SCHNEPP, J.

This appeal concerns the liability of the State of New York for its alleged failure to use acceptable medical procedures in the care and treatment of Richard Carl Larkin while he was a prisoner in the Attica Correctional Facility and particularly for its alleged failure to diagnose and treat what was later determined to be a berry aneurysm. Larkin died on January 3, 1976 at the age of 28. The autopsy report indicates that the immediate cause of his death was a "[r]uptured berry aneurysm with massive subdural and subarachnoid hemorrhage". At the conclusion of the trial of the claim to recover damages for Larkin's wrongful death, the Court of Claims held that the claimant failed to prove medical malpractice and dismissed the claim.

The events leading to Larkin's death are largely undisputed. On December 2, 1975 he went to the prison clinic complaining of a headache that started the night before. Dr. Vratislav Kejzlar, a graduate of a European medical school who was not licensed but permitted to practice in New York State pursuant to a statutory exemption (see Education Law, § 6526, subd 1), examined Larkin and made an entry in the medical records that Larkin had spasms in the back of his neck and a headache. He prescribed medication consisting of Roboxin, a muscle relaxant, and Darvon, an analgesic. On December 8 Larkin, complaining that he had "headaches for [the] last week or so", was again seen by Dr. Kejzlar. He was given more Darvon and Orinade, an antihistamine. On December 12 Larkin, complaining of vomiting, was seen at the clinic by a physician's assistant. An entry was made on his medical records that he had a history of ulcers. On December 19 Larkin was treated for urethritis, an unrelated condition. On December 21 Larkin went to the clinic where he stated that he had a headache and had passed out. His blood pressure, pulse and temperature were normal. A registered nurse prescribed Darvon and told him to report on sick call. On December 22 Larkin was again seen at the clinic complaining of a headache. A registered physician's assistant prescribed Roboxin and Tylenol. On December 23 Larkin again went to the clinic with complaints of a stiff neck and pain. A registered nurse gave him Roboxin and Darvon and told him to report in the morning. On December 24 Larkin appeared at the clinic and told Dr. Kejzlar about his continuing headaches and the neck pain and stiffness; Darvon was prescribed for a one-week period along with Roboxin. On December 25 Larkin again went to the clinic where a registered nurse gave him one injection of dihydroergotamine for probable migraine headaches. Later the same day Larkin passed out in the dining hall and was carried to the clinic on a stretcher. The medical records state that he was conscious but unresponsive. A registered physician's assistant tested Larkin's tendon reflexes (Babinski test) and found them to be normal. He also performed tests on Larkin's eyes with similar negative results. There was no indication of a neurological deficit and the physical examination was within normal limits. Larkin was admitted to the prison hospital with the diagnosis of possible drug reaction to the injection he had received earlier in the day. The nurse's record for December 25 contains the entry under "remarks" that Larkin "[c]omplained frequently of severe headaches * * * [and] [s]tates he thinks he may have a stroke". That night Larkin, according to the nurse's record, was "[a]wake all night [and complained of] pain behind eyes, at temples and neck muscles. Very restless". On December 26 Larkin, who was still complaining of headaches and neck stiffness, was examined by Dr. Kejzlar who performed a simple neurological test (Kernig test) to determine whether Larkin had any kind of irritation of the central nervous system. The results of the test were negative. The physical examination was also negative except for the finding that the "muscles of dorsum of neck [were] palpably tender". Dr. Kejzlar discontinued Valium which had been prescribed on December 25 and prescribed Darvon. He discharged Larkin from the prison hospital. The health record signed by Dr. Kejzlar reflects that no abnormality was found for Larkin's complaint of headaches and that his vital signs were normal. The hospital summary for December 25 and 26 signed by Dr. Kejzlar indicates that Larkin had a history of headaches for three weeks; that the headaches remained constant during his hospital stay; and, that Darvon provided "mild relief" from the headaches. The hospital summary contains a final diagnosis of "[t]ension headaches". On December 29 Larkin returned to the prison clinic complaining of a headache. Dr. Jelinek, a certified radiologist, clinic physician and immediate supervisor to Dr. Kejzlar, examined Larkin and X-rayed his cervical spine. The X-rays showed a straightening of the normal lordotic curvature which was consistent with muscle spasm. The doctor noted in the record that Larkin was "better, still some muscle spasm". Larkin next appeared at the clinic on January 3, 1976, the date of his death. He complained of a headache and neck stiffness. The health record notes that Larkin "appears very apprehensive". Darvon was continued. Later that day, Larkin was found unconscious by a nurse. Dr. Kejzlar summoned an ambulance and Larkin died en route to the hospital in Batavia.

Claimant contends that the State failed to employ proper medical techniques or adequate medical personnel in the care, treatment and diagnosis of Larkin's complaints and failed to exercise professional judgment in the diagnosis of his condition by simply treating him for headaches and releasing him time after time. The proof to support this claim of medical malpractice came from Dr. Charles Salamone, a qualified neurologist, who testified that berry aneurysms are treatable, that a severe headache, along with associated stiffness and neck pain and possibly vomiting, are the symptoms of a hemorrhaging aneurysm, and that the condition does not usually cause a neurological deficit. He said that altered consciousness or loss of consciousness "may occur immediately with the initial bleed or with subsequent leakage or bleeds, if the patient survives." In his opinion the general physician should have realized, at least within the first week, that he was not dealing "with a simple muscle contracture [and] headache on the basis of the patient's history, complaints and findings"; that the continuation and worsening of the headaches were significant; that good and accepted medical practice demanded either that Larkin be referred to a specialist such as a neurologist or a neurosurgeon or that a spinal tap to detect the presence of a hemorrhaging berry aneurysm be performed; and, that the treatment afforded Larkin was not in accordance with good medical practice. He testified that a negative Babinski test would not be unusual and that the Kernig test "is an involuntary sign of meningeal irritation" which has the "same implication that a stiff neck does". The "ultimate test", according to Dr. Salamone, is a spinal tap which can be performed by a general physician. Such a test, he said, "would have shown blood the first day * * * on the 2nd of December" because Larkin was hemorrhaging at that time and continued to hemorrhage until his death. Dr. Salamone testified that surgery or conservative treatment such as bed rest and sedation are the two possible methods of treating a berry aneurysm. According to Dr. Salamone, Larkin's aneurysm could have been treated and cured anytime prior to the "blowout" on January 3. He said that the aneurysm was accessible and that with surgical treatment there was "at least a 75 per cent chance of survival". Dr. Salamone conceded that the frequency of aneurysms generally increased with age with the highest percentage occurring in the fifth to sixth decades of life and that Larkin's symptoms were common and could be caused by other ailments. Nonetheless, he testified that aneurysms are not difficult to diagnose; that the critical factor in this diagnosis is the patient's history; and, that with Larkin's symptoms the physician's obligation was to consider at least the possibility of an aneurysm. He said that after Larkin passed out on December 21 the physician should have considered this symptom in conjunction with the "whole spectrum of symptoms" and recognized that there was "something more organically wrong than simple muscle contraction with a headache". Dr. Salamone also opined that the injection of dihydroergotamine on December 25 was not good and accepted medical practice because it "works by constricting the arteries which is exactly what you don't want to happen in this situation" and because the drug was ostensibly prescribed by a registered nurse who had no authority to do so.

The only proof that the State presented at trial consisted of Larkin's prison and medical records and certain portions of the examination before trial of Dr. Kejzlar. Although Dr. Kejzlar, in his testimony, conceded that it would have been good and accepted medical practice to have referred Larkin to a neurologist on December 26, he refused to concede that the failure to do so violated accepted medical standards. Dr. Kejzlar also testified that a referral to a licensed physician or to his immediate supervisor (Dr. Jelinek) on December 26 would have constituted good and accepted medical practice. He claimed that he consulted with and referred Larkin to Dr. Jelinek, although the medical records do not reflect this conduct. It is noted that the December 29 medical records reflect that Larkin demanded to see a medical doctor or a physician's assistant. It was on this date that Larkin was examined by Dr. Jelinek. Dr. Kejzlar maintained that under the circumstances of the case the care and treatment rendered to Larkin was in accordance with good and accepted medical practice.

At the conclusion of the trial the State moved to dismiss on the ground that claimant failed "to prove a prima facie" case and established nothing more than a "misdiagnosis". The court granted the motion but in doing so stated that it had observed the manner and demeanor of the witnesses and had reviewed all the evidence presented. It found that the claimant submitted no proof at trial that "the doctor failed: (1) to possess a reasonable degree of learning and skill; (2) to use reasonable care and diligence in exercising his skill; (3) to exercise his best judgment in exercising his skills and applying his knowledge ( Pike v. Honsinger, 155 N.Y. 201)." The court held that the Attica medical staff was faced with a complicated medical problem and that the proof failed to show that the treatment accorded Larkin was other than "treatment based honestly and conscientiously on careful examination. Improper diagnosis honestly arrived at is an error of professional judgment for which the State cannot be held liable ( St. George v. State of New York, 283 AD 245, aff'd 308 N.Y. 681)."

We construe the trial court's determination as a dismissal, not because claimant failed to prove a prima facie case as a matter of law, but because claimant failed to prove on the entire record that the State was negligent in the medical treatment it accorded Larkin. However, whether the trial court dismissed the claim as a matter of law, or based its determination upon the conflicting evidence adduced at trial is of little moment (see Siegel, New York Practice, § 529, p 732), since we have the power, in any event, to set aside the trial court's findings if they are contrary to the weight of the evidence and to render whatever judgment the Trial Judge ought to have granted (see, generally, CPLR 5501). Since the record here is complete, we can make new findings of fact on the question of liability if the trial court incorrectly assessed the evidence and its conclusion could not be reached under any fair interpretation of the evidence (see Cohen v. Hallmark Cards, 45 N.Y.2d 493, 498; Hale v. State of New York, 53 A.D.2d 1025; Billington v. State of New York, 33 A.D.2d 822, 823; see, also, McCauley v. State of New York, 8 N.Y.2d 938; 7 Weinstein-Korn-Miller, N.Y. Civ Prac, pars 5522.04, 5522.05). Although the credibility of witnesses is an issue for the trier of the facts (see Bandike Assoc. v. B.B.M. Realty Corp., 44 A.D.2d 622), our inquiry is not confined to an examination of whether a finding is supported by some credible evidence. If it appears that upon all the credible evidence a different finding would have been reasonable, we must "weigh the relative probative force of conflicting testimony and the relative strength of conflicting inferences that may be drawn from the testimony." ( People ex rel. MacCracken v. Miller, 291 N.Y. 55, 62; see, generally, 7 Weinstein-Korn-Miller, N.Y. Civ Prac, par 5501.20.)

We recognize that "[t]he rule requiring [a doctor] to use his best judgment does not hold him liable for a mere error of judgment" ( Pike v. Honsinger, 155 N.Y. 201, 210, supra), but we do not classify this case as one of improper diagnosis within the meaning of St. George v. State of New York ( 283 App. Div. 245, supra). St. George involved an erroneous diagnosis of the true nature of a mental illness and the doctrine of that case does not apply to the case at bar. In this case the negligence, if any, lies not in an improper diagnosis, but in the failure to recognize the possibility of a hemorrhaging berry aneurysm from what the expert witness described as "classic symptoms". The question is whether the failure to recognize the latent neurological condition constitutes negligence.

Our analysis of the evidence adduced at trial does not sustain the trial court's resolution of the factual issues and its conclusion of nonliability on the part of the State. Larkin had complained of headaches for 24 days when Dr. Kejzlar discharged him from the clinic on December 26. His complaints were daily after December 21. Even Dr. Kejzlar conceded that the symptoms were increasing. Nothing that was done for Larkin significantly improved his condition. Despite the existence of recognized symptoms, the possibility of a hemorrhaging aneurysm was not even considered. In fact, the registered physician's assistant was unaware of the symptoms of the condition.

Although Larkin's headaches had persisted over this period of time and his symptomatology had increased and had become more severe, Dr. Kejzlar could find no neurological sign on December 26 and in his own words proceeded "in simple fashion". The chronic and accelerating nature of the symptoms indicated that Larkin's condition was more than transitory and demanded immediate diagnosis and treatment. It is no defense that the symptoms were common and could be caused by other ailments or that aneurysms are more prevalent in later life, since normal and prudent medical procedures dictate that available tests be used to clarify the cause of the symptoms. A spinal tap, which is the "ultimate test", was not even thought of. The State merely treated Larkin for headaches and released him; it failed to exercise any professional judgment whatsoever (see Pigno v Bunim, 43 A.D.2d 718). At the very least, on December 26 Dr. Kejzlar had reason to doubt that he had the skill and experience to handle the case himself and a referral to a more skilled doctor, a neurologist or neurosurgeon, was called for (see Benson v. Dean, 232 N.Y. 52, 59; Homere v. State of New York, 48 A.D.2d 422). We find in this record that no referral was made until the unconscious Larkin was sent to the hospital on January 3.

Dr. Salamone's opinion that the failure to test or refer deviated from accepted medical practice was based on the undisputed fact that Larkin suffered from severe headaches and other associated symptoms from December 2 which continued unabated until his death. The opinion was not based on speculation and we find it to be credible (see Scott v. Brookdale Hosp. Center, 60 A.D.2d 647).

So long as a physician remains within the bounds of accepted medical practice, he is immune from liability for an error in judgment or for lack of success in his medical treatment. However, once a physician departs from accepted medical practice he is subject to malpractice liability "[h]owever good his intentions may have been" ( Pike v Honsinger, 155 N.Y. 201, 210, supra). Although liability may not be imputed from mere errors of professional medical judgment, it may be predicated upon an inadequate and careless medical examination. "Physicians are not liable for mistakes in professional judgment, provided that they do what they think best after careful examination * * * However, liability can ensue if their judgment is not based upon intelligence and thus there is a failure to exercise any professional judgment" ( Pigno v. Bunim, 43 A.D.2d 718, supra).

The overwhelming weight of credible evidence is that good and accepted medical practice required the State to test or refer Larkin. The State did neither; in fact, the State failed to exercise any professional judgment in this case.

"While the line between medical judgment and deviation from good medical practice is not easy to draw" ( Topel v. Long Is. Jewish Med. Center, 55 N.Y.2d 682, 684), the only inference that can be drawn from the proof in this case is that the failure to refer Larkin to a specialist or perform a spinal tap by December 26, at the latest, deviated from good and accepted medical practice and constitutes malpractice. According to the unrebutted testimony, Larkin could have been saved any time prior to the "blowout" and surgery provided "at least a 75 per cent chance of survival" (cf. Kallenberg v. Beth Israel Hosp., 45 A.D.2d 177, 179-180, affd 37 N.Y.2d 719). We also note that the State did not refute the charge that the medication administered on December 25 for migraine headaches by nonauthorized personnel was a drug which adversely affected Larkin's condition.

The trial court's resolution of the factual issues and its conclusion of nonliability on the part of the State is not supported by any fair interpretation of the evidence (cf. Hale v State of New York, 53 A.D.2d 1025, supra). Accordingly, judgment should be granted in favor of claimant on the issue of liability and the matter remitted to the Court of Claims on the issue of damages.

DILLON, P.J., SIMONS, HANCOCK, JR., and DOERR, JJ., concur.

Judgment unanimously reversed, on the law and facts, with costs, judgment granted in favor of claimant on the issue of liability, and matter remitted to Court of Claims on the issue of damages.