Printer Friendly

Analysis of diagnostic error in paid malpractice claims with substandard care in a large healthcare system.

Objective: Although claims databases are not representative of all care delivery, their predisposition toward serious unintended injury can complement resource-intensive chart reviews and guide patient safety initiatives.

Materials and Methods: Non-Veterans Health Administration (VA) practitioners reviewed 1,949 VA malpractice claims paid during fiscal years 1998 through 2003. The portion associated with substandard care, the severity of harm, and types of negligence were identified.

Results: Negligent adverse events occurred in 37% (n = 723) of paid VA malpractice claims. These had high proportions of serious injury (55%) and morbidity (37%). Diagnostic negligent adverse events were most frequent (45%) and with 41% associated morbidity. The annual incidence of diagnosis-related paid VA malpractice claims was 1.95 per 100,000 patients and predicts that 122 of every 100,000 patients may have diagnostic negligent adverse events. Comparisons against non-VA data suggest this to be a healthcare industry problem.

Conclusions: Diagnosis-related negligent adverse events are a serious problem in the healthcare industry.

Key Words: Adverse event, diagnostic error, legal, liability, malpractice, medical error, National Practitioner Data Bank


Adverse events and medical errors have become an increasing concern. The Institute of Medicine (IOM) has concluded that between 44,000 and 98,000 Americans may die from medical errors annually and thus cause more deaths than breast cancer or motor vehicle accidents. (1) However, a private organization recently published findings that rates of death from medical error in Medicare patients between 2000 and 2002 were notably higher than rates estimated by the IOM report. (2) Major sources of the data cited in the IOM report were studies based on reviews of hospital admission records. (3-5) Although such comprehensive reviews provide uniquely useful information, they require significant investments of time and resources; thus, few such studies have been published.

In the Department of Veterans Affairs (VA), there is no repository of information on adverse events and negligent medical care directly comparable to published medical record reviews. (3-6) However, there does exist a centralized database of paid malpractice claims that may be reviewed to search for substandard care.

The VA processes approximately 1,200 claims and suits annually; about 38% of those receive some payment. We presumed that in most cases, nonpayment of claims or suits were due to the plaintiffs' failure to prove elements of malpractice, as required by law. However, payment may be made in the absence of these elements if it is concluded that the government's case is sufficiently weak, making adjudication inadvisable, for example, if the medical record is deficient.

Malpractice claims data have generally not been emphasized in evaluations of quality or safety published in the medical literature, in part because claims represent only a subset of all adverse events that may be discovered in record reviews or because of a belief that many claims bear little relation to demonstrated negligence. However, if analyses of claims were restricted to those actually paid, wherein a medical review panel determined that the care was negligent or substandard, this could provide useful information, with the additional benefit of identifying medical errors most likely to be associated with significant patient harm. This benefit comes at a cost well below traditional inpatient reviews, where less than 1% of reviews will reveal a negligent adverse event and less than half of these will result in a serious patient injury. (3-5)

In 1986, Congress passed the Health Care Quality Improvement Act (Public Law 99-660), which authorized the creation of the National Practitioner Data Bank (NPDB) to collect and store mandatory information from malpractice claims paid in the private sector. Data from VA malpractice claims that were paid and reviewed during the 6 VA fiscal years from 1998 through 2003 are the basis for this report.

Materials and Methods

Per VA regulations, (7,8) paid VA malpractice claims were reviewed by panels composed of a minimum of three individuals, including at least one member of the profession/occupation of the practitioner whose actions were under review. The panelists were non-VA practitioners. The conclusions of the panels, at a minimum, were based on a review of documents pertinent to the care that led to the claim. These documents included the patient's medical records, any available administrative investigation board report, and the opinion of any consultant requested by the panel. The documents reviewed did not include those generated primarily for consideration or litigation of the claim of malpractice. In addition, the documents included written statements of the individual involved in the care that led to the claim. The practitioner whose actions were under review received a written request for pertinent information from the VA facility director at the time that a VA malpractice claim had been paid.

The VA Director of Medical-Legal Affairs oversaw the panels for compliance with regulations and requirements and summarized the panels' conclusions. These summaries were sent forward for appropriate NPDB-related submission of data. In addition, the Office of Medical-Legal Affairs created a database of paid and reviewed VA malpractice claims, along with a means of identifying those associated with a negligent adverse event. The date of the negligent incident, the type of negligent adverse event, and associated severity of injury were coded and entered into the VA database.

Severity of injury was coded as none, minimal, severe, and death. Injuries were minor in nature if they did not require any medical intervention and they did not extend the hospital length of stay (except for observation or to obtain laboratory or radiology results). Injuries were major in nature if they required medical or surgical intervention, increased hospital stay, or were disabling or disfiguring to a degree that the patient would have permanently lessened function or require surgical repair. Injuries were classified as resulting in death if the claimant alleged the patient died as a consequence of the incident for which payment was sought.

The types of negligence were coded by using the established NPDB negligence categories, which were also used in the private sector for NPDB malpractice payment reports. The 10 NPDB negligence categories are as follows: anesthesia, diagnosis, equipment/product, intravenous/blood, medication, obstetrics, patient monitoring, surgery, treatment, and miscellaneous. The NPDB allowed their data to be coded into only one mutually exclusive negligence category. Although we used the same NPDB negligence categories to code the VA negligent adverse events, we allowed for more than one negligence category, as applicable.


During the 6 VA fiscal years of 1998 though 2003, the VA Medical Review Panels evaluated 1,949 malpractice claims. This represents about 75% of all claims paid during that time. The remainder have not yet come to the review process for a number of reasons, including recent payment, delay on the part of medical centers in forwarding reports of payment, incomplete or missing medical records, and so forth. Although the payment dates ranged from VA fiscal years 1998 through 2003, the incident dates for the negligent adverse events were widely distributed from VA fiscal years 1989 through 2002, with a median and modal VA fiscal year of 1997.

The review panels determined substandard care to have been present in 723 (37%) of the cases reviewed, or 120.5 paid VA malpractice claims associated with substandard medical care annually. These were associated with minor injury in 8% (n = 58), serious injury in 55% (n = 268), and patient death in 37% (n = 397).

Sixty-six percent of the paid VA malpractice claims with substandard care had one NPDB negligence category assigned, 29% had two, 4% had three, and less than 1% had four. The four most common preventable adverse events involved errors in diagnosis (45%, n = 324), treatment (28%, n = 205), surgery (26%, n = 187), and medication (16%, n = 119). Combined, these accounted for 91% of the reviewed VA paid malpractice claims and were associated with 88% of the associated patient deaths. Table 1 illustrates the distribution of substandard care across the single and combined negligence categories and the proportion of patient death within each category.

As the diagnosis-related subtotal shows (Table 1), diagnosis-related negligent adverse events, alone and in combination with other negligent errors, were associated with 45% of all paid claims, with an associated mortality rate of 41%. Diagnostic negligent adverse events were a factor in 49% of all deaths in paid VA malpractice claims associated with substandard care. By comparison, surgical negligent adverse events were a factor in 16% of all deaths in paid VA malpractice claims associated with substandard care.


The VA paid malpractice claims review process determined substandard care to have been present in 723 (37%) of the 1,949 malpractice claims paid and reviewed during VA fiscal years 1998 through 2003. The average annual number of paid malpractice claims associated with substandard care was 120.5. The median incident year was VA fiscal year 1997, during which the VA provided inpatient, outpatient, and emergent care to 2,762,549 patients. The number of patients served was used to calculate an incidence rate. The annual incidence rate of paid malpractice claims associated with substandard care was 4.36 per 100,000 veterans served.

Large inpatient, non-VA record reviews found that only 3% of adverse events caused by medical negligence resulted in a malpractice claim, and only 1.6% of negligent adverse events will result in a paid malpractice claim. (9-11) If this holds true for paid VA malpractice claims, extrapolation from our data predicts that the annual incidence of VA adverse events caused by medical errors was 272 per 100,000 veterans served.

In paid VA malpractice claims that were determined to be associated with substandard care, the adverse effects were significant: 55% were associated with serious injury or disability and 37% with death. Our findings of significant harm in cases of paid claims are consistent with other investigators, who have shown positive relations among the resultant degree of disability while receiving medical care, the likelihood of a malpractice claim, and the probability of payment of the claim. (9-11)

Diagnostic errors were determined to be present in nearly half (45%) of all VA substandard-care paid malpractice claims, with a 41% associated mortality rate. Annually, 54 paid VA malpractice claims were associated with diagnosis-related substandard care, alone or in combination with other types of error. The annual incidence rate of paid malpractice claims associated with diagnosis-related substandard care was 1.95 per 100,000 veterans served. Extrapolation from our data predicts that the annual incidence of VA adverse events caused by diagnosis-related errors was 122 per 100,000 veterans served.

From 1998 through 2003, the NPDB collected and recorded data on more than 91,000 paid malpractice claims from across the nation and from all healthcare venues. Thirty-eight percent of the malpractice payment reports were associated with alleged diagnostic negligent adverse events. (12) Because the NPDB only allows for one malpractice allegation category code and the VA data set allows for multiple error categories, it is not surprising that the VA proportion, although comparable, is slightly higher. In both data sets, diagnosis-related incidents were the most frequent.

Diagnostic adverse events are reported to be negligence related at a rate greater than other adverse event types. (3-6) Leape et al (4) found 74.7% of all diagnostic adverse events were negligent, Thomas et al (5) reported that incorrect or delayed diagnoses were negligent in 93.8% of cases, and a large Australian medical record study (6) classified 81% of diagnostic adverse events as being "highly preventable." Although diagnostic error was the most common category in both VA and NPDB data, the latter data base included all paid malpractice claims, whereas the VA data for this report were restricted to paid VA malpractice claims in which a review panel determined that the care rendered was substandard. Thus, because of the restriction of the VA data in regard to paid substandard care malpractice claims, the proportion of diagnostic error in the VA data was predicted to be somewhat higher than the unrestricted NPDB data. Removing this VA data restriction, 39% of paid VA malpractice claims included allegations of diagnostic related adverse events.

Diagnostic negligent adverse events were the most frequent events found among paid VA malpractice claims and, per NPDB data, the most frequent type of adverse event present in paid malpractice claims across the general healthcare industry. In addition, the NPDB proportional occurrence of diagnostic adverse events allegations were nearly equal in paid VA malpractice claims and were just slightly lower than the proportion of negligent adverse events found in substandard care paid VA malpractice claims. This suggests that diagnosis-related adverse events, and specifically diagnostic negligent adverse events, are a general healthcare industry problem.

Reportedly, the specialties with the largest proportion of highly preventable adverse events are general medicine (ie, family practice and internal medicine) and emergency medicine. (3-6) Negligent errors in emergency and urgent care are usually due to delayed or incorrect diagnoses, and negligent adverse events for general internists and family practitioners are most frequently related to failure to diagnose. (3) In the VA data, about 80% of negligent diagnostic errors occurred in outpatient settings, with 49% in clinics and 31% in the emergency area. Not surprisingly, inpatient chart reviews have reported that only 3% of their detected adverse events were associated with care delivered in the emergency area and another 7 to 9% in the doctor's office. (3-6) This resulted in a previously underestimated proportion of negligent adverse events that are diagnosis related.

Because of its selection bias toward severe outcomes, the review of paid malpractice claims does not necessarily provide a precise representation of all adverse events. However, malpractice claims data better represents a broader range of healthcare delivery modalities than traditional inpatient chart reviews. Some investigators have noted advantages to malpractice claims analyses in evaluating the quality of medical care. (13-16) The usefulness of malpractice claim analyses is based on events that are uncommon but associated with significant injury or death, the provision of information on out patient care, and a higher correlation with substandard care and preventable error than is the case with hospital chart review. Our evaluation of paid claims in the VA suggested that these observations are generally accurate. Analyzing paid malpractice claims associated with negligent adverse events not only allows for the evaluation of error and processes of care, it provides a unique opportunity to focus on preventable adverse events associated with relatively serious consequences.


Substandard care was present in a significant proportion of paid VA malpractice claims, with more than half associated with serious injury or disability and more than one third with death. Although substandard-care paid malpractice claims do not necessarily provide a precise representation of all negligent adverse events, the information obtained from their analyses can focus patient safety initiatives on preventable adverse events with serious consequences.

We found that diagnostic negligent adverse events were the most frequent type of event, present in nearly half of all paid VA malpractice claims associated with substandard care. As well as being the most frequent, diagnostic negligent adverse events were also associated with a high rate of patient death. More than three quarters of the diagnostic errors occurred in the emergency care and outpatient clinic settings; healthcare venues underrepresented by traditional inpatient chart reviews.

NPDB data represent paid malpractice claims for the general healthcare industry. Both the NPDB and VA datasets had similar proportion of paid malpractice claims associated with diagnostic error. This suggests that diagnostic errors, although previously not reported as such, are a widespread problem in the healthcare industry. By analyzing the traits of the diagnostic negligent adverse events identified through the review of paid VA malpractice claims, information can be obtained to inform patient safety initiatives and reduce diagnostic errors.

We have initiated a detailed analysis of diagnostic negligent adverse events to identify associated negligent patient care patterns and interventions most appropriate to improve quality of care and reduce patient injury. Specific analyses have included the relative impact of not ordering appropriate diagnostic tests, inappropriate follow-up of test results, failure to order appropriate follow-up testing, and/or failure to order appropriate specialty consults.


1. Institute of Medicine (US). To err is human: building a safer health system. Washington, DC: National Academies Press; 1999.

2. Patient Safety in American Hospitals. Lakewood, CO: Health Grades, Inc; 2004.

3. Brennan TA, Leape LL, Laird NM, Hebert L, Localio AR, Lawthers AG, et al. Incidence of adverse events and negligence in hospitalized patients: Results of the Harvard Medical Practice Study I. N Engl J Med 1991;324:370-376.

4. Leape LL, Brennan TA, Laird N, Lawthers AG, Localio AR, Barnes BA, et al. The nature of adverse events in hospitalized patients: Results of the Harvard Medical Practice Study II. N Engl J Med 1991;324:377-384.

5. Thomas EJ, Studdert DM, Burstin HR, Orav EJ, Zeena T, Williams ET, et al. Incidence and types of adverse events and neglect in Utah and Colorado. Med Care 2000;38:261-271.

6. Wilson RM, Runciman WB, Gibberd RW, Harrison BT, Hamilton JD. The quality in Australian healthcare study. Med J Aust 1995;163:458-471.

7. Department of Veterans Affair. 38 CFR Part 46. Washington, D.C.: Office of the Federal Register, Vol. 67, No. 78, April 23, 2002, pp 19678-19679.

8. Veterans Health Administration. National Practitioner Data Bank Reports. VHA Handbook 1100.17. Washington, D.C.: Veterans Health Administration, November 13, 2002.

9. Brennan TA, Sox CM, Burstin, HR. Relationship between negligent adverse events and the outcome of medical malpractice litigation. N Engl J Med 1996;335:1963-1967.

10. Localio AR, Lawthers AG, Brennan TA, Laird NM, Hebert LE, Peterson LM, et al. Relationships between malpractice claims and adverse events due to negligence: Results of the Harvard Medical Practice Study III. N Engl J Med 1991;325:245-251.

11. Studdert DM, Thomas EJ, Burstin HR, Zbar BI, Orav EJ, Brennan TA. Negligent care and malpractice claiming behavior in Utah and Colorado. Med Care 2000;38:250-260.

12. National Practitioner Data Bank Public Use File [June 2004], US Department of Health and Human Services, Health Resources and Services Administration, Bureau of Health Professions, Division of Practitioner Data Banks.

13. Morris JA, Carrillo Y, Jenkins JM, Smith PW, Bledsoe S, Pichert J, et al. Surgical adverse events, risk management, and malpractice outcome: morbidity and mortality review is not enough. Ann Surg 2003;237:844-851.

14. Kravitz RL, Rolph JE, McGuigan K. Malpractice claims data as a quality improvement tool. I. Epidemiology of error in four specialties. JAMA 1991;286:2087-20892.

15. Gawande AA, Studdert DM, Orav EJ, Brennan TA, Zinner MJ. Risk factors for retained instruments and sponges after surgery. N Engl J Med 2003;348:229-235.

16. Rothschild JM, Federico FA, Gandhi TK, Kaushal R, Williams DH, Bates DW. Analysis of medication-related malpractice claims. Arch Intern Med 2002;162:2414-2420.
The wise man doesn't give the right answers, he poses the right
--Claude Levi-Strauss

Thomas V. Holohan, MD, Janice Colestro, DNS, RN, John Grippi, MD, Jane Converse, JD, and Michael Hughes, BMET

From the National Institutes of Health, Bethesda, MD; Veterans Health Administration, Buffalo NY VAMC; and Veterans Health Administration Central Office, Washington, DC.

Reprint requests to John B. Grippi, MD, Office of Medical-Legal Affairs (IIML), Buffalo VA Medical Center, 3495 Bailey Avenue, Buffalo, NY 14215, Email:

Accepted January 31, 2005.

T.V.H., formerly the VA Director of Patient Care Services, initiated and oversaw the study of VA paid malpractice claims as a means to identify preventable adverse events for the purpose of improving patient care.

The opinions expressed in this document are those of the authors and do not represent the official position of the Department of Veterans Affairs or the United States Government.

The authors are employees of the Federal Government. No funds were received from nongovernmental sources.


* Medical records of Veterans Health Administration paid claims are reviewed by non-Veterans Health Administration practitioners to determine the presence of substandard care for purposes of reporting to the National Practitioner Data Bank.

* Negligent or substandard care was found in 37% of all paid claims.

* When substandard care occurred, adverse events suffered by patients were severe.

* Categorization into negligence classifications established by the National Practitioner Data Bank revealed that diagnostic error was the most frequent error type and was also associated with a high proportion of fatal outcomes.

* Review of malpractice cases complements medical record reviews of adverse events and provides an opportunity to focus on serious errors and to institute corrective actions to improve quality of care.
Table. Substandard-care paid VA malpractice claims by types of
negligence and associated deaths

 Paid claims Deaths
 (n = 723) (n = 268)
Negligence categories n (%) (a) n (%) (b)

Diagnosis 177 (24.5) 72 (40.7)
Diagnosis and treatment 110 (15.2) 50 (45.5)
Diagnosis and surgery 26 (3.6) 7 (26.9)
Diagnosis, treatment and surgery 9 (1.2) 2 (22.2)
Diagnosis and medication 2 (0.3) 1 (50.0)
Subtotal: Diagnosis related 324 (44.8) 132 (40.7)
Surgery 134 (18.5) 27 (20.1)
Medication 114 (15.8) 27 (23.7)
Treatment 67 (9.3) 42 (62.7)
Treatment and surgery 16 (2.2) 8 (50.0)
Medication and treatment and/or surgery 3 (0.4) 0 (0)
Not diagnosis, treatment, surgery, or 65 (9.0) 32 (49.2)
Total 723 (100) 268 (37.1)

(a) Percent of 723; (b) death within negligence category.
COPYRIGHT 2005 Southern Medical Association
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2005, Gale Group. All rights reserved. Gale Group is a Thomson Corporation Company.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:Original Article
Author:Hughes, Michael
Publication:Southern Medical Journal
Date:Nov 1, 2005
Previous Article:Population-based study of the geographic variation in colon cancer incidence in Alabama: relationship to socioeconomic status indicators and...
Next Article:Late stage (III and IV) non-small cell cancer of the lung: results of surgical resection at Inova Fairfax Hospital.

Related Articles
Data bank incomplete and future cloudy.
The National Practitioner Data Bank: the first 18 months.
The medical industry: the errors of its ways.
Clinical practice guidelines and medical malpractice: Guidelines gaining credibility in courtrooms, may eliminate expert testimony. (Doctors, Lawyers...
The truth about medical malpractice.
Utilization review and managed health care liability.
Malpractice claims: finding the silver lining.
Who's sorry now?
Public Citizen report debunks med-mal myths.

Terms of use | Copyright © 2017 Farlex, Inc. | Feedback | For webmasters