Published: 03/02/2005, Volume II4, No. 5941 Page 25
How can the NHS ensure the delivery of its key goal of improving patients' quality of life when it has no routine system for measuring the very thing it is trying to improve?
Over 100 years ago, Florence Nightingale devised a simple three-point outcome measure for her patients: relieved, unrelieved and dead. Since then, the NHS has failed even to match this. Indeed, for many years the hospital activity statistics were reported as a composite of two measures - discharges and deaths - making no overt distinction between the two (see charts 1 and 2).
However, measuring health outcomes rather than healthcare outputs is not a technically difficult task.
Over the last 30 years there has been a tremendous growth in the development of instruments, ie questionnaires, designed to quantify health status.
There are currently over 1,200 different measures available; some focus on specific diseases and others on population groups.
Generic measures have also been developed - the short form 36, for example - which aim to capture a person's general health state (see chart 3).
So far, a key application for this sort of measure has been as outcome or effectiveness indicators in clinical trials and other types of evaluations of healthcare interventions. But there is no reason not to use such measures to evaluate other aspects of healthcare services.
For example, what we get for our spending, which hospital or clinical team is best and whether the NHS is becoming more or less productive could all be looked at from the data generated.
Combined with other data - such as the characteristics of patients - health state measures can assess the seriousness of conditions, how this changes over time and how clinical decisions to treat may vary. This last use is particularly important in monitoring changes in the clinical thresholds at which decisions are taken. For example, will doctors lower their admission threshold - and hence generate more demand - as waiting times reduce?
New measures are also needed because the National Institute for Clinical Excellence bases its assessments of the effectiveness of healthcare technologies largely on clinical trial data and the patient groups used in trials are often not representative of the real world.
Routine measurement of patients' health states in the NHS would provide data not only on patient reactions in practice, but also on variations between hospitals, clinicians or settings in the way technologies are delivered.
John Appleby is chief economist at the King's Fund.