Ibroci et al v. Idrovo et al

Filing 64

MEMORANDUM AND ORDER: For the foregoing reasons, plaintiffs shall either withdraw Dr. Davatzikos' opinion or disclose the numerical data for all brain regions for the 710 members in the control group selected by Dr. Davatzikos without personally identifying information. Plaintiffs shall inform defendants and the Court of their decision within two (2) weeks of the entry of this Memorandum and Order. SO ORDERED. (Signed by Judge Naomi Reice Buchwald on 4/9/2020) (ks)

UNITED STATES DISTRICT COURT
SOUTHERN DISTRICT OF NEW YORK
--------------------------------------X
LINDITA IBROCI, and
NEIL IBROCI,
Plaintiffs,
MEMORANDUM AND ORDER
- against -

17 Civ. 1449 (NRB)

IVAN SERGIO IDROVO, and
JC TRANSPORT INC.,
Defendants.
--------------------------------------X
NAOMI REICE BUCHWALD
UNITED STATES DISTRICT JUDGE
The Court is in receipt of the parties’ letters dated November
13, 2019; November 15, 2019; November 19, 2019; December 5, 2019;
December 11, 2019; and December 19, 2019, and the exhibits attached
thereto.

The Court also held a teleconference on November 25,

2019 and an in-person conference on March 5, 2020 to discuss this
discovery dispute with the parties.
Plaintiffs resist disclosure of the entire data from which
plaintiffs’ expert extracted certain data upon which he bases his
opinion.

The underlying reason for this resistance is that the

expert received the data under the condition that he hold it in
confidence and use it only for his academic endeavors.

However,

plaintiffs’ expert, Dr. Davatzikos, used the data to prepare an
expert report in a private litigation.

The issue posed here is

whether plaintiffs and their expert may rely on the derivations
from the data without disclosing the entire data set.

The answer

is no for two independently dispositive reasons.
First,

plaintiffs’

counsel

did

not

dispute

the

Court’s

observation that, if the data had not been received under a
confidentiality agreement, it would have been disclosed in the
normal discovery process.
25.

See Tr. for March 5, 2020 Conference at

Plaintiffs may not change the discovery rules by selecting an

expert who has entered into confidentiality agreements.
Second, defendants have demonstrated a need for the entire
data set in order to adequately challenge plaintiffs’ expert
opinion.

The gist of Dr. Davatzikos’ opinion is a comparison of

where plaintiff Ms. Ibroci stands in relation to other women in
her age group in terms of the volume of certain brain regions
selected by Dr. Davatzikos.

See Davatzikos Amended Report (ECF

No. 56-6) at 2. In determining whether a specific brain region of
Ms.

Ibroci’s

brain

qualifies

as

an

“abnormal”

region,

Dr.

Davatzikos relied on the z-scores (i.e. the difference between the
control group’s mean and Ms. Ibroci’s brain region volume for a
particular brain region, divided by the control group’s standard
deviation for the volume of that brain region).

Accordingly, at

the foundation of Dr. Davatzikos’ opinion is the accuracy of the
mean and standard deviation of the control group members’ brain
2

region volumes.

Requiring defendants to simply assume the accuracy

of Dr. Davatzikos’ calculations of the control group’s mean and
standard deviation would deny defendants a meaningful opportunity
to cross-examine Dr. Davatzikos as contemplated under the Federal
Rule of Evidence 705.

See Lee Valley Tools, Ltd. v. Indus. Blade

Co., 288 F.R.D. 254, 267 (W.D.N.Y. 2013) (“[E]ffective crossexamination is dependent upon access to the underlying data and
assumptions upon which the expert relied.”) (citing R.F.M.A.S.,
Inc. v. So, 748 F. Supp. 2d 244, 269 (S.D.N.Y. 2010)).
Further, Dr. Davatzikos also cites Ms. Ibroci’s percentile
rank in terms of the volume of certain brain region as evidence of
abnormality.

Davatzikos Amended Report (ECF No. 56-6) at 1.

Despite Dr. Davatzikos appears to derive this percentile rank from
the z-score, nowhere in his report does he provide the basis for
translating the z-score into a percentile rank or deriving it
otherwise.

At the March 5, 2020 conference, plaintiffs argued

that their disclosure of the control group mean and standard
deviation

as

to

each

brain

region

would

allow

defendants

to

recreate the distribution of the control group members’ brain
region

area

for

each

region,

which

would

presumably

defendants to verify Ms. Ibroci’s percentile rank.
rejects this argument.

allow

The Court

Because different distributions can have

the same mean and standard deviation, see F. J. Anscombe, Graphs
3

in Statistical Analysis, 27 AM. STATISTICIAN 17, 19-20 (1973), a
disclosure of mean and standard decisions is insufficient to
replace a disclosure of the actual distribution.

Without brain

region volume data as to individual members of the control group
being disclosed, defendants would be deprived of a meaningful
opportunity

to

challenge

Dr.

Davatzikos’

opinion

on

the

abnormality of plaintiff’s brain. 1
For the foregoing reasons, plaintiffs shall either withdraw
Dr. Davatzikos’ opinion or disclose the numerical data for all
brain regions for the 710 members in the control group selected by
Dr.

Davatzikos

without

personally

identifying

information.

Plaintiffs shall inform defendants and the Court of their decision
within two (2) weeks of the entry of this Memorandum and Order.
SO ORDERED.

Dated:

New York, New York
April 9, 2020
____________________________
NAOMI REICE BUCHWALD
UNITED STATES DISTRICT JUDGE

1
Plaintiffs’ argument based on the Health Insurance Portability
and Accountability Act of 1996 (“HIPPA”) is mooted by the fact that
defendants are not seeking any personally identifiable information, including
the MRI brain scan images of the control group members.

4