The plaintiff Kimberly McClelland sued the defendant Dr. Larry Ozenberger for medical malpractice. Dr. Ozenberger attended McClelland during her pregnancy and delivery. In order to prevent adjacent tears in the birth canal when the baby’s head emerged, Dr. Ozenberger performed an episiotomy, which is an incision in the genital area. After she left the hospital, McClelland noticed occasional loose stool and drainage from her vagina. She saw Dr. Saleh, a specialist in obstetrics and gynecology, who diagnosed the condition as a rectovaginal fistula. Dr. Saleh referred McClelland to Dr. Campbell, a colon/rectal surgeon. Dr. Campbell confirmed the diagnosis of Dr. Saleh and removed the fistula.
*229In response to the inquiry of McClel-land’s counsel as to his opinion concerning her fistula difficulty, Dr. Campbell responded by letter, . the episiotomy evidently cut into the rectal wall and was not repaired. I think that to have a complication like this would not be unusual, but to do nothing about it is strictly poor medical practice.” Thereafter, McClelland executed an authorization that allowed Dr. Campbell to release medical records to defense counsel but specifically disallowed discussions between Dr. Campbell and defense counsel. Notwithstanding, there were discussions between them about the case, twice. After these ex parte discussions, plaintiffs counsel designated Dr. Campbell as an expert witness to be called at trial. The deposition of Dr. Campbell was admitted at the trial on behalf of the defendant.
The plaintiff McClelland before trial moved to suppress the deposition of Dr. Campbell, a treating physician, as incompetent evidence on the ground that he breached the physician-patient relationship established by § 491.060(5) by his ex parte discussions with defense counsel without the consent of the patient or order of the court. The plaintiff moved also that the deposition testimony be excluded on the ground that the ex parte conversations were breaches of the fiduciary duties owed by the physician to the plaintiff as patient.
McClelland’s theory of recovery was that Dr. Ozenberger either negligently cut into the roof of her rectum or negligently failed to recognize and treat the fistula at her post-delivery check-up. The theory of the defense was that there were several possible causes for her rectovaginal fistula: such as, cryptitis, necrosis or mucosal prolapse.
McClelland contended that Dr. Campbell’s opinion of the episiotomy episode changed after the ex parte discussions with defense counsel. She argued that Dr. Campbell retreated from his earlier stated position that Dr. Ozenberger’s treatment “constituted poor medical practice.” Dr. Campbell’s office file recorded a visit by defense counsel with the notation: “Discussion on Kimberly: Told that epi was not a malpractice thing but probably skimmed over the crypts and glands which then resulted in the abscess and fistula.” The defendant contended that Campbell’s notation and his deposition testimony did not reflect a change in his characterization of McClelland’s case. His use of the phrase “skimming over crypts,” rather was merely another way to express “a tiny cut or entry into the crypt.” The defendant contended also that Dr. Campbell used “abscess” to describe the natural history of anal crypti-tis as a cause of rectovaginal fistula.
The case was tried to the jury and a verdict was returned for defendant Dr. Oz-enberger. The plaintiff appealed the judgment and asserted as the ground of error the denial of her motion to suppress the evidence testimony of Dr. Campbell. This court reviewed the contentions of error and rendered opinion in McClelland v. Ozenberger, 805 S.W.2d 264 (Mo.App.1991).1 The court could not determine from the record “how much Dr. Campbell’s position changed, if at all.” Id. at 267. The opinion noted that Dr. Campbell testified at deposition “that a rectovaginal fistula is not an unusual episiotomy complication.” Id. It noted also Dr. Campbell’s opinion that “if it does develop, then a doctor’s failure to treat falls below accepted standards of medical practice.” Id. It was conceded that the fistula developed just after the episiotomy.
The court posed the question of law on appeal as whether “an unauthorized ex parte conversation between defense counsel and the plaintiff’s treating physician in a medical negligence case ipso facto render the physician incompetent to testify at trial?” Id. We responded: “The answer to this issue depends on whether the physician changed his theory of diagnosis or opinion as a consequence of the ex parte conversations.” Id. The opinion concluded from the law then extant that ex parte *230discussions between a plaintiffs treating physician and counsel for defendant were not authorized “without the grant of a motion to compel a medical authorization allowing for such discussions.” Id. at 268. The opinion noted also that the law as thereafter developed “shows a judicial philosophy that discourages these sorts of discussions with plaintiffs doctor.” Id.
The issue of ex 'parte contacts was remanded for evidentiary hearing with directions, [id] at 270:
If the trial court determines prejudice to the plaintiff, it shall grant a new trial. The burden to prove no prejudice at this hearing shall be on the defendant. If it grants a new trial, then the trial court must determine whether Dr. Campbell’s testimony may be presented by the defendant. If the trial court finds the ex parte discussions involved counsel presenting facts to Campbell that would have eventually been posed to him under oath, it then shall determine an absence of prejudice and the judgment on the verdict will be allowed to stand.
Conformably with our mandate, the trial court conducted an evidentiary hearing to determine whether prejudice resulted to the plaintiff from the ex parte discussions between defendant’s counsel and plaintiff’s treating physician, Dr. Campbell. The trial court determined that the defendant carried his burden to show that there was no prejudice to the plaintiff from the ex parte conversations of Dr. Campbell with defense counsel. The court found specifically that there was “no evidence that defendant’s counsel presented any facts to Dr. Campbell that would not have been presented to Dr. Campbell under oath.” The trial court decided, accordingly, that the plaintiff was not entitled to a new trial. The plaintiff appeals the denial of a new trial, an order our opinion in McClelland specifically subjects to appeal by the aggrieved party. Id. at 270[12]. McClelland seeks reversal of the order on three grounds. (1) The trial court abused its discretion by allowing defense counsel to put leading questions to Dr. Campbell on direct examination. (2) The trial court improperly overruled her motion for new trial because the findings of fact are not supported by substantial evidence and are against the weight of the evidence, and the conclusions of law erroneously apply the law to the facts. (3) The “change in testimony” remedy devised by McClelland to determine whether the unauthorized ex parte contacts in violation of the patient-physician privilege was prejudicial. to the plaintiff should be reconsidered and replaced by a rule of per se exclusion.
McClelland imposed upon the defendant the burden to prove that the ex parte conversations between Dr. Campbell and defense counsel did not result in prejudice to the plaintiff. The initiative to adduce evidence on the issue, therefore, was that of the defendant. To that end, the defendant presented for testimony the two principals in the ex parte encounter, Dr. Campbell and defendant’s attorney, Kenner. The plaintiff complains that the court allowed the defendant to pose leading questions to witness Campbell in a manner calculated to dispose answers that the opinions given by Campbell to defense counsel at the ex parte interviews were altogether consistent with the earlier diagnosis rendered by Campbell when he repaired McClelland’s fistula. In a word, the plaintiff complains that defendant’s counsel by tendentious questioning led his own witness towards the proof of the very issue remanded for adjudication: whether in any event “the ex parte discussions involved counsel presenting facts to Campbell that would have eventually been posed to him under oath.” McClelland, 805 S.W.2d at 270[12],
The points on appeal that the testimony elicited by leading questions was prejudicial to the plaintiff and that, in any event, the remedy devised to determine on remand any prejudice from the unauthorized ex parte contacts should be replaced by a rule of per se exclusion are unavailing and need not be labored. Whatever efficacy the appeal promises rests on the validity of the contention that the facts found by the trial court have no substantive basis in the evidence, and so induced distorted conclusions of law and erroneous adjudication.
*231The point on appeal that the “change in testimony” remedy devised by McClelland to determine on remand any prejudice to the plaintiff from the ex parte encounters be reconsidered by this court and replaced by a rule of per se exclusion is simply precluded by the doctrine of the law of the case. The doctrine governs successive appeals involving the same issues and facts. Under the doctrine, the appellate decision becomes the law of the case in subsequent proceedings in the same cause. Davis v. J.C. Nichols Co., 761 S.W.2d 735, 738 (Mo.App.1988). Its operation precludes re-examination of issues decided in the original appeal. Laclede Inv. Corp. v. Kaiser, 596 S.W.2d 36, 40 (Mo.App.1980). The argument for the per se exclusion of Dr. Campbell’s testimony, as a witness rendered incompetent by the breach of the physician-patient relationship, was presented by plaintiff McClelland on the first appeal, argued, and expressly rejected by this court. See McClelland, 805 S.W.2d at 270. We elected, rather, “to create another remedy for this unique problem.” Id. The argument for a rule of per se exclusion, moreover, was presented to this court in McClelland on the motion for rehearing, and then to the Supreme Court on the motion to transfer. The motions were rejected in both courts, and with them, the argument for per se exclusion.
The contention of prejudice from the leading questions to Dr. Campbell is also separately unavailing, and for a separate reason. A leading question is an interrogation on direct examination that suggests to the witness the specific answer desired. It is a question “which instructs the witness how to answer on material points, or puts into [the mouth of the witness] words to be echoed back.” 3 WIG-MORE, EVIDENCE § 769 (Chadbourn rev. 1970). The illegitimacy of the practice is not the question alone, but also the disposition of the witness in the circumstances to unreflectively assent to the answer the question proposes. Id. It is because of the “broad and flexible character” of that controlling principle, “that its application must rest largely, if not entirely, in the hands of the trial court.” [original emphasis] 3 WIGMORE, EVIDENCE § 770. It is the discretion of the trial court, therefore, that controls the use of leading questions, and error will not be found unless that discretion is abused. Harrison v. St. Louis-San Francisco Ry., 339 Mo. 821, 99 S.W.2d 841, 845-46[5-7] (1936).
The plaintiff complains of three incidents of undue questioning of Dr. Campbell by counsel for defendant. The first elicited from the witness that what he described as “possible anal cryptitis” in his original, post-operative diagnosis of McClel-land was “anal cryptitis.” Whatever the vice in the form of the question, the answer only repeated in summary what the witness had already given as responses to discrete, proper questions. That antecedent testimony was a description by the witness of the anal cryptitis process in terms of the post-operative diagnosis given McClelland. In that testimony Dr. Campbell explained his post-operative diagnosis of “possible anal cryptitis” in terms of the etiology of a patient who develops anal cryptitis which eventually leads to rectovaginal fistula. He testified that crypts are pockets in the anal canal that lead to glands that become inflamed when cut into, as during an episio-tomy. Cryptitis is merely inflammation of the crypts in the anal canal. Cryptitis may be localized or generalized, the witness explained. Thus, there may be only inflammation, but where there is a cut so that the crypt is exposed to bacteria, infection and abscess may follow. “Anal cryptitis,” therefore, is the inflammation of a crypt that, when infected, is followed by abscess, rupture and then fistula. The witness confirmed that the postoperative diagnosis “possible anal cryptitis” he gave to McClel-land was in terms of that process of etiology-
The “leading question” to which McClelland objects elicited no more than that. The response was in any event cumulative of testimony properly elicited. It is the presumption also that a judge sitting without a jury will not give weight to incompetent testimony, and so is accorded even greater latitude in the admission of *232evidence. Pike v. Pike, 609 S.W.2d 397, 403[10,11] (Mo. banc 1980).
The second incident of questioning relates to the contents of a letter Dr. Campbell had written to McClelland’s counsel for the purpose of litigation. Counsel had requested Campbell’s opinion concerning McClelland’s fistula from the episioto-my. The last paragraph of the response commented: “I think that to have a complication like this would not be unusual, but to do nothing about it is strictly poor medical practice.” The questioning elicited from the witness, Dr. Campbell, that the opinion he expressed in that paragraph of the letter was conditioned upon whether or not Dr. Ozenberger knew of McClelland’s symptoms of fistula and still “did nothing about it.”
The vice of the leading questions, McClelland argues, is their tendency to predispose the witness to answers that establish consistency between his deposition testimony after the ex parte encounters and his diagnosis and office entries before. The objection of leading question, however, is so within the discretion of the trial court to control that prejudice and hence, abuse of discretion is not readily found. That is particularly so in circumstances, such as these, where the witness is informed on the subject matter and so not readily led, where the responses are manifestly given in the witness’s own terms, and where there is no jury to influence. Sheets v. Kurth, 426 S.W.2d 103, 105[2] (Mo.1968); Pike v. Pike, 609 S.W.2d at 403[10,11].
The objection of leading question is all the more unavailing where, as here, the court made independent inquiry of the witness to clear a doubt as to why Campbell expressed the opinion in the letter unless Ozenburger “didn’t do something or follow up on what he was supposed to do.” The witness made explanation, the court inquired further, and the testimony stood. The objection is more unavailing still where, as here, the findings of fact as entered by the court implicitly find the testimony of the witness to be true.2 It is íale that reason that whatever efficacy the appeal promises rests on the validity of whether the facts as found by the court to sustain the adjudication rest on probative evidence, and not on leading questions or a reordering of remedy. The final incident of questions relates to an ostensible contradiction between the opinion of Dr. Campbell before the ex parte encounters, that the episiotomy cut into or “skimmed” the rectal wall which, he acknowledged, was not the same thing as a cut into the anal crypt and his opinion thereafter, that the fistula evolved from a cut into the anal crypt. The objection was that the witness was led into that response by the leading questions of defendant’s counsel. The objection is also that absent the responses suggested by these questions, the defendant could not acquit his burden to prove that the ex parte encounters did not cause Dr. Campbell to change his theory of diagnosis or opinion.
Q. My question is when you used the phrase “skimmed the rectal wall,” did it mean the same thing as a cut or entry into the crypt?
A. No.
Q. How was it different, Doctor?
A. If you would skim over some tissue that is formed if a little bud gets cut, that’s the difference between cutting into it primarily.
Q. But it wasn’t entry into the crypt, correct?
A. That’s my assumption.
Q. And you always believed there had been an entry into the crypt.
Mr. Nichols: Object to that as leading and suggestive.
The Court: Overruled.
A. [By the witness] That’s the only way I can try to make heads or tails out of this situation.
Q. That’s the only way you could—
A. Visualize it happening.
Q. Visualize this fistula developing, is that correct?
A: Yes.
*233Q. Doctor, when you used “skimming the crypt,” you did not change your theory of your diagnosis of Kim McClelland, did you?
A. No.
Mr. Nichols: Object as leading and suggestive.
The Court: Overruled.
Q. [By Ms. Warden] You did not change any opinion you had about Mrs. McClel-land’s care and treatment, did you?
A. No.
Mr. Nichols: Same objection.
The Court: Overruled. Your answer was? No.
It is evident from the findings of fact that the trial court accepted the answers by Dr. Campbell to the interrogations on direct examination by the defendant, on whose behalf he was called, as competent and probative evidence. Those findings3 acknowledge the evidence that neither Dr. Campbell’s office records nor the North Kansas City Hospital records at the time of the surgery to McClelland contained the phrase “a cut into the rectal wall,” but that Dr. Campbell used the phrase thereafter, once, in his opinion letter to McClelland’s counsel, Stoup. They find from the account given by Dr. Campbell, that the only way the fistula could be explained was by an entry into the anal crypt during the episiotomy by a cut to the glands, which are extensions of the crypts that proceed subcutaneously into the anal canal. Infection entered the anal canal in the cut. Inflammation followed and caused an abscess which ruptured leaving the fistula in its place. It was also the finding that it was Dr. Campbell’s opinion that the episiotomy caused an entry into an anal crypt and that was the most likely cause of the anal cryp-titis.
It follows from the facts found by the trial court, therefore, that the responses by Dr. Campbell to direct examination inquiry, that when he used the phrase “skimmed the rectal wall” he did not mean the same thing as a cut or entry into the crypt and that he always believed there had been an entry into the crypt, were not complaisant echoes of the questions. Nor were they unreflective assent to suggestion. It is implicit from the findings that the court determined they were the independent opinions of the physician witness. There was no abuse of discretion by the court in the allowance of the questions and answers. The prerogative of the factfinder to believe and draw inferences from the testimony of the witness on a material issue granted, the antecedent and more fundamental question remains: whether the facts found rest on evidence that is probative of them, and hence sufficient to sustain the order of adjudication. That review is under Murphy v. Carron, 536 S.W.2d 30 (Mo. banc 1976). McClelland acknowledges that under that standard of review credibility is for the trial court as factfinder and we treat the evidence in a manner most favorable to the adjudication. She recalls to us nevertheless that judgment must rest on probative evidence, so that where the cross-examination of a witness shows that the testimony previously given is only guess, speculation or contradiction, then neither probative fact nor substantial evidence sustains the adjudication and the order may not stand. See, e.g., Stearns v. Be-Mac Trans. Co., 621 S.W.2d 539, 541[1] (Mo. App.1981). McClelland argues that the testimony of Dr. Campbell considered integrally is not probative of the findings of fact entered by the trial court. The conclusions of law, accordingly, are erroneously induced and end in an order that erroneously applies the law.
In this task, it is well to recall the end of the evidentiary hearing under review. As the defendant Ozenberger aptly comments:
[McClelland ] instructed the trial court that there were two threshold issues to be determined at the evidentiary hearing: (1) whether Dr. Campbell changed his theory of diagnosis or his opinions and, if so, (2) whether that change occurred as a consequence of the ex parte conversa*234tions ... In the event these threshold issues revealed a wrong done to the plaintiff, [.McClelland ] instructed the trial court to consider whether or not Dr. Campbell’s testimony should be deemed incompetent as a remedy to right the wrong.
The order on remand denying McClelland a new trial rests on thirty-three findings of fact and four conclusions of law. [See Appendix A] She takes no issue with findings 1, 2, 3, 4, 5, 7, 15, part of 16, 20, 21, 22, 23, 26, 31 and 33. They are perfunctory, she observes and, in any event, do not bear materially to acquit the burden of proof. The plaintiff does contest findings of fact 6, 8, 9, 10, 11, 12, 13, 14, part of 16, 17, 18, 19, 24, 25, 27, 28, 29, 30 and 32. The plaintiff also contests all four conclusions of law.
The contention as to finding 6, as McClel-land concedes, is “somewhat of a minor issue,” and does need our review. Findings 17, 25 and 30 are of the same ilk, and need no separate review. The other findings are for the most part inextricable and are treated integrally.
Finding 8 reads:
When Dr. Campbell wrote ‘possible anal cryptitis’ in his office records dated August 21, 1987 and in the North Kansas City records dated August 21, 1987, he meant ‘anal cryptitis’ as defined as the process of inflammation of a crypt, followed by infection, followed by abscess, rupture and then fistula.
It is the argument that the postoperative diagnosis, “possible anal cryptitis,” that Dr. Campbell gave to McClelland was by his own acknowledgement “speculation.” Cryptitis, Dr. Campbell had testified, was the process of inflammation of an anal crypt, followed by infection, abscess, then rupture of the abscess, after which a fistula remains. In response to direct examination inquiry: “Was it part of your diagnosis back in August of 1987 that Kimberly McClelland had had possible anal cryptitis with the process occurring just as you just described?” the witness answered: “That is my speculation.” The defendant complains that the plaintiff misuses that response, that taken in the full context of his testimony, Dr. Campbell meant only that “the most likely cause of the fistula was anal cryptitis.” He meant the term, “speculative,” rather, as “to all other possibilities posed to him by plaintiffs, including “cut into the rectal wall.”
It is so, as the defendant points out, that Dr. Campbell offered no opinion in terms of “reasonable medical certainty.” He did give postoperative opinions that the cause of the rectovaginal fistula he removed by surgery was “possible anal cryptitis.” That opinion was given in terms of “possible” rather than “reasonable medical certainty,” he explained, because that was the “most likely” explanation of the etiology of the fistula. Whether the medical witness casts response in terms of “most likely” explanation rather than “reasonable medical certainty,” however, does not dispense with the necessity that the opinion be probative of evidential fact. It may be that the finder of fact could understand from the full testimony on direct examination that the entry into the crypt was the only way Dr. Campbell could “try to make heads or tails out of the situation.” It was the only way he could “visualize” the fistula developing. If left to those responses alone, the finder of fact could have reasonably understood that what the witness meant by “[t]hat is my speculation” was the process of entry, inflammation, abscess, rupture and fistula that he visualized as the most medically valid explanation of etiology for the fistula.
That surmise of probativeness, however, is undermined by unqualified responses by Dr. Campbell to cross-examination that the opinion of etiology of the fistula by cut into the crypt was no more certain than by cut into the rectal wall. The subject of inquiry was the letter Dr. Campbell wrote on May 23, 1988, in answer to the request by Arthur Stoup, counsel for plaintiff McClel-land, for a report with his opinion that Dr. Ozenberger did not provide McClelland that degree of care that should have been provided by a practitioner of medicine in Missouri. The letter from Stoup made clear that the requested opinion was necessary *235in order to comply with the Missouri statutes regarding the prosecution of a medical negligence suit. See § 538.225, RSMo 1986. The response by Dr. Campbell included the statements:
[T]he episiotomy evidently cut into the rectal wall and was not repaired. I think that to have a complication like this would not be unusual, but to do nothing about it is strictly poor medical practice.
Counsel for plaintiff confronted Dr. Campbell with his report of May 23, 1988, to Stoup and his opinion that “the episioto-my evidently cut into the rectal wall and was not repaired.” Dr. Campbell acknowledged the contents of the report and acknowledged as well that “a cut into the rectal wall is something totally different from skimming a crypt.” Dr. Campbell agreed also “one would arrive at [the] conclusion,” as a fair assessment of those letters, “that a cut was made into the rectal wall that resulted in a fistula.” He agreed as well that the statements in the opinion were a “true and accurate assessment of the case.”
The cross-examination continued:
Q. But there is nothing in your letter that talks about an abscess, is there? A. No.
Q. There is nothing in the records of North Kansas City Hospital in August of 1987 that talks about an abscess, is there?
A. No.
Q. Nothing in the letter or in the record prior to January 6 of ’89 when you talked to the defense lawyers. Nothing in any of those records that talks about skimming of the the crypt, is there?
A. I think there’s mention of skimming of a crypt someplace.
Q. I said prior to January 6 of 1989? A. No, not that I recall.
Q. The first time that we see any mention of a crypt being skimmed with the formation of an abscess is in your office records of January 6 of 1989, true?
A. What’s the date on that?
Q. January 6, 1989.
A. Yes.
Q. That was when you were having a discussion with David Kenner, one of the defense attorneys for Dr. Ozenberger, correct?
A. Yes.
Q. Doctor, isn’t it true that at no time did you find any evidence when you were treating Kim McClelland that there had ever been an abscess?
Ms. Warden; Objection—
The Court: Overruled.
A. [By the witness] Only the fact that a fistula usually follows some sort of infection which can be termed abscess or just an infection.
Q. But you would have had no way of knowing whether that infection related to a crypt or a cut into the rectal wall that wasn’t repaired on the rectal side? A. No, I wouldn’t.
Q. So it’s really total speculation that the fistula occurred from some type of skimming of the crypt?
A. Yes.
Q. In fact, you don’t have any evidence and aren’t aware of any evidence to suggest that Dr. Ozenberger didn’t actually cut into the rectal wall, isn’t that true?
Ms. Warden: Again, Your Honor, I’d like to voice an objection to using the term evidence. I think it has legal connotations and there is no foundation with this witness and it’s an improper question. The Court: Overruled. You may answer.
A. [By the witness] I have no way of knowing the exact—
[emphasis added]
There is nothing to relieve these responses from the stark and literal effect that the opinion that “the fistula occurred from some type of skimming of the crypt” was “really total speculation.” The fact that fistula follows an infection “which can be termed abscess,” these responses admit, does not suffice to tell “whether that infection related to a crypt or cut into the rectal wall that wasn’t repaired.” Where a party relies on the testimony of a single witness *236to prove a given issue, and the testimony of the witness is contradictory and conflicting, one version tending to prove and the other tending to disprove the issue, and no other circumstance in the case tends to show which version is true, the finder of fact may not speculate which statement of the witness should be accepted. The testimony is not probative, and the issue is simply not proven prima facie. Adelsberger v. Sheehy, 332 Mo. 954, 59 S.W.2d 644[6,7] (1933); Griggs v. A.B. Chance Co., 503 S.W.2d 697, 703[l-5] (Mo.App.1973). It is not only the operation of a rule of evidence that marks the prior testimony and opinion as speculative, but the frank concession of the witness, Dr. Campbell, himself. See Stearns v. Be-Mac Trans. Co., 621 S.W.2d at 541.
There is no substantial evidence to sustain finding of fact 8, and hence that material element of the adjudication, that the fistula was the result of an anal cryptitis and that Dr. Campbell believed that was its etiology when he wrote “possible anal cryp-titis” in his notes and in the hospital records on August 21, 1987.
Nor are there circumstances in the case that meliorate, from speculation to an opinion of evidential value, the explanation of etiology of the fistula. The explanations that Dr. Campbell made at the hearing on remand that the opinion in the letter to counsel. Stoup that the fistula developed from a cut into the rectal wall was “a poor choice of words” and his deposition testimony, that “I probably should also have said, ‘or skimmed over the rectal wall and involved her crypt or gland’ ” do not settle, but simply reassert, the ambivalent opinion of etiology as either one or the other.
The witness Campbell acknowledged a change of opinion after his two meetings with defense counsel from cut into the rectal wall to skimming of the crypts and abscess as the cause of the fistula. The findings of the court, made out from his testimony, nevertheless, that it was the doctor’s opinion that “the most likely cause of plaintiff’s anal cryptitis, was a cut or entry made into the gland during the epi-siotomy,” the doctor “always believed that there had been an entry into the crypt,” and that he never changed his diagnosis4. Those findings are not only undermined by the acknowledged change in opinion, but effectively nullified by the admission that the theory of fistula caused by a skimming of the crypt was “really total speculation.”
Those findings are undermined, as well, by Dr. Campbell’s testimony that at the ex parte occasion with defendant’s counsel Kenner, “[evidently there was some discussion in trying to figure out what might have caused this fistula to develop,” and that after the ex parte conversations, he changed his opinion from cut into the rectal wall to entry by skimming of the crypts. That Dr. Campbell had an actual change of mind after the ex parte interviews is made quite plain by his response that, although he knew plaintiff’s counsel Stoup was relying on his statements and opinions in the letter to counsel of May 23, 1988, the reason that he did not tell Stoup about the change of opinion was because Stoup did not ask him.
We determine as a matter of law that Dr. Campbell changed his theory of diagnosis and opinion from fistula caused by infection from an unattended cut into the rectal wall to fistula caused by infection from an unattended cut into the rectal wall or fistula by infection from an unattended cut into an anal crypt one cause as uncertain as the other. That is to say, the theory of diagnosis was changed from an opinion of medical cause that is probative, to two opinions of medical cause neither of them more probative than the other, and therefore both speculative. The plaintiff has been deprived of a critical opinion of medical malpractice from a treating physician by the wrongful ex parte conversations of counsel for the defendant. The defendant has not met his burden to prove that there was no prejudice to the plaintiff from the ex parte conversations between *237Dr. Campbell and defendant’s counsel. The conclusions of law are overruled. The preconditions delineated in McClelland for the grant to plaintiff of a new trial are met. Id. at 267[1], 270[12].
There remains the fashioning of the remedy. We have determined that Dr. Campbell changed his theory of diagnosis and opinion as to the cause of the fistula as a consequence of the ex parte discussions. We have determined also that the defendant did not acquit his burden to prove that no prejudice resulted to the plaintiff from this change of diagnosis and opinion. The question is whether Dr. Campbell remains a competent witness as to any material facet of the litigation. That depends upon how pervasively the change of diagnosis and opinion affects the testimony of Dr. Campbell that bears on those aspects of the trial proof. It bears comment again that the legal insufficiency of finding of fact 8 [that by the postoperative entry “possible anal cryptitis” Dr. Campbell meant “anal cryptitis” to explain the etiology of the fistula he repaired surgically] as a probative basis for the trial court adjudication so compromises other critical findings as to undermine the structure of the adjudication.
The legal insufficiency of finding of fact 8, for instance, so adversely implicates finding of facts 11, 12, 13 and 24 as to compromise finding of fact 29, their material concomitant. The summary of facts 11, 12 and 13, that it was Dr. Campbell’s opinion that the most likely cause of the anal cryptitis was a cut into the gland made during the episiotomy, that he always believed that there had been an entry into the crypt, and that Dr. Campbell’s later use of the phrase “skimming the crypt” to describe the etiology of the fistula “in no way” changed his diagnosis, simply do not stand as probative once the legal sufficiency of finding of fact 8 is rejected. The lack of probative evidence as to etiology of the fistula, either as a residue of entry into the crypt or of cut into the rectal wall, undermines as well finding of fact 29. That finding has it that “Dr. Campbell’s opinion regarding the standard of care in the last paragraph of his letter to plaintiff’s counsel dated May 23, 1988, was conditioned on whether or not the plaintiff told Dr. Ozen-berger of her symptoms.”
It is against the weight of the evidence, however, to conclude as does finding 29 that Campbell’s opinion regarding the standard of care in his letter to plaintiff’s counsel was conditioned on whether or not plaintiff informed Dr. Ozenberger of her symptoms. It may have validity as to the skimming of the crypt etiology, where the cut is into microscopic glands that extend from the crypts and so are not diagnosable without palpable or reported symptoms. [See findings of fact 10 and 11]5 It has much less validity where the cut is into the rectal wall, and so should be evident to the ministering practitioner who nevertheless does nothing about it. In the latter case, the medical care provider should have known about it, and acted.6 In the first, it is neglect to act after the medical care provider knows.
The letter by Dr. Campbell to attorney Stoup written prior to the ex parte conversations with the opinion that “to have a complication like this would not be unusual, but to do nothing about it is strictly poor medical practice,” refers immediately to the prior sentence, “the episiotomy evidently cut into the rectal wall and was not repaired.” Whatever the post hoc explanation of a harbored meaning, the opinion that the neglect to do something about a *238cut in the rectal wall is “strictly poor medical practice” is plain and unconditional expression. The change in testimony by Dr. Campbell from this medically probative opinion of etiology of the fistula, to a theory of “skimming of the crypts,” and then to an ambivalence about both, deprived the plaintiff of critical evidence by a treating physician.
The changes in the testimony by Dr. Campbell as a consequence of the ex parte conversations with defendant’s counsel so pervade the elements of proof that confront the plaintiff as to virtually nullify the competency of his testimony at a retrial under McClelland.
The cause is reversed and remanded for a new trial. At the retrial, the defendant may not use or present any testimony or medical records offered or authored by Dr. Campbell, except only that the parties may present as evidence the entries in his office records and in the records of the hospital concerning the diagnoses made in August of 1987 and the treatment given the plaintiff between August and September of 1987. The defendant shall not take further testimony from Dr. Campbell concerning the plaintiff in this cause.
All concur.
APPENDIX A
FINDINGS OF FACT AND CONCLUSIONS OF LAW
This case came on for an evidentiary hearing on the 23rd day of July, 1991. Plaintiff appeared by her counsel, Arthur H. Stoup and Stephen W. Nichols. Defendant appeared by his counsel, Michaela M. Warden and Linda S. Dickens. The Court heard evidence in the form of testimony from witnesses John Campbell, M.D. and David Kenner. The Court heard oral arguments from counsel for the parties. The Court considered all of the briefs, citations and authorities filed by the parties. After giving the matter due consideration, the Court makes the following Findings and Conclusions.
FINDINGS OF FACT
1. John G. Campbell, M.D., a colorectal surgeon, repaired plaintiff Kimberly McClelland’s rectovaginal fistula on August 21, 1987.
2. On the day Dr. Campbell repaired plaintiff’s rectovaginal fistula, Dr. Campbell placed the following diagnosis in plaintiff’s medical chart at North Kansas City Memorial Hospital: “Rectovaginal fistula plus possible anal cryptitis and incomplete development of the anal canal with mucosal prolapse.”
3. Dr. Campbell placed the following postoperative diagnosis in his office medical chart on August 21, 1987: Rectovaginal fistula “plus possible anal cryptitis and incomplete development of the anal conal [sic] with mucosal prolapse.”
4. Neither Dr. Campbell’s office records nor the North Kansas City Hospital records at the time of the plaintiff’s surgery contained the phase “a cut into the rectal wall.”
5. Cryptitis is the inflammation of the crypts in and around the anal canal.
6. If a crypt is involved in an episiotomy cut external to the muscle, the crypt can develop an abscess.
7. No muscles were involved in the extension of plaintiff’s episiotomy.
8. When Dr. Campbell wrote “possible anal cryptitis” in his office records dated August 21, 1987 and in the North Kansas City records dated August 21, 1987, he meant “anal cryptitis” as defined as the process of inflammation of a crypt, followed by infection, followed by abscess, rupture and then fistula.
9. Dr. Campbell found evidence of an abscess when he was treating Kimberly McClelland because he found a fistula and a fistula usually follows infection which can be termed “abscess” or termed “infection”.
10. The only way Dr. Campbell could determine that the plaintiff developed a fistula following her delivery was that there must have been glands going to the muscles into the subcutaneous and perirec-*239tal tissue. Glands are extensions of the crypts. The glandular tissue extends from the crypts subcutaneously into the anal canal.
11. In Dr. Campbell’s opinion, the most likely cause of plaintiffs anal cryptitis was a cut or entry into a gland made during the episiotomy performed by the defendant, which subsequently allowed infection to enter the anal crypt.
12. Dr. Campbell always believed that there had been an entry into the crypt, that is the only way he could visualize this fistula developing.
13. In August, 1987, when Dr. Campbell wrote his medical records it was his opinion that plaintiff had possible anal cryptitis. When he later used the phrase “skimming the crypt” in his January 6, 1989 office note he in no way changed that diagnosis.
14. Dr. Campbell used the word “abscess” in his January 6, 1989 office note to explain the process of how anal cryptitis can lead to a rectovaginal fistula. There is no evidence to support plaintiffs claim that Dr. Campbell’s use of the word “abscess” described a different etiology for the fistula than Dr. Campbell had described in his August, 1987 medical records.
15. Defendant’s counsel David Kenner had an ex parte meeting with Dr. Campbell on January 6, 1989.
16. Exhibit 41 contains David Kenner’s notes regarding his meeting with Dr. Campbell. The notes on the last page of Exhibit 41 were Mr. Kenner’s notes taken during his meeting with Dr. Campbell on January 6, 1989. These notes were Mr. Kenner’s notes of Dr. Campbell’s description of how the episiotomy lead to the fistula.
17. Dr. Campbell’s opinion at the time of plaintiff’s surgery was that the episioto-my caused an entry into an anal crypt which became inflamed. The inflammation caused an abscess which ruptured leaving the fistula in its place.
18. Defendant’s counsel David Kenner did not suggest the use of the word “abscess” to Dr. Campbell.
19. On April 13, 1988 Dr. Campbell had an ex parte conversation with plaintiffs counsel Arthur Stoup. It is unclear whether this was in person or over the telephone.
20. Defendant’s counsel Michaela Warden had an ex parte meeting with Dr. Campbell on May 26, 1989.
21. On May 9, 1988 plaintiff’s counsel Arthur Stoup sent a letter to Dr. Campbell requesting that Dr. Campbell provide him with an opinion letter.
22. On May 23, 1988, Dr. Campbell wrote a letter, marked as Trial Exhibit 7, in response to Mr. Stoup’s letter of May 9, 1988.
23. The only mention by Dr. Campbell of a “cut into the rectal wall” in any document was in the letter he wrote to plaintiff’s counsel Arthur Stoup on May 23, 1988.
24. The court finds that Dr. Campbell’s use of the phrase “cut into the rectal wall” in his letter to Mr. Stoup was consistent with Dr. Campbell’s later testimony at his deposition and his testimony at the eviden-tiary hearing because Dr. Campbell testified that the most likely cause of plaintiffs anal cryptitis was a cut or entry into a gland made during the episiotomy performed by the defendant, which subsequently allowed infection to enter the anal crypt.
25. No representatives of defendants have had any other private conversations with Dr. Campbell since May 26, 1989.
26. At her deposition on September 8, 1989, plaintiff Kimberly McClelland testified that she never told Dr. Ozenberger about the fistula or the symptoms caused by the fistula.
27. It was at Dr. Campbell’s deposition on October 6, 1989 that defendant’s lawyer told Dr. Campbell that Kimberly McClel-land had testified she did not tell Dr. Ozen-berger about the symptoms of her fistula.
28. After Dr. Campbell learned that plaintiff did not tell Dr. Ozenberger about *240the symptoms of her fistula, Dr. Campbell was no longer critical of Dr. Ozenberger’s care and treatment of plaintiff.
29. The court finds that Dr. Campbell’s opinion regarding the standard of care in the last paragraph of his letter to plaintiff's counsel dated May 23, 1988 was conditioned on whether or not the plaintiff told Dr. Ozenberger of her symptoms. Since defendant’s counsel did not learn that plaintiff had not told Dr. Ozenberger of her symptoms until plaintiff gave her deposition on September 6, 1989, it was impossible for defendant’s lawyers to have presented that fact ex parte to Dr. Campbell, as the last ex parte visit took place on May 26, 1989.
30. The medical authorization furnished by the plaintiff which purported to deny ex parte visits was in Dr. Campbell’s file at the time he met with lawyers for defendant.
31. There was no conversation about the medical authorization between Dr. Campbell and either of the defendant’s lawyers.
32. Dr. Campbell made the note “Medical Protective” only because it was significant to him that Dr. Ozenberger’s lawyer was not a sole practitioner.
33. Neither David Kenner or Michaela Warden had any discussions with Dr. Campbell about medical malpractice eases generally or about insurance rates on doctors.
CONCLUSIONS OF LAW
1. The court finds that there was no prejudice to plaintiff caused by the ex parte conversations with defendant’s counsel.
2. In view of the court’s resolution of the threshold issues as defined by the Court of Appeals and the court’s determination that no prejudice resulted to plaintiff, the court finds that plaintiff is not entitled to a new trial.
3. The court finds no evidence that defendant’s counsel presented any facts to Dr. Campbell that would not have been presented to Dr. Campbell under oath.
4.The court finds that defendant has met his burden of proof.
Dated: AUG. 23 1991 /§/
THE HONORABLE R. KENNETH ELLIOTT
. The facts of the litigation are fully recited in that opinion. This narrative recapitulates only those aspects of the full evidence that bear directly on the issues under review in this second appeal.
. See Appendix A, Findings of Fact and Conclusions of Law, Findings of fact 27, 28 & 29.
. See Appendix A, Findings of Fact and Conclusions of Law, Findings of fact 4, 9, 10, 11, 12, 13, 14, 17, 18, 23 and 24.
. The want of evidence and opinion probative of an etiology of the fistula not only undermine and nullify finding of fact 8, but also implicate and compromise findings of fact 11, 12, 13, 24, 28 and 29 at the least.
. In the manner told by Dr. Campbell to defense attorney Kenner during their ex parte discussion, a skimming of the crypts and glands during an episiotomy which then resulted in abscess and fistula "was not a malpractice thing,” unless known to the physician who "does nothing about it.”
. That was the theory submitted by the plaintiffs verdict director as an alternative ground of recovery: that defendant Ozenberger cut into the rectum while performing an episiotomy on plaintiff and failed to recognize and repair the cut. That submission was based' upon the opinion of plaintiffs expert, Dr. Janes, an authority in the treatment of rectovaginal fistulas, that at the time the episiotomy was performed a cut was made into the rectum that was sewed up on the vaginal side but not on the rectal side, which then was broken through, causing the fistula.