This case arises upon the affirmance by the Minnesota Court of Appeals of a dismissal of appellant’s complaint by Ramsey County District Court. We affirm.

Respondent Dr. James Bellomo was the primary care physician for Appellant Delores Fabio from 1977 until he retired in 1986. Fabio alleges that on at least two occasions during this period Dr. Bellomo noticed a lump in her left breast, but told her not to worry about it because it was a “fibrous mass.” These visits occurred once between 1982 and 1984, and on March 10, 1986. Each time Dr. Bellomo allegedly noticed this mass, Fabio had gone to see him for an unrelated ailment.

After Dr. Bellomo’s retirement, Fabio went to Dr. Keith Chilgren, complaining of ailments unrelated to breast cancer. During the course of Dr. Chilgren’s examination, he noticed the lump in Fabio’s breast and ordered that a mammogram be performed on her. The mammogram showed two 2-centimeter tumors. A biopsy determined that the mass was cancerous, and that the cancer had metastasized to four lymph nodes. The tumor was subsequently excised, and Fabio underwent chemotherapy treatment.

Fabio then sued Dr. Bellomo, alleging that he committed medical malpractice when he failed to palpate the lump or order a mammogram when he noticed it on March 10, 1986. Prior to trial, Fabio sought to amend her complaint to include charges of malpractice against Dr. Bellomo for failing to properly treat her when he previously noticed the lump between 1982 and 1984. Fabio offered the testimony of *761Dr. Chilgren that a reasonably careful physician, under similar circumstances, would have ordered a mammogram.

She also offered her oncologist, Dr. Caldwell, as an expert witness to establish causation and damages. If allowed to testify, Dr. Caldwell would give his opinion, to a reasonable degree of medical certainty, that it is more probable than not that Fa-bio’s cancer spread from her breast tumor to the lymph nodes between 1984 and 1987. Dr. Caldwell would also testify that, in his opinion and to a reasonable degree of medical certainty, that it is more probable than not that additional lymph nodes became involved between March 10, 1986, the date of the last examination by Dr. Bellomo, and June 1, 1987, when Fabio saw Dr. Chilgren.

Fabio argued three forms of damage: First, she argued that the delay in treatment resulting from Dr. Bellomo’s failure to diagnose her cancer caused her to undergo chemotherapy; second, that the delay resulted in a “loss of chance” of life expectancy and a greater risk of recurrence of cancer; third, that the delay negligently aggravated her preexisting cancerous condition.

After briefs were submitted and depositions taken, the trial court dismissed Fa-bio’s complaint at a pre-trial conference. The trial court ruled, as a matter of law, that no cause of action existed for loss of chance or negligent aggravation of a preexisting condition. The trial court also dismissed her motion to amend her complaint, ruling that any malpractice occurring in the examinations during 1982-1984 was barred by the statute of limitations. Minn.Stat. § 541.07(1) (1992). These rulings were affirmed by the court of appeals 489 N.W.2d 241.

Essentially, two issues are before this court. First, we must determine whether the trial court abused its discretion by denying Fabio’s motion to amend her complaint to allege malpractice against Dr. Bel-lomo between 1982 and 1984. Second, we must determine whether Fabio has put forth sufficient evidence of causation and damages against Dr. Bellomo for his alleged malpractice.

When the trial court granted Dr. Bellomo’s motion to dismiss, it considered matters outside the pleadings, e.g., the deposition of Dr. Caldwell, Fabio’s oncologist. When matters outside the pleadings are presented to a court considering a motion to dismiss, and those external matters are not excluded by the court when it makes its determination, the motion to dismiss shall be treated as one for summary judgment. Minn.R.Civ.P. 12.02.

A motion for summary judgment shall be granted when the pleadings, depositions, answers to interrogatories, and admissions on file, together with the affidavits, if any, show that there is no genuine issue of material fact and that either party is entitled to a judgment as a matter of law. Minn.R.Civ.P. 56.03. On appeal, the reviewing court must view the evidence in the light most favorable to the party against whom judgment was granted. Abdallah, Inc. v. Martin, 242 Minn. 416, 424, 65 N.W.2d 641, 646 (1954). For purposes of this case, therefore, we must accept as true the factual allegations made by Fabio.

Fabio argues that the trial court erred by not granting her motion to amend her complaint. A party may amend a pleading by leave of court, and amendments should be freely granted, except where to do so would result in prejudice to the other party. Minn.R.Civ.P. 15.01; Hughes v. Micka, 269 Minn. 268, 275, 130 N.W.2d 505, 510 (1964). The trial court has wide discretion to grant or deny an amendment, and its action will not be reversed absent a clear abuse of discretion. LaSalle Cartage & Johnson Brothers Wholesale Liquor, 302 Minn. 351, 357-58, 225 N.W.2d 233, 237-38 (1974).

The trial court denied Fabio’s motion to amend because it found that “the proffered allegation of negligence in 1984 shows that Defendant’s ‘treatment’ of Plaintiff * * * was terminated upon the Defendant’s determination that Plaintiff had no condition of her breasts requiring further treatment on his part.” Because it found that there was no continuing course *762of treatment, the trial court denied Fabio’s motion to amend her complaint as barred by the statute of limitations. Whether the trial court abused its discretion by denying the amendment, therefore, turns on whether it was correct that there was no continuing course of treatment.

An action for medical malpractice is barred if not commenced within two years of the date on which the cause of action accrued. Minn.Stat. § 541.07(1). Generally, the cause of action accrues when the physician's treatment for the particular condition ceases. Johnson v. Winthrop Laboratories Division of Sterling Drug, Inc., 291 Minn. 145, 149, 190 N.W.2d 77, 80 (1971); Schmit v. Esser, 183 Minn. 354, 358, 236 N.W. 622, 624-25 (1931). The statute of limitations will be extended when a doctor’s negligence is part of a continuing course of treatment, such as when a doctor consistently fails to properly treat a fracture.

When Dr. Bellomo examined Fabio’s breast between 1982 and 1984, he did not recommend any further treatment. His treatment of her condition ceased at the time he told her not to worry about it. We therefore hold that the trial court was correct to rule that Dr. Bellomo’s examinations of Fabio’s breast that occurred between 1982 and 1984 are barred by the statute of limitations, because these examinations were not part of a continuing course of treatment. Because these examinations were barred by the statute of limitations, we also hold that the trial court did not abuse its discretion by denying Fabio’s motion to amend her complaint. Absent the amendment, therefore, Fabio may recover in medical malpractice from Dr. Bel-lomo only if she can prove damages attributable to his failure to diagnose her breast cancer in 1986.

To establish a prima facie case of medical malpractice, a plaintiff must introduce expert testimony demonstrating (1) the standard of care recognized by the medical community as applicable to the particular defendant, (2) that the defendant departed from that standard, and (3) that the defendant’s departure was a direct cause of the plaintiff’s injuries. Plutshack v. University of Minnesota Hospitals, 316 N.W.2d 1, 5 (Minn. 1982). In this case, Fa-bio has offered evidence by Dr. Chilgren that a reasonably careful physician would have ordered a mammogram upon feeling a lump like that described by Fabio. His testimony would establish the standard of care and show Dr. Bellomo’s breach of that standard. Fabio may recover from Dr. Bel-lomo, therefore, if she can show that his failure to order a mammogram was a direct cause of her damages.

To make out a prima facie case of causation in medical malpractice against Dr. Bellomo, Fabio must present expert testimony that establishes that it is more probable than not that damages resulted from his malpractice. Harvey v. Fridley Medical Center, 315 N.W.2d 225, 227 (Minn. 1982). Fabio argues that Dr. Bello-mo’s negligence caused her to suffer three forms of damages: chemotherapy, “loss of chance,” and “negligent aggravation of a preexisting condition.”

Fabio’s first theory of recovery is for damages caused by undergoing chemotherapy. At oral argument, however, Fa-bio admitted that chemotherapy would have been necessary even if Dr. Bellomo had diagnosed her cancer in 1986. Her complaint, therefore, fails to establish that Dr. Bellomo’s alleged malpractice was a direct cause of her need to undergo chemotherapy, and we hold that summary judgment for Dr. Bellomo was correct on this issue.

Fabio’s second theory of recovery is for “loss of chance.” She argues that her increased chance of a recurrence of cancer and her decreased chance of living another 20 years are compensable injuries. We have never recognized loss of chance in the context of a medical malpractice action, and we decline to recognize it in this case. Fabio argues that recoveries akin to loss of chance have been upheld in traditional tort actions, but we find those recoveries easily distinguishable; e.g., Mack v. McGrath, 276 Minn. 419, 423, 150 *763N.W.2d 681, 684 (1967) (holding that recovery for risk of malfunction of remaining kidney was “generous but not excessive” because it was “fair comment with respect to the implications of * * * being left with only one kidney”), Dunshee v. Douglas, 255 N.W.2d 42, 47 (Minn. 1977) (upholding recovery for risk of future stroke resulting from accident causing scar formation in carotid artery when stroke proven to be reasonably certain to occur). In those cases, when we allowed damages for potential ill effects from initial injuries caused by the defendants, the future effects flowed directly from the initial injuries, the initial injuries were the sole cause of the future effects, and the probabilities of their occurrence were proven with reasonable medical certainty. In this case, however, Fabio’s initial “injury,” her cancer, did not result from a misdiagnosis by Dr. Bellomo, and a misdiagnosis by Dr. Bellomo could not have been the sole cause of any future ill effects.

Further, we note that even if we were to adopt loss of chance as a theory of recovery in medical malpractice actions, it would not apply in this case because Fabio has failed to present evidence either that it is more probable than not that her cancer will recur or that she has a diminished life expectancy. Dr. Caldwell, if allowed to testify, would state that Fabio’s risk of cancer recurrence is about 30 percent and that her prognosis to survive at least 20 years is “at least 50-50.” In other words, if allowed to testify, Dr. Caldwell essentially would testify that it is more probable than not that her cancer would not recur and that it is more probable than not that she will survive at least 20 years. These are hardly the type of damages envisioned by the court in Dunshee when it allowed damages for the possibility of a future stroke when that possibility was established with reasonable medical certainty. We therefore hold that summary judgment against Fabio was appropriate on the issue of loss of chance.

Fabio’s third theory of recovery is for “negligent aggravation of a preexisting condition.” This theory was recently rejected by this court, under very similar circumstances. Leubner v. Sterner, 493 N.W.2d 119, 122 (Minn. 1992). We see no reason to reexamine Leubner at this time, and hold that summary judgment against Fabio is appropriate on the issue of negligent aggravation of a pre-existing condition. Because we reject all three theories of recovery argued by Fabio, we affirm the judgment of the lower courts.

Affirmed.

(dissenting).

This case involves two misdiagnoses made by Dr. Bellomo of a malignant tumor in Delores Fabio. The first misdiagnosis occurred sometime between 1982 and 1984, the second in 1986. The correctness of the trial court’s decision to disallow amendment of Fabio’s complaint turns on whether the statute of limitations for medical malpractice for the 1982-84 misdiagnosis was tolled until the 1986 misdiagnosis. Because I believe that the limitation period was tolled until the second diagnosis, I must respectfully dissent on this issue.

The common law imposed no limitation period on the bringing of a claim; any limitation on the time in which an action may be brought is statutory. R. Muscara, Note, Tort Law—Federal Tort Claims Act—Accrual of Medical Malpractice Action, United States v. Kubrick, 444 U.S. 111, 100 S.Ct. 352, 62 L.Ed.2d 259 (1979), 4 W.New.Eng.L.Rev. 155, 158 (1981). Statutes of limitation are based on the public policy encouraging the quick resolution of disputes. They are designed to “promote justice by preventing surprises through the revival of claims that have been allowed to slumber until evidence has been lost, memories have faded, and witnesses have disappeared.” Order of R.R. Telegraphers v. Railway Express Agency, Inc., 321 U.S. 342, 348-49, 64 S.Ct. 582, 586, 88 L.Ed. 788 (1944). Clearly, the main consideration underlying these statutes is one of fairness toward the defendant. “There comes a time when [a defendant] ought to be secure in his reasonable expectation that the slate has been wiped clean of ancient obligations * * Developments in the Law—Stat*764ute of Limitations, [hereinafter Developments] 63 Harv.L.Rev. 1177, 1185 (1950).

Normally, a statute of limitation begins to run when the cause of action accrues. Id. at 1200. Although it seems obvious that a cause of action would accrue at the time a suit could theoretically be maintained (i.e., at the time tortious act occurred), courts have often delayed the beginning of the statutory period until an event occurs without which a suit would be impossible or improbable (e.g., until the plaintiff learned of the wrong1 or substantial damage occurred). Id.

In many kinds of tort actions, the act alleged to cause injury and the injury resulting therefrom occur simultaneously and are known to the injured party immediately. However, in medical malpractice cases, the injury may be totally unknown to the injured party for years after the tortious act occurs. The customary application of the rule can result in unfairness if the plaintiff does not know she has been harmed until the limitations period has run.

There is an obvious injustice in holding a plaintiff with a meritorious malpractice claim to a strict application of the statute of limitations, i.e., that his cause [of action] accrued at the time of the physician’s wrongful act or omission. A patient, who places his trust in his doctor, may know nothing of what is affecting him. He may know only that he is not getting well. He may be treated for years before learning that, initially, his case was misdiagnosed, or that at the onset of his treatment he was given the wrong drug, one which worsened rather than bettered his condition.

D. Harney, Medical Malpractice, § 8.3 at 260 (1973).

The “continuing course of treatment” doctrine has evolved as a compromise to ameliorate potential unfairness resulting from delayed discovery of an injury.2 Under this doctrine, the statute of limitations is tolled until treatment is ceased by the negligent physician for the injury which formed the basis for the cause of action. The rationale underlying the doctrine is to foster the physician-patient relationship by allowing the patient to seek corrective treatment while maintaining her legal remedy,3 thus affirming the trust and confidence a patient has in her doctor.4 In *765Grondahl v. Bulluck, 318 N.W.2d 240 (Minn. 1982) we stated that three factors should be considered to determine when a continuing course of treatment ceased:

(1) whether there is a relationship between physician and patient with regard to the illness; (2) whether the physician is attending and examining the patient; and (3) whether there is something more to be done.

Id. at 243 (citing Schmit v. Esser, 183 Minn. 354, 358-59, 236 N.W. 622, 625 (1931)).

It is clear that the second factor was satisfied in this case; Dr. Bellomo saw Fabio many times between 1982 and 1986. Further, the third factor was also met in this case; it is obvious that accurate diagnosis and subsequent treatment of Fabio’s cancer was required. Therefore, the only relevant inquiry concerns the first factor.

The majority states that Dr. Bellomo “treated” Fabio’s cancer when he told her “not to worry about it,” op. at 762, in effect, saying that the relationship between Dr. Bellomo and Fabio as to Fabio’s cancer began and ended each time he made a misdiagnosis. However, this view of the relationship begs the question as to whether our requirements for a continuing course of treatment are met. Treatment by a physician must necessarily include diagnosis. Proper treatment requires an accurate diagnosis — which, of course, was not present in this case. Once the admission is made by the majority that the misdiagnosis constituted treatment, and that he so “treated” her illness several times over the years, one must ask why the requirement of Grondahl is not met. It can only be that the majority implicitly affords less value to the “diagnosis” phase of treatment than to other forms of treatment.

Although it seems incongruous that subsequent treatment can occur without affirmative action by the physician since the term treatment connotes the presence of action, in certain situations “treatment” can occur by omission. This treatment by omission arises when a patient returns to the treating physician complaining of problems in the mistreated area but the physician disregards the complaints. The significant factor is that even though the physician may not have provided literal treatment to the afflicted area, the patient, by returning to the physician, has provided him with an opportunity to correct his previous error. The patient’s return visit has not only continued the physician-patient relationship, but it has also continued “treatment” of the specific injury.

Dana David Peck, Comment, The Continuous Treatment Doctrine: A Toll of the Statute of Limitations for Medical Malpractice in New York, 49 Alb.L.Rev. 64, 79 (1984) (citations omitted). If Dr. Bellomo, through his physician/patient relationship with Fabio, was “treating” her by examining the lump in her breast with an eye to diagnosis, I can only conclude that there was a continuing course of treatment during this time.

This court has never faced the question of whether multiple misdiagnoses, acknowledged by the majority to constitute “treatment,” are sufficient to toll the limitations period. However, other jurisdictions have reached this issue. In Fonda v. Paulsen, 46 A.D.2d 540, 363 N.Y.S.2d 841 (1975), the plaintiff underwent a biopsy in May 1969. No evidence of cancer was found and the plaintiff was so informed. The plaintiff saw the defendant doctor five times during 1970 for injuries unrelated to the cancer, but nevertheless complained about the spot from which the biopsy was taken on at least two occasions. In 1972, the defendant referred the plaintiff to another doctor for another biopsy which showed that a malignancy had existed since 1969 in the spot previously examined, and which should have been detected by the 1969 biopsy-

*766The plaintiff did not bring suit until 1974. The physicians claimed the action was barred by the three-year limitations period measured from the time of the negligent diagnosis. The trial court dismissed the suit finding that no treatment occurred after May 1969. In reversing the trial court, the appellate court stated:

If a patient continues under post-operative observation by his physician and is advised that his condition is being cured, this is as much “treatment” as affirmative acts such as surgery, therapy or prescription of medicines. Any other view must be based on the erroneous supposition that only acts of commission, but not acts of omission, can constitute negligence; there would be little chance for legal redress by a patient who has been the victim of an alleged malpractice who is advised that time is the only barrier to a complete cure, when in reality, time is a barrier to a cause of action.

Id., 363 N.Y.S.2d at 844 (citation omitted).

In McDermott v. Torre, 452 N.Y.S.2d 351, 437 N.E.2d 1108, in May 1974, the defendant physician examined a mole on the plaintiffs ankle, excised a portion of the mole and sent it to a lab for a biopsy, which came back negative. The plaintiff saw the defendant physician at least eight times between May 1974 and September 1976 for various other conditions, but also complained about continuing pain in her ankle and the grayish color present in the area. Each time she complained, the physician told her there was no cause for concern. Eventually, the plaintiff learned that the diagnosis was wrong and a malignant melanoma was removed from the site of the mole.

On appeal from summary judgment in favor of the defendant physician, the plaintiff argued that the continuing course of treatment acted to toll the statute of limitations until 1976. In addressing that argument the court stated:

[A] complete discharge by a physician [does not] forever [bar] a finding of continuing treatment. Included within the scope of “continuous treatment” is a timely return visit instigated by the patient to complain and seek treatment for a matter related to initial treatment. Thus, there will be continuing treatment when a -patient, instructed that he or she does not need further attention, soon returns to a doctor because of continued pain in that area for which medical attention was first sought.

Id. at 1111 (emphasis added). The rule that has evolved in New York is that when a physician tells a patient there is nothing more to be done for a particular injury, and the patient returns to that doctor for other reasons, but also complains about the earlier injury, as long as there is an ongoing relationship with the physician, the physician’s non-treatment is sufficient to toll the statute of limitations.

In Shumway v. DeLaus, 152 A.D.2d 951, 543 N.Y.S.2d 777 (1989), motion for leave to appeal dismissed 75 N.Y.2d 946, 555 N.Y.S.2d 693, 554 N.E.2d 1281 (1990), the Appellate Division of the Supreme Court of New York, Fourth Department, refused to hold that a 2¾⅛ year gap in treatment, the length of the medical malpractice limitations period in New York, barred the continuing course of treatment doctrine as a matter of law. Instead of adopting a rigid approach, it articulated the following standard:

[W]e adhere to the principal that “where the physician and patient reasonably intend the patient’s uninterrupted reliance on the physician’s observation, directions, concern, and responsibility for overseeing the patient’s progress, the requirement of continuous treatment for the purpose of the Statute of Limitations is certainly satisfied.”

Id. (quoting Richardson v. Orenteich, 64 N.Y.2d 896, 487 N.Y.S.2d 731, 477 N.E.2d 210 (1985).

Between 1982 and 1986, Fabio saw Bello-mo approximately 60 times for various ailments. She saw no other doctor during that time. She did not complain of continuing pain in the area where Bellomo made the misdiagnosis, but the malignant tumor, which Bellomo determined to be a fibrous *767mass, was nevertheless brought to his attention by his own examination. Surely, the only reason' for requiring a patient to complain about the area previously misdiagnosed is to bring it to the attention of the physician. The key .inquiry must be whether the physician was put on notice, through the actions of her patient or herself, that an abnormal condition existed. Dr. Bellomo certainly possessed such knowledge.

A physician’s diagnosis that nothing further is necessary does not end “treatment” if the physician is subsequently given notice within a reasonable time that the initial diagnosis was wrong. Treatment by omission can toll the statute of limitations if the subsequent misdiagnosis occurred within a reasonable time after the initial diagnosis, and there was a continuing relationship between the physician and the patient.

I respectfully dissent from the majority opinion on the issue of whether a continuing course of treatment existed, so as to toll running of the statute of limitations until 1986.

(dissenting).

I join in the dissent of Justice Gardening.

(dissenting).

I join in the dissent of Justice Gardebr-ing.