Appellants, plaintiffs below, lost their medical malpractice suit in a jury trial. They argue here that it was error for. the trial court to require their medical expert to state his opinion in terms of a “reasonable medical certainty” as to what was “the” (as opposed to “a”) proximate cause of their injury. Because the record as a whole does not reveal that appellants’ case was restricted in this manner, we affirm.
Judith Ann Sponaugle obtained an abortion from Pre-Term, Inc., performed by its employee Dr. Kim, both appellees here. At the time of the abortion, an intra-uterine device (IUD) was present in Mrs. Sponau-gle’s uterus. The device was not removed during the procedure, although she had informed a member of Pre-Term’s staff of its existence, and was told it would be removed. There were no complications from the abortion. More than four months later, Mrs. Sponaugle was hospitalized for abdominal pain. During the course of the ensuing treatment for a uterine infection, x-rays revealed the presence of the IUD. It was surgically removed. Since the infection was still evident, a dilation and curettage was also performed. When the infection did not respond to treatment, Mrs. Sponau-gle was again hospitalized and underwent a total hysterectomy and bilateral salpingo-oophoreetomy.
Mrs. Sponaugle and her husband brought suit against Pre-Term and Dr. Kim1 for their negligence in failing to remove the IUD prior to the abortion, and for not informing her that it was still in place. They alleged that the presence of the IUD aggravated the uterine infection, necessitating the total hysterectomy. The jury rendered a verdict in favor of defendants.2
A key component of appellants’ case at trial was the testimony of their expert medical witness, Dr. Sewell, on the issues of standard of care and causation. Dr. Sewell testified that the failure to inform the patient of the continuing presence of the IUD did not conform to the standards of care in the community.3 Counsel then asked Dr. Sewell to give his opinion, based on “a reasonable degree of medical probability,” as to whether defendants’ failure to so inform was “a” proximate cause of the hysterectomy. In response to defense counsel’s objections, the court directed the witness to testify to “a reasonable medical certainty” that the failure to remove the IUD, and its presence, was “the” proximate cause of the total hysterectomy. Appellants contend that these rulings were error; that Dr. Se-well was prepared to testify with reasonable medical probability that the IUD was a proximate cause of the hysterectomy.
I.
An expert witness opinion must be based on fact or adequate data. It is properly received so long as it is not a mere guess or conjecture. While absolute certainty is not required, opinion evidence that is conjectural or speculative is not permitted. Emmet v. American Insurance Co., *368D.C.App., 265 A.2d 602 (1970); St. Lewis v. Firestone, D.C.Mun.App., 130 A.2d 317 (1957). The precise degree of certainty required varies from one jurisdiction to another. See Annot., 66 A.L.R.2d 1082 (1959). Ordinarily, in a medical malpractice case, expert testimony is required in order to prove the proper standard of care and causation. Kosberg v. Washington Hospital Center, 129 U.S.App.D.C. 322, 324, 394 F.2d 947, 949 (1968); Washington Hospital Center v. Butler, 127 U.S.App.D.C. 379, 384 F.2d 331 (1967).
Because these essential elements of proof can only be shown through expert testimony, there is good reason for courts to take steps to assure that reliable opinions are given.4 This is particularly important where the causation element is unclear. To that end, many courts have held that “medical testimony as to the mere possibility of a causal relation ... is not sufficient, standing alone, to establish such relation.” 31 Am.Jur.2d Expert and Opinion Evidence § 185 p. 753 (1967). In a medical malpractice case, where legal and factual causation are major issues, there may well be a sound basis for the requirement that expert medical testimony be given in terms of “reasonable medical certainty.” However, in some situations, such as where factual causation is indirect or the doctrine of res ipsa loquitur applies, that standard might be unduly restrictive. Cf. Quin v. George Washington University, D.C.App., 407 A.2d 580, 585 (1979) (circumstantial evidence must make plaintiff’s theory reasonably probable to establish causation as an element of res ipsa loquitur).
We do not read the cases, cited by the parties here, as having decided this evidentiary issue in medical malpractice cases;5 moreover, we need not decide it here. The record does not reveal that Dr. Sewell was in any way hampered by being instructed to express his opinion based on a reasonable medical certainty. Whether in terms of certainty or probability, it appears that his difficulty was with testifying that the IUD was a proximate cause, without further qualification, of the hysterectomy.
Appellants proffered that their expert “was willing to testify as to the medical probability of the IUD making the infection more severe, and to a medical probability that the IUD was a proximate cause of the hysterectomy.” But at trial, when deposed, or in his written report to appellants, their expert never stated causation so unequivocally. We think the following excerpt from the transcript demonstrates this point.
Q. [Plaintiffs’ Counsel] Now, doctor, do you have an opinion, based upon a reasonable degree of medical probability, as to whether or not the failure of Pre-Term, and its employees, to inform Plaintiff, Mrs. Sponaugle, that her IUD was still in *369her body after she left their clinic, was a proximate cause of the hysterectomy that she had on June 2, 1974? ******
A. It’s my opinion that that is a probable proximate cause of the patient’s ultimate—
[Defense Counsel]: I object to the doctor’s answer, and ask that it be stricken.
THE COURT: What did you call it?
THE WITNESS: A probable proximate cause. I can’t be for certain without having—
THE COURT: . . Doctor, we talk in terms of reasonable medical certainty.
I gather you understand that. THE WITNESS: Right.
[Our emphasis.] -.
It is at this point that the court asked whether the doctor could give his opinion, within the realm of reasonable medical certainty, that the failure to inform was the proximate cause. The expert responded no. But it is clear that prior to this limiting direction by the judge, appellants’ expert had answered an inquiry in terms of reasonable medical probability. That opinion was not as conclusive as appellants suggest.
Appellants have failed to establish that their case was “weakened” by the court’s ruling on the medical certainty standard. The causal connection between defendants’ alleged negligence and plaintiffs’ injury was tenuous at best. Under these circumstances we think requiring expert testimony on causation under a reasonable medical certainty standard, if error at all, was not reversible error.6 The very nature of the opinion being given was skirting the line between contributing cause and speculation. The trial judge did not commit reversible error in asking that this critical evidence be stated in terms of certainty, particularly since appellants have not shown actual prejudice.
II.
Turning to appellants’ second assignment of error, we again conclude that the record simply does not evidence the restriction complained of. Although the court directed Dr. Sewell to testify in terms of “the” proximate cause in a colloquy noted as error here, the record reveals that the witness freely gave his opinion in terms of “a” proximate cause. It appears to us that plaintiffs’ efforts to show causation were thwarted by their own expert, rather than limitations imposed by the trial judge. They simply failed to establish the causal nexus between defendants’ negligence and the hysterectomy.7
Dr. Sewell’s testimony as to causation was always couched in terms of “a probable proximate cause,” “a proximate contributing cause,” and “a contributory cause” (our emphasis). He testified that the infection was the proximate cause of the hysterectomy; the IUD was a proximate contributing cause to the infection and the subsequent hysterectomy; the IUD made the infection more difficult to treat or more severe. He could not state, however, that had the IUD been removed, the hysterectomy would have been unnecessary. He further elaborated: “[Y]ou asked me whether the IUD caused the infection or not. The IUD didn’t cause the infection. I can’t say that the IUD caused the hysterectomy. The infection caused the hysterectomy. The IUD made the infection worse. If all of that *370means ‘proximate cause,’ then there was a proximate cause.”8 The full extent of the expert’s opinion was thus presented to the jury. He was repeatedly allowed to state his opinion without qualification as to “the” or “a” proximate cause.
Based on this record, we cannot agree with appellants that their expert was limited to testifying as to “the” proximate cause.9 He repeatedly refused to causally connect the IUD to the hysterectomy under any test, be it “the” proximate cause or “a” proximate cause. His answers would have been the same under either rule.
The causation in appellants’ case was fatally weak. The jury simply did not accept their theory. It concluded not only that the defendants’ negligence did not cause the hysterectomy, but also that it did not aggravate the infection. Appellants have made no showing that their case was restricted or harmed by the trial court’s rulings.10 Under all the circumstances and weighed in light of the entire record, we cannot say that the rulings of the trial judge, if erroneous, were of such magnitude as to require reversal.
Affirmed.
. Prior to trial, appellants dismissed their case against a third defendant, Dr. Sites, who treated Mrs. Sponaugle for the infection and performed the surgery.
. Because of the difficult causation issue, two special interrogatories were given to the jury: 1) did the defendants’ negligence proximately cause the hysterectomy?; 2) did their negligence proximately cause the infection? The jury answered no to both questions.
.It was uncontroverted that no part of the IUD was visible to the doctor performing the abortion. There was no evidence that the failure to remove the IUD under these circumstances was itself negligent.
. See McMahon v. Young, 442 Pa. 484, 486, 276 A.2d 534, 535 (1971) (“[I]f the plaintiff’s 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.”).
. In none of the cases cited by appellants was the evidentiary standard before the court. Testimony based on a “reasonable medical probability” was allowed without objection. Mitchell v. Woodworth, 146 U.S.App.D.C. 21, 449 F.2d 1097 (1971) (worker’s compensation case; expert opinion based on incomplete facts was not sufficient to support Department of Labor decision); Raza v. Sullivan, 139 U.S.App.D.C. 184, 432 F.2d 617 (1970), cert. denied, 400 U.S. 992, 91 S.Ct. 458, 27 L.Ed.2d 440 (1971) (dental malpractice; doctrine of res ipsa loquitur of inferred negligence can be applied to expert’s testimony that particular result would not “ordinarily” happen); Zimmerman v. Safeway Stores, Inc., 133 U.S.App.D.C. 342, 410 F.2d 1041 (1969) (wrongful death and medical malpractice; appeal from directed verdict; sufficient evidence of causation to put case to jury; nonexpert evidence on conditions and results can be used); Garfield Memorial Hosp. v. Marshall, 92 U.S.App.D.C. 234, 204 F.2d 721 (1953) (medical malpractice; sufficiency of evidence to support finding of negligence; jury not required to believe speculative evidence of possible other causes); Christie v. Callahan, 75 U.S. App.D.C. 133, 124 F.2d 825 (1941) (medical malpractice case on sufficiency of evidence to support verdict for plaintiff); but cf. Baerman v. Reisinger, 124 U.S.App.D.C. 180, 363 F.2d 309 (1966) (medical malpractice; whether general medical practitioner can be expert on hypothyroidism; reasonable degree of medical certainty used).
. We do not decide the issue of which standard is correct, since appellants have not shown that their expert’s opinion would have been different under the “probability” standard.
. Throughout the record of this case, and without objection, the issue of factual causation has erroneously been characterized as “proximate cause,” meaning “that cause which, in natural and continuous sequence unbroken by an efficient intervening cause, produces the injury, without which the result would not have occurred.” Proximate cause relates to legal, not factual, causation. Appellants were evidently using it to mean “but-for” causation-in-fact. Construing the evidence in a light most favorable to the appellants, we think factual causation was not shown. Defendants’ negligence clearly was not a “but-for” cause of the hysterectomy; nor was there a showing that their failure to inform plaintiff of the presence of the IUD was, under any other test, a factor in producing the hysterectomy.
. This case demonstrates the problems that arise when an expert witness states his opinion on the ultimate issue of a case. Appellants’ expert stated his opinion on what was the proximate cause of the hysterectomy — a legal conclusion. His testimony reveals that he did not understand the legal meaning of the term. His opinions would have been more useful to the jury had they been confined to the issue of causation-in-fact.
. Had the witness actually been held to this limitation, it might have been error.
.“[E]rror is not to be presumed but must be made affirmatively to appear by the party asserting it.” Berenter v. Staggers, 124 U.S.App. D.C. 141, 143 n.2, 362 F.2d 970, 972 n.2 (1966) (citations omitted); a reviewing court will not reverse for harmless error, nor for any error which does not affect the substantial rights of the parties. Sullivan v. Yellow Cab Co., D.C. App., 212 A.2d 616 (1965).