COLLECTIVE REVIEW
Impact of Medical Malpractice Environment on Surgical Quality and Outcomes Christina A Minami, MD, Jeanette W Chung, Karl Y Bilimoria, MD, MS, FACS
PhD,
Jane L Holl,
a systematic review of the literature to examine the association between malpractice environment and outcomes in surgical specialties.
The annual cost of the medical malpractice system has been estimated to be $55 billion.1 The medical malpractice crisis that began unfolding in the late part of the 20th century continues to be a major concern.2-5 The intent of the malpractice system, based on classic tort deterrence theory, assumes that the looming threat of a malpractice suit will deter poor care because providers will be more vigilant and responsible.6 Proponents of the liability system believe that the threat of malpractice suits will encourage providers to adhere to standards of care, which, in turn, should lead to better patient outcomes. However, critics argue that the system is ineffective in encouraging quality and better outcomes and that it generates only unintended consequences, such as defensive medicine, which may lead to worse patient outcomes because of overuse of services or avoidance tactics that may result in overtesting, overtreatment, iatrogenic injury, and soaring health care costs.7,8 In addition to the problems of malpractice insurance availability and affordability, personal strain from experiencing a malpractice claim can affect clinicians. A recent study found that onequarter of American surgeons had been the subject of a malpractice suit over the preceding 2 years, resulting in notable stress personally and professionally.9 Perceived malpractice risk, real or not,10 may force providers to leave high malpractice risk environments, particularly those in high risk fields of medicine, which, in turn, affects access to care and could potentially adversely affect patient outcomes.3,11,12 Surgical specialties are among the fields of medicine with a large proportion of malpractice claims13-18 and have been a primary focus of the malpractice crisis. The objective of this study was to perform
METHODS Search strategy The literature search included Medline, PubMed, and the Cochrane Database of Reviews of Effectiveness to capture articles published between 1980 and 2012. A comprehensive search using terms that included medical malpractice, tort reform, quality of care, outcomes, and litigation in relation to these surgical and procedural subspecialties was performed. Studies on surgical subspecialties and procedure-based fields, including interventional cardiology/cardiac surgery, urology, neurosurgery, obstetrics and gynecology (OB/GYN), general surgery, and orthopaedic surgery were included. A manual secondary search of reference papers cited in these initial studies was used to expand the database and capture studies published in journals outside the range of Medline and PubMed (eg, law and economic journals). Study selection (exclusion and inclusion criteria) The literature search included review articles, observational studies, case reports, survey studies, and retrospective data analyses. To be included, a study had to be an original research article, written in English, published between 1980 and 2012 in peer-reviewed journals, and performed in the United States. Only those that demonstrated some link or lack thereof between malpractice environment and physician practice or patient outcomes were included. Those that focused solely on claims prevalence, judgments, or amounts and their association with various adverse events were excluded. Studies captured in the initial search and reported on invasive procedures (eg, cardiac catheterization or endoscopy) in addition to surgical operations were retained.
Disclosure Information: Nothing to disclose. Received July 23, 2013; Revised August 30, 2013; Accepted September 16, 2013. From the Surgical Outcomes and Quality Improvement Center, Department of Surgery, Feinberg School of Medicine, Northwestern University and Northwestern Memorial Hospital (Minami, Chung, Holl, Bilimoria), and the Center for Healthcare Studies, Feinberg School of Medicine, Northwestern University, Chicago, IL. Correspondence address: Karl Y Bilimoria, MD, MS, FACS, Surgical Outcomes and Quality Improvement Center, Department of Surgery, Feinberg School of Medicine, Northwestern University, Northwestern Memorial Hospital, 676 St Clair St, Arkes Pavilion Suite 6-650, Chicago, IL 60611. email: [email protected]
ª 2014 by the American College of Surgeons Published by Elsevier Inc.
MD, MPH,
Data extraction Data from each study regarding data sources, study population, and endpoints were abstracted. Because of the heterogeneity in methods, patients, fields, interventions,
271
ISSN 1072-7515/13/$36.00 http://dx.doi.org/10.1016/j.jamcollsurg.2013.09.007
272
Minami et al
J Am Coll Surg
Impact of Medical Malpractice Environment
Table 1. Conceptual Framework for Examining the Effect of Malpractice Environment on Physician Practice and Patient Outcomes Intended consequences
Unintended consequences
Physician practice Hypothesis/argument: Hypothesis/argument: Liability threat creates incentives for physicians to adhere Liability threat creates incentives for providers to (a) to prevailing standards, resulting in high quality of care. substitute low-risk procedures for high-risk procedures; (b) avoid high-risk patients; and (c) provide clinically unnecessary tests and procedures due to fear of missed diagnoses. Hypothesis/argument: Patient outcomes Hypothesis/argument: Liability threat is associated with poorer patient outcomes Liability threat encourages physician adherence to because (a) patients who may benefit or need procedures standards of care and conscientiousness, resulting in perceived to be "high risk" won’t get the care that they better patient outcomes. need. They will either do without care or travel further for it. This may lead to poorer outcomes; (b) risky patients may experience worse outcomes if they can’t get the care they need because physicians fear poor outcomes and liability; and (c) overprovision of care (testing and treatment) may increase risk of iatrogenic injury, resulting in poorer outcomes.
and measured endpoints, data pooling was not possible and a narrative data summary was instead completed. Due to the nature of this research, no randomized controlled studies were found; data were from only retrospective cohort and survey studies. Analysis The PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) checklist was used to structure the review. Data pooling was not possible due to the heterogeneity of study populations and endpoints, and a formal meta-analysis was not performed. This article represents a systematic review. A 2 2 classification was used to categorize studies based on whether the study focused on the effects of physician practice or patient outcomes, and whether the effects were intended or unintended consequences of the malpractice liability system. This framework is built on the opposing arguments regarding the intent of the liability system and the resulting effects. The malpractice system is intended to deter negligence and improve patient outcomes. This approach resulted in 4 categories of studies (Table 1):
1. Studies of the intended consequences of malpractice liability on physician practice were those that identified an association between some measure of malpractice risk (eg, laws, claims frequency, claims severity) and an improvement in a measure of adherence to an explicit standard of care or a performance on a standard quality indicator. 2. Studies of the intended consequences of malpractice liability on patient outcomes were those that directly
tested the hypothesis that malpractice risk improves patient outcomes as measured by mortality, other outcomes, or patient satisfaction. 3. Studies of the unintended consequences of malpractice liability on physician practice were those that directly tested whether some measure(s) of malpractice risk were associated with greater use of defensive medical practices (eg, imaging, ordering of laboratory tests, substitution of less risky procedures for more risky procedures, patient selection [ie, avoidance of highrisk complicated patients], referrals). Studies using self-reported survey data on defensive practices were included in this group. 4. Studies of the unintended consequences of malpractice liability on patient outcomes were those that directly tested whether some measure of malpractice environment was associated with poorer patient outcomes (possibly as a result of defensive medicine).
RESULTS The initial literature search identified 349 studies, of which 94 were excluded immediately by their title, given their irrelevance to malpractice and/or surgical and nonsurgical procedures (Fig. 1). Further review excluded 92 studies, based on their international nature and primary focus on nonsurgical topics. A total of 163 studies were further scrutinized and, after final application of inclusion and exclusion criteria, 29 final studies were included in the systematic review. These 29 studies examined the intended and unintended effects of malpractice environs on changes in physician practice and patient outcomes.
Minami et al
Vol. 218, No. 2, February 2014
349 titles/abstracts identified by initial search (both literature search and reference lists)
94 excluded (topic not malpractice or surgery)
255 studies selected and ordered
92 excluded (international studies, nonsurgical topics)
163 reviewed
134 excluded -24 addressed tort reform in general -15 not related to malpractice - 47 not related to outcomes or changes in practice - 24 were review/background articles -24 addressed claims data only
29 included
Figure 1. Flowchart of eligible studies addressing the association between medical malpractice environment and surgical quality.
Physician practice measures included use of physician services (n ¼ 2), physician supply (n ¼ 5), procedure rates (n ¼ 18), and test ordering (n ¼ 7); patient outcomes examined included adverse events (n ¼ 1) and survival (n ¼ 4). Seven studies were analyses of both medical and surgical data (as opposed to solely surgical data), but included surgical/procedural subset analyses.3,12,19-23 Three cardiology studies were also included due to their interventional/procedural-based nature. Obstetrics (n ¼ 13), neurosurgery (n ¼ 2), urology (n ¼ 1), orthopaedic surgery (n ¼ 1), and vascular surgery (n ¼ 1) were the surgical subspecialties represented. There was considerable variation in the procedures and results of the included studies (Appendix, online only). Data sources included multistate (n ¼ 14), single-state, or single-institution databases (n ¼ 10), as well as survey studies (n ¼ 5). A distinction was made between studies that evaluated patient outcomes (n ¼ 4; eg, mortality, fetal well-being) and those that reflected changes in physician practice thought to be indicative of defensive
Impact of Medical Malpractice Environment
273
medicine (n ¼ 28; eg, use of physician services, test ordering, likelihood of proceeding with cesarean section). Liability premiums were the most commonly chosen indicator of malpractice environment though physicians’ past claims experience and malpractice payments were also used in some studies. Four of the 9 studies that used premiums as a proxy for malpractice risk, used Medical Liability Monitor (MLM) data, an annual nationwide survey of physician malpractice insurance premiums. Other sources included data from the Urban Institute Survey and state-level insurance underwriters. Data on malpractice claims payments came from the National Practitioner Data Bank (NPDB), a record of all payments made on behalf of physicians for medical liability settlements or judgments. Given the limited number of studies that addressed our specific aims, there was a wide range of study design quality and statistical rigor. In addition, some studies were more rigorous in controlling for the temporal relationship between exposure and outcomes than others, and the sophistication of economic modeling varied widely. Intended consequences e physician practice As the Appendix (online only) demonstrates, no studies directly tested whether greater levels of malpractice risk were associated with greater physician adherence to standards of care. There were no studies that showed a direct association between malpractice risk and subsequent changes in national measures of quality-of-care or patient safety. Intended consequences e patient outcomes Three studies attempted to directly test the association between the malpractice system and patient outcomes. Baicker and colleagues19 investigated the effect of malpractice liability on the use of physician services and mortality using data on Medicare fee-for-service beneficiaries from 1993 to 2001.19 Although state malpractice liability indicators appeared to be associated with increased Medicare spending and malpractice costs, the authors found no association between average malpractice payments per physician capita or average premiums and mortality in this patient population. Similarly, Kessler and McClellan24 looked at Medicare patients treated for acute myocardial infarction and found that there was no difference in mortality in this population by malpractice environment. Dhankhar and associates,25 however, found that increased malpractice risk led to lower mortality among acute myocardial infarction patients using the Nationwide Inpatient Sample (NIS).25
274
Minami et al
Impact of Medical Malpractice Environment
Unintended consequences e physician practice All 29 studies included in this review addressed the association between malpractice risk and defensive medicine practices. These studies explore the prevailing hypothesis that malpractice pressure leads to defensive medicine, which may manifest as avoidance tactics (ie, steering clear of risky patients and procedures) or assurance tactics (eg, ordering more diagnostic tests). Of the 29 studies, 5 were surveys of clinicians and 24 were retrospective studies. The 5 survey studies20,21,26-28 investigated the link between physician perceptions of risk and subsequent clinical actions and/or decision-making. Studdert20 and Glassman and coworkers21 examined the relationship between physician practice and malpractice history or perceived malpractice risk. Both found that physicians’ past liability exposure was not associated with defensive practices. Individual physicians’ perception of risk, however, did influence their treatment decisions. Glassman and colleagues21 found that although physicians’ past experiences with the legal system did not correlate with patterns of resource use (eg, testing, decision to admit or discharge, and decision to proceed with or delay surgery), their self-perceived claims risk did affect resource use. Four studies found that a significant proportion of physician respondents admitted to ordering unnecessary tests due to malpractice concerns.20,26-28 The large 2005 survey of defensive practices by Studdert20 included a sample of 824 emergency medicine physicians, general surgeons, neurosurgeons, obstetrician/gynecologists, orthopaedic surgeons, and radiologists from the American Medical Association Physician Masterfile and found that 93% of physicians reported practicing defensive medicine. Specialists who lacked confidence in their insurance coverage were more likely to order unnecessary diagnostic testing, refer to other physicians unnecessarily, and suggest invasive procedures. Subgroup analyses revealed field-specific trends, with high rates of possibly unnecessary referral to other physicians among OB/GYNs (59%) and high rates of unnecessary invasive procedures among general surgeons (44%). A survey of physicians in orthopaedic surgery by Sethi and coauthors26 showed that defensive imaging represented 19% to 24% of all imaging ordered. In a study of neurosurgeons, Nahed and colleagues27 reported that a majority of respondents engaged in defensive medicine practices, such as unnecessary testing and referrals. In addition, respondents reported that liability concerns limited their performance of high-risk procedures for traumatic head and spine injuries, intracranial hemorrhages, tumor resections, hydrocephalus (45%), as well as the duration of their clinical practice (71%). Lucas and colleagues28 focused
J Am Coll Surg
on cardiac catheterization and found that 24% of general practitioners reported ordering a cardiac catheterization due to malpractice fears.28 Although survey research has established a correlation between physician perception of malpractice risk and subsequent practice patterns and treatment decisions, 18 observational studies offered an alternate approach to examining this issue. Fourteen of these studies came from the obstetrics literature, examining cesarean section rates, vaginal birth after cesarean (VBAC) rates, and late pre-term induction.29-42 The majority of data on the effect of malpractice premiums on cesarean section rates found increased rates of cesarean section with increased malpractice premiums,29,31,32,35,36,39,42,43 while only 1 study found the opposite,33 and 2 found no significant effect of state malpractice premiums on cesarean section rates.29,41 Higher premiums were associated with higher rates of late pre-term induction37 and with lower vaginal birth after cesarean rates and instrumental deliveries.31,32 Five other studies examined procedures beyond obstetrics. Konety and associates44 found that greater proportions of patients with stage III and IV bladder cancer underwent radical cystectomy, a high-risk procedure, in regions with malpractice payment caps. A study by Dranove and Gron45 focused on craniotomy and cesarean section rates in Florida during the malpractice crisis period of the early 2000s, and found that after Florida premiums rose precipitously, many neurosurgeons decreased their volume of brain operations, and craniotomy patients had increased travel times to definitive care. Women undergoing highrisk deliveries, however, did not see increases in travel times. A 2004 study by Baicker and coworkers46 used Medicare data to examine the use of diagnostic testing as well as rates of 6 specific procedures: angiography, coronary artery bypass graft, percutaneous coronary intervention, radical prostatectomy, and transurethral prostatectomy. Little evidence of net increases in the procedure rates was found in response to state-level increases in malpractice costs. Dhankar and associates,25 when evaluating resource use in the treatment of acute myocardial infarction, found decreased resource use in their patient population (ie, in environments with greater malpractice risk, patients were more likely to undergo medical management as opposed to angioplasty). Similarly, using the Nationwide Inpatient Samples and State Inpatient Databases, Ng and colleagues47 found that increased use of endovascular aneurysm repair (EVAR), a less invasive procedure than open repair, increased with the number of paid claims and mean malpractice premiums. Surgeon supply in response to the presence of state malpractice crises or liability reforms has also been assessed.
Vol. 218, No. 2, February 2014
Minami et al
Unlike the study by Dranove and Gron,45 the analysis by Mello and coauthors3 of physician supply in Pennsylvania before and after the crisis of the early 2000s found little change. The presence of damage caps, however, was associated with an increase in local supply of surgical specialists in 2 studies.22,23 Klick and Stratmann12 came to a similar conclusion, although the “high-risk” specialties group in their analysis included nonsurgical fields, such as emergency medicine and general practice. Unintended consequences e patient outcomes Two studies investigated the association between malpractice risk and patient outcomes. Konety and colleagues44 found that the presence of caps on noneconomic damages resulted in an improvement in diseasespecific survival in advanced bladder cancer patients because surgeons were more willing to perform a risky procedure that resulted in better patient outcomes. Focusing on the effects of malpractice risk, Dubay and associates29 used data from the National Natality Files to test whether greater malpractice risk was associated with use of cesarean section delivery and birth outcomes and found that higher malpractice premiums were associated with lower 5-minute Apgar scores.
DISCUSSION Rooted in classic tort deterrence theory, the medical malpractice system is intended to discourage negligent care, encourage physicians to be conscientious and adhere to guidelines, and result in better health care outcomes. However, there may also be unintended consequences in which malpractice liability results in defensive medicine practices. We sought to synthesize the literature to assess what is currently known about the impact of malpractice environment on physician practice and patient outcomes. We found a limited body of literature based on a number of different data sources, surgical subspecialties, geographic regions, and eras. Information and selection bias were, to varying degrees, present in all of the studies. Although some studies consistently demonstrated an association between malpractice environment and defensive medicine, there was less consistency in the data regarding the ability of malpractice liability to affect physician practice and patient outcomes. Physician practice patterns The bulk of the literature on the effect of the malpractice system has focused on examining the impact on provider practice patterns and is based on surveys and empirical assessments. These studies show that physician perception of malpractice pressures can have a measurable effect on
Impact of Medical Malpractice Environment
275
treatment decisions. It may be that the fear of missing an occult pathology or being accused of failing to intervene in areas of greater liability risk are manifested by a propensity to obtain an unnecessary diagnostic workup, even though such interventions carry their own inherent risk. However, survey studies have considerable limitations. First, physicians with strong opinions, likely negative, may be more likely to respond to the survey. Second, survey respondents may exaggerate the effect when they have an opportunity to voice their opinion on a controversial topic. Third, surveys reflect only perceptions, but it is unknown whether this actually reflects the empirical relationship between malpractice environment and physician practice. A large number of empirical assessments have focused on defensive medicine, reflecting the medical community’s clear interest in this aspect of the malpractice system. All 29 studies commented on the consequences of malpractice liability on physician practice, including whether the legal environment can sway a practitioner’s decision to perform a procedure, order diagnostic testing, or limit his or her practice. In aggregate, the data on malpractice influence on procedure use are mixed. In obstetrics, higher malpractice risk may result in obstetricians intervening sooner and more often than otherwise (ie, have a lower threshold for taking a patient for cesarean section) and in fewer women with a previous cesarean section subsequently trying to deliver vaginally. The obstetrics literature showed trends toward increased cesarean delivery rates and decreased rates of vaginal birth after cesarean in higher liability environments, though there was a lack of complete agreement among these 13 studies. Studies of other operations also showed that malpractice environment can affect whether surgeons perform riskier procedures (eg, radical cystectomy). Overall, practice patterns do appear to be affected by malpractice environment, although these empirical studies had several limitations regarding the population examined, time frames, geographic areas, and indicator of malpractice risk (eg, premiums vs caps). Malpractice and outcomes Although many studies sought to profile the types of adverse events that lead to claims, especially in highliability fields like surgery, we failed to find any existing studies that examined the intended consequences of the malpractice system on greater provider adherence to standards of care. The 5 studies 19,24,25,29,44 that examined the consequences of malpractice environment on patient outcomes offered differing conclusions; 2 found that patient outcomes were improved in higher liability environments,25,44 but 3 found that outcomes were worse or no better.19,24,29 These studies focused on individual
276
Minami et al
Impact of Medical Malpractice Environment
procedures, fields of medicine, or diseases using different measures of malpractice environment, making direct comparison difficult. Some of these studies had very limited representation of different geographic areas and those that broadened the geographic breadth of the study population by using Medicare data, are, by default, limited to an aging population. So it is unclear whether malpractice environment has a beneficial effect on patient outcomes. There is undoubtedly a need for a comprehensive assessment of the association between malpractice environment and patient outcomes across multiple surgical fields using various indictors of malpractice environment. There are several issues that may affect the association between malpractice environment and quality. Defense tactics may be classified as “assurance” vs “avoidant” behaviors, the former alluding to the practice of ordering medical services with little clinical value, and the latter referring to practitioner avoidance of certain high-risk patients and procedures.20 Perhaps an extrapolation of these avoidant behaviors is the issue of relative reticence throughout surgical fields regarding adverse events and errors. Although morbidity and mortality conferences are an embedded part of surgical culture, they are, by definition, protected discussions, open only to an institution’s surgical faculty and housestaff. There are disclosure-and-offer programs being piloted in certain states, such as those in Illinois, New York, Texas, and Washington State, which are modeled on a program out of the University of Michigan Health System, but this approach is by no means ubiquitous.48 Another adverse consequence is that specialty, in-demand, or high quality physicians may be driven out of states due to the malpractice environment. These problems may all feed into methodologic difficulties in examining the association between malpractice environment and outcomes, and may make high malpractice risk environments appear to have poorer quality. There have also been longstanding theoretical criticisms of the deterrence theory, the most relevant to medical malpractice being the idea that insurance insulates individual physicians from the economic repercussions of a lawsuit, therefore blunting the deterrent effect and failing to act as a motivator for improvements in patient safety.7 Moreover, only about 2% of negligent care results in a lawsuit.13 These issues may partly explain why no deterrence effect is observed. Methodologic limitations In assessing the effect of local malpractice environment on medical care, it is important to consider data sources and their inherent strengths and weaknesses. Observational studies generally have the advantage of covering
J Am Coll Surg
large populations and time frames. On the other hand, a drawback of observational studies is that they rely on proxy measures for key variables of interest. For example, claims rates, average award size, and premium rates were frequently used as indicators of providers’ perceived liability threat. Yet it is the strength of the correlation between these indicators of perceived liability threat that is empirically unknown, and it is unclear whether a single indicator or a composite measure of malpractice environment would be more meaningful. As a methodologic approach in the malpractice literature, survey studies offer the advantage of direct measurement of exposure to liability risk and levels of perceived malpractice threat. However, there have been few large nationally representative studies of physicians and malpractice, and many of these surveys have been limited with respect to geographic coverage, provider types or specialties, and/or time periods. Additional drawbacks of survey research are the potential for selection bias in response rates (if physician responses are correlated with underlying malpractice risk) and the potential for respondents to provide socially desirable answers that may not accurately reflect their actual behavior or intentions to act. These inherent strengths and weaknesses shed some light on why stronger associations between malpractice environment and physician behavior that have been typically uncovered by survey studies.
CONCLUSIONS The existing body of data on the effect of malpractice environs on surgical outcomes and quality of care is small, equivocal, and has notable temporal and geographic limitations. Though there is evidence that medical malpractice liability influences physicians’ clinical choices, there is little to support the theory that the threat of medical litigation improves physician adherence to quality care indicators or improves patient outcomes. To facilitate a productive discussion on the malpractice system and tort reform, we need to better understand whether the liability system in the US actually serves its intended purpose of being a deterrent to negligent care and improving safety and quality. A comprehensive evaluation of the effect of malpractice environment on quality of care is needed. Author Contributions Study conception and design: Minami, Bilimoria Acquisition of data: Minami, Bilimoria Analysis and interpretation of data: Minami, Chung, Holl, Bilimoria Drafting of manuscript: Minami, Chung, Bilimoria Critical revision: Holl, Bilimoria
Vol. 218, No. 2, February 2014
Minami et al
REFERENCES 1. Mello MM, Gawande A, Studdert D. National costs of the medical liability system. Health Affairs 2010;29:1569e1577. 2. Paik M, Black B, Hyman D. The receding tide of medical malpractive litigation. 7th Annual Conference on Empirical Legal Studies Paper 2013. 3. Mello M, Studdert D, Brennan T. Changes in physician supply and scope of practice during a malpractice crisis: evidence from Pennsylvania. Health Affairs 2007;26: 425e435. 4. Mello MM. Understanding Medical Malpractice Insurance: A Primer. Available at: http://www.rwjf.org/content/dam/farm/ reports/reports/2006/rwjf17974. Accessed October 8, 2013. 5. Association AM. America’s Medical Liability Crisis: a National View. 2005. http://www.ama-assn.org/ama1/pub/upload/mm/ 450/med_liab_jan07.pdf. Accessed October 8, 2013. 6. Bell P. Legislative intrusions into the common law of medical malpractice: thoughts about the deterrent effect of tort liability. Syracuse Law Review 1984;XXXV:939e993. 7. Mello M, Brennan T. Deterrence of medical errors: theory and evidence for malpractice reform. Tex Law Rev 2002;80: 1595e1637. 8. Studdert D, Mello M, Brennan T. Defensive medicine and tort reform: a wide view. J Gen Intern Med 2010;25: 380e381. 9. Balch CM, Oreskovich MR, Dyrbye LN, et al. Personal consequences of malpractice lawsuits on American surgeons. J Am Coll Surg 2011;213:657e667. 10. Carrier E, Reschovsky J, Mello M, et al. Physicians’ fears of malpractice lawsuits are not assuaged by tort reforms. Health Affairs 2010;29:1585e1592. 11. Mello MM, Studdert DM, DesRoches CM, et al. Caring for patients in a malpractice crisis: physician satisfaction and quality of care. Health Aff (Millwood) 2004;23:42e53. 12. Klick J, Stratmann T. Medical malpractice reform and physicians in high-risk specialties. J Legal Studies 2007;36: S121eS142. 13. Brennan T, Leape L, Laird N. Incidence of adverse events and negligence in hospitalized patients: results of the Harvard Medical Practice Study I. N Engl J Med 1991;324: 370e376. 14. Kane C. Policy research perspectivesemedical liability claim frequency: a 2007-2008 snapshot of physicians. American Medical Association 2010:1e7. http://www.ama-assn.org/resources/doc/ health-policy/prp-201001-claim-freq.pdf. Accessed October 8, 2013. 15. Studdert D, Mello M, Gawande A, et al. Claims, errors, and compensation payments in medical malpractice litigation. N Engl J Med 2006;354:2024e2033. 16. Jena AB, Seabury S, Lakdawalla D, Chandra A. Malpractice risk according to physician specialty. N Engl J Med 2011; 365:629e636. 17. Sloan F, Mergenhagen P, Burfield W, et al. Medical malpractice experience of physicians: predictable or haphazard? JAMA 1989;26:3291e3297. 18. Orosco R, Talamini J, Chang D, Talamini M. Surgical malpractice in the United States, 1990-2006. J Am Coll Surg 2012;215:480e488. 19. Baicker K, Fisher ES, Amitabh C. Malpractice liability costs and the practice of medicine in the Medicare program. Health Affairs 2007;26:841e852.
Impact of Medical Malpractice Environment
277
20. Studdert D. Defensive medicine among high-risk specialist physicians in a volatile malpractice environment. JAMA 2005;293:2609e2617. 21. Glassman PA, Rolph JE, Petersen LP, et al. Physicians’ personal malpractice experiences are not related to defensive clinical practices. J Health Polit Policy Law 1996;21:219e241. 22. Matsa D. Does malpractice liabiity keep doctors away? Evidence from tort reform damage caps. J Legal Studies 2007;36:S143eS182. 23. Encinosa W, Hellinger F. Have state caps on malpractice awards increased the supply of physicians? J Health Affairs 2005. Suppl Web Exclusives: W5-250-W5-258. 24. Kessler D, McClellan M. Do doctors practice defensive medicine? Q J Econ 1996;111:353e390. 25. Dhankar P, Khan M, Bagga S. Effect of medical malpractice on resource use and mortality of AMI patients. J Empirical Legal Studies 2007;4:163e183. 26. Sethi MK, Obremskey WT, Natividad H, Mir H, Janhgir A. Incidence and costs of defensive medicine among orthopedic surgeons in the United States: a national survey study. Am J Ortho 2012;41:69e73. 27. Nahed BV, Babu MA, Smith TR, Heary RF. Malpractice liability and defensive medicine: a national survey of neurosurgeons. PloS one 2012;7:e39237. 28. Lucas FL, Sirovich BE, Gallagher PM, et al. Variation in cardiologists’ propensity to test and treat: is it associated with regional variation in utilization? Circulation Cardiovascular quality and outcomes 2010;3:253e260. 29. Dubay L, Kaestner R, Waidmann T. The impact of malpractice fears on cesarean section rates. J Health Econ 1999;18: 491e522. 30. Gimm GW. The impact of malpractice liability claims on obstetrical practice patterns. Health services research 2010; 45:195e211. 31. Zwecker P. Effect of fear of litigation on obstetric care: a nationwide analysis on obstetric practice. Am J Perinatol 2011;28:277e284. 32. Yang T, Mello M, Subramanian S, Studdert D. Relationship between malpractice litigation pressure and rates of cesarean section and vaginal birth after cesarian section. Medical Care 2009;47:234e242. 33. Tussing A, Wojtowycz M. The cesarean decision in New York State, 1986. Medical Care 1992;30:529e540. 34. Sloan F, Entman S, Reilly B, et al. Tort liability and obstetricians’ care levels. Int Review Law Economics 1997;17:245e260. 35. Rock S. Malpractice premiums and primary cesarean section rates in New York and Illinois. Public Health Rep 1988; 103:459e463. 36. Murthy K. Association between rising professional liability insurance premiums and primary cesarean delivery rates. Obstet Gynecol 2007;110:1264e1269. 37. Murthy K. Obstetricians’ rising liability insurance premiums and inductions at late preterm gestations. Med Care 2009; 47:425e430. 38. Kim B. The impact of malpractice risk on the use of obstetrics procedures. J Legal Studies 2007;36:S79eS120. 39. Baicker K, Buckles K, Chandra A. Geographic variation in the appropriate use of cesarean delivery. Health Affairs 2006; 2006. 40. Baldwin L, Hart L, Lloyd M, et al. Defensive medicine and obstetrics. JAMA 1995;274:1606e1610.
278
Minami et al
Impact of Medical Malpractice Environment
41. Edwards C. The impact of a no-fault tort reform on physician decision-making: a look at Virginia’s Birth Injury Program. Revista Juridica UPR 2010;80:285e310. 42. Localio A, Lawthers A, Bengtson J, et al. Relationship between malpractice claims and cesarean delivery. JAMA 1993;269:366e373. 43. Ryan K, Schnatz P, Greene J. Change in cesarean section rate as a reflection of the present malpractice crisis. Conn Med 2005;69:139e141. 44. Konety B, Dhawan V, Allareddy V, Joslyn S. Impact of malpractice caps on use and outcomes of radical cystectomy for bladder cancer: data from the surveillance, epidemiology, and end results program. J Urol 2005;173:2085e2089.
J Am Coll Surg
45. Dranove D, Gron A. Effect of the malpractice crisis on access to and incidence of high-risk procedures: evidence from Florida. Health Affairs 2005;24:802e810. 46. Baicker K, Chandra A. The Effect of Malpractice Liability on the Delivery of Health Care. Cambridge, MA: National Bureau of Economic Research; 2004. 47. Ng TT, Mirocha J, Magner D, Gewertz BL. Variations in the utilization of endovascular aneurysm repair reflect population risk factors and disease prevalence. J Vasc Surg 2010;51: 801e809. 48. Kachalia A, Mello M. New directions in medical liability reform. N Engl J Med 2011;364:1564e1572.
Summary of Reviewed Studies Examining the Effect of Malpractice Environment on Physician Practice and Patient Outcomes
Consequences of malpractice liability
First author (y)
None Intended consequences of malpractice liability on physician practice Intended consequences of malpractice liability on patient outcomes Baicker (2007)19
Data sources
NPDB; MLM; Quantify the effect of Medicare Fee-forMedPAR; Service malpractice Medicare Part B liability on the use beneficiaries, Claims 1993-2001 of physician services, Medicare payments, and mortality. Quantify the impact Patients in the NIS NPDB; NIS dataset in 2002 of medical who suffered acute malpractice on MI mortality probability from AMI. Health Care Analyze the effects of Medicare Financing liability reforms on beneficiaries Administration treated for serious Medicare patients enrollment files heart disease in with serious heart 1984, 1987, 1990 disease.
Malpractice environment measures
Outcomes
Key findings
Average dollar value of malpractice payments per physician capita Average premiums
d
Mortality
d
No association between mortality from various causes and measures of malpractice costs.
d
Claims frequency and claims severity in 2000e2002
d
d
d
Malpractice reforms by state
Resource use (medical management vs angioplasty vs bypass surgery) Mortality Mortality
Increased medical malpractice risk leads to improvement in mortality among AMI patients. Malpractice reforms do not lead to consequential differences in mortality.
d
d
d d
d
Unintended consequences of malpractice liability on physician practice (defensive medicine) (Continued)
Impact of Medical Malpractice Environment
Kessler (1996)24
Study population
Minami et al
Dhankhar (2007)25
Study aim
Vol. 218, No. 2, February 2014
Appendix.
278.e1
Continued
Consequences of malpractice liability
First author (y)
Study aim
Study population
Data sources
Malpractice environment measures
NCHS linked birth and infant mortality data; NPDB; MLM
d
Baicker (2007)19
Quantify the effect Medicare Fee-forService of malpractice liability on the use beneficiaries, 1993e2001 of physician services and costs.
NPDB; MLM; MedPAR; Medicare Part B Claims
d
Number and size of malpractice payments per physician d Average malpractice premiums
d
Baldwin (1995)40
Content of Low-risk obstetric Examine effect of Obstetrical Care patients of malpractice Study (COOC), Washington State experience on Washington State providers who prenatal resources Physicians initiated care and cesarean Insurance between 9/1/1988 section rates. Exchange and and 8/31/1989 Association Database (WSPIEA)
d
d
Outcomes
Key findings
d
Cesarean rates
d
Average dollar value of malpractice payments per physician capita Average premiums
d
d
Personal malpractice experience County-based malpractice defendant rate
d
Total Medicare spending/ beneficiary Spending/beneficiary by service type Utilization rates of physician services (screening, diagnostic and imaging procedures) and major elective surgical procedures reported in the Dartmouth Atlas Use of obstetric ultrasound, referral and consultation Delivery method
d
d
d
d
Malpractice environment explained an estimated 14.8% and 13.9% of variation in county-level cesarean rates for normal and low/ very low birth weight deliveries, respectively. A 10% increase in malpractice payments increases the use of carotid duplex, ECHO, EKG, CT and MRI by 1.5%e1.8%.
Impact of Medical Malpractice Environment
US births, Examine whether 1995e1998 geographic variation in appropriate use of cesarean section is correlated to geographic variation in malpractice liability.
Minami et al
Baicker (2006)39
278.e2
Appendix.
No association between malpractice claims experience and increased prenatal resources or cesarean sections in low-risk obstetric patients. (Continued) J Am Coll Surg
Continued
Consequences of malpractice liability
First author (y)
Study aim
Study population
Data sources
Dhankhar (2007)25 Quantify the impact Patients in the NIS NPDB; NIS dataset in 2002 of medical who suffered acute malpractice on MI mortality probability from AMI.
Malpractice environment measures d
Claims frequency and claims severity in 2000-2002
Edwards (2010)41
U-S births, Examine effect of 1990e1992 malpractice claims risk on cesarean section, birth outcomes. Assess the effect of All deliveries in Virginia in 2006 the Birth Injury Program (BIP) on cesarean section rates in Virginia.
d
d
National Natality Files
d
OB/GYN malpractice premiums
d
Virginia Health Information
d
Physicians participating in the BIP vs those who didn’t
d
Key findings
Increase in malpractice risk leads to reduced resource use in patients with less severe medical conditions. d Increase in travel Number of physicians times for performing highcraniotomy risk procedures patients. Incidence of high- d High-volume risk procedures neurosurgeons Patient travel times reduced number of procedures for high-risk performed. procedures d Low-volume neurosurgeons increased rate of exit. d No effect on incidence of craniotomies. d Greater malpractice Probability of cesarean risk was associated with higher probability of cesarean section. Cesarean-section d No significant rates decrease in cesarean section rates among physicians who participated in the program versus those who didn’t (Continued) d
Impact of Medical Malpractice Environment
Dubay (1999)29
d
Resource use (medical management vs angioplasty vs bypass surgery)
Minami et al
Florida State Center Comparison of 2 Dranove (2005)45 Compare the effect of Neurosurgeons, different time for Health obstetricians, and change in periods to identify Statistics Hospital malpractice risk on patients the effect of inpatient data undergoing activity levels of changing levels of neurosurgeons and neurosurgery or malpractice risk. obstetric obstetricians, the 1997e2000 procedures incidence of high(lower malpractice in Florida, risk surgery, and premiums) was 1997e2000 and patient travel compared to 2000e2003 times. 2000e2003 (higher malpractice premiums)
Outcomes d
Vol. 218, No. 2, February 2014
Appendix.
278.e3
Continued
Consequences of malpractice liability
First author (y)
Study aim
Study population
Data sources
Florida births, Assess whether 1992e2000 malpractice claims are associated with cesarean section rates and delivery volume.
Glassman (1996)21 Assess how 1,540 cardiologists, surgeons, OB/ physicians’ malpractice history gyns, internists from a single affects clinical insurer in one state decision-making.
Florida Inpatient Hospital Discharge Data; Florida Medical Professional Liability Insurance Claims File Insurance claims; physician survey data
Key findings
Presence of caps
d
Physician supply rates
d
d
Physician claims history
d
Cesarean delivery Delivery volume
d
Annual number of physician claims Recent claim experience Multiple claim experience Serious claim experience
d
Physician responses to clinical scenarios
d
d
d
d
d
d
d
d
Kessler (1996)24
Health Care Analyze the effects of Medicare Financing liability reforms on beneficiaries Administration treated for serious Medicare patients enrollment files heart disease in with serious heart 1984, 1987, 1990 disease.
d
Malpractice reforms by state
d
Hospital expenditures
d
J Am Coll Surg
Counties in states with caps of $250,000 had an increased supply of surgical specialists but those with caps above $250,000 saw now difference in supply. Physician claims history was not associated with choice of cesarean delivery, but had a small effect on delivery volume. No association between malpractice exposure and resource use No association between malpractice exposure and influence by litigation threat or malpractice concern or risk perception Positive association between perceived clinical risk and resource use Direct reforms reduce hospital expenditures while indirect reforms are not associated with effects on expenditures. (Continued)
Impact of Medical Malpractice Environment
Gimm (2010)30
Outcomes
d
Minami et al
Encinosa (2005)23 Examine the effect of Physicians in practice Area Resource File nationwide from (ARF) damage caps on 1970e2000 geographic distribution of physicians.
Malpractice environment measures
278.e4
Appendix.
Continued
Consequences of malpractice liability
First author (y)
Kim (2007)38
Klick (2007)12
Study aim
Assess the impact of malpractice risk on cesarean section, VBAC, antenatal testing, and delivery techniques. Examine the causal effect of reforms on physician supply.
Study population
Data sources
Malpractice environment measures
Outcomes
Key findings
d
Claims paid per birth
d
Cesarean section, VBAC, antenatal testing, and delivery techniques
d
Doctors’ procedure choice is insensitive to malpractice risk.
Physicians in practice, nationwide, 1980e2001
American Medical Association Physician Masterfile
d
Reforms including damage caps, collateral source reform, joint and several liability, caps on attorney contingency fees, mandatory periodic payment of future damages States with caps on non-economic damages (compared to states with no caps)
d
Physician supply rates
d
Caps on noneconomic damages have a significant effect on high-risk specialties.
d
Whether a patient receives radical cystectomy
d
Premium levels Perceived risk by geographic area Claims against hospitals Claims against hospital obstetric staff as a group Claims against individual physicians
d
Cesarean section rates
d
Patients with stage III/IV bladder cancer were more likely to undergo radical cystectomy in states with caps than patients in states with no caps. Both premiums and perceived risk had a positive association with probability of cesarean delivery.
Konety (2005)44
62,464 patients with SEER Evaluate effect of bladder cancer caps on noneconomic damages on use of radical cystectomy and survival.
d
Localio (1993)42
All deliveries in 31 Medical Practice Assess association acute care hospitals Study between in New York in malpractice claims 1984 risk and probability of a cesarean delivery.
d d
d
d
d
(Continued)
Impact of Medical Malpractice Environment
NPDB, Natality Detail File
Minami et al
U.S births, 1992e1998
Vol. 218, No. 2, February 2014
Appendix.
278.e5
Continued
Consequences of malpractice liability
First author (y)
Study aim
Study population
Data sources
Matsa (2007)22
Physicians in Estimate effect of practice, damage caps on nationwide, physician supply. 1970e2000
Mello (2007)3
Assess the extent to Physicians in high- MCARE: Medical Care Availability and low-risk which liability and Reduction of specialties in costs cause Pennsylvania from Error physicians to 1993e2002 restrict their scope of practice or cease to practice. Illinois singleton NCHS; ISMIE Assess association births, between OB 1991e2003 malpractice premiums and late preterm induction (LPI).
Murthy (2009)37
Murthy (2007)36
Random sample of Survey cardiologists from AMA Masterfile
Examine whether OB Illinois singleton births, malpractice 1998e2003 premiums are associated with cesarean delivery rates.
ARF, Regional Economic Information System, NPDB
NCHS; ISMIE
Outcomes
d
Self-reported influence of malpractice on clinical practice/ decisions
d
d
Presence of caps
d
Key findings
Intensity of testing/ treatment as measured using clinical vignettes d Dartmouth Hospital Referral Region (HRR) Medicare End of Life Expenditure Index (EOL-EI)
d
d
Physician supply rates
d
“Before the malpractice crisis” vs “after the malpractice crisis”
d
Physician scope of practice and supply rates
d
d
OB malpractice premiums
d
LPI rate
d
d
OB malpractice premiums
d
Cesarean delivery rate
d
Self-reported influence of malpractice risk was significantly associated with individual physician treatment/testing intensity as well as Medicare EOL-EI at the HRR level. The presence of damage caps is correlated with a significant increase in surgical specialists in rural areas. Overall supply of high-risk specialists did not drop during the crisis.
J Am Coll Surg
Rising premiums were associated with increased frequency of LPI among women with singleton gestations. Higher rates of primary cesarean delivery are associated with increased medical professional liability premiums for obstetriciangynecologists in Illinois. (Continued)
Impact of Medical Malpractice Environment
Investigate factors associated with variation in cardiologists’ use of invasive tests/ treatment.
Minami et al
Lucas (2010)28
Malpractice environment measures
278.e6
Appendix.
Continued
Consequences of malpractice liability
First author (y)
Study aim
Study population
Data sources
Ng (2010)47
Investigate factors influencing geographic variation in endovascular aneurysm repair (EVAR).
US inpatient discharges for abdominal aortic aneurysm, 2001e2006
Rock (1988)35
Test whether OB malpractice premiums were associated with primary cesarean delivery rate.
Births in New York MLIC; ISMIE; IL & and Illinois, NY Department of 1982e1983 Public Health natality data
Sethi (2012)26
Evaluate self-reported 1,214 orthopaedic surgeons from defensive tactics AAOS among orthopaedic surgeons.
Sloan (1997)34
NIS
Physician survey
Evaluate effect of tort 963 women who gave Surveys, phone birth in 1987 in interviews, liability on Florida counties hospital records, method of Florida delivery, antenatal Department of testing, maternal Insurance satisfaction with care.
Outcomes
Key findings
d
Personal malpractice history and premium payments d Perceived level of malpractice risk
d
Self-reported influence of malpractice on practice decisions
d
State-level mean malpractice award d Paid claims rate per 1000 physicians d Average malpractice premiums for general surgery
d
EVAR rate
d
d
d
OB malpractice premiums
d
Cesarean delivery rate
d
d
Self-reported history of malpractice claims Risk of lawsuit Impact of liability premiums
d
Proportion of tests, consultation, hospital admissions for defensive reasons
d
Individual claims experience, county claims experience, total payments incurred by county
d
Frequency of antenatal testing Cesarean rate
d
d d
d
d
72%, 67%, and 66% respondents reported using imaging, laboratory tests, and referrals/ consults for defensive purposes, respectively. Paid claims rate and average general surgery malpractice premiums were significantly and positively correlated with EVAR rates. In both the New York and Illinois samples, a substantial impact was found on delivery decisions resulting from fear of malpractice suits. Orthopaedic surgeons’ defensive medicine is a significant factor in health care costs and is of marginal benefit to patients. High suit rates did not affect antenatal testing and cesarean section rates.
278.e7
(Continued)
Impact of Medical Malpractice Environment
Assess neurosurgeons’ US members of the Survey AANS perceived malpractice risk and defensive medicine practices.
Minami et al
Nahed (2012)27
Malpractice environment measures
Vol. 218, No. 2, February 2014
Appendix.
Continued
Consequences of malpractice liability
First author (y)
Study aim
Study population
Data sources
d
d
Personal history of malpractice claims Liability insurance
Outcomes d
d
d d
d
Tussing (1992)33
Yang (2009)32
Deliveries in NY Examine various excluding NYC factors influencing cesarean delivery, including malpractice fear. US births, Assess relationship 1991e2003. between malpractice pressure and cesarean and VBAC rates.
New York State Live Birth File, SPARCS
d
Malpractice claims per physician
d
Natality Detail File; MLM; NCSL; ATRA
d
Malpractice premiums State malpractice tort reforms
d
d
d
Defensive practice, including both avoidance and assurance behaviors, correlated strongly with respondents’ lack of confidence in their liability insurance and perceived burden of insurance premiums.
d
Cesarean rate VBAC rate
d
Malpractice “fear” was negatively associated with probability of cesarean delivery. Malpractice premiums were positively associated with rates of cesarean section and primary cesarean section, and negatively associated with VBAC rates. Two types of tort reform–caps on noneconomic damages and pretrial screening panels–were associated with lower rates of Csection and higher rates of VBAC. (Continued)
J Am Coll Surg
d
Impact of Medical Malpractice Environment
d
Key findings
Frequency of ordering more tests than medically indicated Frequency of prescribing more medications than warranted Unnecessary referrals Avoid procedures. Avoid high-risk patients. Reduce/eliminate high-risk elements of clinical practice due to liability. Probability of cesarean delivery
Minami et al
Physician survey Studdert (2005)20 Describe prevalence 924 physicians (emergency and correlates of defensive medicine medicine, general surgery, among high-risk orthopaedic specialists during surgery, period of neurosurgery, OB/ malpractice Gyn, radiology) in volatility. Pennsylvania, 2003
Malpractice environment measures
278.e8
Appendix.
Continued
Consequences of malpractice liability
First author (y)
Study aim
Zwecker (2011)31 Investigate effect of OB malpractice premiums on obstetric care.
Study population
Data sources
Women who HCUP NIS; MLM; delivered across 37 NAIC states in 2006
Malpractice environment measures d
Malpractice premiums
Outcomes
Key findings
Fear of litigation appears to have a marked effect on obstetric practice, particularly cesarean section, VBAC, and instrumental delivery, when malpractice premiums rise above $100,000 per annum.
d
Survival
d
Caps on damages were associated with improved survival.
d
Probability of 5-min Apgar
Impact of Medical Malpractice Environment on Surgical Quality and Outcomes Christina A Minami, MD, Jeanette W Chung, Karl Y Bilimoria, MD, MS, FACS
PhD,
Jane L Holl,
a systematic review of the literature to examine the association between malpractice environment and outcomes in surgical specialties.
The annual cost of the medical malpractice system has been estimated to be $55 billion.1 The medical malpractice crisis that began unfolding in the late part of the 20th century continues to be a major concern.2-5 The intent of the malpractice system, based on classic tort deterrence theory, assumes that the looming threat of a malpractice suit will deter poor care because providers will be more vigilant and responsible.6 Proponents of the liability system believe that the threat of malpractice suits will encourage providers to adhere to standards of care, which, in turn, should lead to better patient outcomes. However, critics argue that the system is ineffective in encouraging quality and better outcomes and that it generates only unintended consequences, such as defensive medicine, which may lead to worse patient outcomes because of overuse of services or avoidance tactics that may result in overtesting, overtreatment, iatrogenic injury, and soaring health care costs.7,8 In addition to the problems of malpractice insurance availability and affordability, personal strain from experiencing a malpractice claim can affect clinicians. A recent study found that onequarter of American surgeons had been the subject of a malpractice suit over the preceding 2 years, resulting in notable stress personally and professionally.9 Perceived malpractice risk, real or not,10 may force providers to leave high malpractice risk environments, particularly those in high risk fields of medicine, which, in turn, affects access to care and could potentially adversely affect patient outcomes.3,11,12 Surgical specialties are among the fields of medicine with a large proportion of malpractice claims13-18 and have been a primary focus of the malpractice crisis. The objective of this study was to perform
METHODS Search strategy The literature search included Medline, PubMed, and the Cochrane Database of Reviews of Effectiveness to capture articles published between 1980 and 2012. A comprehensive search using terms that included medical malpractice, tort reform, quality of care, outcomes, and litigation in relation to these surgical and procedural subspecialties was performed. Studies on surgical subspecialties and procedure-based fields, including interventional cardiology/cardiac surgery, urology, neurosurgery, obstetrics and gynecology (OB/GYN), general surgery, and orthopaedic surgery were included. A manual secondary search of reference papers cited in these initial studies was used to expand the database and capture studies published in journals outside the range of Medline and PubMed (eg, law and economic journals). Study selection (exclusion and inclusion criteria) The literature search included review articles, observational studies, case reports, survey studies, and retrospective data analyses. To be included, a study had to be an original research article, written in English, published between 1980 and 2012 in peer-reviewed journals, and performed in the United States. Only those that demonstrated some link or lack thereof between malpractice environment and physician practice or patient outcomes were included. Those that focused solely on claims prevalence, judgments, or amounts and their association with various adverse events were excluded. Studies captured in the initial search and reported on invasive procedures (eg, cardiac catheterization or endoscopy) in addition to surgical operations were retained.
Disclosure Information: Nothing to disclose. Received July 23, 2013; Revised August 30, 2013; Accepted September 16, 2013. From the Surgical Outcomes and Quality Improvement Center, Department of Surgery, Feinberg School of Medicine, Northwestern University and Northwestern Memorial Hospital (Minami, Chung, Holl, Bilimoria), and the Center for Healthcare Studies, Feinberg School of Medicine, Northwestern University, Chicago, IL. Correspondence address: Karl Y Bilimoria, MD, MS, FACS, Surgical Outcomes and Quality Improvement Center, Department of Surgery, Feinberg School of Medicine, Northwestern University, Northwestern Memorial Hospital, 676 St Clair St, Arkes Pavilion Suite 6-650, Chicago, IL 60611. email: [email protected]
ª 2014 by the American College of Surgeons Published by Elsevier Inc.
MD, MPH,
Data extraction Data from each study regarding data sources, study population, and endpoints were abstracted. Because of the heterogeneity in methods, patients, fields, interventions,
271
ISSN 1072-7515/13/$36.00 http://dx.doi.org/10.1016/j.jamcollsurg.2013.09.007
272
Minami et al
J Am Coll Surg
Impact of Medical Malpractice Environment
Table 1. Conceptual Framework for Examining the Effect of Malpractice Environment on Physician Practice and Patient Outcomes Intended consequences
Unintended consequences
Physician practice Hypothesis/argument: Hypothesis/argument: Liability threat creates incentives for physicians to adhere Liability threat creates incentives for providers to (a) to prevailing standards, resulting in high quality of care. substitute low-risk procedures for high-risk procedures; (b) avoid high-risk patients; and (c) provide clinically unnecessary tests and procedures due to fear of missed diagnoses. Hypothesis/argument: Patient outcomes Hypothesis/argument: Liability threat is associated with poorer patient outcomes Liability threat encourages physician adherence to because (a) patients who may benefit or need procedures standards of care and conscientiousness, resulting in perceived to be "high risk" won’t get the care that they better patient outcomes. need. They will either do without care or travel further for it. This may lead to poorer outcomes; (b) risky patients may experience worse outcomes if they can’t get the care they need because physicians fear poor outcomes and liability; and (c) overprovision of care (testing and treatment) may increase risk of iatrogenic injury, resulting in poorer outcomes.
and measured endpoints, data pooling was not possible and a narrative data summary was instead completed. Due to the nature of this research, no randomized controlled studies were found; data were from only retrospective cohort and survey studies. Analysis The PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) checklist was used to structure the review. Data pooling was not possible due to the heterogeneity of study populations and endpoints, and a formal meta-analysis was not performed. This article represents a systematic review. A 2 2 classification was used to categorize studies based on whether the study focused on the effects of physician practice or patient outcomes, and whether the effects were intended or unintended consequences of the malpractice liability system. This framework is built on the opposing arguments regarding the intent of the liability system and the resulting effects. The malpractice system is intended to deter negligence and improve patient outcomes. This approach resulted in 4 categories of studies (Table 1):
1. Studies of the intended consequences of malpractice liability on physician practice were those that identified an association between some measure of malpractice risk (eg, laws, claims frequency, claims severity) and an improvement in a measure of adherence to an explicit standard of care or a performance on a standard quality indicator. 2. Studies of the intended consequences of malpractice liability on patient outcomes were those that directly
tested the hypothesis that malpractice risk improves patient outcomes as measured by mortality, other outcomes, or patient satisfaction. 3. Studies of the unintended consequences of malpractice liability on physician practice were those that directly tested whether some measure(s) of malpractice risk were associated with greater use of defensive medical practices (eg, imaging, ordering of laboratory tests, substitution of less risky procedures for more risky procedures, patient selection [ie, avoidance of highrisk complicated patients], referrals). Studies using self-reported survey data on defensive practices were included in this group. 4. Studies of the unintended consequences of malpractice liability on patient outcomes were those that directly tested whether some measure of malpractice environment was associated with poorer patient outcomes (possibly as a result of defensive medicine).
RESULTS The initial literature search identified 349 studies, of which 94 were excluded immediately by their title, given their irrelevance to malpractice and/or surgical and nonsurgical procedures (Fig. 1). Further review excluded 92 studies, based on their international nature and primary focus on nonsurgical topics. A total of 163 studies were further scrutinized and, after final application of inclusion and exclusion criteria, 29 final studies were included in the systematic review. These 29 studies examined the intended and unintended effects of malpractice environs on changes in physician practice and patient outcomes.
Minami et al
Vol. 218, No. 2, February 2014
349 titles/abstracts identified by initial search (both literature search and reference lists)
94 excluded (topic not malpractice or surgery)
255 studies selected and ordered
92 excluded (international studies, nonsurgical topics)
163 reviewed
134 excluded -24 addressed tort reform in general -15 not related to malpractice - 47 not related to outcomes or changes in practice - 24 were review/background articles -24 addressed claims data only
29 included
Figure 1. Flowchart of eligible studies addressing the association between medical malpractice environment and surgical quality.
Physician practice measures included use of physician services (n ¼ 2), physician supply (n ¼ 5), procedure rates (n ¼ 18), and test ordering (n ¼ 7); patient outcomes examined included adverse events (n ¼ 1) and survival (n ¼ 4). Seven studies were analyses of both medical and surgical data (as opposed to solely surgical data), but included surgical/procedural subset analyses.3,12,19-23 Three cardiology studies were also included due to their interventional/procedural-based nature. Obstetrics (n ¼ 13), neurosurgery (n ¼ 2), urology (n ¼ 1), orthopaedic surgery (n ¼ 1), and vascular surgery (n ¼ 1) were the surgical subspecialties represented. There was considerable variation in the procedures and results of the included studies (Appendix, online only). Data sources included multistate (n ¼ 14), single-state, or single-institution databases (n ¼ 10), as well as survey studies (n ¼ 5). A distinction was made between studies that evaluated patient outcomes (n ¼ 4; eg, mortality, fetal well-being) and those that reflected changes in physician practice thought to be indicative of defensive
Impact of Medical Malpractice Environment
273
medicine (n ¼ 28; eg, use of physician services, test ordering, likelihood of proceeding with cesarean section). Liability premiums were the most commonly chosen indicator of malpractice environment though physicians’ past claims experience and malpractice payments were also used in some studies. Four of the 9 studies that used premiums as a proxy for malpractice risk, used Medical Liability Monitor (MLM) data, an annual nationwide survey of physician malpractice insurance premiums. Other sources included data from the Urban Institute Survey and state-level insurance underwriters. Data on malpractice claims payments came from the National Practitioner Data Bank (NPDB), a record of all payments made on behalf of physicians for medical liability settlements or judgments. Given the limited number of studies that addressed our specific aims, there was a wide range of study design quality and statistical rigor. In addition, some studies were more rigorous in controlling for the temporal relationship between exposure and outcomes than others, and the sophistication of economic modeling varied widely. Intended consequences e physician practice As the Appendix (online only) demonstrates, no studies directly tested whether greater levels of malpractice risk were associated with greater physician adherence to standards of care. There were no studies that showed a direct association between malpractice risk and subsequent changes in national measures of quality-of-care or patient safety. Intended consequences e patient outcomes Three studies attempted to directly test the association between the malpractice system and patient outcomes. Baicker and colleagues19 investigated the effect of malpractice liability on the use of physician services and mortality using data on Medicare fee-for-service beneficiaries from 1993 to 2001.19 Although state malpractice liability indicators appeared to be associated with increased Medicare spending and malpractice costs, the authors found no association between average malpractice payments per physician capita or average premiums and mortality in this patient population. Similarly, Kessler and McClellan24 looked at Medicare patients treated for acute myocardial infarction and found that there was no difference in mortality in this population by malpractice environment. Dhankhar and associates,25 however, found that increased malpractice risk led to lower mortality among acute myocardial infarction patients using the Nationwide Inpatient Sample (NIS).25
274
Minami et al
Impact of Medical Malpractice Environment
Unintended consequences e physician practice All 29 studies included in this review addressed the association between malpractice risk and defensive medicine practices. These studies explore the prevailing hypothesis that malpractice pressure leads to defensive medicine, which may manifest as avoidance tactics (ie, steering clear of risky patients and procedures) or assurance tactics (eg, ordering more diagnostic tests). Of the 29 studies, 5 were surveys of clinicians and 24 were retrospective studies. The 5 survey studies20,21,26-28 investigated the link between physician perceptions of risk and subsequent clinical actions and/or decision-making. Studdert20 and Glassman and coworkers21 examined the relationship between physician practice and malpractice history or perceived malpractice risk. Both found that physicians’ past liability exposure was not associated with defensive practices. Individual physicians’ perception of risk, however, did influence their treatment decisions. Glassman and colleagues21 found that although physicians’ past experiences with the legal system did not correlate with patterns of resource use (eg, testing, decision to admit or discharge, and decision to proceed with or delay surgery), their self-perceived claims risk did affect resource use. Four studies found that a significant proportion of physician respondents admitted to ordering unnecessary tests due to malpractice concerns.20,26-28 The large 2005 survey of defensive practices by Studdert20 included a sample of 824 emergency medicine physicians, general surgeons, neurosurgeons, obstetrician/gynecologists, orthopaedic surgeons, and radiologists from the American Medical Association Physician Masterfile and found that 93% of physicians reported practicing defensive medicine. Specialists who lacked confidence in their insurance coverage were more likely to order unnecessary diagnostic testing, refer to other physicians unnecessarily, and suggest invasive procedures. Subgroup analyses revealed field-specific trends, with high rates of possibly unnecessary referral to other physicians among OB/GYNs (59%) and high rates of unnecessary invasive procedures among general surgeons (44%). A survey of physicians in orthopaedic surgery by Sethi and coauthors26 showed that defensive imaging represented 19% to 24% of all imaging ordered. In a study of neurosurgeons, Nahed and colleagues27 reported that a majority of respondents engaged in defensive medicine practices, such as unnecessary testing and referrals. In addition, respondents reported that liability concerns limited their performance of high-risk procedures for traumatic head and spine injuries, intracranial hemorrhages, tumor resections, hydrocephalus (45%), as well as the duration of their clinical practice (71%). Lucas and colleagues28 focused
J Am Coll Surg
on cardiac catheterization and found that 24% of general practitioners reported ordering a cardiac catheterization due to malpractice fears.28 Although survey research has established a correlation between physician perception of malpractice risk and subsequent practice patterns and treatment decisions, 18 observational studies offered an alternate approach to examining this issue. Fourteen of these studies came from the obstetrics literature, examining cesarean section rates, vaginal birth after cesarean (VBAC) rates, and late pre-term induction.29-42 The majority of data on the effect of malpractice premiums on cesarean section rates found increased rates of cesarean section with increased malpractice premiums,29,31,32,35,36,39,42,43 while only 1 study found the opposite,33 and 2 found no significant effect of state malpractice premiums on cesarean section rates.29,41 Higher premiums were associated with higher rates of late pre-term induction37 and with lower vaginal birth after cesarean rates and instrumental deliveries.31,32 Five other studies examined procedures beyond obstetrics. Konety and associates44 found that greater proportions of patients with stage III and IV bladder cancer underwent radical cystectomy, a high-risk procedure, in regions with malpractice payment caps. A study by Dranove and Gron45 focused on craniotomy and cesarean section rates in Florida during the malpractice crisis period of the early 2000s, and found that after Florida premiums rose precipitously, many neurosurgeons decreased their volume of brain operations, and craniotomy patients had increased travel times to definitive care. Women undergoing highrisk deliveries, however, did not see increases in travel times. A 2004 study by Baicker and coworkers46 used Medicare data to examine the use of diagnostic testing as well as rates of 6 specific procedures: angiography, coronary artery bypass graft, percutaneous coronary intervention, radical prostatectomy, and transurethral prostatectomy. Little evidence of net increases in the procedure rates was found in response to state-level increases in malpractice costs. Dhankar and associates,25 when evaluating resource use in the treatment of acute myocardial infarction, found decreased resource use in their patient population (ie, in environments with greater malpractice risk, patients were more likely to undergo medical management as opposed to angioplasty). Similarly, using the Nationwide Inpatient Samples and State Inpatient Databases, Ng and colleagues47 found that increased use of endovascular aneurysm repair (EVAR), a less invasive procedure than open repair, increased with the number of paid claims and mean malpractice premiums. Surgeon supply in response to the presence of state malpractice crises or liability reforms has also been assessed.
Vol. 218, No. 2, February 2014
Minami et al
Unlike the study by Dranove and Gron,45 the analysis by Mello and coauthors3 of physician supply in Pennsylvania before and after the crisis of the early 2000s found little change. The presence of damage caps, however, was associated with an increase in local supply of surgical specialists in 2 studies.22,23 Klick and Stratmann12 came to a similar conclusion, although the “high-risk” specialties group in their analysis included nonsurgical fields, such as emergency medicine and general practice. Unintended consequences e patient outcomes Two studies investigated the association between malpractice risk and patient outcomes. Konety and colleagues44 found that the presence of caps on noneconomic damages resulted in an improvement in diseasespecific survival in advanced bladder cancer patients because surgeons were more willing to perform a risky procedure that resulted in better patient outcomes. Focusing on the effects of malpractice risk, Dubay and associates29 used data from the National Natality Files to test whether greater malpractice risk was associated with use of cesarean section delivery and birth outcomes and found that higher malpractice premiums were associated with lower 5-minute Apgar scores.
DISCUSSION Rooted in classic tort deterrence theory, the medical malpractice system is intended to discourage negligent care, encourage physicians to be conscientious and adhere to guidelines, and result in better health care outcomes. However, there may also be unintended consequences in which malpractice liability results in defensive medicine practices. We sought to synthesize the literature to assess what is currently known about the impact of malpractice environment on physician practice and patient outcomes. We found a limited body of literature based on a number of different data sources, surgical subspecialties, geographic regions, and eras. Information and selection bias were, to varying degrees, present in all of the studies. Although some studies consistently demonstrated an association between malpractice environment and defensive medicine, there was less consistency in the data regarding the ability of malpractice liability to affect physician practice and patient outcomes. Physician practice patterns The bulk of the literature on the effect of the malpractice system has focused on examining the impact on provider practice patterns and is based on surveys and empirical assessments. These studies show that physician perception of malpractice pressures can have a measurable effect on
Impact of Medical Malpractice Environment
275
treatment decisions. It may be that the fear of missing an occult pathology or being accused of failing to intervene in areas of greater liability risk are manifested by a propensity to obtain an unnecessary diagnostic workup, even though such interventions carry their own inherent risk. However, survey studies have considerable limitations. First, physicians with strong opinions, likely negative, may be more likely to respond to the survey. Second, survey respondents may exaggerate the effect when they have an opportunity to voice their opinion on a controversial topic. Third, surveys reflect only perceptions, but it is unknown whether this actually reflects the empirical relationship between malpractice environment and physician practice. A large number of empirical assessments have focused on defensive medicine, reflecting the medical community’s clear interest in this aspect of the malpractice system. All 29 studies commented on the consequences of malpractice liability on physician practice, including whether the legal environment can sway a practitioner’s decision to perform a procedure, order diagnostic testing, or limit his or her practice. In aggregate, the data on malpractice influence on procedure use are mixed. In obstetrics, higher malpractice risk may result in obstetricians intervening sooner and more often than otherwise (ie, have a lower threshold for taking a patient for cesarean section) and in fewer women with a previous cesarean section subsequently trying to deliver vaginally. The obstetrics literature showed trends toward increased cesarean delivery rates and decreased rates of vaginal birth after cesarean in higher liability environments, though there was a lack of complete agreement among these 13 studies. Studies of other operations also showed that malpractice environment can affect whether surgeons perform riskier procedures (eg, radical cystectomy). Overall, practice patterns do appear to be affected by malpractice environment, although these empirical studies had several limitations regarding the population examined, time frames, geographic areas, and indicator of malpractice risk (eg, premiums vs caps). Malpractice and outcomes Although many studies sought to profile the types of adverse events that lead to claims, especially in highliability fields like surgery, we failed to find any existing studies that examined the intended consequences of the malpractice system on greater provider adherence to standards of care. The 5 studies 19,24,25,29,44 that examined the consequences of malpractice environment on patient outcomes offered differing conclusions; 2 found that patient outcomes were improved in higher liability environments,25,44 but 3 found that outcomes were worse or no better.19,24,29 These studies focused on individual
276
Minami et al
Impact of Medical Malpractice Environment
procedures, fields of medicine, or diseases using different measures of malpractice environment, making direct comparison difficult. Some of these studies had very limited representation of different geographic areas and those that broadened the geographic breadth of the study population by using Medicare data, are, by default, limited to an aging population. So it is unclear whether malpractice environment has a beneficial effect on patient outcomes. There is undoubtedly a need for a comprehensive assessment of the association between malpractice environment and patient outcomes across multiple surgical fields using various indictors of malpractice environment. There are several issues that may affect the association between malpractice environment and quality. Defense tactics may be classified as “assurance” vs “avoidant” behaviors, the former alluding to the practice of ordering medical services with little clinical value, and the latter referring to practitioner avoidance of certain high-risk patients and procedures.20 Perhaps an extrapolation of these avoidant behaviors is the issue of relative reticence throughout surgical fields regarding adverse events and errors. Although morbidity and mortality conferences are an embedded part of surgical culture, they are, by definition, protected discussions, open only to an institution’s surgical faculty and housestaff. There are disclosure-and-offer programs being piloted in certain states, such as those in Illinois, New York, Texas, and Washington State, which are modeled on a program out of the University of Michigan Health System, but this approach is by no means ubiquitous.48 Another adverse consequence is that specialty, in-demand, or high quality physicians may be driven out of states due to the malpractice environment. These problems may all feed into methodologic difficulties in examining the association between malpractice environment and outcomes, and may make high malpractice risk environments appear to have poorer quality. There have also been longstanding theoretical criticisms of the deterrence theory, the most relevant to medical malpractice being the idea that insurance insulates individual physicians from the economic repercussions of a lawsuit, therefore blunting the deterrent effect and failing to act as a motivator for improvements in patient safety.7 Moreover, only about 2% of negligent care results in a lawsuit.13 These issues may partly explain why no deterrence effect is observed. Methodologic limitations In assessing the effect of local malpractice environment on medical care, it is important to consider data sources and their inherent strengths and weaknesses. Observational studies generally have the advantage of covering
J Am Coll Surg
large populations and time frames. On the other hand, a drawback of observational studies is that they rely on proxy measures for key variables of interest. For example, claims rates, average award size, and premium rates were frequently used as indicators of providers’ perceived liability threat. Yet it is the strength of the correlation between these indicators of perceived liability threat that is empirically unknown, and it is unclear whether a single indicator or a composite measure of malpractice environment would be more meaningful. As a methodologic approach in the malpractice literature, survey studies offer the advantage of direct measurement of exposure to liability risk and levels of perceived malpractice threat. However, there have been few large nationally representative studies of physicians and malpractice, and many of these surveys have been limited with respect to geographic coverage, provider types or specialties, and/or time periods. Additional drawbacks of survey research are the potential for selection bias in response rates (if physician responses are correlated with underlying malpractice risk) and the potential for respondents to provide socially desirable answers that may not accurately reflect their actual behavior or intentions to act. These inherent strengths and weaknesses shed some light on why stronger associations between malpractice environment and physician behavior that have been typically uncovered by survey studies.
CONCLUSIONS The existing body of data on the effect of malpractice environs on surgical outcomes and quality of care is small, equivocal, and has notable temporal and geographic limitations. Though there is evidence that medical malpractice liability influences physicians’ clinical choices, there is little to support the theory that the threat of medical litigation improves physician adherence to quality care indicators or improves patient outcomes. To facilitate a productive discussion on the malpractice system and tort reform, we need to better understand whether the liability system in the US actually serves its intended purpose of being a deterrent to negligent care and improving safety and quality. A comprehensive evaluation of the effect of malpractice environment on quality of care is needed. Author Contributions Study conception and design: Minami, Bilimoria Acquisition of data: Minami, Bilimoria Analysis and interpretation of data: Minami, Chung, Holl, Bilimoria Drafting of manuscript: Minami, Chung, Bilimoria Critical revision: Holl, Bilimoria
Vol. 218, No. 2, February 2014
Minami et al
REFERENCES 1. Mello MM, Gawande A, Studdert D. National costs of the medical liability system. Health Affairs 2010;29:1569e1577. 2. Paik M, Black B, Hyman D. The receding tide of medical malpractive litigation. 7th Annual Conference on Empirical Legal Studies Paper 2013. 3. Mello M, Studdert D, Brennan T. Changes in physician supply and scope of practice during a malpractice crisis: evidence from Pennsylvania. Health Affairs 2007;26: 425e435. 4. Mello MM. Understanding Medical Malpractice Insurance: A Primer. Available at: http://www.rwjf.org/content/dam/farm/ reports/reports/2006/rwjf17974. Accessed October 8, 2013. 5. Association AM. America’s Medical Liability Crisis: a National View. 2005. http://www.ama-assn.org/ama1/pub/upload/mm/ 450/med_liab_jan07.pdf. Accessed October 8, 2013. 6. Bell P. Legislative intrusions into the common law of medical malpractice: thoughts about the deterrent effect of tort liability. Syracuse Law Review 1984;XXXV:939e993. 7. Mello M, Brennan T. Deterrence of medical errors: theory and evidence for malpractice reform. Tex Law Rev 2002;80: 1595e1637. 8. Studdert D, Mello M, Brennan T. Defensive medicine and tort reform: a wide view. J Gen Intern Med 2010;25: 380e381. 9. Balch CM, Oreskovich MR, Dyrbye LN, et al. Personal consequences of malpractice lawsuits on American surgeons. J Am Coll Surg 2011;213:657e667. 10. Carrier E, Reschovsky J, Mello M, et al. Physicians’ fears of malpractice lawsuits are not assuaged by tort reforms. Health Affairs 2010;29:1585e1592. 11. Mello MM, Studdert DM, DesRoches CM, et al. Caring for patients in a malpractice crisis: physician satisfaction and quality of care. Health Aff (Millwood) 2004;23:42e53. 12. Klick J, Stratmann T. Medical malpractice reform and physicians in high-risk specialties. J Legal Studies 2007;36: S121eS142. 13. Brennan T, Leape L, Laird N. Incidence of adverse events and negligence in hospitalized patients: results of the Harvard Medical Practice Study I. N Engl J Med 1991;324: 370e376. 14. Kane C. Policy research perspectivesemedical liability claim frequency: a 2007-2008 snapshot of physicians. American Medical Association 2010:1e7. http://www.ama-assn.org/resources/doc/ health-policy/prp-201001-claim-freq.pdf. Accessed October 8, 2013. 15. Studdert D, Mello M, Gawande A, et al. Claims, errors, and compensation payments in medical malpractice litigation. N Engl J Med 2006;354:2024e2033. 16. Jena AB, Seabury S, Lakdawalla D, Chandra A. Malpractice risk according to physician specialty. N Engl J Med 2011; 365:629e636. 17. Sloan F, Mergenhagen P, Burfield W, et al. Medical malpractice experience of physicians: predictable or haphazard? JAMA 1989;26:3291e3297. 18. Orosco R, Talamini J, Chang D, Talamini M. Surgical malpractice in the United States, 1990-2006. J Am Coll Surg 2012;215:480e488. 19. Baicker K, Fisher ES, Amitabh C. Malpractice liability costs and the practice of medicine in the Medicare program. Health Affairs 2007;26:841e852.
Impact of Medical Malpractice Environment
277
20. Studdert D. Defensive medicine among high-risk specialist physicians in a volatile malpractice environment. JAMA 2005;293:2609e2617. 21. Glassman PA, Rolph JE, Petersen LP, et al. Physicians’ personal malpractice experiences are not related to defensive clinical practices. J Health Polit Policy Law 1996;21:219e241. 22. Matsa D. Does malpractice liabiity keep doctors away? Evidence from tort reform damage caps. J Legal Studies 2007;36:S143eS182. 23. Encinosa W, Hellinger F. Have state caps on malpractice awards increased the supply of physicians? J Health Affairs 2005. Suppl Web Exclusives: W5-250-W5-258. 24. Kessler D, McClellan M. Do doctors practice defensive medicine? Q J Econ 1996;111:353e390. 25. Dhankar P, Khan M, Bagga S. Effect of medical malpractice on resource use and mortality of AMI patients. J Empirical Legal Studies 2007;4:163e183. 26. Sethi MK, Obremskey WT, Natividad H, Mir H, Janhgir A. Incidence and costs of defensive medicine among orthopedic surgeons in the United States: a national survey study. Am J Ortho 2012;41:69e73. 27. Nahed BV, Babu MA, Smith TR, Heary RF. Malpractice liability and defensive medicine: a national survey of neurosurgeons. PloS one 2012;7:e39237. 28. Lucas FL, Sirovich BE, Gallagher PM, et al. Variation in cardiologists’ propensity to test and treat: is it associated with regional variation in utilization? Circulation Cardiovascular quality and outcomes 2010;3:253e260. 29. Dubay L, Kaestner R, Waidmann T. The impact of malpractice fears on cesarean section rates. J Health Econ 1999;18: 491e522. 30. Gimm GW. The impact of malpractice liability claims on obstetrical practice patterns. Health services research 2010; 45:195e211. 31. Zwecker P. Effect of fear of litigation on obstetric care: a nationwide analysis on obstetric practice. Am J Perinatol 2011;28:277e284. 32. Yang T, Mello M, Subramanian S, Studdert D. Relationship between malpractice litigation pressure and rates of cesarean section and vaginal birth after cesarian section. Medical Care 2009;47:234e242. 33. Tussing A, Wojtowycz M. The cesarean decision in New York State, 1986. Medical Care 1992;30:529e540. 34. Sloan F, Entman S, Reilly B, et al. Tort liability and obstetricians’ care levels. Int Review Law Economics 1997;17:245e260. 35. Rock S. Malpractice premiums and primary cesarean section rates in New York and Illinois. Public Health Rep 1988; 103:459e463. 36. Murthy K. Association between rising professional liability insurance premiums and primary cesarean delivery rates. Obstet Gynecol 2007;110:1264e1269. 37. Murthy K. Obstetricians’ rising liability insurance premiums and inductions at late preterm gestations. Med Care 2009; 47:425e430. 38. Kim B. The impact of malpractice risk on the use of obstetrics procedures. J Legal Studies 2007;36:S79eS120. 39. Baicker K, Buckles K, Chandra A. Geographic variation in the appropriate use of cesarean delivery. Health Affairs 2006; 2006. 40. Baldwin L, Hart L, Lloyd M, et al. Defensive medicine and obstetrics. JAMA 1995;274:1606e1610.
278
Minami et al
Impact of Medical Malpractice Environment
41. Edwards C. The impact of a no-fault tort reform on physician decision-making: a look at Virginia’s Birth Injury Program. Revista Juridica UPR 2010;80:285e310. 42. Localio A, Lawthers A, Bengtson J, et al. Relationship between malpractice claims and cesarean delivery. JAMA 1993;269:366e373. 43. Ryan K, Schnatz P, Greene J. Change in cesarean section rate as a reflection of the present malpractice crisis. Conn Med 2005;69:139e141. 44. Konety B, Dhawan V, Allareddy V, Joslyn S. Impact of malpractice caps on use and outcomes of radical cystectomy for bladder cancer: data from the surveillance, epidemiology, and end results program. J Urol 2005;173:2085e2089.
J Am Coll Surg
45. Dranove D, Gron A. Effect of the malpractice crisis on access to and incidence of high-risk procedures: evidence from Florida. Health Affairs 2005;24:802e810. 46. Baicker K, Chandra A. The Effect of Malpractice Liability on the Delivery of Health Care. Cambridge, MA: National Bureau of Economic Research; 2004. 47. Ng TT, Mirocha J, Magner D, Gewertz BL. Variations in the utilization of endovascular aneurysm repair reflect population risk factors and disease prevalence. J Vasc Surg 2010;51: 801e809. 48. Kachalia A, Mello M. New directions in medical liability reform. N Engl J Med 2011;364:1564e1572.
Summary of Reviewed Studies Examining the Effect of Malpractice Environment on Physician Practice and Patient Outcomes
Consequences of malpractice liability
First author (y)
None Intended consequences of malpractice liability on physician practice Intended consequences of malpractice liability on patient outcomes Baicker (2007)19
Data sources
NPDB; MLM; Quantify the effect of Medicare Fee-forMedPAR; Service malpractice Medicare Part B liability on the use beneficiaries, Claims 1993-2001 of physician services, Medicare payments, and mortality. Quantify the impact Patients in the NIS NPDB; NIS dataset in 2002 of medical who suffered acute malpractice on MI mortality probability from AMI. Health Care Analyze the effects of Medicare Financing liability reforms on beneficiaries Administration treated for serious Medicare patients enrollment files heart disease in with serious heart 1984, 1987, 1990 disease.
Malpractice environment measures
Outcomes
Key findings
Average dollar value of malpractice payments per physician capita Average premiums
d
Mortality
d
No association between mortality from various causes and measures of malpractice costs.
d
Claims frequency and claims severity in 2000e2002
d
d
d
Malpractice reforms by state
Resource use (medical management vs angioplasty vs bypass surgery) Mortality Mortality
Increased medical malpractice risk leads to improvement in mortality among AMI patients. Malpractice reforms do not lead to consequential differences in mortality.
d
d
d d
d
Unintended consequences of malpractice liability on physician practice (defensive medicine) (Continued)
Impact of Medical Malpractice Environment
Kessler (1996)24
Study population
Minami et al
Dhankhar (2007)25
Study aim
Vol. 218, No. 2, February 2014
Appendix.
278.e1
Continued
Consequences of malpractice liability
First author (y)
Study aim
Study population
Data sources
Malpractice environment measures
NCHS linked birth and infant mortality data; NPDB; MLM
d
Baicker (2007)19
Quantify the effect Medicare Fee-forService of malpractice liability on the use beneficiaries, 1993e2001 of physician services and costs.
NPDB; MLM; MedPAR; Medicare Part B Claims
d
Number and size of malpractice payments per physician d Average malpractice premiums
d
Baldwin (1995)40
Content of Low-risk obstetric Examine effect of Obstetrical Care patients of malpractice Study (COOC), Washington State experience on Washington State providers who prenatal resources Physicians initiated care and cesarean Insurance between 9/1/1988 section rates. Exchange and and 8/31/1989 Association Database (WSPIEA)
d
d
Outcomes
Key findings
d
Cesarean rates
d
Average dollar value of malpractice payments per physician capita Average premiums
d
d
Personal malpractice experience County-based malpractice defendant rate
d
Total Medicare spending/ beneficiary Spending/beneficiary by service type Utilization rates of physician services (screening, diagnostic and imaging procedures) and major elective surgical procedures reported in the Dartmouth Atlas Use of obstetric ultrasound, referral and consultation Delivery method
d
d
d
d
Malpractice environment explained an estimated 14.8% and 13.9% of variation in county-level cesarean rates for normal and low/ very low birth weight deliveries, respectively. A 10% increase in malpractice payments increases the use of carotid duplex, ECHO, EKG, CT and MRI by 1.5%e1.8%.
Impact of Medical Malpractice Environment
US births, Examine whether 1995e1998 geographic variation in appropriate use of cesarean section is correlated to geographic variation in malpractice liability.
Minami et al
Baicker (2006)39
278.e2
Appendix.
No association between malpractice claims experience and increased prenatal resources or cesarean sections in low-risk obstetric patients. (Continued) J Am Coll Surg
Continued
Consequences of malpractice liability
First author (y)
Study aim
Study population
Data sources
Dhankhar (2007)25 Quantify the impact Patients in the NIS NPDB; NIS dataset in 2002 of medical who suffered acute malpractice on MI mortality probability from AMI.
Malpractice environment measures d
Claims frequency and claims severity in 2000-2002
Edwards (2010)41
U-S births, Examine effect of 1990e1992 malpractice claims risk on cesarean section, birth outcomes. Assess the effect of All deliveries in Virginia in 2006 the Birth Injury Program (BIP) on cesarean section rates in Virginia.
d
d
National Natality Files
d
OB/GYN malpractice premiums
d
Virginia Health Information
d
Physicians participating in the BIP vs those who didn’t
d
Key findings
Increase in malpractice risk leads to reduced resource use in patients with less severe medical conditions. d Increase in travel Number of physicians times for performing highcraniotomy risk procedures patients. Incidence of high- d High-volume risk procedures neurosurgeons Patient travel times reduced number of procedures for high-risk performed. procedures d Low-volume neurosurgeons increased rate of exit. d No effect on incidence of craniotomies. d Greater malpractice Probability of cesarean risk was associated with higher probability of cesarean section. Cesarean-section d No significant rates decrease in cesarean section rates among physicians who participated in the program versus those who didn’t (Continued) d
Impact of Medical Malpractice Environment
Dubay (1999)29
d
Resource use (medical management vs angioplasty vs bypass surgery)
Minami et al
Florida State Center Comparison of 2 Dranove (2005)45 Compare the effect of Neurosurgeons, different time for Health obstetricians, and change in periods to identify Statistics Hospital malpractice risk on patients the effect of inpatient data undergoing activity levels of changing levels of neurosurgeons and neurosurgery or malpractice risk. obstetric obstetricians, the 1997e2000 procedures incidence of high(lower malpractice in Florida, risk surgery, and premiums) was 1997e2000 and patient travel compared to 2000e2003 times. 2000e2003 (higher malpractice premiums)
Outcomes d
Vol. 218, No. 2, February 2014
Appendix.
278.e3
Continued
Consequences of malpractice liability
First author (y)
Study aim
Study population
Data sources
Florida births, Assess whether 1992e2000 malpractice claims are associated with cesarean section rates and delivery volume.
Glassman (1996)21 Assess how 1,540 cardiologists, surgeons, OB/ physicians’ malpractice history gyns, internists from a single affects clinical insurer in one state decision-making.
Florida Inpatient Hospital Discharge Data; Florida Medical Professional Liability Insurance Claims File Insurance claims; physician survey data
Key findings
Presence of caps
d
Physician supply rates
d
d
Physician claims history
d
Cesarean delivery Delivery volume
d
Annual number of physician claims Recent claim experience Multiple claim experience Serious claim experience
d
Physician responses to clinical scenarios
d
d
d
d
d
d
d
d
Kessler (1996)24
Health Care Analyze the effects of Medicare Financing liability reforms on beneficiaries Administration treated for serious Medicare patients enrollment files heart disease in with serious heart 1984, 1987, 1990 disease.
d
Malpractice reforms by state
d
Hospital expenditures
d
J Am Coll Surg
Counties in states with caps of $250,000 had an increased supply of surgical specialists but those with caps above $250,000 saw now difference in supply. Physician claims history was not associated with choice of cesarean delivery, but had a small effect on delivery volume. No association between malpractice exposure and resource use No association between malpractice exposure and influence by litigation threat or malpractice concern or risk perception Positive association between perceived clinical risk and resource use Direct reforms reduce hospital expenditures while indirect reforms are not associated with effects on expenditures. (Continued)
Impact of Medical Malpractice Environment
Gimm (2010)30
Outcomes
d
Minami et al
Encinosa (2005)23 Examine the effect of Physicians in practice Area Resource File nationwide from (ARF) damage caps on 1970e2000 geographic distribution of physicians.
Malpractice environment measures
278.e4
Appendix.
Continued
Consequences of malpractice liability
First author (y)
Kim (2007)38
Klick (2007)12
Study aim
Assess the impact of malpractice risk on cesarean section, VBAC, antenatal testing, and delivery techniques. Examine the causal effect of reforms on physician supply.
Study population
Data sources
Malpractice environment measures
Outcomes
Key findings
d
Claims paid per birth
d
Cesarean section, VBAC, antenatal testing, and delivery techniques
d
Doctors’ procedure choice is insensitive to malpractice risk.
Physicians in practice, nationwide, 1980e2001
American Medical Association Physician Masterfile
d
Reforms including damage caps, collateral source reform, joint and several liability, caps on attorney contingency fees, mandatory periodic payment of future damages States with caps on non-economic damages (compared to states with no caps)
d
Physician supply rates
d
Caps on noneconomic damages have a significant effect on high-risk specialties.
d
Whether a patient receives radical cystectomy
d
Premium levels Perceived risk by geographic area Claims against hospitals Claims against hospital obstetric staff as a group Claims against individual physicians
d
Cesarean section rates
d
Patients with stage III/IV bladder cancer were more likely to undergo radical cystectomy in states with caps than patients in states with no caps. Both premiums and perceived risk had a positive association with probability of cesarean delivery.
Konety (2005)44
62,464 patients with SEER Evaluate effect of bladder cancer caps on noneconomic damages on use of radical cystectomy and survival.
d
Localio (1993)42
All deliveries in 31 Medical Practice Assess association acute care hospitals Study between in New York in malpractice claims 1984 risk and probability of a cesarean delivery.
d d
d
d
d
(Continued)
Impact of Medical Malpractice Environment
NPDB, Natality Detail File
Minami et al
U.S births, 1992e1998
Vol. 218, No. 2, February 2014
Appendix.
278.e5
Continued
Consequences of malpractice liability
First author (y)
Study aim
Study population
Data sources
Matsa (2007)22
Physicians in Estimate effect of practice, damage caps on nationwide, physician supply. 1970e2000
Mello (2007)3
Assess the extent to Physicians in high- MCARE: Medical Care Availability and low-risk which liability and Reduction of specialties in costs cause Pennsylvania from Error physicians to 1993e2002 restrict their scope of practice or cease to practice. Illinois singleton NCHS; ISMIE Assess association births, between OB 1991e2003 malpractice premiums and late preterm induction (LPI).
Murthy (2009)37
Murthy (2007)36
Random sample of Survey cardiologists from AMA Masterfile
Examine whether OB Illinois singleton births, malpractice 1998e2003 premiums are associated with cesarean delivery rates.
ARF, Regional Economic Information System, NPDB
NCHS; ISMIE
Outcomes
d
Self-reported influence of malpractice on clinical practice/ decisions
d
d
Presence of caps
d
Key findings
Intensity of testing/ treatment as measured using clinical vignettes d Dartmouth Hospital Referral Region (HRR) Medicare End of Life Expenditure Index (EOL-EI)
d
d
Physician supply rates
d
“Before the malpractice crisis” vs “after the malpractice crisis”
d
Physician scope of practice and supply rates
d
d
OB malpractice premiums
d
LPI rate
d
d
OB malpractice premiums
d
Cesarean delivery rate
d
Self-reported influence of malpractice risk was significantly associated with individual physician treatment/testing intensity as well as Medicare EOL-EI at the HRR level. The presence of damage caps is correlated with a significant increase in surgical specialists in rural areas. Overall supply of high-risk specialists did not drop during the crisis.
J Am Coll Surg
Rising premiums were associated with increased frequency of LPI among women with singleton gestations. Higher rates of primary cesarean delivery are associated with increased medical professional liability premiums for obstetriciangynecologists in Illinois. (Continued)
Impact of Medical Malpractice Environment
Investigate factors associated with variation in cardiologists’ use of invasive tests/ treatment.
Minami et al
Lucas (2010)28
Malpractice environment measures
278.e6
Appendix.
Continued
Consequences of malpractice liability
First author (y)
Study aim
Study population
Data sources
Ng (2010)47
Investigate factors influencing geographic variation in endovascular aneurysm repair (EVAR).
US inpatient discharges for abdominal aortic aneurysm, 2001e2006
Rock (1988)35
Test whether OB malpractice premiums were associated with primary cesarean delivery rate.
Births in New York MLIC; ISMIE; IL & and Illinois, NY Department of 1982e1983 Public Health natality data
Sethi (2012)26
Evaluate self-reported 1,214 orthopaedic surgeons from defensive tactics AAOS among orthopaedic surgeons.
Sloan (1997)34
NIS
Physician survey
Evaluate effect of tort 963 women who gave Surveys, phone birth in 1987 in interviews, liability on Florida counties hospital records, method of Florida delivery, antenatal Department of testing, maternal Insurance satisfaction with care.
Outcomes
Key findings
d
Personal malpractice history and premium payments d Perceived level of malpractice risk
d
Self-reported influence of malpractice on practice decisions
d
State-level mean malpractice award d Paid claims rate per 1000 physicians d Average malpractice premiums for general surgery
d
EVAR rate
d
d
d
OB malpractice premiums
d
Cesarean delivery rate
d
d
Self-reported history of malpractice claims Risk of lawsuit Impact of liability premiums
d
Proportion of tests, consultation, hospital admissions for defensive reasons
d
Individual claims experience, county claims experience, total payments incurred by county
d
Frequency of antenatal testing Cesarean rate
d
d d
d
d
72%, 67%, and 66% respondents reported using imaging, laboratory tests, and referrals/ consults for defensive purposes, respectively. Paid claims rate and average general surgery malpractice premiums were significantly and positively correlated with EVAR rates. In both the New York and Illinois samples, a substantial impact was found on delivery decisions resulting from fear of malpractice suits. Orthopaedic surgeons’ defensive medicine is a significant factor in health care costs and is of marginal benefit to patients. High suit rates did not affect antenatal testing and cesarean section rates.
278.e7
(Continued)
Impact of Medical Malpractice Environment
Assess neurosurgeons’ US members of the Survey AANS perceived malpractice risk and defensive medicine practices.
Minami et al
Nahed (2012)27
Malpractice environment measures
Vol. 218, No. 2, February 2014
Appendix.
Continued
Consequences of malpractice liability
First author (y)
Study aim
Study population
Data sources
d
d
Personal history of malpractice claims Liability insurance
Outcomes d
d
d d
d
Tussing (1992)33
Yang (2009)32
Deliveries in NY Examine various excluding NYC factors influencing cesarean delivery, including malpractice fear. US births, Assess relationship 1991e2003. between malpractice pressure and cesarean and VBAC rates.
New York State Live Birth File, SPARCS
d
Malpractice claims per physician
d
Natality Detail File; MLM; NCSL; ATRA
d
Malpractice premiums State malpractice tort reforms
d
d
d
Defensive practice, including both avoidance and assurance behaviors, correlated strongly with respondents’ lack of confidence in their liability insurance and perceived burden of insurance premiums.
d
Cesarean rate VBAC rate
d
Malpractice “fear” was negatively associated with probability of cesarean delivery. Malpractice premiums were positively associated with rates of cesarean section and primary cesarean section, and negatively associated with VBAC rates. Two types of tort reform–caps on noneconomic damages and pretrial screening panels–were associated with lower rates of Csection and higher rates of VBAC. (Continued)
J Am Coll Surg
d
Impact of Medical Malpractice Environment
d
Key findings
Frequency of ordering more tests than medically indicated Frequency of prescribing more medications than warranted Unnecessary referrals Avoid procedures. Avoid high-risk patients. Reduce/eliminate high-risk elements of clinical practice due to liability. Probability of cesarean delivery
Minami et al
Physician survey Studdert (2005)20 Describe prevalence 924 physicians (emergency and correlates of defensive medicine medicine, general surgery, among high-risk orthopaedic specialists during surgery, period of neurosurgery, OB/ malpractice Gyn, radiology) in volatility. Pennsylvania, 2003
Malpractice environment measures
278.e8
Appendix.
Continued
Consequences of malpractice liability
First author (y)
Study aim
Zwecker (2011)31 Investigate effect of OB malpractice premiums on obstetric care.
Study population
Data sources
Women who HCUP NIS; MLM; delivered across 37 NAIC states in 2006
Malpractice environment measures d
Malpractice premiums
Outcomes
Key findings
Fear of litigation appears to have a marked effect on obstetric practice, particularly cesarean section, VBAC, and instrumental delivery, when malpractice premiums rise above $100,000 per annum.
d
Survival
d
Caps on damages were associated with improved survival.
d
Probability of 5-min Apgar
