History of the Learned Intermediary Doctrine
The term “learned intermediary” was first used in Sterling Drug, Inc. v. Cornish, with the court recognizing that properly warning the doctor of a drug’s side effects leads to an “excellent chance that injury to the patient can be avoided.” Although seemingly straightforward, states have yet to come to a consensus on the scope of the defense. The doctrine generally recognizes physicians as “learned intermediaries” who have specialized medical knowledge and are therefore in the best position to assess the risks and benefits of a treatment for their patients. Some states have codified the principle, many have adopted it in their highest courts, and a few have failed to explicitly accept or reject it. Regardless of the scope of its application, this doctrine remains a valuable defense for drug and device manufacturers in every state.
All jurisdictions have recognized some form of the learned intermediary doctrine, typically under the following rationales:
- Physicians are uniquely aware of a patient’s special medical history and therefore the risks and benefits of a particular drug or device.
- Drugs and medical devices are not normal consumer items and therefore may require complex medical and technical information on the risks and benefits, which are difficult for patients themselves to understand.
- Limiting manufacturer warnings to physicians allows for consistency in warning duties as required by the Food and Drug Administration (FDA).
- Routing information through trained medical professionals avoids the practical difficulties associated with drug and device companies communicating directly with patients.
Scope of the Learned Intermediary Doctrine
Drug and device manufacturers are required to adequately inform the physician about the risks associated with their products. Once this is done, the manufacturer is typically not held liable if the physician fails to properly inform the patient. The operative questions are therefore (1) who constitutes a “physician” and (2) what constitutes an “adequate warning” for purposes of the doctrine.
Who is a physician? The learned intermediary doctrine set out in Restatement (Third) of Torts: Products Liability describes the rule in terms of warnings “provided to . . . prescribing and other health-care providers.” There have been numerous interpretations of “other health-care providers,” but some courts have expanded this to include those outside the “traditional” doctor-patient relationship, such as nurses:
The modern healthcare system, however, places far less emphasis on these traditional relationships, and patients today often receive the majority of their care from nurses, nurse practitioners, physicians’ assistants, and physicians other than the prescribing physician. Appropriately, in situations where the underlying rationale of the doctrine—the traditional doctor-patient relationship—is no longer present, the doctrine has adapted to fit the realities of the situation. . . . [T]he evolution of the doctrine [has been] through an expansion of the possible class of learned intermediaries. This development, likewise spurred by the fact that the medical professionals with whom patients most commonly interact are often no longer primary physicians, has led courts and secondary sources such as the Restatement to suggest that a variety of different healthcare providers may be considered learned intermediaries, depending on the unique facts of the patients’ treatment scenario. . . . We concur with authorities who consider the learned intermediary to be the healthcare professional actually responsible for making decisions related to the patient’s care.
This expansion of the doctrine has similarly been recognized with respect to physician assistants who prescribe drugs.
The evolving “physician” definition has strong implications for drug and device defendants. This is particularly true where the duties of the medical professional are blurred. For example, in Holley v. Burroughs Wellcome Co., a court held that a nurse anesthesiologist was a learned intermediary to whom the manufacturer owed a duty to warn. The nurse anesthesiologist failed to recognize a known adverse reaction to a drug, which ultimately caused brain damage. The defendant manufacturer argued that its duty to warn was satisfied because the learned intermediary doctrine was limited to the prescribing physician. The court disagreed based on the context, reasoning that while the defendants’ argument “would be appropriate in cases . . . where a nurse only administered a medicine . . . but was otherwise not responsible for the patient’s care,” the nurse anesthesiologist in Holley was responsible for all aspects of patient treatment. Thus, the court concluded, the duty to warn ought to apply to her as the primary health care provider standing in between the patient and the manufacturer.
Certain professions, on the other hand, such as pharmacists, are not considered learned intermediaries for these same reasons—they are not the sole health care provider standing between the patient and the manufacturer. However, some courts have made exceptions. For example, in DiVincenzo v. Molinaro, the Connecticut Superior Court held that pharmacists have no duty to warn under the learned intermediary doctrine absent one of the following exceptions: (1) the pharmacist has specific knowledge of potential harm to specific persons in particular cases, (2) the pharmacist makes a representation that they will evaluate possible effects by the administration of a drug or combination of drugs, or (3) there is something unambiguously wrong with the prescription itself.
The question of how far drug and device warning liability goes is not boundless. However, drug and device manufacturer defendants are well served by keeping apprised of the expanding categories of medical practitioners considered to be learned intermediaries, given the significant ambiguities as to whom warnings must be directed.
What is an adequate warning? Even though a manufacturer’s duty to warn runs only to the learned intermediary, that warning must still be adequate. An “adequate warning” has been defined as one “sufficient to apprise the general practitioner as well as the ‘unusually sophisticated medical man’ of the dangerous propensities of the drug.” “[I]t is incumbent upon the manufacturer to bring the warning home to the doctor.” If a physician has not read the product warnings, there can be no liability because nothing about the adequacy of the warnings caused the injury.
A warning that is deemed inadequate (through failing to “apprise the general practitioner . . . of the dangerous propensities of the drug”) must also be found to be causally related to the adverse patient outcome. In order to demonstrate that an inadequate warning proximately caused the plaintiff’s injury, she “must show that adequate warnings would have altered her doctors’ decision to prescribe [the drug].” Finally, the learned intermediary doctrine shields the manufacturer from liability even if the doctor learns of the risks from a source other than the manufacturer.
Adequate Warning and Instructions for Mitigating Risk: Blackburn v. Shire
The notion that a physician is best equipped to guide a patient through an ideal treatment plan suggests that instructions on minimizing risks extend beyond a manufacturer’s duty. However, the New York Court of Appeals has held that “[t]he warning must provide sufficient information to that category of prescribing physicians who may be expected to have the least knowledge and experience with the drug.” Depending on the case, this may set the bar so low as to require more than just a recitation of the risks.
Recently, the Alabama Supreme Court has gone even further by holding that an adequate warning regarding a drug’s risks not only entails listing known side effects but may also include guidance on mitigating those effects, especially when such instructions inform physicians of the severity of a specific risk. In Blackburn v. Shire U.S., Inc., the plaintiff had been prescribed Lialda as an “off-label” treatment for Crohn’s disease. Though Lialda was not FDA-approved for Crohn’s, it was approved for ulcerative colitis, a closely related condition. After taking Lialda for 12–16 months, the plaintiff developed an advanced form of kidney disease. The plaintiff brought suit against defendant manufacturer Shire for failure to warn on the basis that the Lialda label should have included instructions for monthly renal function testing for the first three months of treatment, followed by quarterly monitoring for at least a year. According to the plaintiff, the absence of these testing recommendations in the labeling proximately caused his kidney damage.
The district court granted summary judgment in favor of Shire based on the testimony of the plaintiff’s physician that he did not read the Lialda label before prescribing it to the plaintiff. Although the plaintiff’s physician testified that he would have followed a more explicit instruction, the district court dismissed this testimony as “unsubstantiated speculation.”
On appeal, the U.S. Court of Appeals for the Eleventh Circuit refused to affirm summary judgment based on how the district court interpreted the physician’s testimony. However, Shire had argued alternatively that the plaintiff’s liability theory as to instructions for risk mitigation was not recognized under Alabama law. In light of this alternative argument, the Eleventh Circuit certified two questions for the Alabama Supreme Court’s review: (1) Does a pharmaceutical company’s duty to warn include an obligation to provide instructions on minimizing warned-of risks? (2) Can a plaintiff establish causation in a failure to warn claim by proving that their doctor would have implemented different monitoring procedures, even if they would have still prescribed the same drug?
In response to the first question, the Alabama Supreme Court ruled that an adequate warning may need to incorporate instructions for mitigating side effects. The court concluded that its previous adoption of Restatement (Second) of Torts section 402A comment k suggests that a prescription drug manufacturer’s duty to warn is not necessarily limited to listing known side effects but may also include instructions on reducing those risks.
Shire argued that allowing instructions for mitigating warned-of risks to be part of a manufacturer’s duty would intrude upon a physician’s practice of medicine. The court dismissed this argument, stating that if the Lialda label’s existing instruction for monitoring “does not interfere with a physician’s practice of medicine, it is difficult to see why Blackburn’s desired instruction of monitoring a patient’s renal function”—monthly for the first three months and then quarterly for at least a year—“represents the drastic intrusion upon the physician-patient relationship that Shire claims it to be.” The court went on to note that although Lialda’s label explicitly states that it is approved for ulcerative colitis, this did not prevent the plaintiff’s physician from exercising his medical judgment in prescribing it for Crohn’s disease. Similarly, prescription drugs typically include dosage recommendations, but physicians often adjust those based on patient needs and medication tolerance. The court found that this reasoning applies to monitoring instructions—if Lialda’s label had specified renal function testing intervals instead of merely advising “periodic” testing, physicians would still retain the discretion to alter those recommendations according to their medical judgment. According to the court, the key issue is not whether monitoring instructions intrude on physician responsibility but whether simply listing side effects is enough to sufficiently inform doctors of a drug’s risks.
The court answered the second certified question in the affirmative, stating that since a failure to warn claim may include an allegation that the labeling was inadequate because it did not include instructions for mitigating use, a plaintiff may prove causation by establishing that their physician would have adopted a different course of testing or mitigation, even if the physician prescribed the same drug.
On return from the Alabama Supreme Court, the Eleventh Circuit rejected Shire’s “argument that it was precluded from changing the warning because it was contained in the ‘Highlights’ section of the LIALDA label.” Notably, the type of label change demanded by the plaintiff requires FDA approval under 21 C.F.R. § 314.70(b)(2)(v) and should have been preempted because it requires FDA approval. Nevertheless, the U.S. Supreme Court denied certiorari on October 2, 2023.
The Blackburn decisions require attention on multiple levels. The preemptive line between FDA labeling regulations and state law claims involving risk mitigation instructions is clear as mud. Setting preemption aside, there is no bright line establishing the point at which instructions for risk mitigation should be included. There are also physician-patient implications. The learned intermediary doctrine is based on the premise that physicians, not drug manufacturers, are best suited to prescribe medication and determine how a patient should be monitored. By imposing a duty to include instructions on how patients should be monitored while on prescription medication, the manufacturer is essentially taking the place of a medical provider. Physicians must choose whether to follow the drug manufacturer’s instructions or to instead rely on their own education and experience with each individual patient, with whom the drug manufacturer has had no contact.
Causation and a Plaintiff’s Burden: Himes v. Somatics
Historically, the learned intermediary doctrine has focused on physician conduct as opposed to patient conduct. In 2024, the California Supreme Court flipped this calculus by focusing on what a patient would have done with additional warnings. In Himes v. Somatics, LLC, the plaintiff underwent electroconvulsive therapy (ECT) using a device manufactured by defendant Somatics as a treatment for severe depression. She alleged that Somatics failed to properly warn her physician about the risks of permanent brain damage and memory loss associated with the use of the device. The plaintiff contended that had she been fully informed of these risks, she would have refused the treatment.
The federal trial court granted summary judgment for Somatics, finding that the plaintiff failed to establish causation because she lacked evidence that stronger warnings would have influenced her physician’s decision to prescribe the treatment. The trial court reasoned that since the manufacturer’s duty was to warn the physician—and the physician testified that he would have still recommended the treatment regardless of stronger warnings—the plaintiff could not establish causation.
On appeal, the U.S. Court of Appeals for the Ninth Circuit agreed that the plaintiff had not provided evidence showing that her physician would have changed his decision to prescribe ECT if Somatics had issued stronger warnings. However, the Ninth Circuit found that the causation analysis remained unresolved, particularly as to whether a stronger warning would have affected the plaintiff’s choice to proceed with treatment. The court certified the following question to the California Supreme Court:
Under California law, in a claim against a manufacturer of a medical product for a failure to warn of a risk, is the plaintiff required to show that a stronger risk warning would have altered the physician’s decision to prescribe the product? Or may the plaintiff establish causation by showing that the physician would have communicated the stronger risk warnings to the plaintiff, either in their patient consent disclosures or otherwise, and a prudent person in the patient’s position would have declined the treatment after receiving the stronger risk warning?
The California Supreme Court accepted certification. Contrary to the federal trial court’s holding, the California Supreme Court concluded that a plaintiff does not need to show that a stronger warning would have changed the physician’s prescribing decision to establish causation. Instead, a plaintiff may meet this burden by proving that (1) the prescribing physician would have communicated the stronger warning to the patient, and (2) an objectively prudent person in the patient’s position would have declined treatment upon receiving the stronger warning, regardless of the physician’s advice.
While underscoring the importance of informed consent and patient autonomy, the court also recognized that a plaintiff will likely claim that they would have refused treatment had they received a stronger warning. In response, the court adopted an “objectively prudent person” standard: “the test . . . is not determined by the plaintiff’s subjective belief as to what he or she might have done with the benefit of hindsight.” Under this new theory, a plaintiff must prove that an objectively prudent person in their situation would have declined treatment despite the physician’s judgment that the benefits outweighed the risks.
In applying this standard, the court provided a nonexhaustive list of factors to consider, including: (1) whether the physician assessed the treatment’s risks and benefits and, after discussing them with the patient, still recommended proceeding; (2) whether the treatment was an established practice or a novel approach for the patient’s condition; (3) the availability and effectiveness of alternative treatments and whether they had been attempted; (4) the severity of the patient’s condition; (5) the likelihood of the treatment providing more than marginal benefits; and (6) the patient’s personal characteristics or unique circumstances.
While the Himes court’s holding is premised upon what an objectively prudent person would do, the list of factors adds subjectivity and uncertainty. There are also preemption implications. The ECTs at issue in Himes were initially subject to the premarket approval (PMA) process but were later reclassified as class II devices under the 510(k) pathway. This regulatory shift creates uncertainty not only in how express preemption will be argued but also in how express preemption precedent will be applied. Ultimately, the ruling’s impact on manufacturers’ litigation risks will depend on how Himes is interpreted on remand. Until then, several aspects of Himes limit its reach.
First, the California Supreme Court reaffirmed the learned intermediary doctrine and its applicability to both prescription drugs and medical devices. The ruling also confirmed that the physician’s treatment recommendation remains central to any learned intermediary analysis and preserves existing defense arguments for disputing causation. For instance, a plaintiff cannot survive summary judgment using this new theory if their physician based their decision on independent knowledge of the risks. Similarly, a defendant is entitled to summary judgment if the physician never read the warning, meaning that a stronger warning would not have influenced their prescribing decision.
To succeed under this new theory, a plaintiff must offer evidence that an objectively prudent person would have disregarded their physician’s recommendation. Notably, the court rejected the notion that a plaintiff could defeat summary judgment solely by submitting a self-serving affidavit. Instead, the court adopted an objective approach:
The causation analysis must . . . consider whether an objectively prudent person in the patient’s position would have declined the treatment even where his or her physician would have advised the patient that the treatment would still be in the patient’s best interest, notwithstanding the risks conveyed by a stronger warning.
Consequently, even if the prescribing physician would have revised their consent process in response to a hypothetically stronger warning, if they would have still recommended the treatment to the plaintiff, a defendant may argue that no objectively prudent patient would have refused the treatment under those circumstances.
Second, the question before the California Supreme Court was a narrow one: whether a plaintiff can meet the causation requirement by showing that (1) the doctor would have received a stronger warning, (2) the doctor would have relayed that warning to the plaintiff, and (3) the additional risk information would have led an objectively prudent patient in the plaintiff’s position to decline treatment against their physician’s advice. Since the plaintiff in Himes obtained testimony from her prescribing doctor that he would have modified his consent procedures if Somatics had disclosed additional risk details, the California Supreme Court did not need to address whether the plaintiff could have succeeded without such affirmative testimony. That said, the court made clear that proving causation is the plaintiff’s responsibility, strongly implying that a plaintiff must present affirmative testimony showing that their prescribing doctor would have adjusted their consent procedures to utilize this new theory:
The causation analysis should therefore begin by determining what, if anything, the patient’s physician would have communicated to the patient regarding the relative risks and benefits of the prescription drug or medical device in response to a stronger warning, and should then turn to whether an objectively prudent person in the patient’s position would have declined the treatment . . . . In other words, the plaintiff must prove that an objectively prudent person in [their] position would have declined treatment despite the physician’s assessment that the benefits of the treatment for the patient would still outweigh any risks disclosed by a stronger warning.
Third, the court explicitly rejected the argument that the learned intermediary doctrine does not apply when the manufacturer fails to provide adequate warnings to physicians. The court correctly recognized that such a position would conflate the separate legal elements of duty and causation:
If the manufacturer fails in its duty to adequately warn a physician . . . , [that duty] does not transform, retroactively, into a duty to warn the patient. Instead, the manufacturer’s failure to warn the physician results in a breach of its general duty of care to the patient under negligence principles or a breach of its obligation to market a product free from defects under strict liability principles.
However, a plaintiff still has to demonstrate that this breach caused their alleged injuries.
In August 2024, the Himes case returned to the Ninth Circuit, where the court ultimately vacated the California Supreme Court’s summary judgment and remanded the case back to the district court. The court therefore will need to assess whether an “objectively reasonable person in the plaintiff’s position” would have opted against the treatment had they been provided with a stronger warning.
Takeaways
While your jurisdiction might not be California or Alabama, there are a few overarching takeaways to remember. Ultimately, the decision to prescribe a prescription medication or medical device boils down to a risk/benefit analysis that is applied to each individual patient, based on their personal health history and other factors related to that particular patient. Regardless of jurisdiction, it is important to confirm that a treating physician applies the risk/benefit analysis for each individual patient, including the plaintiff. Additionally, ask whether the treating physician routinely discusses the side effects of drugs with each individual patient, and whether that same process was followed with the plaintiff.
Application of the learned intermediary doctrine may be far from predictable, but clear manufacturer-physician communications can be the best safeguard for defendants against potential liability.
