[467]*467OPINION
BOGGS, Circuit Judge.
Nearly fifty years after conducting an experiment in which over eight hundred pregnant women ingested a liquid substance that — unbeknownst to the women— contained radioactive iron isotopes, Vanderbilt University and the Vanderbilt University Medical Center (“Vanderbilt”) were sued in a class action lawsuit. Vanderbilt settled the lawsuit and then sought indemnity from its insurers, St. Paul Fire & Marine Insurance Company and St. Paul Mercury Insurance Company (“St. Paul”) and United States Fire Insurance Company (“U.S.Fire”). St. Paul entered into a settlement agreement with Vanderbilt. U.S. Fire, on the other hand, refused to indemnify Vanderbilt and sought a court judgment to determine the extent of its liability under the terms of the appropriate policies. The district court ruled in favor of U.S. Fire on the basis that Vanderbilt did not give U.S. Fire timely notice of an occurrence under the terms of the policies. Vanderbilt now appeals. For the following reasons, we affirm.
I
Between 1945 and 1949, faculty and staff at Vanderbilt conducted an experiment in which more than eight hundred pregnant women were fed radioactive iron isotopes in order to facilitate the scientific tracking of iron absorption in pregnant women (“the Experiment” or “the 1940s study”). See Craft v. Vanderbilt Univ., 18 F.Supp.2d 786, 789 (M.D.Tenn.1998). The Experiment was conducted as part of the Tennessee-Vanderbilt Nutrition Project. Participants in the project later claimed that they were not informed of the radioactive nature of the iron solution they ingested. The participants later stated that they were told the solution was a “cocktail” or “vitamin drink.” See ibid.
During the 1960s, Vanderbilt conducted a survey of the women treated in the Experiment, as well as their children (“the Survey” or “the 1960s study”). See ibid. The Survey involved some interviews and answering questionnaires. No physical touching or medical procedures were involved. The participants in the Survey were not informed of the fact that they had been exposed to radiation, nor were they informed of the results of the Survey. See ibid. In 1969, Vanderbilt published the findings from the Survey in the American Journal of Epidemiology. Ruth M. Hagstrom et al., Long Term Effects of Radioactive Iron Administered During Human Pregnancy, 90 Am. J. Epidemiology 1 (1969). The Survey identified a “small, but statistically significant increase” in the incidence of cancer in the exposed children compared to an unexposed control group of children. Id. at 1. The Survey found that four children who were exposed in útero during the Experiment had died of cancer during the 1950s, compared to none of the control group. Id. at 4. In one ease, the researchers concluded that the incidence of cancer was probably not related to the radiation since other children in the family unexposed to radiation had also died of the same form of cancer. Ibid.
In December 1985, the United States Department of Energy (DOE) requested information regarding the Experiment from Vanderbilt for use in a congressional hearing at which some of the Experiment participants would testify. Vanderbilt informed its attorneys of the request.
On February 1, 1994, over 800 women and their children filed a class action lawsuit against Vanderbilt (“the Craft Litigation”). The Craft plaintiffs asserted numerous claims against Vanderbilt and other defendants related to Vanderbilt’s [468]*4681940s study and its 1960s study. Vanderbilt points out that the initial and amended complaints asserted distinct liability based on the 1960s study. The Craft plaintiffs alleged that in undertaking the 1960s study and failing to disclose the nature of the original study, Vanderbilt violated the plaintiffs’ civil rights, acted negligently, and failed to disclose material information.
In 1998, Vanderbilt settled the Craft Litigation for $10 million (“the Craft Settlement”). This included a co-defendant’s $900,000 payment, leaving Vanderbilt’s portion of the Craft Settlement at $9.1 million. The Craft Settlement specifically mentioned both the 1940s study and the 1960s study, stating that:
“Settled Claims” means any and all claims, ... causes of action (in law or in equity), ... against the Settling Defendants arising out of, related to, or as a result of, Settling Defendants’ funding and/or participation of any land or nature in the [1940s study] and in the follow-up [1960s] study, and all of the facts, omissions and/or events alleged in the Complaint or First or Second Amended Complaints or in Pretrial Order No. 4.
The district court approved the Craft Settlement and dismissed all claims against Vanderbilt with prejudice. The court found that any subsequent orders distributing the Craft Settlement would not affect the final judgment. Vanderbilt did not participate in the hearings to distribute the lump-sum settlement proceeds. The Craft Settlement was distributed in the following major portions by the district court: approximately (1) $3.8 million in plaintiffs’ attorneys’ fees and costs; (2) $4 million to the battery claims arising out of the Experiment; and (3) $1.25 million to the wrongful death claims arising out of the children’s deaths in the 1950s.
Vanderbilt had obtained extensive primary insurance with St. Paul, including hospital professional liability coverage, dating from at least the 1950s to the present. Vanderbilt supplemented this coverage with excess liability insurance from U.S. Fire. The relevant excess policies cover the periods July 1, 1965-July 1, 1968 and July 1, 1968-July 1, 1971. The policies included provisions requiring exhaustion of primary and “other” insurance before their coverage could be invoked and required that a covered event must cause injury during the policy period. The policies listed St. Paul as the underlying insurer.
After the Craft Settlement, St. Paul initially disputed coverage of the claims, in part because Vanderbilt had not met its burden of proving the terms and conditions of the coverage provided by St. Paul. U.S. Fire notes and St. Paul admits that neither St. Paul nor Vanderbilt has ever found copies of any alleged policies, although St. Paul stresses that Vanderbilt did locate documentary evidence of liability insurance coverage provided by St. Paul in the 1960s. St. Paul and Vanderbilt entered into a Settlement Agreement (“the St. Paul Settlement”). Under the terms of the St. Paul Settlement, St. Paul identified five separate and non-consecutive policy years, from 1963-1967 and 1968-69, during which part of the Survey was conducted. St. Paul paid Vanderbilt a little more than $2 million in defense costs and $2.5 million as indemnity. Vánderbilt in turn released St. Paul “for any claims which have been or may in the future be brought arising out of any studies involving the use of radioactive isotopes which were completed prior to July 1, 1969.” This specifically included the Craft Litigation and other litigation involving an experiment conducted by Vanderbilt on nearly 200 elementary school children in Nashville, Tennessee, [469]*469who were given lemonade that was laced with radioactive iron. See Hughes v. Vanderbilt Univ., 215 F.3d 548, 545 (6th Cir.2000).
The St. Paul Settlement stated that “nothing in this Agreement ... is intended as or may be construed to be an admission of coverage, fault, liability or wrongdoing of any party.” In addition, the St. Paul Settlement stated that Vanderbilt would not seek any further indemnity payments from St. Paul in the settlement of the Craft Litigation “in the event that any court of competent jurisdiction determines that the payment made by St. Paul ... did not exhaust any coverage which was or may have been provided by St. Paul to Vanderbilt.”
After the Craft Litigation was filed, Vanderbilt notified U.S. Fire. U.S. Fire accepted notice and acknowledged receipt of the lawsuit. U.S. Fire refused to defend Vanderbilt, however, on the basis that its policies were “following-form” excess policies. U.S. Fire took no role in defending the Craft Litigation. Vanderbilt provided U.S. Fire with status reports and continually requested that U.S. Fire provide a defense. Vanderbilt notified U.S. Fire of the Craft Settlement prior to its execution and court approval.
After the Craft settlement, Vanderbilt sought indemnity from U.S. Fire at the same time it sought indemnity from St. Paul. Vanderbilt sought indemnity pursuant to a provision in U.S. Fire’s following-form policies, stating:
IT IS AGREED THAT THE INSURANCE COVERAGE AFFORDED BY THIS POLICY FOR HOSPITAL PROFESSIONAL LIABILITY SHALL, IN NO EVENT, BE LESS THAN THE COVERAGE PROVIDED BY UNDERLYING INSURANCE LISTED IN SCHEDULE A.
U.S. Fire informed Vanderbilt that it was “working” with St. Paul and awaiting its coverage position. Upon request, Vanderbilt provided U.S. Fire with all non-privileged documents from the Craft Litigation.
After St. Paul and Vanderbilt settled claims, U.S. Fire sued Vanderbilt. On January 20, 1999, U.S. Fire filed a declaratory judgment action against Vanderbilt and St. Paul regarding its rights and duties under the policies that U.S. Fire issued to Vanderbilt in the 1960s and early 1970s. On April 14, 1999, St. Paul filed a motion to dismiss the lawsuit or, in the alternative, for summary judgment. U.S. Fire filed a motion for leave to amend its complaint to add claims for indemnity and contribution, breach of, the duty of good faith, and tortious interference.
On July 19, 1999, U.S. Fire moved for summary judgment on the issue of whether Vanderbilt gave U.S. Fire notice of an occurrence under the terms of its policies. On July 20, 1999, Vanderbilt moved for summary judgment on the issue of coverage, ie., U.S. Fire’s contractual obligation to indemnify Vanderbilt for the amounts Vanderbilt paid toward the Craft Settlement. On August 20, 1999, Vanderbilt cross-moved for summary judgment on the issue of notice. On January 27, 2000, the district court granted U.S. Fire’s summary judgment motion, denied Vanderbilt’s motions, granted St. Paul’s motion to dismiss, and denied U.S. Fire’s motion for leave to amend its complaint. On February 18, 2000, Vanderbilt filed a notice of appeal. On March 2, 2000, U.S. Fire filed a notice of conditional appeal related to the district court’s decision granting St. Paul’s motion to dismiss.
II
Vanderbilt and U.S. Fire filed cross-motions for summary judgment on the issue of notice. U.S. Fire contends that it is [470]*470not required to indemnify Vanderbilt because Vanderbilt did not inform U.S. Fire of the events upon which the Craft Litigation was based — the 1940s study and the 1960s study — until the Craft Litigation was initiated. Vanderbilt contends that the notice it gave to U.S. Fire in 1994 was proper and that, even if the notice was untimely, U.S. Fire was not prejudiced by the delay. Vanderbilt argues that the district court erred in granting summary judgment to U.S. Fire on the issue of notice and in denying Vanderbilt’s cross-motion for summary judgment on this issue. Vanderbilt contends that, at the least, the case should proceed to trial on this issue.
A
This court reviews a district court’s grant of summary judgment de novo, using the same standard employed by the district court. See National Enters., Inc. v. Smith, 114 F.3d 561, 563 (6th Cir.1997). Summary judgment is not necessarily appropriate solely because the parties filed cross-motions for summary judgment. See B.F. Goodrich Co. v. U.S. Filter Corp., 245 F.3d 587, 593 (6th Cir.2001). A court may conclude, after reviewing the undisputed material facts agreed upon by the parties and drawing all inferences, in turn, for each non-moving party, that a genuine issue exists, in which case the court is not permitted to resolve the matter, but rather, must allow the case to proceed to trial. See ibid. Or the court may determine that one party has demonstrated that no genuine issue exists and, therefore, that party is entitled to judgment as a matter of law. See ibid.
B
The parties do not dispute that Tennessee law governs our interpretation of the insurance policies at issue in this case. Therefore, in analyzing the parties’ cross-motions for summary judgment on the issue of notice, we are guided by the Tennessee Supreme Court’s recent decision in Alcazar v. Hayes, 982 S.W.2d 845, 856 (Tenn.1998). According to Alcazar, under Tennessee law, “once it is determined that the insured has failed to provide timely notice in accordance with the insurance policy, it is presumed that the insurer has been prejudiced by the breach.” Ibid. Although the insurer is presumed to be prejudiced, the court in Alcazar adopted a rebuttable presumption rule, according to which the insured has the burden to proffer “competent evidence that the insurer was not prejudiced by the insured’s delay.” Ibid. The court rejected the state’s previous approach, which recognized that notice was a condition precedent to recovery under a policy and which required automatic forfeiture of the policy-without a showing of prejudice-if timely notice was not given. See Talley v. State Farm Fire & Casualty Co., 223 F.3d 323, 327 (6th Cir.2000). Therefore, in reviewing the parties’ cross-motions for summary judgment on the issue of notice, we must engage in a two-step inquiry. First, we must determine if Vanderbilt gave timely notice to U.S. Fire in accordance with the terms of Vanderbilt’s insurance policies with U.S. Fire. Second, if we determine that Vanderbilt did not give timely notice, we must decide if Vanderbilt has rebutted the presumption of prejudice to U.S. Fire.
Vanderbilt asserts that the district court erred when it concluded that Vanderbilt’s notice to U.S. Fire was not timely.
According to the terms of its policies with U.S. Fire, Vanderbilt was to give U.S. Fire notice of an “occurrence” “as soon as practicable.” In full, the provision states:
Upon the happening of an occurrence reasonably likely to involve the company [471]*471hereunder, written notice shall be given as soon as practicable to the company or any of its authorized agents. Such notice shall contain particulars sufficient to identify the insured and the fullest information obtainable at the time.
The policies define an “occurrence” as:
either an accident or happening during the policy period or a continuous or repeated exposure to conditions which unexpectedly and unintentionally cause injury to persons ... during the policy period. All damages arising out of such exposure to substantially the same general conditions shall be considered as arising out of one occurrence.
In addition to the provision regarding notice of an occurrence, the policies also included the following provision regarding notice of claims:
The insured shall give like notice of any claim made on account of such occurrence. If legal proceedings are begun the insured, when requested by the company, shall forward to it each paper thereon, or a copy thereof, received by the insured or the insured’s representatives, together with copies of reports of investigations made by the insured with respect to such claim proceedings.
According to the terms of the policies, Vanderbilt had an obligation to notify U.S. Fire “as soon as practicable” of an occurrence under the policy and then had an obligation to notify U.S. Fire as to claims made on account of an occurrence.
The Tennessee Supreme Court has stated that when an insured is required to give notice of an occurrence “as soon as practicable,” the insured has a duty to “give notice when he becomes, or should become, aware of facts which would suggest to a reasonably prudent person that the event for which coverage is sought might reasonably be expected to produce a claim against the insurer.” Reliance Ins. Co. v. Athena Cablevision Corp., 560 S.W.2d 617, 618 (Tenn.1977). The court stated that “ ‘as soon as practicable’ are not words of precise and definite import. They are roomy words. They provide for more or less free play. They are in their nature ambulatory and subject under the guiding rule, to the impact of particular facts on particular cases.” Ibid, (quoting Transamerica Ins. Co. v. Parrott, 531 S.W.2d 306, 312-13 (Tenn.Ct.App.1975)). The court stated that the words “must be construed as requiring the notice within a reasonable time under all the circumstances, to effectuate the objects and purposes of the notice clause.” Reliance Ins., 560 S.W.2d at 618 (quoting Transamerica, 531 S.W.2d at 313).
Vanderbilt’s notice of an occurrence to U.S. Fire came in 1994, immediately upon service of the Craft complaint. Vanderbilt offers several arguments to support its contention that the 1994 notice was given as soon as practicable. Vanderbilt relies heavily on its assertion that it sought coverage from U.S. Fire solely for the claims arising out of the 1960s study, not those arising out of the 1940s. study, since Vanderbilt had policies with U.S. Fire during the time of the 1960s study and not during the time of the 1940s study. Vanderbilt claims that prior to 1994, it neither knew, nor could have known, that it faced liability for the 1960s study. In the alternative, Vanderbilt argues that even if the notice provision were to extend to the 1940s study, it neither knew, nor could have known, that it could be held liable for that study as well. Vanderbilt relies on testimony of the authors of the 1960s study that they were unaware of any potential liability to Vanderbilt as a result of either study. They testified that the 1960s study did not demonstrate a cause-and-effect relationship and that the findings were not considered dramatic.
[472]*472Vanderbilt argues that the district court improperly weighed the evidence, reached its own factual conclusions, and created a heightened standard of notice for a “sophisticated” insured. Specifically, Vanderbilt points to the following statement in the district court’s opinion:
[w]hile the authors of the article [relating to the 1960s study] may have testified years after the fact in the midst of litigation that these findings were not dramatic, the potential for liability arising out of the experiments seems obvious and should have been obvious, especially to a sophisticated institution like Vanderbilt.
Vanderbilt contends that by establishing a “sophisticated insured” test, the district court misapplied the “reasonably prudent person” test from Reliance Insurance. At the least, Vanderbilt argues, there is a genuine issue of material fact as to the issue of notice.
Vanderbilt also argues that, as an excess insurer, U.S. Fire was entitled to notice only if Vanderbilt was aware that a claim was likely to exceed the limits of St. Paul’s primary hospital professional liability coverage. See Evanston Ins. Co. v. Stonewall Surplus Lines Ins. Co., 111 F.3d 852, 860-61 (11th Cir.1997); Transamerica, 531 S.W.2d at 314. Vanderbilt claims this is consistent with U.S. Fire’s policies, which state that notice is required “[u]pon the happening of an occurrence reasonably likely to involve the company hereunder.” Vanderbilt asserts that notice would be required of a claim likely to exceed $500,000, since this was the aggregate policy limit of the coverage provided by St. Paul. Vanderbilt points out that actual claims against Vanderbilt between 1965-69, even for wrongful death, never exceeded $100,000. Vanderbilt contends that the district court erred in considering other Tennessee wrongful death cases from the 1960s that had involved awards of over $100,000, and then concluding that it was possible that the four wrongful deaths alleged as a result of the 1940s study could collectively have exceeded $500,000.
Vanderbilt’s arguments do not persuade us either that Vanderbilt’s notice to U.S. Fire was timely or that there is a genuine issue of material fact as to whether Vanderbilt’s notice was timely. Vanderbilt’s arguments are focused on whether Vanderbilt gave timely notice of a claim under the policy, rather than whether Vanderbilt gave timely notice of an occurrence under the policy. It is undisputed that Vanderbilt gave U.S. Fire timely notice of a claim under the policy since it informed U.S. Fire of the Craft Litigation soon after the case was filed. What is disputed is whether Vanderbilt gave timely notice of an occurrence.
Under Tennessee law, notice of an occurrence is required when an insured, “becomes, or should become, aware of facts which would suggest to a reasonably prudent person that the event for which coverage is sought might reasonably be expected to produce a claim against the insured.” Reliance Ins., 560 S.W.2d at 618. Vanderbilt contends that the event for which it seeks coverage is the 1960s study, since the Craft plaintiffs asserted claims of negligence, violation of civil rights, and failure to disclose material information based on the 1960s study, in addition to the causes of action the plaintiffs asserted based on the 1940s study. Vanderbilt also points out that the 1960s study occurred during U.S. Fire’s policy coverage years. We hold that there is no genuine issue of material fact as to whether, prior to 1994, Vanderbilt was aware or should have been aware of facts that would suggest to a reasonably prudent person that the 1960s study might reasonably be expected to produce a claim [473]*473against it. Furthermore, we hold that there is no genuine issue of material fact that, since Vanderbilt was aware of an occurrence as early as 1969, Vanderbilt’s twenty-five year delay in not giving U.S. Fire notice of an occurrence constituted untimely notice.
Vanderbilt was aware of, or should have been aware of, a number of facts that would suggest to a reasonably prudent person1 that the 1960s study might reasonably be expected to produce a claim against it. The 1960s study was inextricably related to the 1940s study. This is demonstrated by the premise of the 1960s study, which was to investigate and report the consequences arising from the 1940s study, not to conduct a new experiment. Furthermore, the Craft Litigation was based on both the 1940s study and the 1960s study. By the time the 1960s study was conducted, Vanderbilt was already aware — for over twenty years — of the experiment it had conducted in the 1940s and the fact that it had not informed the participants in the Experiment — either at the time or later — that they ingested radioactive isotopes. Then, by the time the results of the 1960s study were published in 1969, Vanderbilt was aware that: (1) it had not informed the participants in the 1960s study of the purpose of the survey that was sent to them; (2) it had not informed the participants in the 1960s study that they had ingested radiation over twenty years ago; and (3) it had not informed the participants in the 1960s study of the results of the study. Moreover, Vanderbilt was aware of the findings of the 1960s study, which revealed a heightened incidence of cancer in the children involved in the Experiment and the knowledge that four children of the pregnant mothers who had ingested radiation as part of the Experiment had died of cancer in the 1960s.
Even if it could be argued that, with all of this knowledge, Vanderbilt could not have been aware of a claim based on the 1960s study, U.S. Fire has presented undisputed evidence of a 1986 United States Department of Energy request for information from Vanderbilt regarding the Experiment for use in a congressional hearing at which some of the Experiment participants would testify. Vanderbilt immediately informed its attorneys of the inquiry. If Vanderbilt had not been aware of an occurrence under the policy in 1969, the DOE investigation should have made Vanderbilt aware of an occurrence by 1985, yet Vanderbilt still waited to give notice to U.S. Fire until the Craft lawsuit was filed against Vanderbilt in 1994.
Vanderbilt relies on testimony of the researchers who conducted the 1960s study who stated that they did not anticipate liability from either study and that the 1960s study did not demonstrate a cause-and-effect relationship. However, the researchers’ testimony as to a cause- and-effect relationship is belied by their own published results of the 1960s study, which stated that:
for 634 exposed children, one case of leukemia and two cases of sarcoma were discovered. No malignancies occurred in the 655 children in the comparison [474]*474group. This represents a small, but statistically significantly, increase ..., and is consistent with radiobiologic experience ....
Hagstrom et al., supra page 4, at 1. Although this does not constitute overwhelming proof of a link between the Experiment and increased malignancy, it does demonstrate that there was some link and that the researchers involved in the 1960s study were aware of it at the time they published their results in 1969.
The fact that the Vanderbilt researchers testified that they did not expect a claim to be produced from either study is not, by itself, sufficient to prevent summary judgment in favor of U.S. Fire. An insured cannot rely solely on its subjective beliefs as to whether it expected a claim to arise out of an accident. That would divest the insurer of the opportunity to make that determination for itself.
The trouble with the contention of the [insured that they reasonably believed that neither a claim nor a lawsuit would arise out of an accident], is that they assumed to become the judges of the prospect for claim and suit, whereas one of the purposes of the notice of accident is to afford opportunity to the insurer, the responsible party in case the policy is to continue in effect, to decide that very question and to act in accordance with its own judgment.
Nationwide Mut. Ins. Co. v. Shannon, 701 S.W.2d 615, 620 (Tenn.Ct.App.1985) (quoting Malloy v. Head, 90 N.H. 58, 4 A.2d 875, 878 (N.H.1939)). Although the beliefs of Vanderbilt’s lawyers regarding whether they reasonably believed a claim would arise might be given some weight, the opinions of scientific researchers not trained in the law must be given little weight. The testimony relied upon by Vanderbilt is not sufficient to create a genuine issue of material fact as to whether Vanderbilt had knowledge of an occurrence prior to 1994.
The remaining question is whether Vanderbilt had knowledge of an occurrence that was “reasonably likely to involve” U.S. Fire. In other words, we must determine if there is no genuine issue of material fact as to whether Vanderbilt had knowledge of an occurrence that was reasonably likely to produce a claim over $500,000, and therefore trigger U.S. Fire’s excess policy. The district court did not err in relying on other Tennessee wrongful death cases with damages over $100,000 to conclude that, in the aggregate, wrongful death claims based on the deaths of four children born of pregnant mothers who had participated in the Experiment could exceed $500,000 in damages. Moreover, Vanderbilt was not only aware of the four deaths, but it was aware that hundreds of pregnant women and their children had participated in the studjr. A reasonably prudent person could conclude that it might reasonably be expected that these hundreds of individuals could assert claims, as the Craft plaintiffs did in 1994, based on battery, negligence, failure to disclose material information, violation of civil rights, and other causes of action. The four potential wrongful death claims and the hundreds of other potential claims that could have been asserted against Vanderbilt on the basis of the 1940s study and the 1960s study could easily have amounted to over $500,000 in damages. We conclude that there is no genuine issue of material fact as to whether Vanderbilt had knowledge of an occurrence that was reasonably likely to involve U.S. Fire’s excess coverage.
Under Tennessee law:
delay is excusable in the case of an accident which is trivial and results in no apparent harm, or which furnishes no ground for insured, acting as a reason[475]*475able and prudent man, to believe at the time that a claim for damages will arise or that the injury is one insured against.
Munal Clinic v. Applegate, 38 Tenn.App. 280, 273 S.W.2d 712, 716 (1954) (quoting 45 C.J.S., Insurance, § 1056, at 1281). Vanderbilt’s delay is not excusable. The “accident” in this case was neither trivial nor involved no apparent harm. Vanderbilt was aware of numerous facts related to the 1960s study that would indicate to a reasonably prudent person that it could reasonably expect claims to arise that would exceed $500,000 and trigger U.S. Fire’s excess policy. Vanderbilt had knowledge of these facts and their import as early as 1969. The import of these facts was reinforced by a federal investigation in 1985. Nevertheless, Vanderbilt did not give notice to U.S. Fire until 1994. By waiting this lengthy period of time, Vanderbilt did not give notice “as soon as practicable” as required by its policies with U.S. Fire. Therefore, Vanderbilt’s notice was not timely.
Since Vanderbilt gave untimely notice of an occurrence under the terms of its policies with U.S. Fire, Vanderbilt bears the burden of establishing that U.S. Fire was not prejudiced by Vanderbilt’s delay. See Alcazar, 982 S.W.2d at 856. According to Alcazar, factors to consider when assessing prejudice are: (1) availability of witnesses; (2) ability to discover other information; (3) existence of official reports concerning the occurrence; (4) the preparation and preservation of demonstrative and illustrative evidence; and (5) the ability of experts to reconstruct the occurrence. Ibid. The court in Alcazar pointed out that “attempting to prove what information the insurer would have been able to discover had notice been promptly provided would be difficult for either party.” Ibid. However, the court stated, it was “less sympathetic to the insured in this instance, since the insured bears sole responsibility for breaching a term of the contract that was intended to preserve fairness to the insurer.” Ibid.
Vanderbilt asserts that the district court erred in concluding that Vanderbilt was unable to rebut the presumption that U.S. Fire was prejudiced by Vanderbilt’s untimely notice. Vanderbilt points out that witnesses related to the 1960s study were still available and that every author of the 1960s study was deposed. In addition, Vanderbilt states that U.S. Fire had access to one of the directors of the 1940s study, Dr. William Darby. Vanderbilt also asserts that U.S. Fire received all the discovery evidence in the Craft Litigation, consisting of 13,000 pages of documents and 24 expert reports and deposition transcripts reconstructing the events at issue.2
We are not persuaded by Vanderbilt’s arguments. Vanderbilt has not sufficiently rebutted the presumption that U.S. Fire was prejudiced by Vanderbilt’s untimely notice. Vanderbilt points out that a vari[476]*476ety of evidence still exists regarding the Experiment and Survey, but U.S. Fire convincingly responds by demonstrating that a great deal of documentary evidence does not exist, many potential witnesses are no longer living, and there are reliability problems related to those individuals still living. This is particularly true of evidence concerning the 1940s study, which was inextricably related to the 1960s study.
The record indicates that those witnesses who are still available present reliability problems in light of the years that have passed since both studies. As U.S. Fire points out, Dr. Hagstrom, the lead author of the 1960s study, stated 130 times that she did not remember potentially relevant facts during deposition testimony and Dr. Darby, one of the directors of the 1940s study, responded 82 times that he could not remember during his testimony. In addition to the unreliability of those witnesses that are still living, all the individuals who designed and administered the 1940s study, with the exception of Dr. Darby, are no longer alive.
Even though Vanderbilt was forthcoming with the discovery and evidence in the Craft Litigation, this cannot make up for the fact that many documents relating to the Experiment and Survey were destroyed by Vanderbilt, including the raw data for the 1940s study and the 1960s study, which were destroyed by Vanderbilt as early as 1975. As the Tennessee Supreme Court stated in Alcazar:
The purpose of a policy provision requiring the insured to give the company prompt notice of an accident or claim is to give the insurer an opportunity to make a timely and adequate investigation of all the circumstances. An adequate investigation often cannot be made where notice is long delayed, because of the possible removal or lapse of memory on the part of witnesses, the loss of opportunity for examination of the physical surroundings and making photographs thereof for use at the trial, and the possible operation of fraud, collusion, or cupidity. Such a requirement tends to protect the insurer against fraudulent claims, and also against invalid claims made in good faith. If the insurer is given the opportunity for a timely investigation, reasonable compromises and settlements may be made, thereby avoiding prolonged and unnecessary litigation.
982 S.W.2d at 849 (quoting 1 Eric Mills Holmes & Mark S. Rhodes, Appleman on Insurance § 4.30 (2d ed.1996)). Even though U.S. Fire had the information that was produced in the Craft Litigation, Vanderbilt’s delay in giving notice prevented U.S. Fire from obtaining additional information and preserving other information. U.S. Fire was left with the evidence that the Craft Litigation generated rather than the evidence it could have obtained and preserved in preparation for the Craft Litigation or any other litigation related to the studies. U.S. Fire was entitled to this information in order to “prepare for the defense of the claim [and] to protect its interests in an area susceptible to the presentation of spurious claims.” Hartford Acc. & Indem. Co. v. Creasy, 530 S.W.2d 778, 779 (Tenn.1975). This information also would have put U.S. Fire “in a position to settle claims on a knowledgeable basis.” Ibid.
In circumstances that are analogous, although not dispositive, this court recently noted the problems that arise with the passage of time. In Hughes v. Vanderbilt University, 215 F.3d 543 (6th Cir.2000), this court dismissed, on statute of limitations grounds, the claims of a plaintiff who brought suit against Vanderbilt claiming that she had been a participant in a 1945 [477]*477Vanderbilt experiment in which 200 Nashville school children ingested radioactive isotopes. The court noted that statutes of limitations protect defendants and courts in “cases in which the search for truth may be seriously impaired by the loss of evidence, whether by death or disappearance of witnesses, fading memories, disappearance of documents, or otherwise.” Id. at 549 (quoting United States v. Kubrick, 444 U.S. 111, 117, 100 S.Ct. 352, 62 L.Ed.2d 259 (1979)). The requirement of timely notice of an occurrence and the presumption of prejudice to an insurer when timely notice is not given serve similar purposes. Although some evidence relating to the 1940s study and the 1960s study still exists, “the search for truth” is still “seriously impaired” in this case because of the evidence that has been lost over time.
Relying solely on non-Tennessee cases, Vanderbilt argues that U.S. Fire must demonstrate that it was specifically prejudiced by Vanderbilt’s untimely notice. Tennessee law, however, presumes that the insurer is prejudiced and imposes on the insured the burden of rebutting this presumption. Vanderbilt has neither rebutted this presumption nor demonstrated that a genuine issue of material fact exists as to this question. U.S. Fire is entitled to summary judgment on the issue of notice.3
Ill
U.S. Fire conditionally appealed the district court’s decision granting St. Paul’s motion to dismiss and denying U.S. Fire’s motion for leave to amend its complaint. U.S. Fire included St. Paul as a defendant in its original declaratory judgment action on the basis that St. Paul was Vanderbilt’s primary insurer during the years of U.S. Fire’s excess policies. U.S. Fire later filed a motion for leave to amend its complaint to add claims against St. Paul for indemnity and contribution, breach of the duty of good faith, and tortious interference to the extent U.S. Fire is held liable to Vanderbilt. The district court denied U.S. Fire’s motion for leave to amend and granted St. Paul’s motion to dismiss on the basis that U.S. Fire had no liability to Vanderbilt. Since we, too, hold that U.S. Fire has no liability to Vanderbilt, we affirm these decisions of the district court.
IV
For the foregoing reasons, the district court’s grant of summary judgment to U.S. Fire on the issue of notice, denial of U.S. Fire’s motion for leave to amend its complaint, and grant of St. Paul’s motion to dismiss are AFFIRMED.