In this diversity case, Santiago Crisosto-mo, M.D., and his wife, Flor de Liza Crisos-tomo, M.D., brought an action against Charles 0. Stanley, M.D., the Stanley Medical Group, Ltd. (Stanley Group) and the Burroughs Wellcome Company (BWC) for injuries sustained when Dr. Santiago Cri-sostomo took the prescription drug Zylo-prim.1 The Crisostomos’ claims against Dr. Stanley and the Stanley Group were based on a theory of medical malpractice resulting from Dr. Stanley's treatment of Dr. Crisostomo. The Crisostomos’ claims against BWC, the manufacturer of Zylo-prim, were based on a theory of strict liability. At the close of the plaintiffs’ ease-in-chief, the district court struck the Crisostomos’ specific allegations of malpractice. It then granted a directed verdict for all the defendants. We reverse as to Dr. Stanley and the Stanley Group, but affirm as to BWC.
I
Background
Since this case was terminated at the end of the plaintiffs’ case, our rendition of the facts is necessarily limited to the record developed up to that point. In December 1976, after playing tennis, Dr. Santiago Crisostomo complained of pain in his big toe. After first consulting an orthopedist, Dr. Crisostomo met with his internist, Dr. Stanley. Dr. Stanley diagnosed an acute attack of gout, for which he prescribed the drug Indocin. A few days later, Dr. Stanley and Dr. Crisostomo met by chance in the hospital where they both worked.2 Dr. *1148Stanley briefly examined Dr. Crisostomo’s toe and wrote out a prescription for Zylo-prim and Colchicine to treat the gout. Although he mentioned that Colchicine can cause stomach upset, Dr. Stanley did not warn of any side effects with respect to Zyloprim.
Dr. Crisostomo began taking the new medication on January 1, 1977. On January 6, he noticed mouth sores and later experienced chills. Dr. Crisostomo testified that, when he notified Dr. Stanley of these symptoms by telephone, Dr. Stanley advised him to “sit tight.” Dr. Crisostomo stopped taking the Colchicine at this point. The next day, his chills and mouth sores continued to bother him and Dr. Crisosto-mo worried that he might have contracted diphtheria, a disease he often had seen in his native country, the Phillipines. That evening, in order to counteract the possible onset of diphtheria, Dr. Crisostomo took some penicillin V which had been prescribed for his son on a prior occasion.
When Dr. Crisostomo awoke on January 8, his mouth was very sore. In addition, his eyes were in such pain that he could hardly open them. Dr. Stanley agreed to meet Dr. Crisostomo in the hospital’s emergency room. An examination of Dr. Cri-sostomo’s upper chest revealed crusty skin lesions, a condition Dr. Stanley diagnosed as Stevens-Johnson Syndrome.3 At that point, Dr. Stanley discontinued the Zylo-prim treatment and ordered that penicillin be administered along with corticosteroids. At about the same time, Dr. Carney, a dermatologist, began treating Dr. Crisosto-mo.
During Dr. Crisostomo’s hospitalization, the skin lesions spread over his entire body. His vision also diminished, even though Dr. Crisostomo was under an ophthalmologist’s care.4 By the time of his release from the hospital on January 28, 1977, the corneas of Dr. Crisostomo’s eyes had scarred and his tear ducts had been damaged. Dr. Cri-sostomo’s eyesight is now 20:200 in one eye and 20:100 in the other. He complains that he is unable to tolerate bright light and that his eyes are always painful. He further contends that his eye problems have prevented him from returning to his profession and that his wife has had to give up her psychiatric practice to care for him in his depressed state.
The Crisostomos filed this suit on January 2, 1979. In their third amended complaint, they alleged that Dr. Crisostomo suffered Stevens-Johnson Syndrome as a result of his taking the drug Zyloprim. Counts I and II sought damages on behalf of Dr. Crisostomo for his injuries, and on behalf of his wife for loss of consortium, against Dr. Stanley and the Stanley Group5 under a theory of medical malpractice. The complaint alleged that Dr. Stanley had failed to perform certain diagnostic tests and that the dosage for Zyloprim was excessive. It also charged that Dr. Stanley had failed to provide any warnings with respect to the drug’s use and its side effects, and that his post-diagnostic care was deficient.
In support of these claims, the Crisosto-mos offered the testimony of Mark Jarrett, M.D., an internist. Dr. Jarrett stated that Zyloprim should be prescribed for only ten *1149percent of those patients who have a predisposition to gout. To identify these patients, Dr. Jarrett recommended that a doctor perform two diagnostic tests, a synovial fluid test and a 24-hour urine test. Dr. Stanley did not perform either of these two tests. On cross-examination, Dr. Jarrett agreed that the average physician does not normally conduct a synovial fluid test. He also conceded that the 24-hour urine test was not diagnostic of gout.
Dr. Jarrett testified that Zyloprim should be given only to those patients who exhibit chronic symptoms of gout, not merely one acute attack. He also found the dosage prescribed by Dr. Stanley of 300 mg./day to be excessive. He recommended instead that an initial dosage should be no greater than 100 mg./day. Dr. Jarrett admitted, however, that he had no way of knowing whether the dosage was a contributing factor in producing the Stevens-Johnson Syndrome. Nonetheless, Dr. Jarrett stated that a doctor initially should warn of a prescription drug’s side effects and that, once he learns of a patient’s adverse reaction, the doctor should advise the patient to stop taking the medication.
In Counts III and IV, the Crisostomos alleged that BWC, as the manufacturer of the Zyloprim, was strictly liable for Dr. Crisostomo’s injuries and his wife’s loss of consortium. The Crisostomos alleged that BWC knew or should have known that penicillin, when combined with Zyloprim, increases the occurrence of, or aggravates the symptoms of, Stevens-Johnson Syndrome. In this regard, the Crisostomos offered the testimony of Dr. James O’Donnell.6 Dr. O’Donnell testified that the ingestion of Zyloprim alone had caused Dr. Crisostomo to contract Stevens-Johnson Syndrome. However, Dr. O’Donnell indicated that, based on his knowledge of clinical cases and research reports, the combined use of ampicillin, a type of penicillin, and Zyloprim increases the likelihood of a drug-induced rash.7 He also stated that the condition of a patient who already had contracted Stevens-Johnson Syndrome could be aggravated by the ingestion of penicillin. According to Dr. O’Donnell, all penicillins work in approximately the same manner and the effects reported from the interaction of ampicillin and Zyloprim would probably extend to other penicillins. He conceded, however, that the reports only discussed the interaction of ampicillin and Zyloprim, and that no extrapolation from the data could be made to other peni-cillins and Zyloprim.
Dr. O’Donnell agreed that the Zyloprim label had warned that Stevens-Johnson Syndrome could result from use of the drug. However, he testified that, in 1976, the label failed to warn of either the increased incidence of rash or the aggravation of Stevens-Johnson Syndrome when the drug is used in combination with peni-cillins.8 On cross-examination, Dr. O’Donnell admitted that a rash is not the same as Stevens-Johnson Syndrome, and that the package insert for Zyloprim advised that its use be discontinued at the first appearance of skin rash or any other adverse reaction.
At the close of the plaintiffs’ case, the district court, ruling verbally from the bench, dismissed the specific pleadings of malpractice against Dr. Stanley for insufficient evidence. As a result, only the “general” allegation of negligence remained.9 The plaintiffs then submitted a fourth *1150amended complaint (which included specific allegations), under Fed.R.Civ.P. 15(b). Concluding that an amendment would be prejudicial to the defendants, the district court rejected the plaintiffs’ motion to amend their complaint. The court then granted a directed verdict for all of the defendants.
II
Analysis
In a diversity action, this court must apply the state standard of review to the district court’s decision to grant or deny a directed verdict. Cincinnati Ins. Co. v. City of Taylorville, 818 F.2d 1345, 1348 (7th Cir.1987). Under Illinois law, a directed verdict for a defendant should not be entered unless all of the evidence, viewed in the light most favorable to the plaintiff, so overwhelmingly favors the defendant that no contrary verdict could ever stand. Pedrick v. Peoria & E.R.R., 37 Ill.2d 494, 510, 229 N.E.2d 504, 513-14 (1967). “This formulation does not require a complete absence of evidence supporting the side against whom the verdict is directed; however, there must be a substantial factual dispute before a jury trial is required.” Cincinnati Ins. Co., 818 F.2d at 1348 (citing Pedrick, 37 Ill.2d at 504-05, 229 N.E.2d at 510).
A. Medical Malpractice
Under Illinois law, mere “[pjroof that a good result was not achieved is not proof of a breach of a physician’s duty.” Spike v. Sellett, 102 Ill.App.3d 270, 273, 50 Ill. Dec. 565, 568, 430 N.E.2d 597, 600 (1981). To show medical malpractice, the plaintiff must establish that the defendant-doctor’s conduct fell below the requisite standard of care and that this factor caused injury to the plaintiff. Borowksi v. Von Solbrig, 60 Ill.2d 418, 423, 328 N.E.2d 301, 304-05 (1975). Unless the negligence in question is grossly apparent to the lay observer, the plaintiff must introduce expert testimony to establish the requisite level of medical care. Mayer v. Baisier, 147 Ill.App.3d 150, 155, 100 Ill.Dec. 649, 652, 497 N.E.2d 827, 830 (1986); Boey v. Quaas, 139 Ill.App.3d 1066, 1071, 94 Ill.Dec. 345, 347, 487 N.E.2d 1222, 1224 (1986). If the plaintiff fails to establish a required element of the prima facie case, then a directed verdict for the defendant is proper. Mayer, 147 Ill.App.3d at 155, 497 N.E.2d at 830. We now turn to an examination of the plaintiffs’ specific claims to determine whether the district court erred in its decisions.
The plaintiffs’ third amended complaint made the following allegations with respect to Dr. Stanley’s alleged medical malpractice:
(a) Defendant failed to use that degree of care and skill in the diagnosis, care and treatment of the plaintiff that a reasonably well qualified physician practicing in the medical specialty of Internal Medicine would have used in the same or similar circumstances.
(b) Defendant negligently failed to make, or cause to have made, tests to determine whether the plaintiff was suffering from gout, including but not limited to:
(1) Synovial fluid examination using polarized-light microscopy.
(2) 24 Hour Urine — uric acid excretion test.
(c) Defendant negligently started plaintiff out on a too-high dosage of allo-purinol (300 mgs/day) without first ascertaining whether the allopurinol could be tolerated by plaintiff in lesser amounts.
(d) Defendant negligently prescribed said allopurinol for plaintiff’s use without advising and warning the plaintiff of any of the following:
(1) that allopurinol might cause Stevens-Johnson Syndrome of a potentially severe and devastating degree.
(2) that allopurinol should be discontinued at the first sign of a rash or other adverse reaction.
(3) that penicillin should not be taken along with allopurinol since the use of penicillin might exacerbate the adverse reactions caused by allopurinol.
(e) Defendant negligently failed to provide appropriate care of plaintiff after *1151the appearance of the Stevens-Johnson Syndrome in one or more of the following respects:
(1) He continued plaintiff on penicillin when its use was contra-indicated.
(2) He did not obtain medical consults with specialists knowledgeable in the care and treatment of drug-induced Stevens-Johnson Syndrome.
(3) He did not obtain medical consults with ophthalmologists knowledgeable in the care and treatment of injuries to the eyes caused by drug-induced Stevens-Johnson Syndrome.
(4) He failed to use appropriate steroid therapy after having diagnosed the plaintiffs condition as drug-induced Stevens-Johnson Syndrome.
R.82.
As a preliminary matter, we note that, when the plaintiffs’ counsel argued against the defendants’ directed verdict motion, with one exception that previously had been waived,10 he agreed to drop the allegations regarding inadequate post-diagnostic care (allegation (e)), the failure to warn of Stevens-Johnson Syndrome (allegation (d)(1)) and the failure to warn of using penicillin in combination with Zyloprim (allegation (d)(3)). Thus, we review only allegations (b), (c), and (d)(2) to determine whether the plaintiffs provided sufficient evidence to support these claims.
We find that the district court’s decision to strike the specific allegations was, for the most part, a correct one. For example, we find no proof that the failure to conduct certain diagnostic tests (allegation (b)) led to Dr. Crisostomo’s injury. Similarly, even assuming that the prescribed amount of the drug was excessive, allegation (c) fails because there is no indication of a causal link between the dosage and the appearance of Stevens-Johnson Syndrome.11
On the other hand, we believe that the plaintiffs have produced sufficient evidence to support allegation (d)(2). Allegation (d)(2) states that Dr. Stanley should have warned Dr. Crisostomo to stop taking the drug at the appearance of a rash or other adverse reaction. Dr. Jarrett testified as follows:
[Plaintiffs’ attorney]: What is the purpose of discussing with the — of the physician discussing with the patient the possibility of these major adverse reactions? [Dr. Jarrett]: There are two purposes. One, it’s the patients — the patient themselves (sic), the one who’s going to be taking the medication, should be aware of what they’re taking and the potential risk to them and, number two, it protects the patient somewhat because if they develop any of these adverse reactions they’ll know it’s at least something they should alert their doctor about.
Tr. at 651-52.
Notwithstanding this evidence, Dr. Stanley insists that the plaintiffs have failed to establish a prima facie case. Dr. Stanley contends that there is no causal relation between his alleged failure to warn and the resulting injury. Indeed, he maintains that a warning at the time he prescribed the drug would have been of little use because *1152Dr. Crisostomo did not discover his rash until he entered the hospital. We cannot accept this argument. At the least, a warning might have alerted Dr. Crisostomo to examine himself more closely for an adverse reaction. If so, Dr. Crisostomo might have stopped taking the Zyloprim earlier.
Furthermore, Dr. Jarrett stated that a physician should tell the patient immediately to discontinue use of the drug once he learns of any adverse reactions. Dr. Cri-sostomo testified that, when he telephoned Dr. Stanley to discuss his chills and mouth sores, Dr. Stanley made no mention of the need to stop using the Zyloprim.12 Although Dr. Stanley argues that no such call was made, we need not resolve this factual dispute at this juncture.13
Thus, the issue of Dr. Stanley’s negligence remains unresolved. We therefore believe that the district court erred in striking allegation (d)(2).14 Dr. Crisostomo produced sufficient evidence to support his allegation that Dr. Stanley had failed to advise him to discontinue the drug should an adverse reaction occur and when an adverse reaction actually occurred. We believe that the court erred in striking this allegation. Accordingly, we reverse the court’s directed verdict for Dr. Stanley on this issue.15
B. Strict Liability
The Crisostomos alleged that BWC knew or should have known that the combination of penicillin V and Zyloprim would increase the risk of, or exacerbate the appearance of Stevens-Johnson Syndrome. They also alleged that the warnings given to doctors with respect to the combined use of these drugs were inadequate.
Under Illinois law, a drug manufacturer’s strict liability turns on whether it sufficiently warned of a drug’s dangerous propensities. See Lawson v. G.D. Searle & Co., 64 Ill.2d 543, 551, 1 Ill.Dec. 497, 501, 356 N.E.2d 779, 783 (1976). A drug is unreasonably dangerous “not because of some defect inherent in the product itself, but because of the absence of an adequate warning accompanying the product.” Woodill v. Parke Davis & Co., 79 Ill.2d 26, 30, 37 Ill.Dec. 304, 306, 402 N.E.2d 194, 196 (1980).16 However, the drug manufacturer is only liable for failure to warn if it had knowledge or reason to know of the danger that caused the injury. Id. at 35, 402 N.E.2d at 198.17
*1153Dr. Crisostomo introduced the testimony of Dr. O’Donnell against BWC.18 It is clear that Dr. O’Donnell’s testimony failed to satisfy the plaintiffs’ burden of proof. Dr. O’Donnell admitted that the Zyloprim alone probably caused the Stevens-Johnson Syndrome, and conceded that its label had warned of this possibility. Furthermore, the medical reports on which his opinion was based suggest that the use of ampicillin and Zyloprim may lead to an increased incidence of rash. Aside from the chemical distinctions between ampicillin and penicillin Y (the drug Dr. Crisostomo took),19 the reports point to an increased risk of rash, not Stevens-Johnson Syndrome. Nor is there any evidence to support Dr. O’Donnell’s view that penicillin can aggravate the symptoms of Stevens-Johnson Syndrome. Therefore, the district court correctly granted a directed verdict for BWC.
Conclusion
We affirm the judgment for BWC. We also agree with the district court that Dr. Stanley cannot be found liable on the majority of the specific malpractice allegations. However, we hold that the plaintiffs established a prima facie case against Dr. Stanley for failing to advise him to stop taking the drug if and when an adverse reaction appeared. A directed verdict for Dr. Stanley should not have been granted with respect to this matter. Accordingly, we reverse and remand to the district court for a new trial on this issue. BWC shall recover its costs in this court. In all other respects, the parties shall bear their own costs of this appeal.
AFFIRMED IN PART; REVERSED AND REMANDED IN PART.
. The drug’s trade name is Zyloprim; it is sometimes referred to by the chemical term allopuri-nol.
. At the time, Santiago Crisostomo was a practicing pathologist.
.According to The Merck Manual, Stevens-Johnson Syndrome is a severe form of erythe-ma multiforme (inflammatory eruption). It is characterized by the appearance of bullous lesions on the oral mucosa, pharynx, anogenital region, and conjunctiva. In addition,
[t]ypical erythema multiforme lesions may or may not be present elsewhere on the skin. The patient may be unable to eat or close his mouth properly and drools [sic] continually. The eyes may become very painful, and conjunctivitis with swelling and pus may make it impossible for the patient to open them. The conjunctival lesions may leave residual corneal scarring. The condition is occasionally fatal.
The Merck Manual of Diagnosis and Therapy 2290 (15th ed. 1987).
. The ophthalmologist, initally a defendant, is no longer a party to this suit.
. The third amended complaint alleged that the Stanley Group was a professional corporation of which Dr. Stanley was one of the owners, officers and agents. It further alleged that at all relevant times, Dr. Stanley was acting as the duly authorized agent for the Stanley Group. For clarity’s sake, only Dr. Stanley will be named hereinafter.
. Dr. O’Donnell is a pharmacist, not a physician. However, as discussed in note 18 infra, the district court accepted his testimony as an expert.
. In particular, Dr. O’Donnell cited medical studies conducted by Dr. Jick of the Boston Collaborative Drug Surveillance Program which link the drug interaction of Zyloprim and ampi-cillin to allergic reactions.
. Dr. O’Donnell also stated that no such warnings were made by other means, such as package inserts, advertising, the Physician’s Desk Reference, letters to doctors, and so forth.
.This allegation stated:
(a) Defendant failed to use that degree of care and skill in the diagnosis, care and treatment of the plaintiff that a reasonably well qualified physician practicing in the medical specialty of Internal Medicine would have used in the same or similar circumstances.
R.82.
. We recognize that the plaintiffs’ counsel did not agree to drop allegation (e)(4). Rather, he insisted that the testimony of Dr. Carney (the treating dermatologist) had inferentially established Dr. Stanley’s negligence with respect to the steroid therapy. However, Dr. Carney’s testimony was not offered as opinion testimony against any of the defendants. Indeed, counsel for the plaintiffs introduced Dr. Carney’s testimony as follows: "Judge, lest there be any confusion we are offering Dr. Carney as a factual medical witness only, not for any opinions as to any defendants.” Tr. at 694. Furthermore, the plaintiffs do not raise this allegation on appeal. Hence, it has been waived.
. Dr. Stanley notes that Dr. Jarrett’s testimony often referred to methods that a more prudent physician would have followed, or to a practice that he himself would have performed. Dr. Stanley contends that, under Illinois law, such testimony does not establish a breach of duty. Rather, he argues, duty is measured by the acts of a reasonable physician in the same specialty. We doubt the validity of such a hypertechnical distinction. Moreover, Dr. Jarrett had identified himself as a reasonably skilled practitioner in the field of internal medicine. In any case, we need not rely on this distinction because we hold that the plaintiffs failed to establish a causal connection between the alleged breach of a duty and Dr. Crisostomo’s injuries.
. It is clear that both parties believed that this second instance of Dr. Stanley’s alleged failure to warn was before the court. First, there was no attempt to eliminate this issue in the pretrial stage. Second, both parties referred to Dr. Cri-sostomo’s telephone call to Dr. Stanley in their opening statements. Finally, the Crisostomos’ testimony, as well as that of Dr. Jarrett, was admitted with respect to this issue.
. During the opening statement to the jury, Dr. Stanley’s trial counsel hinted that no call was made. Because the directed verdict was granted at the end of the plaintiffs’ case, the record does not contain Dr. Stanley’s account.
. Alternatively, we believe that the general allegation of negligence is an adequate basis upon which to litigate these issues. Dr. Stanley argues that the general allegation is conclusory. We disagree. Ideally, subsequent amendments to pleadings should be more specific than the initial complaint. See C. Wright, Law of Federal Courts § 68 (4th ed. 1983). Nonetheless, the federal rules require only that the pleadings provide sufficient notice to the defendant of the facts and substance of the complaint, id., an objective which certainly has been satisfied here. In a negligence case, such as this one, "a detailed statement of the circumstances ... is not necessary or desirable, and a general allegation of negligence will be sufficient.” 5 C. Wright & A. Miller, Federal Practice and Procedure § 1249 (1969); cf. Eizerman v. Behn, 9 Ill.App.2d 263, 132 N.E.2d 788 (1956) (general allegation of negligence sufficient to support verdict). Moreover, our review of the pretrial documents fails to reveal any narrowing of the issues that might have precluded consideration of the plaintiffs’ failure to warn claim.
. Given our disposition of this matter, we need not determine whether the plaintiffs' motion to amend the pleadings properly was denied.
. See Restatement (Second) of Torts § 402A, comment k (1965) (unavoidably unsafe products, such as drugs, are not unreasonably dangerous so long as they are properly prepared and accompanied by directions and warnings).
. The duty to provide an adequate warning for prescription drugs can be discharged by application of the "learned intermediary doctrine.” As the Illinois Supreme Court recently stated in Kirk v. Michael Reese Hospital & Medical Center, 117 Ill.2d 507, 519, 111 Ill.Dec. 944, 100, 513 *1153N.E.2d 387, 393 (1987), cert. denied, — U.S. -, 108 S.Ct. 1077, 99 L.Ed.2d 236 (1988), prescription drug manufacturers are absolved of liability so long as adequate warnings of a medication’s adverse side effects have been imparted to treating physicians. Thus, manufacturers owe no duty to warn patients directly. Id.
. BWC maintains that Dr. O’Donnell is unqualified to render an opinion on whether the warnings were necessary. BWC points out that Dr. O’Donnell is a pharmacist, not a pharmocologist or a medical doctor. When asked to describe his educational background, Dr. O’Donnell indicated that he had both a Bachelor’s degree and Doctorate in Pharmacy, as well as a Master’s degree in clinical nutrition. Additional background information was elicited on cross-examination as follows:
[Attorney for BWC]: I believe that you indicated ... that [a] pharmacologist is involved in the identification, development, testing and evaluation of the effects of drugs on humans as well as other organisms, is that correct?
[Dr. O'Donnell]: Yes, sir.
Q: You have never received a degree in pharmacology, have you?
A: No.
Tr. at 346.
The district court accepted Dr. O’Donnell’s testimony as an expert witness, although the court expressed its reservations as to his qualifications. We will only reverse a decision to admit expert testimony if we find the court abused its discretion. See Spesco v. General Elec. Co., 719 F.2d 233, 238 (7th Cir.1983). On the basis of Dr. O’Donnell’s background and qualifications, we find no error in the decision to recognize him as an expert witness. Cf. Harris v. Smith, 372 F.2d 806, 813-14 (8th Cir.1967) (the testimony of a qualified medical doctor cannot be excluded simply because he is not a specialist in a particular school of medical practice). There is no reason why a witness who is not a medical doctor should be automatically disqualified as a medical expert, if he is otherwise well schooled in the field. E.g., Baches v. Valspar Corp., 783 F.2d 77, 79 (7th Cir.1986) (nonphysicians can opine on the causes of illness); Dawsey v. Olin Corp., 782 F.2d 1254, 1262 (5th Cir.1986) (nonphysician may testify about the effects of phosgene). See generally Annotation, Products Liability: Admissibility of Expert or Opinion Evidence as to Adequacy of Warning Provided to User of Product, 26 A.L.R.4th 377 (1983). At any rate, we conclude that Dr. O’Donnell’s testimony was inadequate to satisfy the plaintiffs’ burden of proof.
. BWC notes that the chemical property of ampicillin differs from other penicillins. Thus, it claims to have had no advance knowledge of any danger posed by use of penicillin V with Zyloprim.