Plaintiff, Gus Taylor, appeals as of right from a judgment of no cause of action entered following a directed verdict for defendant, Wyeth Laboratories, Inc.
This case arose from the death of plaintiff’s wife, Carrie Sue Taylor. Plaintiff alleged that his wife died from a pulmonary embolism (a blood clot in the lung) which was caused by her consumption of the oral contraceptive Ovral-21, manufactured by defendant Wyeth Laboratories. The medication was prescribed by defendant Dr. Brockington. Defendant Dr. Johnson treated decedent for a broken ankle shortly before her death.
Plaintiff claimed defendant Wyeth had been negligent in two respects: (1) in failing to adequately test Ovral-21 before and after marketing with regard to its use and the occurrence of "thromboembolic disorders” (blood clotting); and (2) in failing to provide physicians with adequate warnings of the specific relationship between Ovral-21 and the occurrence of blood clots, in particular, the increased risk to women with certain blood types. In addition, according to plaintiff, defendant Wyeth failed to advise physicians of available diagnostic tests which should be administered to patients using the substance. Plaintiff alleged that each of these acts or omissions "was a proximate and foreseeable cause” of his wife’s death. The defendant physicians reached a settlement with plaintiff during trial.
*393At trial, plaintiff testified that his wife broke her ankle on January 4, 1979. Dr. Johnson applied a second cast to decedent’s leg on January 19, after which decedent continued to complain about pain in her foot and leg. On February 26, 1979, plaintiff took decedent to Dr. Johnson’s office for an appointment, at which time decedent collapsed outside the doctor’s office and was dead on arrival at the hospital. The medical examiner concluded that she died from a "massive pulmonary embolism or blood clot” which migrated from her leg to the large blood vessel connecting her heart to the lungs. The contraceptive was "a factor in the forming of the clot in this case”. Decedent was 34 years old.
At the close of plaintiff’s proofs, the trial judge granted defendant’s motion for directed verdict. On the claim of failure to test, the judge concluded there was insufficient evidence of the information available to defendant and of "what the testing was going to be all about”. Moreover, in the judge’s view, "there is no indication that the testing would have changed what’s in the labeling”. On the failure to advise of the availability of the blood tests, the judge relied on the testimony of Dr. Laird on cross-examination. The judge recalled that Laird "indicated that these tests were available but he himself [sic] would not advise that these tests be given”. The judge also stated:
"And without such a showing, it would seem the fact that the manufacturer did not warn or advise that these tests were available is not of import, because even if they were available, if in fact they were too expensive, that expense outweighed the risk and would make no difference.
"So I find that particular issue does not weigh in plaintiff’s favor and that no reasonable person can say *394that that testing or advising of the test should have been in the warnings to the physician.”
The judge further held that, even if the warnings were inadequate in the manner averred by plaintiff, there was no showing of proximate cause; in other words, that Dr. Brockington’s "actions would have been any different”. The judge noted that Dr. Johnson was decedent’s physician at the time of her death and that Dr. Johnson testified that he never asked if decedent used the pill. Both doctors "knew of the associated increase of thromboembolic disorder”. The judge concluded that no reasonable person could find that any inadequacy in the warnings was the proximate cause of decedent’s death.
On August 10, 1983, the trial judge heard argument on and denied plaintiff’s motion for new trial. The judge reiterated some of her findings made on the motion for directred verdict. In addition, the judge held that Dr. Johnson’s conduct was an intervening cause, the sole proximate cause of the injury complained of.
On appeal, plaintiff raises two issues. First, he contends that the trial judge erred in granting a directed verdict on plaintiff’s claim that defendant negligently failed to adequately test for and research the cause of the injurious effects of its product. We agree.
If the evidence, viewed in a light most favorable to plaintiff, establishes a prima facie case, a defense motion for directed verdict should be denied.1 If material issues of fact remain, upon which reasonable minds might differ, they are for the *395jury.2 A plaintiff has the right to ask the jury to believe the case as he presented it, however improbable it may seem.3 Directed verdicts are particularly disfavored in negligence actions, since the ultimate questions are so often the reasonableness of defendant’s conduct and whether that conduct was a proximate cause of the plaintiffs injury.4
A manufacturer’s duty of reasonable care includes a duty of product inspection and testing during and after the course of manufacture as is reasonably necessary to render the product safe for its users.5 As in other negligence actions, generally it is a question of fact for the jury whether a manufacturer’s failure to test or inspect the product constitutes negligence.6 Plaintiff must show there was a danger which could have been detected and that the injury was proximately caused by the discoverable defect.7 In this respect also, such a claim is no different from any negligence claim for which proximate cause is one of the elements of a prima facie case.8 This issue is one for the jury, provided there is evidence from which reasonable persons could draw a "fair inference” that the injury was caused by the negligent omission.9_
*396In this case, plaintiff claimed that defendant should have conducted further testing and research on the relationship between Ovral-21 and blood clotting. Plaintiff submitted expert testimony that a prudent drug manufacturer would have explored the more specific relationship between blood types and the risk of blood clotting. Therefore, we believe that it was error for the trial judge to conclude that there was no evidence that defendant had information indicating such a relationship existed. Defendant’s employee, Dr. Lewis, admitted that defendant, along with the other oral contraceptive manufacturers, had such information but did not act upon it because they felt the evidence was insufficient. Plaintiff submitted medical research articles from which the jury could infer what the results of the tests would have been. In addition, Dr. Laird stated that he knew of 13 similar cases in which all of the women concerned had type A blood. Accordingly, reasonable minds could differ as to whether a causal connection existed.10
Defendant argued that there is no study which establishes a direct link between blood type and coagulation factors in the blood, a fact which was admitted by plaintiff’s expert, Dr. Laird. While the lack of a study is cause for academic interest, it does not dispose of the issue. Plaintiff submitted evidence, in the form of medical research articles and Dr. Laird’s testimony, of the heightened danger of thrombosis for women with type A blood. That such women experienced a disproportionate share of the pulmonary embolisms is enough to create a jury question on the "failure to test” claim.
*397The trial court held that there was no showing of proximate cause, i.e., that defendant Wyeth would have changed its warning as a result of the tests which it failed to perform. However, plaintiffs "failure to test claim”, like all those of its kind, requires a certain degree of inferential reasoning. We believe that, on the basis of the evidence submitted, a reasonable jury could properly infer that the failure to test or conduct research on the blood type question was negligence.
Plaintiff also claims on appeal that the trial court erred in granting a directed verdict for defendant on plaintiff’s claim that defendant negligently failed to convey adequate warnings and instructions on the dangers associated with the use of Ovral-21.
A manufacturer of a prescription drug has a duty to warn the medical profession of any risks inherent in the use of the drug which the manufacturer "knows or should know” exist.11 If warnings or instructions are required, the information provided must be adequate, accurate and effective.12 Determination of whether this duty has been breached in the context of a negligence claim necessitates that the warnings given be examined as to their reasonableness under the circumstances.13 Whether referred to as "the specific standard of care”, or simply as "the reasonableness of the defendant’s conduct”,14 the issue is for the jury, unless all reasonable persons would agree or there is some overriding legislative policy to the con*398trary. This Court has repeatedly held that the adequacy of a warning and the reasonableness of a failure to warn are questions of fact.15 Moreover, "the duty to warn may exist even where the danger associated with the product is so slight that few have ever been injured by it”.16
In the within case, plaintiff submitted expert testimony on three alleged inadequacies in the warnings accompanying defendant’s product. First, according to the testimony, the warning concerning the danger posed by prolonged immobilization was inadequate, since it was vague and "buried” deep within the printed material. The record does not, therefore, support defendant’s argument that there was no evidence of the warning’s inadequacy. This was a jury question, to be answered according to the jury’s view of the evidence, including the expert testimony and exhibits at hand.
There was evidence that defendant was aware of the correlation between blood type and blood clotting and chose not to warn about it. In addition to the admission of Dr. Lewis that defendant was aware of this danger, this Court has held that "the manufacturer is held to the knowledge of an expert and is presumed to know of scientific studies and articles concerning the safety of its products”.17 Dr. Laird testified that the patient’s blood coagulation factors should be tested and monitored and that two such tests were available at a combined cost of $24 to $40.
When defense counsel impeached Dr. Laird with his testimony in a 1978 trial, in which Laird said *399the tests should not "be required on a routine basis”, and that the tests "are expensive”, the witness stated he had changed his opinion since 1978. Further, Dr. Laird stood by his prior statement, explaining that his reference to "the advisability of using the tests” is "different than doing it in every single patient”. The trial court held, however, that Dr. Laird’s prior testimony, that in 1978 he felt the tests were too expensive to be routinely required, was conclusively favorable to defendant on this point. We believe that this was error.
On a motion for directed verdict, "the trial judge may not select among actual or seeming contradictory statements of a witness given on direct examination and cross-examination what he believes should be applied to that motion”.18 The inconsistency in the expert’s testimony presented a question of credibility for the jury.19 Moreover, the conflict was not an absolute one, given the evidence in the case. Even if Dr. Laird still held the view that a doctor need not give the tests routinely when he prescribes an oral contraceptive, a woman with type A blood would hardly present a "routine” case in which a doctor might reasonably omit these tests, given the information that women of that blood type were at greater risk.
More basically, the question of whether the expense outweighed the risk is precisely the kind of issue which must be decided by the jury. Dr. Laird’s personal view is perhaps relevant, but certainly not dispositive. This is a products liability case, not a professional negligence or malpractice case in which expert testimony would be required to demonstrate a breach of the pertinent *400standard of care.20 Determination of whether a battery of tests costing $40 is too expensive where a person’s life may be at stake does not necessarily call for a strictly professional judgment. The jury had evidence of the risk and the means available to obviate that risk. It was free to reject the expert’s opinion.21
Finally, the trial judge held that there was no showing of proximate cause. Specifically, she ruled that there was no evidence that Dr. Brockington would have changed his advice to decedent if defendant’s warnings were altered to plaintiffs specifications, and that Dr. Johnson’s conduct was an intervening, superseding cause of the injury. In addition, defendant argues on appeal that the evidence conclusively established that decedent would have continued taking Ovral-21 even if the warnings had been made adequate.
Defendant notes that decedent continued taking the pill even though she experienced headaches, implying that this action shows that she would have also disregarded a warning linking her blood type to an increased danger of fatal blood clots. Apparently, this Court is asked to equate chronic headaches with the risk of fatal thromboembolic disorder. Defendant further notes that the danger of blood clotting is stated in the warnings. The trial judge also attached relevance to this fact:
"The Court would note that it is in fact a thromboembolic disorder from which Mrs. Taylor died and that this type of disorder is clearly warned about in the label and that reasonable people could not differ that this particular warning was contained.”_
*401Reasonableness is a question of fact.22 Here, the trial judge studied the warnings and found that they were adequate. However, this issue is not one for the judge to resolve. The jury could reasonably find that a generalized risk of blood clotting in women is not the same as a specific, enhanced risk, applicable to a specific class of individuals with a certain blood type. Defendant argues (and the trial court found) that Wyeth was entitled to the directed verdict because there was no evidence that Dr. Brockington advised decedent of the danger or that he would have relayed a revised warning to her, if the warning had been provided to him by defendant Wyeth. In fact, the record plainly contains the doctor’s statement that it was his custom to give his patients such warnings. There was no direct evidence of what he told the patient in this case because he had no recollection of her. However, evidence of his routine practice, though uncorroborated, is relevant to prove the conformity of his conduct thereto.23 The jury could reasonably infer that, if it had been provided to him, Dr. Brockington would have given a revised warning, that he would have tested decedent’s blood coagulation factors, and that her problem was discoverable and her death not inevitable.
"The proximate cause of an injury has been defined as that which in a natural and continuous sequence, unbroken by any new, independent cause, produces the injury, without which such injury would not have occurred.”24 An act of negligence does not cease to be a proximate cause of the injury because of an intervening act of negli*402gence, if the prior negligence is still operating and the injury is not different in kind from that which would have resulted from the prior act.25 The courts of this state have held that whether an intervening negligent act of a third person constitutes a superseding proximate cause is a question for the jury.26 An intervening cause is not an absolute bar to liability if the intervening event is foreseeable, though negligent or even criminal.27
Judged by this standard, it is clear that whether Dr. Johnson’s conduct was a superseding cause of the injury was a jury question. Plaintiff admitted by his complaint that Dr. Johnson (and Dr. Brockington, for that matter) was negligent.28 That is not the point. The court apparently ignored Dr. Brockington’s prescription of the medication based on defendant’s allegedly inadequate warning and concentrated instead on Dr. Johnson’s negligence. However, decedent’s use of the drug might have been stopped before she broke her ankle if she had been tested or warned.
As in Muilenberg v The Upjohn Co,29 plaintiff has presented a case in which a jury might find that the prescribing physician was not warned of the drug’s side effects with sufficient particularity. In such a situation, the physician, being ignorant of the enhanced risk to women with type A blood, would have no reason to hesitate.
For these reasons, we reverse the trial court’s *403judgment. We hold that plaintiff is entitled to a new trial on each of the claims raised against defendant Wyeth Laboratories.
Reversed.