery room impact events and their definition is very similar to that of our IEC of grade II or above. The corresponding rate in our population was 10.7%. The incidence of moderate, serious or catastrophic sequelae in Coopers study was between 2.7 and 3.6%, compared with 3.3% in our study. However, our patient population, in contrast to Coopers, included patients transferred to an ICU immediately after surgery (see definition of grade IV). Cohen et al.16 found a rate of overall events of 2532% and a rate of major postoperative complications of 0.4%. Forrest et al.21 used a definition similar to ours and also five grades of severity. They found at least one severe adverse outcome in 4.9% of the study population. The total complication rate in the Gothenburg study of perioperative risk22 was 30%. However, it covered a longer postoperative period. Generally, the rate of severe adverse events is difficult to compare because of different severity grading systems and criteria for inclusion in studies. Pedersen et al.23 reported a complication rate of 3.6% when they included various severe cardiopulmonary events that were quite similar to our IEC of grade III and above.
In terms of specific events, Cohen et al. found more arrhythmia.14 16 Our incidences of severe cardiovascular events (hypotension, hypertension, tachycardia and arrhythmia) were slightly below the frequencies reported by Forrest et al.21 Unexpected difficult intubation is reported to occur with a frequency of about 1%,13 16 which is similar to our result. Reporting for bronchospasm was also similar. In our population, the incidence of hypoxaemia was higher (0.9 vs 0.4%) than in the report by Cooper et al.13 The study from Gothenburg24 revealed a lower incidence of respiratory events.
Other events play a minor role in our population, especially if we focus on severity grades above II. Fewer than 25% of all severe IEC did not involve the cardiovascular or respiratory system. For example, severe equipment IEC were rare (0.53%). Cooper et al. found that equipment was involved in only 4% of avoidable severe incidents.25 Their investigation, however, was carried out more than 10 years ago and technical standards and system checks may have improved in the meantime. Although incidence is clearly defined as the new occurrence of an event, this may be difficult to determine in practice. Consider, for example, a patient with a recent myocardial infarction undergoing a short emergency procedure. This procedure may be uneventful but the patient would be subject to close observation in the PACU for a prolonged time and then be transferred to an intermediate care unit or ICU for further cardiac monitoring. We suspect that many anaesthetists will document this situation incorrectly. This may explain the relatively high frequencies of low severity grades with intrinsically severe conditions. Near incidents, however, are included in the methodology.
Our incidence of disconnection, nausea and vomiting, and disorders of laboratory parameters was probably too low. In a recent study we found that the long-term variance of IEC incidences with a low severity grade are heavily dependent on discipline in staff documentation, and may therefore not be reliable indicators for quality comparisons.26
General problems in incident reporting
Many aspects will influence reporting. If, in addition to the usual anaesthesia record, any further reporting activity is required (e.g. finding and filling in a particular report form), there is a tendency among anaesthetists to report only the most severe events.27 For this reason, we have installed a record-keeping system with an integrated module for incident reporting. To prevent major bias, the reporting of incidents should be influenced neither by the anaesthetists fears nor by their individual opinions about what is worth documenting. Our department has kept a strict policy of confidentiality. Clearly, there is a trade-off between clinical education and departmental supervision capabilities for the validity of quality data.
Before starting the project, our department commissioned a documentation working group. This group instructed all medical staff in our department. Training in handling the record and classifying and grading IEC were carried out by the provision of formal guidelines and educational meetings. From then on, new staff members were tutored at regular intervals. Critical incidents were discussed in daily meetings. We learned that IEC reporting is highly sensitive to positive and negative educational effects on documentation discipline and that anaesthesia providers have to know the sources of influence and to be aware of the possibility of manipulating the system.26
We have learned that leading a project of this size is no task for those who are easily discouraged. It requires determination and enthusiasm, patience and persistence, complete support from the department, and considerable resources. This is necessary because motivating staff in what will lead to a substantial cultural change in documentation and practice is a tough barrier,28 particularly if the benefit is indirect. Detecting and reducing unreasonable variation in practice, however, should reduce costs and provide a predictable quality of patient care, which in turn helps to make a clinical facility with its employees more competitive.29 A good argument in discussions about the project has always been that it might be better to contribute proactively to ones own quality project than to wait until somebody else imposes their regulations. Later, during the course of the project, we experienced that passing coded information about the occurrence of IEC was an excellent way of streamlining information transfer in hand-over situations, and thus it aided decision-making even in individual cases (e.g. specific preparations for repeated anaesthesia if previous IEC are reported during preoperative assessment).
Another barrier is the need to convince medical professionals of the necessity of encoding information. Free text data is worthless for automated decision support and scientific work-up (e.g. statistics, epidemiology, economic analyses). Therefore, space for handwritten remarks, although ranked highly on the staff request survey, is limited on our anaesthesia record. Our preanaesthesia record, however, allows detailed coding of historical, clinical, laboratory and physiology data organized within functional systems and stratified according to severity. If information is stripped to the bare facts, the record is processed faster and more accurately, especially in hand-over situations.
Educating staff members to check optical mark-reading protocols properly may appear to be a simple task, but it is hard to achieve. In an initial workflow study, we found that most of the errors were of a technical nature related to these protocols.30 Therefore, we have relied on an integrated validating system involving point-of-care clinicians, dedicated documentation physicians and trained data clerks.
The technology employed represents devices commercially available at the time (1991) the project started: one optical mark reader, a relational database running on a single workstation, and batch processing of the paper forms, with considerable delay in the updating of information. Beginning in October 1997, we have introduced a new real-time system with workstations and printers in all preanaesthesia offices, several optical mark readers and printers in all PACUs, and a fully computerized patient record. It is still possible to document an anaesthetic completely on paper and to feed the data into the system later, because we have many anaesthetic locations that have no network connection and may not have one in the near future.
General critique of the IEC system
Comprehensive quality assurance in anaesthesia is not accomplished with external benchmark comparisons alone. In addition, there has to be an internal auditing and supervising system within each department which is concerned with the specific errors or mistakes of the individual care-provider and the clinical system environment (e.g. wrong drug or equipment usage). These occurrences are usually rare and are difficult to evaluate with standard statistics. The IEC system is not targeted at the thorough evaluation of rare, stochastic errors. The main purpose of the IEC system is to evaluate common problems in day-to-day anaesthesia patient care.
Unfortunately, there is no tradition and no clear move in Germany towards the national reporting and discussion of stochastic anaesthesia incidents, such as the Confidential Enquiry into Perioperative Death (CEPOD)31 in the UK or the Australian Incident Monitoring Study.15 An ideal quality assurance system for anaesthesia would have to include at least three components: a CEPOD-style reporting system; a benchmarking system operating on a state or national scale, and a culture of maintaining and animating local quality peer review groups.
![]() |
Acknowledgements |
---|
![]() |
References |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
2 Abenstein JP, Warner MA. Anesthesia providers, patient outcomes, and costs. Anesth Analg 1996; 82: 127383[ISI][Medline]
3 Deutsche Gesellschaft für Anästhesiologie und Intensivmedizin. Massnahmen zur Qualitätssicherung von Anästhesieverfahren. Anaesth Intensivmed 1992; 33: 7883
4 Deutsche Gesellschaft für Anästhesiologie und Intensivmedizin. Kerndatensatz Qualitätssicherung in der Anästhesie. Anaesth Intensivmed 1993; 34: 3316
5 Cooper JB, Newbower RS, Long CD, McPeek B. Preventable anesthethic mishaps. Anesthesiology 1978; 49: 399406[ISI][Medline]
6 Kersting T. Qualitätssicherung in der AnästhesiologieVerfahrensentwicklung und Verfahrenserprobung. Anaesth Intensivmed 1989; 30: 3437
7 Bothner U, Schwilk B, Steffen P, Eberhart LH, Becker U, Georgieff M. Perioperative Anästhesieverlaufsbeobachtungen, postanästhesiologische Visite und Befragung zur Patientenzufriedenheit. Anaesthesiol Intensivmed Notfallmed Schmerzther 1996; 31: 60814
8 Friesdorf W, Konichezky S, Gross Alltag F, Schwilk B. Ergonomics applied to anaesthesia record keeping. Int J Clin Monit Comput 1993; 10: 2519[ISI][Medline]
9 Georgieff M, Schirmer U. Klinische AnästhesiologiePraxisorientierter Leitfaden unter besonderer Berücksichtigung des operativen Vorgehens. Berlin: Springer-Verlag, 1995
10 Wang LP, Hagerdal M. Reported anaesthetic complications during an 11-year period. A retrospective study. Acta Anaesthesiol Scand 1992; 36: 23440[ISI][Medline]
11 Tiret L, Desmonts JM, Hatton F, Vourch G. Complications associated with anaesthesiaa prospective survey in France. Can Anaesth Soc J 1986; 33: 33644[ISI][Medline]
12 Andrews B. Reporting of anesthesia-related incidents: the New Jersey experience. J Clin Monit Comput 1998; 14: 712[ISI][Medline]
13 Cooper JB, Cullen DJ, Nemeskal R, et al. Effects of information feedback and pulse oximetry on the incidence of anesthesia complications. Anesthesiology 1987; 67: 68694[ISI][Medline]
14 Cohen MM, Duncan PG, Pope WDB, et al. The Canadian four-centre study of anaesthetic outcomes: II. Can outcomes be used to assess the quality of anaesthesia care? Can J Anaesth 1992; 39: 4309[Abstract]
15 Runciman WB, Sellen A, Webb RK, Williamson JA, Currie M, Morgan C. The Australian Incident Monitoring Study. Errors, incidents, and accidents in anaesthesia practice. Anaesth Intens Care 1993; 21: 50619[ISI][Medline]
16 Cohen MM, Duncan PG, Pope WD, Wolkenstein C. A survey of 112,000 anaesthetics at one teaching hospital (197583). Can Anaesth Soc J 1986; 33: 2231[ISI][Medline]
17 Sanborn KV, Castro J, Kuroda M, Thys DM. Detection of intraoperative incidents by electronic scanning of computerized anesthesia records. Anesthesiology 1996; 85: 97787[ISI][Medline]
18 Kersting T. Qualitätssicherung und Qualitätskontrolle in der Anästhesie. Anaesth Intensivmed 1991; 32: 30814
19 Schmidt K, Ladleif P, Hommerich P, Harke H. Qualitätssicherung: Klinikprofile im Vergleich. Anaesth Intensivmed 1991; 32: 3159
20 Tsekos E, Leier M, Boeden G, Blinzler L, Jagschies I, Heuser D. Qualitätssicherung in der Anästhesiezwischen Akzeptanz und Realisierung. Anaesth Intensivmed 1993; 34: 3259
21 Forrest JB, Cahalan MK, Rehder K, et al. Multicenter study of general anesthesia. II. Results. Anesthesiology 1990; 72: 2628[ISI][Medline]
22 Arvidsson S, Ouchterlony J, Nilsson S, Sjostedt L, Svardsudd K. The Gothenburg study of perioperative risk. I. Preoperative findings, postoperative complications. Acta Anaesthesiol Scand 1994; 38: 67990[ISI][Medline]
23 Pedersen T, Eliasen K, Henriksen E. A prospective study of cardiopulmonary complications associated with anaesthesia and surgery: risk indicators of cardiopulmonary morbidity. Acta Anaesthesiol Scand 1990; 34: 14455[ISI][Medline]
24 Ouchterlony J, Arvidsson S, Sjostedt L, Svardsudd K. Peroperative and immediate postoperative adverse events in patients undergoing elective general and orthopaedic surgery. The Gothenburg study of perioperative risk (PROPER). Part II. Acta Anaesthesiol Scand 1995; 39: 64352[ISI][Medline]
25 Cooper JB, Newbower RS, Kitz RJ. An analysis of major errors and equipment failures in anesthesia management: considerations for prevention and detection. Anesthesiology 1984; 60: 3442[ISI][Medline]
26 Bothner U, Georgieff M, Schwilk B. Validation of routine incidence reporting of one anaesthesia provider institution within a nation-wide quality of process assessment program. J Clin Monit Comput 1998; 14: 30511[ISI][Medline]
27 Jayasuriya JP, Anandaciva S. Compliance with an incidence report scheme in anaesthesia. Anaesthesia 1995; 50: 8469[ISI][Medline]
28 Lorenzi NM, Riley RT, Blyth AJ, Southon G, Dixon BJ. Antecedents of the people and organizational aspects of medical informatics: review of the literature. JAMA 1997; 4: 7993
29 Berwick DM. Controlling variation in health care: a consultation from Walter Shewhart. Med Care 1991; 29: 121225[ISI][Medline]
30 Schwilk B, Hahnel J, Friess L, Blessing S, Friesdorf W. Ergebnisse zur Verfahrensentwicklung und zum personellen Aufwand im Zusammenhang mit der Einführung qualitätssichernder Massnahmen. Anasth Intensivmed 1994; 35: 34956
31 Campling EA, Devlin HB, Lunn JN. Reporting to NCEPOD. BMJ 1992; 305: 252