In Nelson, the plaintiff alleged the defendant hospital was negligent when, having placed the plaintiff's left foot on a birthing bar, a nurse allowed the plaintiff's leg to drop, causing injury.
Determining the cause of the plaintiff's injury was described as a contest between two experts, based on the experts' assumptions and the plaintiff's evidence regarding the specific position the plaintiff was in at the time of the injury. There were two possible positions: the frog-leg position and a position with the leg on the birthing bar.
The trial judge rejected the evidence of the defendant's expert, Dr. McGraw, on the basis Dr. McGraw's opinion was not based on the scenario described by the plaintiff at trial a position with the leg on the birthing bar (see: para. 50). The trial judge accepted the evidence of the plaintiff's expert, Dr. Gilbart, on the basis that, unlike Dr. McGraw, Dr. Gilbart ostensibly based his opinion on the scenario described by the plaintiff at trial (see: paras. 51-52).
The Court of Appeal found the trial judge erred in his assessment of the expert evidence. Contrary to the trial judge's assessment, Dr. Gilbart did not base his opinion on an assumption consistent with the plaintiff's evidence at trial. Both experts based their opinions on an assumption the plaintiff was in the frog-leg position. That assumption was inconsistent with the plaintiff's evidence at trial she had her leg on the birthing bar. The plaintiff's description of her injury at trial was "that her leg had dropped from the top of a birthing bar" but that was not the description of the injury she provided to Dr. Gilbart. The trial judge failed to appreciate Dr. Gilbart’s opinion was based on a scenario different than the scenario described by the plaintiff at trial:
 The judge rejected Dr. McGraw’s evidence, in part because his under standing of the frog-leg position was not the position the judge found Ms. Nelson’s left leg was in when it dropped. The same appears to be true of Dr. Gilbart’s understanding of the frog-leg position.
The judgment was fatally flawed because of this error in weighing the evidence:
 … The judge found there was a temporal connection between the labour and delivery and the emergence of symptoms. The key causation issue, however, required him to determine whether dropping the respondent’s leg from the birthing bar would generate sufficient shearing force to result in a de novo tear of the acetabular labrum. I cannot say there was no evidence in support of the respondent’s claim but it is clear there was an error in weighing the causation evidence.
 In what was largely a contest between two experts, the trial judge rejected one, believing that expert based his opinion on an inaccurate assumption, and accepted the other expert’s views in the mistaken belief the latter had not done the same thing. It cannot be said that the trial judge regarded the difference between the positions as immaterial; it was expressly relied upon as one basis for rejecting the evidence of Dr. McGraw.
The trial judge’s manifest error in his findings regarding the evidence of one expert as against the other destroyed the basis for his findings. The Court of Appeal referred to the test in Whitehouse v. Reimer (1980), 116 D.L.R. (3d) 594 (Alta. C.A.), adopted by both the Ontario Court of Appeal and British Columbia Court of Appeal :
 In the circumstances, in my view, the appellant can fairly say that the case meets the test described by Laycraft J.A. in Whitehouse v. Reimer (1980), 116 D.L.R. (3d) 594 (Alta. C.A.) at 595; and adopted by Doherty J.A. in R. v. Morrissey (1995), 22 O.R. (3d) 514 (C.A.); and by Bennett J.A. in Tambosso v. Holmes, 2016 BCCA 373:
Where a principal issue on a trial is credibility [or I would say, reliability] of witnesses to the extent that the evidence of one party is accepted to the virtual exclusion of the evidence of the other, it is essential that the findings be based on a correct version of the actual evidence. Wrong findings on what the evidence is destroy the basis of findings of credibility (or reliability).
The Court of Appeal ordered a new trial.