Summary
finding predecessor database to FAIR Health could not be relied on to establish reasonableness of medical charges, in part, because "it relie[d] on the voluntary submission of data on medical costs from the limited universe of insurance companies who choose to participate in the program"
Summary of this case from Verci v. HighOpinion
SJC–11284.
2013-09-10
N.E. PHYSICAL THERAPY PLUS, INC. v. LIBERTY MUTUAL INSURANCE COMPANY.
Daniel P. Tighe (Douglas R. Tillberg with him), Boston, for the defendant. Francis A. Gaimari, Needham, (Stephen B. Byers with him) for the plaintiff.
Daniel P. Tighe (Douglas R. Tillberg with him), Boston, for the defendant. Francis A. Gaimari, Needham, (Stephen B. Byers with him) for the plaintiff.
The following submitted briefs for amici curiae:
E. Michael Sloman for Automobile Insurers Bureau.
Peter A. Biagetti, Needham, & Andrew Nathanson for Fair Health, Inc.
Paul R.Q. Wolfson, Shirley Cassin Woodward, & Dina B. Mishra, of the District of Columbia, & John J. Regan & Mark C. Fleming, Boston, for American Insurance Association & others.
David L. Arrington & Jodi L. Howick, of Utah, & Todd S. Holbrook & Jeffrey D. Adams, Boston, for Mitchell International, Inc.
Present: IRELAND, C.J., SPINA, CORDY, BOTSFORD, GANTS, DUFFLY, & LENK, JJ.
LENK, J.
After being billed by the plaintiff, N.E. Physical Therapy Plus, Inc. (NEPT), for certain medically necessary chiropractic services provided to the passenger of its insured, the defendant, Liberty Mutual Insurance Company (Liberty Mutual), maintained that the cost of those services was unreasonably high and thus refused to pay the full amount invoiced.
At the ensuing trial on the dispute, Liberty Mutual sought to introduce statistical evidence from a commercial database to show that NEPT's charges exceeded the eightieth percentile of reported charges for the same procedures and were thus unreasonable. Liberty Mutual claimed that the statistical evidence was admissible pursuant to G.L. c. 233, § 79B ( § 79B), which creates a limited exception to the hearsay rule for factual statements contained in commercial publications. Citing a decision of the Appellate Division of the District Court that concluded the same database was unreliable, a District Court judge denied Liberty Mutual's motion to admit the database evidence.
We acknowledge the amicus briefs of the American Insurance Association, Massachusetts Insurance Federation, National Association of Mutual Insurance Companies, and Property Casualty Insurers Associations of America, and of Mitchell International, Inc., on behalf of Liberty Mutual Insurance Company (Liberty Mutual); and the amicus briefs of the Automobile Insurers Bureau and Fair Health, Inc.
We are asked to determine whether a trial judge has discretion to consider the reliability of evidence offered under § 79B where the evidence otherwise satisfies the requirements of § 79B, and, if so, whether the judge here abused such discretion in excluding the database. Based on the explicit language of § 79B, and the gatekeeper role of a trial judge, we conclude that it is within a judge's discretion to consider the reliability of evidence offered pursuant to § 79B. Because the judge's conclusion that the evidence was not sufficiently reliablefinds support in the record, we conclude also that the judge did not abuse his discretion in excluding its admission.
1. Background. a. Facts. Prior to trial, the parties stipulated to the following facts. On August 2, 2003, Gitane Dalencourt was injured in a motor vehicle accident while a passenger in a vehicle insured by Liberty Mutual. The policy provided $8,000 in personal injury protection (PIP) coverage, as required by G.L. c. 90, § 34A. On October 10, 2003, Liberty Mutual received a PIP claim application from Dalencourt.
Liberty Mutual received bills totaling $4,465 from NEPT, for chiropractic treatment provided to Dalencourt from September 16, 2003 to December 17, 2003. It is undisputed that all treatment given by NEPT to Dalencourt was medically necessary to treat the injuries he sustained in the motor vehicle accident. Liberty Mutual paid a total of $3,730,68, leaving an unpaid balance of $734.32. In a letter to NEPT dated July 21, 2004, Liberty Mutual denied NEPT's request that it pay the remaining balance.
On June 17, 2008, NEPT filed an action against Liberty Mutual in the District Court, seeking payment of the outstanding balance pursuant to G.L. c. 90, § 34M, and damages pursuant to G.L. c. 93A and G.L. c. 93, § 12. The only disputed issue at the jury-waived trial was whether the amount charged for the treatment by NEPT was reasonable.
b. Prior proceedings. Prior to trial, Liberty Mutual filed a motion in limine to admit in evidence, pursuant to the hearsay exception contained in § 79B, statistical summaries prepared by Ingenix, Inc. (Ingenix) of billing patterns in NEPT's geographical area for similar medical procedures. Liberty Mutual intended to introduce the Ingenix data at trial to rebut the reasonableness of NEPT's medical service charges by showing that NEPT's charges exceeded the average amount billed by providers in the area.
In support of its motion, Liberty Mutual submitted both an affidavit of Susan Seare, a senior director of legal and government liaisons at Ingenix, and the trial transcript from a trial in the District Court, Davekos vs. Liberty Mutual Ins. Co., Docket No. 0518CV0156, a case involving a nearly identical dispute. At that trial, Carla Gee, a statistician employed by Ingenix, testified to what the Ingenix database is and how Ingenix produces its data.
In sum, she testified that Ingenix is a wholly-ownedsubsidiary of United Health Group, the owner of United Health Care, one of the largest health insurers in the United States. Ingenix collects data from insurance companies throughout the country on the costs of various medical procedures for which they have been billed. This contribution of data by insurance companies is entirely voluntary and Ingenix relies on its clients to submit complete and accurate data. Ingenix then groups this compiled data according to geographical regions and uses a proprietary “relative value” and “conversion factor” approach to produce the resulting data concerning billing patterns in a given region. The database is thus “based on [Ingenix] methodologies and the data that [it] collect[s].” Gee acknowledged that Ingenix has never verified that the data produced as a result of its methodologies accurately correspond to what providers are actually charging. The judge in that case permitted the introduction of the Ingenix evidence, and Liberty Mutual prevailed at trial.
Susan Seare's affidavit largely echoed the testimony of Carla Gee at the Davekos trial, with the addition of an extensive list of customers who use Ingenix, Inc. (Ingenix) products.
Here, in opposition to Liberty Mutual's motion, NEPT brought to the judge's attention the subsequent decision of the Appellate Division of the District Court, reversing the judgment in the Davekos trial on the ground that the trial judge erred in admitting the Ingenix evidence. See Michael Davekos, P.C. v. Liberty Mut. Ins. Co., 2008 Mass.App. Div. 32, 36, 2008 WL 241613 ( Davekos ). Although its focus was on the requirements for admissibility under the business records exception, G.L. c. 233, § 78, rather than § 79B, reliability of the information contained in the documents is one of those requirements. Id. at 34. Concluding that the “Ingenix statistical analyses are derivative only, and ... that the Ingenix raw data itself ... lacks the requisite indicia of reliability
to be admissible,” the Appellate Division determinedthat the Ingenix documents “do not fall within the broadened business records exception for scientific studies containing primarily factual data.” Id. at 35. NEPT also submitted, in opposition to Liberty Mutual's motion, news articles documenting legal settlements between Ingenix and both the State of New York and the American Medical Association regarding allegations that Ingenix databases reported artificially low charges to justify low insurance payments.
Among the factors undermining the reliability of the Ingenix data, the Appellate Division of the District Court noted: the disclaimer on Ingenix products of “any endorsement, approval or recommendation of data in the database”; that the Ingenix data does not “disclose the total number of providers whose charges may make up the Ingenix database at any point”; that “Ingenix cannot guarantee that all of the bills received for a particular [medical procedure] at any given time have been reported, much less accurately reported, by its volunteer insurers”; and that “[a]t best, the Ingenix database includes the bills of an unspecified number of medical providers who, within a specific period of time, happened to have billed only those health insurers that were not only Ingenix clients, but also Ingenix clients that elected to participate in its voluntary data contribution program.” Michael Davekos, P.C. v. Liberty Mut. Ins. Co., 2008 Mass.App. Div. 32, 34, 36, 2008 WL 241613 ( Davekos ).
As part of a settlement agreement entered into with the Attorney General of New York, Ingenix agreed to fund the creation of “a new independent database, not controlled by any insurer.” The independent, not-for-profit corporation created as a result of this agreement, Fair Health, Inc., filed an amicus brief in this case in which it states that it now maintains “an unbiased and statistically-sound database of healthcare charges.” This database is accessible to consumers through a free public Web site and the data contained therein reflects “approximately 70% of the plan participants in, and family members covered by, employer-provided private health insurance in the United States.” We express no view as to the admissibility of data from the database maintained by Fair Health.
After a hearing on June 18, 2009, a District Court judge denied Liberty Mutual's motion, citing Davekos, supra. The jury-waived trial took place before the same judge on June 24, 2010. Liberty Mutual renewed its motion to introduce Ingenix data,
and the judge affirmed his earlier decision to exclude the data.
Although Liberty Mutual made an offer of proof as to what the Ingenix evidence would show, it never produced for the judge's consideration the actual Ingenix information or the records that it sought to have admitted.
Liberty Mutual presented no other evidence to refute the reasonableness of NEPT's charges. Accordingly, the judge found for NEPT on its G.L. c. 90, § 34M, claim for the outstanding balance of the medical charges.
The judge stated in his written findings that, “[t]he Appellate Division of the District Court dealt squarely with the issue of the admissibility of records from Ingenix in the case of [ Davekos ] and ruled that Ingenix's records are not admissible to establish the truth of the underlying hearsay they contain. Based on the Appellate Division's ruling, this court denied the defendant's motion in limine.”
The plaintiff, N.E. Physical Therapy Plus, Inc. (NEPT) satisfied its burden to establish the reasonableness of the charges by introducing medical bills and records under G.L. c. 233, § 79G. General Laws c. 233, § 79G, states that evidence which complies with the requirements of the section “shall be admissible as evidence of the fair and reasonable charge for such services.”
Liberty Mutual appealed to the Appellate Division, contending that the judge erred in excluding the Ingenix data. The Appellate Division affirmed the judgment, and Liberty Mutual appealed to the Appeals Court. The Appeals Court affirmed the judgment, 81 Mass.App.Ct. 1135, 2012 WL 1672475 (2012), and we granted Liberty Mutual's application for further appellate review.
2. Discussion. We review a trial judge's evidentiary decisions under an abuse of discretion standard. See Commonwealth v. Polk, 462 Mass. 23, 32, 965 N.E.2d 815 (2012). In applying that standard, “we look for decisions based on ‘whimsy, caprice, or arbitrary or idiosyncratic notions,’ ” and “do not disturb the judge's ruling ‘simply because [we] might have reached a different result; the standard of review is not substituted judgment.’ ” Cruz v. Commonwealth, 461 Mass. 664, 670, 963 N.E.2d 1172 (2012), quoting Bucchiere v. New England Tel. & Tel. Co., 396 Mass. 639, 641, 642, 488 N.E.2d 1 (1986).
General Laws c. 233, § 79B, provides:
The Federal rules of evidence contain an analogous provision to G.L. c. 233, § 79B, that allows for the admission of “[m]arket quotations, lists, directories, or other compilations that are generally relied on by the public or by persons in particular occupations.” Fed.R.Evid. 803(17).
“Statements of facts of general interest to persons engaged in an occupation contained in a list, register, periodical, book or other compilation, issued to the public, shall, in the discretion of the court, if the court finds that the compilation is published for the use of persons engaged in that occupation and commonly is used and relied upon by them, be admissible in civil cases as evidence of the truth of any fact so stated.”
A party seeking to admit evidence pursuant to § 79B has the burden of “at the minimum, persuading the trial judge to make the preliminary findings that the proposed exhibit is (1) issued to the public, (2) published for persons engaged in the applicable occupation, and (3) commonly used and relied on by such persons.” Mazzaro v. Paull, 372 Mass. 645, 652, 363 N.E.2d 509 (1977)( Mazzaro ).
Two primary rationales give rise to this exception to the hearsay rule. The first is that, as a practical matter, it would be inconvenient, if not impossible, to issue summonses for all of the authors and compilers of any given publication to testify at trial. See id. at 650, 363 N.E.2d 509, citing 6 J. Wigmore, Evidence § 1702 (Chadbourn rev. ed. 1976). The second is that “the trustworthiness of such lists is assured through general reliance on them by those in a particular profession, and the consequent motivation of the compiler to foster this reliance by his accuracy.” Mazzaro, supra at 650, 363 N.E.2d 509, citing 6 J. Wigmore, Evidence § 1702 (Chadbourn rev. ed. 1976).
Liberty Mutual claims that the judge erred in considering the reliability of the Ingenix data as a factor in his admissibility analysis. According to Liberty Mutual,if a publication satisfies the three statutory elements enumerated above it is thereby admissible, since the reliability of the publication is already sufficiently assured by the fact that it is commonly used and relied on by persons in a particular occupation. See Mazzaro, supra at 650, 363 N.E.2d 509. Such a reading of the statute, however, disregards both the discretion explicitly afforded the judge in the language of § 79B, and the judge's traditional role as gatekeeper of admissible evidence. See Commonwealth v. Fitzpatrick, 463 Mass. 581, 603–604, 977 N.E.2d 505 (2012).
Previous decisions construing § 79B have recognized and emphasized that the statute affords a trial judge discretion to determine the admissibility of an offered publication. See Mazzaro, supra at 653, 363 N.E.2d 509 (“The matter [of admissibility] is entrusted by statute to the discretion of the trial judge”); Torre v. Harris–Seybold Co., 9 Mass.App.Ct. 660, 673, 404 N.E.2d 96 (1980) (“both the express terms of the statute and case law in the area reserve a large measure of discretion to the trial judge in passing on admissibility”). See also Leach & McNaughton, Developments in Massachusetts Evidence 1940–55, 40 Mass. L.Q. 14, 15 (1955) ( “ § 79B gives the court discretion to admit certain commercial lists”). If the Legislature had intended evidence to be admissible, without more, upon satisfaction of the elements enumerated in the statute, it would not have included the language providing judges with discretion to determine admissibility. Where the Legislature has included such language, we are not free simply to ignore it. See Commonwealth v. Jones, 417 Mass. 661, 664, 632 N.E.2d 408 (1994) (“We are not free to ignore or to tamper with [a] clear expression of legislative intent”).
Moreover, reading § 79B to preclude a judge's consideration of the reliability of evidence offered under that statute produces an absurd result insofar as, on that view, a judge would be without authority to exclude obviously flawed and unreliable data simply because it is relied upon in a given industry. Such a reading, in addition to ignoring the clear statutory language, lends itself to obvious abuse by a given industry where the use of manipulated and flawed data is in its interest. See Shepherd v. American Broadcasting Cos., 862 F.Supp. 505, 508 n. 13 (D.D.C.1994), rev'd on other grounds, 62 F.3d 1469 (D.C.Cir.1995) (evidence of reasonable rate for attorneys fees contained in published compilation inadmissible under analogous Federal rule because “[t]he court is not yet convinced that published fee surveys reliably reflect rates actually billed and not rates that surveyed lawyers have artificially inflated for the [compilation's] audience”).
A more straight-forward reading of § 79B recognizes that the judge has discretion as to the admissibility of the evidence, even where the evidence satisfies the preliminary statutory requirements. Such discretion necessarily encompasses consideration of the reliability or trustworthiness of the proffered evidence, the “general rationale underlying exceptions to the hearsay rule.” See Commonwealth v. Evans, 438 Mass. 142, 155, 778 N.E.2d 885 (2002), cert. denied, 538 U.S. 966, 123 S.Ct. 1763, 155 L.Ed.2d 521 (2003). Although § 79B's requirement that the offered publication be commonly used and relied upon is intended to ensure a certain level of reliability, a judge is not foreclosed by this statutory language from addressing reliability concerns where they are evident notwithstanding such use. See Shepherd v. American Broadcasting Cos., supra.
Consideration of the reliability of evidence offered pursuant to § 79B is especially warranted where, as here, the proffered evidence contains complex, extrapolated data rather than simple objective facts. See Mazzaro, supra at 652, 363 N.E.2d 509 (“We think ‘compilation,’ as used in the statute, connotes simple objective facts, and not conclusions or opinions”). See also White Indus., Inc. v. Cessna Aircraft Co., 611 F.Supp. 1049, 1069 (W.D.Mo.1985) (“the kinds of publications contemplated by the [analogous Federal] rule are those which deal with compilations of relatively straightforward objective facts not requiring, for their statement, a subjective analysis of other facts ” [emphasis in original] ); 4 C.B. Mueller & L.C. Kirkpatrick, Federal Evidence § 8:101 (3d ed.2007) (noting that “[t]he commercial list exception is actually narrow” but “sometimes it reaches even more elaborate empirical data of specialized sorts”). To be sure, we do not suggest that all publications that contain data derived from the application of complex methodologies to raw data are categorically excluded from § 79B. Rather, we recognize that judges may consider whether particular compilations are sufficiently trustworthy to be admissible in the circumstances as an exception to the hearsay rule. See 5 J.B. Weinstein & M.A. Berger, Weinstein's Federal Evidence, § 803.19 (J.M. McLaughlin ed., 2d ed. 2013) (“there are publications relied on by the public or individuals in particular occupations that are not concerned with simple objective facts. Whether such a publication meets the requisite standard of trustworthiness entitling it to hearsay exemption must be determined on a case by case basis”).
Turning to the present case, the judge had before him the transcripts from the Davekos trial and Seare's affidavit, the latter largely duplicative of the Davekos trial testimony. This evidence demonstrated that Ingenix is a sister company to one of the largest insurance providers in the country; it relies on the voluntary submission of data on medical costs from the limited universe of insurance companies who choose to participate in the program; it applies a proprietary relative value and conversion factor to the raw data; and it has never verified that the data produced as a result of this formula accurately correspond with actual charges for medical procedures.
In other words, the data contained in the Ingenix database derives from raw data that is voluntarily submitted by participating insurance companies, not fully verified, and to which Ingenix applies its proprietary methodologies. On a largely identical record, the Appellate Division of the District Court held “there is nothing in the record to establish the accuracy or reliability of Ingenix's raw data and, thus, its statistical extrapolations.” Davekos, supra at 34. In addition to this evidence, the judge had before him evidence that the New York Attorney General conducted an investigation of Ingenix and found that the “rates produced by Ingenix were remarkably lower than the actual cost of typical medical expenses.” On such a record, we cannot say that the judge abused his discretion in excluding the evidence.
Because we conclude that the judge did not abuse his discretion in determining that the evidence was not admissible under G.L. c. 233, § 79B, we need not reach NEPT's additional arguments that Liberty Mutual failed to preserve the issue for review or that Liberty Mutual failed to comply with the procedures of G.L. c. 90, § 34M, as required of an insurance company when refusing to pay a bill for medical services.
See Cruz v. Commonwealth, 461 Mass. 664, 670, 963 N.E.2d 1172 (2012).
In considering the admission of the Ingenix data, the judge stated that the Appellate Division of the District Court “dealt squarely with the issue of admissibility of records from Ingenix” in Davekos and denied the admission of the data based on that ruling. To the extent that this creates any question as to whether the judge mistakenly believed that he was without discretion and bound to deny the admission of the evidence because of the decision in Davekos, we note that the evidence offered by Liberty Mutual consisted primarily of the trial transcript from Davekos. The affidavit of Susan Seare, the only additional piece of evidence submitted by Liberty, was largely duplicative of testimony in Davekos and offered little to augment the reliability of the Ingenix data. Although Davekos, supra, concerned admissibility under the business records exception, G.L. c. 233, § 78, the court broadly concluded that the Ingenix data was inadmissible because the record failed to establish that the methodologies employed by Ingenix produced reliable data. Id. at 34. Because reliability is also, as previously discussed, a significant criterion for admissibility of evidence under G.L. c. 233, § 79B, the trial judge here appropriately could and did rely on the Davekos decision in declining to admit any Ingenix database evidence.
Judgment affirmed.