ST. JOSEPH'S HOSPITAL, APPELLANT,

v.

WILLIAM COX AND MARTHA COX, APPELLEES; ERIC CASTELLUCCI, M.D., AND EMERGENCY MEDICAL ASSOCIATES OF FLORIDA, L.L.C., APPELLANTS, V. WILLIAM COX AND MARTHA COX, APPELLEES

NORTHCUTT, Chief Judge.

At the age of sixty-nine, William Cox suffered a stroke with devastating consequences. He contends that the effects of the stroke would have been prevented or reversed if he had been given a clot-busting drug when he was seen in the emergency room at St. Joseph’s Hospital. Mr. Cox and his wife sued the hospital, the emergency room doctor, and the doctor’s medical practice. The defendants have appealed following a jury verdict that awarded the Coxes substantial damages. They raise a number of issues, but we find it necessary to address only the trial court’s refusal to grant the defense motion for a directed verdict. Because the Coxes failed to meet their burden of proving causation, the defendants were entitled to a directed verdict. Accordingly, we reverse the judgment entered against them.

The testimony at trial provided an overview about the treatment of strokes. Whereas a hemorrhaging stroke is caused by bleeding on the brain, an ischemic stroke — the kind suffered by Mr. Cox- — is caused by a blood clot. Doctors are able to treat ischemic strokes with a tissue plasminogen activator drug, referred to as tPA, which dissolves blood clots. However, the drag must be administered within a short window of opportunity, no more than three hours after the onset of the stroke if administered intravenously and no more than six hours after onset if administered intra-arterially. Therefore, in order to consider using tPA, the treatment team must know when the stroke began. A further complication is that tPA therapy poses a significantly increased risk for hemorrhaging.

On the morning of his stroke, Mr. Cox was at a friend’s car dealership. He appeared normal when he first spoke with a visitor, but he was incapacitated and unable to speak or walk when the visitor returned fifteen to twenty minutes later. This person thus was able to narrow the onset of Mr. Cox’s stroke to a fifteen- or twenty-minute period, and this information apparently was relayed to the responding emergency medical personnel. However, the ER staff and doctor never obtained this information. Consequently, although Mr. Cox arrived at the hospital within the window for administering' tPA, the ER doctor never considered treating Mr. Cox’s stroke in that fashion because he did not know the time of onset.

*1126Another fact pertinent to this case was that Mr. Cox’s medical records indicated he had suffered a subdural hematoma approximately two-and-a-half years before his stroke. Given the increased risk for hemorrhaging associated with tPA, the drug is contraindicated for, among others, any patient who previously has suffered a subdural hematoma.

The case was submitted to the jury under section 768.13(2)(b)(1), Florida Statutes (2000), which provides for immunity from civil liability when a patient entered a hospital through the emergency room unless damages resulted “from providing, or failing to provide, medical care or treatment under circumstances demonstrating a reckless disregard for the consequences so as to affect the life or health of another.” The Coxes claimed that the emergency room doctor and staff acted with reckless disregard by failing to determine the stroke’s time of onset and, lacking this critical information, failing to administer tPA to Mr. Cox. They maintained that the drug would have prevented or reversed the effects of Mr. Cox’s stroke.

In moving for a directed verdict, the hospital and ER doctor argued that the Coxes failed to prove that Mr. Cox more likely than not would have benefitted from tPA. We review this issue de novo, evaluating the evidence and inferences therefrom in the light most favorable to the nonmoving party. Sims v. Cristinzio, 898 So.2d 1004 (Fla. 2d DCA 2005); see also Hancock v. Schorr, 941 So.2d 409 (Fla. 4th DCA 2006).

The plaintiffs’ evidence on causation came from two sources: trial testimony by Dr. Nancy Futrell and an opinion letter, later recanted, from Dr. Eddy Berges, a neurologist who treated Mr. Cox. Dr. Futrell is board certified in psychiatry and neurology with a particular specialty in strokes. She testified that she has given tPA forty to fifty times and that, on three occasions, she has given it to patients with contraindications. She testified about the successful outcomes for those three patients, but she said nothing about the results in her other cases.

On cross examination, Dr. Futrell admitted that she had never to her knowledge given tPA to a patient with a prior intra-cranial hemorrhage. Still, she questioned the accuracy of the reports indicating that Mr. Cox previously had suffered a subdural hematoma. She had not been able to review the older CT scans on which those reports were based because apparently they had been long since destroyed. But, when reviewing the CT scan taken at the time of Mr. Cox’s stroke, she did not see any signs that he had had a significant subdural hematoma. Therefore, she did not think he faced an increased risk with the use of tPA.

Dr. Futrell testified that she believed “to a high degree of medical probability” that “Mr. Cox, had he received tPA, would have had a very good recovery and have minimal or no neurologic deficit.” She based her conclusion on “a combination of what we can find in the medical literature and clinical experience that [she has] from having given the drug and seeing which patient it works on and, of course, we have the additional resource of our colleagues that we know what’s happened to their cases.” She did not elaborate further. Dr. Futrell then reiterated that her opinion was based on all she knows and all she has seen about this case. The singular fact she noted about Mr. Cox was that his brain appeared younger than his age because it showed minimal or no evidence of the shrinkage that typically occurs with aging.

On cross examination, the medical literature was discussed. Defense counsel ques*1127tioned Dr. Futrell about the “NINDS study,” which Dr. Futrell acknowledged was the seminal 1995 study that first established the efficacy of tPA therapy for the treatment of ischemic strokes.1 Dr. Futrell concurred in the NINDS study results, which showed that twenty percent of stroke patients recover without any medication and that the rate of successful outcomes increases to thirty-one percent when tPA is given.

In his initial opinion letter, Dr. Berges asserted that Mr. Cox had been a suitable candidate for tPA treatment and that he likely would have recovered from his stroke if the treatment had been given. Dr. Berges withdrew his opinion when he learned of the medical record of Mr. Cox’s subdural hematoma.

In negligence actions, Florida courts follow the “more likely than not” standard of causation, i.e., they require proof that the negligence “probably caused” the plaintiffs injury. Gooding v. Univ. Hosp. Bldg., Inc., 445 So.2d 1015, 1018 (Fla.1984). “‘A mere possibility of such causation is not enough; and when the matter remains one of pure speculation or conjecture, or the probabilities are at best evenly balanced, it becomes the duty of the court to direct a verdict for the defendant.’ ” Id. (quoting Prosser, Law of Torts § 41 (4th Ed.1971) (footnotes omitted)).

A medical expert’s opinion is not exempt from this rule. The supreme court has explained that “[t]he opinion of an expert is not sufficient to eliminate the necessity of proving the foundation facts necessary to support the opinion.” Harris v. Josephs of Greater Miami, Inc., 122 So.2d 561, 562 (Fla.1960) (approving rejection of worker’s compensation claim when one doctor connected claimant’s condition with occupation but that doctor’s opinion was based on pure speculation).

It has been said that the more-likely-than-not standard for causation is satisfied with evidence of a “fifty-one percent or better chance.” Jackson County Hosp. Corp. v. Aldrich, 835 So.2d 318, 328 (Fla. 1st DCA 2002). In this case, Dr. Futrell was the only expert witness at trial who testified that Mr. Cox probably would have had a good recovery from the stroke if he had received tPA therapy. We have carefully reviewed her testimony, but we can glean no facts that support her assertion that Mr. Cox would have had a fifty-one percent or greater chance of benefitting from tPA treatment.

Even assuming that Mr. Cox was an ideal candidate for the treatment — thus indulging for these purposes Dr. Futrell’s doubts that the treatment was contraindicated in Mr. Cox’s case because he previously suffered a subdural hematoma — -we find nothing in Dr. Futrell’s testimony or anywhere else in the evidence to suggest that Mr. Cox’s chances of benefiting from tPA therapy exceeded those of other patients. Dr. Futrell herself never testified that she had enjoyed a greater success rate with tPA than that documented in the NINDS study, i.e., thirty-one percent. She never compared any aspects of Mr. Cox’s physical condition to those of patients who had successful interventions in order to suggest that he, as opposed to sixty-nine percent of all patients, was predisposed to a positive outcome from tPA therapy. In short, Dr. Futrell’s opinion on causation was purely speculative. See id. *1128(reversing denial of directed verdict when expert’s opinion on causation was based on speculation). r

Dr. Berges’ initial opinion letter also suffered from the same fatal flaw, in that he offered nothing to support his bare assertion that Mr. Cox more likely than not would have recovered if he had been treated with tPA. See Harris, 122 So.2d at 563 (“ ‘[T]he basis for a conclusion cannot be deduced or inferred from the conclusion itself. The opinion of the expert cannot constitute proof of the existence of the facts necessary to the support of the opinion.’ ”) (quoting Arkin Constr. Co. v. Simpkins, 99 So.2d 557, 561 (Fla.1957)). Dr. Berges later rejected any suggestion that he would have administered tPA to someone with Mr. Cox’s medical history. But even notwithstanding this recantation, the doctor’s initial opinion that Mr. Cox was injured by his failure to receive tPA was speculative and therefore insufficient to establish causation.

The cases cited by the Coxes are distinguishable. In Marsh v. Valyou, 977 So.2d 543 (Fla.2007), the court addressed the admissibility of evidence under Frye v. United States, 293 F. 1013 (D.C.Cir.1923). In Hughes v. Slomka, 807 So.2d 98 (Fla. 2d DCA 2002), Lawrinson v. Bartruff, 600 So.2d 22 (Fla. 2d DCA 1992), and Olsten Health Services, Inc. v. Cody, 979 So.2d 1221 (Fla. 3d DCA 2008), the plaintiffs expert testimony on causation was not constrained by statistical evidence revealing success rates of less than fifty percent, as in this case. And the motion for directed verdict in Goss v. Permenter, 827 So.2d 285 (Fla. 5th DCA 2002), challenged the force of conflicting evidence on the question of whether the doctor had deviated from the standard of care — the doctor claimed that he had not been informed about the patient’s condition but a nurse’s testimony and her notes in the medical record provided a basis for the jury to reject the doctor’s testimony. That is distinguishable from a case such as this one, in which the best known rate for medical success is less than fifty percent.

In the absence of competent evidence to prove that the failure to treat Mr. Cox with tPA more likely than not affected his outcome, the trial court should have granted the defense motion for a directed verdict. Under these circumstances, we must reverse the judgment.

Reversed.

KELLY, J., and DAKAN, STEPHEN L., Associate Senior Judge, Concur.