The district judge dismissed this diversity personal-injury suit on the pleadings, on the ground that it was preempted by an investigational-device exemption issued by the Food and Drug Administration under the Medical Device Amendments of 1976 to the Food, Drug and Cosmetic Act of 1938, 21 U.S.C. §§ 301 et seq. 756 F.Supp. 370 (N.D.Ill.1991). He thus disagreed with Mitchell v. Iolab Corp., 700 F.Supp. 877 (E.D.La.1988), the only other case concerning the preemptive force of an exemption for an investigational device — which makes ours a case of first impression at the appellate level.

The Amendments gave the FDA comprehensive regulatory authority over medical devices for the first time. H.R.Rep. No. 853, 94th Cong., 2d Sess. 1, 6-13 (1976). Among other .things, the Amendments authorize the agency to promulgate implementing regulations, 21 U.S.C. § 371(a), and forbid states to subject a medical device to “any requirement (1) which is different from, or in addition to, any requirement applicable under the [Amendments] to the device, and (2) which relates to the safety and effectiveness of the device or to any other matter included in a requirement applicable to the device under” the Amendments. § 360k(a). In regulations the validity of which neither party questions, the FDA has declared that preemption, though equally applicable whether the state is acting “by statute, ordinance, regulation, or court decision,” is limited to cases in which the FDA “has established specific counterpart regulations [to the challenged state requirements] or there are other specific requirements applicable to a particular device under” the federal act. 21 C.F.R. §§ 808.1(b), (d).

In the year following the enactment of the Medical Device Amendments, the FDA issued “Investigational Device Exemption Regulations” for intraocular lenses. 21 C.F.R. pt. 813. Such exemptions, which the Amendments expressly authorize the FDA to grant upon application, 21 U.S.C. § 360j(g), .are intended to encourage innovation by exempting promising experimen[*1332] tal devices from the usual requirement of establishing the safety and efficacy of a medical device before it can be sold. Cf. H.R.Rep. No. 853, supra, at 12, 42. The regulatory exemption is not total. (See United States v. An Article of Device, 731 F.2d 1253, 1258 (7th Cir.1984), and with specific reference to the exemption for in-traocular lenses Public Citizen Health Research Group v. FDA, 704 F.2d 1280, 1283 (D.C.Cir.1983), and Surgidev Corp. v. Eye Technology, Inc., 648 F.Supp. 661, 684 (D.Minn.1986), aff’d, 828 F.2d 452 (8th Cir.1987).) A detailed application is required, describing the device and setting forth a plan for studying its use in human subjects (“clinical investigation”) during the experimental period. 21 C.F.R. § 813.20. The application must be reviewed by an institutional review committee as well as by the FDA before it can be approved. §§ 813.20, 813.30. After approval, the committee has a duty to monitor the clinical investigation. § 813.65. Congress expected the FDA to grant an investigational-device exemption for intraocular lenses; we know this because the Medical Device Amendments contain a specific provision intended to require the FDA to make intraocular lenses “reasonably available to physicians meeting appropriate qualifications” if an investigational exemption is granted. 21 U.S.C. § 360j(i )(3)(D)(iii); H.R.Conf.Rep. No. 1090, 94th Cong., 2d Sess. 63 (1976); American Society of Cataract & Refractive Surgery v. Sullivan, 772 F.Supp. 666, 673 (D.D.C.1991).

Intraocular lenses are plastic lenses implanted in the eyes of persons who have undergone surgery for the removal of cataracts. Richard B.S. Packard & Fiona C. Kinnear, Manual of Cataract and Intrao-cular Lens Surgery (1991). The removal of a cataract destroys the (natural) lens on which the cataract has formed, and restoration of focused vision requires either special cataract glasses, a contact lens, or an intraocular lens (behind the cornea). Alfred Mamelok, “Cataract,” in 3 Roscoe N. Gray & Louise J. Gordy, Attorneys’ Textbook of Medicine till 56.45, at pp. 56-50 to 56-52 (3d ed. 1991). Now standard treatment for cataract patients, intraocular lenses were experimental back in the 1970s when the investigational-device exemption for such lenses was issued, and for some years afterward.

Optical Radiation Corporation, the defendant in this case, manufactured an intrao-cular lens called “Stableflex.” This was an anterior-chamber intraocular lens (that is, it was installed in the anterior chamber of the eye, which is a space between the cornea and the iris), a type of lens manufactured by several other companies as well. All of these lenses, as well as intraocular lenses of different design, received investigational-device exemptions. In 1984, Slater, the plaintiff in this case, had a cataract removed and a Stableflex implanted in the eye. Before the operation he signed a form in which he consented to participate in the clinical investigation of this experimental device. After the operation, according to the allegations of the complaint, the vision in Slater’s left eye deteriorated, he experienced continuous pain, the eye became infected, and eventually the implant had to be removed, leaving him with permanent damage to the eye (beyond the damage done by the cataract itself). The surgeon who performed the removal found the lens “cocooned” in the eye — which he thought due to the fact that the lens had an awkward design — and therefore difficult to remove. This model of the Stableflex was later withdrawn from the market. David J. Apple et al., Intraocular Lenses: Evolution, Designs, Complications, and Pathology 88 (1989).

The suit is for the injury to Slater’s eye allegedly caused by the implantation and removal of the lens. The complaint alleges inadequate clinical testing, defective design of the intraocular lens, and failure to warn the public of the lens’s dangers. The district judge dismissed the suit on the ground that the plaintiff was seeking to impose in the name of Illinois tort law requirements relating to safety or effectiveness that were different from the requirements imposed by the FDA’s regulations governing the experimental use of intraocular lenses.

The plaintiff’s counsel wisely conceded at argument that insofar as the complaint[*1333] might be construed to allege that Illinois tort law required the defendant to do more or different clinical testing of Stableflex than required by the FDA’s regulations governing the investigational exemption for intraocular lenses, the suit is indeed preempted; for it would then be unquestionably a matter of imposing in the name of state law a different requirement relating to safety and effectiveness from that imposed by federal law. If the FDA in a valid regulation under the Medical Devices Amendments requires X, and X relates to safety or effectiveness, a state cannot, whether through statute, regulation, or common law decision, require non-X, or X ± Y.

But the plaintiffs real gripe is not inadequate testing; it is defective design. Not only is the design of the Stableflex said to be defective, but there is a suggestion that the entire concept of an anterior-chamber lens is unsound and we are told that lawsuits are pending against other manufacturers of such lenses as well. The FDA’s regulations impose no requirements concerning the design of intraocular lenses. Therefore, the plaintiff argues, there is no preemption of state products liability law relating to design defects. Moore v. Kimberly-Clark Corp., 867 F.2d 243 (5th Cir.1989). He adds that if he has no damages remedy under state law, he has no damages remedy, period, as there is no federal damages remedy that he might invoke.

The argument assumes that the only federal requirement that might relate to the safety or effectiveness of the design of a medical device would be an actual specification of that design. The preemption provision does not say this, and so cramped an interpretation would cripple the exemption for investigational devices. The FDA can hardly be expected to specify the safe and effective design of a device when it is still experimental. If there were a known safe and effective design, the device would no longer be experimental. The point of the experiment is to find out whether it is safe and effective. Public Citizen Health Research Group v. FDA, supra, 704 F.2d at 1283. In the experimental phase the appropriate regulations of safety and effectiveness are procedural rather than substantive ones. They do not specify the safe and effective design; they specify the procedures for determining whether the experimental design is safe and effective. These are requirements relating to safety and effectiveness and they can therefore have preemptive effect.

Of course the FDA will not grant an investigational-device exemption unless it believes that the device has sufficient promise of being proved safe and effective to justify the risk of its being used on human beings. Although that belief is different from a certification that the design of the device is safe and effective, it is a certification that the design is sufficiently safe and effective to allow experimental use on human beings. The theory underlying the complaint is that the design of the Stableflex was not sufficiently safe and effective to allow it to be used on human beings. This theory sets up a direct collision with federal policy. The FDA decided, whether rightly or wrongly, but pursuant to regulations the validity of which the plaintiff does not question, that the Stable-flex could be sold, subject only to requirements, procedural in character and, so far as appears, fully complied with, designed to assure that this experimental distribution was in fact a worthwhile experiment. The plaintiff wishes in the name of state tort law to impose additional requirements— namely that the Stableflex have had design characteristics that it lacked — and this en-grafting of additional requirements relating to safety or effectiveness is forbidden by the preemption provision in the Medical Devices Amendments.

It would be a mistake to conclude that preemption in these circumstances leaves the consuming public remediless, at least if we have concern for economic substance rather than legal formality and do not suppose that the only “remedies” (preventives, protections, correctives) are those that the law provides. The consent form that Mr. Slater signed informed him that he had a choice of treatments, one of which was experimental; he chose the experimental. The experimental procedure was riskier be[*1334] cause, by definition, there was limited experience with it. Provided the risks, including any loss of tort remedies, were adequately explained to Slater, he cannot complain that the risks materialized. We add that if experimental procedures are subject to hindsight evaluation by juries, so that failed experiments threaten to impose enormous tort liability on the experimenter, there will be fewer experimental treatments, and patients will suffer.

The scope of the preemption, moreover, is limited. It is limited to efforts by states to impose sanctions for compliance with federal regulations relating to the safety or efficacy of the experimental lenses. It does not affect cases charging negligence in the implantation or removal of a lens, or complaining of contamination of the lens by bacteria or fungi or of failure to obtain the patient’s informed consent to the procedure. Kus v. Sherman Hospital, 204 Ill.App.3d 66, 149 Ill.Dec. 103, 561 N.E.2d 381 (1990) (implantation of lens of which the FDA had withdrawn its approval); Green v. Wedowee Hospital, 584 So.2d 1309 (Ala.1991) (implantation in wrong eye); Hudson v. Taleff, 546 So.2d 359 (Miss.1989) (unspecified negligence in implantation); Kershaw v. Reichert, 445 N.W.2d 16 (N.D.1989) (lack of informed consent); United States v. Torigian Laboratories, Inc., 577 F.Supp. 1514 (E.D.N.Y.), affirmed without opinion, 751 F.2d 373 (2d Cir.1984) (unsterilized lens); In re Luminex International, Inc. Products Liability Litigation, 434 F.Supp. 668 (Judicial Panel on Multidistrict Litigation 1977) (per curiam) (lens contaminated by fungus). It is not the sort of blanket preemption that we find in ERISA, which supersedes “any and all State laws insofar as they may now or hereafter relate to any employee benefit plan.” 29 U.S.C. § 1144(a). See Pilot Life Ins. Co. v. Dedeaux, 481 U.S. 41, 107 S.Ct. 1549, 95 L.Ed.2d 39 (1987); Pohl v. National Benefits Consultants, Inc., 956 F.2d 126, (7th Cir.1992). The limited scope of the preemption explains the prohibition incorporated by reference in the investigational-device exemption for intraocular lenses, 21 C.F.R. § 813.1 et seq., against the inclusion of “any exculpatory language” in consent forms. § 50.20. The prohibition preserves the patient’s common law rights outside of the limited scope of the preemption provision. It does not, as the district court held in the Mitchell case, 700 F.Supp. at 878-79, repeal the preemption provision itself.

Even within the scope of the preemption there is an important qualification. The risks of the experimental procedure must bé adequately explained to the patient. 21 U.S.C. § 360j(g)(3)(D); 21 C.F.R. § 813.-66(a)(6). The tort law of medical battery, which is violated when an operation is performed without the patient’s informed consent, is not preempted. Slater’s complaint' alleges that Optical Radiation Corporation failed to warn the public of the dangers of the Stableflex lens. We are not sure what he is driving at. The lens was not marketed to the public. It was an experimental device provided to patients together with such warnings as might be necessary to elicit informed consent to the implantation of it. We do not understand Slater to be arguing that he did not give his informed consent to the implantation. The consent form that he signed adequately discloses the medical risks of the procedure. True, it does not indicate that by signing the form he will be giving up any legal rights. It says, “I understand that in the event of physical injury resulting from this clinical study, there is no compensation available from Hinsdale Sanitarium and Hospital [where the cataract operation and lens implantation were performed] for such injury.” This could be taken to imply that of course the signator is not waiving any right to seek compensation from the manufacturer of the lens for having designed or manufactured it defectively. But Slater has made nothing of this, either in our court or in the district court. His complaint, variously expressed, is that the Sta-bleflex was too dangerous even for experimental use. That complaint is barred by the preemption provision of the Medical Devices Amendment.

Affirmed.