dissenting.
I at first thought the result agreed upon by my learned colleagues was the correct one, even the obvious one. A close reading of the record, however, and especially the expert medical evidence, brought me to the opposite conclusion. Only by ignoring the evidence and by substituting its own assumptions about the effects of cocaine use could the Commission deny compensation to claimant.
The Commission’s alternative finding1 that the claimant was so intoxicated at the time of his fall “that it would have been impossible for the claimant to engage in his employment” is based upon speculation.2 So far as this record shows (and it includes the testimony of three experts — two pathologists and the claimant), cocaine does not *408intoxicate. I use the term “intoxicate” to mean something resembling alcoholic intoxication, manifested in its advanced stages by lack of coordination and by stupefaction — the kind of disabling intoxication, in other words, which would make it impossible for one to do his work, so that he could be said to have abandoned his employment. It is that level of intoxication that furnishes a defense to a workers’ compensation claim. Phillips v. Air Reduction Sales Co., 337 Mo. 587, 85 S.W.2d 551, 554-55 (1935); Page v. Green, 686 S.W.2d 528, 532 (Mo.App.1985); Brown v. Mid-Central Fish Co., 641 S.W.2d 785, 787-88 (Mo.App.1982); Coonce v. Farmers Insurance Exchange, 228 S.W.2d 825, 827-28 (Mo.App.1950); O’Neil v. Fred Evens Motor Sales Co., 160 S.W.2d 775, 779 (Mo.App.1942); see also 1A A. Larson, Larson’s Workmen’s Compensation Law § 34.21 (1985). Cocaine does not have that effect upon a person; it works upon the mood, by way of the cardiovascular system. A broad definition of “intoxication” might include the “highs” and the “lows” induced by cocaine use, but it is not that kind of intoxication that is a defense to a workers’ compensation claim.
I can account for the Commission’s finding of disabling intoxication only by an assumption on its part that cocaine has the same effect upon the user as alcohol. If one makes that assumption, then the Commission’s finding is perfectly logical on the evidence. The claimant is seen (by co-employee Jamison) reeling, staggering and collapsing outside the plant at 6:30 a.m. It is learned that on the previous evening and night, between 9:00 p.m. and 2:00 a.m., he took as many as seven injections of cocaine. If we assume that cocaine works upon the person like alcohol, then it is logical to conclude that claimant was intoxicated, and that the intoxication was caused by the cocaine.
Eliminate the basic premise that cocaine intoxicates like alcohol intoxicates, however, and the conclusion fails that claimant’s condition was cocaine-induced intoxication. The evidence is quite clear and undisputed that there are no similarities between alcohol and cocaine in their effects upon the user. Our knowledge of the familiar alcoholic intoxication is irrelevant in dealing with a case of cocaine usage. There is no evidence whatever that cocaine would have any tendency to cause a person to act in the way claimant was acting as he arrived at work on the morning of his injury.
Launching from its finding that claimant’s condition observed by Jamison was cocaine-induced intoxication, the Commission takes a second speculative leap. It reasons that one who is helplessly drunk at 6:30 a.m. is still too drunk to work at 8:30 a.m. when his fall takes place. This perhaps would be a logical inference if we were dealing with alcohol, but not where we are dealing with a condition caused in some other way. The evidence in fact makes it impossible that claimant’s condition as described by Jamison continued to the time of the fall.
It seems to me (if I may speculate along with the Commission) that the Commission was on the right track when it found that claimant’s fall was caused by an “idiopathic fall or blackout” which was “likely related to cocaine injections”. The condition earlier observed by Jamison was also an idiopathic fall likely related to cocaine injections. Claimant had omitted the evening meal the evening before and had had no breakfast on the morning of his fall. He had not slept the night before, although he had lain in bed. He was addicted to cocaine. This was not an isolated instance of abuse of his body; it represented a pattern. Fatigue and malnourishment caused the “idiopathic fall or blackout”, and also caused the earlier behavior observed by Jamison. In that way claimant’s actions were “related” to his cocaine use. But that is not to say that he was at any time “intoxicated”.
The testimony of employer’s pathologist Dr. Oxley does not support the Commission’s finding of claimant’s disabling intoxication. Dr. Oxley pointedly declined to give the answer tendered to him by employer’s counsel’s question “whether it was impossible for Mr. Haynes to physically and mentally engage in the course of his *409employment at R.B. Rice”. “Yes”, he answered, “I believe it would have been impossible for him to perform his work appropriately.” The addition of the word “appropriately” changes the lightning to the lightning bug. Dr. Oxley went ahead to amplify his answer, and his answer does not show disabling intoxication on claimant’s part. He gave it as his opinion that the claimant “had in his system enough cocaine to impair functioning”, and again that his findings were “consistent with being under the influence of cocaine”. That falls short of showing the kind of incapacitating intoxication that is a defense to a workers’ compensation claim, and it is clear from Dr. Oxley’s testimony that he intended no such meaning.
I would reverse the Commission’s award and would remand to the Commission for determination of benefits owing to claimant.
. The Commission's finding of claimant’s intoxication is referred to as an alternative finding because the primary finding of the Commission was that the claimant had sustained his injuries as the result of an "idiopathic fall" and were therefore not compensable. The Commission’s conclusion on this point was erroneous, because an injury received in an idiopathic fall is com-pensable if it is caused by a special hazard of the employment (in this case, scalding water under high pressure). Matthews v. Roadway Express, Inc., 660 S.W.2d 768, 770 (Mo.App.1983); Collins v. Combustion Engineering Co., 490 S.W.2d 394, 396 (Mo.App.1973); Howard v. Ford Motor Co., 363 S.W.2d 61, 67 (Mo.App.1962).
. Missouri has never dealt directly with the burden of proof with respect to the intoxication defense, but implicitly has treated it as an affirmative defense to be proved by the employer, see Page v. Green, 686 S.W.2d 528 (Mo.App.1985). All states which have dealt with the question, with the single exception of Texas, see Texas Employers Ins. Ass'n v. Monroe, 216 S.W.2d 659 (Tex.App.1948), have treated intoxication as an affirmative defense and have held that the burden is on the employer to prove the worker’s intoxication. See 100 CJ.S. Workmen’s Compensation § 521(d) (1958).