Decedent Sherri Koeppel was diagnosed with colon cancer in February 1992 and died three months later. This action was brought against the gynecologist, internist, gastroenterologist and radiologist who saw decedent at various times during the last two years of her life and who, according to plaintiff, failed to follow accepted medical practice and procedures that might have led to an earlier diagnosis of the cancer, thereby preventing her untimely death.
Upon oral argument on the motions for summary judgment by defendants Albert (the radiologist) and Silverman (the gynecologist), the IAS Court, focusing on the issue of proximate cause, granted summary judgment as to the former but denied it as to the latter. In denying summary judgment to Dr. Silver-man, the court cited the continuing relationship between Dr. Silverman and decedent, although the court conceded that the basis of the denial was "slim.” On this appeal by Dr. Silver-man, we find that there is no proximate cause between Dr. Silverman’s treatment—or alleged lack thereof—and decedent’s death and therefore reverse and grant summary judgment dismissing the complaint as to him.
During a May 8, 1990, annual gynecological check-up with Dr. Silverman, a blood test revealed that decedent’s hemoglobin level was low, an indication of low grade anemia. Dr. Silver-man instructed her to take iron supplements and return in a month for another blood test, at which time the hemoglobin level was still below normal, and Dr. Silverman referred her to her internist, Dr. Park. Following this advice, she saw Dr. Park in June 1990, at which time her hemoglobin level was within the normal range.
When decedent next saw Dr. Park in September 1990 for an *289annual physical examination, one of three stool samples showed traces of blood. Dr. Park thereupon referred decedent to a gastroenterologist, Dr. Nagler, for a colonoscopy. This procedure was performed in November 1990, after which Dr. Nagler reported by letter to Dr. Park that the colonoscopy revealed only internal hemorrhoids; Dr. Nagler noted that, due to decedent’s "very tortuous and redundant” colon, he was able to pass the instrument into but not through the ascending colon. The letter advised that, upon Dr. Nagler’s recommendation, decedent was about to undergo an upper GI and lower bowel series of X-rays; in the same letter, he recommended that guaiac testing (for blood traces) be performed periodically. The X-rays were taken by Dr. Albert in late November 1990, who reported the results, revealing no abnormalities, to both Dr. Nagler and Dr. Park by letter.
Decedent saw Dr. Silverman again on February 19, 1991, and February 6, 1992, for her annual gynecological check-ups, which showed normal hemoglobin levels each time. Dr. Silver-man was unaware of the results of decedent’s visits to Dr. Park, but it was his custom at every visit to inquire into any changes or complaints of a gastrointestinal nature (including, specifically, any change in bowel habits or any type of bleeding); decedent never had complaints of this nature, nor did she mention her visits to Dr. Park or the ensuing tests that revealed no abnormalities. In late February 1992, decedent experienced severe back pain, and hospital tests revealed that she had colon cancer that had metastasized into the bones of and around her spine, causing her death in May 1992.
According to plaintiff, who submitted an expert’s affidavit in opposition to summary judgment, Dr. Silverman’s liability lies in his departure from accepted medical practice in failing to take a complete medical history during decedent’s 1991 visit; failing to follow up with decedent after the referral to her internist; and failing to follow up with the internist after said referral. Had Dr. Silverman taken these steps, plaintiff argues, the cancer would have been diagnosed at an earlier stage, presumably early enough to be treated successfully. Plaintiff even seems to claim at one point that Dr. Silverman should have performed the guaiac testing recommended by Dr. Nagler to Dr. Park, but this duty is not alleged by the expert.
In order to establish a prima facie case of medical malpractice, a plaintiff must show not only that the doctor deviated from accepted medical practice but also that the alleged deviation proximately caused the patient’s injury (Fridovich v David, 188 AD2d 984, lv dismissed 86 NY2d 759, rearg dismissed *29086 NY2d 839). Here, even if Dr. Silverman deviated from accepted practice as alleged by plaintiff, there is no causal nexus to the failure to make an early diagnosis of the cancer. Indeed, had Dr. Silverman "followed up” with decedent or Dr. Park, he would have learned that, upon following his recommendation, decedent’s June 1990 visit to Dr. Park resulted in a normal blood test. Had he inquired into the matter during decedent’s February 1991 annual check-up, he would have learned that tests performed at her September 1990 visit to Dr. Park revealed traces of blood in one of three stool samples, leading first to the colonoscopy and then the X-ray series, neither of which revealed anything out of the ordinary. Thus, nothing would have alerted the gynecologist of any condition with which he could be charged with monitoring or which he would have pursued in such a way that the cancer would have been discovered. Moreover, since the gastroenterologist had recommended guaiac testing to the internist, Dr. Silverman would only have referred decedent back to her internist, as he had done in the first instance, upon notice of any such nongynecological problems. While Dr. Silverman continued to monitor decedent’s hemoglobin levels, as noted above, decedent never mentioned any gastrointestinal complaints.
Plaintiff insists that the submission of his expert’s opinion that Dr. Silverman’s conduct not only deviated from accepted practice but also contributed to the failure to make an early diagnosis of the cancer is sufficient to raise a question of fact to defeat summary judgment. The mere offering of such expert opinion, however, does not suffice absent a showing of "the requisite nexus between the malpractice allegedly committed * * * and the demise of plaintiffs decedent” (Ferrara v South Shore Orthopedic Assocs., 178 AD2d 364, 366). In this case, the requisite nexus is missing. We note also that, while conflicting medical opinions may serve to defeat summary judgment (see, e.g., Farkas v Saary, 191 AD2d 178, 180-181), the issue of the duty owed as between physicians, and, ultimately, to the patient, is a question of law (Lipton v Kaye, 214 AD2d 319, 322; Markley v Albany Med. Ctr. Hosp., 163 AD2d 639, 640-641).
Gitlin v Cassell (107 AD2d 636), cited by plaintiff, is inapplicable. The psychiatrist in Gitlin was denied summary judgment because, although the patient was under a physician’s care for her physical ailments, the stroke she suffered was due in part to medication prescribed by the psychiatrist where symptoms developed that contraindicated such medication. Thus, an issue of proximate cause was clearly present. Here, under the circumstances described in detail above, Dr. Silverman’s treat*291ment as decedent’s gynecologist cannot be said to have been a proximate cause of the delayed diagnosis. Accordingly, his motion for summary judgment should have been granted. Concur—Murphy, P. J., Milonas, Williams, Tom and Mazzarelli, JJ.