Vol. IX, No. 8
Medical Malpractice: Strengthening
the Evidence Base
For decades, the myriad issues associated with not investigated by the researchers. These
medical malpractice have been hotly debated disagreements are enhanced because there
by advocates, researchers, and policymak- is limited consensus in the research on the
ers alike. Historically, legislative efforts have underlying factors driving the various mal-
focused primarily on tort reform and insur- practice problems. The contentious nature of
ance regulation. More recent proposals have the malpractice debate has resulted in vague
begun to explore alternative policy approaches policy goals, misinterpretations of findings,
such as apology-compensation programs, and the selective use of research with ques-
health courts, and patient safety initiatives. tionable conclusions.
Despite continued attention, most policy pro- In order to provide policymakers with credible
posals are based on research findings facing evidence required to develop policy reforms that
the same drawbacks—data limitations and can address clearly identified malpractice prob-
disparate methodologies. Missing and inac- lems, researchers need better data and more rig-
curate data and poorly designed studies lead orous methodologies. The Robert Wood Johnson
to inconsistent findings, which have made the Foundation, under its Changes in Health Care
malpractice debate vulnerable to exaggerated Financing and Organization (HCFO) initiative,
and invalid claims and ideological rhetoric. conducted a small invitational meeting to provide
an opportunity for researchers and other stake-
The other significant barrier to successful mal- holders to take a closer look at malpractice stud-
practice reform is the failure of researchers and ies and to discuss why findings diverge, where
policymakers to be clear about which malprac- there may be common ground, and how to over-
tice problem they are trying to address, since come current research limitations.
different problems can require very different
solutions. Malpractice problems most often iden- In an off-the-record, facilitated discussion, par-
tified include high malpractice insurance rates, ticipants reviewed the state of the evidence and
reduced physician supply, defensive medicine, explored whether and how research results sup-
increased overall health care costs, invalid law- port or call into question a variety of proposed
suits, lack of compensation for injured patients, policy solutions. The session was a positive step
and patient safety. toward understanding the strengths and weak-
nesses of the underlying evidence in order to
Confusion about malpractice problems creates help policymakers as they develop workable mal-
AcademyHealth is the national program an environment where research findings can practice reforms.
office for HCFO, an initiative of
the Robert Wood Johnson Foundation.
be misinterpreted to support policy solutions
issue brief — C h a n g e s i n H e a lth Financing & Organization page 2
Background on the Medical specific side of the debate. Experts agree that Information on jury awards is also problematic
Malpractice Process conducting malpractice research is exceed- and limited. Jury verdict reporters, a primary
The vast majority of patients injured by medi- ingly difficult due to limitations in the avail- source for verdict information, can contain sig-
cal care do not seek compensation. Research able data and an existing library of studies nificant inaccuracies, and may be more likely to
estimates that approximately 2 percent of that lack methodological rigor. include large verdict awards than small verdict
patients injured by medical malpractice pursue awards.7 Other sources such as the RAND
legal recourse.1, 2 Medical malpractice litigation One major problem with many studies is Institute for Civil Justice Jury Verdict Database
is primarily governed by state law. While legal that they tend to compare states with differ- and the Civil Justice Survey of State Courts
procedures vary from state to state, the overall ent legal environments without controlling collect information on court verdicts, but they
process for adjudicating a malpractice claim is for other ways in which states may differ. do not track cases in every county, making it
consistent across the country. When an injury Additionally, missing information from spe- difficult to develop generalizable conclusions.
occurs and the patient or family decides to pur- cific groups of physicians, settings, and/or Some researchers question the usefulness of
sue legal action, their lawyer files a claim with regions can lead to invalid or over generalized jury verdict data because very few cases make it
the court against the provider who in turn noti- findings. Recognizing and addressing the to verdict and often the initial amount awarded
fies the malpractice insurance carrier. weaknesses in data and methodologies most is not final.
commonly used in malpractice research will
The discovery phase, which can last several improve efforts to generate new and more The richest data on malpractice claims and
years, follows with an exchange of informa- sophisticated studies and provide more reli- the injuries that led to them are collected by
tion between the plaintiff and defendant, able support for policy reforms. insurance companies for business purposes.
including expert opinions. There are several Unfortunately, they are rarely accessible to
opportunities for a claim to be resolved prior Limitations in Data Accuracy and researchers. States require insurers to publicly
to trial. Resolution can occur through media- Availability report a limited amount of claims information,
tion or arbitration, dismissal or summary In general, much of the most pertinent but data from a growing market of self-insured
judgment by the court, or early settlement information needed to examine the sever- and physician owned-and-operated mutual
between the plaintiff and defendant. If the ity, frequency, and causes of malpractice companies may not be captured because these
claim moves to trial it will be decided by losses at state-specific levels is not collected. organizations are not held to the same report-
either a judge or, more commonly, a jury. Examples of helpful data not currently avail- ing requirements as commercial malpractice
able include the insured amounts covered insurers. For example, a data source like the
A trial can motivate the parties to reach a by premiums and breakdowns of paid losses NAIC databank, which primarily collects infor-
settlement, or may result in a verdict for the due to settlement and trail verdicts and eco- mation for financial purposes, excludes infor-
defendant or for the plaintiff, with the latter nomic and noneconomic damages. Available mation from self-insured groups.
accompanied by an award of economic and/or data include closed-claims, insurance com-
non-economic damages. A jury award may be pany information, jury awards, physician Limitations in Methodologies
amended by the court in some cases, especially surveys, and patient medical records. and Standardizations
if it is very large. It may also be overturned if Further complicating sound research is the lack
the defendant wins an appeal in a higher court. Existing databases often suffer from short- of standardization of databases and variations
comings in completeness and accuracy. in study methodologies making it difficult to
A malpractice claim takes an average of For example, researchers have reported compare one study to another. Comparing data
four to five years to resolve, with the most that most closed-claims databases have across states and between data sources is dif-
difficult cases taking significantly longer.3 significant limitations. The Government ficult since there is no standardized definition
Approximately 50 percent of all claims are Accountability Office found that the National for events like injuries, claims, and settlements.
paid.4 In addition to any potential award, a Practitioner Data Bank (NPDB), purportedly Researchers also face difficulties analyzing
significant amount of additional costs are one of the better sources for malpractice the impact of changing laws and regulations.
incurred in the form of litigation expenses. data due to federal mandates that require For example, establishing the effect of a tort
High administrative costs for carriers general- reporting of adverse actions and malpractice reform is complicated since the date a law is
ly mean that only 40 cents of every dollar paid payments, suffers from underreporting of put into place often does not create an immedi-
in malpractice insurance premiums goes to claims, exclusion of institutional providers, ate reaction. Rather, insurance adjustors and
the patient.5 Long resolution time, variation in and omission of legal and administrative physicians may respond to a new reform only
awards, unpredictable returns on investment, costs associated with claims.6 Some states after a test case is resolved by the highest court.
and the insurance cycle are some of the fac- also collect closed-claim data, but variations Changes in common law, which is not written
tors that make it extremely difficult to predict in reporting requirements, definitions, into statute, can make significant differences
and stabilize premium rates. and quality standards frustrate aggrega- on malpractice cases and are seldom assessed
tion of data across states and complicate since they are difficult to track.
Research on Medical Malpractice interstate comparisons. Other closed-claim
is Challenging data sources that are not publicly available Different methodological approaches also can
The challenges associated with medical include the Physician Insurers Association result in disparate findings. For example, dif-
malpractice research make it vulnerable of America (PIAA) Data Sharing Project ferent approaches are often taken to measure
to unreliable, exaggerated, or misleading and the National Association of Insurance the effect of malpractice on defensive medicine.
conclusions that are often used to support a Commissioners (NAIC) Special Survey. One study that surveys doctors to determine
HCFO Issue Brief Malpractice fin2 2 1/30/07 3:45:14 PM
issue brief — C h a n g e s i n H e a lth Financing & Organization page 3
their motivation for performing certain proce- Premiums are sensitive to returns on finan- Strategies researchers can take to improve the
dures will likely differ from another study that cial investments AND reward size. quality of information in the public domain
reviews patient medical records to determine if include promoting generally accepted research
services rendered were medically necessary. Due Mello’s review found that caps on non-eco- principles and holding others accountable for
to the limitations of any single study, it is essen- nomic damages reduce award size but do not following these principles. One way policymak-
tial to consider multiple studies with different affect claims frequency. Caps induce a modest ers who use this information can be better
approaches to ensure a comprehensive body of decrease in premium growth and slight increase informed about the quality of the research is if
reliable evidence. in physician supply but disproportionately bur- a score for quantitative policy research were for-
den the most severely injured patients. Though mulated based on research principles and other
Consensus on Findings caps on awards can take some pressure off key attributes.
One strategy to differentiate reliable studies has physicians by reducing premium growth, it is
been to critically evaluate and summarize the uncertain whether savings generated from tort Some are worried that no matter how good the
current body of research on medical malprac- reforms have any effect on total health care data or how much consensus is formed, there
tice.8, 9 The Synthesis Project at the Robert Wood costs. Other state tort reforms, including attor- will never be a compromise between special
Johnson Foundation (RWJF) recently focused on ney contingency-fee reforms, collateral-source interests on each side of the debate. While there
medical malpractice. Researcher, Michelle Mello, rule reforms, and pre-trail screening panels, had are very different opinions on medical mal-
J.D., Ph.D. of Harvard University assessed an no significant impact. practice, a concerted effort to improve the data,
array of peer-reviewed studies as part of RWJF’s ensure rigorous methodologies, and clarify the
synthesis. She examined research on the causes Alternative Approaches to Reform problem specific reforms are able to address
of the malpractice crises, its effect on health care A host of alternative reform approaches are can make a significant contribution to inform-
delivery and costs, and the impact of state tort being explored by researchers who are rede- ing the debate and moving closer to a workable,
reforms. The synthesis clarifies which topics lack fining the malpractice problem. Several of effective solution. Despite the hesitancy of many
enough support to draw valid conclusions and these reforms include administrative courts researchers to get involved in a contentious
corrects some widely held assumptions. Mello or health courts, full disclosure programs, political debate, many policymakers and scholars
found that: and patient safety initiatives. Health courts recognize the value of additional independent
would use judges with health care expertise voices. Researchers are in an optimal position to
u There are no reliable estimates of the and expert witnesses to adjudicate malpractice distinguish between reliable and unreliable stud-
national costs of defensive medicine. claims.10 A schedule of damages would be ies, educate policymakers on where there is con-
established to guide reward allocation. Full sensus in the evidence, and clarify how different
u Rising claims costs are driven by an disclosure and “early offer” programs sup- policy options can fulfill a policy goal.
increase in average payouts. The number port doctors who apologize for medical errors
of filed claims has stayed stable. and offer compensation to families.11,12 Full Recommendations
disclosure programs have been implemented Researchers need high quality malpractice
u The strongest studies suggest that the in several hospital systems including those data that are complete, accurate, standard-
malpractice crisis has little or no effect in Illinois, Michigan, and Minnesota. Many ized, and comprehensive in order to produce
on physician supply. No solid evidence patient safety initiatives have begun to focus superior research that accurately represents
has found that access to high-risk ser- on fixing organizational systems and facilitat- the causes and effects of medical malpractice.
vices has declined. ing physician action.13,14 One example of a suc- This requires an investment in establishing
cessful patient safety initiative is the collective guidelines and practices that either create or
u There is very limited evidence that the physician efforts of anesthesiologists in the augment data collection systems that can be
medical liability system deters negligent mid-1980s to successfully reduce medical used for cross regional and interstate analysis.
care. Instead, the current system has per- injuries and deaths.15 Researchers should also meticulously con-
verse effects on patient safety initiatives. sider and acknowledge their assumptions and
Role of Researchers the strengths and weaknesses of their data
Mello examined studies that analyzed a host How researchers can most effectively address and methodological approaches. In the face of
of tort reforms and insurance regulations the “noise” that surrounds the medical mal- such varying quality in the data and research,
introduced to address less affordable and practice debate remains an important ques- researchers can facilitate the responsible
less available malpractice insurance. Many of tion. A critical starting point is better access use of high quality, peer-reviewed findings
these reforms are intended to address very to pertinent, accurate, and complete data. through increased participation in the educa-
different underlying causes for the increased While there will never be a perfect dataset, tion of policymakers.
costs associated with malpractice. Parties in improvements are possible. Researchers can
the debate argue that premium growth is highlight limitations of existing data and, There exists a great deal of confusion about
due to either an increased number of claims when possible, advocate that more states the goals of tort, insurance, and patient safety
and higher awards OR insurance cycles that collect essential claims level data and make reform. It is important for researchers and poli-
fluctuate based on investment returns and it publicly available. It is also necessary to cymakers to clarify what goals different reforms
pricing decisions. Evidence gathered by Mello ensure accurate data through audits, a very are meant to accomplish. Many policy goals
suggests that both arguments have merit. labor intensive activity. focus on the overall cost of the malpractice sys-
HCFO Issue Brief Malpractice fin3 3 1/30/07 3:45:16 PM
issue brief — C h a n g e s i n H e a l t h Financing & Organization page 4
tem and how to lower these costs. Some goals 3 Studdert, D.M., et al. “Claims, Errors, and 10 Mello, M.M., et al. “’Health courts’ and accountability
Compensation Payments in Medical Malpractice for patient safety,” Milbank Quarterly, Vol. 84, No. 3,
focus on identifying a better system to com-
Litigation,” New England Journal of Medicine, Vol. 354, 2006, pp. 459-92.
pensate individuals injured by adverse medi- No. 19, 2006, pp. 2024-33.
cal events, while others try to improve patient 11 O’Connell, J. “Offers That Can’t be Refused:
safety and reduce medical errors. Whatever the 4 ibid. Foreclosure of Personal Injury Claims by Defendants’
Prompt Tender of Claimants’ Net Economic Losses,”
objective, researchers and policymakers should Northwest University Law Review, Vol. 77, 1982, pp.
5 Danzon, P.M. “Liability for Medical Malpractice,” In:
be clear and specific about the problems they Culyer AJ, Newhouse JP, eds. Handbook of Health 589-632.
are addressing and the goals of the interven- Economics. New York: Elsevier, 2000.
12 “COPIC’s 3Rs Program: A Success Story,” COPIC
tions they are discussing. Finally, the national
6 ibid. Insurance 3Rs Program Newsletter, Vol. 3, No. 1, June
research agenda needs to include targeted dem- 2006. Also see: http://www.callcopic.com/resources/
onstrations and evaluations to test the effective- 7 Medical Malpractice: Implications of Rising
ness of different approaches. Premiums on Access to Health Care, Report GAO-
03-836. Washington, DC: Government Accountability
Office, August 2003. (no author given) Also see: 13 Kohn, L.T., J.M. Corrigan and M.S. Donaldson, eds.
About the Author http://www.gao.gov/new.items/d03836.pdf. To err is human: building a safer health system.
Christal Stone is an associate with the HCFO Washington, DC: National Academy Press, 2000.
initiative. She can be reached at 202.292.6700 8 Mello, M.M. “Medical Malpractice: Impact of the
Crisis and Effect of State Tort Reforms,” Robert 14 Schoenbaum, S.C. and R.R. Bovbjerg. “Malpractice
or email@example.com. Reform Must Include Steps to Prevent Medical
Wood Johnson Foundation, Research Synthesis
Report No. 10, May 2006. Also see: http://www.rwjf. Injury,” Annals of Internal Medicine, Vol. 140, No. 1,
Endnotes org/publications/synthesis/reports_and_briefs/pdf/ January 6, 2004, pp. 51-53.
1 Localio A.R., et al. ”Relation between Malpractice no10_researchreport.pdf.
Claims and Adverse Events Due to Negligence. 15 Cheney F.W. “The American Society of
Results of the Harvard Medical Practice Study III,” 9 Bovbjerg, R.R. and R.A. Berenson. “Surmounting Anesthesiologists Closed Claims Project: What Have
New England Journal of Medicine, Vol. 325, No. 4, 1991, Myths and Mindsets in Medical Malpractice,” The We Learned, How Has It Affected Practice, and How
pp. 245-51. Urban Institute, October 2005. Also see: http://www. Will It Affect Practice in the Future?” Anesthesiology,
urban.org/UploadedPDF/411227_medical_malprac- Vol. 91, No. 2, August 1999, pp. 552-556.
2 Studdert, D.M., et al. “Negligent Care and Malpractice tice.pdf.
Claiming Behavior in Utah and Colorado,” Medical
Care, Vol. 38, No. 3, 2000, pp. 250-60.
1801 K Street, NW, Suite 701-L Nonprofit
Washington, DC 20006 Organization
tel: 202.292.6700 PAID
fax: 202.292.6800 Permit No. 3999
HCFO Issue Brief Malpractice fin4 4 1/30/07 3:45:16 PM