Economic analysis of Ottawa knee rule

In Bandolier 49 we carried an article about the way in which the Ottawa knee rule was determined and tested in a large trial. The clinical decision rule helps determine whether an X-ray is necessary in acute knee injuries [1]. It examines five simple clinical findings and has sensitivity of 100% for detecting clinically important fractures. Use of the rule reduced use of X-rays by 26% and reduced average waiting time by 33 minutes per patient.

The question is whether it not only saves in X-rays and waiting times, but what is the economic impact of using the rule. An economic evaluation demonstrates substantial cost reductions of about $33 (£20) per patient in a north American setting.


A decision model was used that compared usual practice with practice based on the Ottawa knee rule. The model was populated with information determined from large comparative studies. A probability estimate was specified for each variable, and simulations performed by randomly drawing the value for each variable from its corresponding distribution. This was done 100,000 times to estimate variability. Various US and Canadian costs were used, including the costs of medicolegal awards that could result from a missed fracture. Sensitivity analyses were also performed.


The results are shown in the Table. Over $30 per patient were saved by using the Ottawa knee rule in either US or Canadian systems. Sensitivity analyses showed that the use of the rule was cost saving unless the sensitivity fell to levels below those seen in studies of the rule.

Table: Impact of Ottawa knee rule

(1996 US dollars per patient, with 95% confidence interval)


Cost of usual practice

Cost of Ottawa knee rule


USA Medicare

420 (114 to 1097)

386 (118 to 1028)

34 (24 to 47)


320 (95 to 835)

289 (97 to 763)

31 (22 to 44)



Use of the Ottawa ankle rule was also shown to be cost saving ( Bandolier 21 ). The $30 saved on a per patient basis may not seem much, but it is resources used without any perceptible benefit for the individual patient, and with the possibility of harm ( Bandolier 12 ). In UK terms the £20 equivalent may be a bit high because UK healthcare costs are lower than those in north America.

The issue is the number of people who come into casualty departments with injured knees. In the USA it is over one million a year. If the equivalent number for the UK on a population basis was about 20%, it could still amount to £2 million in the UK (200,000 people multiplied, say £10 saved per patient).

This may be a small amount compared with the total spent in the NHS, but it is just one example of how doing a small thing well frees up resources to benefit patients.


1 IG Stiell, GA Wells, RH Hoag et al. Implementation of the Ottawa knee rule for the use of radiography in acute knee injuries. JAMA 1997 278: 2075-9.

2 G Nichol, IG Stiell, GA Wells et al. An economic analysis of the Ottawa knee rule. Annals of Emergency Medicine 1999 34: 438-447.