Darlene Metzen, individually and as administratrix of the estate of her deceased husband, Matthew J. Metzen, appeals a judgment of the United States District Court for the Eastern District of New York, Leonard D. Wexler, Judge, in favor of the United States. Judge Wexler found that the Department of Veteran Affairs/Medical Center in Northport, New York (the "Northport VA"), a duly operated facility of the United States of America, was not subject to liability for medical malpractice under the Federal Tort Claims Act, 28 U.S.C. Secs. 2671-2680 (1988 & Supp. IV 1992), for failing to put Mrs. Metzen's husband on a low-cholesterol diet because of a concern that a low-calorie diet would be harmful to his liver condition. We now vacate and remand to the district court for proceedings not inconsistent with this opinion.

On July 10, 1989, Darlene Metzen filed a wrongful death claim with the Department of Veterans Affairs (the "DVA"). The DVA denied this claim on March 16, 1990. On April 25, 1990, she filed an amended complaint because the DVA mistakenly had sent the records of another patient to her. On April 30, 1990, the DVA denied the amended complaint. She filed a request for reconsideration. Over the next six months, the DVA took no action. Therefore, on April 10, 1991, she brought this action in the United States District Court for the Eastern District of New York, Leonard D. Wexler, Judge. Following trial, on April 15, 1993, Judge Wexler granted judgment for the defendant. On June 4, 1993, Mrs. Metzen filed a notice of appeal. In an order entered on June 22, 1993, the appeal was dismissed for failure to comply with the Civil Appeals Management Plan, but was reinstated pursuant to a scheduling order dated August 2, 1993.

The district court had jurisdiction to hear the case under 28 U.S.C. Sec. 1346(b) (1988). This court has jurisdiction pursuant to 28 U.S.C. Sec. 1291 (1988).

Under the Federal Rules of Civil Procedure, "[f]indings of fact, whether based on oral or documentary evidence, shall not be set aside unless clearly erroneous, and due regard shall be given to the opportunity of the trial court to judge the credibility of the witnesses." Fed.R.Civ.P. 52(a). " '[A] finding is "clearly erroneous" when although there is evidence to support it, the reviewing court on the entire evidence is left with the definite and firm conviction that a mistake has been committed.' " Anderson v. Bessemer City, 470 U.S. 564, 573, 105 S. Ct. 1504, 1511, 84 L. Ed. 2d 518 (1985) (quoting United States v. United States Gypsum Co., 333 U.S. 364, 395, 68 S. Ct. 525, 542, 92 L. Ed. 746 (1948)). Moreover, "[i]f the district court's account of the evidence is plausible in light of the record viewed in its entirety, the court of appeals may not reverse it even though convinced that had it been sitting as the trier of fact, it would have weighed the evidence differently. Where there are two permissible views of the evidence, the factfinder's choice between them cannot be clearly erroneous." Id. at 573-74, 105 S.Ct. at 1511. Finally, "[w]hen findings are based on determinations regarding the credibility of witnesses, Rule 52(a) demands even greater deference to the trial court's findings; for only the trial judge can be aware of the variations in demeanor and tone of voice that bear so heavily on the listener's understanding of and belief in what is said." Id. at 575, 105 S.Ct. at 1512. Nevertheless, the trial judge may not "insulate his findings from review by denominating them credibility determinations, for factors other than demeanor and inflection go into the decision whether or not to believe a witness. Documents or objective evidence may contradict the witness' story.... Where such factors are present, the court of appeals may well find clear error even in a finding purportedly based on a credibility determination." Id.

In this case, we are interested in several of Judge Wexler's findings of fact and the conclusions he drew by applying medical malpractice law to these findings. See Metzen v. United States, (Findings of Fact and Conclusions of Law), No. 91-1278, Findings of Fact at pp 6-11, and 15-16 (April 13, 1993) ("Findings of Fact" or "Conclusions of Law"). We have examined the record before us and conclude that these findings are clearly erroneous because the court has confused concerns about a "low-calorie " diet with concerns about a "low-cholesterol " or "cholesterol-reducing " diet. This confusion taints the entire reasoning of the opinion leaving us with the definite and firm conviction that a mistake has been committed. Also, the court's conclusion that earlier, more aggressive treatment of Metzen's hypercholesteremia (high cholesterol level) would have been ineffective because such methods take "at least one to two years" and perhaps even "five years" to be effective cannot be supported by the record.

To explain our conclusion, we set forth the facts of this case as before us in the record, giving "due regard" to the court's findings of fact and indicating where these findings are clearly erroneous and where certain inferences therefrom also would be clearly erroneous.

Matthew J. Metzen was a Marine Corps veteran of the Vietnam War. While in Vietnam, Metzen was exposed to herbicides and other horrors of war. From the time of his return from military duty until the date of his death, he suffered from post traumatic stress syndrome punctuated by flashbacks, sleeplessness, and nightmares. On June 4, 1988, at the age of 39, Metzen died from a heart attack. The autopsy report lists the cause of death as arteriosclerotic and hypertensive heart disease.

On February 14, 1986, Metzen sought emergency medical treatment from the Northport VA. At that time, a doctor at the Northport VA physician and faculty member at Stony Brook Medical Hospital, saw Metzen, "identified [him] as having exposure to herbicides when he was in the military and ... as having hypertension." Transcript of Trial at 60, Metzen v. United States, No. 91-1278 (E.D.N.Y. Dec. 9-10, 1992) ("Transcript of Trial"). At that time, the doctor ordered a variety of blood tests and chemical analyses that indicated that Metzen had a cholesterol level of 394 and an elevation of one enzyme indicating some liver damage. Hospital records indicated that Metzen's cholesterol level was high.

Metzen was seen again on March 18, 1986. At that time, Dr. Karabedian, who was in charge of the herbicide clinic, and a medical physician in charge of the general medicine clinic, did an initial history and physical examination. The history revealed only that Metzen had hypertension (high blood pressure). The doctors reviewed the results of the February 14 tests and subsequently treated Metzen for hypertension with a variety of drugs including hydrochlorothiazide, "a diuretic and an effective medication for initial treatment of hypertension in many cases." Id. at 61. They did not treat Metzen for high cholesterol.

B. Overview of Northport VA's Treatment of Metzen

Based on these initial visits, the Northport VA diagnosed Metzen as having several medical problems including post traumatic stress syndrome, medical complications due to exposure to herbicides while on military duty, hypertension, and elevated liver enzymes. See, e.g., Veterans Administration Hospital Records, Joint Appendix at 294a. Although Metzen was obese (he was 6 feet 2 inches tall and weighed 243 pounds), and hyperlipidemic (high cholesterol and triglyceride levels), the Northport VA failed initially to diagnose or treat either of these medical conditions. See Findings of Fact at p 6; Transcript of Trial at 71. In fact, the Northport VA did not treat Metzen's hyperlipidemia or obesity until December 1987 at the earliest, a year and ten months after he was first examined. See Transcript of Trial at 72 (testimony of Dr. Weinstein, a Northport VA physician and faculty member at Stony Brook Medical Hospital).

C. Northport VA's Treatment of Decedent's Medical Condition

There is no question that Northport VA physicians treated Metzen for his post traumatic stress syndrome and his exposure to herbicides. Furthermore, neither of these conditions was or is considered to be a major cause of coronary heart disease, Metzen's cause of death. Thus, the hospital's treatment of Metzen for these medical problems is not involved.

We do, however, examine the hospital's treatment of Metzen for hypertension, hyperlipidemia, and obesity because it is these conditions that are, and have been known to be, contributing causes of fatal coronary heart disease. See Transcript of Trial at 5; see, e.g., Max L. Feinman, M.D. & Josleen Wilson, Live Longer: Control Your Bloodpressure at 23, 62-63, 83-84 (1977) (citing the now-famous 1949 Framingham, Massachusetts Study by the National Heart Institute as one of the first studies to document hypertension, "[o]besity, smoking, high levels of cholesterol and other fats in the blood, and lack of physical exercise" as the "risk factors for heart disease"); Jon Van, Cholesterol Danger Is Reiterated, CHI.TRIB., Nov. 28, 1986, at 4 ("The link between reduced cholesterol levels and a lower risk of heart attack was reiterated Friday in a series of reports published in the Journal of the American Medical Association"); U.S. Dept. of Health and Human Servs., The Cholesterol Connection, FDA CONSUMER, Feb. 1986, at 10 available in LEXIS, News library, Allnws file (linking high cholesterol levels with risk of heart attack and citing a 1984 National Institute of Health Study which recommends diet to reduce fat and cholesterol intake); Al Rossiter, Jr., Heart Disease Risk Goes Up with Cholesterol, UPI, Nov. 13, 1985, available in LEXIS, News library, Allnws file ("The risk posed by high levels of cholesterol in the blood is well known"). We also review the hospital's treatment of Metzen's elevated liver enzymes condition to the extent such treatment influenced the hospital's treatment of Metzen's hypertension and hyperlipidemia.

On June 19, 1986, Metzen returned to the Northport VA for a checkup, "at which time it was Dr. Karabedian's sense that he might need more aggressive treatment, and his medication dosages were increased. Furthermore, he was referred to the hypertension clinic at that time with the expectation that this special treatment which concentrates on that would more effectively and appropriately lower his blood pressure." Transcript of Trial at 62 (testimony of Dr. Weinstein).

There is some dispute in the evidence at this point. Dr. Weinstein testified that Metzen failed to show up at the hypertension clinic for further modification of his medication and stopped taking his medication altogether, thus contributing to the raising of his blood pressure to a level characterized as "dangerously high" by 1987. On the other hand, the contemporaneously written progress notes by a Northport VA nurse on January 29, 1987 reveal that Metzen told the hospital official that he could no longer attend the hypertension clinic because it interfered with his work schedule. See Veterans Administration Hospital Records, Joint Appendix at 252a. In any event, by August 1987, Metzen's blood pressure was finally brought under control. At his last visit to the Northport VA, in June 1988, just before he died, his blood pressure was 120 over 90 or 130 over 80, both readings within a normal range.

The court made specific findings as to Metzen's hypercholesteremic condition and the Northport VA's treatment of this condition. First, we note that the court did not make any findings as to Metzen's hyperlipidemia, a more general condition which includes both hypercholesteremia and high triglyceride levels. Instead, the court focused on the hospital's treatment for Metzen's hypercholesteremia. The record reveals that Metzen suffered from both conditions.

Second, the court found that Metzen's cholesterol level was measured five times between February 1986, and April 1988, four times by the Northport VA, and once by an outside physician ("Healthnet"). The results were as follows:

See Findings of Fact at pp 6-13.

Finally, the court found that "[i]n December 1987 ... Northport VA physicians [determined] that the best approach to treating decedent's hypercholesteremia (elevated cholesterol levels) was to first use diet, and then, if necessary, to consider drugs." See id. at p 11. The court did not specifically find whether or when Northport VA physician Dr. Weinstein placed Metzen on a diet. It only found that "caution had to be used in placing decedent on a stringent low calorie diet since this diet could contribute to worsening his liver disease." Id. (emphasis added). The court further found that although drug treatment was available for patients suffering from hypercholesteremia, "[s]everal of the cholesterol lowering drugs which were available at that time caused damage to the liver and others elevated the triglyceride level." Id.

The record amply sustains the cholesterol measures. The record does not, however, sustain certain of the court's findings regarding the type of treatment and the timing of treatment. We take each in turn to determine whether any of these findings are clearly erroneous.

a. Type of Treatment For Hypercholesteremia

As the court found, between 1986 and 1988 there were two standard medical treatments available for patients suffering from hypercholesteremia: drugs and diet. We begin our analysis with the hospital's actions regarding drug therapy. We then turn to the hospital's diet treatment. In each instance, we address three questions: (1) whether the hospital utilized this form of treatment in Metzen's case, (2) whether any of Metzen's other medical conditions affected the hospital's administration of treatment for hypercholesteremia, and (3) whether or not the Northport VA's failure to treat Metzen's hypercholesteremia more aggressively was a substantial factor in decedent's sudden coronary death.

(1) Drug Therapy

The record is clear that Northport VA physicians did not treat Metzen's hypercholesteremia with drugs, although the court did not make an explicit finding to this effect. Given our review of the record as a whole, and the court's finding that cholesterol lowering drugs available at the time either caused liver damage or elevated triglyceride levels, the hospital's failure to treat Metzen's hypercholesteremia with drugs cannot be faulted. We cannot, therefore, characterize this implicit finding as clearly erroneous.

(2) Diet

We do take issue with the Northport VA's treatment of Metzen's hypercholesteremia by diet, and whether any of Metzen's other medical conditions affected the hospital's administration of the treatment.

As we stated earlier, the court did not make any specific findings as to whether Northport VA physicians ever treated Metzen by diet for his hypercholesteremia. The record reveals that the court tended to discount diet as an effective treatment in hypercholesteremia:

Court: Well, a prior witness, a member of your profession said if one was not on a diet control, it only would go 10 or 20 points difference. Do you agree with that?

Witness: No. I think it depends on the particular disorder ... and on the particular individual as to the extent to which diet can be effective.

Court: I mean I have here on 12/10/87, 586. On ... 4/8/88, 272. That is a tremendous difference.

Witness: Yes.

Court: Is there any justification to that.

Witness: Only the possibility that a diet was effective, but we have no absolute one-hundred percent proof. I'm not aware that how well he was following the diet, except he was losing weight and he knew what he was doing.

Court: Diet couldn't bring it down that much, in my opinion. What do you think about that? I'm not the doctor.

See Transcript of Trial at 116 (testimony of Dr. Weinstein). Later, in questioning the defendant's expert, the court reiterated the same concerns:

Court: ... how do you justify or how do you explain the cholesterol on 12/10/87 of 586, going down on 4/8/88, that's four months, to 272?

Witness: I heard Dr. Weinstein's testimony so I can comment on that. It's a very marked drop. And I think that this either indicates very effective treatment for his hyperchole-steremia--

Court: Do you really believe it could drop that radically on diet?

Witness: Yes, it can drop that fast, if one is very rigid. As a matter of fact, it happened to my son who has hyperlipidemia and this can occur if one is very rigid about the diet. So that is one possibility.

Court: Very rigid, means not eating.

Witness: It means virtually eating no animal fat, etcetera. You are right. It's a very rigid diet, only somebody who is very compulsive and not a hedonist would be able to follow.

Id. at 139 (testimony of Dr. Dolgin). This testimony foreshadows the court's ultimate confusion between a cholesterol-reducing diet and a low-calorie diet, which is explained in greater detail below. We now return to what the record reveals about diet.

Dr. Weinstein testified that on December 11, 1987, "I informed him, instructed him on a low cholesterol diet, and gave him some literature to follow in order that over the ensuing months before he returned for his next visit, that he would begin to slowly lose weight and would begin to start to control his elevated cholesterol which from a numerical estimate was clearly elevated." Id. at 66 (emphasis added). Dr. Weinstein further testified that "I instructed him on not only a low fat diet by a modest reduction in calories and how to go about doing that." Id. at 72 (emphases added).

From this testimony we understand that at least three different diets were available treatments for an obese, hypertensive, hyperlipidemic man: low-calorie, low-cholesterol, and/or low-fat diet. These terms are fairly self-explanatory. A low-calorie diet consists of eating below a certain number of calories a day; a low-cholesterol diet consists of eating less than a certain number of milligrams of cholesterol a day; and a low-fat diet consists in eating a low percentage of fat in proportion to percentage of protein and carbohydrate intake. Furthermore, a cholesterol-reducing diet might consist of a low-cholesterol, a low-fat, or both a low-fat and a low-cholesterol diet because low fat intake has been shown to lower blood cholesterol. See generally Robert E. Kowalski, The 8-Week Cholesterol Cure: How to Lower Your Blood Cholesterol By Up to 40 Percent Without Drugs or Deprivation (Rev. ed. 1989) 237-46 (citing various low fat diets available in the mid-1980s, including his own and the Pritikin diet, as methods for reducing blood cholesterol); Al Rossiter, Jr., Doctors Sometimes Neglect High Cholesterol, UPI, Dec. 4, 1986, available in LEXIS, News library, Allnws file ("Most people can reduce their blood cholesterol levels by eating less saturated fats and cholesterol, found particularly in meat and dairy products, and by substituting polyunsaturated fats such as vegetable fats for part of the saturated fat").

While a person can be on a low-calorie, low-cholesterol, low-fat diet, these diets are not coterminous. For example, a diet solely consisting of a couple of pounds of whole-grain rice and pasta per day would be high in calories but low in fat and cholesterol.

Thus, according to Dr. Weinstein, he instructed Metzen to go on a low-cholesterol, low-fat, calorie-reducing (but not necessarily low-calorie ) diet. Yet, there are no notations in the hospital records that any doctor or nurse ever placed Metzen on a low-cholesterol, a low-fat, or a calorie-reducing diet. In fact, at trial, Dr. Weinstein admitted that the hospital records do not reveal that he placed Metzen on a diet. See Transcript of Trial at 66. The overall record does support the possibility that Metzen did go on a cholesterol-reducing diet sometime between December 11, 1987, when his cholesterol count was 586, and April 8, 1988, when his cholesterol count was down over three hundred points to 272. After all, something must have caused this dramatic reduction in Metzen's cholesterol level and we know that it was not drugs because everyone concurs that Metzen was not prescribed and did not take drugs known to reduce cholesterol levels. The record also supports the proposition that Metzen might have gone on a calorie-reducing diet because during this time Metzen lost up to 14 pounds. See id. at 72. In short, it is unclear whether Dr. Weinstein himself placed Metzen on some sort of diet. The facts also are consistent with a finding that an outside source (or that Metzen himself) placed Metzen on a diet. Any of these findings, therefore, would not be clearly erroneous.

What is clearly erroneous is the court's finding that the Northport VA physicians needed to wait to place Metzen on any kind of diet because of his elevated liver enzyme condition. See Findings of Fact at p 11 ("caution had to be used in placing decedent on a stringent low calorie diet since this diet could contribute to worsening his liver disease") (emphasis added). Obviously, a problem with any low-calorie diet is that if the calorie intake is too low, the person might not intake enough nutrients necessary to sustain life. This problem is particularly acute in the case of a person with elevated liver enzymes who is placed on a low-calorie diet. The testimony makes this point clear:

Q: Did you have any concerns upon first seeing Mr. Metzen about putting him on a low calorie diet?

A: Yes. Initially he presented with his abnormal liver function tests and it is an accepted standard that patients who have abnormal liver functions, particularly if they have active hepatitis or otherwise, should not be on a low calorie diet; that it could contribute to worsening of their liver disease; they should be well nourished during the repair of their liver.

Transcript of Trial at 68.

Immediately after this testimony, the Northport VA lawyer began to question Dr. Weinstein about Metzen's cholesterol level. Therefore, it is perfectly understandable that a court might confuse "concerns about a low-calorie diet" with "concerns about a low-cholesterol diet," as the judge did in this case. See Findings of Fact at p 6. The court apparently latched on to the idea that "patients who have abnormal liver functions ... should not be on a low calorie diet ... [because] they should be well nourished during the repair of their liver." He then confused low-calorie diet with low-cholesterol or cholesterol-reducing diet. He misinterpreted the testimony to mean that doctors must exercise caution in placing patients like Metzen who have elevated liver enzymes on a low-cholesterol diet because such diets can aggravate the liver condition.

The confusion was compounded by later testimony. When the court asked Dr. Dolgin, "[d]o you really believe [cholesterol level] could drop that radically [from 586 to 272 in four months] on diet," the doctor replied, "[y]es ... if one is very rigid." The court asked the doctor to clarify: "Very rigid, means not eating," to which the doctor replied, "[i]t means virtually eating no animal fat." The doctor continued by stating, "You are right. It's a very rigid diet." Transcript of Trial at 139. The above testimony, read in the context of the entire record, and the court's finding that "caution had to be used in placing decedent on a stringent low calorie diet since this diet could contribute to worsening his liver disease," Findings of Fact at p 11 (emphasis added), reveals that the court confused a low-calorie diet with the type of diet on which a doctor would place a hypertensive patient with an astronomically high cholesterol count.

Thus, in reading the record as a whole, we find that it was clearly erroneous for the court to have concluded that Metzen's liver condition would have precluded earlier treatment of Metzen's hypercholesteremia because, at the time Metzen was in the care of the Northport VA, there were several cholesterol-reducing diets that were not necessarily low-calorie.

Furthermore, the court's conclusion that a more aggressive cholesterol-reducing diet would not have been a "substantial factor in reducing decedent's risk of sudden coronary death" because such methods take "at least one to two years" and perhaps even "five years" to be effective was also clearly erroneous. See Rossiter, Doctors Sometimes Neglect High Cholesterol, (citing a National Heart, Lung and Blood Institute Study finding that twenty-three percent of surveyed physicians in the United States are withholding diet therapy from patients with abnormally high levels of blood cholesterol and quoting Survey Director, Beth Schuker, as stating that "this is a concern because these patients are at especially high risk and could benefit from diet change"). After all, had Northport VA physicians placed Metzen on a cholesterol-reducing diet to begin with, they would have had a minimum of twenty-six months (over two years) to reduce a cholesterol level of 394 to a normal level of 200 rather than only six months to reduce a cholesterol level of 586 to 200. We think it can be presumed that had the Northport VA physicians immediately placed Metzen on such a diet in conjunction with treatment for his hypertension, the combined treatment would have bought time for Metzen. See, e.g., Larry Thompson, A Maze of Cholesterol-Reducing Drug; Doctors Stress That Diet Is Still the First Defense, WASH. POST, Nov. 17, 1987, at Z9 ("The National Cholesterol Education Program urges diet as the first line of defense against high cholesterol. The recommended diet alone should succeed in reducing an individual's cholesterol by 10 to 15 percent. And that reduces the risk of heart disease by 20 to 30 percent, a significant amount, said Dr. James Cleeson, program coordinator").

In short, the court's summary dismissal of early treatment because such methods take "at least one to two years" and perhaps even "five years" to be effective, Findings of Fact at p 16, especially in light of Dr. Weinstein's concerns about "preventative" medicine (discussed infra ), is not supported by the record read as a whole. We also hold that the court's finding that the Northport VA failed to put Metzen on a low-cholesterol diet because of a concern that a low-calorie diet would be harmful to his liver condition was clearly erroneous.

b. Timing of Treatment--The Fasting Issue

With regard to the timing of treatment, the record does not clearly sustain (1) which of these tests were taken in a fasting state; (2) if a test was not taken in a fasting state, why not; and (3) whether this finding makes a difference in the diagnosis and treatment of Metzen.

The court found that none of the tests were taken in a fasting state, and therefore blames Metzen for any delay in his medical treatment. Yet, only the first lab report indicates that Metzen's cholesterol count was taken in a non-fasting state. This corresponds with the fact that on the day this test was taken, Metzen went to the hospital for emergency treatment. Dr. Weinstein testified that none of the counts were taken in a non-fasting state. When asked how he knew that, he testified that Metzen told him. See Findings of Fact at pp 6-9; Transcript of Trial at 65.

The progress notes do not relay any information concerning the fasting state of tests subsequent to the initial emergency blood tests run. Furthermore, although Metzen's medical problems, in particular, "hypertension," "post traumatic stress syndrome," and "exposure to herbicides," are listed repeatedly throughout the hospital's medical records, there is no indication in any place in the record that any doctor or health care provider ever noted, treated or considered the lab report findings of abnormally and critically high cholesterol counts.

We find in the record testimony of Dr. Weinstein which might have confused the court on this point causing the court to give more weight to the significance of non-fasting than warranted:

Q: Doctor, now turning to Mr. Metzen's high cholesterol or hyperlipidemia, you noted before this was initially indicated by the VA in February of '86 when he was first seen there?

A: Umm-hmm.

Q: What was done by the VA for Mr. Metzen's high cholesterol?

A: Attempts were made to try to characterize the pattern of his high cholesterol, his high lipid state in his blood. It was initially important to determine which category of high fat state he was in. Was he somebody who had simply high cholesterol without high triglycerides? Was he an individual that has very high triglycerides? Was he somewhere in between? How exactly high was his cholesterol? And Dr. Karabedian made at least two attempts to try to get fasting blood tests during 1986 in order to characterize that. Mr. Metzen actually never came in in the fasting state and, therefore, the triglyceride level which we were trying to determine what the resting-fasting state would be, was never properly evaluated.

Transcript of Trial at 64-65.

Thus, Dr. Weinstein never testified that it was the cholesterol level that was not properly evaluated, but the triglyceride level that was "never properly evaluated." Regardless of whether either level was properly evaluated, the uncontroverted testimony of the expert witnesses reveals that Metzen's levels were so high that any deviation resulting from the tests having been taken in a non-fasting state would be negligible. This being so, it cannot be explained why Dr. Weinstein or another Northport VA physician failed to put Metzen on a cholesterol-reducing diet immediately.

Where the evidence is contradictory and could sustain two permissible views, "the [court]'s choice between them cannot be clearly erroneous." Anderson, 470 U.S. at 574, 105 S.Ct. at 1511. Nevertheless, the evidence also indicates that for a person with very high cholesterol, such as Metzen, the average error in cholesterol counts taken in a non-fasting state would be "20 or 30 points." Transcript of Trial at 20 (testimony of Dr. Burack); see also id. at 137 (testimony of Dr. Dolgin) ("All you have to do is starve yourself for twenty-four hours, and your cholesterol will go down 20, 30 milligrams perhaps"). In Metzen's case, with a level hundreds of points above normal, such errors would be negligible. Thus, the inference that decedent's hypercholesteremia "was not properly evaluated because Metzen did not take the test in a fasting state" is clearly erroneous. See, e.g., Findings of Fact at pp 7-8.

In short, although it may not be clearly erroneous to find that none of the tests were taken in a non-fasting state, the inference deduced from this proposition is clearly erroneous.

The court found that at "[a]n initial screening ... done at the Northport VA on March 18, 1986 ... an electrocardiogram was performed ... to see if decedent suffered from any other medical problems. The electrocardiogram was normal." See Findings of Fact at p 6. There is no dispute with this finding of fact. We do, however, dispute the implication the court made by making this finding of fact: that the Northport VA was unaware of Metzen's heart disease because the electrocardiogram failed to reveal the problem. To explain the problem with this implication we must examine the testimony of Metzen's treating physician.

Dr. Weinstein testified that even though he treats between one and two thousand patients a year, he specifically remembers Metzen because Metzen "was a young man [who] ... I was concerned about because of my concern about preventative medicine.... Most of the patients we see are patients that are elderly, have already irreversible disease and it concerns me a great deal to try to prevent many of these diseases." Transcript of Trial at 57. Dr. Weinstein also admits that he had questioned Metzen about "the possible consequences of high bloodpressure or any consequence in change in his cardiac status," because "that is the standard kind of questioning that you ask a patient who has hypertension, high cholesterol, and is being followed for any of those manifestations of end organ damage, damage to the heart and so forth." Transcript of Trial at 74. The record also indicates that the Northport VA invested a great deal of time and effort in treating Metzen's hypertension.

Hypertension coupled with hypercholesteremia presents a particularly heightened risk of coronary heart disease: "A high level of cholesterol in the blood can be extremely destructive to the body because it contributes to atherosclerosis. Obviously, when you have high blood pressure and high cholesterol you are going to be particularly susceptible to heart attack and stroke." Id. at 83 (emphasis in original). Furthermore, the medical community has known about this link for several decades now:

Most of our information [about the link between hypercholesteremia and heart attack] comes from the results of certain linked events from which we have drawn these conclusions. Immediately following World War II, during years of prosperity when there was butter in the larder and a steak on every plate, the number of deaths in the United States from atherosclerosis-related heart disease steadily increased. In the early 1950s medical science began to recognize the link between atherosclerosis and foods high in cholesterol, and there was a widespread public education program to reduce the American's use of butterfat and other foods high in cholesterol. Unexpectedly, the effort to reduce coronary deaths also triggered a reduction in deaths from strokes. Wadsworth's Veterans Administration study confirmed the effect of diet on the incidence of both heart attack and stroke.

Id. at 83-84 (emphasis added).

Thus, given that Dr. Weinstein was particularly concerned about Metzen because he believed that Metzen was too young to suffer from some irreversible ailment, and given that Dr. Weinstein knew that Metzen was hypertensive, obese, and hyperlipidemic, three known causative factors in the development of coronary heart disease, it follows that Dr. Weinstein must have had some concern that Metzen might suffer from heart disease. These commonly known medical facts, coupled with Dr. Weinstein's testimony, compel us to conclude that the Northport VA must have been on notice that Metzen was a high risk candidate for a heart attack.

The court also makes several findings of fact which imply that even if the hospital were liable for malpractice, Metzen was contributorily negligent. We now find these findings to be clearly erroneous.

After his initial visit to the Northport VA, Metzen relied on the Northport VA for treatment. This practice was in conformity with Northport VA policy that a patient under its care may not be seen by an outside physician. Dr. Weinstein explained the importance of this policy: "it is hazardous to the patient's health that two physicians not communicating effectively would be responsible for the same condition." Transcript of Hearing at 59.

Once again, we find that Dr. Weinstein's words come back to haunt him. Northport VA policy makes a patient completely dependent on the hospital for treatment. The hospital holds itself out as comprised of a staff whose members communicate with one another. Yet, we find numerous instances on the record in which a failure to communicate among hospital caretakers contributed to the waste of Metzen's life.

For example, Dr. Weinstein testified that Metzen stopped coming to the clinic. Yet, hospital records indicate that Metzen continued to attend his regularly scheduled appointments either at the outpatient department, the medical clinic or the herbicide clinic. Of the 37 scheduled appointments, Metzen missed only five. Moreover, he was always seen within two weeks of a missed appointment.

The court's findings, which proclaim that decedent "failed to keep his appointments at Northport VA," Findings of Fact at p 10, attribute more significance to Metzen's actions than the record supports.

2. Failure To Take Cholesterol Tests In a Fasting State

We have discussed this problem in detail in section IV.C.2.b. of this opinion. We concluded that although it may not be clearly erroneous to find that none of the tests were taken in a non-fasting state, the inference deduced from this proposition is clearly erroneous. Thus, supposing that the tests were taken in a non-fasting state, and given Metzen's astronomically high cholesterol levels, it was clearly erroneous to conclude that the defendant could not properly evaluate the test findings.

The court also found that in April 1988, Metzen went to see a private physician, Dr. Munro, to complain about chest pains, and that Dr. Munro's tests "showed a pattern of abnormality suggesting decedent was having acute myocardial infarction" at about that time. Findings of Fact at p 12. Yet, even though Metzen visited the Northport VA on three occasions subsequent to his visit with Dr. Munro, "[o]n no occasion did he mention to anyone at the Northport VA that he experienced chest pain, that he had been seen by a Dr. Munro, or that he had an abnormal electrocardiogram.... Rather, plaintiff resumed his work as a manual laborer, doing strenuous work at Calverton cemetery." See Findings of Fact at p 13.

The court's finding at paragraph 12 is supported by the record. However, the district court apparently inferred that Metzen's failure to impart to a Northport VA caretaker the details of his April 1988 doctor's visit somehow nullified the hospital's responsibility beginning in February 1986 to treat properly Metzen's hypertension, hypercholesteremia, obesity, and heart disease. Such an inference would be clearly erroneous. A lack of some medical information late in the day cannot make up for the fact that Northport VA caretakers were on indubitable notice that Metzen was at grave risk of a heart attack. From the very beginning, the hospital knew that Metzen exhibited three of the five major warning signs of heart disease: hypertension, hypercholesteremia, and obesity. Metzen did not smoke and the record is silent about physical exercise except that his job as a cemetery caretaker seemed to require some exercise on his part. Although Metzen exhibited these warning signs, during the last two years of his life, while he was a Northport VA patient, neither his medical history nor physical was ever repeated. Although at the initial visit, on February 14, 1986, an electrocardiogram was performed, it was not repeated. See Veterans Administration Hospital Records, Joint Appendix at 208a.

Our review of the record has " 'left us with the definite and firm conviction that a mistake has been committed.' " See Anderson, 470 U.S. at 573, 105 S.Ct. at 1511 (quoting Gypsum, 333 U.S. at 395, 68 S.Ct. at 542). This case presents to us a relatively young Vietnam veteran who relied on the Northport VA for medical treatment of a variety of distinct medical conditions, one being hypercholesteremia. And although this man's cholesterol count was measured at critically high levels and continued to rise, the defendant did not place Metzen on one of the cholesterol-reducing diets available at the time. See, e.g., Allan Parachini, Heart Association Serves Up a Diet You Can Live With, L.A. TIMES, Sept. 4, 1986, at 16 (citing American Heart Association's "Healthy Heart" dietary guidelines).

We have detailed where we believe the court's findings are clearly erroneous. We now assess briefly how these facts would be applied to medical malpractice law, leaving the actual conclusions of law for remanded proceedings.

Under the Federal Tort Claims Act, New York State law governs the substantive law of this case. 28 U.S.C. Sec. 2674. To prevail, Mrs. Metzen must show by a preponderance of the evidence that the treating physicians breached their duty of care to her husband, thereby causing his injuries, see, e.g., Stuart v. Ellis Hosp., 198 A.D.2d 559, 560-61, 603 N.Y.S.2d 212, 214 (3rd Dept.1993), because (1) they lacked the knowledge and skill possessed by the average physician in the community at the time; or (2) they failed to exercise ordinary and reasonable care in the application of their knowledge and skill; or (3) they failed to use their best judgment in the application of this knowledge and skill. Littlejohn v. State, 87 A.D.2d 951, 451 N.Y.S.2d 225, 226 (3d Dept.1982); see also Sitts v. United States, 811 F.2d 736, 739 (2d Cir.1987); Faughnan v. Big Apple Car Serv., 828 F. Supp. 155, 165 n. 7 (E.D.N.Y.1993).

A. The Northport VA's Duty and Breach of that Duty

We have established that the Northport VA physicians were on notice that Metzen was a prime candidate for heart disease. Medical knowledge since at least the late 1970s has declared hypertension to be the single most important cause of heart attacks. Further, at least since the early 1980s, medical knowledge also has declared high cholesterol and obesity to be significant contributory causes of heart attack. See Feinman & Wilson, Live Longer, supra, at 23, 62-63, 83-84. Yet, the Northport VA failed to use ordinary and reasonable care in the application of this knowledge in treating Metzen. Although the Northport VA medical staff provided some treatment for hypertension, they neglected to place Metzen on a low-cholesterol diet, thereby failing in this respect to lessen the risk of a heart attack.

B. Northport VA's Breach Caused Decedent's Injuries

The court held that the Northport VA was not liable for the wrongful death of Metzen because the "Northport VA physicians complied with the applicable standards of medical care in the New York community. Between February 1986 and April 1988 it was appropriate medical practice to treat hypercholesteremia conservatively and not to prescribe medication considering his liver problems." See Conclusions of Law at p 3. Therefore, the court never explicitly reached the issue of causation.

We do not reach the issue of causation on appeal. On remand, however, we direct the court to examine this issue in light of our discussions of (1) Metzen's remarkable response to diet between December 1987 and April 1988, (2) the uncontroverted testimony of the experts that diet can bring cholesterol counts down from dangerously high to normal levels, and (3) the timing of treatment.

For the foregoing reasons, the judgment below is vacated and the case is remanded for proceedings not inconsistent with this opinion.