UNITED STATES DISTRICT COURT WESTERN DISTRICT OF NEW YORK _______________________________________
JASON BUHRMAN, M.D., Ph.D., DECISION AND ORDER Plaintiff, 24-CV-6029 FPG CDH v.
UNIVERSITY OF ROCHESTER,
Defendant _______________________________________
INTRODUCTION Plaintiff Jason Buhrman, M.D., Ph.D. (“Plaintiff”), claims that defendant the University of Rochester (“Defendant” or the “University”) discriminated against him due to a medical disability and retaliated against him for seeking reasonable accommodations and taking approved medical leave. (See Dkt. 1). Defendant denies these claims. (See Dkt. 10). Presently before the Court is Plaintiff’s motion to compel. (Dkt. 35). For the reasons that follow, Plaintiff’s motion to compel is granted in part and denied in part. FACTUAL BACKGROUND I. The Program In 2017, Plaintiff began a program at the University known as the American Board of Internal Medicine Research Pathway (the “Program”). (Dkt. 41-1 at ¶ 4). Plaintiff’s enrollment, employment, and participation in the Program were pursuant to a contract with the University that was renewed on a year-by-year basis. (Dkt. 1 at ¶¶ 25-26). The Program combines traditional medical residency training with fellowship research. (Dkt. 1 at ¶ 24). For Plaintiff, the residency training portion of the Program was in general internal medicine (“IM”), and the fellowship research portion of the
Program was in his chosen specialty of hematology and oncology (“HemeOnc”). (Id.; Dkt. 41-1 at ¶¶ 5,7). Whereas traditionally a trainee1 would spend three years in an IM residency, then sit for the general IM boards before moving onto a fellowship, the Program “essentially compress[es] the [IM] residency portion into two years.” (Dkt. 41-1 at ¶ 6). The parties present slightly different explanations of Plaintiff’s transition from the residency portion of the Program to the fellowship portion. According to Plaintiff,
the first two years of the Program consisted of IM residency, the third year was a combined IM residency and HemeOnc fellowship clinical training, and the fourth through seventh years consisted of the HemeOnc fellowship clinical and research training. (Dkt. 35-2 at 5; Dkt. 42 at 8). Defendant, however, states that Plaintiff was required to complete two years of IM residency, followed by 18 months of clinical training in HemeOnc and three years of research training—though, to become board
eligible in IM, Plaintiff was required to complete both the two years of IM residency and the 18 months of clinical training in the HemeOnc fellowship. (Dkt. 41-1 at ¶ 7). In other words, there is a factual dispute about whether Plaintiff’s third year in the Program was a combination of the final year of his IM residency and the first year of his HemeOnc fellowship, as Plaintiff asserts, or whether Plaintiff’s third year was
1 The Court uses the term “trainee” to refer collectively to residents and fellows. strictly the first year of his HemeOnc fellowship, albeit a component of becoming board eligible for IM, as Defendant represents. II. Trainee Evaluations
Each University department has its own Clinical Competency Committee (“CCC”) that conducts bi-annual reviews of its residents and fellows based on the department’s own training milestones and evaluation standards, which are established by the Accreditation Council for Graduate Medical Education (“ACGME”). (Dkt. 35-2 at 5; Dkt. 41-1 at ¶¶ 13-14). For each bi-annual review, the trainee receives two documents: (1) a CCC action plan and (2) an ACGME milestone evaluation. (Dkt. 35-2 at 5; Dkt. 41-1 at ¶ 16). For the ACGME milestone evaluation,
there are four possible ratings: Superior, Satisfactory, Conditional on Improvement, and Unsatisfactory. (Dkt. 41-1 at ¶ 23). Conditional on Improvement means the trainee “meets some developmental milestones but occasionally falls short of the expected level of development for this year of training.” (Id.). III. Plaintiff’s Performance in the Program According to Defendant, Plaintiff completed “two years of the IM residency
training component of the Program” and “[h]is record reveals no performance deficiencies.” (Id. at ¶ 19). However, Plaintiff then received a Conditional on Improvement rating on his year-end evaluation for his third year in the Program (2019-20). (Id. at ¶ 22; Dkt. 41-2). Plaintiff states that at the end of his third year, the University awarded him his “Certificate of Completion of the Internal Medicine Residency Training Program.” (Dkt. 1 at ¶ 32). During his fourth year in the Program (2020-21), Plaintiff was suspended from clinical responsibilities due to what Defendant describes as “his struggles with patient care and serious concerns about his ability to safely treat patients.” (Dkt. 41-
1 at ¶ 25). Plaintiff again received a Conditional on Improvement rating on his fourth- year year-end evaluation. (Id. at ¶ 26; Dkt. 41-5). At the start of his fifth year in the Program (2021-22), Plaintiff, after previously experiencing symptoms related to what was eventually diagnosed as liver failure, suffered a grand mal seizure and “was placed on medical, disability, and FMLA leave.” (Dtk. 35-2 at 5). Plaintiff returned to the Program in January of 2022, when he again received a Conditional on Improvement rating on his fifth-year mid-
year evaluation. (Dkt. 41-1 at ¶¶ 28-29; see Dkt. 41-6). Plaintiff received a liver transplant in May of 2022. (Dkt. 1 at ¶ 59). At the end of his fifth year in the Program, the University decided not to offer Plaintiff a contract for his sixth year. (Dkt. 41-1 at ¶ 34). PROCEDURAL BACKGROUND This case has been referred to the undersigned for all non-dispositive pretrial
matters. (Dkt. 13; Dkt. 22). Plaintiff commenced this action on January 11, 2024, asserting federal and state law claims of discrimination and retaliation based on medical disability, as well as breach of contract. (Dkt. 1). Defendant filed an answer, asserting as its sixth affirmative defense that “[a]ny employment action taken by the University toward Plaintiff was taken for legitimate, non-discriminatory reasons,” and asserting as its seventh affirmative defense that “[a]t no time did the University take any action toward Plaintiff that it would not have taken toward similarly situated individuals regardless of disability.” (Dkt. 10 at ¶¶ 148-49). Plaintiff states that “[b]ased upon
the University’s Answer, written discovery responses, and deposition testimony, it has become clear that the University’s affirmative defenses are based on its contention that Dr. Buhrman was not satisfactorily performing in the Program.” (Dkt. 35-2 at 6). During the course of discovery, Plaintiff requested records related to other trainees at the University. (Dkt. 35-1 at ¶ 10). Defendant largely objected to these requests for production (“RFPs”) as overbroad, unduly burdensome, not relevant, and
disproportionate to the needs of the case, and refused to produce the requested documents. (Dkt. 35-1 at ¶ 11; Dkt. 35-11; Dkt. 35-12). For some documents that were produced, including HemeOnc CCC meeting notes, Defendant made extensive redactions, claiming that the redacted information was irrelevant or non-responsive. (Dkt. 35-1 at ¶ 12; Dkt. 35-13 at 4; Dkt. 35-14). After exchanging further correspondence with Defendant, Plaintiff submitted
a letter to the Court advising it about the discovery dispute and requesting a conference. (Dkt. 35-19). Defendant submitted a responsive letter. (Dkt. 35-20). The Court held a telephonic discovery conference with the parties on December 22, 2025. (Dkt. 30). During the conference, the Court and the parties discussed narrowing Plaintiff’s comparator requests (Dkt. 35-1 at ¶ 17) and the Court instructed the parties to continue to confer (Dkt. 30). Plaintiff subsequently narrowed his requests to four categories of documents related to the evaluations and personnel files of IM residents and HemeOnc fellows over a 7–10-year period. (Dkt. 35-21). Defendant agreed to produce one of the four
categories of documents for a five-year period with redactions and on an attorneys’- eyes-only (“AEO”) basis, but refused to produce documents in response to the other three categories, again objecting on grounds of burden and relevance. (Dkt. 35-22). Plaintiff rejected Defendant’s proposal to produce the redacted documents on an AEO basis and made his own proposals to narrow the scope of the three other categories of documents. (Dkt. 35-23). Eventually, the parties reached an impasse. Plaintiff filed the instant motion to compel on February 19, 2026. (Dkt. 35).
Plaintiff seeks an order compelling Defendant to produce: (1) Hematology/Oncology fellow evaluations, consisting of the ACGME milestones and CCC Action plans, for a period of 5 years (academic years beginning 2018-2023);
(2) Internal medicine resident evaluations, consisting of the ACGME milestones and CCC Action plans, for a period of 4 years (academic years beginning 2016-2020);
(3) Personnel files of Hematology/Oncology fellows and internal medicine residents who received a “conditional upon improvement” rating on his/her ACGME milestones, for a period of 5 years (academic years beginning 2018-2023); and
(4) Personnel files of Hematology/Oncology fellows and internal medicine residents that were placed on probation, suspension, and/or had their contract non-renewed, for a period of 5 years (academic years beginning 2018-2023).
(referred to herein as “RFP 1,” “RFP 2,” “RFP 3,” and “RFP 4,” respectively). (Dkt. 35- 2 at 4). Plaintiff also asks the Court to compel Defendant to “reproduce the unredacted versions of all documents where redactions were made solely on the basis of relevancy.” (Dkt. 35-2 at 17). Defendant filed its opposition to Plaintiff’s motion to compel on March 27,
2026. (Dkt. 41). Defendant argues that: (1) other IM residents are not appropriate comparators for Plaintiff and therefore their evaluations and personnel files are not relevant; (2) other HemeOnc fellows are also not appropriate comparators unless they are/were in the Program because the Program involves an additional research component; (3) RFPs 1 and 2 are overbroad insofar as they include evaluations for trainees who had no history of performance deficiencies; (4) Plaintiff’s requests, as a whole, are unduly burdensome and disproportionate to the needs of the case because
they would require searching for and reviewing a significant number of evaluations and personnel files that are not maintained in a single location or uniform manner; and (5) production of the requested documents would also be burdensome because the University would have to comply with the Family Educational Rights and Privacy Act (“FERPA”) and redact all personally identifying information (“PII”) and provide notice to the trainees whose records are produced. (Id. at 15-24). Plaintiff filed a reply
on April 3, 2026. (Dkt. 42). DISCUSSION I. Legal Standard on Motion to Compel Federal Rule of Civil Procedure 26(b)(1) provides that “[p]arties may obtain discovery regarding any nonprivileged matter that is relevant to any party’s claim or defense and proportional to the needs of the case[.]” Fed. R. Civ. P. 26(b)(1). Relevance is “construed broadly to encompass any matter that bears on, or that reasonably could lead to other matter that could bear on any issue that is or may be in the case.” Albino v Glob. Equip. USA, Ltd., No. 6:14-CV-06519(MAT), 2017 WL
3130380, at *1 (W.D.N.Y. July 24, 2017) (quoting Oppenheimer Fund, Inc. v. Sanders, 437 U.S. 340, 351 (1978)). “The burden of demonstrating relevance is on the party seeking discovery . . . Once relevance has been shown, it is up to the responding party to justify curtailing discovery.” Go New York Tours, Inc. v. Aurora Tourism Servs., LLC, No. 22-CV-10633 (DEH) (JW), 2023 WL 9111158, at *1 (S.D.N.Y. Dec. 20, 2023) (quotation omitted). The Court “has broad discretion in determining relevance for discovery purposes.” Michael Kors, L.L.C. v. Su Yan Ye, No. 1:18-CV-2684 (KHP),
2019 WL 1517552, at *2 (S.D.N.Y. Apr. 8, 2019). “Proportionality focuses on the marginal utility of the discovery sought. . . . [T]he greater the relevance of the information in issue, the less likely its discovery will be found to be disproportionate.” Vaigasi v. Solow Mgmt. Corp., No. 11-CIV-5088- RMB-HBP, 2016 WL 616386, at *14 (S.D.N.Y. Feb. 16, 2016) (citations omitted). When determining whether discovery is proportional to the needs of the case, the
Court considers “the importance of the issues at stake in the action, the amount in controversy, the parties’ relative access to relevant information, the parties’ resources, the importance of the discovery in resolving the issues, and whether the burden or expense of the proposed discovery outweighs its likely benefit.” Fed. R. Civ. P. 26(b)(1). Rule 34 provides that “[a] party may serve on any other party a request to produce documents within the scope of Rule 26(b).” Fed. R. Civ. P. 34(a). If a party fails to produce documents requested pursuant to Rule 34, then the party seeking
discovery may move for an order to compel production. Fed. R. Civ. P. 37(a)(3)(B)(iv). “Motions to compel, pursuant to Rule 37, are left to the sound discretion of the court.” Syntel Sterling Best Shores Mauritius Ltd. v. TriZetto Grp., Inc., 328 F.R.D. 450, 452 (S.D.N.Y. 2018) (quotation omitted). II. FERPA The Court first addresses the parties’ disagreement over the extent to which FERPA impacts this discovery dispute. “FERPA conditions the receipt of certain
federal funds by universities . . . on the limitation of circumstances under which those universities may release educational records.” Ware v. Univ. of Vermont & State Agric. Coll., No. 2:22-CV-212, 2025 WL 2304724, at *2 (D. Vt. Aug. 11, 2025); see U.S.C. § 1232g(b). Generally, “FERPA prohibits ‘the release of education records’ or ‘personally identifiable information contained therein other than directory information’ concerning ‘students’ without permission. Doe v. Yale Univ., 564 F. Supp.
3d 11, 20 (D. Conn. 2021) (quoting 20 U.S.C. § 1232g(b)(1)). “FERPA does, however, allow the disclosure of students’ educational records if ‘such information is furnished in compliance with judicial order, or pursuant to any lawfully issued subpoena, upon condition that parents and the students are notified of all such orders or subpoenas in advance of the compliance therewith by the educational institution.’” Metcalf v. Yale Univ., No. 15-CV-1696 (VAB), 2017 WL 627423, at *4 (D. Conn. Feb. 15, 2017) (quoting 20 U.S.C. § 1232g(b)(2)(B)). Courts have applied FERPA to the release of personnel files of medical residents. See, e.g., Craig v. Yale Univ. Sch. of Med., No. 3:10 CV 1600 JBA, 2012 WL 1579484, at *2 (D. Conn. May 4, 2012).
Defendant claims that compliance with FERPA will significantly increase the burden of responding to Plaintiff’s document requests, because the University will be required to redact all PII and notify the trainees whose records will be subject to production. (Dkt. 41 at 23-24). In his reply brief, Plaintiff argues that FERPA’s notice requirement does not apply when all PII is redacted—redactions to which he does not object—and that his need for the requested records outweighs the privacy interests of the residents and fellows whose records will be subject to production. (Dkt. 42 at
6-7, 13-14). While there is some case law supporting Plaintiff’s argument that FERPA does not require an institution to give advance notice to a student when all PII in their records has been redacted, see Doe v. Wesleyan Univ., No. 3:19-CV-01519 (JBA), 2021 WL 4704852, at *8 n.4 (D. Conn. Oct. 8, 2021), even assuming that is the case, the University would nevertheless be subject to a legitimate burden in having to make
such redactions. Essentially, under Plaintiff’s view, Defendant can follow one of two alternatives: “either comply with FERPA’s [notice] requirement . . . or . . . redact all personally identifiable information.” Doe v. Ohio, No. 2:91-CV-0464, 2013 WL 2145594, at *6 (S.D. Ohio May 15, 2013). Either way, however, “compliance with FERPA will place some burden on Defendant[].” Id. The burden of redacting the PII also cannot be dismissed out of hand given FERPA’s expansive definition of PII, which encompasses information beyond basic direct and indirect identifiers. See 34 C.F.R. § 99.3;2 see also Daniels v. Univ. of Notre Dame, No. 3:22-CV-698-PPS-JEM, 2025 WL 1899726, at *2 (N.D. Ind. June 4, 2024) (denying motion to compel in part
due to the burden imposed by the “more in-depth redactions . . . required” by FERPA). Plaintiff’s second argument—that his need for the requested records outweighs the privacy interests of the students whose records will be produced—refers to a threshold showing that some courts in this Circuit require when a party moves to compel disclosure of educational records covered by FERPA. See, e.g., Doe, 564 F. Supp. 3d at 21. However, as noted recently by another court in this Circuit, “nothing in the FERPA statutory text or implementing regulations, nor in any binding
precedent, . . . indicates that courts considering discovery requests must deploy a discrete standard for educational records as a threshold inquiry before implementing Rule 26.” Ware, 2025 WL 2304724, at *3. That is, Rule 26(b)’s proportionality requirement already provides adequate grounds for a court to balance Plaintiff’s need for the requested records against both the burden imposed on Defendant in complying with FERPA and the privacy interests of the third parties whose records would be
subject to production. See id.
2 Specifically, “personally identifiable information” under FERPA includes “[o]ther information that, alone or in combination, is linked or linkable to a specific student that would allow a reasonable person in the school community, who does not have personal knowledge of the relevant circumstances, to identify the student with reasonable certainty,” and “[i]nformation requested by a person who the educational agency or institution reasonably believes knows the identity of the student to whom the education record relates.” 34 C.F.R. § 99.3. Likewise, while Plaintiff argues that FERPA is not a bar to disclosure (Dkt. 42 at 13), Defendant’s point here is not that FERPA operates as a bar, but merely that “[c]ourts factor the burden and expense of collecting and redacting records subject to
FERPA and providing the required notice . . . into the proportionality analysis under Rule 26” (Dkt. 41 at 14). See, e.g., Ware, 2025 WL 2304724, at *3 (“Not only must Defendants locate and review each requested document, but they must also often redact PII and notify the involved students prior to production. These necessary expenses factor into the proportionality analysis for Plaintiffs’ requests.”). For the foregoing reasons, the Court agrees with Defendant that FERPA imposes a cognizable burden, and the Court has appropriately factored that burden
into its analysis below of whether each of Plaintiff’s discovery requests satisfies Rule 26(b)’s proportionality requirement. III. Comparator Evidence The Court turns next to the parties’ dispute over whether other IM residents and HemeOnc fellows are similarly situated to Plaintiff for the purpose of obtaining comparator evidence. In addressing this dispute, it is helpful for the Court to discuss
some of the underlying law governing employment discrimination claims. In employment discrimination cases, “comparator evidence is generally used to show that a plaintiff was treated less favorably than a similarly situated employee outside his protected group.” Metcalf, 2017 WL 627423, at*4 (quotation omitted). “Such comparator evidence can be especially relevant to [demonstrating] pretext . . . as it may show that the protected characteristic accounted for the differential treatment of two otherwise alike employees.” Whitney v. Montefiore Med. Ctr., No. 21 CIV. 9623 (PAE), 2023 WL 7386400, at *14 (S.D.N.Y. Nov. 8, 2023) (quotation and original alteration omitted), aff’d, No. 23-7961-CV, 2025 WL 2101512 (2d Cir. July
28, 2025). But “[c]omparator evidence supports an inference of discrimination only if the comparator is similarly situated to the plaintiff in all material respects.” Carey v. New York State Dep’t of Health, No. 23 CIV. 3061 (LGS), 2025 WL 2732918, at *10 (S.D.N.Y. Sept. 25, 2025) (quotation omitted). “An employee is similarly situated to co-employees if they were (1) subject to the same performance evaluation and discipline standards and (2) engaged in comparable conduct.” Ruiz v. Cnty. of
Rockland, 609 F.3d 486, 493-94 (2d Cir. 2010) (quotations omitted). In the context of employee discipline, “show[ing] that similarly situated employees ‘engaged in comparable conduct,’” requires a “‘reasonably close resemblance of the facts and circumstances of plaintiff’s and comparator’s cases,’ such that ‘the conduct for which the employer imposed discipline was of comparable seriousness’ to the conduct of the similarly situated but undisciplined employees.” Jordan v. United Health Grp. Inc.,
783 F. App’x 31, 33 (2d Cir. 2019) (quoting Graham v. Long Island Railroad, 230 F.3d 34, 40 (2d Cir. 2000)). “Discovery into potential comparators is a well-recognized purpose of discovery in an employment discrimination or retaliation case.” Felder v. Warner Bros. Discovery, No. 23 CIV. NO. 08487 (AT) (GS), 2025 WL 1718098, at *4 (S.D.N.Y. June 20, 2025). “The standard for allowing discovery of potential comparators is more liberal” than “at the summary judgment stage.” Id. at *5. This is because the “the purpose of [such] discovery is, in part, to help identify the universe of proper comparators.” DeAngelo v. Yellowbook Inc., No. 3:12-CV-520 GWC, 2015 WL 588627,
at *4 (D. Conn. Feb. 10, 2015) (quotation omitted). Nevertheless, when resolving a dispute over comparator discovery, courts commonly undertake at least a preliminary analysis of whether the proposed comparators are similarly situated to the plaintiff, in order to assess whether the evidence sought constitutes relevant and proportional discovery for purposes of Rule 26(b). See, e.g., Metcalf, 2017 WL 627423, at *3 (applying law governing disparate treatment claims to determine whether comparator evidence sought was relevant);
Johnson v. Riverhead Cent. Sch. Dist., No. CV147130DRHAKT, 2016 WL 4507002, at *10 (E.D.N.Y. Aug. 26, 2016) (“because Plaintiff's counsel has not provided information tending to show that Plaintiff and the identified employees are similarly situated, Plaintiff has not established the relevance of the personnel files”); Smalls v. New York Hosp. Med. Ctr. of Queens, No. 13 CV 1257 (RRM), 2013 WL 12333083, at *5 (E.D.N.Y. Oct. 15, 2013) (“there should be a ‘close resemblance of the facts and
circumstances’ surrounding the plaintiff and comparator in order for personnel files to be discoverable”). Here, Plaintiff asserts that “comparator evidence is relevant to [his] discrimination and retaliation claims” because he “can show that the University’s basis for terminating him was pretextual by comparing the University’s treatment of him to that of similarly situated trainees[.]” (Dkt. 35-2 at 9). However, consistent with the foregoing case law, to be entitled to discovery of the evaluations and personnel files of other IM residents and HemeOnc fellows, Plaintiff must make at least some showing that those trainees could be similarly situated to himself.
A. IM Residents RFPs 2, 3, and 4 seek, in whole or in part, employment records related to IM residents. Defendant argues that Plaintiff is not similarly situated to other IM residents because the non-renewal of Plaintiff’s contract was based on performance issues that occurred during and in connection with his HemeOnc fellowship, where Plaintiff was subject to different performance standards, had different responsibilities and a different supervisor, and was evaluated by different
decisionmakers than IM residents. (Dkt. 41 at 17-18). Plaintiff, on the other hand, points to the fact that he first received a Conditional on Improvement rating in his third year in the Program, which, as discussed above, he contends was a combination of the final year of his IM residency and the first year of his HemeOnc fellowship. (Dkt. 42 at 8). Plaintiff also argues that “during his time in the [IM] residency portion of the Program, [he] was subject to the exact same performance standards as all other
[IM] residents, making them proper comparators.” (Dkt. 35-2 at 12). The Court agrees with Defendant that the employment records of IM residents are beyond the scope of relevant discovery. Plaintiff’s asserted need for comparator evidence is to show that the performance issues Defendant cites as the reason for his termination were pretextual. (See, e.g., Dkt. 42 at 5-6). But Plaintiff’s performance issues arose during and exclusively in connection with his HemeOnc fellowship. During the relevant timeframe, Plaintiff was only evaluated by the HemeOnc CCC based on HemeOnc’s specific milestones and standards, and there is no overlap between the members of the HemeOnc CCC and the members of the IM CCC. (Dkt.
41 at 17). Plaintiff also had a different supervisor beginning in his third year—Dr. Carla Casulo, M.D., of the HemeOnc department—than he had during his IM residency. (Id.; see also Dkt. 41-13 at 10-11). While not necessarily dispositive, “a key determinant as to the employees who are proper comparators with an employment discrimination plaintiff is whether they shared a common supervisor.” Sanders v. SUNY Downstate Med. Ctr., No. 22 CV 4139 (KAM) (CLP), 2024 WL 4198355, at *8 (E.D.N.Y. Sept. 16, 2024) (quotation omitted).
Although Plaintiff points to evidence indicating the first year of his HemeOnc fellowship counted, in a technical sense, toward his IM residency (see Dkt. 42 at 8), he has neither shown that IM’s evaluation or disciplinary standards nor his conduct during his residency played any role in his termination, such that evidence of the treatment of other IM residents would be relevant. See Artunduaga v. Univ. of Chicago Med. Ctr., No. 12 C 8733, 2017 WL 11887926, at *3 (N.D. Ill. Jan. 6, 2017)
(finding general surgery residents were not proper comparators to plaintiff, a former plastic and reconstructive surgery (“PRS”) resident, even though PRS was considered a “combined residency” with general surgery and PRS residents “spend the first part of their residency under [the] General Surgery Program”). For this reason, it is immaterial that Plaintiff was subject to the same performance standards as IM residents during his residency, as the relevant performance standards are those that he was subject to during his fellowship.3 The University’s “revoking” Plaintiff’s Certificate of Completion for his IM
residency (see Dkt. 35-2 at 11) is similarly immaterial for present purposes. Plaintiff acknowledges that this was a retroactive action taken because he received a Conditional on Improvement rating on his year-end evaluation during his third year in the Program and as a byproduct of the fact that the third year counted dually toward his residency and fellowship. (See id; Dkt. 41 at 17; Dkt. 35-6 at 19; Dkt. 41- 8 at 2). Plaintiff’s ACGME milestone form for this evaluation reflects that it was completed by HemeOnc, not IM. (See Dkt. 41-2). While the revoking of his Certificate
of Completion ultimately had implications for Plaintiff’s IM residency, Plaintiff has not articulated its relevance for the purpose of obtaining comparator discovery. In addition to lacking relevance, the IM-related records Plaintiff is seeking are disproportionate to the needs of the case. RFP 2, which seeks ACGME milestones and CCC Action plans for all IM residents over a four-year period, would, according to Defendant, require the production of approximately 1,216 evaluations. (Dkt. 41 at
21). As discussed in more detail below in connection with RFP 1, the vast majority of these evaluations will, in all likelihood, be irrelevant because RFP 2 is not tailored to
3 Plaintiff also states that the University has “imposed Disciplinary Procedures and Grievance/Appeals Policy, which sets forth the disciplinary mechanisms applicable to all trainees.” (Dkt. 35-2 at 13). However, the copy of the policy that Plaintiff has provided only states that “[t]hese procedures are applicable to all residents” and does not make any reference to fellows. (See Dkt. 35-5 at 2). Because the relevant period of time is his fellowship, Plaintiff has not shown that this policy could serve as a basis for comparability. identify residents who engaged in comparable conduct. The privacy interests of these residents thus outweigh the likely benefit of the discovery. Defendant further represents that determining which IM personnel files would
be responsive to RFP 3 would require the University to review approximately 790 evaluations. (Id. at 22.). These personnel files, as well as those produced in response to RFP 4, may require extensive redactions pursuant to FERPA. While the scope of comparator discovery may be broad, Plaintiff’s requests for records related to IM residents more closely resemble a fishing expedition than properly tailored discovery. For these reasons, the Court will not compel the evaluations and personnel files of other IM residents. Specifically, Plaintiff’s motion is denied as to RFP 2 and
as to RFP 3 and RFP 4 to the extent these requests seek documents related to IM residents. B. HemeOnc Fellows Defendant also argues that Plaintiff is not similarly situated to and therefore not entitled to comparator discovery regarding other HemeOnc fellows who were not in the Program, because such fellows did not have “the same research component as
Plaintiff [and] their evaluations did not include a research component.” (Dkt. 41 at 18-19). The Court agrees with Plaintiff that, on this point, Defendant is taking an overly narrow view of the range of potential comparators. (See Dkt. 42 at 9). “[T]he standard for comparing conduct requires a reasonably close resemblance of the facts and circumstances of plaintiff’s and comparator’s cases, rather than a showing that both cases are identical.” Graham, 230 F.3d at 40. Defendant points to evidence that it says shows “Plaintiff’s poor performance and lack of attendance in the research lab played a role in his evaluations and in the University’s decision not to renew his
contract.” (See Dkt. 41 at 18-19). But Plaintiff notes that “the University has never identified any ‘additional research-based obligations and performance measures’ that [he] was subject to simply because he was in the Program.” (Dkt. 42 at 9-10 (quoting Dkt. 41 at 6)). Plaintiff also points out that his performance was evaluated based on the standards set by ACGME, and Defendant does not assert that such standards were different for traditional HemeOnc fellows and HemeOnc fellows in the Program. (Id. at 10).
At this stage of the proceedings, Plaintiff’s research responsibilities as a participant in the Program do not so clearly distinguish his situation from that of traditional HemeOnc fellows that it would be reasonable to limit discovery on this basis. Indeed, because Plaintiff was the only HemeOnc fellow in the Program during the years covered by the RFPs (see Dkt. 41 at 19), adopting Defendant’s position would deny Plaintiff any comparator discovery, which could in turn prevent him from
even attempting to show disparate treatment and/or pretext. See Chen-Oster v. Goldman, Sachs & Co., No. 10-CV-6950 (AT) (RWL), 2019 WL 3294145, at *2 (S.D.N.Y. June 7, 2019) (“Denying all discovery of comparators, however, would prevent Plaintiffs from even attempting to make [a] showing [that a comparator is similarly situated].”). Accordingly, the Court will consider Plaintiff’s requests for evaluations and personnel files to the extent they relate to HemeOnc fellows. IV. Plaintiff’s Requests for Production
A. RFP 1 RFP 1 seeks “Hematology/Oncology fellow evaluations, consisting of the ACGME milestones and CCC Action plans, for a period of 5 years (academic years beginning 2018-2023).” (Dkt. 35-2 at 4). In the parties’ most recent conferral efforts preceding the motion to compel, Defendant offered to produce these documents with all PII redacted and “on the condition that [the parties] amend [their] protective order to add an [AEO] provision,
understanding that these materials will be marked AEO.” (Dkt. 35-22 at 2). In response, Plaintiff advised Defendant that he “is amenable to this proposal with the exception of the [AEO] designation as the redaction of all identifying information in the evaluations makes this redundant.” (Dkt. 35-23 at 2). Defendant now states that its offer to produce these documents was made “solely for the purpose of avoiding the cost and delay of motion practice[.]” (Dkt. 41 at 12). Plaintiff, however, argues that
Defendant “has acknowledged the relevance of this information by agreeing to produce it, albeit in a limited fashion.” (Dkt. 35-2 at 10). Defendant’s current objection to RFP 1 appears to be that the request would encompass evaluations of HemeOnc fellows irrespective of whether they received a Conditional on Improvement rating or were disciplined for performance deficiencies, like Plaintiff. (See Dkt. 41 at 20). Defendant argues that Plaintiff’s failure to so limit his request ignores the second prong of the similarly situated analysis, which requires comparators to have “engaged in comparable conduct.” (Id.). The Court finds some merit to Defendant’s argument, in the sense that RFPs
3 and 4—which seek the personnel files of HemeOnc fellows who received a Conditional on Improvement rating and who were placed on probation, suspension, and/or had their contract non-renewed—are likely to encompass the entire universe of potential comparators, as the personnel files responsive to those requests would reflect the University’s treatment of other HemeOnc fellows who were disciplined for conduct of comparable seriousness. Plaintiff argues that until he is “privy to other trainees’ evaluations he has no way of knowing whether they engaged in similar
conduct as he did.” (Dkt. 42 at 11). However, Plaintiff fails to articulate how the evaluations would enable him to identify a potential comparator who would not already be identified through Defendant’s responses to RFPs 3 and 4. Plaintiff suggests that, “[f]or example, a review of a trainee’s ACGME milestone evaluation could show that: (1) the trainee received similar comments within a competency and then a different overall rating than Dr. Buhrman or (2) that the trainee received the
same ratings as Dr. Buhrman but received no subsequent disciplinary action.” (Id.). But the first example is highly speculative, would be of minimal probative value, and “the likelihood that discovery of [the evaluations] will be overly burdensome outweighs the scant possibility of uncovering admissible evidence.” Chen-Oster v. Goldman, Sachs & Co., 293 F.R.D. 557, 567 (S.D.N.Y. 2013). The second example would seemingly be addressed by RFP 3, as Plaintiff would be able to determine from their personnel file whether a fellow who received a Conditional on Improvement rating received any subsequent disciplinary action. The Court accordingly concludes that RFP 1 is overbroad insofar as it seeks a
significant number of evaluations—approximately 240, according to Defendant (Dkt. 41 at 22)—without any attempt to limit its scope to fellows who engaged in comparable conduct. Even courts that recognize the need for broad comparator discovery still require the plaintiff to reasonably limit their requests to comparators who engaged in similar conduct. See, e.g., Doe, 2021 WL 4704852, at *7 (recognizing plaintiff’s entitlement to comparator discovery but noting “[t]his of course does not mean that a plaintiff is entitled to discovery about every fellow student, however
dissimilarly situated . . . [b]ut in limiting her requests to students accused of cheating, the plaintiff has made a reasonable effort to avoid inquiring after obviously dissimilar cases”); Oliver v. New York State Police, No. 115CV0444BKSDJS, 2019 WL 1324040, at *6 (N.D.N.Y. Mar. 22, 2019) (“[T]here is no question that information concerning similarly situated male State Troopers may well be relevant to the Plaintiff’s claim that she was treated more harshly because of her gender or because she filed a sex
discrimination complaint. It does not follow, however, that Plaintiff is therefore entitled to each and every disciplinary file that she chooses[.]”) (citations omitted). Nonetheless, Plaintiff has articulated a rationale for compelling production of a narrower subset of evaluations. Specifically, Plaintiff notes the importance of reviewing a trainee’s evaluations and personnel file “in conjunction,” so that Plaintiff “can review the fellow[’s] . . . evaluations in comparison to how the University imposed disciplinary actions[.]” (Dkt. 35-2 at 13, 14). A fellow’s evaluations also likely contain specific comments or criticism—beyond an overall rating—that could be relevant to the fellow’s treatment by the University, as gleaned through their personnel file.
The Court therefore finds that RFP 1 is more appropriately limited to the evaluations of HemeOnc fellows who received a Conditional on Improvement rating or were placed on probation, suspension, and/or had their contract non-renewed during the academic years beginning 2018-2023. The evaluations produced in response to RFP 1 would thus correspond to the fellows whose personnel files are produced in response to RFP 3 and 4.4 See, e.g., Chen-Oster, 2019 WL 3294145, at *3 (ordering production of personnel and disciplinary files and evaluation documents for
each identified comparator). To this end, Defendant states that identifying the personnel files responsive to RFPs 3 would necessarily require the University to review the ACGME milestone evaluations for all HemeOnc fellows during the applicable five-year period. (See Dkt. 41 at 22-23). The overlapping nature of the search would thus reduce the burden of responding to RFP 1 as limited herein. B. RFP 3
RFP 3, as modified by the Court’s determination above regarding IM residents, seeks “[p]ersonnel files of Hematology/Oncology fellows . . . who received a ‘conditional upon improvement’ rating on his/her ACGME milestones, for a period of 5 years (academic years beginning 2018-2023).” (Dkt. 35-2 at 4).
4 To the extent the personnel files and evaluations are stripped of any identifying information, Defendant must ensure that Plaintiff can determine for each anonymized fellow their corresponding personnel file and evaluations. The Court finds that the documents sought by RFP 3 are relevant and proportional to the needs of the case. “[P]re-trial discovery of personnel files of similarly situated employees should be permitted where disclosure is reasonably
likely to lead to the discovery of admissible evidence of disparate treatment and/or employment discrimination directed at the plaintiff.” Smalls, 2013 WL 12333083, at *5. As Plaintiff explains, “by receiving a ‘conditional upon improvement’ rating on their evaluations . . . these fellows[] . . . engaged in comparable conduct” to himself. (Dkt. 35-2 at 13). Plaintiff’s request for personnel files of HemeOnc fellows who received a Conditional on Improvement rating is therefore reasonably calculated to identify similarly situated fellows and obtain relevant comparator evidence.
The Court finds Defendant’s objections unavailing as applied to RFP 3. As explained above, the Court disagrees that Plaintiff cannot be similarly situated to other HemeOnc fellows who were not in the Program. Defendant takes a similarly restrictive view of who could constitute a proper comparator in arguing that “even if the Court concludes that other HemeOnc fellows could be comparators . . . any resulting discovery should be limited to any HemeOnc fellows with a comparable
profile as Plaintiff, i.e., those who had received ‘conditional on improvement’ ratings for multiple consecutive years and were removed from clinical work due to patient safety concerns.” (Dkt. 41 at 20). This limitation would, again, improperly turn the comparable conduct standard into a requirement that a comparator have engaged in identical conduct. Defendant’s remaining objections, insofar as they are applicable, concern the burden of producing the requested documents. Specifically, Defendant points to the burden of having to (1) review all HemeOnc ACGME milestone evaluations
(approximately 120 evaluations) from the requested five-year period to identify which personnel files are responsive to RFP 3, (2) “manually search for, extract, and review” HemeOnc’s evaluations and personnel files, which the department maintains in hard copy only, and (3) comply with FERPA. (Id. at 22-23). While the Court rejects Plaintiff’s characterization of these burdens as “disingenuous at worst and conclusory at best” (Dkt. 42 at 12), Defendant has not shown that responding to RFP 3 would be so burdensome as to outweigh the likely benefit to Plaintiff of obtaining relevant and
necessary comparator discovery. Moreover, an employer resisting discovery “may not take advantage of self-made burdens” related to its own recordkeeping practices. Carter v. Logan Bus Co., No. 15-CIV-5217 (ENV) (JO), 2016 WL 5231800, at *2 (E.D.N.Y. Sept. 21, 2016) (quotation omitted). For these reasons, the Court grants Plaintiff’s motion as to RFP 3 to the extent it seeks HemeOnc personnel files.
C. RFP 4 RFP 4, as modified by the Court’s determination above regarding IM residents, seeks “[p]ersonnel files of Hematology/Oncology fellows . . . that were placed on probation, suspension, and/or had their contract non-renewed, for a period of 5 years (academic years beginning 2018-2023).” (Dkt. 35-2 at 4). Like RFP 3, the Court finds that RFP 4 is relevant and proportional to the needs of the case. As Plaintiff explains, “any trainee who was placed on probation, suspension, and/or had their contract non-renewed likely engaged in comparable
conduct to [himself].” (Dkt. 35-2 at 14). Plaintiff’s request for personnel files of HemeOnc fellows who were placed on probation, suspension, and/or had their contract non-renewed is therefore reasonably calculated to identify similarly situated fellows and obtain relevant comparator evidence. The Court’s analysis of Defendant’s objections above in the context of RFP 3 applies with the same force in the context of RFP 4. Further, while Defendant claims that identifying which personnel files are responsive to RFP 4 would be burdensome
insofar as it would require (like RFP 3) a review of all 120 HemeOnc ACGME milestone evaluations from the relevant five-year period (Dkt. 41 at 23), it does not appear that these evaluations would indicate whether a fellow was placed on probation, suspension, and/or had their contract non-renewed. Plaintiff notes that his own evaluations did not include such information (Dkt. 35-2 at 14). Defendant has thus not clearly shown that this particular burden is applicable to RFP 4.
For these reasons, the Court grants Plaintiff’s motion as to RFP 4 to the extent it seeks HemeOnc personnel files. V. Relevancy Redactions Lastly, the Court addresses Plaintiff’s request to compel Defendant to “reproduce the unredacted versions of all documents where redactions were made solely on the basis of relevancy.” (Dkt. 35-2 at 17). The extent of this dispute is somewhat unclear. Plaintiff only points to Defendant’s redaction of CCC meeting notes (see Dkt. 35-14) but describes the notes as an “example” (Dkt. 35-1 at ¶ 12). For its part, Defendant omits any mention in its motion papers of the relevancy
redactions dispute or Plaintiff’s request to compel unredacted versions of redacted documents. “In this Circuit, the weight of authority goes against allowing a party to redact information from admittedly responsive and relevant documents based on that party’s unilateral determinations of relevancy.” McConkey v. Churchill Sch. & Ctr., No. 24-CV-06091 (LJL), 2025 WL 1928552, at *1 (S.D.N.Y. July 14, 2025) (quotation omitted). This is, in part, because “such unilateral determinations of relevancy breed
suspicions and may deprive the reader of context[.]” Swanson v. Manhattan Beer Distributors, LLC, No. 15-CV-5383 (ENV), 2019 WL 13220137, at *1 (E.D.N.Y. Jan. 24, 2019) (quotations omitted). A related “concern with relevancy redactions is that they can lead to motion practice . . . which often creates additional expense and delay.” Kaiser Aluminum Warrick, LLC v. US Magnesium LLC, No. 22CV3105JGKKHP, 2023 WL 2482933, at *2 (S.D.N.Y. Feb. 27, 2023).
Courts do permit relevancy redactions in limited circumstances (1) where the producing party satisfies the good cause standard for a protective order under Rule 26(c), or (2) pursuant to some form of discovery protocol that limits the producing party’s discretion to redact material it deems irrelevant. See Marcelletti v. GEICO Gen. Ins. Co., No. 6:23-CV-06211 EAW CDH, 2025 WL 1811668, at *6, *6 n.8 (W.D.N.Y. July 1, 2025) (discussing approaches by courts in this Circuit to permitting relevancy redactions). But under either approach, advance permission or approval by the Court is generally required before a party produces documents with redactions based on relevancy.
Here, Defendant has not at any time sought the Court’s permission to redact any discovery materials. And, in the course of litigating the instant motion, Defendant, having not addressed the matter at all, has not shown good cause for making such redactions. Moreover, by virtue of their production, albeit with redactions, the applicable documents are “admittedly responsive.” There is accordingly no basis for the Court to allow Defendant to redact these documents based on its own unilateral determinations of relevancy.
Within 30 days of entry of this Decision and Order, Defendant must provide Plaintiff with unredacted versions of all documents that it previously produced with redactions based solely on relevancy. If Defendant believes there is a legally justifiable basis for these redactions that is consistent with this Decision and Order and case law in this Circuit, then it must obtain the Court’s approval for those redactions through an affirmative request for relief within 30 days of entry of this
Decision and Order. CONCLUSION For the reasons stated above, Plaintiff’s motion to compel is granted in part and denied in part. Within 30 days of entry of this Decision and Order, Defendants must produce: Evaluations of Hematology/Oncology fellows who received a Conditional on Improvement rating on their ACGME milestones or were placed on probation, suspension, and/or had their contract non-renewed, consisting of the ACGME milestones and CCC Action plans, for a period of 5 years (academic years beginning 2018-2023); Personnel files of Hematology/Oncology fellows who received a Conditional on Improvement rating on their ACGME milestones, for a period of 5 years (academic years beginning 2018-2023); and Personnel files of Hematology/Oncology fellows who were placed on probation, suspension, and/or had their contract non-renewed, for a period of 5 years (academic years beginning 2018-2023). Also within 30 days of entry of this Decision and Order, Defendant must provide Plaintiff with unredacted versions of all documents that it previously produced with redactions based solely on relevancy. If Defendant believes there is a legally justifiable basis for its redactions that is consistent with this Decision and Order and case law in this Circuit, then it must obtain the Court’s approval for those redactions through an affirmative request for relief within 30 days of entry of this Decision and Order. SO ORDERED. (ow DHlbonl COLLEEN D. HOLLAND United States Magistrate Judge Dated: Rochester, New York August 6, 2026
29.