In February last year an anaesthetist was convicted of manslaughter, when the patient under his care suffered severe brain damage leading to coma and death. What should have been a simple technical mishap, easily recognised and corrected, went unnoticed during his absence from the operating theatre.
Damages of almost 1 million Pounds were awarded in September last year to a young woman left with severe physical and mental handicap after an anaesthetic error during an operation to extract wisdom teeth. This tragedy was all the more poignant because her parents were later killed in a train disaster.
Advertisement
Such headlines seem to appear with ever increasing frequency. Medical defence societies, who insure doctors for professional liability, have reacted to the rising tide of court settlements by increasing premiums dramatically. The time-honoured principle of flat-rate premiums for all doctors was broken by a new insurance company offering cut-rate premiums for doctors in ‘low risk’ specialties. Differential premiums were introduced and anaesthetists found theirs had increased fourfold in a year. Projected increases in premiums threatened to devour whole salaries and the government was forced to act.
Since 1 January 1990, health authorities have indemnified hospital doctors for their NHS work. this has significant implications for safety standards in anaesthesia.
Are the headline cases insolated incidents, or has the practice and science of anaesthesia failed to keep up with the increasing demands of modern surgery? Accidents have always happened, but patients are now much more willing to sue, attracting the attention of the press. Safety standards have in fact improved and the chances of dying of an anaesthetic accident in Britain were recently estimated as 1 in 180,000. Can the toll of deaths and injuries be reduced further?
Case studies are instructive. In one recent medical disaster, the patient was 33, a fit man admitted for emergency surgery after the sudden loss of sight in one eye. Delicate surgery was required to repair a torn retina, the light-sensitive membrane at the back of the eye. Wearing magnifying glasses, the surgeon was aware only of the minute details of his task. Indeed the operating theatre was in semi-darkness to allow the surgeon a clear view inside the eye.
To prevent any movement of the eyeball in the hours of painstaking surgery, the anaesthetist had administered a muscle-paralysing drug, a routine procedure in many operations. Unable to breathe for himself, the patient was connected to a ventilator which gently pumped oxygen and anaesthetic gases in and out of the lungs.
Hidden beneath the surgical drapes covering the patient’s neck and chest, the vital breathing tube became disconnected. Starved of oxygen, the brain starts to die within minutes. As those minutes passed, the heart tried desperately to supply more oxygen to the brain, rapidly increasing the pulse and blood pressure. Finally, even the heart faltered, slowed and stopped. And where was the anaesthetist in those precious minutes when disaster might have been averted? Out of the theatre, getting a drink.
‘I was rather tired,’ he said, having gone to bed at 3.30 am on the day of the operation. Returning to the theatre, he suspected a malfunction in the monitoring equipment, as no pulse or blood pressure was registered. Not until the surgeons threw off the surgical drapes to reveal the blue and lifeless body did he realise the enormity of his mistake. Cardiac massage was started by the surgeon who then noticed the disconnected breathing tube. When oxygen was supplied once more, the heart was made to beat but the patient remained comatose and died six months later. The sad aspect of this tragedy is the all-too-familiar sequence of events. Inadvertent disconnection of the breathing circuit is one of the main causes of accidental anaesthetic deaths.
What could be done to prevent such a disaster? An analogy is often made between the process of piloting an aircraft and anaesthetising a patient. Take off and landing are busy and demanding tasks involving the full skills of the pilot, contrasting with the longer periods of level flight when only vigilance is required. likewise, anaesthetists are fully occupied at the start and end of an anaesthetic, and most maintain a high degree of vigilance as they monitor and chart the course of the patient during the surgery.
Unlike airline pilots, however, anaesthetists do not have their hours of work restricted by law, and shifts lasting 48 hours or more are commonplace. Also in marked contrast to the airline industry, accidents are not generally subject to inquiry and ‘near misses’ often go unreported. Indeed, no formal procedure exists for the documentation or study of these critical incidents in which potential disaster is averted by the prompt action of the anaesthetist.
Where there is an inquiry, usually in a court of law, the purpose is to apportion blame, not unravel the many contributing factors which led to the disaster, or make recommendations for prevention. The jury does not rule on the design of breathing tube connectors, nor on the hours of work of duty anaesthetists. So any attempt by hospitals to improve patient safety will inevitably be a pragmatic one.
Litigation and awards for damages became common in American courts several years before they did in Britain. As a result in 1988, the nine hospitals at Harvard, who fund ther own medical insurance (the Controlled Risk Insurance Company), commissioned a retrospective study of litigation associated with anaesthesia. A detailed analysis was made of all the deaths and serious injuries attributable to anaesthesia, over the preceding 12 years. The cases sought were those in which a patient came to sudden and unexpected harm. Those who died as a result of known medical or surgical conditions were excluded from the study. Eleven cases were found out of 1.3 million anaesthetics.
By far the commonest cause of injury in these cases was inadequate ventilation of the patient’s lungs. This happened in patients accidentally disconnected for a ventilator, and in cases where the anaesthetist failed to notice that the patient’s breathing had become severely depressed.
The Harvard study concluded that patient safety monitoring could have averted disaster by giving early warning to the anaesthetist and allowing corrective action to be taken. Indeed, the report estimated that some $5 million in projected insurance payouts could have been saved had patient safety monitors been routinely applied. This pragmatic approach ignores the underlying causes of the various mishaps but seeks to prevent catastrophe by raising the alarm before harm is done.
In 1985, within the period of study, the Harvard Department of Anaesthesia had already formulated standards for minimum patient monitoring. These were mandatory, to be applied to every anaesthetic administered at Harvard. A summary of the standards is shown in the Table below. Not all the standards demand the use of technology. Anaesthetists are also expected to use their basic senses, to feel for a pulse, observe the colour and breathing of the patient, or listen to the heartbeat with a stethoscope.
In the three years after the introduction of the mandatory standards, 319,000 anaesthetics were administered without a major preventable anaesthetic injury, and the rate of anaesthetic accidents fell threefold. With some changes, the Harvard standards were adopted by the American Society of Anesthesiologists (ASA) in 1986. Malpractice insurance premiums for anaesthetists have actually fallen at a time when all other doctors have seen large rises.
Some of the monitoring devices, such as the ECG (electrocardiogram) and automatic blood pressure machines, are well established. However, two new instruments, the pulse oximeter and the capnograph, represent major advances in monitoring technology. The importance attached to these two devices can be judged by the 20 per cent insurance discount offered in Massachusetts for those anaesthetists who use these instruments as part of their implementation of the ASA standards.
Continuous measurement of the oxygen content of arterial blood by a non-invasive instrument was the holy grail of monitoring technology. Decades of research produced only machines that were large, unreliable and required recalibration at frequent intervals. The goal seemed unattainable. But the pulse oximeter is just such an instrument. It is a compact, portable device which is connected to the patient by a simple probe that fits over a finger or earlobe. Shining light through the tissues, the pulse oximeter detects and analyses the flush of oxygenated blood that pulses through the body with each heart beat. A liquid crystal display (LCD) shows the pulse waveform, together with numerical data on the heart rate and oxygen content of arterial blood. The latter is expressed as the percentage oxygen saturation of haemoglobin, the specialised protein molecule that carries oxygen in the bloodstream.
In health, and during anaesthesia, arterial haemoglobin is almost fully saturated with oxygen and the pulse oximeter gives a reading between 95 and 100 per cent. Any interruption to supply of oxygen results in a rapid fall in saturation, displayed on the screen but also conveyed audibly. Each pulse is signalled by an audible ‘beep’, of a pitch proportional to the oxygen saturation. So the anaesthetist is instantly warned by the change in pitch, without having to watch the display. In addition, alarms are set on acceptable limits for oxygen saturation and pulse rate. These give audible and visual warnings if the limits are transgressed.
The pulse oximeter could have saved the life of the young man who died during eye surgery. Even if the audible beeps were switched off, the alarm would have sounded before oxygen saturation fell to dangerous levels, allowing time for the anaesthetist to investigate the cause of the change. Disconnection of the breathing circuit is a common and well recognised hazard, easily found with a simple check.
Of course, an attentive anaesthetist would have detected the problem without the aid of a pulse oximeter, for instance by the lack of chest movements, or later by the rapid rise in pulse rate and blood pressure. But it is impossible for even the best anaesthetist to be 100 per cent alert all the time and devices such as the pulse oximeter provide invaluable back up.
The second important instrument is the capnograph. This detects carbon dioxide in exhaled breath. With every breath we take, oxygen is absorbed into the blood stream, conveyed to the tissues, and exchanged for carbon dioxide in the process of metabolism. The carbon dioxide is carried to the lungs and exhaled in place of the volume of oxygen absorbed. The capnograph samples gas continuously from the windpipe and displays the concentration of carbon dioxide as a square wave form on a screen. Fresh inhaled gas contains no carbon dioxide, so when the patient inhales the concentration falls to zero. With exhalation, the gas from deep within the lung is sampled, and the carbon dioxide concentration rises to a plateau. Both the pattern and absolute values convey much essential information which is used to generate alarms.
The respiratory rate is measured from the frequency of the square wave. Appropriate alarm limits are set to detect dangerously slow breathing or a disconnection in the breathing circiut (zero breaths). A failure of fresh gas supply during anaesthesia is signalled by a rise in the amount of carbon dioxide inhaled, which is normally close to zero.
The capnograph can also prevent one other uncommon but potentially fatal error, oesophageal intubation. A patient who is paralysed and ventilated during surgery will have a breathing tube passed into the windpipe at the beginning of the anaesthetic. For a variety of technical reasons, it is sometimes impossible for the anaesthetist to have a direct view of the windpipe, so the breathing tube is introduced blind. If the tube passes into the adjacent oesophagus, the stomach will be ventilated rather than the lungs, depriving the patient of oxygen. Although a variety of tests are done to check the correct placement of the tube, none give as reliable an answer as the capnograph, which detects carbon dioxide with each breath from the lungs, but will detect none form the stomach.
I have illustrated some of the ways in which patient monitors can make anaesthesia safer by giving early warning of unexpected events or accidents. Alarms are also fitted to equipment monitors, such as the oxygen concentration meter, which guards against the supply of too little oxygen to the patient. But patient monitors have a much wider function in providing feedback about the patient’s condition, so that a physiological norm can be maintained during the strain of illness and surgery. The analogy is pilots adjusting the controls to maintain level flight during turbulence. They cannot perform the task without continuously watching the altimeter.
The body mounts a stress response to surgery with profound effects on the pulse, blood pressure, hormone levels and so on. The anaesthetist must oppose these effects with anaesthetic drugs and gases to maintain physiological variables within a tolerance that depends on the fitness of the patient. When a patient has severe narrowing of the coronary arteries, for example, failure to maintain physiological stability may result in a heart attack during surgery. This maintenance is a complex and demanding task – it is both the art and science of anaesthesia. During major surgery, more than 10 physiological variables may have to be monitored and these results continuously integrated with the anaesthetist’s knowledge of physioilogy and pharmacology to modify the anaesthetic and maintain homeostasis.
In the era before intensive monitoring, surgery within three months of a heart attack carried a 40 per cent risk of a repeat ‘coronary’, which was often fatal. The very best modern practice, with intensive care and monitoring continuing into the postoperative period, reduces this risk to below 5 per cent.
But Britain has lagged behind the US, both in the litigation stakes and in standards of monitoring. So just how safe is anaesthesia in Britain? Where Britain excels is in answering that question. The Confidential Enquiries into Peri-Operative Deaths (CEPOD) was set up under the auspices of the Associations of Anaesthetists and Surgeons of Great Britain and Ireland. Published in December 1987, the report contained a detailed analysis of every death within 30 days of an operation in three representative health regions. In the 12-month period of study, more than half a million operations were performed, and 4034 deaths were reported to the CEPOD. Confidential reports were obtained from the surgeons and anaethetists involved in the care of the patients who died. These reports were scrutinised by independent surgical and anaesthetic assessors who judged the quality of care and listed the factors which contributed to the death of the patient.
Most patients who died were elderly, and succumbed to the natural progression of their illness, often following emergency surgery. But in 14 per cent of the cases, deficiencies in anaesthetic practice made some contribution to the death of the patient. More often, poor surgical treatment was blamed.
Only three patients died solely as a result of anaesthetic factors, in the sort of accident we have already considered. This represented a risk of 1 in 180,000, the figure quoted earlier. No comparable study has been undertaken anywhere in the world and the Confidential Enquiries are now extended nationally.
In 1987 when the report was published, my hospital possessed only one pulse oximeter, and that was on loan from the manufacturer for a trial period. Since then, both pulse oximeters and capnographs have come into widespread, but not universal, use. The Association of Anaesthetists of Great Britain and Ireland has published monitoring guidelines, similar to the American standards, but these are not mandatory. Restraints on capital expenditure in the NHS still limit the supply of new monitoring equipment.
But how long will it be, I wonder, before a health authority is faced with 1 million Pounds bill for damages in an anaesthetic accident? Pressure for mandatory monitoring standards is going to come, I suspect, from the new NHS managers.
———————————————————————— Harvard’s minimum monitoring standards ———————————————————————— The anaesthetist must be present throughout the administration of the anaesthetic. ———————————————————————— The blood pressure and heart rate must be recorded at least every five minutes. ———————————————————————— The electrocardiogram shall be displayed continuously during the whole anaesthetic. ———————————————————————— Both the breathing and circulation of the patient shall be continuously monitored, by a variety of available means. ———————————————————————— When a ventilator is used, a disconnection warning device must be employed. ———————————————————————— The concentration of oxygen in the breathing circuit will be measured continuously, by a device with a minimum concentration alarm. ———————————————————————— There must be readily available means to measure the patient’s body temperature. ————————————————————————
* * *
An electronic finger on the pulse
Haemoglobin is the substance which carries oxygen in the blood. Saturated with oxygen, it has the bright red colour of arterial blood, and stripped of oxygen, the bluish colour of venous blood. This colour change is the key to pulse oximetry. Absorption of light is measured at two wavelengths, 660 nanometres (red light) and 940 nanometres (infrared light). The ratio of these absorption coefficients describes the proportion of saturated and desaturated haemoglobin.
Light emitting diodes (LEDs) provide high intensity light at the two selected wavelengths. Shone through a fingertip, the transmitted light is detected by a silicon photodiode. But early attempts at optical oximetry were confounded by the variable background absorption of light by the skin, pigments and tissues. Recalibration was required for every subject and at frequent intervals thereafter, making the device inconvenient and unreliable.
The breakthrough, which led to the modern self-calibrating instrument, was the observation that light transmitted through tissues has a small pulsatile element superimposed on the background intensity. Amplification of this pulsatile signal reveals the signature of the arterial pressure waveform and this component of absorption is found to depend only on arterial blood within the tissue. In contrast, venous blood flows steadily and contributes only to the static background absorption.
The silicon diode converts the transmitted light into electrical current, with DC and AC components corresponding to the static and pulsatile components of light transmission. Electronic signal processing separates out the AC component which contains information only about the colour changes in arterial blood. The AC signal has a different amplitude at the two wavelengths and the ratio of these amplitudes predicts the degree of oxygen saturation of arterial blood.
Total light transmission varies greatly, depending on the thickness of a finger or pigmentation of the skin. The DC, or static, transmission is measured at each wavelength and used as a scaling factor to correct the corresponding variation in amplitude of the pulsatile signal. Optimum performance of the photodioed is ensured by varying the brightness of the LEDs so that the intensity of tansmitted light falls within the dynamic range of the photodiode.
The two LEDs are strobed so that the photodiode measures red, infrared and ambient light in turn, in a cycle repeated many hundred of times during each pulse. Ambient light levels can then be substracted from the measurements of transmitted light.
Operating theatres are full of electrical equipment that produces high levels of electromagnetic interference. In addition, artifacts may be produced by movement by the patient (such as shivering) or flickering lights. The best pulse oximeters employ sophisticated algorithms for the detection and rejection of artifact and also generate warning messages such as ‘Probe off patient’ or ‘Pulse signal small’.
The net result is an instrument of remarkable accuracy which gives continuous non-invasive measurement of arterial oxygen saturation in a wide variety of clinical settings.
* * *
A balancing act for carbon dioxide
Carbon dioxide has an importance in the body far beyond that of mere by-product of metabolism. Released from cells during the production of energy from glucose, carbon dioxide diffuses rapidly into the bloodstream and hence into the gas spaces in the lungs. Here the rising concentration of carbon dioxide is diluted with each fresh breath, finding a steady state in equilibrium with arterial blood which circulates back to the body. A proporation of the carbon dioxide disolved in plasma is hydrated to form carbonic acid, H2CO3, which lowers plasma pH. Chemoreceptors in the brain stem detect any change in pH and take corrective action by signalling the respiratory control centre. A fall in pH leads to hyperventillation which lowers the concentration of CO2, in the lung and hence the blood. Less carbonic acid is formed and the pH rises back to normal.
These compensatory mechanisms are demonstrated by the rapid increase in breathing during exercise. Cellular processes depend on enzymes which are exquisitively sensitive to changes in pH, such that a fall from the normal 7.4 to below 7 may prove rapidly fatal.
During anaesthesia, the patient’s breathing is often controlled by a mechanical ventilator, shortcircuiting the normal pH control mechanism. Over or under ventilation by the ventilator will disturb the patient’s plasma pH with potentially serious consequences in the unfit or elderly, producing for instance, irregularities of the heart rhythm. These hazards are avoided by the use of a capnograph.
Gas samples at the end of exhalation comes from deep within the lung and has a partial pressure of carbon dioxide close to that of arterial blood. This ‘end-tidal’ carbon dioxide is shown as the plateau of the capnograph waveform in the figure. Ventilation is adjusted to maintain a normal end-tidal carbon dioxide and hence arterial pH.
Capnography also warns of several uncommon but serious complications of anaesthesia. A sudden fall in end-tidal C02 occurs when blood flow to the lung is obstructed, for instance when air accidentally enters the circulation and heart chambers. A rapid rise is a sign of malignant hyperpyrexia, a very rare but potentially fatal reaction to certain anaesthetic drugs. Early recognition allows an antidote to be administered.
The capnograph was designed as a physiological monitor. Its added role, in the detection of tehnical mishaps such as disconnection of the breathing circuit, is part of the new concept of patient safety monitoring.
Dr Robin Youngson is an anaesthetist now working in New Zealand