COFFIN, Senior Circuit Judge.

Plaintiff-appellant Sharon Primus claims in this medical malpractice case that her breast cancer progressed undetected to a serious level as a result of mis-diagnosis and substandard care by an Air Force doctor. The district court concluded otherwise and granted judgment for the United States after a non-jury trial on Primus's claim under the Federal Tort Claims Act (FTCA), 28 U.S.C. §§ 1346(b), 2671-2680. See Primus v. United States, 287 F. Supp.2d 119 (D.Mass. 2003). On appeal, Primus contends that the evidence does not support the court's ruling, and she further asserts that her case was unfairly prejudiced by the exclusion of crucial expert testimony. Finding no reversible error, we affirm.

I. Background

The medical care underlying this case began with a routine physical examination in 1989, which revealed a two-centimeter lump in appellant's breast. At the time, appellant's husband was an Air Force officer stationed at Luke Air Force Base in Arizona, and appellant was referred to a surgeon at the base. He concluded that there was no evidence of cancer. Nearly two years later, in July 1991, appellant first saw Dr. Earl Walker, another base surgeon and the doctor whose treatment is at issue in this case, because of a cyst in her right breast. Dr. Walker detected a four-millimeter lump by palpation, prompting him to make a preliminary finding of fibrocystic disease. He ordered a mammogram, and at a follow-up appointment in early August, again diagnosed the mass as fibrocystic disease. He described it then as a six-millimeter, smooth lump. According to medical evidence presented at trial, "a cancer is usually hard and irregular."

Appellant was examined by Dr. Walker for a third, and final, time in January 1992. His notes refer to a four-to-six-millimeter lump, and he recommended that appellant have another mammogram in July. In June, appellant saw a nurse at the base, Diane Musselwhite, who noted that appellant had a ten-millimeter lump in her breast. The record indicates that no follow-up mammogram was done.

In July 1992, appellant and her family moved to Massachusetts. She became pregnant the next month, and during a series of subsequent physician visits, the condition of her breasts — including a hardened area in her right breast — was attributed to the normal effects of pregnancy. In October 1993, Dr. Richard Galgano, a primary care doctor at a private medical clinic, saw appellant and noted a lump in the outer portion of her right breast. Appellant told Galgano that doctors had diagnosed a lump in that breast as noncancerous two years earlier, and she also told him that the mass had been stable in size and was not painful. No diagnostic tests were ordered.

Nearly a year and a half later, in March 1995, appellant consulted a general practitioner about the possible removal of the lump in her right breast. He referred her for testing and then to a surgeon when a mammogram indicated the presence of cancer. On March 29, the surgeon, Dr. Kevin O'Donnell, diagnosed breast cancer based on the mammogram and palpation of the lump, which was described in his subsequent report as "mobile" and "hard," and two square centimeters in size. A biopsy confirmed the presence of cancer, and on May 12, appellant's breast and twenty-one lymph nodes were removed in a radical mastectomy. She began chemotherapy the next month.

Appellant filed two civil actions stemming from her medical treatment, both claiming that she suffered personal injuries because her cancer was not timely diagnosed and treated. The first lawsuit was filed in Massachusetts against Dr. Galgano and his employer, Brighton Marine Health Center, Inc. The second action — the one currently before us — was originally filed in Arizona pursuant to the FTCA, and it sought damages from the United States based on Dr. Walker's treatment in 1991 and 1992. The two cases were consolidated after the Arizona case was transferred to Massachusetts. A simultaneous trial was held, with the case against the United States tried to the court, as required by 28 U.S.C. § 2402, and the case against Dr. Galgano tried to a jury. Appellant prevailed in the jury trial and was awarded $500,000 for negligence and $960,000 for future pain and suffering. That judgment was affirmed on appeal. See Primus v. Galgano, 329 F.3d 236 (1st Cir. 2003).

In the non-jury portion of the trial, the district court ruled in favor of the United States. It concluded that appellant had failed to show either causation — i.e., that the breast cancer detected in 1995 was traceable to the lump palpated by Dr. Walker four years earlier — or that Dr. Walker deviated from the applicable standard of care by failing to investigate the breast mass more aggressively.

Appellant's challenge on appeal is two-pronged. First, she claims that the court improperly excluded the testimony of an expert radiologist, severely damaging her ability to prove causation. Second, she claims that the evidence does not support the court's factual determinations as to causation and standard of care. We address each of these in turn.

II. Exclusion of Additional Expert Witness

Appellant claims that the district court abused its discretion in denying her motion to allow late designation of an additional expert witness. Three months after the deadline for disclosure of new experts, see Fed.R.Civ.P. 26(a)(2)(C), appellant had sought to add the testimony of Dr. Darrell Smith, an expert radiologist, to counter the testimony of the government's expert pathologist, Dr. James Connolly. The court, after conducting a hearing on appellant's motion, ruled that she had failed to meet her burden of showing "substantial justification" for her tardiness. See Fed.R.Civ.P. 37(c)(1). The court, however, did allow appellant to supplement the testimony of her previously designated expert, Dr. Mary Jane Houlihan, to rebut Connolly's testimony.

Appellant complains that the court failed to consider all of the relevant factors in reaching its decision, including that the government would have suffered no prejudice from the late submission while the harm to her from exclusion was substantial. She further argues that the court's decision is due limited deference because excluding it was tantamount to dismissal of her case, and such a "drastic sanction" should be imposed only in limited circumstances, Anheuser-Busch, Inc. v. Natural Beverage Distribs., 69 F.3d 337, 352 (9th Cir. 1995).

We decline to second-guess the district court in this instance. The adoption of Rule 37(c)(1) in 1993 "gave teeth to a significantly broadened duty" to comply with case management orders, Wilson v. Bradlees of New England, Inc., 250 F.3d 10, 19 (1st Cir. 2001), and "mandatory preclusion [is] 'the required sanction in the ordinary case,'" id. (quoting Klonoski v. Mahlab, 156 F.3d 255, 269 (1st Cir. 1998)). Although "preclusion of expert testimony is a grave step, not to be undertaken lightly," Thibeault v. Square D Co., 960 F.2d 239, 247 (1st Cir. 1992), the court here acted with due deliberation.

The court noted that plaintiff had been granted multiple extensions to complete her discovery and expert designations. Ultimately, the United States disclosed its expert reports on January 12, 2001, obliging appellant to respond by February 12. No response was forthcoming. The government filed a motion for summary judgment at the end of March. Although appellant sought and obtained extensions to respond to that motion, none of the requests stated a need to obtain additional expert testimony. Not until she filed her opposition to the motion for summary judgment, on May 21, did she move to designate an additional expert witness.

The lack of urgency in the request was further reflected at the hearing on the motion to designate a new expert, where appellant's counsel told the court that appellant's previously named expert, Dr. Houlihan, "was comfortable with rebutting Dr. Connolly herself." Counsel stated that strategically it would be "better to have an extra person there and available," but also said that "we have every reason to believe [Dr. Houlihan] can effectively professionally counter Dr. Connolly." In a related ruling, the court in fact assisted appellant in that effort. Describing as "close" the question whether to grant appellant's belated request to supplement Dr. Houlihan's testimony, the court allowed the submission so that appellant could more fully respond to the government's experts.

The court, moreover, considered, but was unimpressed with, appellant's explanations for her tardiness. Counsel attributes the delay to multiple factors: difficulties in securing an available expert; the unexpected weight given to Dr. Connolly's testimony in the government's motion for summary judgment; and the discovery after the deadlines of Dr. Smith's availability. But in light of appellant's representation to the court that, in effect, Dr. Smith's testimony would be cumulative of Dr. Houlihan's, none of these bespeaks an abuse of discretion on the part of the district court. Difficulties in securing witnesses are presumably a standard problem; appellant may have been surprised by the government's use of Dr. Connolly's testimony, but it had been available to her; and late discovery of a new potential witness cannot on its own require the court to disregard its previously set timetable. Deadlines would have no meaning if such rationales were sufficient to support reversing a trial court's judgment as an abuse of discretion.

The district court also was concerned that the United States had prepared a summary judgment motion in reliance on appellant's earlier disclosure of her expert evidence. Whether or not appellant is correct that the case management rules and Rule 37(c)(1)'s sanctions were not aimed at this sort of pre-trial prejudice, we cannot fault the court for considering the time and expense involved in the government's having prepared a dispositive motion. To be sure, evidentiary changes on the eve of trial are much more problematic and disruptive of trial preparation. See, e.g., Laplace-Bayard v. Batlle, 295 F.3d 157, 162 (1st Cir. 2002) (upholding exclusion where plaintiffs disclosed new expert witness "barely a week before trial"); Thibeault, 960 F.2d at 246-47 ("Many courts — this court included — have recognized the introduction of new expert testimony on the eve of trial can be seriously prejudicial to the opposing party."). But contrary to appellant's contention, what occurred here was not simply a matter of "inconvenience" or timing; real resources were expended on legal work that was premised on the expert evidence submitted before the deadline. That work was relevant not only for the summary judgment motion, but for trial preparation as well.

Given these circumstances, and particularly counsel's verbal assurance that, in effect, appellant's case would not be significantly compromised by exclusion of the additional witness, we cannot view the court's decision as "so wide of the mark as to constitute an abuse of discretion," Macaulay v. Anas, 321 F.3d 45, 51 (1st Cir. 2003). That it acted within reasonable bounds is also reflected in the court's willingness to allow belated supplementation of Dr. Houlihan's testimony to meet appellant's need. We therefore affirm the court's decision to exclude Dr. Smith's testimony from the trial.

III. Causation and Standard of Care

To prove that the medical care she received from Dr. Walker was negligent under Arizona law — and thus was compensable under the FTCA — appellant needed to show both a departure from a reasonable standard of care and proximate cause between his treatment and her injury. See Ariz.Rev.Stat. §§ 12-561-563 (Medical Malpractice Act); Thompson v. Sun City Cmty. Hosp., Inc., 141 Ariz. 597, 608, 688 P.2d 605, 616 (1984) (emphasis omitted) (trier of fact "must find for the defendant unless [it] find[s] a probability that defendant's negligence was a cause of plaintiff's injury."). The district court's findings that she satisfied neither of these obligations are reviewed for clear error. See La Esperanza de P.R., Inc. v. Perez y Cia. de Puerto Rico, Inc., 124 F.3d 10, 15-16 (1st Cir. 1997) (bench trial findings of fact subject to clear error review); Fed.R.Civ.P. 52(a). We deem a finding clearly erroneous only when, after reviewing the entire record, we are "'left with the definite and firm conviction that a mistake has been committed.' " García Pérez v. Santaella, M.D., 364 F.3d 348, 350 (1st Cir. 2004) (quoting Anderson v. City of Bessemer City, 470 U.S. 564, 573, 105 S.Ct. 1504, 84 L.Ed.2d 518 (1985) (internal citation omitted)).

Appellant's task on appeal is thus daunting. She not only bears the initial burden of proving the probability of a link between negligent care by Dr. Walker and the severity of her breast cancer, but she also must firmly convince us that the district court's ruling against her was unsupportable. Our review of the record leaves us unpersuaded.

We begin with appellant's contention that the district court erred in finding that Dr. Walker's treatment did not fall below the standard of care reasonably expected of health care providers in similar circumstances. Appellant specifically claimed that Dr. Walker should have pursued more testing to evaluate the persistent lump in her breast, ordering either a targeted ultrasound or fine needle aspiration. The record, however, contains ample evidence to support the court's determination that, with respect to breast lumps as small as appellant's, these techniques were not sufficiently effective in the early 1990s that they were a mandatory element of reasonable care. The court cited the testimony of Dr. Susan Pories, a breast cancer surgeon at Beth Israel Deaconess Medical Center ("Beth Israel"), who explained that ultrasonography was generally incapable of visualizing a mass of less than two centimeters in diameter and that fine needle aspiration of a mass as small as appellant's at the time Dr. Walker treated her (i.e., four to six millimeters) would have been "close to impossible."

That view was reinforced by other testimony. Dr. Janet Baum, a radiologist and Director of Breast Imaging at Beth Israel, testified that ultrasound was "just beginning to be used in 1991," and that she would not have recommended its use for such a small lesion because, even if the lesion were visible, "you really couldn't evaluate it adequately." Appellant's expert, Dr. Houlihan, a breast surgeon and Director of the Breast Care Program at Beth Israel, acknowledged that performing fine needle aspiration on a mass of that size "technically can be difficult."

Appellant contends that the court's conclusion on the standard of care is flawed because the court failed to consider the history of the growth of her breast mass, a progression that "should have informed the standard of care and required the employment of additional diagnostic procedures." Dr. Walker did not ignore the implications of appellant's history, however, having recommended another mammogram and further evaluation six months following appellant's last appointment with him. Appellant did not return to Dr. Walker, and the record indicates that a mammogram was not done again for three years.

In sum, the district court's conclusion, based on expert testimony, that the small size of the mass rendered ultrasound and fine needle aspiration of doubtful value, combined with Dr. Walker's articulated intent that appellant and her doctors remain watchful, does not permit us to term "clearly erroneous" the court's finding that Dr. Walker did not deviate from the applicable standard of care.

That conclusion on standard of care effectively disposes of the case. Appellant's theory of recovery is that Dr. Walker should have done more investigation, which would have revealed the non-benign nature of the small mass in her breast; with such early detection, appellant claims she would have avoided the more serious cancer and risk of death that subsequently occurred. Without a finding that Dr. Walker negligently treated appellant, however, we need not consider the causal link between his care and appellant's injury. Even if the lump Dr. Walker palpated in July 1991 was the source of the cancer found in her right breast in 1995, there is no medical malpractice and no recovery if Dr. Walker's diagnostic approach reflected the "degree of care, skill and learning expected of a reasonable, prudent health care provider" in similar circumstances. Ariz.Rev.Stat. § 12-563.

We nonetheless choose to address briefly the district court's finding — intensely debated by the parties on appeal — that appellant fell short of proving that the cancerous lump detected in 1995 was the same mass, in an early stage, encountered by Dr. Walker. The district court cited as "crucial" the testimony offered by Dr. O'Donnell, the surgeon who performed appellant's mastectomy. Dr. O'Donnell, who had reviewed appellant's medical history, testified that he was unable to say whether there was a connection — or not — between the earlier lump and the tumor he removed.

Dr. O'Donnell's testimony that the medical indicators were inconclusive was balanced on either side by the appellant's and the government's experts. Testifying for appellant, Dr. Houlihan stated that a relationship between the two masses was indicated by the similarity in location. The government's experts, meanwhile, opined that the post-operative pathology reports showed a rapid growth rate for the excised tumor, suggesting that it was of more recent development. Dr. Connolly, Chief of Anatomical Pathology at Beth Israel and the expert whose testimony appellant had sought to counter with Dr. Smith's excluded evidence, concluded that the tumor likely was doubling in size every thirty days — a rate too fast for the two masses to be related. Supporting the theory that the cancer was of more recent origin, Dr. Baum testified that her review of appellant's 1989 and 1991 mammograms did not reveal any abnormal masses. In response to the rapid growth theory, Dr. Houlihan "testified, essentially, that one cannot work backward from the tumor itself to know how long it was there," Galgano, 329 F.3d at 245, and noted that "these tumors will frequently change their behavior during their lifespan." She acknowledged, however, that the tumor was "very fast-growing . . . at the time that it was removed."

The district court ultimately found the government's scenario to be more persuasive, noting among other reasons that Dr. Houlihan's opinion was based on less than the full medical record. The court also found "particularly telling" Dr. O'Donnell's testimony that the lesion he removed was deep in appellant's breast. Dr. Walker testified that the mass he had palpated was superficial, and his recollection of the location of the lump was substantiated by Dr. Houlihan's testimony that palpating a mass less than one centimeter in diameter can be difficult if it is not near the surface of the breast.

In short, as we read the record, both sides offered cogent evidence in support of their positions through the testimony of highly qualified medical experts, all of whom were colleagues at the same hospital. We see no clear error in the district court's judgment that the government's experts were, on the whole, more persuasive. Although Dr. Connolly's testimony about the growth of the tumor factored into the court's analysis, we cannot accept the notion that exclusion of appellant's proposed rebuttal witness was decisive. As described above, Dr. Connolly was only one of multiple experts who offered opinions cited by the court negating the link between the mass palpated by Dr. Walker and the cancerous tumor that was found three years later. And, as counsel had predicted, Dr. Houlihan effectively responded to Dr. Connolly's suggestion that the post-operative pathology inevitably meant the two lumps were unrelated.

Even if Dr. Walker's treatment had constituted a breach of the applicable duty of care, and even if we deemed the battle of the experts a draw on the issue of causation — in our view, the best case for appellant supported by the record — the government would prevail in light of appellant's burden to show a probability of harm from Dr. Walker's treatment. Accordingly, we find no basis upon which to disturb the district court's judgment.

Affirmed.