GARY M. GAERTNER, Judge.

Defendants, Dr. John Vanderwoude and St. Luke’s Episcopal-Presbyterian Hospital, Inc., appeal from a jury verdict awarding plaintiff $25,000.00 in damages in this medical malpractice action. On appeal, defendants argue that the trial court erred in submitting the case to the jury because: (1) plaintiff did not adduce substantial evidence showing that Dr. Vanderwoude failed to satisfy the requisite standard of care; and (2) plaintiff did not adduce substantial evidence establishing a causal connection between plaintiff’s injury and Dr. Vanderwoude’s alleged negligence. We reverse.

On December 24, 1979, between eleven and eleven-thirty p.m., plaintiff was replacing a leaking hose that ran from the washing machine to the faucet in the washer room of his home. Plaintiff took a knife from a tool shelf in the washer room to use in prying apart a connection between the hose and the faucet. Plaintiff testified that the knife was about a year old. It had originally been used in the kitchen, but the blade had become heavily rusted and it had thus been moved to the tool shelf. It had thereafter been used for cutting wires and for gardening.

While plaintiff was attempting to pry apart the connection, the knife slipped and cut through his left hand. The knife went in on the palm side of the hand at the base of the little finger and exited a little below the base of the little finger on the back of his hand. Plaintiff wrapped his bleeding hand in a kitchen towel, and his wife drove him to the emergency room at St. Luke’s Hospital West.

A nurse at the emergency room had plaintiff soak his hand in an antiseptic solution for about twenty minutes, and gave him a tetanus shot. Plaintiff then saw Dr. Vanderwoude. Plaintiff testified that he told Dr. Vanderwoude that he had run an old, rusty knife through his hand. Plaintiff’s wife testified that she heard her husband make this statement. Dr. Vander-woude denied that plaintiff told him the wound was from an old, rusty knife. The doctor testified he thought plaintiff had injured himself while preparing something in the kitchen.

Dr. Vanderwoude debrided the wound, then washed it with a saline solution and applied an anesthetic. He sutured the palm side, and covered the back side with a dressing. He told plaintiff to keep his hand elevated and dry, and to see a doctor in seven days to have the stitches removed. He also told plaintiff to inspect the wound in forty-eight hours and look for “cardinal signs of infection,” including redness, swelling, increasing pain or discharge from the wound. The doctor told plaintiff to take Tylenol for the pain, but did not prescribe any other drugs.

Plaintiff testified that he felt increasing pain and discomfort in the hand during the next two days. On December 27, 1979, he partially unwrapped his hand and noticed puffiness and a change of color. He saw Dr. Wade Hammond that same day, who immediately admitted plaintiff to St. John’s Hospital and performed surgery to remove infected tissue from plaintiff’s left hand. Plaintiff was released from the hospital on December 29, 1979. Dr. Hammond administered antibiotics while plaintiff was at the hospital, and plaintiff continued taking *173 them after his release. Although his infection responded positively to the antibiotic treatment, plaintiff has continued to experience pain and numbness in his left hand.

Plaintiff thereafter brought this malpractice action against Dr. Vanderwoude and St. Luke’s Hospital. Plaintiff alleged that Dr. Vanderwoude violated the appropriate standard of care by failing to prescribe prophylactic antibiotics, and that, as a result of such failure, plaintiff developed an infection that caused damage beyond those damages he would have suffered from the original knife wound. At trial, plaintiff offered no expert medical testimony other than reading excerpts from Dr. Vander-woude’s deposition. Dr. Vanderwoude did, however, testify in person as his own witness.

Defendants filed a motion for a directed verdict at the close of plaintiffs case and at the close of the entire case, and filed a motion for judgment after the jury returned a verdict awarding plaintiff $25,-000.00 in damages. Defendants argued on each motion that plaintiff had not made a submissible case because he had failed to establish either that defendants deviated from the established standard of care in treating plaintiffs injury, or that the infection was caused by the failure to administer antibiotics. The court overruled each of defendants’ motions. Defendants appeal the court’s decisions overruling those motions.

If the plaintiff fails to adduce substantial evidence supporting his theory of recovery, then a directed verdict should be granted and the case should not be submitted to the jury. Pinky v. Winer, 674 S.W.2d 158, 159 (Mo.App.1984). Sustaining a motion for a directed verdict is, however, a drastic action which should only be taken if all the evidence and reasonable inferences therefrom are so strongly against the plaintiff that reasonable minds cannot differ. Duke v. Gulf & Western Manufacturing Co., 660 S.W.2d 404, 409 (Mo.App.1983). We must view the evidence in a light most favorable to the plaintiff and disregard defendants’ evidence except insofar as it supports plaintiff’s case. Strebler v. Rixman, 616 S.W.2d 876, 877 (Mo.App.1981). The evidence is sufficient to make a submissible case if a reasonable probability that the defendants were negligent may be fairly inferred from it. Rauschelbach v. Benincasa, 372 S.W.2d 120, 125 (Mo.1963). If, however, viewing the evidence in a light most favorable to the plaintiff, the question of negligence may only be determined by resort to conjecture and surmise, then the plaintiff has failed to make a submissible case. Yoos v. Jewish Hospital of St. Louis, 645 S.W.2d 177, 183 (Mo.App.1982).

To prove a prima facie case of medical malpractice, plaintiff must establish three elements: (1) Proof that the act or omission failed to meet the requisite standard of medical care; (2) Proof that the act or omission was performed negligently; and (3) A causal connection between the act or omission and the claimed injury sustained by the plaintiff. Pinky, supra at 160.

In their first allegation of error, defendants argue that the trial court erred in submitting the case to the jury because plaintiff did not adduce substantial evidence proving that Dr. Vanderwoude failed to satisfy the requisite standard of medical care. To make a submissible case on this issue, plaintiff was required to establish that Dr. Vanderwoude failed to exercise that degree of skill and learning ordinarily exercised under the same or similar circumstances by members of his profession. Kinser v. Elkadi, 674 S.W.2d 226, 229 (Mo.App.1984). The general rule requires that a plaintiff put forth expert medical testimony establishing the appropriate standard of care. Hart v. Steele, 416 S.W.2d 927, 931 (Mo.1967). Expert testimony is not required, however, if the defendant’s own testimony establishes the standard of care. Bateman v. Rosenberg, 525 S.W.2d 753, 757 (Mo.App.1975); Pinky, supra at 160.

Viewing the evidence in a light most favorable to plaintiff, we find that plaintiff made a submissible case regarding the req *174 uisite standard of care. Although plaintiff put forth no independent expert medical testimony, Dr. Vanderwoude’s own testimony — including his deposition excerpts— established the standard of care. At his deposition, Dr. Vanderwoude was asked what factors he considered in determining whether to use an antibiotic in treating a laceration or puncture wound. 1 The doctor said he considered whether the wound was clean or dirty. A “dirty wound,” according to the doctor, included one that had been inflicted by a dirty instrument. A “dirty instrument” included a knife that was rusty or had been in contact with soil or dirt. The following exchange then took place:

[Q] “Doctor, you have told us the reasons why physicians, medical science, might deem it appropriate to administer an antibiotic. Could you tell us any reasons why a physician or a medical practitioner might deem it inappropriate to administer an antibiotic to a person presenting himself for treatment with a laceration or puncture wound?

[A] “The patient maybe [sic] allergic to antibiotics.

[Q] “Okay. Obviously an allergic reaction. Are there any others?

[A] “I am not quite sure what you are reaching for.

[Q] “Well, you told us one reason a physician wouldn’t prescribe an antibiotic, in the situation you just described as an allergic reaction. Are there any other reasons he wouldn’t prescribe them that you can think of?

[A] “If he thinks the wound doesn’t merit it.

[Q] “And what do you mean by not meriting it?

[A] “If it’s a clean wound.

[Q] “How do you determine if it’s a clean wound?

[A] “By the appearance of the wound, and from the mechanism of injury, as we talked about before.

[Q] “By ‘mechanism of injury’ you mean what caused the puncture or the laceration?

[A] “Right.

[Q] “Are there any other reasons that you would not administer an antibiotic? [A] “I can’t say offhand. Those are the primary.

[Q] “Those are the only ones you can think of; is that correct?

[A] “Right.”

Defendants argue that Dr. Vander-woude’s testimony pertained only to his own personal standard of care, and did not establish any general standard applicable to the medical community at large. We find, however, that the testimony quoted above, viewed in a light most favorable to plaintiff, establishes a general standard of care. 2 That standard requires administering antibiotics to a patient with a dirty wound. The jury could, therefore, have found that Dr. Vanderwoude violated the standard of care if it determined: (1) that plaintiff had a dirty wound; (2) that Dr. Vanderwoude knew plaintiff had a dirty wound; and (3) that Dr. Vanderwoude failed to administer antibiotics to plaintiff.

Plaintiff presented substantial evidence to support each of these findings. Plaintiff’s wound was “dirty” because it was inflicted with a “dirty instrument,” namely a rusty knife. Dr. Vanderwoude knew plaintiff’s wound was dirty because plaintiff and his wife each testified that plaintiff informed the doctor at the emergency room that plaintiff had cut himself with a rusty *175 knife. 3 Finally, it is undisputed that the doctor did not prescribe, antibiotics for plaintiff. We hold, therefore, that the trial court properly submitted to the jury the issue of whether Dr. Yanderwoude violated the appropriate standard of care.

In their second allegation of error, defendants argue that the trial court erred in submitting the case to the jury because plaintiff did not adduce substantial evidence establishing a causal connection between Dr. Vanderwoude’s failure to administer antibiotics and plaintiff’s infection. The plaintiff in a medical malpractice action bears the burden of proving that the defendant’s negligence caused plaintiff’s injury. Mercer v. Thornton, 646 S.W.2d 375, 376 (Mo.App.1983). The usual test for a causal connection between a defendant’s negligence and a plaintiff’s injury is whether the facts in evidence show that the injury would not have been sustained but for the negligence. Robbins v. Jewish Hospital of St. Louis, 663 S.W.2d 341, 345 (Mo.App.1983). 4 Although expert testimony is generally required to prove causation, Hart v. Steele, supra at 931, a plaintiff may prove causation by circumstantial as well as direct evidence. Kappel v. Slickman, 401 S.W.2d 451, 453 (Mo.1966). Plaintiff must only adduce substantial evidence showing that the injury is a natural and probable consequence of the negligent act or omission. If the logical conclusion from the evidence is that if certain things were properly done certain results would not have occurred, and such results did occur, the question of causation is sufficient to go to the jury. Steele v. Woods, 327 S.W.2d 187, 195 (Mo.1959).

In the case before us, plaintiff contends that Dr. Vanderwoude’s failure to prophylactically prescribe antibiotics caused plaintiff’s wound infection. Plaintiff argues that the evidence adduced at trial, together with reasonable inferences therefrom, establishes a “logical conclusion” of causation. Plaintiff's logic is essentially as follows: Dr. Vanderwoude stated that one of the reasons antibiotics are generally prescribed is to prevent infection; the doctor did not prescribe antibiotics for plaintiff; forty-eight hours later, plaintiff developed an infection; the infection responded well to antibiotic therapy; therefore, the doctor’s failure to prescribe antibiotics caused plaintiff’s injury. Plaintiff offered no direct evidence to support this theory of causation.

We find the asserted causal connection insufficient to make a submissible case. In Kappel v. Slickman, supra, our Supreme Court faced a similar issue. Seven days after undergoing a hysterectomy, plaintiff developed an infection (later diagnosed as cellulitis) in the incision area. Plaintiff was released from the hospital a few days later, but had to be rehospitalized twenty-four days thereafter. Plaintiff sued the two physicians who had treated her, claiming that their failure to administer antibiotics during her initial hospitalization had caused the infection which required rehospitalization. 5 At trial, plaintiff offered no expert medical testimony other than reading into evidence excerpts from the deposition of one of the defendant doctors. The doctor stated at his deposition that during plaintiff’s first hospitalization he prescribed (although plaintiff did not receive) Achromy-cin(V), a broad spectrum antibiotic, “as a protection against the spread of infection.” Id. at 452. The doctor further stated that

because of its broad spectrum capabilities, [Achromycin(V) ] should have been able to combat whatever organism was involved; that, at that time, his best *176 judgment was that an antibiotic drug should be administered, and that Achro-mycin(V) was the drug which he chose.

Id. The court held this evidence insufficient to make a submissible case on the issue of causation. The jury could not reasonably infer causation

because of the lack of essential proof that omission of the Achromycin(V) on the first hospitalization caused the continuance or recurrence of the infection which necessitated the second hospitalization. True, from Doctor Newman’s testimony the inference may be drawn that his object in prescribing Achromy-cin(Y) was to prevent cellulitis. However, there is no evidence from which the jury could find that the treatment would likely have produced such result. Certainly the propensities of Aehromycin(V) are not such a matter of common knowledge that jurors of ordinary intelligence and experience would know the results reasonably to be anticipated from treatment with such drug. There can be no doubt that infections do exist which resist even broad spectrum antibiotic treatment.

Id. at 454.

In the case before us, plaintiff contends that the jury was able to infer causation from the circumstantial evidence presented. We agree, however, with the Kappel court’s finding that the therapeutic propensities of antibiotics are not such that an average juror can know the results reasonably to be anticipated from them. Accordingly, the mere fact that plaintiff was not administered antibiotics and that he developed an infection two days later is insufficient to establish a causal connection. Causation requires not only proximity in time, but also a showing that plaintiff would not have been injured but for defendant’s negligence. Even when viewed in a light most favorable to plaintiff, the evidence presented fails to raise a reasonable inference that plaintiff’s infection would not have developed but for Dr. Vanderwoude’s failure to administer antibiotics. 6

Plaintiff has made no effort to identify a specific antibiotic which Dr. Vanderwoude could or should have prescribed to prevent the infection. In Robinson v. Gatti, 115 Ohio App. 173, 184 N.E.2d 509 (1961), plaintiff claimed that his doctor’s failure to pro-phylactically administer antibiotics had caused plaintiff’s lung infection. The court held that plaintiff had failed to prove causation:

It is to be noted that the record discloses no evidence of any specific antibiotic to be administered for treatment of the type of injury involved in this case. All testimony relates to the general term “antibiotic” with no further scientific inquiry.

Id. 184 N.E.2d at 513. Given the wide variety of antibiotics currently available, the court concluded that “any decision based solely on opinion evidence of the possible results from the use of such a general class of treatment as ‘the use of antibiotics’ should not be the basis on which to predicate a judgment and return damages to the plaintiff.” Id.

In arguing that Dr. Vanderwoude could have prescribed some antibiotic that would have prevented his infection, plaintiff relies principally upon two facts: (1) antibiotics are administered prophylactically for the purpose of preventing infection; *177 and (2) plaintiff’s infection responded well to antibiotic therapy administered by Dr. Hammond. This evidence is insufficient to raise an inference that antibiotic treatment would have been effective. Agreeing with the Robinson court’s discussion of the nature of antibiotics, we do not find that Dr. Yanderwoude could have known, at the time he treated plaintiff, which antibiotic— if any — would have prevented an infection from developing in plaintiff’s wound. Furthermore, the mere fact that Dr. Hammond effectively administered antibiotic therapy subsequent to the development of the infection does not tend to prove that prophylactic administration of some unidentified antibiotic would have been effective. Submitting the case to the jury based upon this evidence would inevitably lead the jurors into the forbidden realm of conjecture and surmise.

Having concluded that plaintiff presented no substantial evidence showing that Dr. Vanderwoude’s failure to administer antibiotics caused plaintiff’s infection, we hold that the trial court improperly submitted the issue of causation to the jury. Accordingly, we reverse the trial court’s order denying defendants’ motions for directed verdict and for judgment.

Judgment reversed.

KAROHL, P.J., and SIMON, J., concur.

Notes

. All of the deposition testimony referred to herein was read into evidence at trial.

. Defendants observe that Dr. Vanderwoude testified specifically at trial that no general standard of care exists in this regard. Defendants thus argue that, even assuming the doctor referred to a general standard of care at his deposition, such testimony is unpersuasive in light of his testimony at trial. Our scope of review limits us, however, to determining whether plaintiff presented substantial evidence establishing the standard of care. It is the jury's province to weigh conflicting evidence. Therefore, notwithstanding the doctor’s testimony at trial, his deposition testimony constitutes substantial evidence of the standard of care, and a submissible case has been made.

. Dr. Vanderwoude’s testimony that he was not so informed by plaintiff does not affect our determination that plaintiff presented substantial evidence to support such a finding by the jury-

. In addition to satisfying this ‘Rut for" test, generally known as causation in fact, plaintiff must also establish that defendant’s negligence was a “proximate cause” of plaintiffs injury. Bass v. Bi-State Development Agency, 661 S.W.2d 609, 613 (Mo.App.1983).

.The evidence conflicted over whether or not plaintiff in fact received antibiotics during her initial hospitalization. For purposes of its opinion, however, the court assumed that plaintiff had not received antibiotics.

. A Florida court has reached a similar result. In Cude v. Deal, 234 So.2d 711 (Fla.App.1970), plaintiff was hospitalized with a foot injury following an industrial accident. Plaintiff was released the following day. The defendant physician did not prescribe antibiotics. Two days later, plaintiff was rehospitalized, having developed gangrene in the injured foot. Two of his toes were subsequently amputated. Plaintiff sued for malpractice, alleging, inter alia, that defendant's failure to prescribe antibiotics had caused plaintiff’s rehospitalization and amputation. Although plaintiff established that the failure to prescribe antibiotics violated the established standard of care, the court nevertheless held that plaintiff had failed to make a submissible case:

We have carefully analyzed the record here and find there is no proof relating to the failure to prescribe antibiotics and the first hospital discharge, on the one hand, with the second hospitalization and surgical amputation of the left foot, on the other hand. This was necessary to make out a case for the jury on the issue of medical malpractice.

Id. at 712.