Is mortality reporting nearing the end?
WHO KILLED COCK ROBIN?
A. F. Merry
University of Auckland, Auckland, New Zealand
It has been said that the anaesthesia mortality rate is steadily decreasing. A rate of 1 in 200 000 is sometimes cited as current. In the review of anaesthetic mortality in Australia between 1997 and 1999 carried out under the auspices of ANZCA, a rate of 1 in 79509 is reported (with comment that the comparable figure from NSW in 1984 was 1 in 26000). [1] In the recently released review covering the years 2000 to 2002 the rate is 1 in 56 000. [2] The two rates are not comparable because the latter report is based on improved denominator data.
The work done in iteratively improving the quality of the data incorporated into each successive review is impressive. However, the discrepancy between rates in the last two reports illustrates that we did not know the mortality rate from anaesthesia in the 90s, and I suggest we still don’t. A similar point was made recently by Lagasse [3], although his contention that mortality may not be improving was challenged.[4] Arbous, Meursing and others have recently given us the following estimates: incidence of 24-h peri-operative death per 10 000 anaesthetics: 8.8 (95% CI 8.2-9.5); and of anaesthesia-related death 1.4 (95% CI 1.1-1.6) – much higher than the Australian estimates, presumably at least in part because of a different approach in obtaining the data.
There are several difficulties in estimating anaesthetic mortality: most reporting is voluntary; we seldom have a reliable figure for the denominator; sedation is not necessarily captured; we seldom know the casemix to which figures are being applied; there may be concern over confidentiality; and we don’t have an agreed definition of anaesthetic mortality (table 1). [6]
Criticism such as this may prompt the suggestion that we should give up trying to estimate mortality, and that more is gained by critical incident reporting or evaluation of morbidity. In reality the question is not one of or, but rather and. The recent ANZCA “Data Taskforce” concluded that we need data from a variety of sources. Without mortality data we cannot give patients a reliable estimate of risk, and we cannot begin to answer questions such as those relating to non-medical anaesthetists or the value of epidurals in major abdominal surgery. There are many ways in which patients can die during surgery and anaesthesia, and the cause of death is often multifactorial. Identifying these factors can be like finding the killer of Cock Robin. It is human nature for all concerned to deny their own contributions to things which go wrong, and to point to everyone else. It is clear that anaesthetists contribute to outcome after surgery. [5] [6] [7]
The potential for anaesthetists to improve outcome is particularly marked in high risk cases. More data demonstrating this will be critical to justifying the high levels of training and equipment expected for anaesthesia in this part of the world. I have no doubt that anaesthetists in Australia and New Zealand have contributed greatly to improving patient safety, but we need an objective measure of progress. A standardised index of anaesthetic mortality would be such a measure.
The acceptance by the Australian Committees of a uniform glossary of terms is major step forward (table 2): [1] However, it seems there is still no satisfactory definition of which cases should be considered for classification. There is a difference between data related to deaths within the first 24 hours after an anaesthetic and data related to dying in hospital and/or within 30 days (for example). From the perspective of patients and their families, the latter is probably more meaningful, particularly since many peri-operative problems only manifest several days later (myocardial infarction for example[ 8]). Reporting differs between States. A case can be made for mandatory reporting such as occurs in WA. Unified terms of reference are needed. The situation in NZ, where reporting was disbanded after confidential information was subpoenaed by the police, is not acceptable.
A significant contribution has been made by all those who have developed what is probably the best system in the world for quantifying anaesthetic mortality, including those who established the former Committee in NZ, and in particular Pat Mackay. We are hot on the trail of the killer of Cock Robin. Let us not slow down in our pursuit of data critical to the improvement of safety in anaesthesia.
Table 1. Illustrative definitions of anaesthesia related mortality; reporting is only mandatory in Western Australia (it was mandatory in New Zealand before its Committee was disbanded). Data from Tasmania is collected and reviewed by the NSW committee; and from the Northern Territory by the South Australian Committee. The ACT does not have a committee and does not report cases to any other State Committee.
| Region | Definition |
| Victorian Consultative Council on Anaesthetic Mortality and Morbidity (VCCAMM) (1976)* | A death which occurs during an operation or procedure (or within 24 hours of its completion) performed with the assistance of sedative, analgesic, local or general anaesthetic drugs or any combination of these, or, a death which may be the result (either partially or totally) of an incident during or after such operation or procedure even if more than 24 hours has elapsed since its completion. |
| South Australian Perioperative Mortality Committee (1987) | |
| The Anaesthetic Mortality Committee of Western Australia | All deaths occurring within 48 hours of an anaesthetic or deaths where the anaesthetic is thought to have been a contributing factor must be reported to the State Commissioner of Health. If the investigator is of the opinion that the death is likely to have been due in some measure to the anaesthetic, he prepares a case report for the Chairman of the Committee. |
* The Victorian Council is the only State Mortality Council which also has a brief to report on significant morbidity associated with anaesthesia.
Table 2. Part of “Glossary of Terms” [1]
| A. Death Attributable to Anaesthesia | |
| Category 1 | Where it is reasonably certain that death or morbidity was caused by the anaesthesia or other factors under the control of the anaesthetist. |
| Category 2 | Where there is some doubt whether death or morbidity was entirely attributable to the anaesthesia or other factors under the control of the anaesthetist. |
| Category 3 | Where death or morbidity was caused by both medical/surgical and anaesthesia factors. |
Time of Presentation
Sunday 14 May 2006 - 1330-1500
References
1. MacKay, P., Safety of Anesthesia in Australia. A review of Anaesthesia Mortality 1997-1999. 2002, Melbourne: Australian and New Zealand College of Anaesthetists.
2. Gibbs, N., Safety of Anesthesia in Australia. A review of Anaesthesia Mortality 2000-2002. 2006, Melbourne: Australian and New Zealand College of Anaesthetists.
3. Lagasse, R.S., Anesthesia safety: model or myth? A review of the published literature and analysis of current original data. Anesthesiology, 2002. 97(6):
p. 1609-1617.
4. Cooper, J.B. and D. Gaba, No myth: anesthesia is a model for addressing patient safety. Anesthesiology, 2002. 97(6): p. 1335-1337.
5. Slogoff, S. and A.S. Keats, Does perioperative myocardial ischemia lead to postoperative myocardial infarction? Anesthesiology, 1985. 62(2): p. 107-14.
6. Arbous, M.S., et al., Impact of anesthesia management characteristics on severe morbidity and mortality. Anesthesiology, 2005. 102: p. 257-68.
7. Merry, A.F., et al., First-time coronary artery bypass grafting: the anaesthetist as a risk factor. British Journal of Anaesthesia, 1992. 68: p. 6-12.
8. Akhtar, S. and D.G. Silverman, Assessment and management of patients with ischemic heart disease. Critical Care Medicine, 2004. 32(4 Suppl):
p. S126-36.
