dissents in a memorandum as follows: In reviewing the jury’s finding of liability in this psychiatric malpractice action, plaintiff is entitled to the most favorable view of the evidence and the benefit of every reasonable inference which can be drawn from that evidence (Eppel v Fredericks, 203 AD2d 152). When, as here, the record contains conflicting medical testimony, the resolution of such conflicts, as well as the evaluation of the credibility and accuracy of witnesses’ testimony is a matter within the province of the jury (Aunchman v Palen, 186 AD2d 104, lv denied 81 NY2d 702; Shaw v Binghamton Lodge No. 852, 155 AD2d 805). If the jury’s verdict in this matter was not "utterly irrational”, given the evidence in the record, it should not be set aside for *389legal insufficiency (Hoffson v Orentreich, 168 AD2d 243, 245). A jury’s verdict should not be set aside as against the weight of the evidence "unless the jury could not have reached the verdict on any fair interpretation of the evidence” (supra, at 244). With these principles in mind, I cannot agree with the majority that the trial court erred in refusing to grant defendants’ motion to dismiss the complaint at the close of all the evidence because of legally insufficient evidence. Therefore, I dissent.
When plaintiff Josephine Vera, a 34-year-old chronic schizophrenic, was involuntarily committed by her sister to the psychiatric unit at the Beth Israel Medical Hospital on June 27, 1984, she was agitated, disoriented, and experiencing auditory command hallucinations of a lethal nature. Ms. Vera’s admission record notes: "Pt. grossly disorganized and psychotic as well as expressing suicidal thoughts [with] plan to jump off bridge. Pt. is unable to care for herself & sister is unable to care for her.” Plaintiff was given Thorazine to sedate her.
The next day, his first day working for Beth Israel as its inpatient unit chief, Dr. Harold Schwartz conducted his initial examination of Ms. Vera. Dr. Schwartz noted in her record:
"34 year old Hispanic female admitted last night [with] hx of three weeks of increasingly agitated behavior, auditory hallucinations and suicidal thoughts. Last night she was experiencing command auditory hallucinations to jump out a window. She felt compelled to obey & was brought to ER by sister. * * *
"Auditory hallucinations are persistent and she states she will kill herself if the voices don’t stop. She denies symptoms of depression or mania or recent drug abuse.”
Plaintiff had been hospitalized on prior occasions for acute exacerbations of her schizophrenic condition. These past acute exacerbations were believed to be related to plaintiff’s failure to take the medications that had been prescribed for her.
Dr. Schwartz noted on the chart that appropriate medication served to decrease plaintiff’s "anxiety about voices, hallucinations and may be responsible for diminishing their threatening nature.” At trial, Schwartz observed that plaintiff’s "exacerbations, the flareups, were due to, at least in part, to not taking medication.”
Haldol and then Prolixin were administered to plaintiff in increasing dosages. The medication had the desired effect: Ms. Vera’s hallucinations grew "softer” and there was no sign of *390lethal command hallucinations after the third day of hospitalization. Given the plaintiffs history of non-compliance in taking her medicine, a switch from Haldol to oral Prolixin was made in anticipation of administering Prolixin Decanoate, an injectable longer-lasting form of the drug.
The oral Prolixin was increased on July 8th from 25 milligram to 35 milligram doses, taken twice daily. Ms. Vera was stabilized on the latter amount, a total of 70 milligrams daily, from July 9 through July 14. Dr. Schwartz believed that it was having sufficient therapeutic effect.
Dr. Schwartz discontinued the oral Prolixin on July 14, with the last 35 milligram dose given at 6:00 p.m. on that date. He directed that plaintiff receive a 50 milligram injection of Prolixin Decanoate the next morning. This was done in anticipation of plaintiffs discharge. However, as plaintiff’s psychiatric expert testified, and the jury was empowered to believe, that amount was inadequate to maintain plaintiff on the same therapeutic level of oral medication she had been on (70 milligrams daily). In making a transition from oral to inject-able Prolixin, Dr. Tuckman explained, "you need about 25 percent, 20, 25 percent more of the injection to equal the oral form because of the way they are absorbed and the way it is dissipated through the body.” Thus, by Dr. Tuckman’s calculation, plaintiff would require about 87.5 milligrams of Prolixin Decanoate to maintain the same therapeutic effect as the oral dosage.
Dr. Tuckman also noted that the final dosage of oral Prolixin no longer had any therapeutic effect 24 hours later—by the early evening of July 15. While plaintiff had Prolixin in her bloodstream at that time—since oral Prolixin has a half-life of three to five days—it was an amount far less than 70 milligrams. Because of the different, and substantially slower, process by which an intramuscular injection of Prolixin is absorbed into the bloodstream, the 50 milligram injection administered to plaintiff at 11:00 on July 15, had no effect in the following 48 or more hours. For a significant period of time, therefore, plaintiff was no longer receiving the therapeutic benefit of the oral Prolixin and not yet benefiting from the Prolixin Decanoate. It was during this "window period” that Dr. Schwartz discharged plaintiff from the hospital at 9:30 a.m. on July 17.
Notwithstanding defendants’ attempts to characterize it otherwise on appeal, it was not Dr. Tuckman’s testimony that it would have been appropriate for Dr. Schwartz to discharge *391plaintiff if only he had waited exactly 48 hours. As the majority correctly notes, none of the medical witnesses contended that such a mechanical approach would have been compatible with responsible and careful medical practice. The therapeutic effect of antipsychotic medication, all the witnesses agreed, must be determined by clinical examination, that is, by observing the patient’s response over the course of time.
The 48-hour period referred to by plaintiff’s expert represented a minimum period before which there would have been no point at all in attempting to evaluate the therapeutic effect of the injection. It was Dr. Tuckman’s opinion that Dr. Schwartz should have watched plaintiff and assessed her behavior for a period of time following the 48-hour period. He stated that "you wait at least two days or more to make sure that the effect of the Decanoate is going to be at [the same therapeutic] level [of the pills].”
Defendants’ own expert, Dr. Chalemian, agreed that far more than a limited observation is required to make proper assessment of the drug’s effect. One must talk to the patient "repeatedly over time * * * to assess that the command hallucinations of a dangerous nature disappeared” in response to the medication. Thus, while it is true as the majority points out that predictions of future suicidal conduct necessarily involve a degree of calculated risk, the issue before the jury was whether Dr. Schwartz went about computing that risk in Ms. Vera’s case in a manner consistent with the duty of care he owed her. The jury’s finding that Dr. Schwartz committed malpractice, by failing to conduct repeated interviews and observation of Ms. Vera after the medicine could reasonably have been expected to have taken effect, while perhaps not the only conclusion which could have been reached based on the evidence in the record, is one which clearly has a foundation in the evidence.
On July 13 the nursing staff, social workers and other hospital employees went on strike. According to Dr. Henry Pinsker, an Associate Director of Beth Israel’s psychiatric department from 1974 to 1992, "to the extent that it was consistent with good practice, patients would be discharged a little bit earlier [due to the strike].” Dr. Schwartz prepared plaintiff’s discharge resumé and his evaluation of her "discharge condition” on July 15th, two days prior to her actual discharge. He initially decided to discharge plaintiff "to self’ on the morning of July 16—not even 24 hours after the administration of the injectable Prolixin. He subsequently *392changed the discharge time to the morning of the 17th because plaintiffs sister requested additional time to hire a home attendant or aide.
On July 15 Dr. Schwartz observed that plaintiff "no longer has command hallucinations to kill herself, continues to have aud. halluc. which comment on her behavior. Remains withdrawn & seclusive.” Dr. Pinsker testified that "if everybody was doing the job right,” the hospital record should have contained a page, on which psychiatric notes and observations were recorded, for each and every day that plaintiff was confined—that is, from June 27 to July 16. There was no "daily psychiatric notes” page in the record corresponding to July 16 (the day before plaintiffs actual discharge), an omission which Dr. Pinsker regarded as unusual.
Dr. Pinsker suggested that in such circumstances one might look to the nursing flow chart for some indication of how the patient was doing on that date. The nursing flow chart reported that plaintiff had a significant change in behavior during the day shift on the 16th, as a psychiatric resident, Dr. Strumpf, noted by checking the appropriate box on the preprinted form. Based on the testimony about the way the form was intended to be used, the jury could rationally infer from this indication that plaintiff underwent a negative change. Although the corresponding page with psychiatric notes for the 16th was missing, it is notable that, on the only other day that a change in behavior was indicated on the nursing flow chart (June 29), the psychiatric notes for that date reported suicidal ideation, severe depression, and crying and screaming.
On the day of plaintiffs discharge, her sister picked her up at 9:00 a.m. Dr. Schwartz told plaintiffs sister that Ms. Vera should return to the emergency room in a week for an injection of medication. He also noted that plaintiff was still hearing voices. According to plaintiffs sister, Ms. Vera did not look very well and appeared mechanical.
Once home, plaintiff was introduced to the nursing aide, who prepared some food for plaintiff and then sat down to watch television. Ms. Vera remembers hearing voices at that time instructing her: "You better jump out the fire escape.”
Plaintiff ate the food, went to the fire escape at the kitchen window, and jumped.
As a result of the accident, plaintiff suffered severe injuries, including amputation of her right leg above the knee, paralysis from the waist down, and complete loss of control of bladder and bowel functions.
*393This review of the evidence makes clear that, contrary to the position of the majority, the trial court did not err in submitting this case to the jury, as there was legally sufficient evidence upon which a jury could find defendants liable. Furthermore, the verdict which the jury did reach, was not against the weight of the evidence. Dr. Tuckman’s testimony does not represent, as the majority states, " 'at most, a difference of opinion among physicians’ ” which cannot support a finding of malpractice. The majority’s selective focus on Dr. Tuckman’s response to one question on cross-examination, which must be viewed in context, ignores all the other evidence on which a jury could conclude that under all the then existing circumstances, Dr. Schwartz’s July 15 decision to discharge plaintiff was undertaken in a manner that constituted a departure from accepted standards of care. In short, a fair reading of the record supports a rational line of inferences by which a jury could conclude that the decision to release Ms. Vera, made two days after the strike started and less than 24 hours after the injection of Prolixin, and apparently without proper consideration of the negative change noted by Dr. Strumpf on July 16, constituted malpractice. Thus, the evidence was not legally insufficient, and the trial court did not err by refusing to grant the motion to dismiss the complaint before submitting the case to the jury.
Not only would I affirm the trial court’s decision to submit this case to the jury, I would reject the alternative grounds for reversal advanced by defendants, that the verdict reached by the jury was against the weight of evidence. The jury’s finding of liability on the theory that Ms. Vera was released by defendants before a proper determination could be made whether the Prolixin Decanoate was having its desired therapeutic effect and whether she was medically fit to be discharged from the hospital should be affirmed.