The study by Girisgin et al¹ is an unusual one for the Emergency Medical Journal as it is an animal study. It raises some interesting issues:

1. How valid is an animal model of haemorrhage?

2. How valid is a rabbit model when investigating gut absorption of fluids?

3. How valid is blood pressure as a biomarker of volume status?

4. Why investigate the rectal route for fluid resuscitation?

To paraphrase HB Stoner,² one of the fathers of trauma research in this country, a good animal model should fulfil certain criteria:

• It must be reproducible

• It must be measurable

• Its intensity must be controllable.

Models of haemorrhage may be classified into controlled and uncontrolled haemorrhages. Controlled haemorrhage may be either of the “Wiggers” type, in which blood is withdrawn until a specified hypotensive blood pressure is achieved,^3,4 or of a fixed-volume type, in which a predetermined volume of blood is removed irrespective of the blood pressure response.⁵ Uncontrolled haemorrhage models involve the laceration of a major vessel—for example, a 4 mm tear of the aorta.⁶ Although the first two models may seem rather artificial compared with the third, they do fulfil Stoner’s criteria, and the uncontrolled haemorrhage models have shown that with normal clotting, even tears to the aorta are not invariably fatal, so that the models then resemble the controlled haemorrhage models.

The model used by Girisgin et al,¹ in which rabbits were bled out about 30% of their circulating blood volume so that the mean arterial pressure was reduced to about 34% of baseline, fulfils the criteria suggested by Stoner. It is not dissimilar to the …