HARRY BABUSH AND EVELYN BABUSH, APPELLANTS,

v.

AMERICAN HOME PRODUCTS CORPORATION, APPELLEE

HERSEY, Judge.

This is an appeal from a summary final judgment in a products liability case. The trial court held that the action was barred by the statute of limitations.

In 1971, in order to lower his cholesterol level, Harry Babush began taking the drug clofibrate, marketed by appellee as Atrom-id-S. At some point while taking the drug he developed dermatitis. Then in 1977, two months after he stopped taking the drug, his doctors diagnosed primary biliary cirrhosis, a fatal liver disease. There is no evidence that any correlation between the drug and the disease was mentioned in connection with the diagnosis. However, according to the allegations of the complaint, “A World Health Organization study into the benefits and harms of ingestion of clofibrate was undertaken beginning in 1965 and as early as 1973 problems with the drug ... were identified.” Those problems included liver damage and dermatitis.

In 1980 the physicians who were treating Babush estimated that he had only a 50-50 chance of living another five years. Four years later his health had deteriorated so badly that he underwent two liver transplant surgeries and a further corrective surgery.

In 1986 Babush’s neighbor, a retired pharmacist, allegedly told him that Atrom-id-S “was toxic to the liver.”

In 1988, Babush filed the complaint in this action against the drug manufacturer, joined by his wife on her dependent claim for loss of consortium.

The issue is when does the statute of limitations commence to run on a products liability claim. It is appellee’s position that the analysis and result is the same as in a medical malpractice suit. Appellant postulates that there is a difference. We agree with appellant and reverse.

Under earlier law a single statutory provision applied both to medical malpractice claims and to product liability claims. In Nardone v. Reynolds, 333 So. 2d 25 (Fla.1976), our supreme court addressed that statute in response to certified questions from the United States Court of Appeals for the Fifth Circuit. The Nardone court said that, in applying section 95.11(4), an action would be barred four years after either “the plaintiff has notice of the negligent act giving rise to the cause of action or ... the plaintiff has notice of the physical injury which is the consequence of the negligent act.” Id. at 32 (citing City of Miami v. Brooks, 70 So. 2d 306 (Fla.1954)).

Notwithstanding various changes in the statute of limitations, the supreme court has continued to apply the rule fashioned in Nardone that the limitations period in a medical malpractice action commences to run when either “the plaintiff has notice of the negligent act ... or when the plaintiff has notice of the physical injury_” Moore v. Morris, 475 So. 2d 666, 667 (Fla.1985). See also Barron v. Shapiro, 565 So. 2d 1319 (Fla.1990).

The statute of limitations applicable to this products liability case is four years. § 95.11(3)(e). The statute commences to run “from the time the facts giving rise to the cause of action were discovered or should have been discovered with the exercise of due diligence_” § 95.031(2), Fla. Stat. (1989).

The medical malpractice statute provides that the limitations period commences to run from “the time the incident giving rise to the action occurred or within 2 years from the time the incident is discovered, or should have been discovered with the exercise of due diligence_” § 95.11(4)(b), Fla.Stat. (1989).

If these two statutes operate differently one from the other, the distinction would seem necessarily to be occasioned by the use of the phrase “the facts giving rise to the cause of action” in the former as compared with “the incident giving rise to the action” in the latter. We find it difficult to make this esoteric distinction and thus turn to precedent for guidance.

The case of University of Miami v. Bogorff, 583 So. 2d 1000 (Fla.1991), involved both medical malpractice defendants and a products liability defendant. The supreme court applied the Nardone formula to the medical malpractice claims, pointing out that knowledge of medical treatment and knowledge of an injury “ [although they did not know if medical negligence caused that condition, ... was sufficient for their cause of action to accrue, thereby commencing the statutory limitation period....” Id. at 1002.

The court, after disposing of collateral issues, then addressed the products liability claim. In this phase of the opinion, the court again cited Nardone, but only for the limited purpose of defining the extent of knowledge necessary to commence the limitation period. The factors mentioned by the court as apparently bearing on the issue of when the facts giving rise to the cause of action were or should have been discovered, include: (1) awareness of the existence of a serious physical injury; (2) knowledge that the particular drug had been administered; and (3) constructive knowledge of medical opinion in the hospital records that the drug may have contributed to the injury. The court speaks, too, of knowledge by the plaintiffs of the possible invasion of their legal rights.

We interpret the test which the supreme court has fashioned here as having two essential ingredients: an injury distinct in some way from conditions naturally to be expected from the plaintiffs condition, and (as opposed to or in the medical malpractice context) exposure to the product in question. Use of the conjunction “and” in this equation necessarily implies that the connection must be to some extent causal. There could be no “invasion of their legal rights” unless this were so. The court then states that “because knowledge of the contents of accessible medical records is imputed, the Bogorffs had constructive knowledge of medical opinion that the drug may have contributed to the injury in 1977.” Id. at 1004. This speaks clearly to causation: that ingestion of the drug was a contributing cause of plaintiffs injuries. Cf. Goodlet v. Steckler, 586 So. 2d 74 (Fla. 2d DCA 1991) (Lehan, J., concurring) (notice must be such that it connects defendant to the injury in some manner).

Harry Babush certainly was aware that he was being exposed to the product, clofibrate, during the years 1971 through 1977. When he was diagnosed as having fatal liver damage in 1977, two months after he stopped taking the drug, he became aware that he had been injured and that the injury had no relationship to the medical problem for which he was being treated, that is, a heightened cholesterol level. The question is whether Babush knew or should have known of the connection between ingestion of the drug clofib-rate and the onslaught of primary biliary cirrhosis.

The record shows that Babush was told in 1986 that the drug he had taken was toxic to the liver. There is nothing in the record to show that Babush knew, earlier or from any other source, or should have known (by virtue of his own medical records) that there was a causal connection between the drug and the injury. If there were nothing else in the record we would hold that, as a matter of law, this action is not barred by the statute of limitations.

There is, however, one further factor to consider. Babush alleges in the complaint that a World Health Organization study disclosed in 1973 that liver damage and dermatitis were identified as problems resulting from ingestion of clofibrate. This allegation bears on the question whether Babush knew or should have known at a time more than two years before the filing of the complaint that his injury was caused by the drug.

On the state of the record in this case we cannot say as a matter of law that the existence of this study was enough to put Babush on notice that his legal rights had been violated.

Thus we conclude that there remains a genuine, material issue of fact which should not have been resolved by summary judgment.

Accordingly, we reverse and remand for further proceedings in accordance with this opinion.

REVERSED AND REMANDED.

GUNTHER and FARMER, JJ., concur.