Werth v. Hill-Rom, Inc.

MEMORANDUM OPINION AND ORDER RICHARD H. KYLE, District Judge. INTRODUCTION On January 21, 2008, M.W. was born at Mercy Hospital in Coon Rapids, Minnesota, to Plaintiffs Justin and Melissa Werth. Shortly after his birth, M.W. was prescribed supplemental oxygen therapy due to a respiratory condition; he was immediately moved to a nursery, where his head was was placed under an oxygen hood and he was laid in a bassinet under a baby warmer, Stabilet model 1250A, manufactured by Defendant Hill-Rom, Inc. (“Hill-Rom”). Early the following morning, a nurse was checking M.W.’s vital signs, and as she leaned toward him with a stethoscope, a fire erupted in the bassinet. Nurses quickly removed M.W. and extinguished the flames, but not before he had suffered severe burns. Plaintiff Allina Health System (“Allina”), Mercy Hospital’s owner, assembled a team of five experts to determine the fire’s cause. After conducting a time-consuming (and expensive) investigation, the team opined that a hot particle had broken from the baby warmer and fallen into the bassinet, igniting the flammable materials (blankets and the like) below. Together, the Werths and Allina (collectively, “Plaintiffs”) then commenced this action against Hill-Rom, asserting claims for strict products liability, breach of warranty, and negligence. With discovery complete, Hill-Rom now moves to exclude the causation opinion offered by Plaintiffs’ experts, pursuant to Federal Rule of Evidence 702 and Daubert v. Merrell Dow Pharmaceuticals, Inc., 509 U.S. 579, 113 S.Ct. 2786, 125 L.Ed.2d 469 (1993). For the reasons that follow, its Motion will be granted.1 BACKGROUND I. The incident baby warmer (Hill-Rom Stabilet Model 1250A) Allina purchased the incident baby warmer from Hill-Rom in 1994. (Doc. No. 94, Ex. F.) It is in the shape of an inverted “L.” (Id. Ex. A at 6.) The vertical portion of the “L” is referred to as the “column,” and a “heater element assembly” (sometimes referred to as the “warmer head”) extends perpendicularly from the top of the column, radiating heat downward toward a bassinet below. (Id.) The side of the heater element assembly closest to the vertical column is referred to as the “column side,” while the opposite end is referred to as the “front side.” For purposes of the instant Motion, the key components of the heater element assembly are the “heating element” and its incorporated and adjacent parts. The heating element is a hollow, cylindrical quartz tube shaped much like a fluorescent light bulb. (Id. at 172-73.) Passing through the quartz tube is a wire that heats the tube when connected to electricity. (Id.) The heat emitted from the tube rises upward and is then reflected down to the bassinet by a metal heat shield (or “reflector”), which sits above the heating [1054] element and is shaped like an upside-down “U.” {Id. at 172,179.) On each end of the quartz tube is a ceramic “end cap.” One side of the end cap is open (into which the end of the quartz tube is inserted), and the other side of the end cap contains a spring-loaded “male” electrical contact. {Id. at 172-74.) The quartz tube and its end caps are inserted into electrical contacts embedded in the heater element assembly’s housing, resulting in the tube being held in place above the bassinet by tension from the springs in the end caps. {Id.) In addition, the heating element has a slight downward orientation from the front side to the column side-that is, the front side is slightly higher than the column side. The quartz tube, therefore, rests inside the warmer head on an angle approximately 6° below horizontal. (Kemal Dep. at 112-14.)2 Records from the U.S. Food and Drug Administration disclose that certain Hill— Rom baby warmers, including the type at issue here, were recalled in the early 1990s. (Doc. No. 94, Ex. A at 166.) In several instances, hot particles were noted to have fallen from the heating element into the bassinet below, causing burns or melting holes, although no injuries or fires were reported. {Id.) As a result, in May 1994, Hill-Rom modified the design of its baby warmers to include protective “end covers.” End covers are small metal plates with vertical “lips” that attach to the heater element assembly, resting below the quartz tube’s end caps. {Id. at 167; Kemal Dep. at 75-76.) The end covers are intended to catch particles that might break off the end of the quartz tube. No similar incidents have been reported after this modification, and it is undisputed that these end covers were installed on the baby warmer at issue in this case. (Doc. No. 94, Ex. A at 167; Doc. No. 128, Ex. 11; Kemal Dep. at 76.) II. The actions following the fire Immediately after the fire, hospital staff secured the incident baby warmer, bassinet, and all other items being used by M.W. (blankets, IV tubes, oxygen hood, etc.), locking everything in a secure room with controlled access. (Doc. No. 94, Ex. A at 2.) A short time later, the Coon Rapids Fire Department and the Office of State Fire Marshal examined the scene, resulting in some of these items being moved. The conclusions reached by the Fire Department and State Fire Marshal (if any) are undisclosed in the record. Following that examination, however, Ahina re-secured the evidence and then retained a team of experts to develop an investigation protocol and determine the fire’s cause. {Id. at xxiii, 2.)3 [1055] The team first invited “all interested parties,” including Hill-Rom, to attend a physical inspection of the accident scene and the evidence. (Id. at xxiii.) As part of that inspection, hundreds of photographs were taken of the incident baby warmer and the items inside the bassinet, as well as the nursery where the fire occurred. Over the next several months, the team conducted a bevy of laboratory tests on both the incident baby warmer and exemplar warmers, including metallurgical testing, chemical analyses, microscopic examinations, and temperature and electronics testing. (See generally id. at 3-13.) The team also performed a number of experiments that included reconstructing the accident scene and testing the medical and electrical devices connected to the incident warmer (or exemplars thereof). (Id.) Ultimately, the team prepared a 296-page report (the “Report”), including dozens of appendices, detailing its examinations, findings, and conclusions. (Doc. No. 94, Ex. A.) In addition to gathering and examining the evidence, the team considered several hypotheses as to the fire’s cause. (Id. at xxiv.) It analyzed burn patterns in the bassinet to determine where ignition occurred and what ignition sources were present in that location. (Id.) It considered several possible sources, including electro-static discharge (“ESD”) from the nurses tending to M.W., malfunction of a medical device in the vicinity of M.W. (other than the warmer), unstable or reactive chemicals, and the incident warmer itself. (Id. at xxiv-xxv.) Because “[t]here was no surviving evidence or fire pattern to indicate conclusively which competing theory occurred,” the team performed tests and analyses to determine which it believed was the “most likely” cause. (Id. at xxv,) The team’s testing showed that the first three hypotheses could be excluded. All electronic medical devices in the vicinity of M.W. were tested and found to be functioning normally. (Id. at 49.) Thousands of ESD trials were performed to determine what level of energy would be required to ignite the materials in the bassinet when enriched with oxygen. Those tests showed that the maximum amount of energy that could be electrostatically discharged by one of the nurses was below the amount necessary to ignite the materials in the bassinet. (Id. at 158-59.) Chemical analyses of the bassinet’s contents were performed in an attempt to locate chemicals that may have ignited, but none were found. (Id. at 260-61.) The team also considered whether vapor from alcohol disinfectant on the nurses’ hands could have ignited, but the nurses denied using it and an examination of their hands the day after the fire revealed no singeing or injuries consistent with such an ignition. (Id. at 261.) Hill-Rom does not quarrel with any of these conclusions; indeed, it notes that they were tested “thoroughly and exhaustively.” (Def. Mem. in Supp. at 4.) The team next turned its attention to the incident baby warmer. It tested an exemplar warmer, observing that the quartz tube reached approximately 800°C at full power (as the incident warmer apparently had been set before the fire). It then considered five ways in which the incident warmer could have started the fire: (i) dust or lint coming into contact with the heating element, igniting, and fall [1056] ing into the bassinet; (ii) particles flaking off the heat shield (reflector), falling onto the heating element, and dropping into the bassinet; (iii) a heated piece of quartz breaking off the heating element and falling into the bassinet; (iv) a piece of hot or molten metal from the heating element dropping into the bassinet; and (v) heat from a defective crimp in the heating element’s electrical socket igniting nearby materials that, in turn, fell into the bassinet. (Doc. No. 94, Ex. A at xxv, xxxi.) The “dust or lint” hypothesis was excluded based on experiments showing that air tended to flow around the warmer head assembly while in operation, suggesting that dust and lint were unlikely to fall onto the heating element. (Id. at 249-51.) Hill-Rom does not challenge that conclusion. (Def. Mem. in Supp. at 4.) The team also excluded “heat-shield flakes” as a possible cause, as microscopic analysis showed that the largest flake missing from the heat shield “did not contain enough thermal mass to be a competent ignition source.” (Doc. No. 94, Ex. A at xxxi.) The “defective crimp,” too, was ruled an “unlikely” source, because analysis showed the crimp’s temperature was not sufficiently high to ignite nearby material and because its location in the heater was not “well aligned” with the bassinet to provide a path for a heated particle to fall. (Id.) What remained, then, were two theories: a hot piece of quartz from the heating element, or a hot piece of metal from inside or nearby it, falling into the bassinet.4 The team then performed “heat-transfer calculations” to determine the minimum particle size necessary to contain sufficient thermal energy to ignite the cotton bedding materials (the most flammable) in the bassinet. (Id. at 248-49.) It used a mathematical model to calculate the temperature of quartz particles after falling from the warmer head into the bassinet, a distance of approximately 30 inches. Based on those calculations, it determined that a spherical quartz chip 0.695 millimeters wide would reach the bedding at 390°C and contain sufficient thermal energy to ignite it. (Id. at 249.) Examination of the incident warmer’s quartz tube revealed that the front end was chipped in an area approximately 8 millimeters by 1-2 millimeters by 1/2 millimeters wide. In other words, the chipped area was “equivalent in volume to a sphere with a diameter of 1.1 mm” — • larger than a sphere 0.695 millimeters wide. (Id. at 244.) The team also noted that the ceramic end cap was 0.9 millimeters wider than the heating element, and hence “a gap of up to 0.9 mm can exist between the tube and the ceramic [end cap], providing potential resting spots for quartz chips [and] a pathway for chips ... to leave the ceramic end cap.” (Id. at 246.)5 Finally, the team observed that the quartz tube was oriented approximately 6° below horizontal, and hence “[a]ny quartz chips in the ceramic end cap would have a [1057] tendency to move downhill, which is in the direction of the open end of the cap.” (Id. at 247.) From these facts, the heat-transfer calculations, and the exclusion of other hypotheses for the fire’s origin, the team opined that “ignition due to a hot [quartz] particle originating from the ... warmer heater assembly and falling into the bassinet is the most likely ignition source for the ... incident.” (Id. at 264 (emphasis added).) III. Procedural history Armed with the expert team’s opinion, Plaintiffs commenced this action against Hill-Rom on June 3, 2010. In their Second Amended Complaint, they assert that the incident warmer was defectively designed and/or manufactured and failed to contain adequate warnings regarding the risks it posed (Count 1); that Hill-Rom breached its contract of sale to Allina (Count 2), express warranties that the warmer was safe (Count 3), and implied warranties that the warmer was of merchantable quality, safe, and fit for use (Count 4); and that Hill-Rom was negligent in the design, development, testing, manufacture, inspection, assembly, marketing, and distribution of the warmer (Count 5). The Werths assert that as a result of these failures, M.W. has suffered permanent injuries requiring ongoing medical care, pain and suffering, and diminished future earning capacity, all of which exceed $75,000. For its part, Allina asserts that (i) the fire damaged certain hospital equipment and prevented it from offering oxygen therapy in the subject nursery at Mercy Hospital, (ii) it was required to conduct an extensive (and expensive) investigation, and (iii) its reputation was damaged by the fire, all of which entitle it to damages exceeding $75,000. The parties then proceeded with discovery. Plaintiffs designated each member of the aforementioned expert team as a testifying witness in this case, as required by Federal Rule of Civil Procedure 26(a)(2)(A). Plaintiffs also produced an expert report in August 2011, pursuant to Federal Rule of Civil Procedure 26(a)(2)(B). In that report (the “Supplemental Report”), the expert team “adopt[ed] and incorporate[d]” the analyses and conclusions in its previously prepared 296-page Report, but also amplified the team’s opinion regarding the fire’s cause. While the initial Report opined only that “a hot [quartz] particle ... is the most likely ignition source for the ... incident” (Doc. No. 94, Ex. A at 264 (emphasis added)), the Supplemental Report opined, “to a reasonable degree of scientific and engineering certainty,” that the fire actually “was caused by ... a hot particle originating from the warmer heater element assembly falling into the bassinet and igniting the oxygen-enriched solid combustibles” (id. Ex. M at 11, 13 (emphasis added)). The Supplemental Report nowhere explained why it contained a more authoritative opinion on the fire’s cause than the initial Report; it contained no discussion of additional testing or analyses performed following the initial Report’s preparation. With more than a year of discovery now complete, Hill-Rom has moved to “exclude as unreliable the testimony at trial of Plaintiffs’ expert team ... concerning [its] untested and speculative hypothesis that a hot particle from a Hill-Rom infant radiant warmer was the root cause of the fire at issue.” (Doc. No. 74.) The Motion has been fully briefed, the Court held a hearing on March 29, 2012, and the Motion is now ripe for disposition. STANDARD OF REVIEW Federal Rule of Evidence 702 governs the admission of expert testimony. It provides: [1058] If scientific, technical, or other specialized knowledge will assist the trier of fact to understand the evidence or to determine a fact in issue, a witness qualified as an expert by knowledge, skill, experience, training, or education, may testify thereto in the form of an opinion or otherwise, if (1) the testimony is based upon sufficient facts or data, (2) the testimony is the product of reliable principles and methods, and (3) the witness has applied the principles and methods reliably to the facts of the case. The Court, acting as a “gatekeeper,” must evaluate whether proffered expert testimony satisfies this Rule, bearing in mind that the touchstone for admission is assistance to the trier of fact. See, e.g., Lee v. Andersen, 616 F.3d 803, 808 (8th Cir.2010); Larson v. Kempker, 414 F.3d 936, 941 (8th Cir.2005). “The proponent of the expert testimony must prove its admissibility by a preponderance of the evidence.” Lauzon v. Senco Prods., Inc., 270 F.3d 681, 686 (8th Cir.2001) (citation omitted). The Court enjoys “broad discretion” in determining whether expert testimony passes muster. Weisgram v. Marley Co., 169 F.3d 514, 518 (8th Cir.1999); accord, e.g., First Union Nat’l Bank v. Benham, 423 F.3d 855, 861 (8th Cir.2005) (“We give district courts great latitude in determining whether expert testimony meets the reliability requisites of Rule 702.”). Daubert set forth four “general” factors to be considered when evaluating the admissibility of expert testimony: (1) “whether [the expert’s theory] can be (and has been) tested”; (2) “whether the theory or technique has been subjected to peer review and publication”; (3) “the known or potential rate of error”; and (4) “general acceptance.” 509 U.S. at 593-94, 113 S.Ct. 2786. The Court stressed, however, that the inquiry must remain “a flexible one,” id. at 594, 113 S.Ct. 2786, and hence these four factors are neither mandatory nor exclusive. E.g., Shuck v. CNH Am., LLC, 498 F.3d 868, 875 (8th Cir.2007); Unrein v. Timesavers, Inc., 394 F.3d 1008, 1011 (8th Cir.2005) (“[A] court should use, adapt, or reject Daubert factors as the particular case demands.”). Other potentially relevant factors include “whether the expertise was developed for litigation or naturally flowed from the expert’s research; whether the proposed expert ruled out other alternative explanations; and whether the proposed expert sufficiently connected the proposed testimony with the facts of the case.” Presley v. Lakewood Eng’g & Mfg. Co., 553 F.3d 638, 643 (8th Cir.2009) (citations omitted). At bottom, “[t]here is no single requirement for admissibility,” as long as “the proffer indicates that the expert evidence is reliable and relevant.” Unrein, 394 F.3d at 1011. A court must remain cognizant that Rule 702 “reflects an attempt to liberalize” the admission of expert testimony and “clearly is one of admissibility rather than exclusion.” Polski v. Quigley Corp., 538 F.3d 836, 838-39 (8th Cir.2008). Yet, Rule 702’s ultimate goal “is to protect juries from being swayed by dubious scientific testimony.” In re Zurn Pex Plumbing Prods. Liab. Litig., 644 F.3d 604, 613 (8th Cir.2011). When expert testimony is “so fundamentally unsupported that it can offer no assistance to the jury,” it should be excluded. Id. at 614 (citation omitted). ANALYSIS Hill-Rom argues that Plaintiffs’ causation theory — a chip broke from the front end of the warmer’s quartz tube and landed in the bassinet, setting the bedding materials ablaze — is “untested speculation.” (Def. Mem. in Supp. at 6.) It points out that despite spending nine months (and a million dollars) preparing a “colossal” and “meticulous” 296-page report, [1059] Plaintiffs’ experts failed to test their theory to show (1) how a particle could break from quartz tube and (2) how such a particle could exit the end cap, avoid the end cover, and travel to the bassinet. Hill-Rom largely predicates its Motion on the size and orientation of the warmer’s components. It correctly notes that the ceramic end cap is only 0.9 millimeters wider than the quartz tube, and it contends it is “absurd to suggest that ... a large, irregularly shaped chip” — at least 0.695 millimeters wide, as was necessary to ignite the bassinet materials — “could have squeezed through this gap.” (Def. Mem. in Supp. at 8.) At bottom, it argues: Plaintiffs’ [experts’ conclusion as to the origin of the fire includes the hypothesis that a quartz chip, at least.695 mm in diameter, broke from the tip of the quartz tube, squeezed through a gap between the ceramic end cap surrounding the end of the quartz tube, made it out onto the protective end cover, [and] somehow fell into the bassinet, ... [but the] [e]xperts do not even attempt to provide a reasoned explanation of how this could have occurred ... [a]nd they made no attempt to test [it]. (Def. Reply at 7.) Plaintiffs respond that their experts followed the protocol in the National Fire Protection Association (“NFPA”) publication NFPA 921: Guide for Fire and Explosion Investigations, which involves evaluating, testing, and excluding possible fire causes by a process of elimination.6 The team reconstructed the fire scene and conducted tests to determine where the fire started in the bassinet and how it spread. It conjured several fire-cause hypotheses, testing them one by one until only one remained: a particle broke from the quartz tube and dropped into the bassinet. By applying this “generally accepted scientific method of process of elimination,” Plaintiffs argue, their experts’ opinion is admissible. (See PI. Mem. in Opp’n at 26-36.) The Court does not agree. I. The experts failed to disclose their reliance on NFPA 921 The most fundamental problem with Plaintiffs’ argument is that nowhere in the nearly 300-page Report or the Supplemental Report is NFPA 921 ever mentioned. Nor did any of the experts mention that methodology in his deposition. Plaintiffs simply assert in their brief that their experts followed this well-recognized methodology without proffering any evidence to support it. (See, e.g., PL Mem. in Opp’n at 28 (contending, without citation to the record, that the expert team “investigated this fire in accordance with NFPA 921”).) Under the Federal Rules of Civil Procedure, the experts were required to disclose in a written report “a complete statement of all opinions” they would offer at trial “and the basis and reasons for them.” Fed.R.Civ.P. 26(a)(2)(B)(i) (emphasis added). And when “a party fails to provide information ... as required by Rule 26(a) ..., the party is not allowed to use that information ... at a trial, unless the failure was substantially justified or is harmless.” See Fed.R.Civ.P. 37(c)(1). Plaintiffs have not argued that their failure to cite NFPA 921 was substantially justified; indeed, they have offered no explanation for that failure at all. Nor does the Court believe that their failure was harmless. Rule 26(a)(2)’s purpose is to eliminate “unfair surprise to the op [1060] posing party.” Sylla-Sawdon v. Uniroyal Goodrich Tire Co., 47 F.3d 277, 284 (8th Cir.1995). The requirement to provide a report containing all bases for an expert’s opinion is intended to permit opposing counsel to effectively prepare to depose the expert in advance of trial. E.g., Ciomber v. Coop. Plus, Inc., 527 F.3d 635, 642 (7th Cir.2008) (emphasis added); Allstate Ins. Co. v. Electrolux Home Prods., Inc., 840 F.Supp.2d 1072, 1079-80, 2012 WL 13512, at *6 (N.D.Ill. Jan. 4, 2012) (“The disclosure requirements prevent putting counsel in a position where he or she must depose an expert without an understanding as to what the expert will testify.”); Silverstein v. Procter & Gamble Mfg. Co., 700 F.Supp.2d 1312, 1320 (S.D.Ga.2009); Reed v. Binder, 165 F.R.D. 424, 430 (D.N.J.1996) (“Nothing causes greater prejudice than to have to guess how and why an adversarial expert reached his or her conclusion.”); see also Fed.R.Civ.P. 26, advisory committee’s note ¶ 2 (1993 amendments). Here, Hill-Rom was deprived of the opportunity to question Plaintiffs’ experts about their adherence to the specific (and detailed) requirements of NFPA 921, which was raised for the first time in Plaintiffs’ Opposition to the instant Motion. This is hardly harmless. Simply put, the experts’ failure to disclose their reliance on NFPA 921 — in their initial Report, their Supplemental Report, their depositions, or at any other point in discovery — would alone justify excluding their opinion. II. The experts did not apply NFPA 921 reliably Even if the experts had properly disclosed their reliance on NFPA 921, this would not automatically render their opinion admissible. Although NFPA 921 has achieved general acceptance in the fire-investigation community, an expert purporting to invoke it must show he applied the methodology reliably to the facts of the case. Fed.R.Evid. 702; Presley, 553 F.3d at 645. In other words, “[t]he mere assertion that [an expert] applied ... NFPA 921 is insufficient to establish that [his opinion] is based on the methods and procedures of science rather than on subjective belief or unsupported speculation.” United States v. Zhou, No. 06-cr-286, 2008 WL 4067103, at *6 (D.N.J. Aug. 25, 2008) (internal quotation marks, citations, and alteration omitted). In the Court’s view, Plaintiffs’ experts did not apply NFPA 921 reliably here. A. The lack of physical testing A key step under NFPA 921, once an investigator has developed a hypothesis about a fire’s cause, is testing that hypothesis. NFPA 921 § 4.3.6.7 Yet, nothing in the Report or the Supplemental Report indicates that the team tested its hypothesis at all. Indeed, as Bruley candidly acknowledged in his deposition, it never did. (Bruley Dep. at 187; see also Kemal Dep. at 175.) Plaintiffs argue that the team did perform testing, noting inter alia tests conducted on the medical devices present at the fire scene, lab experiments to determine the ignition properties of the bassinet’s contents, thousands of ESD tests to ascertain how much energy was necessary to achieve ignition, and airflow tests to determine the way in which air and particles moved around the warmer head. (PI. Mem. in Opp’n at 7-17.) These arguments miss the mark. The crux of Hill-Rom’s Motion concerns not the tests performed to rule out certain hypotheses; indeed, Hill-Rom acknowledges the experts “thoroughly and exhaustively tested” those the [1061] oríes. (Def. Mem. in Supp. at 4.) Rather, it takes issue with the team’s failure to physically test the theory it ruled in: the quartz-particle hypothesis. The Report gives no indication that the experts performed physical tests to show that it was likely, or even reasonable, to conclude a large chip8 could break from the quartz tube and fall into the bassinet, despite means to prevent that from occurring (the ceramic end cap and protective metal end cover). Such testing was critical, because NFPA 921 requires an investigator to determine (1) “the circumstances that brought the ignition source [the quartz chip] in contact with the first fuel ignited [the bassinet materials],” NFPA 921 § 18.6.3, and (2) “[h]ow safety devices and features designed to prevent fire from occurring [the end cap and end cover] failed to operate,” NFPA 921 § 18.4.4.2.1(5). See also NFPA 921 § 18.4.4.2.1(3), (6). There is no reference in the Report, for example, to testing done to determine the amount of force necessary to break such a large piece of quartz from the tube. There is no mention of testing to show, if chipping were to occur, whether the fragment likely would be expelled from the end cap or simply fall into it. Nor is there any indication of testing to determine (1) the amount of force with which a chip would be expelled and its expected trajectory or (2) if a chip were to fall into the end cap, how much force would be required to move it and whether such movement likely would lead the chip past the end cover and into the bassinet. At bottom, the record contains no indication of testing to support the opinion that a large fragment broke from the quartz tube and somehow reached the bassinet. Plaintiffs note that the Report does, in fact, mention ways in which a quartz chip could have traveled to the bassinet. (PL Mem. in Opp’n at 21-24.) They point to a two-page section of the Report indicating that chips could break from the quartz tube when it is installed and could break off during the warmer’s operation. (Doc. No. 94, Ex. A at 246-47.) They also note that the Report mentions the 0.9-millime-ter width difference between the end cap and the quartz tube, “providing potential resting spots for quartz chips” that, in turn, could move by gravity (since the tube and end caps were on a 6° downward angle) toward the cap’s opening. (Id.) But as the emphasized language above indicates, the Report simply describes ways in which a chip could have broken free of the quartz tube and made its way to the bassinet; it is replete with possibilities and conjecture. (See id. (“Chips could be broken off the quartz tube during installation.”) (“[H]ot quartz chips may also break off during operation the heater.”) (“Experience has shown it is possible to unintentionally install the heater element [incorrectly].”) (“[IJf the quartz tube were to chip, the ceramic end cap might shift.”) (“A cap of up to 0.9 mm can exist.”) (“Any chips exiting the ceramic end cap could potentially drop through.”) (emphases added).)9 Such speculation [1062] how a chip, in theory, could have broken from the tube and traveled to the bassinet is simply not enough. As one court has stated, “the function of expert testimony is to explain how something happened, not to speculate as to how something could possibly have happened.” Estate of Groff v. Aquila, Inc., No. 4:05-cv-0250, 2007 WL 4644707, at *10 (S.D.Iowa Sept. 28, 2007); accord, e.g., Glastetter v. Novartis Pharm. Corp., 252 F.3d 986, 989 (8th Cir.2001) (“[T]he district court’s gatekeeping role separates expert opinion evidence based on ‘good grounds’ from subjective speculation that masquerades as scientific knowledge.”); Concord Boat Corp. v. Brunswick Corp., 207 F.3d 1039, 1057 (8th Cir.2000) (“Expert testimony that is speculative is not competent proof.”) (internal quotation marks and citations omitted); Johnson v. Avco Corp., 702 F.Supp.2d 1093, 1108-09 (E.D.Mo.2010) (“We do not allow witnesses to tell juries that in their expert opinion something happened simply because it is possible.”). And notably, NFPA 921 itself provides that “[sjpeculative information cannot be included” in a fire-cause analysis. NFPA 921 § 18.6.5. The Supplemental Report did not rectify this problem. True, there the experts opined that a dislodged quartz chip was the fire’s cause, rather than speculating that this could have been how the fire occurred. (Doc. No. 94, Ex. M at 13.) Yet, the Supplemental Report nowhere explained why it contained a more definitive opinion than the initial Report, which is particularly problematic given that the Supplemental Report simply “adopt[ed] and incorporate[d]” the initial Report’s reasoning, including its speculation how a chip broke from the quartz tube and traveled to the bassinet. {Id. at 11.) The Supplemental Report, for example, contains no suggestion that additional testing was performed (following the initial Report) to attempt to show that a fragment breaking from the quartz tube was likely to travel to the bassinet.10 Without any evidence or other explanation for the greater level of confidence, the Court concludes that the Supplemental Report is both contrary to NFPA 921 § 18.7.4 (“If the level of certainty of the opinion is only ‘possible’ ..., the fire cause is unresolved and should be classified as ‘undetermined.’ ”) and nothing more than the experts’ litigation say-so, see Gen. Elec. Co. v. Joiner, 522 U.S. 136, 146, 118 S.Ct. 512, 139 L.Ed.2d 508 (1997) (“[Njothing in either Daubert or the Federal Rules of Evidence requires a district court to admit [1063] opinion evidence that is connected to existing data only by the ipse dixit of the expert.”). The Court recognizes, of course, that expert testimony inherently involves some amount of educated guesswork. Grp. Health Plan, Inc. v. Philip Morris USA, Inc., 344 F.3d 753, 760 (8th Cir.2003). But while a degree of “speculation is necessary [and] an even greater amount is permissible (and goes to the weight of the testimony), ... too much is fatal to admission.” Id. Here, Plaintiffs’ experts never attempted to validate their theory by testing how a large chip could break from the quartz tube, escape the end cap and end cover, and land in the bassinet; they simply theorized that this might have happened. It is undisputed that exemplar warmers were available to them and that such testing could have been performed. (Kemal Dep. at 111 (“Q: Is there any way to test the likelihood that a quartz chip ... would have escaped from the ... end cap? A: There is — the dimensions are there.”); see also id. at 118-20, 152-53; but see Caligiuri Dep. at 75.) And perhaps most importantly, the absence of physical tests stands in stark contrast to the extensive testing the team did undertake with regard to other potential causes for the fire.11 At bottom, the Court believes the team’s failure to physically test its hypothesis undermines the reliability of its opinion and renders it too speculative to admit. See Daubert, 509 U.S. at 590, 113 S.Ct. 2786 (“Proposed [expert] testimony must be supported by appropriate validation.”); Solheim Farms, Inc. v. CNH Am., LLC, 503 F.Supp.2d 1146, 1150 (D.Minn.2007) (Magnuson, J.) (excluding as unreliable expert opinion concerning fire’s cause when expert “failed to perform any tests to verify his causation theory”); Am. Family Ins. Grp. v. JVC Ams. Corp., Civ. No. 00-27, 2001 WL 1618454, at *4 (D.Minn. Apr. 30, 2001) (Doty, J.) (excluding opinion that device caused fire because, “[m]ost significantly, although an exemplar of the [device] was provided for his use, [the expert] failed to perform any testing specific to his theory on the origin of the fire”). B. The experts’ so-called “cognitive testing” Plaintiffs contend that instead of conducting physical tests, their experts performed “cognitive testing” — a “thought experiment” — on their theory. (PL Mem. in Opp’n at 32-36 & n. 7.) To be sure, NFPA 921 § 4.3.6 provides that “[a] hypothesis can be tested either physically by conducting experiments or analytically by applying scientific principles in ‘thought experiments.’ ” But courts have recognized that physical testing plays an important role in validating a theory under NFPA 921, even if such testing is not required. See, e.g., Presley, 553 F.3d at 644-45 (“NFPA 921 suggests that fire theories involving an appliance be substantiated by testing of exemplar appliances,” although there exists no “bright-line rule” mandating such [1064] testing); Fireman’s Fund Ins. Co. v. Canon U.S.A., Inc., 394 F.3d 1054, 1058 (8th Cir.2005) (“NFPA 921 requires that hypotheses of fire origin must be carefully examined against empirical data obtained from fire scene analysis and appropriate testing.”) (emphasis added); United Fire & Cas. Co. v. Whirlpool Corp., No. 5:10-cv-199, 2011 WL 4375049, at *3 (N.D.Fla. Sept. 20, 2011) (“Although testing on exemplars is not required, in an area such as the fire sciences, testing is an important way to show reliability.”); Fireman’s Fund Ins. Co. v. Tecumseh Prods. Co., 767 F.Supp.2d 549, 555-56 (D.Md.2011); Meemic Ins. Co. v. Hewlett-Packard Co., 717 F.Supp.2d 752, 763-64 & n. 8 (E.D.Mich.2010); Chester Valley Coach Works, Inc. v. Fisher-Price, Inc., No. 99-CV-4197, 2001 WL 1160012, at *10-11 (E.D.Pa. Aug. 29, 2001). In the Court’s view, such testing was particularly important here, given that the experts’ causation theory (ignition from a hot quartz particle) apparently has never occurred since Hill— Rom modified the warmer’s design to include protective metal end covers in 1994. Nor does the record contain any indication of published studies or other literature substantiating the experts’ theory. See Daubert, 509 U.S. at 593, 113 S.Ct. 2786. In any event, while “cognitive testing” might provide a sufficient foundation for expert testimony in appropriate circumstances, here it falls short. The cognitive testing alluded to by Plaintiffs is their experts’ heat-transfer calculations. (PL Mem. in Opp’n at 33-34.) Those calculations determined (1) the temperature of a quartz particle falling into the bassinet, (2) the amount of thermal energy necessary to ignite the materials there, and (3) the size of a heated quartz chip necessary to achieve such ignition. (Id. at 18-21.) The calculations, therefore, showed only that the hypothesized quartz particle would contain enough thermal energy to ignite the bassinet’s materials; they did nothing to show how a particle could have broken off the quartz tube in the first place or travel from the heating element to the bassinet.12 The heat-transfer calculations simply fail to “bridge the analytical gap between” the experts’ speculation and their ultimate conclusion. Fuesting v. Zimmer, Inc., 421 F.3d 528, 536 (7th Cir.2005), vacated in part on other grounds on rehearing, 448 F.3d 936 (7th Cir.2006); accord, e.g., Gaskin v. Sharp Elecs. Corp., No. 2:05-CV-303, 2007 WL 2572397, at *8 (N.D.Ind. Aug. 31, 2007) (expert’s opinion on fire cause excluded where expert “performed some deductive reasoning to support his hypothesis” but “did not conduct any scientific tests or experiments to bolster his theory”). In their brief and at oral argument, Plaintiffs suggested that their experts’ opinion should be admitted because Hill— Rom had not offered any alternative theory regarding the fire’s cause. This suggestion is misplaced for at least two reasons. First, Hill-Rom is under no obligation to prove how the fire started; rather, it is Plaintiffs’ burden to proffer admissible evidence on that question. E.g., Lauzon, 270 F.3d at 686. Second, the lack of an alternative theory does not mean that Plaintiffs’ theory is correct. The root cause of a fire often cannot be determined. See NFPA 921 §§ 18.6.5.1, 18.7.4. Plaintiffs acknowledged at oral argument that this was “a rare and probabilistic” (unlikely) fire. (3/29/12 Hear. Tr. at 14.) And, Plaintiffs’ experts could not definitively rule [1065] out certain other possible causes, such as ESD; at most, they could conclude only that those causes were “unlikely.” (See Doc. No. 94, Ex. A at 159.)13 C. Post-Motion discovery Also undermining the reliability of Plaintiffs’ experts’ opinion is what occurred after Hill-Rom filed the instant Motion. Due to a scheduling quirk, the Motion was filed on November 18, 2011, several weeks before Hill-Rom deposed the experts. (See Doc. No. 74.) The Motion made clear, in Hill-Rom’s view, that the experts had only “speculated that a small piece could have broken free [from the quartz tube], worked its way through various obstacles, and then dropped into the bassinet.” (Doc. No. 77 at 5.) In their subsequent depositions, however, the experts offered new theories how this could have occurred, in an apparent attempt to rectify the failings pointed out by Hill-Rom. For example, Kemal testified in his deposition that a “particle could be chipped and launched at the same time and, as a result, would fall out” of the warmer head. (Kemal Dep. at 115 (emphasis added).) Caligiuri, too, testified that “if you were to chip a piece of this quartz off, ... it wouldn’t just fall by itself under gravity; it would also be given some degree of momentum,” resulting in it being “ejected” from the ceramic end cap. (Caligiuri Dep. at 36, 61-67.) In fact, Caligiuri stated that he had “test[ed] the situation where [he] actually pushed on the quartz tube against the ceramic end cap, in an exemplar unit, and created a chip that came out of th[e] gap” between them. (Id. at 73.) This testimony is noteworthy in two respects. First, it highlights that despite their arguments to the contrary (see PI. Mem. in Opp’n at 35-36), Plaintiffs’ experts could have tested their theory—indeed, Caligiuri indicated that he did precisely that, albeit at some unknown time and under unspecified conditions.14 Second, there is no mention of chip “ejection” in the experts’ nearly 300-page Report, which posits only that a fragment could break from the tube and land in the end cap, possibly falling out at some later time. (Doc. No. 94, Ex. A at 246-47.) “Rule 26(a)(2) does not allow parties to cure deficient expert reports by supplementing them with later deposition testimony.” Ciomber, 527 F.3d at 642; accord, e.g., Williams v. Daimler Chrysler Corp., No. 4:06CV188, 2008 WL 4449558, at *5 (N.D.Miss. July 22, 2008); Hartford Ins. Co. v. Gen. Elec. Co., 526 F.Supp.2d 250, 252-53 (D.R.I.2007). Nevertheless, the fact that Plaintiffs’ experts tried (belatedly) to interject new explanations for how a chip could have broken from the quartz tube and traveled to the bassinet strongly suggests that they recognize an inherent flaw in their opinion.15 [1066] III. Other considerations A. Presley The Eighth Circuit has repeatedly affirmed the exclusion of expert testimony regarding the cause of fires under circumstances similar to those here. See, e.g., Presley, 553 F.3d at 643-47; Pro Serv. Auto., L.L.C. v. Lenan Corp., 469 F.3d 1210, 1214-16 (8th Cir.2006); Fireman’s Fund Ins., 394 F.3d at 1057-60. Of particular note is Presley, a case with remarkably similar facts. There, the plaintiffs had sued the manufacturer of a space heater following a fire in their home. 553 F.3d at 640. To determine the fire’s cause, they retained an expert who (1) personally examined the fire scene, (2) performed metallurgical testing on evidence collected there, and (3) had flammability tests performed using exemplar heaters. He used a “piecemeal” approach to determine the fire’s cause, purportedly in accordance with NFPA 921, and ultimately opined that a manufacturing defect in the heater had caused the fire. Id. at 641-42. While he did not conduct testing to confirm this theory — ostensibly because “it would be impossible to recreate the fire scene and his ignition scenario” — he asserted that he had verified each step of his theory “cognitively” based on observations, other tests he had performed, and basic principles of science. Id. at 642. The defendant moved to exclude the expert’s testimony and the district court granted the motion, concluding that it lacked a “scientifically reliable basis.” Id. at 645. Of particular concern to the district court were (1) the expert’s failure to test his theory and (2) his contention that it would have been impossible to do so. Id. On appeal, the plaintiff argued that the district court had, in essence, required testing in order for the expert’s opinion to be admissible. The Eighth Circuit disagreed and affirmed. It noted that although the district court had referred to the expert’s lack of testing, “mere reference to this ... does not establish the district court was requiring [the expert] to test. Instead, the court properly weighed [the expert’s] lack of testing in assessing the reliability of [his] opinions ” Id. (emphasis added). The court of appeals also noted that the expert had failed to reliably apply NFPA 921 because, among other things, he had failed to physically test his theory. Id. And “[although plaintiffs contended] [that the expert’s] observations, the metallurgical tests, the ... flammability tests, and the references to NFPA 921 sections [we]re adequate bases for reliability,” the appellate court held that “it was not an abuse of discretion for the district court to find [the expert’s] theory required too great an inferential leap from these bases.” Id. at 646. The parallels between this case and Presley are striking. There, as here, a fire purportedly was caused by a defective heater. There, as here, an expert examined the fire scene and performed laboratory tests and analyses on the evidence and exemplar heaters. There, as here, the expert purported to follow NFPA 921 in his analysis. There, as here, the expert did not physically test his causation hypothesis, asserting that it would have been impossible to do so. And there, as here, the expert claimed that he had “cognitively” validated his opinion. Nevertheless, the district court in Presley excluded the expert’s testimony, and the Eighth Circuit [1067] affirmed. The Court can find no principled basis to deviate from Presley here. B. Rule 403 Finally, even if Plaintiffs had satisfied their burden of showing their experts’ opinion was sufficiently reliable, admission of the opinion would not necessarily follow. As the Supreme Court highlighted in Daubert a judge “assessing a proffer of scientific testimony under Rule 702 should also be mindful of other applicable rules,” including Federal Rule of Evidence 403, which permits the exclusion of evidence “if its probative value is substantially outweighed by the danger of unfair prejudice, confusion of the issues, or misleading the jury.” 509 U.S. at 595, 113 S.Ct. 2786. This is of particular concern with expert testimony, which can be “both powerful and quite misleading because of the difficulty in evaluating it.” Id. (citation omitted); accord, e.g., United States v. Blade, 811 F.2d 461, 465 (8th Cir.1987) (noting that expert testimony enjoys an “aura of special reliability”); United States v. Hines, 55 F.Supp.2d 62, 64 (D.Mass.1999) (“[A] certain patina attaches to an expert’s testimony unlike any other witness; this is ‘science,’ a professional’s judgment, the jury may think, and give more credence to the testimony than it may deserve.”). For this reason, district courts enjoy even wider latitude than normal when determining whether an expert’s testimony should be admitted. Daubert, 509 U.S. at 595, 113 S.Ct. 2786 (“Because of this risk, the judge in weighing possible prejudice against probative force under Rule 403 ... exercises more control over experts than over lay witnesses.”). Given the unreliable basis upon which Plaintiffs’ experts’ predicate their causation opinion, the Court believes there exists great potential to mislead the jury here. See, e.g., United States v. Contentos, 651 F.3d 809, 820 (8th Cir.2011) (no abuse of discretion in excluding expert testimony with little probative value, as it would have resulted in “confusion of the issues [and] misleading the jury”); Tunnell v. Ford Motor Co., 330 F.Supp.2d 731, 741-42 (W.D.Va.2004) (excluding under Rule 403 expert opinion based on speculation). There can be little doubt that Plaintiffs’ experts hold impressive credentials, including two with doctorate degrees from Stanford University and a third who is a former NASA scientist. Moreover, the circumstances giving rise to this case are undoubtedly tragic — a baby badly burned less than a day after birth, in the very hospital where he was born. In light of these sad facts and the experts’ qualifications, it is not difficult to conceive that a jury would blindly accept Plaintiffs’ causation theory while overlooking the shaky foundation upon which it rests. It is the Court’s obligation to prevent that from happening. In re Zurn Pex, 644 F.3d at 613. CONCLUSION Based on the foregoing, and all the files, records, and proceedings herein, IT IS ORDERED that Defendant’s Motion to Exclude Expert Testimony (Doc. No. 74) is GRANTED, and Plaintiffs’ experts are precluded from opining at trial in this matter (or otherwise) that the fire in question was caused by a hot particle escaping from the incident baby warmer and falling into the bassinet. [1068] [[Image here]] [1069] [[Image here]] [1070] [[Image here]] [1071] [[Image here]] [1072] [[Image here]] [1073] [[Image here]] Footnotes [1] By granting Hill-Rom’s Motion, the Court believes it is unnecessary to reach Plaintiffs' own Daubert Motions (Doc. Nos.95, 99). [2] Although the undersigned has endeavored to explain the shape and orientation of the warmer's component parts, as the old saying goes, a picture is worth a thousand words. To better understand the heater element assembly, the reader is directed to the schematics attached to this Memorandum Opinion and Order. [3] The team comprised the following individuals: (1) Mark Bruley, Vice President for Accident and Forensic Investigation at ECRI Institute, an organization that analyzes medical-device failures. Bruley has a degree in biomedical engineering from Temple University and has investigated medical-device fires for over 30 years; (2) Dr. Robert Caligiuri, a metallurgical engineer with Exponent Failure Analysis Associates ("Exponent”). Caligiuri obtained a Ph.D. in material science from Stanford University and has spent 25 years investigating the root cause of failures in a variety of products; (3) Dr. Abid Kemal, who holds a Ph.D. in mechanical engineering from Stanford University and has experience in product failure investigations; (4) John Loud, an electrical engineer with Exponent who holds a Master's degree from San Jose State University. He has conducted investigations into numerous products allegedly causing [1055] fires; and (5) Barry Newton, a registered professional engineer employed by Wendell Hull & Associates, Inc. Newton has a mechanical-engineering degree from New Mexico State University and is a “world-renowned” authority on fire risks in oxygen-enriched environments. He has conducted fire investigations since 1988. Hill-Rom does not challenge the qualifications of any of these individuals. [4] While the Report indicated that a piece of metal falling into the bassinet was a potential cause for the fire, Bruley testified in his deposition that the only type of hot particle likely to have ignited the fire was a chip from the quartz tube. (Bruley Dep. at 118; see also Doc. No. 94, Ex. M at 16 (Plaintiffs' supplemental expert report noting that “the ignition of hot metal particles can be ruled out'').) The Court's analysis, therefore, is confined to the hot-quartz-particle portion of the team's opinion. [5] The quartz tube is not rigidly affixed in the ceramic end cap, but rather can move around inside it. Hence, it is theoretically possible for the gap between the tube and the end cap to be as wide 0.9 millimeters, although it could be narrower depending on how the tube is inserted into the end cap and moves thereafter. (Caligiuri Dep. at 59.) [6] The Eighth Circuit has recognized that NFPA 921 "qualifies as a reliable method” under Daubert. Fireman’s Fund Ins. Co. v. Canon U.S.A., Inc., 394 F.3d 1054, 1057-58 (8th Cir.2005). [7] Relevant portions of NFPA 921 are attached to Doc. No. 128 as Exhibit 12. [8] The spherical chip that Plaintiffs’ experts hypothesize caused the fire was at least 0.695 millimeters wide, meaning its volume was approximately 0.18 cubic millimeters. (This is calculated using the formula % Hr, where r represents the sphere's radius (half its diameter).) The volume of the chipped area on the front end of the tube was 0.7 cubic millimeters. (See Doc. No. 94, Ex. A at 244.) Hence, the hypothesized chip comprised more than 25% of the chipped area’s volume. Yet, pictures of the quartz tube's chipped front end suggest that it suffered several small chips rather than large ones. (See id. at 245.) [9] (See also Kemal Dep. at 91-92 ("We determined that it was possible for hot particles to drop — to drop into the bassinet.”); id. at 94 ("We think it is ... possible ... a hot particle ... could have been a competent ignition [1062] source.”); id. at 110 (“Any vibration could result in a [chip] coining out [of the end cap].”); id. at 114 ("Movement is still possible — there's still a possibility that due to movement and vibration, a piece could have come out.”); id. at 174-75 ("[T]he sum of all this told us that a hot particle ... could have originated from the head and dropped into the bassinet.”) (emphases added); Caligiuri Dep. at 35-36 (noting the Report contains "a discussion of what could happen” and "describes how a piece might fall out”) (emphases added).) This speculation carried over into Plaintiffs' brief (see PL Mem. in Opp’n at 21 ("[c]hipping of the quartz tube can occur during installation”); id. at 22 ("a particle of material could have escaped the ceramic end cap”); id. at 23 ("Even with protective end covers, the Team believes a particle can still fall directly from the heater element assembly into the bassinet.”) (emphases added)), and continued at oral argument {see TíITPIYI Hear. Tr. at 19 ("[T]here is a path by which particles can escape”); id. at 24 ("[W]e have a mechanism by which there can be a fractured surface.”)). [10] The Supplemental Report does indicate that Plaintiffs' experts reviewed certain additional documents after preparing the initial Report. {See Doc. No. 94, Ex. A at 35.) Those documents, however, appear to relate to Allina's maintenance of the incident warmer; there is no indication that they concerned chipping of the quartz tube or fragments making their way to the bassinet. [11] Kemal testified that it would have been extremely difficult and time-consuming to perform ignition tests on hot quartz particles, comparing it to "trying to light a campfire with flint stones. You have to attempt a large number of times.” (Kemal Dep. at 175-80.) Given the amount of time and money Plaintiffs’ experts expended ruling out other potential fire causes, it is difficult for the Court to accept this contention. Regardless, it is beside the point; Hill-Rom focuses not on whether a particle could have ignited the materials in the bassinet, but rather how the particle could have gotten there in the first place. (Def. Mem. in Supp. at 6 ("[H]ow would this quartz chip have made it into the bassinet?”).) And, it would not aid Plaintiffs’ cause even if their experts’ theory were, in fact, impossible to test: "[a]ny hypothesis that is incapable of being tested is an invalid hypothesis.” NFPA 921 § 4.3.6.1. [12] In fact, Caligiuri made clear in his deposition that there is an important distinction between "the potential for ignition” and "the creation of th[e] chip” in the first instance. (Caligiuri Dep. at 61.) [13] Notably, the experts’ ESD tests actually achieved ignition on a small number of occasions. (See Doc. No. 94, Ex. A at 156-57.) [14] Caligiuri actually contradicted himself on this point. He first testified: "I have chipped quartz tubes before, yes. Did I do it on a heating element quartz tube here? No.” (Id. at 71.) Shortly thereafter, he testified that he had "pushed on the quartz tube against the ceramic end cap, in an exemplar unit, and created a chip that came out of th[e] gap ... with kinetic energy.” (Id. at 73.) [15] The experts’ deposition testimony does not alter the Court’s analysis because it suffers from the same defect as the Report itself: it is unduly speculative. Kemal and Caligiuri both testified that ejection was simply one way a quartz chip could travel to the bassinet. (See, e.g., Caligiuri Dep. at 62-63 ("Q: [H]ow do you know that [ejection] happened here as opposed to [a] piece coming—that was already just lying at the bottom of the end cap? A: I don’t know that. I didn’t—I don’t know that, because it could also have been a piece that was just lying there, too. I’m just offer [1066] ing a scenario where apiece can be ejected as well as just fall out ”) (emphasis added); id. at 64 ("I think it’s a scenario that could happen.”); Kemal Dep. at 115 ("[T]he particle could be chipped and launched at the same time.”) (emphasis added).) Nor is there any evidence the “ejection” theory was reliably tested.