Johnson v. Washington Metropolitan Area Transit Authority

764 F. Supp. 1568, 1991 U.S. Dist. LEXIS 7297, 1991 WL 90611

• Court: District Court, District of Columbia
• Decided: May 22, 1991
• Docket: Civ. A. 86-3110-LFO
• Status: Published

Opinion

MEMORANDUM

OBERDORFER, District Judge.

This case arises out of the unfortunate death of Devora Johnson, the daughter of plaintiffs Eleanor and Franklin Johnson. On March 20, 1986, the decedent leaped from the platform in the lower level of the Metro Center station into the path of an oncoming Red Line subway car, the train struck her, and she died soon after. Although an Order of January 27, 1988 originally entered summary judgment in favor of defendant Washington Metropolitan Area Transit Authority (WMATA) and dismissed the complaint, the Court of Appeals affirmed in part, reversed in part, and remanded with instruction to consider whether

the results of the drug tests [given the train operator after the accident] are probative on the issue of whether the train operator behaved wantonly or merely negligently, and if so, whether that relevance is not substantially outweighed by the danger of unfair prejudice.

Johnson v. Washington Metropolitan Area Transit Authority, 883 F.2d 125, 130 (D.C.Cir.1989), cert. den., — U.S. -, 110 S.Ct. 1473, 108 L.Ed.2d 610 (1990). The Court of Appeals added that “[p]roba-tiveness must be considered in light of the evidence of the operator’s false testimony and WMATA’s failure to conduct further tests.” Id. The parties have fully briefed and argued this and other points. For the reasons stated below, the drug test results are both relevant to whether Dixon acted wantonly and admissible. Accordingly, an accompanying order will deny WMATA’s second motion for summary judgment motion and schedule this matter for trial.

I.

The Johnsons have essentially two claims. First, they contend that WMATA employees had a duty to prevent decedent from standing too close to the edge of the platform because she was acting in a strange and frightened manner. Cf. Weinberg v. Johnson, 518 A.2d 985, 988 (D.C.1986) (holding WMATA liable for the torts of its employees under the doctrine of re- spondeat superior). Second, they contend that once decedent was on the tracks, the train operator Keister Dixon could have stopped the train in time to avoid hitting her, but failed to do so because he was impaired by cocaine use at the time.

The original summary judgment order, as explained by a Memorandum filed February 5, 1988, determined that there was no evidence of anything that should have alerted WMATA employees to Devora Johnson’s intentions. The decedent did not tell anyone that she was going to leap in front of an ongoing train. See Memorandum of February 5,1988 at 17 [hereinafter, “Memorandum”]. Indeed, she did not appear unusually upset, although she did seem rather impatient and had a “very distant look.” See id. Most essentially, there was no evidence that she “was leaning into the track area so as to be in danger where she stood before she jumped.” Id. Accordingly, the Memorandum concluded that “[g]iven the evidence proffered, no reasonable jury could find that defendant was negligent because its employees failed to restrain plaintiffs’ decedent from jumping in front of the oncoming train.” Id. at 19; see also Celotex Corp. v. Catrett, 477 U.S. 317, 326, 106 S.Ct. 2548, 2554, 91 L.Ed.2d 265 (1986) (holding that summary judgment must entered “against a party who fails to make a showing sufficient to establish the existence of any element essential to that party’s case, and on which that party will bear the burden of proof at trial”). In the alternative, the Memorandum held that decedent “assumed the risk and was contribu-torily negligent by jumping into the path of an oncoming train.” Memorandum at 19 (citing Sinai v. Polinger Co., 498 A.2d 520, 525 (D.C.1985) and Restatement (Second) of Torts, § 496A, comment d (1965)). The Court of Appeals did not disturb these findings. See Johnson v. WMATA, 883 F.2d at 128.

The original summary judgment order likewise determined that there was not enough credible evidence to support defendants’ theory that Keister Dixon’s violation of his duty of care proximately caused decedent’s death. All the witnesses testified that decedent jumped when the train was no more than twenty feet away from her. See id. at 12. At that point, the train was traveling at least 15.65 miles per hour, and at that speed it would have needed at least 157 feet to stop. See id. Accordingly, the Memorandum concluded that “no reasonable jury could fail to conclude that the plaintiffs’ decedent proximately caused her own death by jumping in front of the oncoming train when it was no more than 20 feet away from her, and that it would have been impossible for defendant or the operator ... to stop the train after she jumped into the track and before the train struck her.” Id. at 11 — 12 (citations omitted).

In reaching this conclusion, the Memorandum dismissed the testimony of two witnesses as incredible. Ronald Thompson and Richard Louis Moore testified that the decedent was on the tracks for 5 to 15 seconds before impact. See id. According to plaintiffs’ expert, if decedent had been on the tracks for 10.3 seconds or more, a reasonably prudent train operator would have been able to stop the train in time and avoid hitting her. See id. at 13. The Memorandum did not deem this testimony to raise a genuine issue of fact because these time estimates are not only inconsistent with the testimony of all the other witnesses; they are also inconsistent with the distance estimates of Thompson and Moore. See id. at 14. Moore testified that the decedent was only ten to twelve feet from the front of the train when she jumped; Thompson testified that the train was no more than twenty feet away at the time. See id. If, however, the decedent had lain on the tracks for fifteen seconds, the train would have been more than a hundred feet away, and possibly not yet in the station, at the time decedent jumped. See id. Moreover, Thompson acknowledged that his time estimates were less reliable than his estimate of the distance “because of the way in which time seems to ‘slow down’ when an accident is witnessed.” Id. at 14 (citing Thompson Deposition at 25). The Memorandum then concluded that no reasonable jury would believe Thompson’s and Moore’s time estimates.

The Court of Appeals disagreed. According to the Court of Appeals, the internal contradictions in the testimony of Moore and Thompson did not render their testimony incredible, but merely demonstrated that there was a factual dispute over how far the train was away from decedent when she jumped. See Johnson v. WMATA, 883 F.2d at 128. In reaching this conclusion, the Court of Appeals relied upon the general principle that judges may set aside testimony as incredible only if it is both self-serving and undermined by other evidence. See id. at 128 — 29. Because both Thompson and Moore “are disinterested witnesses,” the Court of Appeals concluded that there was a “genuine issue as to whether the collision could have been avoided.” Id. at 129.

The Court of Appeals observed that “[t]his is not, however, the end of the matter.” Id. Because Devora Johnson was trespassing when she jumped onto the tracks, under District of Columbia law she and her successors in interest “may only recover for injuries ... that were willfully and wantonly inflicted.” Memorandum at 11 (quotation and quotation marks omitted); accord Johnson v. WMATA, 883 F.2d at 130. The primary evidence of wanton behavior presented by plaintiffs was that shortly after decedent’s death the train operator Keister Dixon tested positive for cocaine, and it was by no means obvious that those results were admissible into evidence.

The Memorandum briefly discussed the issue. It observed that plaintiffs’ expert Dr. Walter Monroe Booker had stated that “impairment could not be determined from any of Mr. Dixon’s drug test results.” Memorandum at 15 (citing Booker Deposition at 59 — 60, 64). The Memorandum then noted that even if Dixon had been under the influence of drugs at the time, “plaintiffs could still not prevail because no reasonable juror could conclude that any operator could have stopped the train between the time decedent jumped and the time she was struck.” Id. (emphasis in original). Finally, the Memorandum concluded that Dixon’s “use of drugs is either inadmissible (because of lack of expert testimony indicating whether he was impaired or affected by drugs at the time of the incident) or it would not support a verdict on proximate cause.” Id.

The Court of Appeals found it “ambiguous whether [the Memorandum] based summary judgment on a finding that the tests were inadmissible or primarily on its conclusion that even a sober operator would have been unable to stop the train in time.” Johnson v. WMATA, 883 F.2d at 130. The Court of Appeals further noted that in discussing the drug tests the Memorandum had not

considered the possible relevance of two other pieces of evidence: (1) the train operator’s presumably false testimony that he had not taken any drugs in the past three months, and (2) WMATA’s failure to have more thorough tests performed which would have indicated more about when the operator took the drugs, and whether he was impaired at the time of the accident.

Id. Accordingly, the Court of Appeals remanded with instructions to consider the admissibility of the drug test results in light of Dixon’s presumably false testimony and WMATA’s failure to conduct additional tests. See id.

II.

Before weighing the arguments for and against submitting the drug test results to the jury, it is first necessary to consider what the Johnsons plan to use those drug test results to prove. See Fed.R.Evid. 401 (judging the relevance of evidence according to its tendency to prove “any fact that is of consequence to the determination of the action”).

A.

While the drug test results are not relevant to or probative of the train’s location when the decedent fell to the tracks, the Johnsons contend that they are probative of whether Dixon violated his duty of care towards decedent. As the Court of Appeals recognized, the relevant standard is wantonness because decedent was a trespasser at the time of the accident and under District of Columbia law the plaintiffs “may, generally speaking, only recover from landowners for injuries that were willful, wanton, or that resulted from maintenance of a hidden engine of destruction.” Holland v. Baltimore & O.R. Co., 431 A.2d 597, 601 (D.C.1981) (en banc). For the most part, the District of Columbia has adopted a definition of wanton behavior that corresponds to the Restatement’s definition of “reckless disregard of safety.” See Copeland v. Baltimore & O.R. Co., 416 A.2d 1, 3 & n. 3 (D.C.1980) (citing Restatement (Second) of Torts § 500, at 587). Under this definition, wanton behavior is more than mere negligence or mistake. It is “characterized by extreme recklessness and utter disregard for the rights of others.” Safeway Trails, Inc. v. Schmidt, 225 A.2d 317, 320 (D.C.1967). Moreover, “it must be foreseeable to a reasonable person that the conduct will result in harm.” Copeland v. Baltimore & O.R. Co., 416 A.2d at 4.

Some jurisdictions, however, define wantonness differently when a landowner injures a trespasser. See Restatement (Second) of Torts § 500, at 586 (scope note). Under this definition, if a landowner is unaware of the presence of the trespasser, the landowner is wanton only if he or she acts with reckless disregard for the safety of others. However, if the landowner is aware of the trespasser’s presence, he must exercise ordinary care to avoid injuring the trespasser. See id. In Copeland, the D.C. Court of Appeals discussed this two-pronged definition of wantonness. Specifically, it cited an Illinois Supreme Court decision that “held that a railroad owes an undiscovered trespasser a duty to refrain from willfully or wantonly injuring him/her.” Copeland, 416 A.2d at 3 (discussing Bremer v. Lake Erie & W.R. Co., 318 Ill 11, 148 N.E. 862 (1925)). That same decision also held that “once the presence of a trespasser is known, the railroad must exercise ordinary care not to injure the trespasser.” Id. “This standard,” the D.C. Court of Appeals observed, “has become generally accepted.” Id. (citations omitted). From this passage, it would appear that the D.C. Court of Appeals has adopted, or would likely adopt, this two-pronged approach.

WMATA does not directly challenge plaintiff’s reading of Copeland. Instead, it argues that Copeland was overruled on this point by Holland v. Baltimore & O.R. Co., 431 A.2d 597 (D.C.1981). Specifically, WMATA points to the statement in Holland that trespassers may only recover for “intentional, wanton or willful injury or maintenance of a hidden engine of destruction.” Id. at 599. It is not, however, clear how this language indicates a rejection or revision of the definition of wantonness in Copeland. The quoted language simply indicates that a landowner’s behavior must be wanton, whatever wantonness may be. Moreover, Holland does not say it overrules Copeland. Indeed, it neither cites Copeland nor discusses the definition of wantonness. In such circumstances, it would be improper to infer that Holland overruled Copeland. “One of the basic duties of an appellate court is to signify when it is departing from precedent and to provide a legal analysis explaining why it is changing course.” United States v. Edmond, 924 F.2d 261, 267-68 (D.C.Cir.1991). By assuming that the D.C. Court of Appeals overruled a precedent without explanation, WMATA implicitly assumes that the D.C. Court of Appeals disregarded this basic duty. “Out of respect for the court of appeals,” such an assumption cannot be maintained without much more compelling evidence. Id. at 268.

Nonetheless, this Court is not free to employ Copeland’s two-pronged definition of wantonness. Our Court of Appeals clearly stated that if “the drug test results are not admissible, then summary judgment for WMATA would be appropriate due to the lack of evidence that WMATA breached the applicable standard of care.” Johnson v. WMATA, 883 F.2d at 130. If the two-pronged definition of wanton behavior were applied, there would be no need to show that Dixon operated the subway car while impaired. It would be enough that he knew that Devora Johnson was on the tracks and failed to exercise ordinary care to avoid her. This Court is not, however, free to ignore the clear instructions of the Court of Appeals. Ac-, cordingly, “wantonness” must be interpreted to mean reckless disregard for the safety of others.

B.

By any definition, if Dixon operated the train while under the influence of cocaine, his conduct was wanton. It is not only foreseeable that patrons will wind.up on WMATA tracks; WMATA trains its operators on how to react when that happens. See Goings Deposition at 35; see also Wood Deposition at 58 (noting that at least twenty people have been killed in collisions with WMATA subway cars). Use of cocaine either during or shortly before operating a train impairs an operator’s reactions, his judgment, and ultimately his ability to ensure the safety of subway patrons when, for example, they end up on the tracks. See Booker Deposition at 62; Bogema Deposition at 61; Wood Deposition at 40; Kimball Deposition at 21. For this reason, WMATA strictly prohibits the use of cocaine and other narcotics by train operators and other employees. See Handbook of Metrorail Operating Rules ¶ 34(a) (attached to Second Supplemental Response). Dixon knew about these regulations, see Dixon Deposition at 83, and through his training he should have known of the possibility that someone would end up on the tracks. As a consequence, if he took cocaine and operated the train while impaired, he was acting with reckless disregard to the safety of others that he knew, or should have known, would be endangered by his impairment.

Nevertheless, WMATA contends that the evidence of cocaine-induced impairment is not enough to prove wantonness. According to WMATA, to demonstrate Dixon’s reckless disregard for the safety of others the Johnsons must show that he was “intoxicated.” This argument is based upon a recent decision by the Court of Appeals of Maryland, Nast v. Lockett, 312 Md. 343, 539 A.2d 1113 (1988). In that case, the Maryland Court found that driving an automobile while impaired by alcohol was not wanton conduct per se. See id.., 539 A.2d at 1124. It does not, however, follow from this decision that a train operator impaired by cocaine must be intoxicated in order to be wanton. Just as it refused to determine that a person driving while impaired by alcohol is always acting wantonly, the Maryland Court of Appeals did not hold that someone who is driving while only impaired is never acting wantonly. Instead, the Maryland Court held that “[a]s the degree of impairment by the voluntary consumption of alcohol increases, the need for other aggravating circumstances lessens, and vice versa.” Id., 539 A.2d at 1122. In this case, a reasonable jury might find extremely aggravating circumstances. As mentioned above, WMATA trains its train operators on how to react to any patron who might fall onto the track, area. To comply with those instructions and avoid injuring patrons who ended up on the tracks, Dixon had to be alert and able to react quickly. If he were to do anything which he knew, or should have known, would impair his ability to do so, he would be putting the patrons of the subway at greater risk of injury. Thus, even under Nast v. Lockett, if there were evidence that Dixon ingested cocaine at some time relevant to the time of the accident, a reasonable jury could infer that Dixon’s conduct was wanton.

In the alternative, WMATA contends that Dixon cannot be found to have been wanton because it was not foreseeable “that a person would commit suicide by flinging herself in front of his train.” WMATA’s Second Motion for Summary Judgment at 29. This observation, however, misses the point. As mentioned before, at least twenty people have been killed by WMATA subway cars after ending up on the tracks. See Wood Deposition at 58 — 59. What caused them to be on the tracks is not the issue. Nor does the D.C. Court of Appeals’ decision in Copeland lend any support to WMATA’s contention. In Copeland, the D.C. Court of Appeals held that while a Baltimore and Ohio Railroad freight train had negligently failed to use headlights of the strength required by Government regulations, the railroad had not acted with reckless disregard for the safety of others because “the area of the tracks was an area on which the railroad could not have expected to find human activity (other than railroad personnel).” Id. at 4. By “area of the tracks” Copeland did not, however, mean the area bounded by the metal rails. Copeland referred instead to the area near enough the tracks that the train's operation could fore-seeably endanger someone within that area. Cf. Restatement (Second) of Torts § 500, comment d, at 589 (referring to a “zone of danger”). Thus, in Copeland the D.C. Court of Appeals contrasted the isolated freight train tracks, “separated from a nine-lane freeway by a wall and a guardrail,” with areas where human activity was likely such as a “a train yard, or area where children, employees, licensees, or trespassers were known to play or otherwise frequent.” Id. There can be little doubt that the Metro Center subway station is the latter sort of area and that it was foreseeable that a person would end up on the tracks.

C.

To summarize, according to our Court of Appeals, because the decedent was a trespasser at the time of the accident, the plaintiffs can only recover for her death if they prove that the train operator acted with reckless disregard for the safety of others. If, however, they can prove that he was impaired at the time of the accident by cocaine use, they will have shown that Dixon impaired his ability to avoid foreseeable injuries in reckless disregard of the safety of others.

III.

The drug tests results are the Johnsons’ primary evidence that Dixon was impaired by drug use at the time of the accident. As our Court of Appeals noted, for purposes of determining their admissibility, it must be assumed that the jury would find that Dixon had enough time to stop the subway car from hitting decedent, but failed to do so. See Johnson v. WMATA, 883 F.2d at 129. Similarly, all the evidence relating to or corroborating the drug test results must be interpreted in the light most favorable to the Johnsons. See generally 2 D. Louisell & D. Mueller, Federal Evidence. § 126, at 48 — 51 & n. 28 (1985). With these principles in mind, it will be useful to briefly summarize the drug test results and related evidence.

Dixon was given the drug tests on the day of the accident as part of his mandatory post-incident medical examination. 1 The tests detected the presence of the primary metabolite of cocaine and the secondary metabolite of marijuana in Dixon’s system. 2 If the results are assumed to be accurate, as they must be for purposes of this motion, Dixon must have ingested cocaine sometime during the previous week. 3 However, because the test was of the “screening type,” it only detects the presence of drug residues in the subject’s body, and no conclusions can be drawn from the test itself as to how much cocaine Dixon consumed, when he did it, or even whether he was impaired by cocaine use at the time of the test. 4

A neurological examination, by contrast, could have determined whether Dixon was impaired at the time of the examination. 5 It is, however, undisputed that even if such an examination established that Dixon was impaired, it alone would not have determined the cause of the impairment. 6 Moreover, neurological exams are not routinely given in the transportation industry: The drug tests given Dixon reflected contemporaneous and current government and industry standards as well as state-of-the-art technology. 7

The Johnsons contend, however, that there is other evidence that Dixon was impaired at the time of the accident. Dixon was apparently singing and whistling before the accident. 8 After the accident, he acted hysterically, moving around the train booth unable to “get his nerves together.” 9 However, neither the police officers who interviewed him right after the accident nor the doctors who examined him later that day noted any signs of drug use or impairment. 10 Moreover, Dixon denied having taken any cocaine during the first three months of 1986. 11

IV.

Now that wanton behavior has been defined and the evidence of such behavior in the record reviewed, it is appropriate to determine whether the drug test results are relevant. Under the Federal Rules of Evidence, “ ‘[r]elevant evidence’ means any evidence having any tendency to make the existence of any fact that is of consequence to the determination of the action more probable or less probable than it would be without the evidence.” Fed.R.Evid. 401. Under this definition, the drug results are clearly relevant. The positive drug test results are some evidence that Dixon may have been operating the train under the influence of, and impaired by, cocaine at the time of Devora Johnson’s death. Negative results would have excluded even that possibility.

Nevertheless, WMATA contends that the drug test results are not relevant because they do not make it probable that he was impaired at the time of the accident. The Johnsons do not, however, have to show that each piece of evidence they plan to present makes it more likely than not that Dixon was impaired at that time of the accident. As McCormack memorably put it, “A brick is not a wall.” McCormack on Evidence § 185, at 436 (2d ed.1972); see also 1 Wigmore on Evidence § 12, at 693 (P. Tillers ed. 1983) (stating that “[t]he quality of a complete demonstration” cannot be expected of each separate “evidential fact”). To be relevant, the evidence presented by the Johnsons need only make Dixon’s drug use more probable than it would be without that evidence.

Relying on Simonson v. White, 220 Mont. 14, 713 P.2d 983 (1986), WMATA also argues that “evidence of prior drug use is inadmissible without independent observation of impairment or some other objective evidence linking the drug test results to the accident.” WMATA’s Supplemental Reply at 4 (citing Simonson v. White). Nothing in Simonson v. White, however, requires “independent observation of impairment.” In that case, the Supreme Court of Montana merely required some evidence laying a foundation for the relevance of the drug use. See Fed.R.Evid. 103(b). 12 Specifically, that court found the relevance of the drug use to be conditioned upon evidence that defendant’s conduct caused the accident in question. See id., 718 P.2d at 988 (“If White’s negligence was not a cause of the accident, [the evidence of the use of drugs] should not be admitted.”). In this case, there is such evidence: As our Court of Appeals ruled, the testimony of Moore and Thompson that Devora Johnson was on the tracks for more than ten seconds before the subway car hit her is sufficient to send the question of proximate cause to the jury.

Coming at the same point from another angle, WMATA argues that “[p]roof of causal connection may not be supplied by an inference which itself rests upon another inference or presumption.” David v. Granger, 35 A.D.2d 636, 312 N.Y.S.2d 963, 965 (App.Div.1970). As Wigmore notes, this hoary maxim is deceiving at best:

There is no such orthodox rule, nor can there be. If there were, hardly a single trial could be adequately prosecuted. For example, on a charge of murder the defendant’s gun is found discharged. From this we infer that he discharged it, and from this we infer that it was his bullet that struck and killed the deceased. Or the defendant is shown to have been sharpening a knife. From this we argue that he had a design to use it upon the deceased, and from this we argue that the fatal stab was the result of this design. In these and innumerable daily instances we build up inference upon inference, and yet no court (until in very modern times) ever thought of forbidding it. All the departments of reasoning, all scientific work, every day’s life, and every day’s trial proceed upon such data. The judicial utterances that sanction the fallacious and impracticable limitation, originally put forward without authority, must be taken as valid only for the particular evidentiary facts therein ruled upon.

1 Wigmore on Evidence § 41, at 1111 (footnotes omitted); see also id. § 41, at 1120 & n. 4 (noting that this “fallacy has been frequently repudiated in judicial opinions” and listing cases). Once the “particular evidentiary facts” ruled upon in David v. Granger are analyzed, it becomes apparent that that case involves another question of conditional relevance. The defendant in David v. Granger was involved in a car accident. The plaintiffs sought to introduce a blood-alcohol test showing that defendant had alcohol in his blood at the time of the accident. The Court reasoned that while the test evidence would permit an inference that he was driving while impaired, it “would still remain to be established whether such impairment was a proximate cause of the accident.” David v. Granger, 35 A.D.2d at 637, 312 N.Y.S.2d at 965. Because “[n]o testimony was adduced at trial relating to the manner of operation of decedent’s car,” the Court found no evidence of proximate cause and set aside the jury’s verdict in favor of the plaintiff. Id. In this case, by contrast, our Court of Appeals has ruled that there is sufficient evidence upon which a reasonable jury could conclude that Dixon’s actions caused the accident. See supra 3 — 5. As a consequence, a foundation has been laid establishing the relevance of the drug test results.

V.

In the alternative, WMATA contends that the drug test results should be excluded due to their potential for prejudice and confusion. See Fed.R.Evid. 403. 13 Those results are, however, quite probative of Dixon’s impairment. Consequently, the danger of unfair prejudice or confusion does not “substantially outweigh[]” the probative value of the evidence, and the results must therefore be admitted. See id.

WMATA’s basic argument is that the results of the drug test do not establish that Dixon was impaired at the time of the accident because, while those results dem onstrate that Dixon ingested cocaine some time within a week or so before the accident, they do not indicate at what point within that week he did so. Thus, while Dixon could have been impaired by cocaine use at the time of the accident, he equally could have ingested the cocaine the weekend before the accident and been unimpaired at the time of the accident. As a consequence, WMATA contends, the drug test results have limited probative value.

While WMATA’s assessment of the probative value of drug test results in isolation is undoubtedly correct, its argument suffers from the same flaw as its relevancy argument. Simply put, there is no reason to consider the drug tests results in isolation. Indeed, it would be deceptive to do so: “Few categories of evidence indeed could ever be ruled admissible if each category had to stand on its own, unaided by the process of cumulating information that characterizes the way any rational person uses evidence to reach conclusions.” Tribe, Trial by Mathematics, 84 Harv.L.Rev. 1329, 1350. It is especially important to look to other information when, as here, multiple inferences can be drawn from a single piece of evidence. When such evidence is considered in isolation, there are many potential inferences, each equally possible. Other evidence can, however, exclude some of the possible inferences, increase the probability of others, and thereby show how probative a piece of evidence is in context. “In short, the balancing test of Rule 403 requires that the judge consider the proffered evidence against the background of all of the evidence in the case.” 22 C. Wright & K. Graham, Federal Practice & Procedure: Evidence § 5214, at 272 (1978).

When considered in the context of the other evidence presented by the John-sons — and when, for purposes of this motion, that evidence is viewed in the light most favorable to the Johnsons — the drug test results become quite probative of Dixon’s impairment at the time of the accident. Viewed in isolation, the drug test results suggest that Dixon may have used cocaine at any time during the previous week without rendering any of those potential inferences more probable than any other inference. The evidence presented by the John-sons, however, increases the probability of one of those inference. The fact (at least according to the Johnsons) that Dixon had more than ten seconds in which to stop the train yet failed to do so is consistent with cocaine use or cocaine-induced impairment. Similarly, a reasonable jury might find the fact that before the accident Dixon was singing and whistling as well as the fact that after the accident he acted hysterically to be consistent with cocaine-induced impairment. Indeed, a reasonable jury might find in Dixon’s denial that he used drugs in the first three months of 1986 to be corroborating evidence. If the drug test results are assumed to be correct, Dixon lied and a reasonable jury might infer that Dixon did so in order to protect himself from potential liability from, or even prosecution for, operating the train while impaired by cocaine. Thus, the corroborating facts supplied by the Johnsons, when viewed in the light most favorable to them, make the inference that Dixon was impaired at the time of the accident more plausible. By contrast, WMATA has failed to present any evidence corroborating any of the other possible inferences. In sum, while the inference that Dixon was impaired at the time of the accident is only one possible inference from the drug test results, the alleged evidence of Dixon’s negligence, his behavior before and after the accident, his apparent prevarication, and the lack of any more probable inference strengthens that inference and might, in a reasonable juror’s eyes, make it more likely than not that Davis was impaired at the time of the accident. .

WMATA argues that the John-sons’ corroborating evidence is contradicted by more credible evidence that Dixon was not impaired. Be that as it may, it would not be appropriate at this stage to consider the credibility of the Johnsons’ evidence vis-a-vis WMATA’s evidence. “Weighing probative value against unfair prejudice under Fed.R.Evid. 403 means probative value with respect to a material fact if the evidence is believed not the degree the court finds it believable.” Bowden v. McKenna, 600 F.2d 282, 284 — 85 (1st Cir.), cert. denied, 444 U.S. 899, 100 S.Ct. 208, 62 L.Ed.2d 135 (1979) (footnote and citation omitted); accord Western Industries, Inc. v. Newcor Canada Ltd., 739 F.2d 1198, 1202 (7th Cir.1984). If a judge were “to exclude evidence on the ground that he thinks it incredible,” he would not only be invading the jury’s function of assessing the credibility of witnesses; he would be doing so without the benefit of viewing either the demeanor of the witnesses or cross-examination by opposing counsel. 22 C. Wright & K. Graham, Federal Practice & Procedure: Evidence § 5214, at 265— 66. In other words, because it is the jury’s duty to resolve genuine issues of fact, the probativeness of a particular piece of evidence must be judged in terms of the John-sons’ reading of the disputed facts, not the opposition’s.

Viewed in this light, the drug test results are quite probative. Without the drug test results, the evidence of Dixon’s negligence, the singing and whistling, and of course Dixon’s testimony about his drug use ate scattered facts of little import. With the drug tests results, however, these scattered facts become important corroborating evidence in a cogent theory of Dixon’s impairment. The evidence of the drug test results is, therefore, quite probative.

The Johnsons contend that the probativeness of the drug test results is enhanced by WMATA’s presumption that any employee with a detectable level of cocaine in his or her urine or blood was working while impaired. The presumption is, however, a red herring. The only evidence of it is a single line in a transmittal memorandum stating that WMATA’s Substance Abuse Policy “defines minimum levels of substances which, when detected, presume impairment.” Memorandum from J. Potts to Officers, Office Directors, and Local 689, ATU, Represented Employees, January 8, 1985, at 1 (Plaintiff’s Supplemental Memorandum, Exhibit 1). That memorandum was in turn written in response to several arbitration rulings holding that WMATA could not discipline employees for drug use without evidence of impairment on the job. See Kelley Deposition at 41, 43. WMATA adopted the presumption to comply with those rulings and to make disciplinary actions based upon drug testing more palatable to the unions. See id. at 51 — 52. The presumption was in no way based upon either scientific evidence or practical experience. See id. at 51 — 52. As a consequence, evidence of the presumption does not make it “more or less probable” that Dixon was impaired at the time of the accident and is therefore irrelevant and inadmissible. See Fed.R.Evid. 402; cf. Richardson v. Richardson-Merrell, Inc., 857 F.2d 823, 831 (D.C.Cir.1988) (holding that epidemiological presumptions may not be admitted because they are based upon caution and prudence rather than a reasonable degree of medical certainty).

The Johnsons also argue that several adverse inferences should be drawn from WMATA’s failure to collect or preserve relevant evidence. They make these arguments under the rubric of the District of Columbia’s spoliation of evidence doctrine. However, even though District of Columbia law governs the Johnsons’ tort claims, federal, not state, rules govern procedural matters such as what evidence may be admitted. Cf. Ricciardi v. Children’s Hosp. Medical Center, 811 F.2d 18 (1st Cir.1987) (holding that even though state substantive law controls in diversity actions, federal evidentiary rules must be applied in federal courts); see generally Hanna v. Plumer, 380 U.S. 460, 85 S.Ct. 1136, 14 L.Ed.2d 8 (1965). The appropriate federal doctrine is the so-called adverse inference rule which “provides that when a party has relevant evidence within his control which he fails to produce, that failure gives rise to an inference that the evidence is unfavorable to him.” Int’l Union v. N.L.R.B., 459 F.2d 1329, 1336 (D.C.Cir.1972). Normally, such inferences are not drawn unless there is evidence of “evil intent, bad faith or willfulness.” See Vick v. Texas Employment Com., 514 F.2d 734, 737 (5th Cir.1975); Friends for All Children, Inc. v. Lockheed Aircraft Corp., 587 F.Supp. 180, 208 (D.D.C.1984). In any event, it is clear that “whether to draw the inference is a matter of discretion for the fact finder,” International Union v. N.L.R.B., 459 F.2d at 1339, and that the facts of this case do not support an exercise of that discretion. First, the Johnsons contend that an adverse inference should be drawn from the destruction of the blood samples drawn from Dixon on the day of the accident. Those samples were not, however, destroyed by WMATA. They were routinely disposed of by the Washington Hospital Center and the American Medical Laboratories, the custodians of the samples. See Kozikowski Deposition at 118; Second Affidavit of Stuart C. Bogema ¶¶ 4-5. Moreover, even though such samples degenerate after a year, the Johnsons failed to request them until September, 1987, more than six months after this suit was filed and almost a year and six months after Dixon’s blood specimen was taken in March, 1986. See WMATA’s Reply ¶ 9. Therefore, it was plaintiffs’ delay in requesting the samples, rather than any desire on the part of WMATA to cover up evidence, that led to the destruction of the blood samples. Accordingly, no adverse inference should be drawn.

Similarly, no adverse inference should be drawn from the fact that WMA-TA did not give Dixon a neurological examination. There is no evidence that such examinations are given as a matter of course. Moreover, the Johnsons have failed to point to any evidence at the time of the accident which would have suggested the need for such an examination. At the time the decision to conduct such an examination would have been made, the drug test results were not available, and neither the police officers interviewing Dixon nor the doctors examining him noted any signs of drug use or impairment. See Supplemental Statement ¶ 15; Second Motion for Summary Judgment at 15. Consequently, there is no reason to infer from WMATA’s failure to conduct a neurological examination on the day of the accident that it intended to suppress relevant evidence.

The destruction of audio tapes poses a closer question. Those tapes recorded the communication between WMATA’s Central Control and its train operators. Unlike the blood sample which probably could not have provided plaintiffs with any helpful evidence, the audio tape could have provided evidence of Dixon’s conduct, possibly confirming that he was singing and whistling as well as recording any manner of speech evidencing impairment. Moreover, unlike the decision to conduct a neurological examination which was made before there was any hint of a lawsuit, the audio tapes were apparently destroyed after thirty days, which would have placed the destruction some time after WMATA knew the results of Dixon’s drug tests. See Miller Deposition at 24-29. It was, however, by no means clear at that point that a suit would be filed: The investigation of the accident suggested that Devora Johnson’s death was an unavoidable suicide, and witnesses had testified that Dixon depressed the emergency brake as soon as possible. Moreover, WMATA had only a limited amount to gain by destroying the tape: The plaintiffs could still obtain evidence of Dixon’s behavior by interviewing persons on the train or at WMATA’s Central Control. In light of these considerations, there is little reason to think that WMATA destroyed, or acquiesced in the destruction of, the tapes because it thought the tapes contained damaging evidence.

Accordingly, no adverse inference will be drawn from the destruction of the audio tape and the blood samples or from WMA-TA’s failure to conduct a neurological examination.

D.

Rule 403 requires courts to consider “the danger of unfair prejudice, confusion of the issues, or misleading the jury.” Fed.R. Evid. 403. As WMATA argues, there is considerable such danger here.

As WMATA notes, “evidence of drug use is, by its very nature, prejudicial.” Simonson v. White, 220 Mont, at 23, 713 P.2d at 988. Such evidence is “likely to stimulate an excessive emotion or to awaken a fixed prejudice as to a particular subject or person involved in the issues.” J. Wigmore, Code of Evidence 355 (3d ed. 1942). Jurors, especially in the District of Columbia, have strong opinions and emotions about drug use and drug users that might obscure their careful consideration of the evidence. Furthermore, jurors can easily envision themselves at the mercy of a drug-impaired train operator. It is therefore possible that a jury would decide to “send a message” to WMATA to eradicate drug use by its employees by holding WMATA liable regardless of whether Dixon’s drug use was the proximate cause of Devora Johnson’s death.

In addition to the dangers of unfair prejudice, there is also considerable danger of confusion. As the previous analysis has shown, the drug test results standing alone have little probative value. They become probative only in connection with the other available evidence. However, some of that evidence — such as whether the subway car could have stopped in time to avoid hitting the decedent and whether Dixon acted as if he was impaired by drugs at the time of the accident — is disputed. As a consequence, in order to assess the drug test results rationally, the jury will first have to resolve the conflicts in the rest of the testimony. Some jurors will be tempted to omit this careful analysis of the facts and jump to the conclusion that the drug tests results show that Dixon was impaired at the time of the accident. Thus, above and beyond the danger of unfair prejudice, there is a considerable danger of confusion.

E.

Despite the danger of prejudice and confusion, the drug test results must be admitted. Under the Federal Rules, relevant evidence may only be excluded if the danger of prejudice and confusion from that evidence “substantially outweigh[s]” its probative value. See Fed.R.Evid. 403. As the keystone of a cogent theory that Dixon acted wantonly, the drug test results are extremely probative and should be admitted.

It is, however, possible to reduce the potential for prejudice. The drug test results only become relevant if the jury determines Dixon’s conduct could have caused decedent’s death. As a consequence, the potentially prejudical effect of the drug test results on the issue of proximate cause could be minimized by bifurcating trial so that the issue of proximate cause is resolved before the introduction of any evidence about the drug test. Under this procedure, if the jury finds in the first phase that Dixon could have stopped the train in time to avoid the accident but failed to do so, the jury would be reconvened to hear evidence about drug use. Alternatively, some other device such as special instructions and a special verdict form could be used in a unitary trial to guide the jury’s decision making.

VI.

WMATA also moves for summary judgment on the ground that the Johnsons have not provided sufficient evidence of impairment to send the question of wantonness to the jury. WMATA’s argument is similar to its arguments on relevancy and probativeness. According to WMATA, because the drug test results only demonstrate that he was impaired sometime in the week before the tests were given, to conclude under these circumstances that Dixon was impaired at the time of the accident would be to “engage in sheer speculation.” Siegel v. Mazda Motor Corp., 878 F.2d 435, 439 (D.C.Cir.1989). It should be clear from the foregoing analysis that this contention must be rejected because the drug test results combined with the corroborating evidence presented by the Johnsons raise a genuine issue as to whether Dixon was impaired by cocaine at the time of the accident, and that genuine issue in turn raises a genuine issue of material fact as to whether Dixon acted wantonly and breached his duty of care toward the decedent.

VII.

One final issue remains. Because Devora Johnson was contributorily negligent, her parents can only recover under the doctrine of last clear chance. See D.C. Transit System, Inc. v. Garman, 301 F.2d 568, 570 (D.C.Cir.1962). 14 Under that doctrine, a plaintiff may recover despite his or her own negligence if four conditions are met:

(1) that the plaintiff was in a position of danger caused by the negligence of both plaintiff and defendant; (2) that the plaintiff was oblivious to the danger, or unable to extricate herself from the position of danger; (3) that the defendant was aware, or by the exercise of reasonable care should have been aware, of the plaintiff's danger and of her oblivion to it or her inability to extricate herself from it; and (4) that the defendant, with means available to him, could have avoided injuring the plaintiff after becoming aware of the danger and the plaintiffs inability to extricate herself from it, but failed to do so.

Felton v. Wagner, 512 A.2d 291, 296 (D.C.1986) (citations omitted). WMATA contends that the Johnsons have failed to satisfy their burden of proving the second element: “that the plaintiff was oblivious to the danger, or unable to extricate herself from the position of danger.” It is clear that Devora Johnson was not “oblivious to the danger” of the oncoming subway car. Railroad tracks are so inherently dangerous that even young children are deemed to understand their danger. See Foshee v. Consolidated R. Corp., 849 F.2d 657, 658 (D.C.Cir.1988). The Johnsons’ claim must, therefore, rise or fall upon whether the decedent was able to extricate herself from the danger of the oncoming subway car.

WMATA argues that the undisputed evidence shows that she was. The decedent was apparently able to move once she reached the tracks, and under plaintiffs’ theory of the facts she had at least ten seconds in which to do so. WMATA contends that in that time she could have rolled into the refuge area located underneath the platform about five feet from the center of the tracks. See Supplemental Response ¶ 24. That area is a space approximately two and one-half feet wide and nearly three feet high that runs underneath the platform at all WMATA subway stations, and, as the photographs supplied by WMATA amply demonstrate, a fair-sized man could fit underneath it. See Supplemental Statement ¶ 24; Second Supplemental Response ¶ 10. WMATA therefore contends that Devora Johnson, who was only five feet, six inches and between 125 and 146 pounds, could easily have fit under the refuge area and thereby extricated herself from danger. See Supplemental Statement ¶ 23; Supplemental Response ¶ 23.

The Johnsons, however, contend that their daughter did not see the refuge area. Its existence was certainly not something about which the ordinary WMATA patron would know. In the lower level of Metro Center, a person standing on the platform would be unaware of the refuge area below them. See Second Wood Deposition at 13. Nor were there any signs or arrows indicating the location of the refuge area. See id. at 11-12. Indeed, even WMATA office employees would be unaware of the refuge area. See Shaeffer Deposition at 8. More fundamentally, even if Johnson had been aware of the refuge area, it is not clear that she would have understood that she would be safe there. WMATA subway, cars have “collector shoes” which extend beyond the track. See Krempasky Deposition at 8. An individual might therefore not realize that the refuge area was big enough for them to hide inside. Accordingly, the plaintiffs have raised a genuine issue as to whether Johnson was aware that she could extricate herself from danger by rolling into the refuge area. It must be resolved by the jury.

VIII.

Because there are genuine issues as to whether the train operator’s conduct proximately caused the accident, and, if so, whether he acted wantonly and whether, in any event, the decedent had the last clear chance to avoid the accident, an accompanying order will deny WMATA’s second motion for summary judgment and schedule further proceedings in this matter.

1 . See Defendant WMATA’s Supplemental Statement of Material Facts Upon Which There Is No Genuine Dispute ¶ 16 [hereinafter, "Supplemental Statement”]; Plaintiffs Responses to Defendant WMATA’s Supplemental Statement of Material Facts Upon Which There Is No Genuine Dispute ¶ 16 [hereinafter, “Supplemental Response”].

2 . See Plaintiff s Supplemental 12 — I(k) Statement of Disputed Facts ¶ 5 [hereinafter, “Second Supplemental Response”]; Defendant WMATA’s Responses to Plaintiff’s Supplemental Statement of Disputed Facts ¶ 5 [hereinafter, "WMATA’s Reply”]-

3 . Supplemental Response ¶ 17.

4 . See Supplemental Statement ¶ 17 (noting that the tests given Dixon "are incapable of determining the quantity of drugs taken, the time they were taken, or impairment’’); Supplemental Response ¶ 17 (not disputing this characterization); Booker Deposition at 59 — 60, 64 (testifying that the test results in isolation do not allow him to determine to a reasonable degree of medical certainty whether Dixon was impaired at the time of the accident).

5 . See Second Supplemental Response ¶ 8; WMATA’s Reply ¶ 8.

6 . See Supplemental Statement ¶ 21; WMATA’s Reply ¶ 8.

7 . See Supplemental Statement ¶ 17; WMATA’s Reply ¶ 8 (noting that plaintiffs have not disputed this evidence).

8 . See Brunina Deposition at 3.

9 . Jordan Deposition at 13; see Supplemental Response ¶ 15.

10 . See Supplemental Statement ¶ 15.

11 . See Dixon Deposition at 83.

12 . That rule provides in full: "When the relevancy of evidence depends upon the fulfillment of a condition of fact, the court shall admit it upon, or subject to, the introduction of evidence sufficient to support a finding of the fulfillment of that condition.” Fed.R.Evid. 103(b).

13 . That rule provides in full:

Although relevant, evidence may be excluded if its probative value is substantially outweighed by the danger of unfair prejudice, confusion of the issues, or misleading the jury, or by considerations of delay, waste of time, or needless presentation of cumulative evidence.

14 . Our Court of Appeals has ruled that there is no "suicide exception” to the doctrine of last clear chance. See Johnson v. WMATA, 883 F.2d at 130-32. The Court of Appeals rejected such an exception because there is no rule in the District of Columbia preventing recovery for wrongful death merely because the decedent was attempting to take his or her life. See id. at 130-31 (discussing Toy v. District of Columbia, 549 A.2d 1 (D.C.1988)). The Court of Appeals found this to be essentially "a question of policy, a decision as to when the value of human life places a heavy responsibility on all comers.” Id. at 131. It did not, however, consider whether such considerations would be counterbalanced in a case such as this one by the fact that decedent may have placed others in danger by her reckless act: By forcing the train operator to apply the mushroom brake, the decedent caused a sudden deceleration which might have injured passengers in the train. Nonetheless, the Court of Appeals’ ruling that the Johnsons have stated a claim under D.C. law must be observed.