Index · Topics · The Growth of Monitoring

04The Growth of Monitoring

Monitoring is the history of replacing an impression with a number. Each instrument in the sequence below answered a question that had previously been settled by looking at the patient and forming a judgement.

The problem created in 1846

Once a patient could be made insensible, the ordinary signs by which a conscious person reports distress were gone. The person attending had to infer the state of a body from outside it, and for the first fifty years the available instruments were a hand on the pulse, an eye on the chest, and experience. The literature of the period is full of descriptions of colour, of the character of the breathing and of the appearance of the eye, which is what careful observation looks like when there is nothing to measure with.

Writing it down

The first real advance was not an instrument at all. In the 1890s the practice began of recording the pulse and respiration at intervals through an operation, on a chart, as the operation proceeded. This sounds trivial and was not. A remembered impression cannot show a trend; a written series can. A chart makes a gradual change visible while it is still gradual, converts the attendant's attention into a record that another person can read, and creates for the first time a document that can be reviewed afterwards to work out what happened. Nearly everything later is an extension of that idea to more variables and shorter intervals.

The instrument sequence

The oximetry problem and its solution

The measurement of blood oxygen is worth setting out because the difficulty is instructive. Light passing through tissue is absorbed differently by oxygenated and deoxygenated haemoglobin, so in principle the ratio of absorption at two wavelengths gives the saturation. In practice the light also passes through skin, fat, bone, pigment and venous blood, all of which absorb, and all of which differ between people. Instruments built on this principle worked in the laboratory and could not be calibrated for a patient.

The solution, worked out in the early 1970s, was to discard the constant part of the signal entirely. Everything except arterial blood is steady from moment to moment; only the arterial component pulses. If the measurement is made on the pulsatile component alone, every fixed absorber cancels out and no calibration for the individual is required. This is a good example of a whole class of advance in the subject, in which the improvement comes not from a better sensor but from noticing which part of a signal actually carries the information.

What monitoring did to the field

Two consequences follow. The first is that continuous measurement turned the attendant's work into something that could be taught, examined and argued about from a record, which mattered enormously for the emergence of a speciality. The second is subtler: each new measurement changed what counted as an event. Before continuous oxygen measurement, a brief fall in saturation that resolved was not an event because nobody knew it had happened. The history of monitoring is therefore also a history of the steady expansion of the set of things considered worth noticing, which is why comparing incident rates across eras is almost meaningless.