The plaintiff Kelly Brusard is the mother of the now eight year old child Kaileigh and brought this medical malpractice action on her own behalf and as Kaileigh’s mother and next friend. The defendant is the obstetrician who delivered Kaileigh at Brockton Hospital late in the evening of April 24, 1990. The gist of the plaintiffs’ complaint is that the defendant’s failure to deliver Kaileigh by cesarean section much earlier that *289day resulted in significant neurological injury to Kaileigh. The case was tried to a jury on special questions, the first of which the jury answered by finding that the defendant had not been negligent, and judgment then entered for the defendant. The trial judge denied the plaintiffs’ new trial motion. The plaintiffs appeal from the trial judge’s exclusion of certain proffered evidence, which they contend was prejudicial error, and seek a new trial.
1. Background. Kelly Brusard’s due date for her first child was April 13, 1990. Eleven days later, she went into Brockton Hospital at about 11 a.m. for the continuation of some fetal testing begun the day before with normal results. On April 24, however, the result of the so-called biophysical test profile was abnormally low (a score of two out of a possible eight2), the mother was thereafter admitted to the hospital, and her labor commenced. The defendant arrived at the hospital at 2 p.m., aware of the fetal biophysical test profile. As labor progressed, the monitored fetal heartbeat dropped from 150 to eighty beats per minute at 5:34 p.m., and the mother was prepared for a cesarean section. The fetal heartbeat returned to normal, labor was progressing, and no surgery was then performed. At 7:30 p.m., Dr. O’Toole broke the mother’s water and found thick green meconium. At 8:45 p.m., after the mother’s cervical dilation decreased from eight to five centimeters and Dr. O’Toole had determined that labor was no longer progressing, he ordered a cesarean section which he performed at 9:45 p.m. At the time of her birth, the child’s Apgar score was three at one minute when normal is ten. Kaileigh was later transferred to Children’s Hospital, received a final diagnosis of, inter alia, birth asphyxia, and sustained significant ongoing brain damage.
The plaintiffs’ experts, an obstetrician and a pediatric neurologist, testified that a biophysical profile is a measure of, among other things, the baby’s oxygen supply, that Kaileigh’s biophysical profile of two out of eight indicated that the baby had suffered a major insult at some point after the testing with normal results the day before, that the baby needed immediate delivery, that she was deprived of oxygen throughout her mother’s labor, and that this continued oxygen deprivation, or asphyxia, to the brain occasioned blockage of arteries to the brain, increasing the risk of thrombosis and brain tissue death. The plaintiffs’ *290pediatric neurologist opined that Kaileigh suffered three strokes from thrombosis, at least one of them at 5:34 p.m. when fetal blood pressure dropped precipitously, and that the bulk of the damage she suffered occurred in the last ten, and particularly the last six, hours of labor. On this view, had Dr. O’Toole delivered the baby sooner, Kaileigh would not have suffered as much, if any, injury.
The defendant’s experts, including a gynecologist, pathologist, and pediatric neurologist, saw the matter quite differently. In essence, their view was that Kaileigh had suffered a stroke causing all the damage prior to the time the biophysical profile was performed, that the biophysical profile score reflected this and not asphyxia or other fetal distress, that the stroke was caused by an embolus, not thrombosis, that there was. no evidence of asphyxia or acute fetal distress during labor, that in any event oxygen deprivation does not cause an embolus to travel, that Kaileigh’s stroke was not caused by asphyxia, and that earlier delivery would have made no difference to Kaileigh’s condition.
2. Evidentiary exclusions. The plaintiffs complain that they were prejudiced by the erroneous exclusion of two pieces of evidence: the written policies and procedures of Brockton Hospital in effect on April 24, 1990, that pertain to the care and treatment of obstetrical patients, and a chart taken from a text, Williams on Obstetrics. We address the latter point first.
The trial judge would neither permit the plaintiffs to introduce in evidence the desired portion of the text nor allow the plaintiffs to use the entirety of a chart relating to the interpretation of fetal biophysical test scores in the cross-examination of the defendant’s expert witnesses. The plaintiffs contend that the portion of the text itself was admissible under G. L. c. 233, § 79C, and that, in any event, pursuant to Proposed Massachusetts Rule of Evidence 803(18), they should have been permitted to use the text portion to cross-examine Dr. O’Toole’s expert witnesses.
The requirements of Proposed Mass.R.Evid. 803(18), adopted by the Supreme Judicial Court in Commonwealth v. Sneed, 413 Mass. 387, 396. (1992), are not the same as those in G. L. c. 233, § 79C. Rule 803(18) provides:
“The following are not excluded by the hearsay rule, even though the declarant is available as a witness: *291. . . (18) Learned treatises. To the extent called to the attention of an expert witness upon cross-examination, statements contained in published treatises . . . on a subject of . . . medicine . . ., established as a reliable authority by the testimony or admission of the witness or by other expert testimony or by judicial notice. If admitted, the statements may be read into evidence but may not be received as exhibits.”
In Commonwealth v. Sneed, 413 Mass. at 396, the court observed that there is “potential benefit, and little risk of harm, in adopting the proposed Massachusetts rule of evidence, which permits a party challenging an expert’s opinion to interrogate that expert about a relevant, divergent opinion expressed in a reputable book.” The Sneed court continued, “Admission in evidence of a statement from a treatise of the kind referred to in proposed rule 803(18), whose authenticity and reliability are shown, which was not written for use in litigation, and which expresses an expert opinion on a subject relevant to the case on trial, will tend to enhance, rather than detract from, the truth-finding function.” Ibid.
The chart at issue here has three columns, the first indicating the biophysical profile score, the second indicating the interpretation of that score, and the third indicating the recommended patient management. The portion of the chart that the plaintiffs attempted to use in cross-examining the defendant’s expert witnesses had in the first column a biophysical profile score of “0-2,” in the second column the interpretation of that score, i.e., “Almost certain fetal asphyxia,” and, in the third column, the recommended patient management, i.e., “Deliver.” Curiously, the judge permitted the plaintiffs to use the first and third columns in the cross-examination of defense experts but not the second column. The judge’s reasoning on this point is not apparent despite our examination of the transcript of extensive side bar discussions.
We think it reasonable to infer from his decision to permit use of two-thirds of the chart that the judge considered the text Williams on Obstetrics to be “a reliable authority.” Had it been otherwise, the judge should have precluded its use altogether. There was basis in the record for considering the treatise reliable, including Dr. O’Toole’s admitted reliance upon it from time to time and the consistent testimony of Dr. O’Toole, his *292obstetrical expert, and his pediatric neurology expert that they had each used the obstetrics treatise in medical school. The plaintiffs should have been permitted to use the second column of the chart during cross-examination.
We turn then to the more difficult question of whether this exclusion of evidence was reversible error. To be sure, the experts differed on whether Dr. O’Toole had followed the standard of care when he delivered Kaileigh at 9:45 p.m., rather than earlier in the day, upon ascertaining the low fetal biophysical profile score. They differ on what the biophysical profile score means and what it requires of a doctor. The experts also differ on what caused Kaileigh’s injuries, the plaintiffs contending that it was brain asphyxia occurring after the biophysical profile testing, with the defendant in contrast maintaining that the embolic stroke or strokes causing all of the damage occurred before the biophysical profile was performed.3 The plaintiffs’ theory of causation suggests the efficacy of early delivery while the defendant’s theory holds that the time of delivery is without significance.
“[T]he erroneous exclusion of relevant evidence is reversible error unless, on the record, the appellate court can say with substantial confidence that the error would not have made a material difference.” DeJesus v. Yogel, 404 Mass. 44, 48-49 (1989). In applying this standard, we are to view the record in a commonsense way, id. at 48, and we look to see if the excluded evidence was noncumulative and “bore directly on one of the essential issues in the plaintiffs’ case.” Sacco v. Roupenian, 409 Mass. 25, 31 (1990). On balance, we cannot say with substantial confidence on the record before us that the error would not have made a material difference.
The defendant’s obstetrical expert, for example, testified on direct examination that a biophysical profile of two out of eight did not necessarily indicate acute fetal distress; instead, it required the obstetrician only to evaluate the baby further, to bring the mother to the labor and delivery setting, but not to *293deliver immediately. He testified that Dr. O’Toole complied with the standard of care and that there was no indication of asphyxia or a worrisome deprivation of oxygen to the baby before delivery. The evidentiary exclusion prevented the plaintiffs from cross-examining the obstetrician with a recognized obstetrical authority’s quite different interpretation of the biophysical profile score of two, viz., almost certain fetal asphyxia, where the recommended patient management is to “Deliver.” We observe that, heard together in context, “Deliver,” when “Almost certain fetal asphyxia” is indicated by the test score, can connote an immediacy that is absent when the interpretation of the score is omitted, as happened at trial. Although the jury earlier heard testimony substantively consistent with, but not identical to, the precluded evidence from the plaintiffs’ own obstetrical expert, the impact upon the jurors of hearing this alternative interpretation from an authority who was not a “hired gun” is something else again. The jury might well have concluded that what in fact caused the low biophysical score could only be known by a doctor in hindsight and that the prudent course was to deliver the baby right, away. The excluded evidence accordingly bore directly on at least one of the essential issues in the plaintiffs’ case: Dr. O’Toole’s alleged deviation from the standard of care. It may also have borne upon the issue of causation.4 The jury were presented with real alternatives as to what caused the strokes: asphyxia or *294an embolus. Hearing the defense obstetrician, for example, respond to a contrary point of view stated in a treatise he used in medical school might well have made a material difference. Much the same can be said about the plaintiffs’ inability to cross-examine the defendant’s pediatric neurologist. There must be a new trial.
Under G. L. c. 233, § 79C, as amended by St. 1965, c. 425, “[statements of fact or opinions on a subject of science or art contained in a published treatise . . . [or] book . . . shall, in so far as the court shall find the said statements are relevant and that the writer of such statements is recognized in his profession or calling as an expert on the subject, be admissible in [medical malpractice] actions ... as evidence tending to prove said facts or as opinion evidence,” provided that particularized information designated in the statute is given in advance to the adverse party.
It appears that the text Williams on Obstetrics has appeared in many editions and is at least in part a compilation of chapters written by different authors. The authors of the section at issue here are identified as “Manning and Colleagues,” and the edition of the text that the plaintiffs attempted to introduce was dated 1993. Kaileigh was born in 1990, and Dr. O’Toole acknowledged that he had a 1990 edition in his office which he consulted from time to time. He also agreed that the text is one used in medical schools to teach obstetrics.
It was the plaintiffs’ burden to lay an adequate foundation for the introduction of the proffered evidence under G. L. c. 233, § 79C. The plaintiffs draw our attention repeatedly (as they had the trial judge’s) to the adequacy of the notice given to the adverse party and to the relevance of the proffered chart. They fail to recognize, however, that this is not all that the statute requires. While the plaintiffs point to Dr. O’Toole’s possession and occasional use of the textbook and to the use of the text in medical schools to teach obstetrics, this does not establish that the author(s) of the chart, “Manning and Colleagues,” are “recognized in [their] profession or calling as . . . expert[s] on the subject.” The plaintiffs never fully identified the authors of the chart, never inquired even of their own experts about the *295authors’ professional qualifications, reputations, or expertise in the field, and, for that matter, never established that the chart appearing in the 1993 edition was the same as what had appeared in the 1990 edition. In these circumstances, the judge was not provided with a basis for finding that “the writer of a relevant statement in the text has the required professional standing.” Reddington v. Clayman, 334 Mass. 244, 247 (1956). The plaintiffs did not satisfy the foundational requirements of G. L. c. 233, § 79C, and the judge did not abuse his discretion in refusing to admit the chart in evidence.
Because the admissibility of the pertinent written policies and procedures of the Brockton Hospital may arise at any retrial, we address it here. During the direct examination of the labor and delivery nurse, the plaintiffs attempted to introduce in evidence written hospital guidelines related to certain Apgar test results and procedures for emergency cesarean sections. They complain that the guidelines were not offered tó prove the truth of the statements contained therein but to explain the significance of the Apgar score in understanding the status of the fetus and the difference at Brockton Hospital between an unscheduled cesarean section and a tme medical emergency cesarean section. The plaintiffs were permitted to use the documents to refresh the nurse’s recollection, but her recollection was not thereby refreshed. There was no error in the judge’s exclusion of this evidence. It contained inadmissible hearsay and did not fall under any exception to the hearsay rale, including G. L c. 233, §§ 78 and 79G.
Judgment reversed.
Testing for fetal heart rate, which would have increased the possible score to a maximum of ten, appears not to have been performed.
The experts differ, for example, as to the significance of the thick green meconium found by Dr. O’Toole at 7:30 p.m. The plaintiffs’ experts take the color and texture as indicative of recent fetal stress, consistent with their chronology of events and theory of causation, while the defendant’s experts are of the view that the color and texture of the meconium, taken with placental staining, are consistent with fetal stress occurring prior to the performance of the biophysical profile test.
The defendant contends that any evidentiary error was harmless. His first position is that the excluded evidence related only to causation and that the jury never had to reach this issue because they concluded that Dr. O’Toole had complied with the standard of care. We reject this analysis for the reasons stated above. The defendant’s second position is that, even if the excluded evidence also bore on the standard of care issue, the error was harmless since the jury, were they to have reached the causation issue after concluding there was a deviation from the standard of care, would also have found an absence of causation. As support for this, the defendant calls our attention to a written communication from the jurors during deliberations to the judge. The jurors sought his guidance because, while they had not reached a verdict on the first jury question pertaining to negligence, twelve out of fourteen deliberating jurors had concluded, as to the second jury question, that there was no causal connection between the defendant’s actions and Kaileigh’s injuries. The judge sealed their question and instructed them to continue to deliberate. They did so and returned a verdict after concluding that Dr. O’Toole had not deviated from the standard of care. The defendant does not call our attention to any appellate authority addressing the significance of such a mishap during deliberations to the determination of whether an evidentiary error may have made a material difference to the case. We think the jury’s communication, in the *294context of deliberations taking place absent the disputed evidence, is without terribly much significance. Because the excluded evidence here bore to some extent on both issues, we cannot be confident as to what the jury would have done on either issue had they heard the evidence.