Leading up to the passage of the Patient Protection and Affordable Care Act in March 2010, federal liability reforms were contemplated as a means of garnering support for the legislation among congressional Republicans and medical professional organizations. Although no liability-reform provisions survived in the final bill, Congress made clear the need for more experimentation. The final legislation authorized $50 million for states and health care systems to test new approaches to the resolution of medical-injury disputes. This authorization supplemented the $23 million that the Agency for Healthcare Research and Quality (AHRQ) awarded in 2010 for projects to advance new approaches to medical-injury compensation and patient safety" (Kachalia and Mello, 2011, p. 1564).Which of the medical liability reform approaches described in Tables 2 and 3 of Kachalia and Mello (2011) do you favor? Why?h e a l t h p o l i c y r e p o r t T h e n e w e ngl a nd j o u r na l o f m e dic i n e n engl j med 364;16 nejm.org april 21, 20111564 New Directions in Medical Liability Reform Allen Kachalia, M.D., J.D., and Michelle M. Mello, J.D., Ph.D. Medical liability reform has maintained a tena- cious hold on the national policy agenda. During the first several years of the 21st century, a mal- practice insurance “crisis” prompted vociferous demands by organized medicine and liability in- surers for tort reforms to curb litigation costs.1 Many observers anticipated that once the insur- ance market calmed, so too would calls for re- form. Instead, a new force for change emerged — health care reform.2,3 Leading up to the passage of the Patient Pro- tection and Affordable Care Act in March 2010, federal liability reforms were contemplated as a means of garnering support for the legislation among congressional Republicans and medical professional organizations.3 Although no liability- reform provisions survived in the final bill, Con- gress made clear the need for more experimen- tation. The final legislation authorized $50 million for states and health care systems to test new approaches to the resolution of medical-injury disputes.4 This authorization supplemented the $23 million that the Agency for Healthcare Re- search and Quality (AHRQ) awarded in 2010 for projects to advance new approaches to medical- injury compensation and patient safety.5 As policymakers’ attention has shifted from winning passage of the health reform bill to de- termining how to implement and pay for it, medical liability reform has remained a focus because of its perceived potential to help “bend the health care cost curve.”6 Liability risks and costs are often cited as drivers of higher health care spending, poorer access to care, and lower quality of care.6,7 Long-standing criticisms of the tort system’s performance — for example, that it compensates only a small proportion of negligently injured patients while generating un- acceptably high overhead costs — and skepti- cism that it has helped improve patient safety add fuel to the fire.8,9 These developments make it timely to review what is known about the effectiveness of various strategies for liability reform and the implications for the future direction of reform. Historically, liability reform has largely been aimed at reduc- ing insurance costs for health care providers.10 It has taken place almost entirely at the state level and has drawn repeatedly on the same set of legislative modifications to tort law. Today, reform is taking place outside of state legisla- tures through federal sponsorship of voluntary policy experiments led by hospital systems, liabil- ity insurers, and state agencies. The experiments target both liability cost control and patient- safety improvement. This transition is a welcome change in light of a mounting body of evidence demonstrating the limitations of traditional ap- proaches to reform. Measuring the Effec tiveness of Medic al Liabilit y Reforms Assessing the effectiveness of liability reforms first requires identification of relevant evalua- tion metrics. Though not an exhaustive list, the metrics presented in Table 1 can be assessed with the use of available data. Discussions about malpractice reform often start with physicians’ and insurers’ complaints about the system, which include the high cost of malpractice insurance coverage, the number of nonmeritorious suits, the size and unpredictabil- ity of jury awards, and the inefficiency of litiga- tion as a mechanism for resolving disputes. Each of these complaints finds an empirical ba- sis in studies of malpractice claims11,12 and in the volatility seen in malpractice premiums over the past 10 years.13 Yet patients and attorneys also reasonably object that the current tort sys- tem is hard for many injured patients to access, takes an unreasonable amount of time and ex- pense to deliver compensation, and often results in different litigation outcomes for patients with similar injuries.9 The best estimates are that only 2 to 3% of patients injured by negligence file claims, only about half of claimants recover money, and litigation is resolved discordantly n engl j med 364;16 nejm.org april 21, 2011 1565 health policy report with the merit of the claim (i.e., money is award- ed in nonmeritorious cases or no money is awarded in meritorious cases) about a quarter of the time.11,12 Thus, from the perspectives of these stakeholders, evaluations of system reforms should consider the frequency with which claims are brought, the amounts that plaintiffs receive, the amounts that are lost to overhead expenses in the litigation process, and the ways in which these factors translate into insurance premiums. These liability-focused measures have long been at the center of tort-reform evaluations. Less studied, but now receiving greater attention, are measures of how the liability system affects clinical care.14 Today, the pressing need to im- prove quality and efficiency in health care man- dates that any liability reform also be evaluated on the basis of clinically relevant metrics. Care- related metrics include those that assess how the liability system affects the cost, quality, and availability of health care (Table 1). The tort system can affect clinical care either by design or because of unintended consequences. A key purpose of the liability system is to en- courage health care providers to deliver care at a socially optimal level of safety. A well-function- ing liability system thus should encourage insti- tutions to adopt safer systems and should spur individual providers to use greater care in prac- tice. These investments, in turn, should result in fewer adverse events and higher-quality care. An oppressive liability environment, on the other hand, can have the unintended effect of “over- deterrence” — causing unwanted provider prac- tices aimed primarily at avoiding liability.7,15 These practices include defensive medicine, in which providers avoid high-risk patients or ser- vices or order extra tests, referrals, and services primarily to reduce their liability risk. A liability reform may perform very different- ly across liability-related and care-related criteria, and improvements in some measures may come Table 1. Metrics for Assessing the Performance of Medical Liability Reforms. Metric Description Problems in the Current System Liability measures Claims frequency The number of malpractice claims filed Physicians and insurers perceive claims frequen- cy as excessive, yet only 2 to 3% of patients who are injured by negligence file claims. Indemnity costs Settlement and verdict amounts among paid claims Awards in some cases are very high. Awards are highly variable across similar cases. Although total claims costs are high, most injured pa- tients receive no compensation. Overhead costs Administrative expenses associated with pursuing and defending litiga- tion and running liability-insurance companies System overhead costs consume an estimated 55% of each malpractice premium dollar. Malpractice insurance costs The premiums paid by health care pro- viders for malpractice insurance coverage Premiums vary broadly by specialty and geograph- ic region, and they can exceed $250,000 per year. Care-related measures Defensive medicine Ordering of tests, referrals, and other services primarily, though not solely, to reduce liability risk; or avoidance of high-risk services or patients Although defensive medicine accounts for only a small proportion of total health care spending, the amount is large in absolute terms (>$45 billion annually). Physician supply The availability of physician services in a state High insurance costs and liability risk may create social inefficiencies if they cause competent physicians to stop practicing medicine, re- duce their scope of practice, or avoid high- risk locations or patient groups. Quality of care The quality of care that patients receive, as indicated by patient outcomes or other measures Defensive practices or loss of trust in the physician– patient relationship due to liability pressure may lead to a lower quality of care. Evidence suggests that the current system does not pro- vide a strong incentive to avoid negligent care. T h e n e w e ngl a nd j o u r na l o f m e dic i n e n engl j med 364;16 nejm.org april 21, 20111566 at the expense of others. For example, caps on damages may be successful in lowering premi- ums and reducing defensive practices, but they may also weaken the incentive to practice safely, resulting in decreased quality of care. Assess- ments of the efficacy of reforms thus need to consider the entire picture to properly calculate the overall value of reforms for all stakeholders. The Evidence on Tr aditional Tort Reforms In step with legislative attention to medical lia- bility reform, there has been a surge in interest among academic researchers in evaluating how well traditional tort-reform laws have worked. The number of well-designed studies has tripled or quadrupled over the past several years. As a re- sult, the body of reliable evidence on which to base policy decisions has grown dramatically. We recently performed a comprehensive review of rel- evant studies published through 2009 in medical, economics, and law journals, as well as reports issued by government agencies and philanthrop- ic foundations.14,16 The review examined studies of eight of the most widely adopted state tort re- forms, which are presented and defined in Table 2. Our review yielded two main conclusions. First, evaluations of traditional tort reforms have remained heavily focused on metrics related to liability costs, with most care-related measures receiving relatively short shrift. Second, the evi- dence reveals that, with few exceptions, tradi- tional tort reforms have not proved to provide many improvements in these liability metrics (Table 3). Caps on noneconomic damages, which limit the amounts that juries may award for pain and suffering, constitute the main exception to the above rule. Studies have nearly uniformly found that caps are an effective means of reducing the size of indemnity payments.17-26 Studies of their effects on premiums have returned mixed find- ings, but a reasonable conclusion based on meth- odologically strong studies using recent data is that caps modestly constrain the growth of in- surance premiums over time.18,23,24,26-31 Evidence concerning the effect of caps on claims frequen- cy is too equivocal to permit a firm conclu- sion,17,25,26,33 and only one study has examined their effect on overhead costs.32 Statutes of limitation and repose, which re- strict the amount of time that plaintiffs have to file a lawsuit, have been associated with lower premiums,18,26,28,30 but the mechanism of effect is unclear, because they are not significantly as- sociated with lower frequency of claims, indem- nity costs, or overhead expenses. Studies of pre- trial screening panels, certificate-of-merit (COM) requirements, limits on attorneys’ fees, joint-and- several liability (JSL) reform, collateral-source rule reform, and periodic payment generally have not identified significant effects on claims fre- quency, indemnity costs, or insurance premi- ums.14 The evidence base for the findings can generally be characterized as moderate to high, considering the quantity and quality of available studies and the consistency of results across studies. The evidence concerning overhead costs is more limited. There is some evidence that screening panels, COM requirements, and fee limits result in increased overhead costs — the first two because they interject an additional layer of procedural requirements into the litiga- tion process and the last possibly because it leads plaintiff’s attorneys to take on more com- plex cases — but this evidence is not conclusive. For all eight traditional reforms, the evidence on care-related metrics is fairly sparse overall. However, the effects of caps on damages, JSL reform, and collateral-source rule reform on de- fensive medicine have been well studied. Caps are associated with lower rates of defensive med- icine, whereas all studies of collateral-source rule reform have found no effect, and findings concerning JSL reform have been mixed.34,35,47,49,50 Little or no evidence is available concerning the effects of the other five reforms on defensive medicine. There is some evidence that caps on dam- ages modestly increase the supply of physicians in a state, although study findings have been mixed.14,29,38-41 There is moderately strong evi- dence that limits on attorneys’ fees,41,42 JSL re- form,41,42 collateral-source rule reform,41,51,52 and periodic payment41,42 do not significantly affect physician supply. Little is known about the effects of the other three reforms on physician supply. Notably, none of these eight reforms have been extensively studied for their effect on the quality of care. A handful of studies have exam- ined limited patient outcomes as proxy measures, and all but one have found no significant asso- ciation.34-36,42 Overall, evidence demonstrating that traditional reforms improve clinical care is lacking. n engl j med 364;16 nejm.org april 21, 2011 1567 health policy report Current Focus on Patient Safet y and Efficienc y of