OPINION

The plaintiff in this case, Mrs. Betty Odom, became sterile as a result of two ectopic pregnancies, which she claims were[*1002] caused by an intrauterine device manufactured by G.D. Searle & Company. She has sued Searle on a failure-to-warn theory. The district court granted Searle’s motion for summary judgment because “plaintiff has presented nothing to establish a causative link between the alleged failure to warn and the decision by the plaintiffs physician to prescribe the product.” We agree that plaintiff has failed to show that her doctor would not have prescribed.the IUD if Searle had phrased its warning differently. We thus affirm the judgment of the district court.

I.

G.D. Searle & Co. first began clinical testing of its Copper-7 intrauterine device (the “Cu-7 IUD”) in December of 1970. After observing more than 16,000 women who used the device, Searle submitted a New Drug Application for the Cu-7 IUD to the Food and Drug Administration. The FDA reviewed this data, and on February 25, 1974, approved the Cu-7 for marketing.

Searle sold the Cu-7 IUD with a package insert warning that the IUD could sometimes cause pelvic inflammatory disease (“PID”) and ectopic pregnancy:

Warnings

2. Ectopic pregnancy: (a) A pregnancy which occurs with an IUD in situ is more likely to be ectopic than a pregnancy occurring without an IUD....

3. Pelvic infection: An increased risk of pelvic inflammatory disease associated with the use of IUDs has been reported.... The decision to use an IUD in a particular case must be made by the physician and patient with the consideration of a possible deleterious effect on future fertility.

Adverse Reactions

Pelvic infection including salpingitis with tubal damage or occlusion has been reported. This may result in future infertility. ...

An ectopic pregnancy occurs when the ovum stays in one of the fallopian tubes to develop, instead of moving to the uterus. This necessitates a salpingectomy — the surgical removal of the fallopian tube. It is not disputed that Searle’s package insert complied with the FDA’s labeling requirements in 21 C.F.R. § 310.502(b).

In 1979, Mrs. Odom, a resident of South Carolina, gave birth to her second child. Later that year, Odom’s obstetrician-gynecologist, Dr. Bernard Credle, prescribed for her a Cu-7 IUD to forestall any further pregnancies. In 1986, Odom wished to have more children, and had the IUD removed. Soon thereafter she became pregnant, but suffered an ectopic pregnancy in her left fallopian tube and had to have a salpingectomy. She became pregnant again in 1988, but this time suffered an ectopic pregnancy in her right fallopian tube. Once again, she had to have a salpin-gectomy, leaving her completely sterile.

On March 15, 1990, Mrs. Odom sued Searle, claiming that the Cu-7 IUD had caused her ectopic pregnancies, and that the package insert for the Cu-7 IUD did not adequately warn of this risk. During discovery, she submitted an affidavit from a Dr. Robert Laird, who posited that her ectopic pregnancies had been caused by PID, which had been caused by the Cu-7. Dr. Laird also asserted that Searle’s warning was inadequate; in -his opinion, a proper warning would have advised physicians not to prescribe the IUD to women who might later want children. In a separate deposition, Dr. Laird estimated the risk of contracting PID from the Cu-7 to be about 1.84 percent.

Searle deposed Dr. Credle, who testified that at the time he prescribed the Cu-7 IUD to Mrs. Odom, he already knew of the risk of PID through his own experience and training. He estimated the risk at “less than five percent.” Notwithstanding what had happened to Mrs. Odom, Dr. Cre-dle stated that he had always believed' and still believed that “the Cu-7 is the best IUD that’s ever been on the market,” and that he would still prescribe it today.

[*1003] Searle subsequently moved for summary judgment. The district court granted Searle’s motion, on the sole ground that Mrs. Odom had provided insufficient evidence of causation. Specifically, the court found that she had failed to provide any evidence that Dr. Credle would not have prescribed the Cu-7 IUD if he had been given the warning proposed by Dr. Laird. Odom now appeals this judgment.

II.

It is plain that Mrs. Odom’s claim is governed by the “learned intermediary” doctrine. See Brooks v. Medtronic, Inc., 750 F.2d 1227, 1230-32 (4th Cir.1984). Under this doctrine, the manufacturer’s duty to warn extends only to the prescribing physician, who then assumes responsibility for advising the individual patient of risks associated with the drug or device. The sole issue in this case, therefore, is whether an adequate warning to Odom’s doctor about the risk of sterility would have deterred him from prescribing the IUD. Mrs. Odom does not dispute that under the “learned intermediary” doctrine, the manufacturer cannot be said to have caused the injury if the doctor already knew of the medical risk. Stanback v. Parke, Davis and Co., 657 F.2d 642, 645-46 (4th Cir.1981). What she does dispute, however, is the district court’s summary judgment finding that Dr. Credle was independently aware of the risk of PID.

Even viewing the facts most favorably to Mrs. Odom, we cannot escape the district court’s conclusion that Dr. Credle would have prescribed the Cu-7 IUD no matter how carefully Searle refined the phrasing of its warning. Dr. Credle testified at length about his independent knowledge of the risk of PID; indeed, his own estimate of the risk actually exceeded that of Mrs. Odom’s expert, Dr. Laird. By 1979, according to Dr. Credle, the risk of PID was “common knowledge, not only [in] the medical literature, but [also in] the lay literature.” Dr. Credle further testified that it was common medical school knowledge that PID could lead to ectopic pregnancy. That knowledge did not alter Dr. .Credle’s stated judgment that in 1979, the Cu-7 IUD was an appropriate contraceptive for women who wanted to have more children in the future, and that it remains so today.

Mrs. Odom presents nothing to controvert this testimony. Instead, she argues that we should simply presume causation in the event she is able to prove that Searle’s warning was inadequate. There is no such presumption under South Carolina law, and we are unwilling to create one here. See Thomas v. Hoffman-LaRoche, Inc., 949 F.2d 806, 812-14 (5th Cir.1992) (declining to create such a presumption under Mississippi law). In Thomas, the Fifth Circuit distinguished between preventable risk warnings, which are commonly associated with mechanical products, and unavoidable risk warnings, which are often associated with prescription' drugs or devices like the IUD. In the former category, a warning, if heeded, would diminish or eliminate the risk. In the latter context, however, an adequate warning “means only that the learned intermediary would have incorporated the ‘additional’ risk into his decisional calculus. The burden remains on the plaintiff to demonstrate that the additional non-disclosed risk was sufficiently high that it would have changed the treating physician’s decision to prescribe the product for the plaintiff.” Id. at 814.

This is the burden Mrs. Odom failed to carry. Mrs. Odom cannot discharge this burden by arguing that a more drastic warning would have resulted in a different course of treatment. According to her expert Dr. Laird, the proper warning would have stated flatly that the Cu-7 IUD was an inappropriate contraceptive device for women who might later want children. This apocalyptic pronouncement, for all intents and purposes, would have removed the Cu-7 from the market, notwithstanding the FDA’s judgment that the device was safe. Indeed, the warning suggested by Odom would have deprived countless patients of the product’s beneficial use for the eighteen years that Searle has marketed the IUD. These are extraordinary costs, not justified by the evidence in this case. Any alternative warning that Mrs.[*1004] Odom proposes must bear some reasonable relation to the 1.84% risk that her own expert confirmed, and Dr. Credle’s testimony makes clear that such a warning would not have changed his decision to prescribe the IUD. *

III.

The judgment of the district court is therefore

AFFIRMED.