Cascone v. United States

370 F.3d 95, 2004 U.S. App. LEXIS 10459, 2004 WL 1197419

• Court: Court of Appeals for the First Circuit
• Decided: May 27, 2004
• Docket: 03-2164
• Status: Published

Opinion

LYNCH, Circuit Judge.

An intensive-care nurse at the Veterans Affairs Medical Center in the Leeds section of Northampton, Massachusetts (the Leeds VAMC) was convicted of murdering four patients and attempting to murder three others between August 1995 and February 1996 by injecting them with epinephrine, a stimulant that in large doses can trigger heart attacks. See Skwira v. United States, 344 F.3d 64, 67 (1st Cir.2003) (describing the conviction in the context of another case); United States v. Gilbert, 229 F.3d 15, 17-18 (1st Cir.2000) (interlocutory appeal from the nurse’s trial). An extensive federal criminal investigation starting in February 1996 had led to the nurse’s indictment on November 19, 1998. Some of the estates of her alleged victims have since sued the United States for wrongful death under the Federal Torts Claims Act (FTCA), 28 U.S.C. § 2671 et seq. See Skwira, 344 F.3d at 67; Cutting v. United States, 204 F.Supp.2d 216, 218 (D.Mass.2002).

The cases have raised the issue, under the FTCA discovery rule, of when the estate, through the family members, discovered, or in the exercise of reasonable diligence should have discovered, the factual basis for the cause of action. Gonzalez v. United States, 284 F.3d 281, 288 (1st Cir.2002). The FTCA requires a plaintiff to file an administrative claim with the appropriate federal agency within two years after that point. 28 U.S.C. § 2401(b); Skwira, 344 F.3d at 70. In Skwira, this court concluded that the plaintiffs’ October 21, 1999 administrative claim for the wrongful death of Leeds VAMC patient Edward Skwira on February 18, 1996 was not timely. 344 F.3d at 67.

The instant case was brought by the estate of another patient, Michele Cascone, who died on January 28, 1996. The theory of the case is that the nurse, Kristen Gilbert, wrongfully caused Cascone’s heart attack and eventual death by injecting him with epinephrine. Here, the administrative claim was filed on November 10, 2000. Applying the principles of Skivira to the different facts of this case, we conclude that the administrative claim was timely because it accrued on or after November 10,1998. See 28 U.S.C. § 2401(b).

There are at least two salient differences from Skivira that lead us to our conclusion. First, there was nothing inherently suspicious in Cascone’s reported death of an apparent heart attack because Cascone had a long history of heart disease, including congestive heart failure, and at the time of admission to the hospital showed heart irregularities. Indeed, the death certificate said he died of, inter alia, “[cjhronic atrial fibrillation” (emphasis added). Second, the considerable publicity, of which the Cascones were not in fact aware, about the investigation at the Leeds VAMC focused on other patient deaths that were identified as inherently suspicious because those patients had no preexisting heart conditions, and no one from the government or the media publicly drew a connection between the nurse’s misdeeds and Cascone’s death before the November 10, 1998 cut-off date. Government investigators never contacted Cas-cone’s family and never sought to exhume Cascone’s body for further testing, and Gilbert was never indicted for Cascone’s murder. We reverse the dismissal of the plaintiffs claim.

I.

The following facts, presented in the light most favorable to the plaintiff, are drawn from the complaint and materials submitted to the district court on the motion to dismiss. See McIntyre v. United States, 367 F.3d 38, 42 (1st Cir.2004); Muniz-Rivera v. United States, 326 F.3d 8, 11 (1st Cir.2003).

On January 26, 1996, Michele Cascone, age 74, was admitted to the Leeds VAMC with a case of pneumonia. He was placed in the intensive-care unit, Ward C. The hospital records indicate that he had a medical history “significant for atrial fibrillation, wide complex tachycardia, and congestive heart failure” and that he suffered from “insulin dependent diabetes mellitus” and “multiple cerebrovascular accidents.” Upon his admission to the hospital, a physical examination showed that his “[hjeart was irregular, with ... a soft systolic murmur,” and a heart monitor revealed “atrial fibrillation with rapid ventricular response.” The treating physician also noted that Cascone’s heart rate had increased because of “his hypoxemia and general condition.”

Two days later, during the evening of January 28, 1996, Cascone suffered a series of heart attacks and died. He was pronounced dead at 11:25 p.m. The death certificate listed ventricular tachycardia, bilateral pneumonia, chronic obstructive pulmonary disease, and chronic atrial fibrillation as the causes of death. None of Cascone’s family members was at the hospital when he died, as best we can tell from the record. Cascone’s daughter, Janice Kenyon, called the hospital that afternoon, learned of Cascone’s first heart attack, and relayed the message to his wife Nancy Cascone. A Leeds VAMC doctor then called Nancy and said that her husband had had another heart attack. He called again later in the evening to say that her husband had passed away. The doctor asked if Nancy wanted an autopsy done. He did not state any reason why she might want one, and Nancy said that she did not. Michele Cascone was survived by Nancy, three sons (Leonard, Michael, and William), two daughters (Sandra Herk and Janice Kenyon), a sister (Romil-da Dow), and a brother (Joseph) (collectively, the Cascones). At the time, none of the Cascones inquired further into the cause of Michele Cascone’s death.

From Cascone’s death until November 10, 1998 (the cut-off date for accrual), the Cascones lived in various towns in Massachusetts. Although the estate, through Nancy Cascone as executrix, is the only plaintiff, the location of various family members is pertinent. Nancy lived in Rowe; Leonard moved a lot but identified his hometown as Greenfield; Michael lived in Shelburne Falls; William was incarcerated at a state prison in the city of Norfolk; Sandra Herk lived in Athol; Janice Kenyon lived in Leominster; Romilda Dow lived in Newburyport from June to December and in Florida for the other months of the year; and Joseph, who passed away in December 2002, lived in North Andover. The Cascones appear to be in communication with one another. Michael and William said in their depositions that they corresponded regularly with Nancy, and deposition testimony indicated that several of the others had communicated with Nancy both after Cas-cone’s death and after Nancy received a telephone call from a reporter on April 25, 2000 raising the possibility that Gilbert might have killed Cascone.

The government’s argument on appeal is not that the Cascones actually knew of any information or read any news coverage before November 10, 1998 that would have given them a reasonable basis to suspect that the government was responsible for Cascone’s death. Nancy, William, Sandra, Janice,'Leonard, and Romilda testified in their depositions that they had not read any of the articles or seen any of the television coverage relied upon by the government. Michael testified that his mother’s phone call to him after she spoke with the reporter on April 25, 2000 was the first that he had heard of the Leeds VAMC investigation. Joseph was not deposed because he passed away in December 2002. The government’s contention is instead that before November 10, 1998, the Cas-cones should have, in the exercise of reasonable diligence, suspected a causal link between Cascone’s death and government misconduct based on the quantity of news coverage. For present purposes, we assume arguendo the validity of the government’s approach: that a plaintiff who wishes to pursue an FTCA claim may not remain in total ignorance of the pertinent press and media reports. We therefore examine several factors: the geographical scope of the coverage vis-a-vis the family members, the content of the stories, and the degree of press and media saturation.

In June and July of 1996, articles began to appear in various newspapers reporting that a federal investigation had been launched by the VA’s Inspector General’s Office after an internal probe at the Leeds VAMC revealed an unusually high number of heart attack deaths at the hospital. Several reported that suspicious deaths had occurred between late 1995 and early 1996 and that investigators had discovered discrepancies in the inventory of epinephrine, a stimulant that could cause cardiac arrest in high doses. In August of 1996, the story broke that the investigation had led to a grand jury probe focusing on Kristen Gilbert, a nurse who worked the 3 p.m. to 11 p.m. shift 1 on Ward C, the intensive-care unit. Some sporadic coverage of the story continued through 1997.

The story regained some prominence in January 1998 when Gilbert stood trial for allegedly calling in a bomb threat to the Leeds VAMC to retaliate against co-workers and a former boyfriend who worked at the hospital for their participation in the investigation, and again in April 1998 when she was convicted of that crime. United States v. Gilbert, 181 F.3d 152 (1st Cir.1999) (affirming Gilbert’s conviction on appeal). The articles covering the bomb threat story often mentioned the string of suspicious heart attacks at the Leeds VAMC but tended to provide few details about them, such as the time frame or the ward in which they occurred.

Most of the articles upon which the government relies appeared in the Daily Hampshire Gazette and the daily edition of the Springfield Union News, two western Massachusetts newspapers. None of the Cascones subscribe to either newspaper. From 1996 to 1998, 8-10% of households in Franklin County, where Nancy, Leonard, and Michael lived, bought the daily edition of the Springfield Union News and 4-5% bought the Daily Hampshire Gazette. 2 Less than 1% of households in areas where the remaining Cascones lived bought those papers. The two newspapers published over fifty articles between 1996 and 1998 on the investigation into the rash of heart attacks at the Leeds VAMC, at least fifteen of which were on the front page. 3 None specifically mentioned Michele Cas-cone. One July 25, 1996 article in the Daily Hampshire Gazette purported to list all deaths by cardiac or cardiopulmonary arrest at the Leeds VAMC between January 1995 and February 1996, but the list had no entry matching Cascone’s description. Many of the articles that discussed individual patient deaths focused on those without any history of heart problems, such as Henry Hudon, a 35-year-old schizophrenic veteran admitted with flu-like symptoms.

The Daily Hampshire Gazette, however, did publish two articles in September 1997 reporting that an FTCA suit had been filed over the December 1995 death of Rita Morel at the Leeds VAMC. Both articles noted that Morel had preexisting heart problems, with one pointing to a history of congestive heart failure, and the other noting that she had struggled for years with hypertension, strokes, and diabetes. Additionally, the newspaper published a July 23, 1998 article stating that federal investigators had exhumed the body of Stanley J. Jagodowski, who died at the Leeds VAMC in August 1995 after suffering for several years from coronary artery disease and severe peripheral vascular disease.

The Athol Daily News, to which Sandra Herk (Cascone’s daughter) and her husband Arthur Herk subscribed at all relevant times, 4 ran four articles, all between August and November of 1996, on the investigation into suspicious heart attack deaths at the Leeds VAMC. 5 None of the articles identified Ward C or the intensive-care unit as the place where the suspicious deaths occurred. And none mentioned Cascone by name or described as suspicious the deaths of other patients who had a history of heart ailments. In fact, Henry Hudon was the only individual patient whose death was discussed. The two articles discussing Hudon’s death reported that he was 35 years old and quoted his sister as saying “[h]e never had a heart problem, never, never.”

The Greenfield Recorder, which was bought at that time by about 69% of households in the Greenfield city zone (where Leonard lived), by about 36% of households in the surrounding rural areas in Franklin County (where Nancy and Michael lived), and by less than 1% of people in the other Cascones’ hometowns, ran five articles on the investigation, all of which were published between September and November of 1996. 6 Again, none singled out Ward C or the intensive-care unit. And none mentioned Cascone or reported suspicious deaths involving patients with preexisting heart problems. Hudon and Skwira were the only individual patients whose deaths were discussed; no medical history was given for Skwira, and it was emphasized that Hudon had no history of heart problems.

The Boston Globe published a lengthy article on the investigation as the cover story to its April 19, 1998 Sunday magazine. 7 At the time, the Sunday edition of the newspaper (which includes the magazine) was read by around 33% of households in Essex County (where Cascone’s brother Joseph lived), 8 48% of those in Norfolk County (where Cascone’s son William was incarcerated), 12% of those in Worcester County (where Cascone’s daughters Sandra and Janice lived), and 5% in Franklin County (where Nancy and Cascone’s sons Leonard and Michael lived). Cascone’s daughter Janice said in her deposition that she “often buy[s] the Sunday” edition of the Boston Globe. The article, which discussed numerous individual patient deaths, did not mention Cas-cone and focused primarily on patients without a history of heart troubles, such as Hudon. The article did, however, describe the death of Rita Morel, who was identified as having a history of congestive heart failure. The article reported that Morel’s brother had filed an FTCA claim with the VA over her death.

The daily edition of the Boston Globe, which Janice (Cascone’s daughter) “sometimes” read and which was purchased at the time by around 31% of those in Norfolk County (where Cascone’s son William was incarcerated), 19% of households in Essex County (where Cascone’s brother Joseph and sister Romilda lived), and less than 7% of those in the other Cascones’ hometowns, also ran eleven articles on the investigation between August 1996 and August 1998. 9 None mentioned Cascone. But one short article, published on July 25, 1998 at page B7, did describe a suspicious death involving a patient with past heart troubles. The story reported that federal investigators had exhumed the body of Stanley Jagodowski, who died at the Leeds VAMC in August 1995 after “suffering] from coronary artery disease and other vascular problems for several years.”

The Boston Herald, which was read by around 12% of households in Essex County at the time and which Arthur Herk and William Cascone occasionally read, ran six articles on the investigation between August 1996 and April 1998. 10 None mentioned Cascone or other deaths involving patients with a history of heart problems.

In addition to the press coverage, Boston-based Channel 5 television news ran four broadcasts between July and September of 1996 on the investigation. 11 Although none of the Cascones live in Boston, Cascone’s two daughters (Sandra Herk and Janice Kenyon) said in their depositions that they sometimes watched Channel 5 news and his sister (Romilda Dow) said that she occasionally watched the local news on Boston-based television stations. Cascone was again not mentioned. The only individual death discussed was that of Henry Hudon, whom the relevant story described as having “no history of heart problems.”

The Cascones stated in their depositions that between Cascone’s death in January 1996 and the November 10, 1998 cut-off date for accrual, no one contacted them with suspicions about Cascone’s death. They were not contacted by reporters, any of the other eight families bringing FTCA suits against the VA, the VA Inspector General’s Office, or the state police, which had been enlisted to help with the federal investigation.

Nor were they contacted by the United States Attorney’s Office, even though an August 9, 1996 article in the Springfield Union News quoted William Welch, the AUSA handling the case, as saying, “We’re going to contact anyone and everyone who may have relevant information [about the deaths at the Leeds VAMC].” The paper noted that Welch’s statement “includ[ed] patient families.” In November 1996, April 1997, and July 1998, three families were contacted for permission to exhume bodies of patients who had died at the Leeds VAMC, according to a July 24, 1998 Springfield Union News article. But Cas-cone’s family was not one of them.

Meanwhile, the VA Inspector General’s Office announced that details of the federal investigation would be kept secret until it was complete. A July 17,1996 Springfield Union News article quoted Jon Wooditch, a spokesman at the VA’s Office of Inspector General, as saying “[b]ecause it’s an ongoing inquiry, it’s not in the best interests of anybody to make comment.” Both the Daily Hampshire Gazette and Springfield Union News reported on July 26, 1996 that veterans’ agents from various local towns had a private meeting with officials at the Leeds VAMC but were unsuccessful in discovering any additional information. The newspapers also reported that AUSA Welch repeatedly refused to confirm that Gilbert was a suspect in the grand jury investigation, to disclose how many deaths were being investigated, or to make public the results of the analysis of several bodies that had been exhumed. On September 17, 1996, the district court, at the request of both Welch and Gilbert’s attorney, sealed hearings related to the grand jury investigation. As late as April 19, 1998, the Boston Globe reported that some families that had brought suit were having trouble uncovering facts about the deaths because of the ongoing criminal investigation.

On November 19, 1998, a federal grand jury indicted Gilbert for murder of four Leeds VAMC patients and the attempted murder of three others. At this point, the November 10, 1998 cut-off date for accrual of Cascone’s claims had already passed. The indictment did not allege Cascone to be one of Gilbert’s victims.

Then, on April 25, 2000, the Springfield Union News reported that Gilbert may have murdered a fifth veteran at the Leeds VAMC, Michele Cascone. As best we can tell, this was the first mention of Cascone’s death in the press or television news coverage surrounding the deaths at the Leeds VAMC. The story said that although Gilbert had not been charged with Cascone’s murder, the prosecution had moved in pre-trial proceedings to be given permission to introduce evidence about his murder to the jury to demonstrate a pattern to Gilbert’s killings. The article noted that “[d]etails relating to [Cascone’s death], previously contained in sealed documents, were revealed in a pretrial hearing” (emphasis added). In particular, the article reported that, according to AUSA Ariane Vuono, a co-worker had observed Gilbert carrying a vial of epinephrine between Cascone’s third and fourth heart attacks and that Gilbert had jokingly asked the co-worker, “Want some epi?”

Later that day, Fred Contrada, who had authored the article, called Nancy Cascone for comment on her husband and the Gilbert trial. According to Nancy Cascone’s deposition testimony, she replied, “I don’t know any Gilbert trial.” Contrada then told her that Gilbert was charged with killing four patients at the Leeds VAMC and that her husband may have been the fifth. Contrada asked Nancy whether anyone had called her from the VA or the hospital, and she said no. The Cascones say that this was the first time that any of them had heard of Gilbert or her possible role in Michele Cascone’s death. After the call, Nancy bought a copy of the Springfield Union News and read Contrada’s article. She then notified the rest of the family and asked William to write a letter on her behalf contacting an attorney.

On November 10, 2000, David Mech, Nancy Cascone’s attorney and special administrator of Michele Cascone’s estate, filed a claim for wrongful death and personal injury with the Department of Veterans Affairs. The claim stated, “[Cascone] was in the care of one Kristen Gilbert when he was injected with a lethal dose of epinephrine which put [him] in cardiac arrest and caused his death. The hospital, its servants, agents, and employees were negligent in the care of Cascone and Gilbert’s supervision.” On April 6, 2001, the VA denied the claim as untimely.

In due course, on June 15, 2001, Mech, as special administrator of the estate, filed suit in federal district court against the United States under the FTCA, claiming negligence, conscious pain and suffering, wrongful death, and recMess conduct. See 28 U.S.C. § 2401(b) (allowing six months from the denial of the administrative claim to file suit). The complaint alleged that the United States had violated its duty of care to Cascone by “failing to provide the medical care required for his health and recovery” and by “failing to take reasonable action to protect him from harm.” The complaint was later amended on November 29, 2001 to substitute Nancy Cas-cone as the executrix of the estate.

On March 3, 2003, the United States moved to dismiss for lack of subject matter jurisdiction, arguing that the administrative claim had not been timely filed. On July 9, 2003, after a hearing and a review of the exhibits submitted by the defendants, which included deposition transcripts and copies of various stories in the press and television news, the district court granted the motion. The court reasoned that “even assuming the plaintiff was subjectively ignorant of any investigation at the VAMC,” the accrual standard is “objective.” The court found that Cas-cone’s “sudden” death after being admitted with a “non-life threatening medical ailment” put the plaintiff on notice of the injury. The court further determined that media reports from July of 1996 onwards would have put a reasonable person on notice before November 10, 1998 of a causal connection between possible government wrongdoing and Cascone’s death. Nancy Cascone, as executrix of Cascone’s estate, timely appealed.

II.

The FTCA provides, in relevant part, that “[a] tort claim against the United States shall be forever barred unless it is presented in writing to the appropriate Federal agency within two years after such claim accrues.” 28 U.S.C. § 2401(b). Because the FTCA is a waiver of sovereign immunity, it is strictly construed. Skwira, 344 F.3d at 73-74. Timely filing of an administrative claim is a jurisdictional prerequisite to suit. Id. at 71.

Although normally a tort claim accrues at the time of injury, the Supreme Court in United States v. Kubrick, 444 U.S. 111, 100 S.Ct. 352, 62 L.Ed.2d 259 (1979), created a “discovery rule” exception for FTCA claims involving medical malpractice. See McIntyre, at 60; Skwira, 344 F.3d at 73. The Court held that such claims accrue when a plaintiff has factual knowledge of both the existence and the cause of his injury. Kubrick, 444 U.S. at 119-120, 100 S.Ct. 352. The plaintiff need not, however, know that the acts legally constitute medical malpractice. Id. at 122, 100 S.Ct. 352. The rationale is that once a plaintiff knows the fact of the injury and the identity of the parties who caused it, he can find out if he was a victim of malpractice by consulting doctors or lawyers. Id. This court has extended Kubrick’s discovery rule to FTCA claims for wrongful death in settings involving deliberate acts, where “the prospects of any claim against the government were so hidden that a reasonable plaintiff would not have been alerted to their existence.” Skwira, 344 F.3d at 84 (Boudin, C.J., concurring); see also McIntyre, at 40.

Under the discovery rule, the test is whether plaintiff knows, or in the exercise of reasonable diligence should have known, the factual basis of the cause of action, including the fact of the injury and the injury’s causal connection to the government. McIntyre, 367 F.3d at 42; Skwira, 344 F.3d at 78 (“[A] medical malpractice claim has accrued ‘once a plaintiff knows of the injury and its probable cause.’ ” (quoting Gonzalez, 284 F.3d at 289)). Whether a plaintiff should, in the exercise of reasonable diligence, have discovered necessary facts is an objective inquiry. McIntyre, at 42. Nonetheless, the particular circumstances of individual plaintiffs can be relevant to the outcome. The issue is whether a reasonable person similarly situated to the plaintiff would have known the necessary facts. Id. at 60. Where the plaintiff resides can be a factor when information about the underlying facts is limited to certain geographical areas. Id. at 60. This is often true, for example, when notice is based on local television or press reports. Id. How often a plaintiff communicates with other family members who have access to more information can also be relevant. Id.

Here, the government, relying on press and media reports, argues that the estate objectively should have been alerted to the fact that Cascone’s seemingly natural death was suspicious. The government contends that the reports should have prompted the estate to conduct a reasonable investigation and eventually to have concluded that there was a reasonable basis to believe that Gilbert, and hence the government, wrongfully caused the death. We hold that a similarly situated person would not have had an objective basis before November 10, 1998 to think that Cascone’s seemingly natural death was suspicious. Thus, no duty to investigate further was triggered before then.

We focus our analysis on the facts available to a reasonable person in Nancy Cas-cone’s position because she is the named plaintiff, as the executrix of the estate. Because she communicates regularly with the other Cascones, and there is no evidence that the Cascones are estranged from one another, we also look to information available to the other Cascones, to the extent that such information might reasonably be expected to have been communicated to Nancy.

At the time of Cascone’s death in January 1996, none of the Cascones had a reasonable basis to suspect foul play. 12 The question is whether media reports after Cascone’s death were sufficient to trigger a duty to inquire before November 10, 1998 into the circumstances of Cas-cone’s death. Most of the news coverage in this case appears to have been in Springfield and Hampshire County newspapers. But none of the Cascones lived in either place, and during the relevant times the Springfield Union News and the Daily Hampshire Gazette were bought by less than 10% and less than 5%, respectively, of households where the Cascones did live. In these circumstances, a plaintiff in Nancy Cascone’s position could reasonably be ignorant of the articles in those two newspapers. The same reasoning applies to stories on the Springfield television news, as there is no evidence in the record that a significant portion of people in the Cas-cones’ hometowns had access to or watched Springfield news stations.

Again, accepting arguendo the government’s premise that people can, in certain circumstances, be charged with knowledge of relevant news, we turn our focus to roughly thirty scattered stories in the Greenfield Recorder, the Athol Daily News, the Sunday Republican, the Boston Herald, the Boston Globe, and Boston-based Channel 5 News, all of which either are read or watched at least occasionally by one or more Cascones or have a significant circulation in counties where they live.

The combination of coverage in Nancy Cascone’s local newspaper, the Greenfield Recorder, and in media read or watched by other Cascones or those in the other Cas-cones’ hometowns was at least arguably enough to put Nancy Cascone on notice that an investigation was being conducted into heart attack deaths at the Leeds VAMC. But most of the coverage from these sources contained few details about the deaths being investigated. Only a few stories mentioned that the suspicious deaths occurred in Ward C or the intensive-care unit or that they happened on the 3 p.m. to 11 p.m. shift, and a significant number did not indicate that the deaths under investigation had occurred between late 1995 and early 1996. To the extent that those details were reported, they were rarely in the lead paragraph or the headline.

Knowledge of an investigation into heart attack deaths at the Leeds VAMC, with few additional details, would not be sufficient to give Nancy Cascone a reasonable basis to suspect that the suspicious deaths included Michele Cascone’s and, thus, to trigger a duty to investigate further. None of the Cascones had a reasonable basis to suspect that Cascone had died of anything but his preexisting heart condition, even if they should have known there was an investigation. “Where the plausible explanation [for death] is one of purely natural causes ..., there is initially no reasonable basis for supposing [misconduct]. It is not the purpose of the discovery rule to encourage or reward simple paranoia.” Thompson v. United States, 642 F.Supp. 762, 768 (N.D.Ill.1986); see also Diaz v. United States, 165 F.3d 1337, 1340 (11th Cir.1999).

That Gilbert may have murdered some patients at the Leeds VAMC is not alone sufficient to provide a reasonable basis for the family of every patient who died of a heart attack at the hospital, including those patients admitted with a history of heart ailments, to suspect that their loved one was murdered. Cf. McIntyre, at 60 (the revelation that the FBI caused the death of one informant against James Bulger and Stephen Flemmi by leaking his identity to them did not provide a reasonable basis to believe that the FBI was responsible for all those who died while informing against Bulger and Flemmi); Skwira, 344 F.3d at 80 (finding accrual not just based on press reports of an ongoing investigation into deaths at the Leeds VAMC, but relying also on suspicions about Skwira’s death voiced by government investigators, an autopsy demonstrating that his death certifícate misstated the cause of death, and an acknowledgment by family members that they were surprised that Skwira had died of a cardiac event).

We ask next whether any of the media reports from these sources described suspicious deaths involving patients who had a history of heart problems and, if so, whether those reports were so prominent or pervasive that Nancy Cascone should be charged with knowledge of them, even though she testified in her deposition that she did not read them. Two Boston Globe articles did mention the deaths of patients with past heart troubles. On April 19, 1998, the Boston Globe Sunday Magazine published a seven-page story on the Leeds VAMC deaths. Toward the end of the article, it reported that Maurice Morel had filed a suit against the VA for the death of his sister Rita Morel at the Leeds VAMC on December 26, 1995. One sentence, on the last page of the article, mentioned that Rita Morel had been “plagued” by a number of illnesses, including “congestive heart failure.” That history was not discussed further or mentioned again in the article. The other mention of patients with preexisting heart conditions came in a July 25, 1998 article. The article, only 201 words long and printed on page B7, reported that the body of Stanley Jagodow-ski, who died at the Leeds VAMC on August 22, 1995, was being exhumed as part of the federal investigation. One sentence noted that Jagodowski “suffered from coronary artery disease and other vascular problems for several years.”

The two references in the Boston Globe in April 1998 and July 1998 do not provide a sufficient basis to charge Nancy Cascone with the knowledge, either directly or through her family members, that the deaths of other Leeds VAMC patients with heart problems were being investigated or viewed as suspicious. Nancy Cascone testified in her deposition that she did not see any of the articles in the record, and the Boston Globe circulates to only about 5% of households in her county on Sundays and about 2% on other days. And none of the other Cascones saw or could be reasonably expected to see the articles. 13

Other than those two articles, none of the stories described suspicious deaths involving patients with a history of heart trouble. Indeed, to the extent that the stories included profiles of individual deaths, they identified those deaths as suspicious because those patients had no known preexisting heart ailments. The Greenfield Recorder, the Athol Daily News, and Channel 5 News ran stories on Henry Hudon, a 35-year-old schizophrenic veteran with no history of heart problems, and the Boston Herald’s story on the death of Kenneth Cutting, a 40-year-old veteran, quoted his father as saying that Cutting “was dying of multiple sclerosis, but he didn’t have a bad heart that we knew of.” One of those stories — an Associated Press story that ran in both the Greenfield Recorder and the Athol Daily News on October 5, 1996 — suggested that the suspicious deaths included only those in good health or with no history of heart disease: “Some [patients] died suddenly at the end of years of good health. Others had endured senility, alcoholism, Parkinson’s disease or multiple sclerosis, but it was their hearts that failed. One schizophrenic veteran died at age 35 for no apparent reason at all.”

A second factor influences the outcome of the case: others, with better sources of information than the Cascones had, drew no connection between Gilbert and the death of Michele Cascone, despite intense interest in Gilbert’s misdeeds. Despite the government’s investigation and its promise to talk to the families of suspected victims, no one from the government ever contacted the Cascones. Furthermore, the government did not seek to exhume Cascone. The VA Inspector General’s Office, the U.S. Attorney’s Office, and the state police, all of'which were working on the government investigation, surely spoke with doctors at the Leeds VAMC about which deaths were suspicious and worth investigating. Apparently, Cascone’s death was not on that list until very late. 14 When Gilbert was indicted, Cascone was not named as a victim. A reasonable plaintiff in Nancy Cascone’s position, having access to much less information, should not be held to a higher standard. See McIntyre, at 60; Attallah v. United States, 955 F.2d 776, 780 (1st Cir.1992) (“The police did not have sufficient information to bring charges against the [government officials involved] until 1987. We believe [plaintiffs] could not have been more efficient.”).

Furthermore, other private individuals who had a strong interest in investigating the full extent of Gilbert’s wrongdoing and had spent years digging up information on the topic were also not led to investigate Cascone’s death. Cascone’s death certifi cate was a matter of public record on file with the city clerk’s office in Northampton and stated that he had died at the Leeds VAMC of, inter alia, “chronic atrial fibrillation” on January 28, 1996. According to the Daily Hampshire Gazette, there were only fourteen heart attack deaths at the Leeds VAMC between January 1995 and February 1996. Yet, no one from the media called the Cascones to inquire about Cascone’s death, even though both television and press reporters had covered the story extensively and interviewed a number of other victims’ families (at least seven families were contacted in connection with the Boston Globe Sunday Magazine article alone). A number of deaths were profiled in media reports, but Cascone’s was never one of them, until April 2000. Similarly, no one from the families of the other eight victims who brought FTCA suits contacted the Cascones to gather information about his death. '

The facts in Skwira were fundamentally different. In Skwira, this court affirmed the dismissal of an FTCA claim by the family of a patient who had died at the Leeds VAMC on the ground that the administrative claim was not timely filed. 344 F.3d at 80-82. There, the court relied primarily on facts revealed to the family by government investigators. Id. at 80. Nothing indicated that Skwira had a history of heart troubles, and government investigators had not only been provoked to inquire further into the cause of death long before the cut-off date for accrual, but had also expressly shared the reasons for their suspicions with the patient’s family, asked for permission to exhume the body, and told the family when the analysis revealed that the death certificate misstated the cause of death. Id. at 68. Moreover, in Skwira, the plaintiffs had actual knowledge of relevant press and media reports; the case did not involve charging them with constructive knowledge. Id. at 81. Indeed, Skwira’s daughter testified that when she read the press reports about the Leeds VAMC investigation, “it was like a light bulb went off because I knew that was exactly what had happened to my father.” Id.

Skwira primarily involved a different question than that in this case: whether plaintiffs, fully cognizant of facts giving rise to suspicions about the death and the cause of injury, could wait for the government to finish its investigation before filing an administrative claim. Id. at 85 (Boudin, C.J., concurring). That is not the situation here.

III.

Nancy Cascone did not have a reasonable basis to suspect before November 10, 1998 that Michele Cascone died of anything other than a heart attack caused by his preexisting heart ailments. Accordingly, we reverse the dismissal of the plaintiffs claim against the United States and remand the case for further proceedings consistent with this opinion.

1 . There are some minor discrepancies in the articles about the exact time of the shift. Although most identify the shift as running from 3 p.m. to 11 p.m., some describe it as running from 4 p.m. to 12 a.m.

2 . Throughout this opinion, the cited circulation numbers include both single copy sales and subscriptions.

3 . At least eleven front-page articles were published in these two papers during the summer of 1996, all of which said that the deaths under investigation happened between late 1995 and early 1996 and about half of which singled out Ward C or the intensive-care unit. There was also a front-page article in the Sunday Republican (the Sunday edition of the Springfield Union News), which circulates to around 25% of households in Franklin County. That article, which focused primarily on a parallel investigation of a VA hospital in Missouri, did not specify the ward or time frame in which the deaths occurred. Nor did it discuss the deaths of any individual patients identified as having a history of heart disease.

4 . Arthur Herk stated in his deposition that he generally read the paper "semi-well” on a nightly basis, provided that he did not get interrupted. Although Sandra Herk testified in her deposition that she herself did not often read the paper, she said that her husband sometimes told her about what he read there.

5 . Only one was on the front page. Three stated that the suspicious deaths were heart attacks in late 1995 and early 1996, and two headlines asked, "Is Government To Blame in [VA] Hospital Deaths?”

6 . It is not clear from the record whether any of the five articles ran on the front page. Three identified the suspicious deaths as occurring in late 1995 and early 1996.

7 .The article was headlined "The Mysterious Deaths on Ward C.” The subheading was "At the VA hospital in Northampton, 63 patients on one ward died in the span of 14 months. Some family members now ask: WHAT KILLED THEM?” The article noted a higher than normal cluster of cardiac-related deaths on Ward C in late 1995 and said that investigators suspected that someone had murdered patients using epinephrine. It stated that suspicion had focused on a nurse who worked the 3 p.m. to 11 p.m. shift on Ward C.

8 . Cascone’s sister, Romilda Dow, lived in Essex County only during the months from June to December, so was not in the area at the time.

9 . Four of the eleven articles were only one paragraph long. None were on the front page of the paper, and only one was on the first page of the Metro section. Only five articles specified that the suspicious deaths were cardiac arrests. And only two identified the ward or shift in which the deaths occurred.

10 . None of the articles were on the front page. Five noted that the suspicious deaths involved heart attacks, two said that the deaths occurred in Ward C on the 4 p.m. to midnight shift, and three noted the time frame as between November 1995 and February 1996.

11 .There was also coverage in the Springfield television news. But none of the Cascones live in Springfield. Michael and Nancy Cas-cone said in their depositions that they do not receive local television news because they use satellite television services. And there is no evidence in the record that any of the other Cascones watched or had access to those television stations.

12 . The government initially argued in its brief that Cascone’s "admittance to a hospital with an illness that the family understood to be non-life threatening, coupled with almost im mediate death from a wholly different condition, would put any reasonable person on notice that further investigation was called for.” The government withdrew this contention at oral argument, conceding that “they weren't on any inquiry notice at that point.” The government was wise to withdraw this argument. Cascone was 74 years old and had serious heart ailments, and his death certificate noted atrial fibrillation, one of his preexisting heart problems, as a cause of death. The fact that Cascone was admitted for pneumonia rather than a cardiac condition does not undermine the inference that he died of natural causes. It was perfectly reasonable for Nancy and the other Cascones to believe, as Sandra Herk (Cascone’s daughter) said in her deposition testimony, that the pneumonia exacerbated his preexisting heart conditions or that his heart problems simply happened to flare up at that point, independently of his other illness.

13 . Hypothetically, Nancy could have learned about the information from Janice (Cascone’s daughter), who “often” bought the Sunday edition and "sometimes” bought the daily edition, from William (Cascone's son), who about once a month read old copies of the Boston Globe that he found on a trash can in the prison where he was incarcerated, or from Joseph (Cascone’s brother), who lived in Essex County, where the Sunday circulation of the Boston Globe was 33% and the daily circulation was 19%. She might also have heard something from Romilda (Cascone’s sister), who lived in Essex County when the July 25, 1998 article (but not the Sunday magazine article) came out and said she "oc casion[ally]" read the Boston Globe. But Nancy did not.

That Nancy heard nothing from the other Cascones was not objectively unreasonable. Janice, Romilda, and William all testified in their depositions that they did not recall reading any news stories about the investigation until after 1998. Joseph passed away in December 2002 at the age of 94 and thus was never deposed. There is no evidence, though, that Joseph read the Boston Globe regularly or that he was sufficiently in touch with his sister-in-law, Nancy, to be expected to convey what he had read to her. It would not be unreasonable for Janice, Romilda, William, • and Joseph to have missed reading the Boston Globe on April 19, 1998 and July 25, 1998. None of the four appears to have had a regular subscription, and Romilda was not even in town when the first article was published.

14 . The April 24, 2000 hearing revealed that the government had secret information, under seal, suggesting that Gilbert might have killed Cascone. But the record does not explain whether the government discovered this information because it had been provoked to investigate Cascone’s death and, even if it had, whether it had been provoked to do so before November 10, 1998.