In 1972, plaintiff consulted her family doctor in Elizabeth City about discharges from her breasts. A biopsy by wedge resection was performed at that time, but plaintiff continued to have soreness and tenderness in her breasts along with a continued discharge. The family doctor referred plaintiff to Georgiade, a physician at Duke Medical Center.
On 16 April 1975, Georgiade examined plaintiff, diagnosed her condition as due to fibrocystic disease, and prescribed treatment by a surgical procedure known as a bilateral subcutaneous mastectomy. Georgiade explained to plaintiff that the procedure “was the only way [she] would ever get rid of [her] problem.” Surgery was performed on 20 May 1975 at which time the diseased tissue was removed. A prosthesis was inserted to replace tissue removed from each breast.
Plaintiff left the hospital on 27 May 1975 and “was doing pretty good.” She eventually returned to work as a hairdresser but found that she could no longer manage all of the physical requirements of dressing hair. Plaintiffs condition improved until December when she began “having a lot of pain again and the left prosthesis, or the left breast had gotten very hard.”
Georgiade saw plaintiff again on 2 March 1976. He diagnosed plaintiffs malady as a capsular formation, or scar tissue buildup around the left prosthesis, which should be surgically released. The capsular release was scheduled for 17 March 1976, with plaintiff as an outpatient. Quillen assisted Georgiade in the procedure by anesthetizing plaintiffs breast area. Fourteen injections were made around the circumference of the breast, which was opened at the old incision. The prosthesis was removed, then the scar tissue, and the prosthesis was replaced. Plaintiff checked out of the hospital after the surgery and went with her husband to a local motel.
At about 8:00 p.m. the same evening, plaintiff awakened thinking her bandage “was cutting [her] breath off.” Plaintiffs breathing difficulty grew progressively worse, and her husband carried her to the emergency room at Duke Medical Center. Plaintiffs lungs were X-rayed, and it was determined that greater than 50°/o of both her lungs had collapsed. Emergency room physi*486cians reinflated plaintiff’s lungs without anesthesia because “[t]ime was a great element.” Following the emergency room treatment, plaintiff “point-blank asked Dr. Quillen did he puncture my lungs with those needles when he gave me the needles the day of the surgery. He told me that they try to be very careful to avoid those things but sometimes they happen anyway.” Plaintiff eventually was discharged from the hospital on 24 March 1976.
Plaintiff returned home to Elizabeth City but continued to have trouble with her breasts and her chest. A few months later the right prosthesis began exhibiting symptoms similar to that of the left breast prior to the capsular release. Plaintiff refused to see Georgiade again and sought treatment at Durham County General Hospital, where additional surgery on both breasts was performed.
Plaintiff s first argument assigns error in the trial judge’s exclusion of her expert witness’s opinion on the nature of the cause of the collapsed lungs. During the voir dire examination of Dr. Gerald Golden, tendered as plaintiffs expert witness, plaintiff asked two hypothetical questions. First, Golden was asked, based on certain assumed facts, “do you have an opinion within a reasonable medical certainty as to the cause of the bilateral pneumothoracies [collapsed lungs]?” Defendants’ objection was sustained, but the record shows that the answer would have been, “It came from a needle puncture of the lung.” Second, Golden was asked, based on certain assumed facts, “do you have an opinion as to whether the causing of the bilateral pneumothoracies was a deviation from standard practice?” Defendants’ objection again was sustained, but the record shows that Golden’s answer would have been, “It is a deviation from accepted standard of care.”
Plaintiff offered further “qualifying questions” which the trial judge deemed to relate back to the first hypothetical question. The questions to Golden sought to elicit further evidence that plaintiff’s lung collapse was not spontaneous. Under a continuing, sustained objection, Golden stated that “the likelihood of anybody having a spontaneous pneumothorax bilaterally is almost nil.” Nevertheless, defendants’ objection to the first hypothetical question again was sustained.
*487Later, before the jury, plaintiff asked Golden the following hypothetical question:
Please assume that a 34 year old female suffering from fibrocystic disease of both breasts, assume that the patient on or about May 20, 1975 had a bilateral subcutaneous mastectomy with prosthesis being implanted and that subsequent to that bilateral subcutaneous mastectomy being performed that it was determined that capsular scarring had taken place and that replacing of the breast prosthesis should be performed, and that capsular release surgery was performed March 17, 1976. Doctor, I ask you to assume further that in the performance of this surgery, that is the replacement of the breast prostheses and the capsular release the anesthesia was injected as follows: assume that approximately 14 injections were made around the circumference of each breast injecting into the patient .5% Xylocaine, and assume further that after the surgery was performed the patient left the hospital and went to a local motel and then approximately 2 hours later the patient began to experience shortness of breath and chest pain, and assume further that the patient was immediately rushed to the emergency room of the hospital and that there it was discovered that she had bilateral pneumothoracies greater than 50% in each lung. Making those assumptions, do you have an opinion as to whether the surgical procedure performed in 1976 could or might have caused the bilateral pneumothoracies?
Defendants’ objection to this question was overruled, and Golden was permitted to answer that his opinion was, “Yes.” However, defendants’ objection to the following question was sustained: “Doctor, I ask you in the causing of the bilateral pneumothoracies you’ve just testified about, if causing that there was a deviation from standard medical practice?” Golden’s answer for the record was, “Yes, it was (WHISPERED).”
Plaintiff argues that the exclusion of the above answers and of Golden’s voir dire testimony concerning the possibility of a spontaneous cause of plaintiff’s collapsed lungs are the bases for his assignment of error. However, we note that although the answer to the first hypothetical question was excluded, the same *488question was asked again, and the answer allowed, apparently upon plaintiffs rephrasing in the emphasized portion of the question quoted supra. We fail to see how plaintiff was prejudiced by the trial judge’s first ruling under these circumstances.
Defendants cross-assign error to the trial judge’s ruling, which allowed Golden to answer the hypothetical question quoted supra, on the ground “that without the crucial fact of entry into the lung by a needle, there was an insufficient predicate so as to allow the basing of an opinion thereon.” We therefore must discuss the propriety of the judge’s ruling on the hypothetical question quoted supra from defendants’ perspective. In addition, we will discuss plaintiff’s exceptions to the judge’s rulings on the hypothetical questions concerning whether the cause of plaintiff’s collapsed lungs was a deviation from standard medical practice.
Hypothetical Questions As To The Cause Of Plaintiff’s Collapsed Lungs
 Defendants argue that the hypothetical question quoted supra provides no basis, except for speculation and conjecture, from which a negligent act can be shown.
When an expert witness testifies as to the facts based upon his personal knowledge, he may testify directly as to his opinion, . . . and when the facts are not within the knowledge of the witness himself, the opinion of an expert must be based upon facts supported by evidence stated in a proper hypothetical question.ii the expert witness has personal knowledge of some of the facts but not all, a combination of these two methods may be employed.
In asking a hypothetical question, it is customary to incorporate in the question the relevant facts in evidence which counsel hopes will be accepted as true by the jury and to ask the witness his opinion based on such facts, if the jury shall believe them to be facts.
... In framing such a question, only such facts as are in evidence or such as the jury will be justified in inferring from the evidence should be included.
*489 Taylor v. Boger, 289 N.C. 560, 564-65, 223 S.E. 2d 350, 353 (1976) (emphasis added). See also Todd v. Watts, 269 N.C. 417, 152 S.E. 2d 448 (1967). In light of these general principles, the question is whether an expert witness may make additional inferences from the facts assumed in a hypothetical question to form the basis of his opinion.
The function of an expert witness is to utilize his superior skill in a particular field to form an opinion which will be helpful to the jury. Such an opinion is admissible “where the witness is better qualified than the jury to draw appropriate inferences from the facts.” 1 Stansbury’s N.C. Evidence (Brandis rev. 1973) § 132, p. 425. “An inference is nothing more than a permissible deduction from the evidence . . ..” Cogdell v. Wilmington & Weldon Railroad Co., 132 N.C. 852, 854, 44 S.E. 618, 619 (1903). Thus, a medical expert must use his “education and training . . . to deal with and express [his] opinion as to what ills and the causes that constantly threaten and affect humanity.” Shaw v. National Handle Co., 188 N.C. 222, 232, 124 S.E. 325, 330 (1924).
In Shaw, a medical expert opined that decedent’s cause of death was due to “gas poisoning, monoxide poisoning” based merely upon a hypothesis of the surrounding external conditions as he observed them. Id. at 231, 124 S.E. at 330. Even though no autopsy was performed, and thereby no direct evidence of the cause of death received, our Supreme Court found that the expert’s opinion was competent since it came “from one who has had personal observation of the facts, and from practical training and experience is qualified to give an opinion which is likely to aid the jury to a correct conclusion . . ..” Id. at 232, 124 S.E. at 330, quoting Davenport v. Railroad Cos., 148 N.C. 287, 294-95, 62 S.E. 431, 433 (1908).
Similarly, in Hargett v. Jefferson Standard Life Insurance Co., 258 N.C. 10, 128 S.E. 2d 26 (1962), admitted evidence showed that plaintiff’s decedent had a discoloration at the end of a finger with a pierce in the middle. Decedent was in intense pain and died minutes after losing consciousness. A medical expert who had not personally attended decedent, but who had great experience in dealing with the effects of insect bites on humans, testified to decedent’s cause of death. The expert opined, based merely upon the “external evidence,” that “death resulted from *490an insect sting.” Id. at 12, 16, 128 S.E. 2d at 28, 31. The Supreme Court sustained the expert’s answer, stating that the above facts were “sufficient predicate for the conclusion reached by the witness.” Id. at 16, 128 S.E. 2d at 31.
Shaw and Hargett support the proposition that an expert witness may employ inferences from facts assumed in a hypothetical question to form an opinion. Defendants, however, have sought to distinguish these cases from the case sub judice on the ground that Golden had no personal observation of plaintiffs condition. It is not necessary that an expert’s opinion be based “solely from matters personally observed. . . . [A]n expert witness has wide latitude in gathering information and may base his opinion on evidence not otherwise admissible.” State v. DeGregory, 285 N.C. 122, 132, 203 S.E. 2d 794, 801 (1974) (emphasis original). Although Golden did not personally examine plaintiff, he testified as follows: “I have reviewed the Duke University Medical Center medical records of Judy Simons for 1975 and 1976. I have also reviewed the Plaintiff’s Exhibit 9, the Durham County General Hospital record and I have also reviewed Plaintiff’s Exhibits 10, 11 and 12 from the Albemarle record.” We find that the expert’s knowledge gained by a review of plaintiff’s medical records is a sufficient basis on which to form an opinion. See Potts v. Howser, 274 N.C. 49, 161 S.E. 2d 737 (1968).
Even so, the analysis of the Third Circuit in Friedman v. General Motors Corp., 411 F. 2d 533 (3d Cir. 1969), is particularly determinative of defendants’ argument. In Friedman, plaintiffs thumb was injured and had to be amputated because of alleged defects in the design and construction of her washing machine. Defendant corporation’s medical expert gave opinion evidence as to how plaintiff Friedman could have been injured. He testified, in response to a question asking for an opinion as to the cause of the accident, “[I]f Mrs. Friedman had opened the lid to the washing machine during the middle of the spin cycle and then reached in and put her hand on the lid safety switch, this would have started to spin again.” Id. at 535. Plaintiffs there argued, as do defendants here, that the opinion was mere “conjecture” and not based on any evidence. The Third Circuit reasoned as follows:
While there was no direct evidence that Mrs. Friedman had put her hand on the lid safety switch, the evidence shows that she could have done so however inadvertently. We can*491not say that [the witness’s] statement was mere conjecture ... for it is the opinion of an expert based on material facts which are part of the record.
Id: at 535-36, n. 3 (emphasis added). In the case sub judice there was evidence that approximately fourteen injections were made around plaintiffs breast in preparation for the capsular release procedure. Approximately three hours after she awakened from surgery, plaintiff began exhibiting the first symptoms of her collapsed lungs. Upon these facts assumed from the evidence, Golden concluded that the surgical procedure caused the collapsed lungs. Since that opinion clearly is based upon material facts in the record, it is not fatal to the hypothesis that Golden must have inferred, or deduced, entry of the needle into plaintiffs lungs to effect the collapse. See generally Lee v. Capital Tire Co., 40 N.C. App. 150, 252 S.E. 2d 252, disc. rev. denied, 297 N.C. 454, 256 S.E. 2d 807 (1979).
Thus, the rule is that “[d]irect testimony supporting the fact assumed is not required. It is sufficient if it is fairly inferable from circumstances proved.” E. Cleary, McCormick’s Handbook of the Law of Evidence (2d ed. 1972) § 14, p. 33. Defendants’ cross-assignment of error is overruled.2
Questions As To The Deviation From Standard Medical Practice
Plaintiff contends that the trial judge’s exclusion of Golden’s answers to the questions of whether the cause of plaintiff’s collapsed lungs was a deviation from standard medical practice was “based upon the court’s decision to exclude Dr. Golden’s testimony concerning causation, or its rationale.” If plaintiffs evaluation of the trial judge’s rulings is correct, our analysis of the previous issue would sustain her exceptions. However, we find a more plausible basis for the rulings in plaintiff s phrasing of the questions, quoted in pertinent part supra, the essence of which is, “do you have an opinion as to whether the causing of the bilateral pneumothoracies was a deviation from standard practice.” Plaintiff made two possible omissions in the phrasing of *492these questions: first, plaintiff did not employ the “could or might” language in her questions; second, plaintiff did not ask her expert witness whether the causing of her collapsed lungs was a deviation from standard medical practice in Durham, North Carolina or in similar communities in 1975 and 1976. If these omissions were the bases for the sustained objections, we must reverse the trial judge’s rulings for the following reasons.
 In plaintiffs first question, couched hypothetically, her failure to employ the “could or might” language is not fatal to the admission of Golden’s answer. “[I]f the expert has a positive opinion on the subject, he should be allowed to express it without using the ‘could’ or ‘might’ formula.” Taylor v. Boger, supra at 565, 223 S.E. 2d at 353 (emphasis original). Since the question relates to the cause and effect of plaintiff’s injury, adoption of the “could or might” language would force Golden into a more speculative answer when the nature of the question required, and the witness would have, a more positive opinion. See Mann v. Virginia Dare Transportation Co. and Tillett v. Virginia Dare Transportation Co., 283 N.C. 734, 198 S.E. 2d 558 (1973). This would be “patently unjust.” 1 Stansbury’s N.C. Evidence (Brandis rev. 1973) § 137, p. 455, n. 97. The same reasoning must apply to plaintiff’s second question, although it was not couched in hypothetical form.3
 The phrasing of both of plaintiffs questions arguably did not establish adequately the standard of care' upon which Golden would have been permitted to testify to defendants’ deviation therefrom. Thus, the question is whether plaintiff had to ask her expert whether the causing of her collapsed lungs was a deviation from standard medical practice in Durham, North Carolina or in similar communities in 1975 and 1976. Under the circumstances of the case sub judice, we find that the standard of care was adequately established.
*493G.S. 90-21.12 codified the standard of health care necessary for the trier of the facts to establish a defendant’s liability. The statute states, in pertinent part, that
the defendant shall not be liable for the payment of damages unless the trier of the facts is satisfied by the greater weight of the evidence that the care of such health care provider was not in accordance with the standards of practice among members of the same health care profession with similar training and experience situated in the same or similar communities at the time of the alleged act giving rise to the cause of action.
G.S. 90-21.12. This standard of care is the same as that developed by our case law. Thompson v. Lockert, 34 N.C. App. 1, 237 S.E. 2d 259, disc. rev. denied, 293 N.C. 593, 239 S.E. 2d 264 (1977). We find the following testimony of Golden in the record:
Q. Doctor, are you familiar with the amount of skill and knowledge of the average plastic surgeon in the Town of Durham, North Carolina or in similar communities in 1975 and 1976?
A. Yes, sir.
Q. Are you familiar with the amount of care ordinarily exercised by the average plastic surgeon in the Town of Durham or other similar communities in 1975 and 1976?
A. Yes, sir.
Q. Doctor, how did you acquire that familiarity?
A. Well, I did my education at two medical centers located in communities of roughly the same size, Winston-Salem for four years, and I spent seven years in Charlot-tesville at the University of Virginia, and in addition, my practice — we’re only an hour from Washington and we have a large metropolitan population within Martinsburg.
*494Q. Are the standards of board certified surgeons the same throughout the country?
A. Yes, sir.
Golden also testified that he is a plastic surgeon. This testimony is sufficient under G.S. 90-21.12 to establish the standard of health care upon which plaintiffs expert could testify to defendants’ deviation therefrom. It is of no material consequence that plaintiffs questions did not include the additional language indicated.
We therefore conclude that if these omissions from plaintiffs questions were the bases for defendants’ sustained objections, they are not supportive of the trial judge’s rulings. Careful examination of the questions reveals error in excluding Golden’s answers.
 Defendants cross-assign error to the trial judge’s ruling which allowed Golden to testify as plaintiffs expert witness on the ground that he was not sufficiently familiar with the standard of medical practice in Durham, North Carolina, or in similar communities in 1975 and 1976. Defendants contend, inter alia, that because Golden had not completed his training as a plastic surgeon in 1975 and 1976, nor had he practiced in this State since 1967, and because he had been practicing in hospitals in communities smaller than that of Durham, he could not be familiar with the appropriate standard of care.
G.S. 90-21.12 mandates the establishment of a standard of health care by “members of the same health care profession with similar training and experience situated in the same or similar communities at the time of the alleged act giving rise to the cause of action.” Golden’s testimony, quoted in part supra, shows that he was trained and experienced in the fields of general and plastic surgery in hospitals and communities similar to Duke Medical Center and Durham in 1975 and 1976. This is sufficient to qualify Golden as an expert witness. See Lowery v. Newton, 53 N.C. App. 234, 278 S.E. 2d 566 (1981); Whitehurst v. Boehm, 41 N.C. App. 670, 255 S.E. 2d 761 (1979). It would be unduly restrictive under G.S. 90-21.12 to require an expert to have knowledge of the standard of care in a similar community at the time of the *495alleged act only by having practiced in the particular field at that time. Cf. Dickens v. Everhart, 284 N.C. 95, 199 S.E. 2d 440 (1973). This cross-assignment of error is overruled.
Plaintiffs second and third arguments assign error to the trial judge’s ruling excluding her testimony that she would not have proceeded with the 1975 surgery had she been fully informed of the possible complications, and the trial judge’s instruction that “[t]he plaintiff must prove by the greater weight of the evidence that a reasonably prudent person in the plaintiffs condition would reasonably have been expected to withhold her consent to the operation if she had been informed of the circumstances and risks of the procedure . . ..” (Emphasis added.)
 Generally, plaintiff contends that the proper standard of proof to apply in this case is the “ ‘true test of causation’ which focused on the motives of the particular patient/plaintiff,” because G.S. 90-21.13 does not apply to actions arising prior to its effective date. We do not agree. G.S. 90-21.13, govening informed consent to health care treatment or procedure, utilizes a “reasonable person” test to show a valid consent. The statute became effective on 1 July 1976 and expressly did not apply to cases pending on that date. See Thompson v. Lockert, supra; see also 1975 N.C. Sess. Laws (2d Sess. 1976) c. 977, §§ 8, 10. We agree with defendants’ statement that “[t]here is no provision in the statute ... regarding the applicability of the act to causes of action which allegedly arose before the effective date of the statute but in which no litigation was pending on the effective date.” G.S. 90-21.13 therefore must apply to the present litigation since it commenced after the effective date, 1 July 1976. The trial judge’s instruction was not in error.
 In sustaining defendants’ objections to plaintiffs question as to whether she would have proceeded with the surgery had she known the possible complications, the trial judge relied upon Watson v. Clutts, 262 N.C. 153, 136 S.E. 2d 617 (1964). Plaintiff contends that the exclusion of her answer is prejudicial error because it removes from the jury’s consideration “the testimony *496of the only person in the courtroom who has actually experienced a situation substantially like the hypothetical one which they must consider . . However, we find Watson to be controlling in this instance.
Plaintiff Watson attempted to testify that if her doctor had advised her that the operation might paralyze her vocal cords she would have withdrawn her consent. The exclusion of that plaintiffs testimony was sustained because she could not “change the decision” after the operation. Id. at 161, 136 S.E. 2d at 622. The same rationale applies in the case sub judice. In essence, the probative value of plaintiff’s testimony is so weak that to receive it would only confuse the jury. See 1 Stansbury’s N.C. Evidence (Brandis rev. 1973) § 77, pp. 234-36. The trial judge’s ruling is sustained.
 Plaintiffs Assignment of Error No. 2 is grounded upon the trial judge’s refusal to admit into evidence the substance of a conversation she heard between her husband and Dr. Walter Wolfe, a thoracic surgeon treating plaintiff’s collapsed lungs. When plaintiff proposed to testify to the substance of the conversation, a voir dire examination of plaintiff was conducted. She testified that Wolfe is a member of Private Diagnostic Clinic and that she was charged for the services of Wolfe by Private Diagnostic Clinic. This testimony and the following answer were not admitted before the jury:
A. Dr. Wolfe came upon request and when he came in Howard asked him was there any goddamn body in that hospital that knew how to do anything right. And Dr. Wolfe gave it back to him just as good as he did. Dr. Wolfe told him, he said: I’m not the one that punched those goddamn holes in your wife’s lungs. If you want to get upset with anybody go find him. He said: all I’ve done is try to save your wife’s life.
Plaintiffs husband’s testimony concerning this conversation with Wolfe also was excluded. Plaintiff argues that these rulings were in error because Wolfe, a partner in Private Diagnostic Clinic and thereby a partner of defendant Georgiade, could make admissions admissible against the named defendants in exception to the hearsay rule. We agree.
*497The extrajudicial declarations of an alleged partner cannot be used (except as against himself) to prove the existence of the partnership [footnote omitted] or that the declarant was engaged in the firm’s business at the time, [footnote omitted] ■ But if these facts are independently established, his declarations within the scope of his authority as a partner are admissible against other members of the partnership as well as against himself, in an action between the partnership and an outsider, [footnote omitted]
2 Stansbury’s N.C. Evidence (Brandis rev. 1973) § 170, pp. 20-21 (emphasis added). Thus, the question is whether the partnership of Wolfe in Private Diagnostic Clinic was sufficiently proven by plaintiffs voir dire testimony to require the admission of his statements.
In Tapp v. Dibrell, 134 N.C. 546, 548, 47 S.E. 51, 51 (1904), plaintiff Tapp testified that defendants Carrington and Dibrell were partners, “that he talked with Carrington in August, 1902, and said that he wanted to settle the claim but he would be liable on garnishment.” Defendants objected to the admission of this testimony, but the Supreme Court affirmed. “[The testimony] was certainly admissible against him, and if there was a partnership, against his co-partner. It was not offered to prove a partnership. The witness had testified to the partnership. While this was not conclusive it was a sufficient basis to admit the declaration of Carrington.” Id. Plaintiff’s testimony in the case sub judice likewise was a sufficient, albeit inconclusive, basis to admit Wolfe’s statements pursuant to the hearsay rule exception quoted supra. The trial judge’s rulings therefore were improper.
Defendants Quillen and Duke University were granted a directed verdict at the conclusion of plaintiff’s evidence. The question raised by a directed verdict motion is whether the evidence is sufficient to go to the jury. Rappaport v. Days Inn of America, Inc., 296 N.C. 382, 250 S.E. 2d 245 (1979); Kelly v. International Harvester Co., 278 N.C. 153, 179 S.E. 2d 396 (1971).
In passing upon such motion at close of plaintiff’s evidence in a jury case, as here, the evidence must be taken as true, considered in the light most favorable to plaintiffs, and may be *498granted only if, as a matter of law, the evidence is insufficient to justify a verdict for the plaintiffs.
Huff v. Thornton, 287 N.C. 1, 10, 213 S.E. 2d 198, 205 (1975), quoting Dickinson v. Pake, 284 N.C. 576, 583, 201 S.E. 2d 897, 902 (1974). Accord, Younts v. State Farm Mutual Automobile Insurance Co., 281 N.C. 582, 189 S.E. 2d 137 (1972).
In light of our rulings upon the critical evidentiary issues discussed at length supra, we find that the evidence improperly excluded would have been sufficient to withstand the directed verdict motions of defendants Quillen and Duke University. For this reason, the granting of those motions was reversible error.
We have carefully examined plaintiffs remaining assignments of error and defendants’ remaining cross-assignments of error and find them to be without merit, not warranting further discussion in this opinion.
For errors found in the trial below, we make the following disposition:
Reverse as to defendants Quillen and Duke University.
New trial as to defendants Georgiade and Private Diagnostic Clinic.
Judges VAUGHN and WHICHARD concur.