RIPPLE, Circuit Judge.
Appellants Mary Santaella and Cary El-dridge are the mother and husband, respectively, of Teresita Eldridge, a woman who died as the result of prescription drug intoxication; they are also the named beneficiaries in her accident insurance policy. When Metropolitan Life Insurance Company (“Met-Life”) neither denied nor paid their claim for accident insurance benefits, they filed suit in federal district court claiming breach of contract. The district court, considering cross-motions for summary judgment, concluded that Teresita Eldridge’s death was not accidental and therefore was not covered by the group policy. It granted summary judgment to MetLife and denied it to Mary Santaella and Cary Eldridge. This appeal followed. For the reasons set forth in the following opinion, we reverse the judgment of the district court and remand the case for the entry of judgment in favor of the appellants.
I
BACKGROUND
A.
Facts
The facts of this ease are undisputed. Teresita Eldridge, a 36 year-old United Airlines flight attendant, was found dead in her home in Bayard, New Mexico, on July 23, 1994. The Chief Medical Investigator for the State of New Mexico, forensic pathologist Ross E. Zumwalt, M.D., performed an autopsy on her body the following day.
He concluded that the immediate cause of her death was “[d]rug (propoxyphene) intoxication.” Certificate of Death, R.16, Ex.A. The Report of Findings and the autopsy report likewise listed the cause of death as drug intoxication of propoxyphene; they stated that the manner of death was accidental. R.16, Exs.B, C.
In his deposition, Dr. Zumwalt elaborated on his conclusions. He testified that the propoxyphene that caused Mrs. Eldridge’s death was a relatively mild synthetic narcotic analgesic or painkiller often sold under the brand name Darvon. It is generally used for pain relief; however, it can cause sedation, confusion and even seizures, particularly in overdoses. The toxicology report stated that the level of propoxyphene in Teresita El-dridge’s blood was 1.42 milligrams per liter. According to Dr. Zumwalt, that level was “fairly low” for a fatal dosage of propoxy-phene. Although anything over one milligram per liter is considered to be within the lethal range, he testified, the average lethal blood level is 4.5 milligrams per liter, more than three times the blood level of Mrs. Eldridge at the time of her death.
Dr. Zumwalt also explained in his deposition that he had considered and ruled out the other manners of deáth — natural death, suicide, homicide and undetermined means— when he concluded that Teresita Eldridge’s death was accidental. He noted that Mrs. Eldridge was found with a can of beer in her hand; however, the alcohol level in her system tested negative. He examined her body for evidence of natural diseases that could kill someone suddenly and unexpectedly, but found none.
After discussing the possibility of suicide with family members, he concluded there was no positive history of suicide attempts or ideation to suggest that cause of death. The family had informed the forensic pathologist, however, that Mrs. Eldridge had been known to take a great deal of prescription medication and to abuse drugs. Dr. Zumwalt noted that Teresita Eldridge had not taken a high toxic level of drugs on the day of her death. Nor was there evidence of undigested residual pills in the stomach, an indication of suicide found in other cases. The Medical Examiner concluded:
And so in my experience, she was the type of person who would or could have a lethal overdose of a prescription drug in an accidental manner. And finally, propoxyphene has been associated very commonly with accidental drug overdoses because there is a very small margin of safety in propoxy-phene. The margin between a therapeutic dose and a toxic or lethal dose is smaller than many other drugs.... The number of pills one would have to take to get a lethal quantity is smaller in propoxyphene than in many other drugs.
R.26, Ex.2 (Zumwalt Dep.) at 21-22.
Consequently, with no evidence of the possibility of suicide or other manners of death, Dr. Zum-walt concluded that Mrs. Eldridge’s death was accidental. In the autopsy report, Dr. Zumwalt listed two other pathologic diagnoses — splenomegaly and lymphadenopa-thy — which indicate an enlargement of the spleen and lymph system. Dr. Zumwalt testified that one cause for such enlargements is chronic drug use. However, these conditions were not a cause of death for Mrs. Eldridge. Dr. Zumwalt then offered his expert opinion that, within a reasonable degree of medical certainty, death was due to a propoxyphene overdose and that the manner of death was accidental.
In its Rule 12(N) Statement, MetLife added the information that on May 6, 1994, about two months before her death, Teresita Eldridge had suffered a seizure. Attached to the 12(N) Statement as an exhibit was a report of the results of a May 10, 1994, magnetic resonance imaging examination indicating that Mrs. Eldridge had “rather severe maxillary sinusitis” but no brain abnormalities. R.26, Ex.l. Also attached to the Statement as an exhibit is a barely legible page, presumably part of a medical history chart, dated May 9, 1994. Although the parties have reported what they believe portions of the handwritten notes say, neither has presented further evidence (in the form of affidavits or depositions, for example) from the medical personnel who wrote the notes or
from other doctors who had treated Mrs. Eldridge.
B.
The Life Insurance Policy
At the time of her death, Mrs. Eldridge was enrolled in the United Airlines’ personal accident group insurance policy issued by MetLife to United Airlines. It insured Tere-sita Eldridge in the amount of $300,000 for accidental loss of life and dismemberment. The group policy is part of United Airline’s employee welfare benefit plan governed by the Employee Retirement Income Security Act of 1974, as amended (“ERISA”), 29 U.S.C. §§ 1001
et seq.
Teresita Eldridge designated as her beneficiaries under that plan her mother, Mary Santaella, and her husband, Cary Eldridge, the appellants in this ease. Teresita Eldridge’s premium payments were made through payroll deductions and were current.
The policy offers 24-hour protection against any accident anywhere in the world, on or off the job. It states that, if an injury results in death within one year of the date of an accident, the full benefit amount is paid. However, benefits are not paid if death results from suicide, self-inflicted injury, illness, disease, bodily infirmity or infection. The policy expressly denies coverage of “in-' tentionally self-inflicted injuries, suicide or any attempt thereof, while sane or insane.” App. at A-79. There are no definitions provided, however, for such terms as “accident,” “injury” or “intentionally self-inflicted.”
C.
The District Court Proceedings
This case was appropriate for summary judgment resolution, stated the district court, because there are no disputed facts and because the interpretation of insurance contracts is a matter of law. The court found that the MetLife policy was not ambiguous because its undefined terms could be construed in a common-sense manner. According to the court, the reasonable interpretation of the “suicide or self-inflicted injury” exclusion of benefits is that coverage can be denied only if an insured purposefully injures herself. The reasonable interpretation of “accidental death” is one of reasonable foreseeability: “[Wjould a reasonable person with Teresita’s characteristics know that serious bodily injury or death was likely to occur as a result of ingesting an overdose of propoxyphene?” R.30 at 21. The district court concluded that the answer, on this record, must be “yes.” The court reasoned that Teresita Eldridge undisputedly abused drugs, as evidenced by her enlarged spleen and lymph system. The court then concluded that these conditions were most likely caused by chronic drug abuse. The court then noted that Mrs. Eldridge had suffered a seizure two months before her death and noted that seizures are a common side effect of taking an overdose of propoxyphene. Although the court acknowledged that there was no evidence that Mrs. Eldridge knew her seizure was a side effect of overdosing on propoxyphene, a reasonable person would have been alerted:
[A] reasonable person would (i) consider a seizure to be a serious physical injury, (ii) question whether the seizure was related to her drug abuse, and (iii) reasonably expect that serious injury or death was highly likely to occur if she continued to abuse drugs, including propoxyphene.
Id.
The court therefore concluded that Tere-sita Eldridge’s death was not accidental and therefore that she was not covered under the policy.
II
DISCUSSION
A.
This case involves the denial of accidental death benefits claimed under an ERISA-regulated employee welfare benefit plan. In these cases, we apply the traditional de novo standard to the review of a summary judgment ruling.
Casey v. Uddeholm Corp.,
32 F.3d 1094, 1096 (7th Cir.1994).
In a summary judgment action, the moving party shoulders the initial burden of production. It must identify “those portions of ‘the pleadings, depositions, answers to interrogatories, and admissions on file, together with the affidavits, if any,’ which it believes demonstrate the absence of a genuine issue of material fact.”
Celotex Corp. v. Catrett,
477 U.S. 317, 323, 106 S.Ct. 2548, 2553, 91 L.Ed.2d 265 (1986) (quoting Fed.R.Civ. P. 56(c)). When the movant satisfies that burden, the nonmovant must “set forth specific facts showing that there is a genuine issue for trial.” Fed.R.Civ.P. 56(e). “If no genuine issue of material fact exists, the sole question is whether the moving party is entitled to judgment as a matter of law.”
Logan v. Commercial Union Ins. Co.,
96 F.3d 971, 978 (7th Cir.1996) (citing
Miranda v. Wisconsin Power & Light Co.,
91 F.3d 1011, 1014 (7th Cir.1996)). We have jurisdiction over the district court’s grant of summary judgment to MetLife; it is a final and appeal-able judgment. The court’s denial of summary judgment to the plaintiffs, even though usually not appealable because it is an interlocutory decision, has merged into the final judgment and is therefore appealable as well.
See Swaback v. American Info. Techs. Corp.,
103 F.3d 535, 543 (7th Cir.1996).
When, as here, cross-motions for summary judgment are filed, we look to the burden of proof that each party would bear on an issue of trial; we then require that party to go beyond the pleadings and affirmatively to establish a genuine issue of material fact.
See Celotex,
477 U.S. at 324, 106 S.Ct. at 2553. In this case, at trial the plaintiffs would bear the burden of proving Mi’s. Santaella’s entitlement to the benefits of the insurance coverage, and the defendant MetLife would bear the burden of establishing Mrs. Santaella’s lack of entitlement because she falls within the “exclusions” section of the insurance contract.
See Fuja v. Benefit Trust Life Ins. Co.,
18 F.3d 1405, 1408 (7th Cir.1994).
Because this insurance plan was established under ERISA, federal common law rules of contract interpretation govern the disposition of this case.
Life Ins. Co. of North America v. Von Valtier,
116 F.3d 279, 283 (7th Cir.1997). Under those federal common law rules we are to
“interpret the terms of the policy in an ordinary and popular sense, as would a person of average intelligence and experience,” and “construe all plan ambiguities in favor of the insured.” Plan language may only be deemed ambiguous where “it is subject to more than one reasonable interpretation.”
Id.
(quoting
Cannon v. Wittek Cos., Int’l,
60 F.3d 1282, 1284 (7th Cir.1995)). With these principles in mind, we begin our de novo review of the district court’s holding that Teresita Eldridge’s death was not covered by the MetLife policy because her death was not accidental.
B.
The appellants submit that the district court erred in determining that MetLife was entitled to summary judgment on the ground that the record would not support a finding by the trier of fact that the death of Mrs. Eldridge was accidental. The appellants raise three contentions in this respect. They submit first that the court’s factual determinations were not supported by the evidence of record. They challenge the district court’s conclusion that a reasonable person would have asked whether the seizure two months prior to her death was related to her drug abuse and would have foreseen that serious injury or death was highly likely to occur if she continued to abuse drugs, including pro-poxyphene. The appellants submit that all the evidence concerning this seizure — whether it occurred and, if so, what caused it — is pure speculation unsupported by the medical record, which contains only illegible notes in a medical chart. The appellants also contend that the district court erred in overlooking or rejecting the professional expert opinion of a highly experienced forensic pathologist, Dr. Zumwalt, who concluded that Mrs. Eldridge died from an accidental overdose of propoxy-
phene. According to the appellants, the professional opinion of the state medical expert should have been treated as dispositive. Without similar rebuttal evidence, Dr. Zum-walt’s opinion stands unimpeached. Finally, the appellants assert that the undisputed facts before the district court warranted summary judgment for them. They point to Dr. Zumwalt’s uncontradicted testimony that Teresita Eldridge accidentally overdosed and to the autopsy, reports and death certificate stating that her death was accidental.
MetLife, on the other hand, submits that the term “accidental” requires an examination of the subjective and objective expectations of the insured to determine whether the insured “knew or should have known that serious bodily injury or death was a probable consequence substantially likely to occur as a result of’ the insured’s conduct.
Wickman v. Northwestern Nat’l Ins. Co.,
908 F.2d 1077, 1088-89 (1st Cir.),
cert. denied, 498
U.S. 1013, 111 S.Ct. 581, 112 L.Ed.2d 586 (1990). MetLife claims that Mrs. Eldridge reasonably should have foreseen that serious injury or death was likely to occur as a result of her voluntary ingestion of propoxyphene. In MetLife’s view, Mrs. Eldridge abused drugs and, as a result, she had an enlarged spleen and damaged lymph system and suffered a seizure two months before her death. Also, it claims the likelihood of death resulting from an overdose of propoxyphene is generally recognized. MetLife urges this court to affirm the district court’s conclusion that Mi's. Eldridge’s death cannot be considered accidental.
C.
We turn first to the terms of the MetLife 24-Hour Personal Accident Insurance policy for United Airlines employees. The policy provides full payment of benefits if an injury results in death within a year of the accident. It does not define “accident” but does exclude from coverage injuries that are intentionally self-inflicted and suicidal acts. Such an exclusion is frequently found in policies of accident insurance; “[i]n the absence of statutory regulation or of public policy to the contrary, the insurer may limit the circumstances and types of accident for which it will be liable” by setting forth exclusion provisions. 1A John Alan Appleman, Insurance Law and Practice § 360, at
445-46
(1981).
Interpreting the policy terms in an ordinary and popular sense, in the way that a person of average intelligence and experience would,
see Von Valtier,
116 F.3d at 283, we treat the term “accidental” as it is commonly defined, as “unexpected or unintentional.”
Casey v. Uddeholm Corp.,
32 F.3d 1094, 1097 (7th Cir.1994).
Wickman,
the First Circuit case on which the district court relied, is not
to the
contrary.
See 908
F.2d at 1085 (“The question comes down to what level of expectation is necessary for an act to constitute an accident.”). When we consider whether a certain result is accidental, “it is customary to look at the casualty from the point of view of the insured.” Appleman, Insurance Law and Practice § 360, at 452-53;
see Todd v. AIG Life Ins. Co.,
47 F.3d 1448, 1456 (5th Cir.1995) (noting that one rule for construing contracts is the principle of
contra proferentem)-, Wickman,
908 F.2d at 1087 (noting that the common law prescribes that policy contract terms should be judged from the viewpoint of the insured). However, when, as is usually the case, the evidence is not sufficient to ascertain with certainty the subjective expectation or intent
of the decedent, an objective reasonable person standard must be employed to determine the insured’s expectation.
See Wickman,
908 F.2d at 1088 (“An objective analysis, when the background and characteristics of the insured are taken into account, serves as a good proxy for actual expectation.”). Applying
Wickman’s
approach, the Fifth Circuit set forth a methodology for analyzing whether a death was accidental:
[F]or death under an accidental death policy to be deemed an accident, it must be determined (1) that the deceased had a subjective expectation of survival, and (2) that such expectation was objectively reasonable, which it is if death is not substantially certain to result from the insured’s conduct.
Todd,
47 F.3d at 1456 (citing
Wickman).
As we noted earlier in this subsection, the policy also contains an exclusion clause which absolves MetLife from liability if it can demonstrate to the satisfaction of the trier of fact that death was the result of an “intentionally inflicted injury.” The plan does not define this term. It is therefore our responsibility to give this term the content that it is given in the ordinary and popular sense, in the way that a person of average intelligence and experience would interpret the terms.
See Von Valtier,
116 F.3d at 283.
We now apply these principles to the record in the case before us.
D.
There is no dispute that Teresita Eldridge ingested a lethal amount of the pain reliever propoxyphene that caused her death. However, there is no evidence that she knew it was a lethal amount. Nothing in the record reveals Mrs. Eldridge’s expectation or intent in taking the overdose. Moreover, the record simply would not support a jury determination that a reasonable person in Mrs. Eldridge’s situation would have considered death either “highly likely to occur” as a result of her ingestion of this particular dosage of the medication,
Wickman,
908 F.2d at 1088, or “substantially certain to result from the insured’s conduct,”
Todd,
47 F.3d at 1456.
The record reveals that Mrs. Eldridge’s family reported a history of drug abuse. The record is silent, however, as to the extent of that condition. More importantly, although the record establishes that a dose of propox-yphene caused her death, it reveals little else about that dose. Notably, we do not know how much propoxyphene Teresita Eldridge took. The concentration in her blood was 1.42 milligrams per liter. We also know from the
record
that Mrs. Eldridge had an enlarged spleen and lymph system, discovered by Dr. Zumwalt during the autopsy, and that she possibly suffered a seizure two months before her death.
On the basis of this record, the district court decided that a reasonable person in Mrs. Eldridge’s situation would have considered her seizure a serious physical injury and, in making the decision to take the drugs, would have assumed that the seizure was causally related to the use of drugs and further would have assumed that taking the dosage created the probability of death or serious injury.
We believe that the district court erred in this analysis. The record contains no evidence to support the theory that Mrs. El-dridge ought to have related her use of drugs, to whatever degree she had a dependency, to her seizure. The record makes clear that Mrs. Eldridge took seriously her seizure; she sought medical treatment and, it appears, was given both MRI and EEG examinations. She received a follow-up examination three days after the seizure. That examination is the subject of the illegible page from her medical history included in the
record. At best, the contents
of
that page are largely speculative. The little that the page does reveal hardly supports the district court’s decision. MetLife, in its Rule 12(N) Statement to the district court, stated that the chart indicates that Mrs. Eldridge had taken cold medicine, R.26 at 3; the appellants tell us she had taken only aspirin. The chart records that she had not used illegal drugs in five years. Propoxyphene apparently is not mentioned on the chart. In short, although Mrs. Eldridge sought medical treatment for her seizure episode, the record reveals nothing about the role that drugs played in that situation. Nor does the record suggest that she had any reason to know that this event made the ingestion of the drugs she took on the day of her death an unreasonable risk of death or serious injury. We must conclude that the district court had no evidentiary basis for believing that Mrs. Eldridge should have related her seizure with her taking of drugs, or specifically with her ingestion of the prescription medication propoxyphene. Its conclusion that Mrs. El-dridge “should have known that death or serious injury was substantially likely to occur” had no foundation in the record.
See Wickman,
908 F.2d at 1089. The record does not support the district court’s imputed cognizance to Mrs. Eldridge.
Indeed, when the entire record is considered, we think that it would permit a trier of fact to reach but one
conclusion
— that Mrs. Eldridge’s death was an accident. The deposition testimony of the Chief Medical Officer of New Mexico, Dr. Zumwalt, is a major component of the record and certainly the most thorough evidence offered. He assessed the facts; he collected the autopsy results, the toxicology report, the medical history, the survey of the scene of death, and the interviews with the family; he then formulated conclusions on the basis of those facts. For example, he explained the toxicology report’s analysis that the level of propox-yphene was 1.42 milligrams per liter. Dr. Zumwalt testified that, even though that quantity is above the lethal base level of 1 milligram per liter, it is three times lower than the average lethal dosage of 4.5 milligrams per liter of propoxyphene. In Dr. Zumwalt’s opinion, the quantity of propoxy-phene found in Mrs. Eldridge was a relatively low level of propoxyphene for a fatal dosage. He noted that the margin between a therapeutic dose and a lethal one is smaller in propoxyphene than in many other drugs and opined that, in his experience, someone like Teresita Eldridge who took prescription medications and abused drugs “was the type of person who would take a lethal overdose of a prescription drug in an accidental manner.” The pathologist commented as well that a person could develop increasing levels of tolerance for propoxyphene and thus could take mistakenly a lethal dose of the drug.
Dr. Zumwalt did discover during the autopsy that Mrs. Eldridge had an enlarged spleen and damaged lymph system and noted in his deposition that such damage can result from drug abuse. However, the pathologist testified that neither splenomegaly nor lym-phadenopathy was a cause of her death. Dr. Zumwalt also mentioned in his deposition that an overdose of propoxyphene can cause a seizure. However, neither party asked him about the actual seizure Teresita Eldridge reported experiencing on May 6, 1994 or asked him to review Mrs. Eldridge’s medical chart concerning that seizure. Consequently, Dr. Zumwalt’s unrebutted testimony provides no link between the seizure and any drug-taking, much less the drugs ingested on the day of Mrs. Eldridge’s death.
In focusing on the evidence of a seizure and in speculating about the possibility of a causal link between Mrs. Eldridge’s seizure and her taking propoxyphene, the district court erred in setting aside the unrebutted opinion of Dr. Zumwalt that Teresita El-dridge’s death was accidental. Its selective use of limited facts is not appropriate in deciding a motion for summary judgment.
Casey v. Uddeholm Corp.,
32 F.3d 1094, 1099 (7th Cir.1994). Employing the methodology of
Todd
and
Wickman,
we must conclude that the record requires the conclusion that Mrs. Eldridge had a subjective expectation of survival and that such an expectation was objectively reasonable because death was not certain or even highly likely to result from her conduct.
Todd,
47 F.3d at 1456. It was an accident.
What we have said with respect to the accidental nature of Mrs. Eldridge’s death makes it clear that MetLife cannot rely upon the “intentionally inflicted self-injury” exclusion in the policy. The record simply will not support a determination by the trier of fact that Mrs. Eldridge did anything other than máke a fatal mistake. Just as the record lacks evidence that her death was not accidental, there is a lack of evidence that Mrs. Eldridge died of intentionally self-inflicted injuries. She voluntarily took propoxyphene in a dangerous overdose amount, but in an amount that, according to the unrebutted testimony of Dr. Zumwalt, usually is indicative of an accidental overdose.
The record before us does not support MetLife’s characterization of Mrs. Eldridge’s ingestion of propoxyphene as a purposeful infliction of injury on herself, one that resulted in her death. It is undisputed that she took the drug voluntarily, but nothing in the record indicates that she intended to take an overdose or that she intended to inflict injury on herself.
See
Casey, 32 F.3d at 1097 (“A self-inflicted injury may be accidental, where accidental is taken to mean unintentional rather than unexpected. For example, it is an accident when someone hits his thumb with a hammer when driving a nail. The injury was self-inflicted but not intended, hence accidental.”). Nor is there any evidence that Mrs. Eldridge knew or should have known that she had a damaged spleen and lymph glands or that the seizure she suffered might be related to an abuse of drugs. Without some evidence that she was aware of the risk of serious injury or death, we must conclude that Mrs. Eldridge did not die from an intentionally self-inflicted injury and thus that the exclusion did not bar her claim.
Cf Life Ins. Co. of North America v. Von Valtier,
116 F.3d 279, 283-84 (7th Cir.1997) (holding that policy exclusion for intentionally self-inflicted injuries barred the claim of defendant who inflicted a fatal wound on himself realizing, admitted the defendant’s expert, that the wound would lead to his death).
We must hold, therefore, that Mrs. Eldridge’s death was accidental and did not result from an intentionally self-inflicted injury. Indeed, given the state of the record, it is appropriate that we not only reverse the decision of the district court granting summary judgment to MetLife, but that we also direct the entry of judgment for the appellants. This case was submitted to the district court on cross-motions for summary judgment. Neither party has indicated the appropriateness, or even the possibility, of further meaningful augmentation of the record. The reasons that often make us reluctant to grant summary judgment upon reversing a judgment for the other party are simply' not present. As we recently explained:
The reason that appellate courts, when reversing a grant of summary judgment, typically do not direct the" district court to enter summary judgment in favor of the appellant is because a genuine issue of material fact remains. But, in instances in which the facts and law establish that the appellant is entitled to judgment as a matter of law, we are free to direct the district court to enter judgment in appellant’s favor.
Swaback v. American Info. Techs. Corp.,
103 F.3d 535, 544 (7th Cir.1996). In this case, as in
Swaback,
both parties have responded to the cross-motions for summary judgment relying on the same undisputed facts; each party asserts that those facts necessarily lead to judgment in its favor.
Conclusion
Accordingly, we reverse the judgment of the district court and remand the case to the district court with directions to enter summary judgment for Mary Santaella and Cary Eldridge. The appellants shall recover their costs in this court.
REVERSED AND REMANDED WITH DIRECTIONS.