Richard A. Knox, a Boston Globe medical reporter, and the Globe Newspaper Company, publisher of The Boston Globe, collectively “the Globe defendants,” appeal from a judgment of civil contempt for violating a Superior *385Court order to disclose Knox’s confidential sources regarding a series of articles published in The Boston Globe (Globe). The series concerned accidental chemotherapy overdoses administered to Globe columnist Betsy Lehman and to another patient, Maureen Bateman. We vacate both the order of February 23, 1998, compelling the disclosure of confidential sources and denying the Globe defendants’ motion for a protective order, and the order of August 13, 1998, imposing sanctions for civil contempt, and remand for further proceedings in the Superior Court.
The events underlying the Globe series were as follows. Lehman and Bateman were patients on an experimental breast cancer treatment protocol (protocol) at the Dana-Farber Cancer Institute (Dana-Farber). The protocol involved the administration of high doses of the drug cyclophosphamide. The plaintiff, Dr. Lois Ayash, a physician at Dana-Farber, was protocol chair and principal investigator for the protocol. In mid-November, 1994, a physician on Lehman and Bateman’s clinical treatment team erroneously administered a fourfold overdose of the drug. Lehman died on December 3, 1994, as a result of the error.2 Bateman suffered severe toxicity, but survived.3
It was not until February, 1995, that a data manager at DanaFarber discovered that the overdoses had been administered. Dana-Farber then suspended the clinical privileges of two physicians who had been directly involved in the patients’ care when the overdoses were given. No action was taken against Ayash at that time. On March 31, 1995, Ayash’s clinical privileges were suspended and she was assigned to administrative duty. She became the subject of two Dana-Farber internal investigations and an investigation by the Massachusetts Board of Registration in Medicine (board of registration).
The Globe published the first of Knox’s articles concerning these events on March 23, 1995. In that initial article, Knox made two statements which Ayash contends are defamatory. First, he erroneously identified her as one of the doctors who had countersigned the mistaken order. Second, he identified her as the “leader of the team,” which, Ayash contends, wrongly *386implied that she had direct patient care responsibilities for Lehman and Bateman when the overdoses were administered. On February 1, 1996, Ayash brought suit against the Globe defendants for libel and for breach of confidentiality and privacy4; against Knox; against Dana-Farber; and against David M. Livingston, M.D.5
Knox concedes that he was told within four days of publication that Ayash did not countersign the overdose orders.6 Ayash energetically sought, and Knox refused to disclose, the identity of the source who had informed him of his mistake. That source has now been identified as Karen Antman, a former physician at Dana-Farber. The disclosure occurred on August 18, 1998, after the Superior Court judge had issued the orders under appeal.7 Knox no longer seeks to maintain confidentiality with regard to Antman.
According to her brief on appeal, Ayash continues to seek the identities of Knox’s sources who disclosed the following information which, she asserts, should have been kept confidential: (1) the information that Ayash was the subjéct of a DanaFarber internal “corrective action” investigation; (2) the information that Ayash was the subject of an investigation by the board of registration; and (3) documents provided to Knox concerning recommendations made by Dana-Farber’s internal peer review committee and Dana-Farber’s responses.8
Upon Knox’s refusal during discovery to provide any *387information which would lead to the disclosure of his confidential sources’ identities, Ayash moved to compel disclosure. The Globe defendants moved for a protective order. On February 23, 1998, the Superior Court judge granted Ayash’s motion and denied the Globe defendants’ motions.
A single justice of this court denied the Globe defendants’ petition, pursuant to G. L. c. 231, § 118, first par., seeking reversal of or leave to take an interlocutory appeal from the Superior Court judge’s orders.9 As Knox persevered in his refusal to reveal his confidential sources, the Superior Court judge issued a contempt order and imposed monetary sanctions on Knox and the Globe on August 13, 1998. Knox was ordered to pay a fine of one hundred dollars per day beginning August 24, 1998, the amount to escalate by one hundred dollars per day each successive week that he failed to comply with the discovery order. The Globe was fined one thousand dollars per day beginning August 24, 1998, the amount to escalate by one thousand dollars per day each successive week that it refused to comply with the discovery order. A single justice of this court ordered a stay of execution on the contempt judgment pending an expedited appeal.
On August 28, 1998, the Globe defendants moved in Superior Court for reconsideration of the discovery and sanction orders as a result of the August 18, 1998, disclosure of Karen Ant-man’s identity. The record before us indicates that the motion is pending. We will nevertheless proceed with our review in the interests of expediting this appeal.
“The validity of an underlying discovery order, disobedience of which has led to an adjudication of contempt, may be challenged on appeal from the adjudication of contempt.” Matter of Roche, 381 Mass. 624, 625 n.l (1980). We now review the rulings of the Superior Court judge denying a protective order and granting the motion to compel in order to determine whether they constituted an abuse of discretion. See Sinnott v. Boston *388Retirement Bd., 402 Mass. 581, 585-586, cert, denied, 488 U.S. 980 (1988).
No constitutional or statutory privilege or rule of court protects news reporters’ sources of information. As the Superior Court judge stated in his memorandum of decision ordering disclosure of Knox’s sources, there is no constitutionally based privilege protecting news reporters’ sources of information. Matter of a John Doe Grand Jury Investigation, 410 Mass. 596, 598 (1991), and cases cited. Nor is there any statutory privilege or rule of court providing for such a privilege. Id. Nevertheless, “[i]t is well settled that, in supervising discovery, a presiding judge is ‘obliged to consider the effect that compelled discovery would have on “the values protected by the First Amendment, [even] though [these values were] entitled to no constitutional privilege.” ’ Matter of Roche, supra at 636, quoting Herbert v. Lando, 441 U.S. 153, 180 (1979) (Powell, J., concurring). The ‘needless disclosure of confidential relationships,’ Matter of Roche, supra at 637, is, therefore, to be avoided.” Sinnott v. Boston Retirement Bd., 402 Mass, at 586.
Developing common-law protection for news reporters’ confidential sources of information. In 1985 the Supreme Judicial Court refused to adopt rules which would provide for an evidentiary privilege for news reporters’ sources of information, advocating instead a common law approach as “more likely to result in principles that are flexible enough to maintain an appropriate balance between the competing interests involved.” Petition for the Promulgation of Rules Regarding the Protection of Confidential News Sources & Other Unpublished Information (hereafter Petition for Promulgation of Rules), 395 Mass. 164, 171 (1985). The court there stated that under common-law principles already developed, a party seeking to avoid disclosure must first “make some showing that the asserted damage to the free flow of information is more than speculative or theoretical.” Id. at 1.72, citing Matter of Roche, 381 Mass, at 635, and Matter of Pappas, 358 Mass. 604, 612 (1971) , aff’d sub nom. Branzburg v. Hayes, 408 U.S. 665 (1972) . Only then will the court go on to balance “the public interest in every person’s evidence and the public interest in protecting the free flow of information,” in order to determine whether disclosure should be prevented. 395 Mass, at 172, citing Commonwealth v. Corsetti, 387 Mass. 1, 5-6 (1982).
The evidence required to make “some showing” of a risk to *389the free flow of information. The Supreme Judicial Court has not discussed in detail the nature and extent of evidence required to make “some showing” that disclosure poses a risk to the free flow of information. In Petition for Promulgation of Rules, it stated that there is no privilege for information that has already been published. 395 Mass, at 172, citing Commonwealth v. Corsetti, 387 Mass, at 4-5. In an earlier case, the court intimated that the bald allegation that disclosure of a news reporter’s confidential sources will impede the free flow of information would not be enough to satisfy this burden. See Matter of Pap-pas, 358 Mass, at 612.
The two most recent Supreme Judicial Court cases concerning this matter, Sinnott v. Boston Retirement Bd., 402 Mass. 581 (1988), and Matter of a John Doe Grand Jury Investigation, 410 Mass. 596 (1991), affirmed the Superior Court’s grant of investigative news reporters’ motions to quash subpoenas for disclosure of confidential sources,10 and touched on the evidentiary requirements for meeting the test of some showing that disclosure poses a risk to the free flow of information (the “threshold showing”). In neither case did the court appear to require a burdensome showing.
In Sinnott v. Boston Retirement Bd., supra, the court commented only indirectly on this issue. Affirming the denial of a motion to compel discovery of a Globe reporter’s sources for an article concerning the Boston retirement board’s alleged abuses in awarding pensions, 402 Mass, at 582-587, it concluded that the Superior Court judge was warranted in applying the balancing test. There was no discussion of the nature of the threshold showing. Id. at 586. In its discussion of the balancing test, however, the Supreme Judicial Court appears to have embraced the view that forced disclosure of investigative news reporters’ confidential sources of information carries an inherent threat to *390the free flow of information. “[T]he public interest in nondisclosure rest[s] upon the concern that ‘the deterrent effect such disclosure is likely to have upon future “undercover” investigative reporting . . . threatens freedom of the press and the public’s need to be informed.’ ” Id. at 587, quoting from Von Bulow v. Von Bulow, 811 F.2d 136, 142-143 (2d Cir.), cert, denied sub nom. Reynolds v. Von Bulow, 481 U.S. 1015 (1987).
Three years later, in Matter of a John Doe Grand Jury Investigation, 410 Mass. 596 (1991), the court considered the nature of a threshold showing. Doe affirmed a Superior Court judge’s allowance of motions to quash grand jury subpoenas issued to two news reporters for the purpose of discovering the identities of their confidential sources regarding a murder investigation. Id. at 596-597. The motion judge concluded that the reporters had met the threshold test. The court thought the judge’s determination warranted, based on his ultimate finding “that the reporters would not have received the information they obtained if they had not promised anonymity to their sources and that the reporters’ future news-gathering ability, both generally and in the case being investigated by the grand jury, would be impaired if they violated their promises.” Id. at 600. The court did not comment further on the evidence required to meet the threshold showing. Its brief treatment of the issue again may signal the view that the court does not consider the initial burden to be heavy.11
The defendants’ initial burden. In the case before us, the Superior Court judge concluded that the defendants had not shown that the disclosure of Knox’s confidential sources, would pose more than a speculative or theoretical risk to the free flow of information. Our reading of the record is that certain of the subsidiary findings on which the judge based his ultimate one are not supported by the evidence.
The Superior Court judge found that Knox’s informants do not continue to provide him with information, and that Knox was not involved in a continuing investigation. The extent and *391significance of Knox’s continuing relation with his sources must be viewed in the context of his long-term relationships with medical experts who provide him information when called upon. Knox testified in his deposition of June 11, 1996, that during his twenty-seven years as a medical reporter he has cultivated numerous sources in the medical community and has established continuing relationships with experts upon whom he relies for information in specialized areas. Among these sources are individuals who have agreed to provide information only upon the promise of confidentiality. If Knox were unable to keep such promises, his future ability to gather and publish information from these and other sources who wish to remain anonymous would suffer. See Matter of a John Doe Grand Jury Investigation, supra at 600.
With regard specifically to Knox’s confidential sources for the series of articles at issue, Knox testified in a deposition on May 7, 1998, after the issuance of the discovery order but before the order for contempt and sanctions was issued, that two of these confidential sources had subsequently provided him with information concerning other matters.12
We do not question the trial judge’s determination that Knox was not actively engaged in a continuing investigation at the time of trial. However, the record reflects that the board of registration’s proceedings concerning the overdose incident were ongoing at least as late as September, 1997; there is no indication in the record that they have been concluded. As these proceedings unfold, Knox will in all likelihood report on further developments and may well turn to his earlier confidential sources for information.
The judge further found that the information was given to Knox over two and one-half years previously and concerned an “isolated incident.” The characterization of the chemotherapy overdosing as an “isolated incident” does not comport with the facts on the record. A report issued by Dana-Farber documenting the results of two committee investigations into the matter13 indicates that the overdosing had long-term repercussions and *392was of widespread public importance.14 The investigators found . structural deficiencies in Dana-Farber’s requirements for research protocols, in its clinical procedures, in its pharmacy’s procedures and technology, and in its quality assurance program. According to the report, Dana-Farber took extensive corrective actions as a result of the investigators’ recommendations, and the report itself was being disseminated both to clarify the situation at Dana-Farber, and to “provid[e] other cancer care providers with insights that may benefit patients and institutions throughout the country.” In order to follow up on the corrective measures taken by Dana-Farber or to report on ways other institutions may have been affected, Knox could seek information from the confidential sources he used earlier.15 Compare Sinnoit v. Boston Retirement Bd., 402 Mass, at 587 n.5.
Like the reporters in Matter of a John Doe Grand Jury Investigation, the Globe defendants have met the threshold showing by demonstrating that “the reporter!] would not have received the information [he] obtained if [he] had not promised anonymity to [his] sources and ... the reporteras] future news-gathering ability, both generally and in the case being investigated . . . would be impaired if [he] violated [his] promises.” 410 Mass, at 600.
The balancing test concerning the libel claim against the Globe defendants. Because the Globe defendants have made “some showing” that disclosure of Knox’s confidential sources presents a danger to the free flow of information that is more than speculative or theoretical, Petition for Promulgation of Rules, 395 Mass, at 172, that danger must be balanced against the public’s need for evidence. The Superior Court judge did, in fact, balance these factors despite having found that the defendants had not met their initial burden. His specific findings, however, appear to have been limited to the libel count against the Globe defendants.
With regard to this claim, Ayash’s arguments before the *393Superior Court judge focused on her need for disclosure of the source who had informed Knox of errors in his March 23 article. Access to this source, she contended, would permit her to assess whether Knox could have obtained the correct information before publication. This source has now been revealed as Karen Animan. Knox no longer seeks to protect her confidentiality, and on appeal Ayash makes no request for further compelled disclosure regarding Antman. Additionally, Knox has not claimed confidentiality for any of his sources for the two statements in the March 23 article which Ayash contends defamed her.16 In light of these facts, the Superior Court judge must determine upon remand whether, as to the libel claim against the Globe defendants, Knox’s disclosure of additional confidential sources would achieve more than the “needless disclosure of confidential relationships.” Petition for Promulgation of Rules, 395 Mass, at 172, and cases cited.
Disclosure of sources concerning the plaintiff’s other claims. Ayash argues that other counts of her complaint require the disclosure of additional confidential sources. These contentions may have merit. However, “the judge’s explanation of his decision on the record [is] a necessary aid to proper appellate review in this field.” Id. at 173, and cases cited. Thus, the Superior Court judge must perform the required balancing test and make findings as to whether disclosure is required with regard to the other counts of Ayash’s complaint.
Conclusion. The orders of February 23 and August 13, 1998, are vacated. The case is remanded for findings and conclusions consistent with this opinion.
So ordered.
Ayash began a rotation as the attending physician for Lehman and Bate-man on December 1, 1994. At that time, Lehman and Bateman had already suffered adverse reactions to the cyclophosphamide treatment, but Ayash did not discover that the overdoses had occurred.
According to the plaintiff, Bateman later succumbed to her breast cancer.
The Globe defendants were granted partial summary judgment on Ayash’s invasion of privacy claim.
Counts against Knox are for intentional interference with advantageous business relations, and intentional and/or negligent infliction of emotional distress.
Counts against Dana-Farber are for gender discrimination, breach of contract, breach of the implied covenant of good faith and fair dealing, breach of confidentiality and privacy, and defamation.
Counts against Livingston, who was physician-in-chief at Dana-Farber during 1994 and through the spring of 1995, are for defamation and intentional interference with advantageous business relations.
The Globe published a correction on June 4, 1995.
Ayash herself named Antman during a deposition conducted by DanaFarber. She stated that Antman, a friend of hers, had told her that she telephoned Knox “shortly after the first article came out from The Globe and told him essentially that he had made a major mistake in what he had reported and in terms of what a [principal investigator] is responsible for.”
The documents are titled, “Synopsis of the Conclusions and Recommendations of the Internal Peer Review Committee” (dated May 23, 1995), and *387“Institute Actions in Response to Synopsis of the Conclusions and Recommendations of the Internal Peer Review Committee of the Board of Tmstees of the Dana-Farber Cancer Institute.”
The single justice’s order was without prejudice to the Globe defendants submitting a motion to the Superior Court for a more narrowly tailored protective order. The proposal they had submitted would have prohibited all discovery Ayash sought to be compelled. The Globe defendants refused to submit a more narrowly tailored protective order.
Earlier cases affirmed orders to disclose. See Matter of Pappas, supra (a motion to quash a grand jury summons to a reporter to testify about a visit to Black Panther headquarters was denied); Dow Jones & Co. v. Superior Ct., 364 Mass. 317 (1973) (a reporter was obligated to identify the source of allegedly defamatory statements quoted in a news article); Matter of Roche, 381 Mass. 624 (1980) (a reporter was ordered to testify at a deposition in a judicial disciplinary action concerning sources of information for a television investigative report on District Court judges); Commonwealth v. Corsetti, 387 Mass. 1 (1982) (a reporter was ordered to testify at a hearing on a motion to suppress evidence in an impending criminal trial regarding a telephone conversation he had had with the defendant).
The court’s approach to the threshold test cannot be ascribed simply to adherence to a deferential standard of review (see Sinnott v. Boston Retirement Bd., 402 Mass, at 586), since it declined to defer to the Superior Court judge when it reviewed his conduct of the balancing test. It disagreed with the judge’s analysis of the evidence in that context and supplanted his balancing test with its own. Matter of a John Doe Grand Jury Investigation, 410 Mass, at 601.
Knox testified that one of these sources had provided him with information during the spring of 1998, and the other, the person who had given him the May 23, 1995, synopsis of Dana-Farber’s internal investigation, had provided information in late 1996 or 1997.
The document, dated October 30, 1995, titled, “A Special Report from the Board of Trustees and Administration of the Dana-Farber Cancer Institute *392to its Various Constituencies,” presents the results of investigations by an internal peer review committee and an external peer review committee.
In dismissing Ayash’s privacy claims against the Globe defendants, the Superior Court judge found that Knox’s reporting in this case concerned events of “public concern” about health care issues.
The judge also found that the information sought was not cumulative or otherwise available. This finding is not relevant to the threshold showing, but rather to the balancing test required after the party seeking to avoid disclosure has met the threshold showing.
In his deposition of June 11, 1996, as well as in his answers to interrogatories, Knox responded to questions concerning both statements. He testified that he mistakenly identified Ayash as a cosignatory on the overdose order because he misread a signature on the medical record as Ayash’s, then consulted with the lawyer for Ms. Lehman’s estate, who also thought that the signature was Ayash’s. He stated that he had characterized Ayash as team leader because she was principal investigator of the protocol, because she was listed on some documents as Lehman’s primary physician, and because of references made by others, whom Knox named, regarding Ayash’s role.