In this case we answer a certified question of law from the United States Court of Appeals for the District of Columbia Circuit. The issue is whether the plaintiff must present medical opinion testimony on causation to establish a prima facie case of negligence. The Circuit Court of Appeals has certified this question of law in accordance with the District of Columbia certification statute. D.C.Code § 11-723 (1995). See generally Penn Mut. Life Ins. Co. v. Abramson, 530 A.2d 1202, 1205-08 (D.C.1987) (describing our statutory scheme for certification of questions of law). This statute authorizes the Circuit Court of Appeals to certify questions of local law to this court when, in that court’s estimation, we have no controlling precedent determinative of the pending cause. D.C.Code § ll-723(a). Although the plaintifiyapp ellant filed his complaint in the United States District Court for the District of Columbia, our local law must provide the rule of decision for this appeal because the claim itself arises under the District of Columbia’s law of negligence. See generally Erie R.R. Co. v. Tompkins, 304 U.S. 64, 58 S.Ct. 817, 82 L.Ed. 1188 (1938) (holding fed*1382eral courts must apply local law to nonfederal actions).
The Circuit Court of Appeals certified the following question of law:
Must a medical malpractice plaintiff, in order to have his claim presented to a jury, present medical opinion testimony on the issue of causation (whether, to a reasonable degree of medical certainty, it is more probable than not that the procedure was a substantial factor in causing the injury), where (1) the injury occurred during the medical procedure itself, (2) the evidence establishes that the resulting injury is a recognized risk of such a procedure, and (3) the evidence establishes that there is some risk, in this case less than a five-percent-per-year chance, that the injury would occur spontaneously in the absence of the procedure?
Certification of Question of Law, No. 94-7168 (D.C.Cir. Sept. 29, 1995) (emphasis in original). We hold that the plaintiff in this medically complex case must establish causation with medical opinion testimony.
I. Factual and Procedural History
This case involves both a complex medical condition and a complicated medical treatment. Appellant Michael A. Lasley first visited his physician, Dr. Ronald Wheeler, when he experienced frequent intense headaches, numbness and tingling on his body’s right side, and impaired vision in his right eye. A CAT scan revealed the presence of a congenital arteriovenous malformation (AVM) approximately four centimeters in size adjacent to the anterior left lobe of Lasley’s brain. An AVM is an abnormal tangle of fragile blood vessels. These vessels substantially disturb cerebral blood flow. After exiting large cerebral arteries, blood normally travels to small veins. An AVM, however, diverts the blood leaving the brain. The accumulation of the diverted blood inside the AVM is dangerous because the AVM’s fragile vessels can easily rupture, causing severe hemorrhaging. Cerebral hemorrhaging in turn causes brain damage that can be fatal.
Upon discovering the AVM, Dr. Wheeler referred Lasley to Dr. Anthony J. Caputy, a Georgetown University Hospital neurosurgeon. Dr. Caputy confirmed the existence of the AVM by inspecting an angiograph of Lasley’s brain. He also learned that several branches of Lasley’s middle and anterior cerebral arteries fed blood into the AVM. Dr. Caputy further referred Lasley to Dr. Alfred J. Luessenhop, the chief of neurosurgery at the hospital. On June 15,1989, Dr. Luessen-hop met with Lasley to discuss treatment options for his AVM. On June 20, 1989, Dr. John P. Deveikis, the head neuroradiologist, also met with Lasley to discuss treatment. Lasley decided to go forward with a particular treatment, and Lasley, who was eighteen years old, and his mother both signed a consent form authorizing Dr. Deveikis to perform the procedure despite the acknowledged risks. Lasley alleges that neither Dr. Luessenhop nor Dr. Deveikis adequately explained the risks of the procedure. Therefore, although he admits consenting to the procedure, Lasley claims his consent was uninformed.
Dr. Deveikis treated Lasley’s AVM by em-bolizing it. In an embolization procedure, a catheter is placed into the carotid artery of the neck. A second catheter is then introduced via the first into the feeder artery that diverts blood leaving the brain into the fragile AVM vessels. An emboli (chemical mixture) is then inserted into the feeder artery via the catheters. Once inside, the emboli occludes (obstructs) the blood supply from the feeder artery to the AVM, facilitating subsequent surgical removal of the tangled AVM vessels. The procedure may be repeated for each of the discrete feeder arteries carrying blood into the AVM. While the procedure is preoperative and was performed in an angiogram suite rather than an operating room, it is nonetheless an invasive procedure akin to surgical operation.
The principal risk of the embolization procedure is identical to that of the AVM condition itself. Both the AVM condition and the embolization procedure pose a danger of vessel rupturing and hemorrhaging. Dr. De-veikis testified that “the risk of the AVMs is that it could bleed, and if you do nothing, you have that risk.” According to the various treating doctors, the probability of a natural AVM rupture and hemorrhage is approxi*1383mately three to five percent per year for anyone with an AVM. Dr. Deveikis testified that Lasley himself presented “a higher risk for hemorrhage.” An embolization treatment of an AVM is dangerous for the same reason. As Dr. Deveikis stated, “Anytime you do anything to an AVM, there’s a possibility it could cause the fragile vessels to burst, and that could cause bleeding, and the bleeding could cause damage to the brain.... That’s the major fear of doing anything to the AVM, and it’s also kind of ironic that that’s the thing we are trying to prevent.” The evidence does not precisely indicate the severity of the risk that the embolization procedure itself might cause the AVM to rupture and hemorrhage. The record, therefore, fails to reveal the comparative risk Lasley purportedly faced between the AVM condition and the embolization treatment.
When Dr. Deveikis embolized Lasley’s second artery feeder, this risk materialized. The artery feeder ruptured and hemorrhaged, causing a severe intracerebral hema-toma (blood clot). Lasley was immediately taken to the operating room, where Dr. Luessenhop performed emergency surgery, saving Lasley’s life. Dr. Luessenhop successfully removed both the hematoma and the AVM. Lasley did, however, sustain injuries, including aphasia (speech difficulty) and right-sided hemiparesis (body weakness).
Lasley filed a negligence claim under District of Columbia law in the United States District Court for the District of Columbia. In his complaint, Lasley premised his claim on the negligent failure of Drs. Luessenhop and Deveikis to obtain Lasley’s informed consent. Lasley, however, did not claim that Dr. Deveikis was negligent while performing the procedure. Lasley’s sole negligence contention was that neither Dr. Luessenhop nor Dr. Deveikis properly advised him of the risks of the embolization procedure. Lasley asserts that despite consenting to the procedure, he would have withheld his consent had the doctors fully explained to him the risks and the alternatives.
The District Court granted the defense motion for judgment as a matter of law after Lasley presented his evidence. The trial judge explained that Lasley failed to present any expert testimony on the causation element of his claim. The trial judge concluded that District of Columbia law required Las-ley to present expert testimony that the em-bolization procedure caused Lasley’s AVM to rupture and hemorrhage. Lasley appealed to the United States Court of Appeals for the District of Columbia Circuit, which subsequently certified the question to this court.
II. Analysis
The Circuit Court of Appeals has certified the narrow question whether, under the circumstances of this case, District of Columbia law requires medical opinion testimony to prove causation. We specifically examine whether it was fatal to Lasley’s claim that none of his witnesses explicitly asserted that the embolization procedure caused his AVM to rupture and hemorrhage. Our rule is that medical opinion testimony is normally required in medically complicated cases. We have, however, acknowledged certain exceptions to an expert witness requirement in other types of negligence cases where the issues are less complicated. We conclude that the rule requiring expert testimony applies to this particular case.
A. Appellant’s Arguments
Lasley urges us to adopt his view that it is unnecessary for an expert medical witness to explain the causation element of his informed consent claim. Although he concedes that neuroscience is complicated in general, he contends the issue of causation pertinent to his negligence claim is quite simple.
Lasley argues that causation in this case is simple because the causation is obvious. He stresses that only two possibilities exist: Either the embolization procedure or the AVM itself caused his blood vessels to rupture and hemorrhage. Moreover, the rupture and bleeding occurred during the embolization procedure. In Lasley’s estimation, the temporal coincidence of the procedure and the rupture reveals the causal link between them.
Lasley also implicitly argues that the issue of causation is simple because of the unlikelihood, or indeed the impossibility, that the *1384embolization procedure is not the cause of the rupture. He reiterates the trial statistic that absent the procedure, there is a mere three to five percent chance that an AVM will rupture and hemorrhage on its own in any given year. He also declares that it is medically impossible for an AVM to rupture spontaneously once an embolization procedure is underway.
Based upon these reasons, Lasley posits that the jury was completely capable of inferring the relevant causal relation without reliance on medical opinion testimony. Lasley additionally states that doctors themselves are puzzled by the precise causes of vessel rupture, so expert testimony could provide nothing more than supposition.
B. Appellee’s Counter-arguments
Georgetown counters Lasley’s simplicity argument by emphasizing the complexity of the causation issue in this case. Georgetown argues that a lay jury is ill equipped to evaluate which of the two medically complicated causal explanations is more persuasive. Medical opinion testimony is necessary, Georgetown contends, because of the overwhelming complexity of this evaluation.
Georgetown concurs with Lasley that there are only two causal candidates: Either the vessels in Lasley’s AVM burst spontaneously, or the embolization procedure was the cause. In Georgetown’s view a sufficient understanding of each possibility requires specialized medical insight jurors do not typically possess. Furthermore, since either explanation is plausible, it is Lasley’s burden to prove that the procedure rather than the preexisting condition is the actual cause. Georgetown contends the jury cannot meaningfully make this sophisticated judgment without evaluating medical reasons described by medical experts.
Georgetown also argues it is insufficient to show solely that the rupture occurred dining the procedure. They reason that an approximate temporal overlap establishes a mere correlation when the law requires proof of causation. Finally, Georgetown notes that if medical experts are unable to discern the true cause, jurors may not even try to do so.
C. Discussion
Before beginning our analysis, we make some preliminary observations. Lasley’s complaint alleges the doctors’ negligent failure to obtain his informed consent. Such a claim initially requires proof of nondisclosure of material risks. Gordon v. Neviaser, 478 A.2d 292, 294 (D.C.1984). Whether Lasley’s proof of nondisclosure was persuasive at trial in the District Court is an issue of fact, and we offer no assessment here of the sufficiency of the evidence in this regard. A claim of uninformed consent also requires proof of causation. “Once there has been a breach of the duty to disclose, the patient must demonstrate a causal relation between the physician’s failure to disclose the material information and the injury sustained.” Id. (citing Canterbury v. Spence, 150 U.S.App. D.C. 263, 281, 464 F.2d 772, 790, cert. denied, 409 U.S. 1064, 93 S.Ct. 560, 34 L.Ed.2d 518 (1972)). The parties here have stipulated that Lasley’s ultimate injuries — such as speech difficulty — are the adverse consequences of the cerebral bleeding. The only disputed causation issue, therefore, is the specific question of whether the embolization procedure caused Lasley’s AVM to rupture and hemorrhage.
Our rule for medically complicated cases is that proof of causation normally requires medical opinion testimony. “Ordinarily, in a medical malpractice case, expert testimony is required in order to prove ... causation.” Id. at 295 (quoting Sponaugle v. Pre-Term, Inc., 411 A.2d 366, 368 (D.C.1980) (citations omitted)). This requirement in medical malpractice cases generally also applies to that class of malpractice cases specifically claiming lack of informed consent. For example, proof of causation remains an element of such a claim. “[A]s in malpractice actions generally, there must be a causal relationship between the physician’s failure to adequately divulge and damage to the patient.” Canterbury, supra, 150 U.S.App. D.C. at 281, 464 F.2d at 790 (citations omitted); Gordon, supra, 478 A.2d at 294 (citing Canterbury). Furthermore, medical experts are still required to establish the causal link. Even “[i]n non-disclosure cases, ... they are normally needed on issues as to the cause of any *1385injury_” Gordon, supra, 478 A.2d at 295 (quoting Canterbury, supra, 150 U.S.App. D.C. at 283, 464 F.2d at 792). Experts are essential in medical negligence cases alleging nondisclosure because the issue of causation is “distinctly related to some science, profession, or occupation....” Washington v. Washington Hosp. Ctr., 579 A.2d 177, 181 (D.C.1990) (quoting District of Columbia v. Peters, 527 A.2d 1269, 1273 (D.C.1987)). As the Circuit Court of Appeals has commented, “medical facts are for medical experts.” Canterbury, supra, 150 U.S.App. D.C. at 283, 464 F.2d at 792 (citations omitted).
“The purpose of expert opinion testimony is to avoid jury findings based on mere speculation or conjecture.” Washington, supra, 579 A.2d at 181. This rationale explains why we have acknowledged limited exceptions to the rule. Expert testimony is not required if the issue of causation can be “resolved wholly within the realm of ordinary human knowledge and experience,” Gordon, supra, 478 A.2d at 295-96 (quoting Canterbury, supra, 150 U.S.App. D.C. at 282-83, 464 F.2d at 791-92), or if “the proof is so obvious as to lie within the ken of the average lay juror,” Washington, supra, 579 A.2d at 181 (citations omitted). We have summarized the exceptions as follows:
To prevent the jury from engaging in speculation, we have held that in the absence of expert testimony, a jury may not consider the causal connection between a defendant’s negligence and a plaintiffs claimed disability unless:
(1) the disability first emerged coincidentally with or very soon after the negligent act, or
(2) the disability was of a type which by its very nature reflected its cause, or
(3) the cause of the injury related to matters of common experience, knowledge, or observation of laymen.
Baltimore v. B.F. Goodrich Co., 545 A.2d 1228, 1231 (D.C.1988) (relying on Early v. Wagner, 391 A.2d 252, 254 (D.C.1978); Jones v. Miller, 290 A.2d 587, 590-91 (D.C.1972); Wilhelm v. State Traffic Safety Comm’n, 230 Md. 91, 185 A.2d 715, 719 (1962)). It is Laslejfs view that his case falls within these readily apparent exceptions rather than the rule itself. We disagree.
We are of the view that the issue of causation in this case is so medically complicated that Lasley was required to introduce medical opinion testimony explaining precisely how and why the embolization procedure caused the vessels in his AVM to rupture and bleed. We agree with Georgetown that a jury may be incapable of rationally determining the actual cause of the rupture without the assistance of expert testimony.
Two reasons motivate our conclusion. First, a thorough understanding of each of the two possible causes is essential yet difficult to attain without the benefit of medical expertise. Spontaneous AVM ruptures and embolization procedures are neither obvious causes nor subjects within the realm of ordinary human knowledge and experience. Second, even if a jury could understand each of the two causal alternatives, without medical opinion testimony, it has no rational basis for evaluating which of the two possible causes actually did cause the AVM to rupture and bleed. A jury is unable to make this determination on its own because it has no logical reason to choose one over the other. Without access to relevant medical facts, the jury has no reasoned basis for concluding whether the AVM or the emboli-zation caused the rupture. “To allow a jury of laymen, unskilled in medical science, to attempt to answer such a question would permit the rankest kind of guesswork, speculation and conjecture.” Baltimore, supra, 545 A.2d at 1231 (quoting Wilhelm, supra, 185 A.2d at 719).
In this ease the plaintiff never presented an expert to explain to the jury precisely how or why the embolization procedure rather than the AVM itself caused any injury. We find this deficiency to be fatal to Lasle/s claim. This exact evidentiary error was fatal to a plaintiff’s prima facie proof of uninformed consent in a previous District Court case. See Blincoe v. Luessenhop, 669 F.Supp. 513, 517 (D.D.C.1987). In a nearly identical factual setting, a patient suffering from an AVM faded to present medical opinion testimony that an embolization procedure caused her injury. The Blincoe court ana*1386lyzed the significance of this omission as follows:
To establish the medical fact of the causal link between Mrs. Blincoe’s March 4, 1983 “stroke” and the embolization procedures, plaintiffs must proffer medical testimony, see Canterbury, 464 F.2d at 792....
[[Image here]]
Plaintiffs refer the court to no portion of the record indicating a causal link between the February embolizations and the March 4 “stroke.” This court, in addition, can find no medical evidence in the medical records, the depositions or the affidavits establishing a causal relationship between the February embolizations and the stroke of March 4. Without any competent medical evidence to establish a causal link, we must conclude as a matter of law that no causal relationship exists between the February embolizations and the March 4 stroke.
Id. at 516-17 (emphasis added). The District Court reached the same conclusion in this case. The trial judge confirmed that “nowhere in the record” did Lasley’s evidence “establish a causal link between the emboli-zation and the bleeding.”
It is not uncommon to encounter this inherent difficulty of evaluating between an ailment and subsequent medical treatment as the cause of an injury. In medical malpractice cases it is often unclear whether an injury is caused by the preexisting malady or the medical procedure performed to alleviate it. This principle is well recognized:
Almost every person who receives the services of a physician is sick or disabled when he first goes to the physician. Thus, in a medical malpractice case, there is the ever-present possibility that it was the 'patient’s original affliction rather than the physician’s negligence which caused the ultimate damage. Malpractice cases often present the difficult problem of distinguishing the effects of the defendant’s negligence from the natural results of the plaintiff’s disease.
Typically, expert testimony will be required to establish causation in fact for the jury.
1 David W. Louisell & Harold Williams, Medical Malpractice § 8.05, at 8-71 (1995) (footnotes omitted) (emphasis added).
We have recognized this difficulty in prior cases. In Gordon, supra, appellant injured his shoulder while skiing. 478 A.2d at 293. After his doctor surgically operated twice, his injury worsened. Id. The patient sued for negligent nondisclosure of the risks of the surgery. Id. We, however, affirmed a directed verdict against him, explaining as follows:
The testimony of an expert was necessary to establish both that there was a risk involved in the surgery and that the condition of appellant’s shoulder deteriorated as a result of the surgery ....
Appellant faded to establish through expert testimony, however, that the surgery was the cause of his aggravated disability. This lack of proof on the issue of causality was fatal to his medical malpractice claim.
Id. at 296 (emphasis added). We reached the same conclusion in Baltimore, supra, where an employee fell five or six feet because of the negligent operation of a hydraulic lift. 545 A.2d at 1230. The employee suffered from “numerous physical and psychological ailments” before the accident. Id. After discussing exceptions to the expert testimony rule, we nonetheless concluded that because of “the multiple preexisting and concurrent possible causes ... this case falls within that category of ‘medically complicated’ cases requiring expert testimony.” Id. at 1231. We also note that Lasley’s case bears no resemblance to McQueen, a case where we did find a legitimate exception to the expert testimony rule. International Sec. Corp. of Virginia v. McQueen, 497 A.2d 1076 (D.C.1985). That case involved personal injury rather than medical malpractice, there was no preexisting medical condition that might have caused the injuries, and the actual cause of the injuries was readily apparent. Id. at 1078 (according to plaintiff, a security guard “pushed me back into the room, ... karated me to the floor and I banged my ribs against the metal bench against the wall”).
We are unpersuaded by Lasley’s arguments that the facts of this case compel application of an exception to the expert testimony requirement. Having explained why *1387the evaluation between the two causal alternatives is complicated, we now turn to the reasons why it is insufficient to prove causation solely with evidence that the rupture occurred during or immediately after the procedure.
In a medically complicated case such as this, contemporaneity between a medical procedure and an injury is too weak a foundation upon which to infer causation. Correlation and causation are hardly synonymous. We find that “a proximate temporal association alone does not suffice to show a causal link” because a mere temporal coincidence between two events does not necessarily entail a substantial causal relation between them. See Hodges v. Secretary of the Dep’t of Health & Human Servs., 9 F.3d 958, 960 (Fed.Cir.1993) (quoting Grant v. Secretary of the Dep’t of Health & Human Sews., 956 F.2d 1144, 1148 (Fed.Cir.1992)). Conse quently, more evidence is required. “Without more, this proximate temporal relationship will not support a finding of causation.” Hosier v. United States, 718 F.2d 202, 205 (6th Cir.1983), cert. denied, 469 U.S. 817, 105 S.Ct. 84, 83 L.Ed.2d 31 (1984) (plaintiff failed to show that her immunological reaction to a swine flu inoculation caused the auto-immune disease irom which she suffered). The need for further proof is especially clear here because it is possible that Lasley’s AVM ruptured spontaneously during the embolization procedure. This possibility complicates the causation issue, and this added dimension of complexity only increases the need for clarity.
We are impressed by the complete absence of clarification during the District Court trial. As the trial judge stated, “We have no one saying it’s more likely than not ... that the intrusion into that part of the AVM by the doctors doing the embolization caused the bleed, as opposed to it bled because it was going to bleed anyway_” No doctor ever detailed exactly why or how Lasley’s rupture actually occurred. Lasley’s doctors failed to identify any specific causal mechanism or articulate any precise sequence of cause and effect. See Blincoe, supra, 669 F.Supp. at 517 (insufficient evidence to establish a causal link); see generally H.L.A. Haet & Tony Honoke, Causation in the Law 410 (2d ed.1987) (if two possible causes, and details of actual cause unknown, burden of proof not met without at least “evidence of the characteristically different processes by which different causes produce their effects”). Informed only of the conjunction of a condition, a procedure, and an injury, a lay jury could not have pinpointed the cause of Lasley’s vessel rupture without blindly guessing.
If we were to conclude otherwise — that contemporaneity proved causation — we might inappropriately shift the plaintiffs burden of proof onto the defendant. Instead of requiring the plaintiff to indicate why the injury occurred, we would in effect be forcing defendants to disprove causality. For example, a jury could easily be tempted to conclude that any injuries sustained during brain surgery necessarily result from the operation itself because brain surgery is inherently risky. Such a conclusion here would constitute an unreasonable application of the doctrine of res ipsa loquitur because both parties agree that there was always an appreciable risk that the AVM might rupture on its own even if the embolization procedure were never performed. See Blincoe, supra, 669 F.Supp. at 517 (“res ipsa loquitur is inapplicable”).
We also disagree that the evidence presented on the improbability of the alternative cause is sufficient to prove causation. While Lasley is correct that he need not disprove the alternative explanation, he must nevertheless affirmatively prove that his explanation is indeed the right one. This proof is not established simply by showing that for everyone suffering from an AVM, a spontaneous rupture in everyday life is statistically unlikely in any given year. We instead require particularized proof that this embolization procedure more likely than not substantially caused this AVM to rupture and hemorrhage. The following exchange during the District Court trial clearly illustrates the insufficiency of Lasley’s proposed inference of causation from comparative risks alone:
PLAINTIFF: Let me ask you [Dr. Caputy], is the risk of hemorrhage greater in an *1388embolization procedure or just in natural history?
DEFENDANT: Objection, Your Honor.
THE COURT: Yes, I’ll sustain the objection. He wasn’t brought in here as an expert to testify.
Without such a crucial medical opinion identifying the relative risks of the two potential causes, a jury could not reliably determine the actual cause solely from probabilities. Moreover, we discount Easley’s declaration in his brief that a spontaneous rupture is medically impossible during an embolization procedure. A medical expert may be uniquely competent to assert such a medical proposition, but in this trial no expert ever presented this opinion.
We also reject Lasle/s reliability argument. He maintains that medical opinion testimony is unnecessary because the exact reasons why AVM embolization can cause vessel rupture are unknown even to the experts. We agree with Georgetown that if this is so, then the jury is foreclosed from making a guess about it. It would appear that
if the plaintiffs medical expert cannot form an opinion with sufficient certainty so as to make a medical judgment, there is nothing on the record with which a jury can make a decision with sufficient certainty so as to make a legal judgment.
McMahon v. Young, 442 Pa. 484, 276 A.2d 534, 535 (1971).
Consequently, we conclude that in this medically complex case, medical opinion testimony on causation was necessary to survive a motion for judgment as a matter of law. The clerk shall certify this answer to the United States Court of Appeals for the District of Columbia Circuit.
So ordered.