This is an appeal from a judgment of the district court in favor of the United States in a tort claim action brought by the appellants, Sergeant and Mrs. Joseph S. Rewis. The action was brought upon the allegations of malpractice by the medical officer on duty at an Air Force base hospital in New Mexico, whose examination of the fifteen month old child of the appellants did not disclose aspirin poisoning, which did, in fact, twenty hours later, cause the death of the child.
On this appeal from the findings of fact and conclusions of law by the trial court, the testimony must be viewed, of course, most favorably to the Government. Viewed in this light, the testimony showed the following tragic story. On September 4, 1963, at 9:34 P.M., Sergeant Joseph Rewis, then on active duty at Walker Air Force Base in New Mexico, brought his fifteen months old daughter, Joann, to the base hospital for treatment. Dr. Robert H. Finley was the Medical Officer on duty. He testified that, upon bringing the child to the hospital, Sgt. Rewis reported to him that the child had been well earlier in the day but had that evening started vomiting and having diarrhea, had not eaten well at the evening meal, and had been blue about the lips. Dr. Finley observed the child as she sat on her father’s lap and thereupon examined the ears with an otoscope, the throat with tongue blade and flashlight, and felt the sides of the neck. He looked specifically for any sign of blueness, but saw no evidence of any blueness or cy-*597anosis. He listened to the chest but found it clear. The back of the throat was red and inflamed, and there was hypertrophy of the lymphoid or tonsular tissue on the back of the throat, but no pus or exudate. The child was “fussy and irritable as you would expect of a sick child at almost ten o’clock at night.” Dr. Finley entered on his record that the child was hyperventilated. Upon giving his testimony at the trial, he explained that in using this term he did so as a routine observation and not as an indication of any excessive or pathological condition. He testified that she was breathing somewhat rapidly and was very excited and frightened at being examined by a doctor at that hour. Dr. Finley asked Sgt. Rewis whether the child had had “any drugs or medicine earlier so that I can take this into account.” He testified that “he very definitely stated that she had not.”1
The question whether the child’s father told Dr. Finley about the possibility of her having eaten aspirin is dealt with by the government only as it reflects on the standard of medical care given by the doctor. The government does not take the position that Rewis knew of it but did not report it as a basis for a defense of contributory negligence or any related defense.
Dr. Finley had seen Joann’s mother on an emergency call approximately two hours previously and had found that she was suffering from a viral infection involving the respiratory tract, and he diagnosed the child’s condition similarly. He prescribed a decongestant for the child, baby aspirin every four to six hours, clear liquid diet, benedril expectorant, cough syrup and Coca Cola syrup for the nausea. After Sgt. Rewis took the child home, it continued to be nauseated and unsettled and cried during most of the night. It became “glassy-eyed” and unresponsive. On the following day, Sgt. Rewis brought the child back to the hospital and by the time Dr. Andrew Ageloff, the base pediatrician, who was called in for consultation, examined the child at about noon, it appeared that the child was semi-comatose, ten to fifteen percent dehydrated, and breathing “very, very rapidly.”
Upon being asked whether it was possible that the child had taken any medicine, specifically aspirin Sgt. Rewis said that she could not have taken any medicine. However, when the doctor, after further examination was taking a blood specimen and explaining to Rewis that there was a strong possibility that the child had gotten a large quantity of aspirin, the father “suddenly recalled” that the child, on the day before, at about three o’clock in the afternoon, had been playing with a bottle containing 300 adult aspirin tablets. Immediately, extensive measures were taken to save the child, including intravenous therapy, catheterization into one of the larger veins of the leg for fluid administration, oxygen and moistening environment in an oxygen tent, and, when an artificial kidney was located at Lackland Air Force Base, Texas, an attempt was made to fly the child there, but Joann died at 7:20 P.M. of that evening at a hospital en route.
Drs. Finley and Ageloff both testified that the examination and treatment by Dr. Finley the night before the child’s death adhered to the standard of care for the area. Similar testimony was given in response to a hypothetical question by a civilian doctor, Thomas W. Collier, M.D., of Brunswick, Georgia.
*598Two doctors called on behalf of the appellants, initially, on the basis of a hypothetical question presented by appellants’ counsel which included an assumption that Sgt. Rewis had told Dr. Finley of the suspicion of aspirin poisoning, both testified that the proper standard of care owed to the patient had not been observed by the medical officer. Upon cross examination, however, upon the assumption that the testimony that the medical officer knew of the aspirin ingestion by the child was not true, these two doctors modified their testimony to agree that, on the hypothesis presented by the Government, they could not fault the doctor for the manner in which he treated the child.
With respect to the cause of death and the question whether adequate treatment at 9:45 P.M. on September 4th would have saved the child’s life, the testimony was such as would have permitted the trial court to have found either way. Dr. Collier, without expressing his own opinion, testified that an article written by a Dr. Doame presented a graph from which he concluded that, with a blood-level of 87.2 milligrams, at 19y2 hours after ingestion, the maximum blood-level count would have been 170%, “and he considers the maximum blood-level above 160 as being incompatible with life.” Dr. Finley testified that, “If it turns out that this child had a blood test, which showed a blood sali-cylate level of 87.5 milligrams percent,” he would not have “been able to save the child.” He stated, “This would indicate that a degree of poison, or an amount of poison had been in the body which would be incompatible with life. While you don’t stop trying, it is recognized that this is a fatal level.” Further, Dr. Finley testified, “It is my impression that she probably would have died, however, as I stated before, therapy would have gone on in the hopes that this would have been one of the cases that would be saved.”
“Q. And there are many cases that have been saved? There are worse cases than this, as a matter of fact?
A. Not with the initial blood sali-cylate as calculated back.
Q. Well, you have just said that she might have been saved.
A. She might have been, yes.”
In Dr. Ageloff’s testimony, taken by deposition and without cross examination, he stated: “Quite likely, no matter what therapeutic measures were instituted at the time [of her examination by Dr. Finley] she might have died.” (Emphasis added.)
Dealing with this same question, Dr. Shepherd testified that after a period of four to seven hours, a child with aspirin would show definite symptoms. In response to the question whether this was because “the aspirin would have already gone to work to deteriorate the system,” he stated, “not to deteriorate the system, but the effects of aspirin poisoning would have begun to occur.”
The trial court denied recovery on two bases. The first, as stated in the Court’s opinion, was, “The plaintiffs cannot recover in the instant case because there was no evidence to show that the alleged malpractice was the cause of the child’s illness and subsequent death. Watson v. United States, 5 Cir., 346 F.2d 52 (1965) * * *”
The Court then said:
“In the Watson case, the United States Court of Appeals for the Fifth Circuit affirmed a holding of the District Court that the United States was not liable for the alleged malpractice of a public health service doctor in the outpatient clinic of the public health service hospital of Savannah, Georgia. The action had alleged negligent failure of the doctor to diagnose a serious disease, asserting that the delay in treatment resulting from the failure to diagnose had caused the loss of the claimant’s leg. The Court of Appeals held that where there was no showing to any reasonable degree of medical certainty, that the loss of the leg could *599have been avoided, there can be no recovery. In the case at bar, there was no testimony by any of the experts showing that the child would have recovered if the diagnosis of aspirin poisoning had been made immediately and treatment begun instantaneously.” (Emphasis added).
The trial court erred in its construction of our decision in the Watson ease. What occurred in that case is that the trial court, deciding the case without a jury, made a finding of fact that:
“The only medical testimony adduced on the effect of the delay between February 24 and April 25 was that there was no causal relationship between the said delay and the failure of the bypass graft. The Court finds that had the bypass graft been attempted in February, 1960, there would have been, to a reasonable degree of medical certainty, no greater likelihood of success of this operation, nor would there have been, to a reasonable degree of medical certainty, any benefit derived by plaintiffs through the use of conservative treatment, even if same were to have been initiated in February of 1960.” (Emphasis added.)
The trial court in that case noted that the only testimony before it was to the effect that there was no causal relationship between the delay and the injury. This Court, based on that finding, affirmed the judgment. Here, it is not completely clear, but it appears that the trial court based its judgment in this aspect of the case upon the failure of any witness to testify in effect that “to a reasonable degree of medical certainty” the child’s life could have been saved. No such specific testimony is required in a case of this kind. The Court, upon the record as a whole, including the testimony of all of the witnesses bearing on the subject, is, as the fact finder, required to determine for itself whether a proper diagnosis and treatment on the evening of September 23rd would likely have prevented the loss of this life. See Sentilles v. Inter-Caribbean Shipping Corporation, 361 U.S. 107, 80 S.Ct. 173, 4 L.Ed.2d 142. The trial court might well have found, in light of the evidence before it that if the aspirin poisoning had been discovered and treated on the evening of September 4th, there was a reasonable degree of medical certainty that the child’s life would have been saved.
As the second basis for entering judgment for the United States, the trial court said:
“Moreover, the testimony and the finding on the death certificate concludes that death was caused by aspirin poisoning, revealing that the ingestion of the aspirin was the proximate cause of the injury and resulting death. Thus, there can be no recovery for the alleged malpractice on the part of the attending physician.”
It is not clear what is intended to be conveyed by this paragraph, for it is certain that the Government itself concedes that, notwithstanding the ingestion of the aspirin by the fifteen months old child, there could still be a recovery under the circumstances of this case if, in fact, the medical treatment received by her at the hands of the medical officer in charge fell below the required standards of care and such failure caused the child’s death. The Court’s language cannot, of course, stand for the proposition that the fifteen months old child cannot recover because of contributory negligence, or any similar doctrine.
Finally, the trial court determined that there was no negligence on the part of the admitting doctor. The Court found that Dr. Finley was not advised as to the possibility of the child’s having ingested aspirin in the afternoon, and the' Court found that “she was determined by the doctor to be crying and fussing and breathing rather rapidly.” Other than the finding that the child was crying, a finding discussed later, the Court’s findings were in accord with the statement of facts which we have made here.
In considering whether the trial court’s finding that there was no negligence on *600the part of the admitting doctor is to be affirmed, it is necessary that we consider the basis on which the trial court arrived at this conclusion.
Dr. Shepherd, the only Board-certified pediatrician who testified in the case, gave a very positive and lucid explanation as to why, when a child of tender years, unable to speak for itself, appears before a doctor in a condition of hyperventilation or overbreathing, it is absolutely essential that the cause of such over-breathing be ascertained, because of the very strong possibility that aspirin poisoning may be its cause. Based on the hypothetical question presented to him by appellant’s counsel, Dr. Shepherd clearly and plainly expressed the opinion that inadequate tests were made of Joann by the medical officer in charge. When his attention, however, was called to the hypothesis of the government, Dr. Shepherd’s opinion was considerably modified. However, there were two assumptions made in the government’s hypothesis that bear no support in the record. Dr. Shepherd finally said that if the examining physician had a reasonable basis for concluding that the overbreathing was attributed to a condition he found to be present, there would be no legal duty upon him to make any further tests of the kind that would have disclosed the presence of aspirin. He made it very plain, however, that this opinion was based upon two specific assumptions. One of them was that the examining physician observed the child while sitting calmly in her father’s lap and breathing normally, followed thereafter with the excitement of an examination and crying as the result of which she appeared to be overbreathing. In his answer to the hypothetical question put by the government, which embodied all of the assumptions most favorable to the United States, including the statement, “she was crying and fussing and breathing rather rapidly”, Dr. Shepherd said, “In my opinion, this doctor did not look into this problem as he should have looked into it on a fifteen month old child. Despite what you have told me, it is my opinion that the child, or any child, that is overbreathing is in difficulty. It is a danger warning.” Then the following took place:
“THE COURT: If the doctor asked the father specific question about the baby getting any aspirin, and the father said, no, it did not, then what?
TEE WITNESS: Judge, if I could elaborate a little.
THE COURT: Yes.
THE WITNESS: Even if you ask him it is frequently true that we rarely with children who can’t talk, and with parents who often don’t see the child get poison we must always got to be alert to poison, [sic] We very frequently fail to get any history of any ingested poison and we very frequently turn up with children who are poisoned without a history, and this is one of the basic reasons, this is one of the keys as to why I feel this way, because it is mandatory in the treatment of children.
******
“THE WITNESS: I would just like to make a fair answer to you that in my opinion yes, he should have gone further because it is my opinion that this is a real danger signal if the child is overbreathing and I don’t think it is the same from overbreathing from crying and overbreathing on a pathological basis and it has to be differentiated and this was not done.
“Q. Now, could overbreathing be caused by the lateness of the hour, by a virus, or by the fear of a doctor, would there be overbreathing there ?
“A. There could be overbreathing from fear in the process of crying and excitation, but as I have just said it is mandatory in a child who is over-breathing because the failure to recognize this is so very important, it is mandatory to separate this overbreath-ing from just crying and fussing and excitement, overbreathing on a pathological basis. This has got to be done.
“Q. Therefore, every patient who comes to you with any form of over-*601breathing you immediately treat it for aspirin poisoning?
“A. I didn’t say that.
“Q. Excuse me. That was the basis of my hypothetical. I am asking you.
“A. I find out why they are over-breathing.
“Q. And how do you find this out?
“A. It is necessary — you don’t stop until you find this out and you put them in the hospital, if necessary.
“Q. But how do you find this out?
“A. You put them in the hospital.
“Q. Every child that is overbreathing in your office you put them in the hospital ?
“A. If you want me to answer the question, I will be glad to.
“Q. Excuse me. I apologize for interrupting you ?
“A. All right. You keep checking until you find to your satisfaction the cause of overbreathing. If this involves lab tests, you do lab tests. If you think necessary, if it involves X-rays of the chest, you do X-rays of the chest. If it involves hospitalization or consultant you do this. This has got to be done in a child who has respiratory difficulty, or a child who is overbreathing. Yes, because this in an infant fifteen months old is life and death. This is different from adults. Children get info acidosis. Their electrolytes get out of balance. They are in trouble when they are overbreathing, if it is on an electrolyte basis. This has got to be ruled out.”
Thereafter, after presenting to Dr. Shepherd the entire hypothetical question which included the statement that the child was crying, he asked for the witness’s opinion again. Thereupon, Dr. Shepherd answered and said, “I qualify the answer, yes, by just one question, and that is, one supposition. If he felt that this breathing was due — this is my only problem in all that we have had in our discussion is this breathing, because I feel that this is the key to the whole thing, but if this breathing, in his mind, is due to nothing more that excitation in this hypothetical case, my answer to your question is yes.” [That is, that proper medical standards were followed.]
Counsel for the Government asked Dr. Shepherd, “Would it be correct to say that if a child is overbreathing, if a child is crying vigorously, you cannot determine if it is hyperventilating?” (Emphasis added.)
To which Dr. Shepherd answered: “At the time it’s crying ?”
“Q. Yes, sir.
“A. It would be very difficult.”
Then, again, approaching the end of the examination, Dr. Shepherd said, “His overbreathing is from another point, and this is typical of aspirin, and with a negative chest and overbreathing child, you are obligated to find out why the child is overbreathing because he can die with aspirin poisoning, and this is where we disagree, and if you tell me that the doctor says that the child was overbreathing to the best of his judgment, because observing him sitting quietly, as he must have observed in his father’s lap, he was breathing normally without stress, but overbreathing with stress, yes, I will say he absolutely was to the best of his judgment.” The doctor then stated, “You stated in your hypothetical that he was sitting there quietly when he was not excited? This is the only point of contention that we have, Mr. Clark, I think.”
“THE COURT: I think you have covered it.
* * *»
Now, the point of all of this discussion is that no where in the record is there any evidence that Joann was crying, much less “crying vigorously”, nor was there any evidence that at any time that she was in Dr. Finley’s presence when she was sitting quietly, if she did at any time sit quietly, her breathing was normal and calm, as is assumed by Dr. Shepherd in his answer to Government counsel’s question, or that her breathing was any different under the examination procedures than when she was merely sitting in her father’s lap. It seems clear that with*602out these qualifying assumptions in the Government’s hypothetical case, Dr. Shepherd’s opinion clearly states that there was a failure to pursue proper medical standards in not ascertaining the true cause of the hyperventilation. In other words, Dr. Shepherd’s testimony would be unchanged from the strongest view expressed by him were it not for the improper assumptions, first, that' vigorous crying made it difficult to ascertain the real cause of overbreathing and, second, that the child was observed by the doctor breathing naturally and calmly while not being examined and then breathing rapidly when subjected to the excitement of the examination.
The significance of this is further shown by the fact that the trial court, in its findings of fact, stated that Joann was “crying and fussing and breathing rather rapidly”, indicating that this aspect of Dr. Shepherd’s testimony may well have been a decisive element of the trial court’s decision.
There is another rather slight error in the findings of the trial court, not •supported by the testimony in the record and that is a statement that the child’s father “also stated that she had been blue in one of her lips.” The actual testimony by Dr. Finley is that Sgt. Rewis “stated that he had brought her in at this time because she had suddenly turned blue around the lips and this caused him concern.” A considerable colloquy occurred at the trial in which appellants’ counsel objected to that part of the question by Government counsel in which he embodied the language later used by the Court that the child “had been blue in one of her lips.” In light of the involved discussion by all of the experts with respect to the significance of “blueness” or “cyanosis” as an indicator of aspirin poisoning, we think this inaccurate finding of the facts is of more than passing concern.
There is no substantial difference between the appellants and the United States as to the standards of medical care required in such a situation as is here present. The expert witnesses’ opinions were recognized by both parties to be the proper basis for the trial court’s determination whether the correct standard of care had been given. These opinions were, except for Dr. Finley’s testimony and except for Dr. Ageloff’s deposition, based on hypothetical questions. Of course, in order for the opinion to have any value it must be based on assumptions which the trier of the facts can find to have been proved. See United States v. Cooper, 5 Cir., 277 F.2d 857; International Paper Co. v. United States, 5 Cir., 227 F.2d 201.
The hypothetical question put by the government to each of the witnesses who gave an opinion as to the propriety of Dr. Finley’s procedures, contained the improper assumption that the child was “crying and fussing.” In view of the degree of importance attached by Dr. Shepherd to the need for isolating the cause for overbreathing, and in view of his positive testimony that it would be very difficult to test for overbreathing, if the child was crying, we think this may have led to the trial court’s accepting Dr. Shepherd’s qualified opinion based on the improper assumptions, rather than on his original opinion, which eliminated these assumptions.
So, too, did the hypothesis upon which Dr. Shepherd finally expressed an opinion favorable to the Government include the assumption that Dr. Finley had observed Joann sitting quietly on her father’s lap breathing normally, and thereafter, while being examined, breathing faster, from which Dr. Shepherd concluded this might be a permissible basis for the examining physician to conclude that he had identified the source of the overbreathing. The record does not support any such assumption. There was no proof that Joann was at any time, while in Dr. Finley’s presence, breathing in a normal fashion. The proof is that she was breathing rather rapidly, sufficient for Dr. Finley to have entered a notation on the record of examination that she was suffering from hyperventilation.
Of course, in making its findings, the trial court could completely disregard *603the testimony of any one or more of the experts who testified for either party. However, since the opinion does not discuss the relative contentions made by the parties with respect to the expert testimony but rather seems to place the decision upon the failure to establish causation and upon the theory that the fact of the ingestion of the aspirin alone exculpated the government whatever the degree of performance by the examining doctor, we are unable to determine to what extent the improper assumptions of fact may have affected the trial court’s judgment.
The ultimate burden of proof is, of course, on the plaintiff, who, to recover, must show both failure to give proper diagnosis or treatment and proximate causation. However, plaintiff’s burden of going forward with evidence as to causation is satisfied and a prima facie case on causation is made out once he introduces evidence that the condition was one which would respond to medical treatment seasonably given and that there was a reasonable medical probability that had treatment been seasonably given the child would have continued to live. The question of reasonable medical probability is, of course, normally the subject of expert testimony.
If plaintiff’s prima facie case is made out, then the burden shifts to the government to refute the evidence of reasonable medical probability that the child would have continued to live. In making out the prima facie case and in discharging the ultimate burden the plaintiff need not exclude every possible hypothesis that the child would have died but need only establish by a fair preponderance of the evidence the reasonable medical probability that it would not have died.
While we think this case requires expert opinion on the issue of causation, the trial court may also consider a fundamental inconsistency in the government’s position which, it appears to us, was not satisfactorily explained either in the district court or on this appeal. On the one hand the government claims the child’s condition when she was first examined by the admitting doctor several hours after ingesting aspirin was not sufficiently far from normal that the doctor could be found to have failed to meet a proper standard of care in not recognizing and diagnosing her condition; but on the other hand it says the child’s condition at that time had reached such stage that there was no reasonable medical probability of her continuing to live. In another trial perhaps the expert testimony can shed clearer light on whether it is possible for an infant of this age to be beyond a reasonable probability of salvation by medical help but at the same time display no such objective and discernible symptoms as would disclose to the examining physician, or even alert him, to the existence of the condition.
In order that the court may apply the correct standard of evidence relating to causation, and in order that the court may reconsider its judgment in light of our decision that the finding that the child was crying was clearly erroneous, and that at least one expert opinion was based on several unproved facts, the judgment is REVERSED and the case is REMANDED for a new trial. Upon such trial the parties should be permitted to introduce new evidence or rely on the present record to the extent they wish to do so. However, the trial court should give serious consideration to the contention of the appellant that the government should not be permitted to use the deposition testimony of Dr. Ageloff if he is still available as an Air Force medical officer.
The judgment is reversed.
. Sgt. Rewis testified that when he took the child to the base hospital and when he was giving the history to the doctor, he told Dr. Finley Joann might have had access to some aspirin in the afternoon. However, the trial court discredited this testimony in view of the contrary testimony by the doctor. The resolution of the conflict is certainly not only permissible, but appears to us to be correct for it can hardly be conceived that, if given this clue, the doctor would have treated the patient as he did.