BREWER, J.

Affirmed.

BREWER, J.

In this obstetrical malpractice case, plaintiffs appeal from a judgment for defendants after a jury verdict. We affirm.

Plaintiffs are the parents of and guardians ad litem for Joshua. During his birth, Joshua's left shoulder became stuck behind his mother's pubic bone, a circumstance known in the medical field as "shoulder dystocia." The doctors who delivered Joshua applied traction to free his shoulder. After he was born, Joshua's left arm was flaccid, and doctors diagnosed an injury to the brachial plexus nerves, which innervate the arm. Despite several surgeries, Joshua has permanent severe damage to his arm known as Erb's palsy. Plaintiffs filed this action, alleging multiple specifications of negligence and seeking damages for Joshua's injury.

A primary issue at trial was whether Joshua's injury was caused by defendants' negligence in applying excessive lateral traction to the shoulder and head during Joshua's birth. The jury heard evidence from a number of experts concerning the possible causes of injuries to the brachial plexus nerves and Joshua's injury specifically. On cross-examination, witnesses were asked about and referred to numerous articles in medical literature. They also referred to articles forming the basis for or supporting their opinions. The jury reached a verdict for defendants, and plaintiffs appeal, raising numerous assignments of error. We have considered and reject each assignment. We write to address a single assignment of error concerning defense counsel's discussion of certain medical literature during closing argument.

In their fourth assignment of error, plaintiffs contend that the trial court erred in allowing defense counsel to read several excerpts of medical literature to the jury in closing argument. In Oregon, medical treatises or articles, like all learned treatises, are deemed to be hearsay and are not admissible as substantive evidence. Eckleberry v. Kaiser Foundation, 226 Or. 616, 620, 359 P.2d 1090 (1961); Scott v. Astoria Railroad Co., 43 Or. 26, 39, 72 P. 594 (1903); Travis v. Unruh, 66 Or. App. 562, 565, 674 P.2d 1192, rev den, 297 Or. 82 (1984); Laird C. Kirkpatrick, Oregon Evidence § 803.18, Art VIII-125 (4th ed 2002). Excerpts from medical literature may, however, be offered as the basis of expert opinion testimony under OEC 703 or may be used to impeach an expert witness if the expert either relied on the treatise in forming an opinion or acknowledges it as a recognized authority in the field. Devine v. Southern Pacific Co., 207 Or. 261, 275-76, 295 P.2d 201 (1956); Kern v. Pullen, 138 Or. 222, 231-32, 6 P.2d 224 (1931), overruled in part on other grounds by Fitze v. American-Hawaiian S.S. Co., 167 Or. 439, 117 P.2d 825 (1941). Here, all the information contained in the excerpted materials had been presented to the jury through witness cross-examination and discussion of the authors' works; however, the quoted excerpts themselves had not been read to the jury, with the exception of those portions specifically noted below.

We first address defendants' contention that the excerpted materials were not hearsay because they were not offered as substantive evidence of their contents but, rather, to show the evolution of medical thought on the likely causes of brachial plexus injuries in shoulder dystocia. We do not perceive the significance of that distinction for present purposes. In any event, whether or not the challenged materials could have been admitted as substantive evidence if timely offered is beside the point. Even if the materials had been timely offered and admitted solely as foundations for expert opinion testimony or to impeach such testimony, the particular excerpts at issue were not included in the evidentiary record. The first question on appeal thus is whether the trial court abused its discretion in allowing defendants' counsel to read them to the jury in closing argument. See R.J. Frank Realty, Inc., v. Heuvel, 284 Or. 301, 306, 586 P.2d 1123 (1978) (rulings concerning jury argument are reviewed for abuse of discretion). As explained below, we conclude that the court erred.

The regulation of jury argument is left largely to the discretion of the trial court. Id. Jury argument properly may include reference to matters that are within the scope of the issues and evidence, but evidence outside the record may not be suggested to the jury by any means. OEC 103(3). A court abuses its discretion by allowing jury argument regarding facts not in evidence. Atlas Copco Industrial v. Karn Repair Service, 172 Or. App. 317, 323-25, 18 P.3d 1102, rev den, 332 Or. 316 (2001). The "facts" that the trial court allowed defense counsel to argue were the quoted excerpts from medical literature that supported defendants' theory as to the cause of Joshua's injuries and were consistent with the opinions expressed by defendants' witnesses. However, the challenged excerpts had not been received in evidence for any purpose. We conclude that the trial court abused its discretion in permitting defense counsel to refer to them in closing argument.

Error is not reversible, however, unless the losing party demonstrates that the error substantially affected his or her rights. See ORS 19.415(2); Jett v. Ford Motor Co., 335 Or. 493, 497, 72 P.3d 71 (2003); Shoup v. Wal-Mart Stores, Inc., 335 Or. 164, 173, 61 P.3d 928 (2003) ("The possibility that an error might have resulted in a different jury verdict is insufficient under [ORS 19.415(2)]. Instead, the court must be able to conclude, from the record, that the error `substantially affect[ed]' the right of the losing party."). Defendants assert that any error in permitting defense counsel to quote the challenged materials did not substantially affect plaintiffs' rights because those materials were cumulative of evidence that had been presented to the jury.

We have examined the record with care, and focus on the five excerpts for which plaintiffs preserved objections. First, defense counsel read from a 1997 article by Dr. Ouzounian:

"We identified four cases that occurred in the absence of shoulder dystocia and four cases that occurred in the posterior arm of infants with anterior shoulder dystocia. These data further support the notion that the etiology of permanent brachial plexus palsy associated with birth may not be related to traction."

Although that particular excerpt was not introduced in evidence, several witnesses mentioned the author and discussed the substance of his opinions on the subject of the excerpt, and a witness identified the article as authoritative and reliable. A witness also discussed the substance of the excerpt. We conclude that the reading of the excerpt during closing argument did not substantially affect plaintiffs' rights.

Second, defense counsel also read an excerpt from an October 1997 "Practice Patterns" publication of the American College of Obstetricians and Gynecologists:

"There is evidence that injuries might and do occur despite application of appropriate obstetrical maneuvers. Injuries are a common outcome associated with shoulder dystocia and may occur despite use of appropriate standard obstetrical maneuvers. Brachial plexus injuries and others, fracture of the humerus, fracture of the clavicle, are the most commonly reported injuries associated with shoulder dystocia."

(Emphasis added.) Several witnesses discussed the October 1997" Practice Patterns" publication in their trial testimony. Only the emphasized sentence of the quoted excerpt was not included in the evidentiary record, and it does not bear on the disputed issue of causation. We conclude that defense counsel's reference to it in closing argument did not substantially affect plaintiffs' rights.

Third, defense counsel discussed an article authored by Dr. Gherman, which was not read at trial:

"Another article by Dr. Gherman before the one that we've been talking aboutcame out — and you heard about this — March of 1998:

"`Brachial plexus injury has often been assumed to be the result of excessive lateral head traction — head flexion applied during delivery, thereby selectively stretching the nerve root. This causes transient dysfunction with restoration of function expected in most cases. There has been a general tendency to presume that most, quote, no-shoulder brachial plexus injuries represent under-reporting of shoulder dystocia. Previously reported serious brachial plexus palsies, however, * * * have consistent findings suggesting otherwise.'

"And then it picks up here:

"`More than half, 633 of 100 `— excuse me —' of 1149, 51 — 55.1 [percent] of the nerve injuries in ten recent reports occurred with a reported shoulder dystocia. Our data, taken together with the preceding reports, provides several lines of evidence to show that not all Erb's palsies are traction related, rather an in-uterine insult, perhaps, combined with a susceptibility to pressure or traction may be etiologic.'"

Fourth, counsel quoted from a June 1998 Gherman article that also was not included in the evidentiary record:

"`The occurrence of brachial plexus injury, therefore, should not necessarily be equated with iatriogenic' — mean[ing] doctor-caused — `birth injury.'"

Although the foregoing excerpts were not included in the evidentiary record, expert witnesses made several references to Gherman, his articles, and his opinions on the causes of brachial plexus injuries. Dr. Edelberg, one of plaintiffs' experts, was cross-examined concerning Gherman's views. He testified that he disagreed with them. Dr. Clark, one of defendants' experts, testified that Gherman's research supported his opinion that more than 50 percent of brachial plexus injuries are due to causes other than obstetrical maneuvers. In his cross-examination of Dr. Sandmire, plaintiffs' counsel allowed Sandmire to refer to the findings of doctors Ousounian and Gherman, as well as to those of Dr. Gonik.

Fifth, defense counsel quoted from a 1998 article by Dr. Gonik, the text of which was not included in the record:

"[W]e have determined that the maternal expulsive forces are from four to nine times greater than the forces of traction that are used in various settings of delivery."

However, several witnesses mentioned Gonik and discussed the substance of his studies, including the quoted article, at trial.

We conclude that, although the trial court erred by allowing defense counsel in closing argument to read written materials to the jury that were not included in the evidentiary record, in light of the fact that the record included the substance of those materials, that error did not substantially affect plaintiffs' rights. We have considered and reject without discussion plaintiffs' remaining assignments of error.

Affirmed.