BATCHELDER, J.
The Superior Court ([O’Neil, J.) has transferred, without ruling, the following three questions of law:
A. Does a cause of action exist against a person who allegedly created false material evidence while acting as defense counsel in a previous case and, after withdrawing as counsel, allegedly gave false testimony advancing such evidence?
B. If so, what are the elements of such a new cause of action?
C. Does such cause of action also exist against such counsel’s firm?
In addition, the trial court transferred two questions of law with ruling:
D. Did the Superior Court err in ruling that defendants are collaterally estopped from litigating the issue of the reasonableness of plaintiffs’ attorneys’ fees in the underlying case?
E. Did the Superior Court err in ordering plaintiffs to produce attorney-client and work product materials?
We address each question in turn.
For purposes of this opinion only, we will assume the underlying facts to be as follows: On March 18, 1988, at a real estate closing at defendant Robert Schroeder’s office, the plaintiffs, Mark and Kathy Aranson, acquired title to a condominium unit in Conway from the Woodland Road Realty Trust. Approximately elgETmonthslate^the Town of Conway (the town) brought suit against the trust,’~the trustees, and the plaintiffs on the theory that the condommium unit had been sold without a certificate of occupancy as required by local ordinance.
Schroeder and his law firm initially represented the trust and the trustees in the town’s suit. The plaintiffs, however, obtained inde[362]*362pendent counsel and brought a counterclaim against the trust and the trustees, seeking, among other things, to rescind the purchase of the unit under the terms of the condominium statute, RSA chapter 356-B (1984 & Supp. 1994), and to recover treble damages under the Consumer Protection Act, RSA chapter 358-A (1984 & Supp. 1994). The plaintiffs’ counterclaim was severed from the action brought by the town.
The town prevailed in its action against the trust, the trustees, and the plaintiffs. The plaintiffs, however, prevailed at a jury-waived trial in Superior Court (Temple, J.) on the counterclaim. They were awarded rescissionary relief, consequential damages of $172,339.06, and attorney’s reasonable fees, determined in a subsequent hearing in Superior Court (Mohl, J.) to be $53,042.75 plus $3,851.85 in disbursements. The trust and the trustees appealed the award against them, and we summarily affirmed by order dated July 5, 1991. Although the plaintiffs have pursued their New Hampshire judgment in the Massachusetts courts where the trustees reside, they have not collected a farthing.
The plaintiffs in this new action ask us to recognize a cause of action for the tort of malicious defense, define its elements, and determine whether an attorney’s law firm is also subject to liability. In effect they seek an aspect of tort reform, the recognition at common law of a remedy where heretofore none has existed. The success of the plaintiffs’ action depends upon what took place at the real estate closing and the subsequent actions taken bv Schroeder in his role as counsel for the trust and the trustees in defense of the counterclaim. In this regard, the Superior Court (Temvle. J.) found:
Crucial to determining most of this case are the facts surrounding the closing that took place March 18, 1988. Sometime before that date, plaintiff Mark Aranson had called the Town offices relative to whether there were any problems in buying the unit from tKellifendants and helaras advised that there [were] not. According to defendant John Thompson, the occupancy permit was discussed at the closing, stating that Mr. Aranson had said that the permit was all set and that he had checked with the building inspector. Mr. and Mrs. Aranson, as well as the selling agent of the defendants, deny that any discussion concerning the permit, or the call Mr. Aranson made to the building inspector’s office, took place at the closing.Ajmemorandum concerning this alleged discussion was prepared by Attorney Robert Schroeder, who represented the financing bank at the closing in his office. AttorneySchroeder had no [363]*363independent recollection of this conversation. The memo’s reference to the presence of defendant Richard Sokolow at the closing, which did not occur, and the fact that the existence of the memo was made known only subsequent to pleadings being filed by the plaintiffs, causes the'Court to find that this undated memo was likely pfepdrecT'dfter consultation between the defendants and Attorney Schroeder subsequent to the defendants learning of the plaintiffs’ position and claims.
The Court accepts the plaintiffs’ contentions that the discussion concerning the permit and telephone call did not take place at the closing. Further, the Court finds that the plaintiffs did not know of the occupancy permit requirement until they were informed of such when suit was brought by the Town of Conway.
The evidence shows that the plaintiffs Aranson did not know at the closing that no certificate of occupancy existed for the unit and that the plaintiffs Aranson did rely upon the representations made by the defendants relative to the unit purchased that all approvals had been obtained, that construction was complete with stated warranty. Testimony proved that the unit would not have been purchased had the plaintiffs known of the lack of certificate and building deficiencies causing the Town to withhold the certificate.
(Emphasis added.)
The essence of the plaintiffs’ argument is captured in three sentences in their brief:
The cause of action [malicious defense] ... is essentially the mirror image of § 674 of the Restatement (2d) of Torts, and would do for false defenses what §674 already does for false claims: provide a remedy. The present state of the law — condemning false evidence from the plaintiff’s side but arguably tolerating it from the defendant’s side — is one-sided and unfair. Both forms of misconduct should be treated the same; both should be condemned, and made the subject of damages.
In Rockhouse Mountain Property Owners Association, Inc. v. Town of Conway, 127 N.H. 593, 503 A.2d 1385 (1986), we were likewise asked to recognize a new cause of action. Rockhouse is helpful in giving guidance for the task at hand:
As in any case in which we are asked to recognize a new cause of action, it is a question of policy whether it would be [364]*364wise to provide the relief that the plaintiffs seek. Reaching an answer to this question requires two quite separate steps, for we must determine whether the interest that the plaintiffs assert should receive any legal recognition and, if so, whether the relief that the plaintiffs request would be an appropriate way to recognize it.
Id. at 597-98, 503 A.2d at 1387-88 (citation omitted).
In deciding this question, we keep in mind that the New Hampshire Constitution provides:
Every subject of this state is entitled to a certain remedy, by having recourse to the laws, for all injuries he may receive in his person, property, or character; to obtain right and justice freely, without being obliged to purchase it; completely, and without any denial; promptly, and without delay; conformably to the laws.
N.H. CONST, pt. I, art. 14.
The analytical predicate to any consideration of the transferred questions must be a determination that the relief sought is not obtainable by other existing legal remedies. The plaintiffs seek consequential and special damages, costs, interest, attorney’s fees, and enhanced compensatory damages — the same damages to which they would have been entitled in a successful suit for malicious prosecution. See RESTATEMENT (SECOND) OF TORTS § 681 (1977).
The plaintiffs characterize their action as a variant of the tort of malicious prosecution, see Hogan v. Robert H. Irwin Motors, Inc., 121 N.H. 737, 739, 433 A.2d 1322, 1324 (1981), which originated in the misuse of the criminal process and has been adopted to redress malicious civil prosecution as well, see Kolka v. Jones, 71 N.W. 558, 559 (N.D. 1897). In the civil arena, wrongs of this sort are frequently presented as abuse of process actions. See, e.g., Long v. Long, 136 N.H. 25, 30, 611 A.2d 620, 623 (1992). Although separate and distinct, malicious prosecution and abuse of process “have the common element of an improper purpose in the use of legal process, and there are many cases in which they overlap and either will lie.” W. KEETON ET AL., PROSSER AND KEETON ON THE LAW OF TORTS 898 (5th ed. 1984).
Absent a remedy for malicious defense, the plaintiffs have, on the facts before us, only a claim that sanctions be imposed by the trial court based upon the use of false evidence by the defendants. See, e.g., Harkeem v. Adams, 117 N.H. 687, 690-91, 377 A.2d 617, 619 (1977). In appropriate circumstances, there may be ample reason to [365]*365extend the reach of the sanctions to counsel who engages in the fostering of an unfounded defense or pursues a defense for an improper purpose. Van Patten & Willard, The Limits of Advocacy: A Proposal for the Tort of Malicious Defense in Civil Litigation, 35 HASTINGS L.J. 891, 926 (1984) (citing cases). The difference between the adoption of the tort of malicious defense and the existing power of courts to levy sanctions is the nature and extent of the damages recoverable by the aggrieved party. Is a plaintiff less aggrieved when the groundless claim put forth in the courts is done defensively rather than affirmatively in asserting a worthless lawsuit for improper purposes? We think not.
We answer the first transferred question in the affirmative. In adopting, as a part of our common law, the tort of malicious defense, as we do today, we are mindful that fundamental changes in our jurisprudence must be brought about sparingly and with deliberation. Cf. Bonte v. Bonte, 136 N.H. 286, 616 A.2d 464 (1992) (Brock, C.J., and Batchelder, J., dissenting). Such a change was precipitated, for example, a quarter century ago, when Clifton Buttrick’s car went off the road after its headlights failed, and we adopted the doctrine of strict liability in tort. Buttrick v. Lessard, 110 N.H. 36, 260 A.2d 111 (1969). The often quoted language of Oliver Wendell Holmes, Jr., from his first lecture to the Lowell Institute in Boston, provides background for what we do today:
The life of the law has not been logic: it has been experience. The felt necessities of the time, the prevalent moral and political theories, intuitions of public policy, avowed or unconscious, even the prejudices which judges share with their fellow-men, have had a good deal more to do than the syllogism in determining the rules by which men should be governed. The law embodies the story of a nation’s development through many centuries, and it cannot be dealt with as if it contained only the axioms and corollaries of a book of mathematics. In order to know what it is, we must know what it has been, and what it tends to become.
O.W. Holmes, Jr., The Common law l (1881).
The defendants assert that no jurisdiction has to date adopted malicious defense as a cause of action. This, however, is only part of the story. The Supreme Court of South Carolina impliedly recognized the tort in Cisson v. Pickens, 186 S.E.2d 822, 825 (S.C. 1972). And relief has been granted where the tortious conduct took root in a cross complaint. Bertero v. National General Corp., 529 E2d 608, 616 (Cal. 1975). In Bertero, the California court declined to adopt [366]*366the tort of malicious defense yet found that a malicious prosecution claim could be grounded on the cross complaint. Id. In such cases the injury and expenses to the individual and the harm to the judicial system are the same irrespective of whether the conduct had its origins in the defense of a claim or in a cross claim in the same case. See Van Patten & Willard, supra at 894.
Moreover, anyone who has been a litigant knows that the fact of litigation has a profound effect upon the quality of one’s life that goes far beyond the mere entitlement to counsel fees. Litigation is a disturbing influence to one degree or another. The litigant may have the benefit of skilled and conscientious counsel as well as a strong and well-founded case on the facts, but until such time as the favorable verdict is in hand beyond the reach of appeal, there is a day-to-day uncertainty of the outcome. One wonders about the availability of witnesses at the appropriate times and whether their information will be adequately imparted. One may have gnawing uncertainty about the myriad things that can go wrong in a lawsuit. Here, although the plaintiffs prevailed in their lawsuit, they did so at a price — in time, money, and uncertainty — that was substantially exacerbated by the alleged actions of Schroeder. If a factual predicate exists to support liability and a measure of the damages thus exacerbated, the plaintiffs are entitled to a remedy to that extent.
In adding malicious defense to our common law, we merely recognize that when a defense is commenced maliciously or is based upon false evidence and perjury or is raised for an improper purpose, the litigant is not made whole if the only remedy is reimbursement of counsel fees. It follows that upon proving malicious defense, the aggrieved party is entitled to the same damages as are recoverable in a malicious prosecution claim.
Van Patten and Willard, supra at 923, note that “[t]he action for malicious defense is also needed to protect the integrity of the judicial process, to deal with dishonest and unethical behavior, and to discourage misuse and abuse of limited judicial resources.” “When the lawyer goes beyond the role of counselor and intentionally initiates defensive action that, harasses the plaintiff and that the attorney knows or should know is without a credible basis, then the attorney, no less than the client, should be liable.” Van Patten & Willard, supra at 9277
Of course, the mere existence of a remedy for malicious defense will not serve as a license for its abuse. Malicious defense, like its counterpart malicious prosecution, is a limited cause of action that will lie only in discrete circumstances, and malicious defense claims [367]*367will accordingly be scrutinized closely and construed narrowly. Cf. Barquis v. Merchants Collection Ass’n of Oakland, 496 P.2d 817, 824 n.4 (Cal. 1972) (characterizing malicious prosecution as a “narrowly circumscribed” cause of action). We would not, for example, look favorably upon a plaintiff’s threatening a malicious defense action when faced with a defendant’s general denial of liability, for a party should not be precluded from legitimately raising a defense for fear of such an action. Cf. Melvin v. Pence, 130 F.2d 423, 426 (D.C. Cir. 1942). Similarly, we would not sanction threats of malicious defense litigation as a means to gain tactical advantage, intimidate counsel, curtail legitimate discovery, or enhance settlement prospects. And a malicious defense action would not be the proper recourse for a plaintiff merely dissatisfied with a monetary judgment, for in that instance sufficient mechanisms for relief already exist in our law. See Belanger v. Teague, 126 N.H. 110, 111, 490 A.2d 772, 772 (1985) (additur may be sought “as alternative relief on a motion for a new trial on the ground of inadequate damages”).
In answer to the second transferred question, regarding the elements of malicious defense, we adopt the following standard:
One who takes an active part in the initiation, continuation, or procurement of the defense of a civil proceeding is subject to liability for all harm proximately caused, including reasonable attorneys’ fees, if
(a) he or she acts without probable cause, i.e., without any credible basis in fact and such action is not warranted by existing law or established equitable principles or a good faith argument for the extension, modification, or reversal of existing law,
(b) with knowledge or notice of the lack of merit in such actions,
(c) primarily for a purpose other than that of securing the proper adjudication of the claim and defense thereto, such as to harass, annoy or injure, or to cause unnecessary delay or needless increase in the cost of litigation,
(d) the previous proceedings are terminated in favor of the party bringing the malicious defense action, and
(e) injury or damage is sustained.
Van Patten & Willard, supra at 933-34. These elements essentially mirror those required to prove the tort of malicious prosecution. See ERG, Inc. v. Barnes, 137 N.H. 186, 190, 624 A.2d 555, 558 (1993); cf. Restatement (Second) of torts § 674 (1976).
It has been suggested that a malicious defense action should not contain as an element termination of the previous proceedings in [368]*368favor of the party asserting the malicious defense claim. In malicious prosecution actions, the “termination” requirement serves “to avoid inconsistent verdicts and to avoid prejudice to the plaintiff’s case by the introduction of evidence of plaintiff’s intent.” Van Patten & Willard, supra at 928 (footnotes omitted). Requiring that the termination be favorable “tends to eliminate unnecessary litigation, for a defendant who loses in the principal action will not institute a malicious prosecution suit, since the adverse judgment operates to establish conclusively that the plaintiff had probable cause.” Babb v. Superior Court of Sonoma County, 479 P.2d 379, 382 (Cal. 1971). We see no reason why these principles would not be equally applicable to the malicious defense context.
The third transferred question asks to what extent an attorney’s law firm could be subject to suit under the same theory. The relationship between Schroeder and his law firm, as well as the status of the firm as an entity, is not contained in the record available to us. Consequently, we decline at this time to respond, as the answer in the first instance is best left to the sound discretion of the trial court. Based upon the facts presented to it and in light of applicable legal principles, the trial court can best determine the liability, if any, of a malicious attorney’s law firm. See RSA ch. 304-A (1984) (UNIFORM PARTNERSHIP ACT); RSA ch. 294-A (1987 & Supp. 1994) (professional corporations).
The fourth transferred question concerns the trial court’s ruling that the defendants could not attack the award of relevant attorney’s fees to the plaintiffs in the underlying case. The plaintiffs argued below, as they do here, that because the court in the underlying case specifically approved the award, the defendants were collaterally estopped from contesting the reasonableness of its appropriate components. We vacate and remand.
The elements of collateral estoppel are well-established: the issue subject to estoppel must be identical in each action and the finding must have been essential to the first judgment; the party to be estopped must have appeared in the first action or have been in privity with someone who did, and must have had a full and fair opportunity to litigate the issue; and the first action must have resolved the issue finally on the merits.
Day v. N.H. Retirement System, 138 N.H. 120, 122, 635 A.2d 493, 495 (1993). As between a party and a non-party, “privity” requires “virtual representation” and “substantial identity,” id. at 123, 635 A.2d at 495, such that the interests of the non-party “were in fact represented and protected in the [prior] litigation,” Daigle v. City of [369]*369Portsmouth, 129 N.H. 561, 571, 534 A.2d 689, 694 (1987) (quotation omitted).
The plaintiffs thus had to demonstrate in the proceedings below that defendant Schroeder exerted such control over the underlying litigation that he should be bound by a determination against the underlying defendants as though he himself were a party. See RESTATEMENT (SECOND) OF JUDGMENTS § 39 (1982). It appears, however, that the plaintiffs presented no evidence in support of this position. The trial court apparently agreed with the plaintiffs’ assertion that Schroeder was in privity with the underlying defendants because he allegedly created the “false” defense and then acted as a co-conspirator in presenting that defense. Such an assertion, standing alone, does not warrant a determination that Schroeder’s involvement in the underlying action was so significant as to constitute control. Accordingly, we vacate the trial court’s decision and remand to the trial court for factual findings on this issue. See id., comment c at 384 (outlining criteria for consideration in determining whether involvement is extensive enough to constitute control).
The final transferred question involves the trial court’s order that the plaintiffs produce all documents relating to the underlying case. The plaintiffs maintain that the requested documents are protected by the attorney-client privilege and work product doctrine. The defendants, on the other hand, argue that the plaintiffs impliedly waived the attorney-client privilege when they placed at issue their litigation costs in the underlying suit.
We have not previously addressed the question whether a party waives the attorney-client privilege by affirmatively pleading a claim that places at issue the subject matter of privileged communications. We have acknowledged that in certain circumstances evidentiary privileges must yield to countervailing considerations. See, e.g., Nelson v. Lewis, 130 N.H. 106, 109, 534 A.2d 720, 722 (1987) (patient partially waives physician-patient privilege by putting medical condition at issue in medical negligence suit); McGranahan v. Dahar, 119 N.H. 758, 764, 408 A.2d 121, 125 (1979) (attorney-client privilege may not be absolute when “there is a compelling need for the information and no alternative source is available”). And the principle of “at-issue” waiver of the attorney-client privilege appears to be well-accepted in American jurisprudence. See Mountain States Tel. & Tel. v. DiFede, 780 P.2d 533, 543 (Colo. 1989) (citing cases); Developments in the Law-Privileged Communications, 98 HARV. L. REV. 1450, 1637 (1985).
The defendants suggest we adopt the test for at-issue waivers of the attorney-client privilege articulated in Hearn v. Rhay, 68 F.R.D. [370]*370574 (E.D. Wash. 1975). In Hearn, the court ruled that an implied waiver of the privilege is appropriate when: (1) assertion of the privilege was a result of some affirmative act, such as filing suit, by the asserting party; (2) the party asserting the privilege placed the protected information at issue; and (3) application of the privilege would deny the opposing party information vital to its case. Id. at 581; accord Mountain States Tel. & Tel., 780 P.2d at 543-44; Conkling v. Turner, 883 F.2d 431, 434 (5th Cir. 1989).
The Hearn test has been criticized for neglecting fairness concerns and the systemic importance of the attorney-client privilege itself. See Remington Arms Co. v. Liberty Mut. Ins. Co., 142 F.R.D. 408, 413-14 (D. Del. 1992); see also Davidson & Voth, Waiver of the Attorney-Client Privilege, 64 OR. L. REV. 637, 646-50 (1986). As one court has observed, the Hearn test “ignores the rationale of the attorney-client privilege: to encourage those in need of legal counsel to seek it in an atmosphere of candor and confidence.” Koppers Co., Inc. v. Aetna Cas. and Sur. Co., 847 F. Supp. 360, 363 (W.D. Pa. 1994).
We find merit in these criticisms of the Hearn standard. We therefore limit the extent of an at-issue waiver of the attorney-client privilege to circumstances “in which the privilege-holder injects the privileged material itself into the case.” Marcus, The Perils of the Privilege: Waiver and the Litigator, 84 MICH. L. REV. 1605, 1633 (1986); cf. Nelson, 130 N.H. at 110, 534 A.2d at 722 (waiver of physician-patient privilege extends only to information relevant to plaintiff’s claim). When the party asserting the privilege has injected privileged material into the case, such that the information is actually required for resolution of the issue, the privilege-holder “must either waive the attorney-client privilege as to that information or . . .be prevented from using the privileged information to establish the elements of the case.” Remington Arms Co., 142 F.R.D. at 415.
In this case, the trial court ordered the plaintiffs to produce “all documents . . . relative to the underlying case.” In light of the standard announced today, this ruling may be overbroad. Accordingly, we vacate and remand to the trial court to determine in the first instance whether the plaintiffs, by initiation of their claim, have necessarily injected privileged material into the case and, further, whether such information is actually required for resolution of the plaintiffs’ claim. With respect to the plaintiffs’ assertion that certain documents are protected under the work product doctrine, we remand to the trial court to determine whether the defendants have demonstrated a substantial need for the requested materials, and [371]*371also whether the defendants could not without undue hardship obtain the materials by other means. See SUPER. CT. R. 35(b)(2).
Vacated and remanded.
THAYER, J., dissented; the others concurred.