Joe Lewis sued podiatrist Charles J. Sidlow and North Georgia Podiatric Medicine & Surgery, P.C., alleging that Sidlow committed malpractice by failing to diagnose a significant foot injury that led to the amputation of Lewis’s leg. After the parties completed discovery, Sidlow moved for summary judgment, arguing that the two-year statute of limitation had run before Lewis filed suit. The trial court denied Sidlow’s motion and granted a certificate of immediate review. This court granted the application, and for the reasons that follow, we affirm the trial court’s denial of summary judgment to podiatrist Sidlow.
Lewis had diabetes, and was treated by Sidlow from June 1996 to June 1, 2000. On June 1, 2000, Sidlow diagnosed Lewis with Charcot’s arthropathy in his right foot and gave him an orthotic prescription to have his right shoe adjusted. Charcot’s arthropathy is an arthritic condition involving the midfoot, heralded by increased blood flow to the affected limb, causing blood pooling, pain, swelling, increased heat, demineralization of the bones in the foot, and bone fractures. After initial onset, the symptoms abate, but left untreated, the condition leads to secondary bone infection and amputation. On June 8, 2000, Lewis was admitted to the hospital with sepsis, an infection in his blood, and was placed on intravenous antibiotics. An MRI of Lewis’s right foot on June 10, 2000, revealed fractures and osteomyelitis, an infection in the bones. He developed gangrene, and his leg was amputated in July 2000. Lewis died in August 2003.
The malpractice complaint alleges that Sidlow failed to exercise a reasonable degree of podiatric care, diligence, and skill ordinarily employed by podiatrists generally under similar conditions, which led to complications and permanent injury to Lewis’s right leg. Sidlow moved for summary judgment, arguing that the complaint was filed more than two years after the statute of limitation began running, *113relying solely on the pleadings and the deposition he took of Lewis’s expert witness. He contended that the expert testified that the injury manifested itself on March 6, 2000, and therefore Lewis’s complaint was barred.
The trial court disagreed, holding that
Defendant Sidlow has defended the case throughout by maintaining there was no negligence. A material issue of fact certainly exists as to whether there was any negligence at all. The Plaintiff is only complaining of negligence arising after the diagnosis of the foot in June of 2000. The fact that there might be some evidence of a misdiagnosis does not mandate a running of the statute of limitations from that date, particularly if there was subsequent negligence of which the Plaintiff complains. A misdiagnosis is not negligence per se.
Sidlow again argues on appeal that the testimony of Lewis’s expert establishes that Sidlow misdiagnosed Lewis in March 2000, and that consequently the statute of limitation had run by the time Lewis filed his complaint. Lewis argues in response that the statute began running on June 1, 2000, when Sidlow actually diagnosed Charcot’s foot and then failed to treat it properly.
A fair reading of the expert’s deposition testimony does not support Sidlow’s argument. The expert testified that Sidlow treated Lewis for years, always making the same notations, and that on March 6, 2000, Sidlow’s notes changed to show that Lewis complained of his orthotic boot breaking down and had signs that differed from previous visits.
Sidlow alleges on appeal that the trial court erred in applying the “continuous treatment” doctrine to find that the statute of limitation had not run when Lewis filed his complaint. He also contends that the trial court erred in failing to find that either Sidlow’s Charcot’s foot manifested in March 2000 or that Sidlow’s “misdiagnosis, itself, is the injury that commences the running of the medical malpractice statute of limitations.”
The two pivotal questions before us are: (1) When did the patient’s injury occur — at the time of the original misdiagnosis or some time after the proper diagnosis was made? and (2) When should the patient have discovered the injury in order to trigger the running of the statute of limitation? The record, as it now exists before this court, does not answer these all-important questions as a matter of law.
While we have found no Georgia case directly on point with this matter, some direction can be attained from our prior decisions.
*114[I]n most misdiagnosis cases, the injury begins immediately upon the misdiagnosis due to the pain, suffering, or economic loss sustained by the patient from the time of the misdiagnosis until the medical problem is properly diagnosed and treated. The misdiagnosis itself is the injury and not the subsequent discovery of the proper diagnosis; thus the fact that the patient did not know the medical cause of his suffering does not affect the applicability of OCGA § 9-3-71 (a).
(Emphasis supplied.) Harrison v. Daly, 268 Ga. App. 280, 283 (601 SE2d 771) (2004).
In this matter, unlike most misdiagnosis cases where the injury manifests itself before the proper diagnosis, the plaintiff alleges that, after the proper diagnosis, Sidlow treated him improperly, resulting in injury. Furthermore, the record, when viewed in the light most favorable to the plaintiff, indicates that, at the time of the proper diagnosis, the plaintiff had no redness or swelling in his foot which would have put him on notice that, at any time before that point, he had been injured. This evidence raises a significant question of fact as to both when the plaintiffs injury occurred and when the plaintiff should have known about this injury. In light of these questions, the trial court did not err in denying the defendant’s motion for summary judgment which would have required it to determine, as a matter of law, the date on which the patient’s injury occurred.
1. This is not a “continuous treatment” case controlled by the Supreme Court of Georgia’s ruling in Young v. Williams, 274 Ga. 845 (560 SE2d 690) (2002). In Young, the Supreme Court overruled this court’s opinion in Williams v. Young, 247 Ga. App. 337 (543 SE2d 737) (2000), which had adopted the “continuous treatment” doctrine and reversed the trial court’s grant of summary judgment. We concluded that a factual issue existed as to whether the doctor had treated the plaintiff in the two years preceding the filing of her lawsuit. The Supreme Court reversed and remanded, holding that “[t] he legislatively-prescribed statute of limitation does not provide for the commencement of the period of limitation upon the termination of the healthcare provider’s treatment of the patient, and the judicial branch is not empowered to engraft such a provision on to what the legislature has enacted.” Young v. Williams, supra, 274 Ga. at 848.
On remand, this court considered whether the trial court erred in granting the defendants’ motion for summary judgment on the basis that the suit was barred by the statute of limitation, and concluded that it did not err. Williams v. Young, 258 Ga. App. 821 (575 SE2d 648) *115(2002). This court noted that Williams’ last visit to Dr. Young was on September 30,1996, and that her suit was not filed until October 28, 1998, concluding that
the evidence before the trial court established that Williams’ foot and leg pain was continuous throughout Dr. Young’s treatment of her, and that the injury was present more than two years before she filed suit. [Cit.] Accordingly, the applicable two-year statute of limitation expired before Williams filed her lawsuit, and, for this reason, the trial court properly granted summary judgment to the defendants.
Id. at 824.1
Unlike the situation in Williams, however, the plaintiff before us filed his complaint within two years of his last visit to Sidlow. Thus we find no ground for reversal in Sidlow’s contention that the trial court erred in applying the continuous treatment doctrine, because that doctrine is inapplicable here.
2. The issue then becomes whether Lewis suffered an “injury” that started the statute running in March 2000. In considering this issue, it is helpful to review the legislative change in the statute of limitation for medical malpractice made in response to the Supreme Court of Georgia’s ruling in Shessel v. Stroup, 253 Ga. 56 (316 SE2d 155) (1984).
In Shessel, the court held that the 1976 medical malpractice statute of limitation, which provided that “an action for medical malpractice shall be brought within two years after the date on which the negligent or wrongful act or omission occurred,” violated the Equal Protection Clause in those medical malpractice actions in which the alleged negligence produced no injury until more than two years after the negligence occurred. Id. at 59. In response, the legislature repealed OCGA § 9-3-71 and substituted a new statute, subsection (a) of which provided that “an action for medical malpractice shall be brought within two years after the date on which an *116injury or death arising from a negligent or wrongful act or omission occurred.” Quinn v. Stafford, 257 Ga. 608, 610 (5) (a) (362 SE2d 49) (1987) .
The Supreme Court of Georgia repeated this concept in Young v. Williams, supra, reiterating that, under OCGA § 9-3-71 (a), the period of limitation “begins with the occurrence of an injury, not the performance of a negligent act.” 274 Ga. at 848. While subsequent cases have stated that “[i]n cases alleging failure to diagnose, the limitation period runs from the date of the misdiagnosis,” Harrison v. Daly, supra, 268 Ga. App. at 283, that proposition is only generally true. “[I]n most misdiagnosis cases, the injury begins immediately upon the misdiagnosis due to the pain, suffering, or economic loss sustained by the patient from the time of the misdiagnosis until the medical problem is properly diagnosed and treated. [Cit.]” Kane v. Shoup, 260 Ga. App. 723, 724 (1) (580 SE2d 555) (2003). “This is not always the case[,] Whitaker v. Zirkle, 188 Ga. App. 706 (374 SE2d 106) (1988) ,” Zechmann v. Thigpen, 210 Ga. App. 726, 729 (3) (437 SE2d 475) (1993), however, because “the focus of OCGA § 9-3-71 (a) is not the date of the negligent act but the ‘consequence of the defendant’s acts on the plaintiff.’ Jones v. Lamon, 206 Ga. App. 842, 845 (426 SE2d 657) (1992).” Vitner v. Miller, 208 Ga. App. 306, 307 (430 SE2d 671) (1993). The result is that “[wjhen an injury occurs subsequent to the date of medical treatment, the statute of limitation commences from the date the injury is discovered.” (Citation and punctuation omitted.) Zechmann v. Thigpen, supra, 210 Ga. App. at 729. In such cases, the misdiagnosis itself is not the injury. Price v. Currie, 260 Ga. App. 526 (580 SE2d 299) (2003) (physical precedent only). Whether the injury occurs before or after the negligence does not alter the analysis.
“Because the statute of limitation is an affirmative defense, OCGA § 9-11-8 (c), the burden was on [Sidlow] to show its applicability.” Walker v. Melton, 227 Ga. App. 149, 151 (1) (b) (489 SE2d 63) (1997). Therefore, Sidlow was required to show that Lewis suffered an injury in March 2000 that started the statute of limitation running. Based on the evidence of record, he has not done so. Sidlow relies on the deposition testimony of Lewis’s expert witness, quoting a portion of that testimony in support of his conclusion that the expert thought the injury began in March 2000. Reading the deposition as a whole, however, it is clear that the expert was merely commenting on the change in Sidlow’s notes in March 2000, before which “Sidlow always did the same thing in his notes.” Sidlow’s March 6,2000 notes varied from the norm by noting Lewis’s pulses and that his “boot is breaking down.” The expert said that the change in the pattern of Sidlow’s notes caught his attention, but despite repeated questioning, never said that Lewis had an injury on that date. In fact, the *117expert said Sidlow testified in his deposition, which is not in the record before us, that Lewis had no redness, swelling, or warmth even on the June 1, 2000 visit, although at that time Sidlow diagnosed Lewis as having Charcot’s foot.
The crux of the expert’s testimony is that Sidlow deviated from the accepted standard of care by failing to treat Lewis after his Charcot’s foot diagnosis on June 1, 2000. While the testimony that Sidlow quotes in his brief seems to indicate that the expert thought that Lewis’s injuries resulted from the Charcot’s foot in March 2000, the expert subsequently clarified in his deposition errata sheet that the opinion he gave regarding Lewis’s injuries referred to Sidlow’s failure to treat Lewis on June 1, 2000.
Based on this scant evidence, Sidlow has not established the absence of an issue of fact for a jury to determine whether he committed malpractice in his treatment of Lewis. Because it is the injury to the patient that starts the statute running, not the misdiagnosis standing alone, and because here, Sidlow has not proven that the injury began in March 2000, the trial court did not err in denying Sidlow’s motion for summary judgment.
Judgment affirmed.
Blackburn, P. J., Ruffin, P. J., Eldridge and Adams, JJ., concur. Andrews, P. J., and Mikell, J., dissent.The Supreme Court of Georgia initially granted certiorari in that case to consider the exact issue before us, asking, “In a medical malpractice action based upon the physician’s alleged misdiagnosis of the patient’s condition, what is ‘the date on which an injury... occurred’ for purposes of commencement of the two-year statute of limitations established by OCGA § 9-3-71 (a)?” Williams v. Young, Case No. S03G0606, May 5, 2003 (unpublished opinion). The court, however, subsequently vacated the grant of certiorari, and the United States Supreme Court denied Williams’s petition for a writ of certiorari (Williams v. Young,_U. S._(124 SC 2838,159 LE2d 267) (2004)), leaving this court’s decision in Williams v. Young, supra, 258 Ga. App. 821, as the last word.