CANADY, J.
We have for review the decision of the Second District Court of Appeal in TSE Industries, Inc. v. Larson & Larson, P.A., 987 So.2d 687 (Fla. 2d DCA 2008), in which [38]*38the district court certified direct conflict with the decision of the Fourth District Court of Appeal in Integrated Broadcast Services, Inc. v. Mitchel, 931 So.2d 1073 (Fla. 4th DCA 2006), regarding when the two-year statute of limitations begins to run on a legal malpractice claim. We have jurisdiction. See art. V, § 3(b)(4), Fla. Const.
The Second District held in Larson that where the judgment underlying a litigation-related legal malpractice claim is final, the statute of limitations does not begin to run until the final disposition of motions for sanctions. On similar facts, however, the Fourth District in Mitchel held that the limitations period on a legal malpractice claim began to run on the underlying judgment when that judgment was final but did not begin to run with respect to a subsequent sanctions judgment until the sanctions judgment became final. As we explain below, we agree with the Fourth District’s decision in Mitchel.
I. BACKGROUND
In 1998, Larson & Larson, P.A., as counsel for TSE Industries, filed a patent infringement suit in United States district court to enforce one of TSE’s patents against Franklynn Industries, Inc. After trial, the jury returned a verdict in favor of Franklynn, finding that TSE’s patent was invalid. The trial court entered judgment against TSE in the case on October 24, 2001. TSE filed timely motions for judgment as a matter of law and for new trial, and Franklynn filed a post-judgment motion for declaration of an exceptional case and for recovery of attorney fees. See 35 U.S.C. § 285 (2006) (“The court in exceptional cases may award reasonable attorney fees to the prevailing party [in patent infringement actions].”)
On August 16, 2002, the federal district court entered two orders. One order disposed of TSE’s motions and affirmed the jury verdict, thus entering judgment against TSE.1 TSE did not appeal this order, and the judgment was final thirty days later, September 16, 2002. In the other order of August 16, 2002, the federal judge granted Franklynn’s sanctions motion. The court found the case exceptional within the meaning of the statute, relying largely on its findings that a TSE employee engaged in inequitable conduct both before the Patent and Trademark Office and before the court and that TSE was aware of a possible problem with the patent and did not candidly disclose this fact. The federal court determined that but for such conduct, the trial would not have been necessary and awarded Franklynn prevailing party attorney fees and expert witness fees. Leaving the amount of the sanctions award for later determination, the trial court ordered Franklynn to file a revised statement of fees and costs. The parties, however, settled the issue before the court issued a final determination, and on October 10, 2002, the parties filed a stipulation of dismissal.
On October 5, 2004 — more than two years after the judgment in the patent case was final but less than two years after the parties filed the stipulation of dismissal — TSE filed a legal malpractice suit against Larson in Florida circuit court. TSE alleged negligence and breach of contract and sought monetary damages for the attorney fees and expenses incurred by TSE to initiate and prosecute the patent infringement suit and for the sums it paid [39]*39for Franklynn’s attorney fees and expenses.2
Larson moved for summary judgment based on the two-year statute of limitations for professional malpractice in section 95.11(4)(a), Florida Statutes (2002). After hearing argument, the trial court entered final summary judgment for Larson. Applying this Court’s decision in Silvestrone v. Edell, 721 So.2d 1173 (Fla.1998), the trial court held the action was barred. The court reasoned that judicial labor regarding the merits of the underlying action ended when the judgment was not appealed. Further, the court concluded that the order granting attorney fees could not alter that final judgment and TSE’s complaint did not allege any legal malpractice occurring after the August 2002 orders were entered.
As explained more fully in the next section, the Second District reversed, concluding that this Court’s Silvestrone opinion did not contemplate the factual scenario in this case. Larson, 987 So.2d at 691. Accordingly, the district court held that the statute of limitations did not begin to run until the parties settled the sanctions claim and filed the stipulation dismissing the case with prejudice.
II. SILVESTRONE AND THE CONFLICT IN THE CASES
Florida law provides that “[a] cause of action accrues when the last element constituting the cause of action occurs,” § 95.031(1), Fla. Stat. (2002), and that a legal malpractice action must be brought within two years “from the time the cause of action is discovered or should have been discovered with the exercise of due diligence,” § 95.11(4)(a), Fla. Stat. (2002). “A legal malpractice action has three elements: 1) the attorney’s employment; 2) the attorney’s neglect of a reasonable duty; and 3) the attorney’s negligence as the proximate cause of loss to the client.” Law Office of David J. Stern, P.A. v. Sec. Nat’l Servicing Corp., 969 So.2d 962, 966 (Fla.2007) (quoting Sec. Nat’l Servicing Corp. v. Law Office of David J. Stern, P.A., 916 So.2d 934, 936-37 (Fla. 4th DCA 2005)).
In Silvestrone, we specifically addressed the question of when the statute of limitations begins to run in a litigation-related legal malpractice claim. The district courts in both Larson and Mitchel expressly relied on Silvestrone. Accordingly, to provide context for these decisions and the conflict presented here, we begin by reviewing Silvestrone.
Silvestrone, the plaintiff in a federal antitrust action, prevailed when the jury returned a verdict in his favor and awarded him money damages. Silvestrone, 721 So.2d at 1174. The court entered final judgment two years later, when it resolved various post-trial motions, including Silves-trone’s motion for attorney fees and costs and a coplaintiffs motion for new trial. The final judgment became final on February 4, 1992, when it was not appealed. Almost a year later, Silvestrone filed a legal malpractice action against his former trial attorney “for exceeding his authority [40]*40to enter into a settlement agreement.” Id. His former attorney raised the statute of limitations as a bar to the suit, arguing that the statute began to run when the jury returned its verdict. Although the district court agreed, we unanimously quashed its decision. Id.
Beginning from the premise that the law was unclear regarding when the statute of limitations period begins to run for a litigation-related legal malpractice claim, we sought to establish a clearly delineated rule. Id. at 1175. We recognized in this context that until final judgment is entered, a trial court may revisit any nonfinal ruling. Accordingly we reasoned and held as follows:
[Wjhen a malpractice action is predicated on errors or omissions committed in the course of litigation, and that litigation proceeds to judgment, the statute of limitations does not commence to run until the litigation is concluded by final judgment. To be specific, we hold that the statute of limitations does not commence to run until the final judgment becomes final.
To be liable for malpractice arising out of litigation, the attorney must be the proximate cause of the adverse outcome of the underlying action which results in damage to the client. Since redressable harm is not established until final judgment is rendered, a malpractice claim is hypothetical and damages are speculative until the underlying action is concluded with an adverse outcome to the client.
[[Image here]]
We therefore hold, in those cases that proceed to final judgment, the two-year statute of limitations for litigation-related malpractice under section 95.11(I)(a), Florida Statutes (1997), begins to run when final judgment becomes final. This bright-line rule will provide certainty and reduce litigation over when the statute starts to run. Without such a rule, the courts would be required to make a factual determination on a case by case basis as to when all the information necessary to establish the enforceable right was discovered or should have been discovered.
Id. at 1175-76 (emphasis added) (citations and footnote omitted). For Silvestrone, this meant that his malpractice claim was timely because the two-year statute did not begin to run until the “final judgment became final,” id. at 1175, some two years after the jury verdict was entered, see id. at 1174.
In the present case, the Second District addressed a somewhat different factual scenario. Unlike Silvestrone — where the damages based on the jury’s verdict and the award of attorney fees were included in the same final judgment — in this case, when the final judgment against TSE Industries in the federal action was final, the sanctions order against TSE was not. The district court applied our Silvestrone decision as follows:
The Silvestrone court’s “bright-line rule” was that that the statute of limitations begins to run when “the litigation is concluded by final judgment.” Under the particular facts in Silvestrone, the litigation was concluded when the final judgment became final. In this case, however, the litigation was not concluded until the parties filed the stipulation to dismiss the underlying action with prejudice. When the final judgment in the patent infringement action became final for purposes of appeal, a second final judgment remained to be rendered in order to conclude the patent infringement action.
Larson, 987 So.2d at 690-91 (emphasis added). The Second District thus reasoned that until the order imposing sane-[41]*41tions was final, the legal malpractice cause of action did not accrue.
The operative facts of Mitchel, the conflict case, are similar. Mitchel represented Integrated Broadcast Services, the plaintiff in a federal lawsuit. The final summary judgment in favor of the defendants was affirmed on appeal, but while the appeal was pending, the defendants moved for sanctions — on grounds that Mitchel does not identify — against Integrated and its attorney, Mitchel.3 Two months after the final judgment was final, the trial court awarded the defendants sanctions, and Integrated appealed. The parties subsequently settled and filed a dismissal of the sanctions appeal in January 2003. Integrated filed its legal malpractice action against Mitchel more than two years after the final summary judgment was final on appeal but within two years of the sanctions judgment becoming final. Mitchel, 931 So.2d at 1074. Integrated’s malpractice complaint alleged that counsel’s negligence caused the adverse judgment in the underlying action and the subsequent sanctions award. The trial court held that the statute of limitations barred the suit.
The district court in Mitchel held that the statute of limitations barred the malpractice claim as to the underlying action but determined that the malpractice claim regarding the sanctions judgment was not barred. Id. The Fourth District cited Sil-vestrone and reasoned that “[bjecause the damages caused by the sanctions did not exist at the time the judgment became final in the underlying litigation, the trial court erred in concluding that the cause of action for legal malpractice for the sanctions accrued before the judgment for sanctions became final.” Mitchel, 931 So.2d at 1074.
In Larson, the Second District certified conflict with Mitchel “to the extent that it holds that the statute of limitations on the underlying judgment runs when the underlying judgment becomes final even when a motion for attorneys’ fees or sanctions remains pending.” Larson, 987 So.2d at 692-93.
III. ANALYSIS
Both Larson and TSE contend that there is only one accrual date for the malpractice claims brought by TSE against Larson, but they disagree about what that date is. Larson now argues, as it did below, that the cause of action accrued when the judgment entered in the underlying action became final. TSE contends that the cause of action accrued when TSE and Franklynn filed a dismissal after entering into a settlement agreement with respect to the claim for attorney fees. Below, we address these arguments and explain why we adopt the reasoning of Mitchel, which identifies two different accrual dates — one for the claims based on the underlying litigation and one for the claims based on the post-trial sanctions.
The statute of limitations requires that a legal malpractice action on a litigation-related claim be brought within two years after the cause of action is or should have been discovered, § 95.11(4)(a), Fla. Stat. (2002), and in Silvestrone we drew the line of accrual at the time final judgment was final to “provide certainty and reduce litigation over when the statute starts to run.” Silvestrone, 721 So.2d at 1176.4 [42]*42Until a final judgment is final, the outcome of the case and the occurrence of harm to the client remains uncertain, and it cannot be said that the cause of action was discovered or should have been discovered. See id. at 1175. Before that point is reached, the “malpractice claim is hypothetical and damages are speculative.” Id. But once a judgment adverse to the client has reached the point of finality, “the last element constituting the [malpractice] cause of action occurs,” § 95.031(1), Fla. Stat. (2002)— that is, the element of “loss to the client,” Law Office of David J. Stern, P.A., 969 So.2d at 966—and the cause of action is or should be “discovered,” § 95.11(4)(a), Fla. Stat. (2002).
The crux of Silvestrone’s reasoning is that it cannot be known with sufficient certainty that the client has suffered any loss caused by the lawyer’s negligence until the finality of a judgment adverse to the client. A favorable result for the client in the lawsuit — which could be the result of appellate proceedings — would, of course, mean that the client had suffered no loss. Silvestrone’s rule thus merely establishes a bright line for establishing when the client has suffered some loss as a consequence of the attorney’s negligence. It does not require that there be a determination of the full extent of all losses suffered by the client due to the lawyer’s negligence.
This is in line with the long-standing rule generally applicable to personal injury claims under which “the cause of action accrues and the statute begins to run from the time when the injury was first inflicted, and not from the time when the full extent of the damages sustained have been ascertained.” Seaboard Air Line R.R. Co. v. Ford, 92 So.2d 160, 164 (Fla.1956) (on rehearing); see also Highland Indus. Park, Inc. v. BEI Defense Sys. Co., 357 F.3d 794, 797 (8th Cir.2004) (“[W]e know of no state whatever in which an injured party must know the full extent of the damages that it may recover before the statute of limitations begins to run on its claim. Indeed, the cases on this issue are legion.”); Goodhand v. United States, 40 F.3d 209, 212 (7th Cir.1994) (stating that it “is a general principle of limitations law” that “[t]he statute of limitations begins to run upon the discovery of the injury, even if the full extent of the injury is not discovered until much later”).
We reaffirmed Silvestrone’s bright-line rule in Fremont Indemnity Co. v. Carey, Dwyer, Eckhart, Mason & Spring, P.A., 796 So.2d 504 (Fla.2001), a case involving a legal malpractice claim arising from a lawyer’s failure to inform the client — the defendant in an architectural malpractice action — of settlement offers. We concluded that the case presented “a classic example of why redressable harm cannot be determined until the conclusion of the litigation.” Id. at 506. We specifically relied on the Silvestrone rule that the limitations period for a litigation-related legal malpractice claim does not commence “until the final judgment becomes final” with a result adverse to the client. We reasoned that until the judgment adverse to the client reached the point of finality, the “possibility existed that [the client] would not suffer any redressable harm” as a consequence of the lawyer’s negligence. As in Silvestrone, our analysis in Fremont turned on the need for sufficient certainty that the client had sustained some loss. Fremont does not hold that the full extent of a client’s loss must be established in [43]*43order for the limitations period to commence.
TSE argues for our approval of the Second District’s alteration of the Silves-trone rule in the decision under review. The Second District determined that the Silvestrone rule of finality applied in this case only when the sanctions claim was resolved by the stipulation of dismissal. According to the Second District, it was only at that point that all matters related to the underlying litigation were concluded and the limitations clock started ticking. Larson, 987 So.2d at 692. “Until the case is fully resolved,” the court reasoned, “there is a chance that the appeals process could result in a reversal of the original decision that established an injury.” Id. (emphasis added). The Second District’s reasoning on this point, however, is not correct.
Under the Silvestrone bright-line rule, redressable harm was established when the final judgment against TSE in the patent infringement action became final in September 2002. At that point, “the information necessary to establish the enforceable right [of TSE against Larson for malpractice] was discovered or should have been discovered.” Silvestrone, 721 So.2d at 1176. The dismissal filed after the sanctions claim was settled had no effect on the finality of the final judgment in the underlying action. Even if the federal court had issued a final judgment on sanctions and it was reversed on appeal, the finality of the judgment in the patent case and the attendant establishment of damages would have been unaffected. The Second District’s statement to the contrary is plainly erroneous.
Under the Second District’s approach, the determination of redressable harm, which was essential to our Silvestrone holding, is effectively removed from the analysis and the accrual of the cause of action occurs only when there is no further ongoing litigation related to the underlying action. This is inconsistent with the provision in section 95.11(4)(a) that “the period of limitations shall run from the time the cause of action is discovered or should have been discovered with the exercise of due diligence.” When the underlying action was finally determined adversely to TSE, TSE had knowledge of the harm it suffered as a result and the malpractice cause of action accrued with respect to the damages entailed by that adverse determination. There is no basis for concluding that TSE’s cause of action for these damages was not discovered or discoverable at that point.5
In concluding that no cause of action accrued in this case until the settlement of the sanctions claim, the Second District expressed concern that otherwise parties “would be forced to argue inconsistent positions if they were required to file the malpractice action before resolution of attorneys’ fees in the underlying action.” Larson, 987 So.2d at 692. The Second District based its decision in part on what it characterized as “[t]he policy concerns articulated” in Peat, Marwick, Mitchell & Co. v. Lane, 565 So.2d 1323, 1324 (Fla.1990). Larson, 987 So.2d at 692. We conclude, however, that the Second Dis-[44]*44triet’s reliance on Peat, Marwick is unwarranted.
In Peat, Marwick we addressed the question of when the limitations period began for an accounting malpractice action related to tax advice and the preparation of a tax return — with receipt of a notice of tax deficiency or when the tax court action was final. In rejecting the date of the receipt of the deficiency notice as the date of accrual, we referred to the decisions of the district courts concerning “attorney malpractice actions,” which had “expressly held that a cause of action for legal malpractice does not accrue until the underlying legal proceeding has been completed on appellate review because, until that time, one cannot determine if there was any actionable error by the attorney.” Peat, Marwick, 565 So.2d at 1325. Our conclusion that “the limitations period for accounting malpractice commenced when the United States Tax Court entered its judgment” flowed from the same reasoning. Id. at 1327. We thus “rejectfed] Peat Marwick’s contention that an IRS deficiency determination conclusively establishes an injury upon which to base a professional malpractice action.” Id. at 1326.
In Peat, Marwick we also observed that accepting Peat, Marwick’s argument would mean that the clients “would have had to have filed their accounting malpractice action during the same time that they were challenging the IRS’s deficiency notice in their tax court appeal.” Id. at 1326. We concluded that “[s]uch a course would have placed them in the wholly untenable position of having to take directly contrary positions in these two actions.” Id. That is, in the tax court the clients would be arguing that the return prepared by the accountant was proper, while in the malpractice action the client would be arguing that the return was prepared negligently.
The rule we rely on here does not set up the same prospect regarding the assertion of inconsistent positions as did the rule urged by Peat, Marwick. Ordinarily, there will not be a substantial gap in time between ultimate resolution of the litigation regarding the underlying judgment and the resolution of the litigation regarding sanctions. In any event, Peat, Mar-wick does not articulate a rule that the running of the statute of limitations for professional malpractice is held in abeyance until the conclusion of any collateral litigation in which the client might assert a position inconsistent with the malpractice claim. Such a rule could not be reconciled with the commencement point — “the time the cause of action is discovered or should have been discovered” — established in section 95.11(4)(a). In Peat, Marwick, 565 So.2d at 1325, we acknowledged that a legal malpractice cause of action accrues as soon as the client can “determine if there was any actionable error by the attorney” — that is, when “the underlying legal proceeding has been completed on appellate review” and the client can know of damages arising from the final judgment suffered as a consequence of the attorney’s malpractice.
TSE also argues that the Second District’s conclusion that none of TSE’s claims were barred is justified by the application of the continuing or continuous representation doctrine, “a doctrine of law which tolls the legal negligence statute of limitations so long as the attorney continues to represent the client in the matter.” Alagia, Day, Trautwein & Smith v. Broadbent, 882 S.W.2d 121, 122 (Ky.1994). Although the Second District concluded that its interpretation of Silvestrone “rendered] TSE’s argument regarding the continuing representation doctrine moot,” the result reached by the Second District is certainly consistent with that doctrine. [45]*45Larson, 987 So.2d at 692; see also id. at 693 (Altenbernd, J., concurring) (“Our holding today probably is a modified version of the continuing representation doctrine.”).
We have not previously adopted the continuing representation doctrine in legal malpractice actions, see Perez-Abreu, 790 So.2d at 1056 (Pariente, J., concurring in result only) (discussing rationale for the continuous representation rule), and for the reasons presented below, we conclude that the doctrine is inconsistent with the pertinent statutory provisions.
The continuous representation doctrine is predicated on the view that because of “the attorney’s superior knowledge of the law and the dependence of the client,” “there can be no effective discovery of the [the attorney’s] negligence so long as the relationship prevails.” Alagia, 882 S.W.2d at 125. Under this view, although the attorney’s negligence and the resulting harm may be manifest, the nature of the ongoing attorney-client relationship requires that the running of the statute of limitations be tolled until that relationship is terminated with respect to the matter at issue. Certain “practical advantages” are cited in support of the doctrine. Id. TSE argues in accordance with the Kentucky court’s reasoning in Alagia that the doctrine allows “a negligent attorney ... to correct or mitigate the harm if there is time and opportunity and if the parties choose such a course.” Id. Further, it is urged that without the doctrine, “the client may be forced, on pain of having his malpractice claim become time-barred, to automatically accept the advice of a subsequent attorney.” Id.
As a practical matter, by holding that redressable harm is not established for purposes of a legal malpractice claim “until the final judgment becomes final,” Silvestrone, 721 So.2d at 1175, we substantially addressed the basic concern regarding undue disruption of the attorney-client relationship underlying the continuing representation doctrine. Because finality requires the conclusion of appellate review, if any, there is no requirement that an ongoing attorney-client relationship be disrupted until two years after the appellate process has run its course. In the usual case, collateral matters will be resolved by then. However, even if they are not, once a judgment adverse to the client has reached the point of finality, there is no justification for tolling the statute of limitations.
In Kelley v. School Board of Seminole County, 435 So.2d 804 (Fla.1983), which we decided long before Silvestrone, we rejected the continuing representation doctrine in an architectural malpractice action. We disapproved the district court majority’s conclusion that “the ‘continuous treatment’ doctrine [is] embodied in Florida’s statute of limitations when a client-professional relationship is involved.” Kelley, 435 So.2d at 805 (footnote omitted); see Almand Constr. Co. v. Evans, 547 So.2d 626, 628 (Fla.1989) (noting that in Kelley the Court “rejected the ‘continuous treatment’ doctrine”). And we stated that the dissenting opinion of “Judge Cowart reached the proper conclusion.” Kelley, 435 So.2d at 805. Judge Cowart squarely rejected the notion that an ongoing client-professional relationship justifies tolling the statute of limitations:
Existing law does not require a client to sue his professional “immediately” after discovering negligently caused injury or damage. Applicable statutes of limitation give either two or ... four years within which time the client should either give up on his amiable but bungling professional, get competent help and sue, or be forever barred from asserting his stale claim.
[46]*46Sch. Bd. of Seminole County v. GAF Corp., 413 So.2d 1208, 1213 (Fla. 5th DCA 1982) (Cowart, J., dissenting), quashed sub nom. Kelley v. Sch. Bd. of Seminole County, 435 So.2d 804 (Fla.1983); see Kelley, 435 So.2d at 805 (citing this text with approval).
TSE’s argument that we should employ the continuous representation doctrine to toll the statute of limitations founders not only on our decision in Kelley but also on the text of section 95.051, Florida Statutes (2002), which restricts the circumstances under which statutes of limitations may be tolled. Section 95.051(1) contains a list of specific circumstances in which the running of the time under statutes of limitations (subject to the exception of certain statutes) is tolled. The existence of an ongoing attorney-client relationship is not on that list. (Nor is the existence of any other ongoing professional-client relationship.) Section 95.051(2) provides that “[n]o disability or other reason shall toll the running of any statute of limitations” except as specifically authorized by statute. See Hearndon v. Graham, 767 So.2d 1179, 1185 (Fla.2000) (“[T]he tolling statute specifically precludes application of any tolling provision not specifically provided therein.”). Accordingly, in the absence of a specific statutory authorization for doing so, we are precluded from tolling the statute of limitations based on the continuous representation doctrine.
In sum, the position urged by TSE and the analysis employed by the Second District are wholly detached from the governing statute, which ties the running of the limitations period for lawyer malpractice claims to “the time the cause of action is discovered or should have been discovered.” § 95.11(4)(a).6 The policy considerations advanced by TSE and the Second District — whatever their merits may be — • cannot be allowed to defeat the policy choice which is embodied in the statutory text enacted by the Legislature.
Under the Second District’s reasoning, if collateral post-trial issues remain pending, a cause of action for malpractice does not accrue even when after the conclusion of any appellate review, the client is subject to an enforceable final judgment which resulted from a lawyer’s malpractice. It requires a willing suspension of disbelief to conclude that there is no “redressable harm” in such circumstances. At the point a judgment adverse to the client becomes irreversibly final, it can no longer be said that the “malpractice claim is hypothetical and damages are speculative.” Silvestrone, 721 So.2d at 1175. It cannot plausibly be argued that a client who is subject to such an enforceable judgment and thus faces the immediate prospect of execution of the judgment has not “discovered” the harm caused by the lawyer’s negligence. The conclusion that the statute of limitations does not begin to run in such circumstances has not the slightest connection with the discovery rule laid down in the statute. And it is inconsistent with the reasoning of Silvestrone, which is predicated on the discovery rule.
Finally, we turn to Larson’s argument that TSE’s sanctions-based claim is barred because it accrued when the underlying judgment became final. We conclude that this argument is no more consistent with the statute of limitations and Silvestrone than is the view adopted by the Second District.
In Mitchel, the Fourth District determined that two statute of limitations peri[47]*47ods applied; thus there were two legal malpractice claims. Mitchel, 931 So.2d at 1074; see Larson, 987 So.2d at 692 (referring to the Fourth District’s finding of a “bifurcated statute of limitations for the two judgments”). Although the client’s legal malpractice claim regarding the adverse judgment was barred by the statute of limitations, the claim that Mitchel negligently caused the judgment for sanctions was not. The district court explained as follows:
In order to be liable for malpractice resulting from litigation, the lawyer “must be the proximate cause of the adverse outcome of the underlying action which results in damage to the client.” Silvestrone v. Edell, 721 So.2d 1173, 1175 (Fla.1998). Because the damages caused by the sanctions did not exist at the time the judgment became final in the underlying litigation, the trial court erred in concluding that the cause of action for legal malpractice for the sanctions accrued before the judgment for sanctions became final.
[[Image here]]
The summary judgment is affirmed as it relates to plaintiffs claims other than the damages resulting from the sanctions, but is reversed in order for plaintiff to file an amended complaint limited to damages caused by the sanctions.
Mitchel, 931 So.2d at 1074 (emphasis added).
Sanctions such as attorney fees are collateral to the underlying judgment and do not prevent judgment in the underlying action from becoming final. However, being “collateral” does not mean they are subsumed by the underlying judgment. A judgment imposing such sanctions is ap-pealable independently of the final judgment in the underlying action. See Janelli v. Pagano, 492 So.2d 796, 796-97 (Fla. 2d DCA 1986) (“A subsequent attorney’s fee order judgment is a final order appeal-able as such apart from the final judgment.”); see also Oregon Natural Desert Ass’n v. Locke, 572 F.3d 610, 614 (9th Cir.2009) (“An award of attorney fees raises legal issues collateral to and separately appealable from the decision on the merits.” (citing Budhnich v. Becton Dickinson & Co., 486 U.S. 196, 108 S.Ct. 1717, 100 L.Ed.2d 178 (1988))). Thus, as the Fourth District concluded, the finality of a determination regarding a claim for sanctions may occur independently of the finality of the underlying judgment. And a determination adverse to the client in the sanctions litigation does not ineluctably flow from an adverse determination in the underlying litigation.
The damage suffered by the client arising from the underlying judgment is discrete from any damage that might be suffered by the client arising from a sanctions claim that is not finally adjudicated in the underlying judgment. The discrete occurrence of damage — marked by a separate final judgment or the equivalent— gives rise to the discrete discovery of loss and the discrete accrual of separate causes of action for malpractice.7 Here, the point [48]*48of finality with respect to the sanctions claim was not reached until the settlement was entered while the sanctions litigation was pending. Until then, it was possible that an appeal in the sanctions litigation would produce an outcome favorable to TSE. Only with the entry of the settlement agreement was the existence of any harm to TSE arising from the sanctions claim determined with sufficient certainty to justify commencement of the limitations period.
Accordingly, we conclude that the Fourth District’s bifurcated approach is most consistent with section 95.11(4)(a) and our reasoning in Silvestrone. Given that the underlying litigation between TSE and Franklynn was concluded before the sanctions litigation was resolved, redressa-ble harm was established regarding the underlying litigation before there was re-dressable harm with respect to the sanctions. The sanctions-based malpractice claim was “hypothetical and [the related] damages [were] speculative until ... an adverse outcome to the client” occurred in the litigation concerning the sanctions. Silvestrone, 721 So.2d at 1175.
IV. CONCLUSION
Under our decision, TSE’s malpractice claim regarding the final judgment in the underlying patent action is barred, but the malpractice action regarding the sanctions claim is not barred. Accordingly, we quash that portion of the Second District’s decision that reversed the summary judgment as to the malpractice claim regarding the underlying patent infringement action, and we approve the Second District’s reversal of the summary judgment as to the malpractice claim regarding sanctions.8
It is so ordered.
POLSTON and LABARGA, JJ., concur.
PERRY, J., concurs in part and dissents in part with an opinion.
LEWIS, J., dissents with an opinion, in which QUINCE, C.J., and PARIENTE, J., concur.