Linda F. WOLLEN, Appellant,
DePAUL HEALTH CENTER, Richard F. Jotte, Sr., M.D., Edwin Ernst, III, M.D., and Ernst Radiology Clinic, Inc., Respondents.
C. Robert Buckley, Independence, for amicus Mata.
Don R. Sherman, Sanford Goffstein, Lori Koch, Joseph Kortenhof, Robert A. Wulff, Mark R. Dunn, St. Louis, for respondents.
Jeffrey P. Ray, Charles W. McKee, Kansas City, for amicus Missouri Org. Defense Lawyers.
Linda F. Wollen appeals the dismissal, for failure to state a cause of action, of her petition for damages filed under the wrongful death statute, § 537.080 RSMo 1986. The decision below is vacated; and this case is remanded for further proceedings in accordance with this opinion.
Assuming for the purposes of appellate review that the allegations in appellant’s pleadings are true, the facts of this case are: On January 26, 1988, David L. Wollen went to respondent Richard F. Jotte, Sr., M.D., for medical treatment. Dr. Jotte referred Mr. Wollen to the other respondents, DePaul Health Center, Ernst Radiology Clinic, Inc., and Edwin Ernst, III, M.D., for tests. If respondents had performed appropriate tests, or had correctly interpreted the tests that they had conducted, respondents would have diagnosed that Mr. Wollen 682*682 was suffering from gastric cancer on January 28, 1988. If Mr. Wollen had been correctly diagnosed, and given appropriate treatment, he would have “had a thirty percent (30%) chance of survival and cure.”
Mr. Wollen died from gastric cancer on July 2, 1989. On January 17, 1990, Linda F. Wollen, Mr. Wollen’s widow, filed a wrongful death action against respondents. Each respondent filed a motion to dismiss the action on the ground that the petition failed to plead a causal connection between respondents’ negligence and the death of Mr. Wollen. The circuit court sustained respondents’ motions, and appellant refused leave to file an amended petition on May 7, 1990. After appeal to the Court of Appeals, Eastern District, and on application by appellant, this Court ordered this case transferred here on September 10, 1991.
This case turns on the issues of proof and causation. At this initial stage of the case, courts are required to assume that the evidence introduced at the trial will prove the facts as pled by the plaintiff. See Ritterbusch v. Holt, 789 S.W.2d 491, 492 (Mo. banc 1990); Lowrey v. Horvath, 689 S.W.2d 625, 626 (Mo. banc 1985). The question remains, however, whether the ultimate legal fact of causation can be inferred from those facts.
As is the case with most petitions, in addition to pleading the alleged facts, the petition before this Court pleads, as facts, elements of the cause of action—negligence, causation, and damages. Such “facts”/elements are, in reality, merely legal conclusions. See Bergel v. Kassebaum, 577 S.W.2d 863, 867 (Mo.App.1978). In the simple negligence case, the inference from the real pleadings of fact to these legal conclusions is simple and straightforward. A statement that the plaintiff slipped on a patch of ice creates an inference that the ice caused the plaintiff to slip. The inference in this case is not that easy.
In the failure-to-diagnose case, the fact pleaded to show causation often has to be a statistic. This problem can be avoided if the plaintiff can plead that there is a reasonable medical or scientific certainty that defendant’s negligence caused the harm. Cf. Schiles v. Schaefer, 710 S.W.2d 254, 261 (Mo.App.1986); Bertram v. Wunning, 417 S.W.2d 120, 125 (Mo.App.1967).
The cases on reasonable medical certainty, however, reflect the lack of clarity that can occur when the legal profession tries to impose its terms on other professions. In this Court’s understanding of the medical issues involved, there are three possibilities: two of which involve “reasonable medical certainty,” and a third that does not.
The first possibility is the circumstance that could be phrased as “but for” causation. In this circumstance, the patient has a disease for which, at the stage the patient seeks diagnosis, there is a cure that works in the overwhelming majority of cases. Death from this type of disease caught at this stage is rare unless either the patient or the doctor is negligent at some stage of the treatment process.
The second possibility is exactly the opposite. In this circumstance, there is no known cure for the disease. Medical science can only, at best, extend the patient’s life a short time. In the overwhelming majority of cases, the disease is fatal except for a small number of spontaneous remissions or cures.
The third possibility is the one that appears in this case. Doctors have a treatment that works in a large number of cases and fails in a large number of cases. Because there is a real chance that the patient will survive and a real chance that the patient will die from the disease—even if it is diagnosed—it is impossible for a medical expert to state with “reasonable medical certainty” the effect of the failure to diagnose on a specific patient, other than the fact that the failure to diagnose eliminated whatever chance the patient would have had.
Appellant proposes analyzing this case under the theory that respondents’ 683*683negligence was a “substantial factor” that contributed to Mr. Wollen’s death. Prior cases do not support appellant’s interpretation of the substantial factor theory. InJackson v. Ray Kruse Const. Co., 708 S.W.2d 664, 667-69 (Mo. banc 1986), this Court examined the issue of multiple tort-feasors and considered situations where, while cause could not absolutely be proven, it is “more likely than not” that “but for” the negligence of the tort-feasors the injury would not have occurred. In Goff v. St. Luke’s Hospital of Kansas City, 753 S.W.2d 557, 563 (Mo. banc 1988), the plaintiff’s evidence showed that the negligence of the physicians and the hospital was “a major cause of death.”
In this case, appellant can only allege that, even when all of the tort-feasors are taken together, the negligence of the respondents might have contributed to the death of Mr. Wollen. In these circumstances, the petition fails to allege facts that indicate that respondents’ negligence was a substantial factor in causing the death of Mr. Wollen.
In argument before this Court, all parties have discussed whether this Court should recognize an alternative theory of recovery based on the Restatement (Second) of Torts, section 323(a). As Missouri is a fact-pleading state, this Court will consider whether the facts as pled entitle appellant to relief under a “lost chance of recovery” theory. See Rule 55.05.
In the past this Court has recognized that section 323 creates a duty of care. SeeHoover’s Dairy, Inc. v. Mid-America Dairymen, Inc., 700 S.W.2d 426, 432-33 (Mo. banc 1985). In creating this duty, this section of the Restatement defines a type of negligent behavior but does not alter the rules of causation. For the reasons discussed in Parts III and V, the petition fails to allege facts that show that respondents’ alleged negligence caused Mr. Wollen’s death. The petition does, however, allege facts that would support an action for lost chance of recovery.
684*684 The arguments against recognizing such a cause of action, however, are not totally without merit. Traditionally, causation has been treated as a “yes-or-no” or “all-or-nothing” question. There has been a trend towards apportioning fault among various parties in law suits but such apportionment occurs only after the jury has decided that the evidence shows causation. See, e.g., Kibbons v. Union Electric, 823 S.W.2d 485 (Mo. banc 1992); Gustafson v. Benda, 661 S.W.2d 11 (Mo. banc 1983). This traditional view of causation does aid the goal—which should be shared by the entire legal profession—of avoiding unnecessary barriers to ordinary people understanding their rights and responsibilities under the law.
On the other hand, there are compelling reasons for granting compensation in this type of cause of action. The traditional yes-no view of the world in causation theory does not match the “maybe” view of the world found in probability, statistics, and everyday life. To both the statistician and the patient seeking care from a doctor, there is no meaningful difference between a 50.001% and a 49.999% chance of recovery.
Medical science has given patients real chances to recover, sometimes only a small chance, but still a chance, in circumstances that used to be hopeless. When patients go to doctors with serious illnesses, they expect to have those chances that medical science has provided. To the individual patient, if the doctor’s negligence destroys the chance of recovery, it is irrelevant what that chance originally was. The courts of this state implicitly recognized this fact as far back as 1923. Cf. Smith v. Mallinckrodt Chemical Works, 212 Mo.App. 158, 251 S.W. 155, 158 (1923).
A patient with cancer, like Mr. Wollen, would pay to have a choice between three unmarked doors—behind two of which were death, with life the third option. A physician who deprived a patient of this opportunity, even though only a one-third chance, would have caused her real harm.
In light of this reality, the patient does suffer a harm when the doctor fails to diagnose or adequately treat a serious injury or disease. The harm suffered is not, however, the loss of life or limb. The harm is the loss of the chance of recovery. While, in the end, damages can only be expressed by multiplying the value of a lost life or limb by the chance of recovery lost, the proper place for such an inquiry is in the damages stage rather than in the liability/causation determination. See Herskovitz, 685*685 664 P.2d at 485-86 (Pearson, J., concurring); cf. William M. Landes & Richard A. Posner,The Economic Structure of Tort Law 263-69 (1987) (recommending proportional damages in increased risk of death cases involving toxic torts). Therefore, rather than adopting a theory of proportional causation, this Court chooses to recognize a cause of action for lost chance of recovery in medical malpractice cases.
The problem with the petition is not the facts of the case but rather the statute under which the case is brought. The most that appellants can prove is that respondents caused Mr. Wollen to lose his chance of survival. The wrongful death statute, § 537.080 RSMo (whether 1986 or Supp. 1991), requires that, before the next of kin of a person can sue for wrongful death, “the death of [that] person result[ ] from” the negligence of the tort-feasors. The question in this case is whether a court or a jury can infer from the loss of a thirty percent chance of survival that Mr. Wollen’s death resulted from respondents’ negligence.
The analysis of this question is simplified by imagining three possible petitions in this type of case. Case 1 (this case): An allegation of wrongful death in which the statistics show a thirty percent chance of survival if the disease had been properly diagnosed. Case 2: A wrongful death petition in which the statistics show a seventy percent chance of survival. Case 3: A petition filed under the survivorship provisions, § 537.020 RSMo 1986.
Assuming for the sake of argument that all three cases are based on a lost chance of recovery, the question is which causes of action are allowable under Missouri law. The language of the survivorship statute and the wrongful death statute are mutually antagonistic. The survivorship statute applies when the injury alleged did not cause death, and the wrongful death statute applies when the injury did cause death.
In past cases, this Court has focused on the issue of whether it was “more likely than not” that the negligence alleged caused or was a substantial factor in causing the injury. See Wors v. Glagow Village Supermarket, Inc., 460 S.W.2d 583, 586 (Mo.1970); Howard v. Kysor Industrial Corp., 729 S.W.2d 603, 606 (Mo.App. 1987). In reality, such an issue is merely a restatement that the burden of proof in tort cases is usually preponderance of the evidence. See Motley v. Colley, 769 S.W.2d 477, 479 (Mo.App.1989); Fujita v. Jeffries, 714 S.W.2d 202, 206 (Mo.App. 1986). The true issue is whether there is a permissible inference from the allegation of a statistical chance of recovery to the conclusion that death was caused by the negligence of the defendant(s).
In discussing this issue, there is a difference between logical inferences, and those inferences that involve a leap of faith that makes such inferences mere speculation.See James v. Sunshine Biscuits, Inc., 402 S.W.2d 364, 375 (Mo.1966); Anderson v. Bell, 303 S.W.2d 93, 96 (Mo.1957); Howard v. Kysor Industrial Corp., 729 S.W.2d at 606; Schiles v. Schaefer, 710 S.W.2d at 261. By its very nature, however, the statistical inference involved in a wrongful death action for failure to diagnose involves a speculative leap of faith.
In the posture of this case, the Court assumes that the plaintiff can prove by a preponderance of the evidence that there was a statistical chance of survival. Regardless of the exact percent of individuals in the circumstances of the decedent, as long as there is a significant chance of either survival or death, the statistic cannot tell whether the decedent would have survived if properly diagnosed. A statistic of this kind typically predicts that, out of a random sample of a large number of people in decedent’s circumstances—if properly diagnosed and treated—Y percent will live686*686 and Z percent will die. A specific individual, however, could be in either group. A jury could speculate as to which group a decedent would fall, but the statistical evidence—without more—does not give a jury a basis to believe that the decedent belongs to either the group that lives, or the group that dies. Therefore, regardless of whether the lost chance of survival is greater than or less than 50%, it is impossible to prove that decedent’s death resulted from the failure to properly diagnose and treat. Thus, neither of the wrongful death actions (cases 1 and 2) are allowed under the wrongful death statute; and only the action filed under the survivorship statute (case 3) is allowed as a matter of Missouri law.
In the first averment of appellant’s current petition, Mrs. Wollen states that she is suing as Mr. Wollen’s widow under the wrongful death statute. As the survivorship statute, § 537.020.1 RSMo 1986, requires that the suit be brought by the personal representative of the decedent’s estate, this averment does not state a cause of action. As appellant refused leave to amend the petition, the circuit court was correct to order this action dismissed.
The record, and the briefs before this Court, show that the failure to amend was apparently based on the assumption that this action could only be brought under the wrongful death statute. On that assumption, there were no possible amendments. This opinion, however, does create the possibility of amending that first averment to state that Mrs. Wollen is bringing the suit under the survivorship statute as the personal representative of decedent’s estate, if she has or can qualify as such. On this basis, this Court remands this case with directions to the circuit court to allow Mrs. Wollen, or any other qualified personal representative of Mr. Wollen’s estate, to be substituted as a party plaintiff and granted leave to amend the petition.
Therefore, the decision below is vacated; and this case is remanded for further proceedings in accordance with this opinion.
ROBERTSON, C.J., RENDLEN, COVINGTON, HOLSTEIN, and THOMAS, JJ., and BLACKMAR, Senior Judge, concur.
 See Voegeli v. Lewis, 568 F.2d 89 (8th Cir.1977) (cause of action under South Dakota law for reduction of chance of recovery); Daniels v. Hadley Memorial Hospital, 566 F.2d 749 (D.C.Cir.1977)(under D.C. law reduced chance of survival sufficient to put causation before jury); McBride v. United States, 462 F.2d 72 (9th Cir.1972) (cause of action under Hawaii law for deprivation of significant chance of recovery); Jeanes v. Milner, 428 F.2d 598 (8th Cir.1970) (wrongful death action allowed under Arkansas law even if chance of survival is less than 50 percent); Hicks v. United States, 368 F.2d 626 (4th Cir.1966) (wrongful death action allowed under Virginia law for lost chance of survival);Thompson v. Sun City Community Hospital, Inc., 141 Ariz. 597, 688 P.2d 605 (1984) (increased chance of harm puts issue of causation before jury); Richmond County Hospital Authority Operating University Hospital v. Dickerson, 182 Ga.App. 601, 356 S.E.2d 548 (1987) (cause of action allowed to go forward even if patient’s chance of survival was less than 50%); DeBurkarte v. Louvar, 393 N.W.2d 131 (Iowa 1986) (patient allowed to sue for reduced chance of recovery caused by late diagnosis);Hastings v. Baton Rouge General Hospital, 498 So.2d 713 (La.1986) (showing of increased risk of harm shows proximate cause); Roberson v. Counselman, 235 Kan. 1006, 686 P.2d 149 (1984)(wrongful death action allowed for lost chance of recovery); Thomas v. Corso, 265 Md. 84, 288 A.2d 379 (1972) (wrongful death action allowed for lost chance of survival); Falcon v. Memorial Hospital,436 Mich. 443, 462 N.W.2d 44 (1990) (medical malpractice action allowed under survival statute for lost chance of survival); Aasheim v. Humberger, 215 Mont. 127, 695 P.2d 824 (1985) (patient allowed to sue for lost chance of recovery); Perez v. Las Vegas Medical Center, 107 Nev. 1, 805 P.2d 589 (1991) (recovery allowed for loss of chance); Evers v. Dollinger, 95 N.J. 399, 471 A.2d 405 (1984)(cause of action based on increased risk allowed); Kallenberg v. Beth Israel Hospital, 45 A.D.2d 177, 357 N.Y.S.2d 508 (1974) (wrongful death action allowed for lost chance of survival), aff’d, 37 N.Y.2d 719, 374 N.Y.S.2d 615, 337 N.E.2d 128 (1975); McKellips v. St. Francis Hospital, Inc., 741 P.2d 467 (Okl. 1987) (wrongful death action allowed for lost chance of survival); Hamil v. Bashline, 481 Pa. 256, 392 A.2d 1280 (1978) (wrongful death action allowed for lost chance of survival); Truan v. Smith, 578 S.W.2d 73 (Tenn.1979) (cause of action based on increased risk allowed); Herskovits v. Group Health Cooperative of Puget Sound, 99 Wash.2d 609, 664 P.2d 474 (1983) (cause of action based on reduced chance of survival allowed); Thornton v. CAMC, Etc., 172 W.Va. 360, 305 S.E.2d 316 (1983)(proving increased risk is sufficient to prove causation); Ehlinger v. Sipes, 155 Wis.2d 1, 454 N.W.2d 754 (1990) (proof of reduction of chance of recovery sufficient to place issue of causation before jury); Joseph J. King, Jr., Causation, Valuation, and Chance in Personal Injury Torts Involving Pre-existing Conditions and Future Consequences, 90 Yale L.J. 1353 (1981) (argues for recognition of loss of chance); see generally John D. Hodson, Annotation, Medical Malpractice: “Loss of Chance” Causality, 54 A.L.R.4th 10 (1987); cf. Richard A. Posner, Economic Analysis of Law § 6.7, at 168 (3rd ed. 1986) (supporting cause of action for increased risk of death from toxic torts). But see Rewis v. United States, 503 F.2d 1202 (5th Cir.1974) (wrongful death action allowed under New Mexico law only if more probable than not that negligence caused the death); Grody v. Tulin, 170 Conn. 443, 365 A.2d 1076 (1976) (causation must be shown by “reasonable probability”); Gooding v. University Hospital Building, 445 So.2d 1015 (Fla.1984) (wrongful death action allowed only if more probable than not that negligence caused the death); Curry v. Summer, 136 Ill. App.3d 468, 91 Ill.Dec. 365, 483 N.E.2d 711 (1985) (action allowed only if more probable than not that negligence caused the death);Walden v. Jones, 439 S.W.2d 571 (Ky.1968) (plaintiff must show reasonable probability of recovery);Cornfeldt v. Tongen, 295 N.W.2d 638 (Minn.1980) (wrongful death action allowed only if more probable than not that negligence caused the death); Clayton v. Thompson, 475 So.2d 439 (Miss.1985) (no recovery for diminishment of chance of recovery); Pillsbury-Flood v. Portsmouth Hospital, 128 N.H. 299, 512 A.2d 1126 (1986) (no recovery based on loss of chance); Cooper v. Sisters of Charity of Cincinatti, 27 Ohio St.2d 242, 272 N.E.2d 97 (1971) (wrongful death action allowed only if more probable than not that negligence caused the death); Sherer v. James, 290 S.C. 404, 351 S.E.2d 148 (1986) (evidence of increased risk insufficient to show proximate cause).
 This cause of action requires only minor changes to current instructions and verdict forms, similar to those involved in comparative fault. The instructions would require the jury to find (and the verdict form would have a place for recording) the value of the lost life or limb; and, in percent, the exact chance of recovery lost. The court would then multiply these numbers to determine the exact net value of the damage award. A case of comparative fault would require that the trial court apply a second percentage. This Court expects that the MAI Civil Instructions Committee will propose implementing instructions.
 This action will be limited to those cases in which the chance of recovery lost was sizeable enough to be material, which must be so found by the jury. In the event statistics constitute relevant evidence, the lost chance must also be statistically significant within applicable statistical standards.