Liability Associated with Obstetric Anesthesia: A Closed Claims Analysis

The ASA Closed Claims Project is a structured evaluation of adverse anesthetic outcomes obtained from the closed claim files of 35 United States professional liability insurance companies. Claims for dental damage are not included in the database.

The data collection process has been previously described in detail.19,20Briefly, a closed claim file was reviewed at the professional liability insurance company by a practicing anesthesiologist. The file typically consisted of relevant hospital and medical records, narrative statements from involved healthcare personnel, expert and peer reviews, summaries of depositions from plaintiffs, defendants, and expert witnesses, outcome reports, and the cost of settlement or jury award. The reviewer completed a standardized form and narrative summary that recorded information about patient characteristics, surgical procedures, sequence and location of events, critical incidents, clinical manifestations of injury, standard of care, and outcome. Forms and summaries completed by the on-site anesthesiologist reviewer were subsequently reviewed by a physician member of the central Closed Claims Committee before incorporation in the database.

The physical or psychological injury for which the patient was seeking compensation was recorded in each claim. There was no apparent injury in some claims, and some claims had multiple injuries. In the case of brain damage followed by death, death was considered the complication. The patient and newborn in each claim were assigned a severity of injury score that was designated by the on-site reviewer using the insurance industry’s 10-point scale that rates severity of injury from 0 (no injury) to 9 (death). For the purposes of analysis, injuries were grouped into temporary/nondisabling (score = 0–5) versus  permanent/disabling (score = 6–8) versus  death (score = 9). The damaging event that caused the injury was determined by the on-site anesthesiologist reviewer and confirmed by the Closed Claims Committee. Appropriateness of anesthesia care was rated as appropriate (standard), substandard, or impossible to judge on the basis of reasonable or prudent practice at the time of the event by the on-site reviewer. A previously published study found reliability of reviewer judgments to be acceptable.21 

From the ASA Closed Claims Project database of 7328 cases, those associated with obstetric procedures that occurred from 1990 to 2003 (1990 or later ) were included for the current study. These claims were reviewed again in detail by two of the authors (JMD, KBD). These claims (1990 or later) were then compared to obstetric claims for procedures that occurred before 1990 (n = 190) and were reported previously by Chadwick et al.  7(pre-1990 ). Most of these claims (92%) reported by Chadwick et al.  involved procedures that occurred between 1975 and 1985.7Obstetric claims were classified as associated with vaginal delivery or Cesarean section on the basis of the delivery mode that occurred. Planned vaginal delivery that was changed to Cesarean section was classified as Cesarean section. If both regional and general anesthesia were administered, the claim was classified according to the technique associated with the injury in the claim.

Claims from 1990 or later for maternal and newborn death or permanent brain damage were independently classified as anesthesia-related by two of the authors (JMD, KBD) if there was a possible anesthesia contribution to the injury. Two claims for newborn permanent brain damage were excluded from this analysis because the alleged newborn brain damage was not substantiated. In claims in which there was disagreement about anesthesia-related injury, classification was resolved after discussion. Specific maternal, obstetric, and newborn factors associated with maternal and newborn death/brain damage were abstracted from the narrative summaries of these claims. The presence of anesthesia factors such as alleged anesthesia delay, neonatal resuscitation by the anesthesiologist, and poor communication between the anesthesiologist and the obstetrician (usually concerning the urgency of the Cesarean section) were also abstracted from the summaries of claims for newborn death/brain damage.

Payments were adjusted to 2007 dollar amounts using the Consumer Price Index.**In a subset of claims in which adequate information was available on the datasheet (for claims collected after 2000 due to a revision of the datasheet) or narrative summary, payments were separated into payments on behalf of the anesthesiologist (i.e. , on behalf of an anesthesiologist, anesthesiology group, resident, or certified registered nurse anesthetist) and total payments by all parties.

Differences in categorical variables in patient characteristics, injuries, and liability between pre-1990 and 1990 or later claims were compared using the Chi-square tests and Fisher exact test, with specific comparisons made using the z test if the overall Chi-square test was statistically significant. Age in years was compared in the two groups using Student t  test. Payment amounts (excluding $0 and missing data) were compared in the two groups using the Kolmogorov-Smirnov test. Kappa statistics were calculated between two raters (JMD, KBD) on the independent classification of anesthesia-related versus  nonanesthesia-related for maternal and neonatal death/brain damage claims in the 1990 or later time period. The relationship between anesthesia-related claims and other factors such as anesthesia delay, poor communication, substandard care, and payment on behalf of the anesthesiologist (and anesthesia care team or anesthesia group) in these claims was determined using Fisher exact test. All statistical analysis was conducted with SPSS 12.0.1 for Windows (SPSS Inc, Chicago, IL). Confidence intervals for proportions were calculated as equal-tailed Jeffreys prior intervals according to SPSS Resolution Number 37680. Statistical significance was accepted at P < 0.05. For multiple comparisons of payments with newborn death/brain damage, maternal nerve injury, and maternal minor injuries compared to maternal death/brain damage, a Bonferroni correction was used.

There were 426 claims associated with obstetric anesthesia from 1990 or later (246 [58%] Cesarean section and 180 [42%] vaginal delivery). Compared to pre-1990 claims (n = 190), the proportion of obstetric claims from 1990 or later associated with Cesarean section decreased (P = 0.029) and the proportion of claims associated with general anesthesia decreased (P < 0.001; table 1).

The pattern of injuries in the claims also changed over time. Compared to pre-1990 claims, the proportion of obstetric claims from 1990 or later associated with maternal death (P = 0.002) and newborn death or brain damage (P = 0.048) decreased (fig. 1). In contrast, the proportion of obstetric claims from 1990 or later associated with maternal nerve injury (P < 0.001) and back pain (P = 0.012) increased (fig. 1). Respiratory causes of injuries decreased in claims from 1990 or later (24% in pre-1990 claims vs.  4% in 1990 or later claims, P < 0.001; table 1). Claims related to inadequate oxygenation/ventilation (P = 0.006), aspiration of gastric contents (P = 0.012), and esophageal intubation (P = 0.007) decreased in 1990 or later claims, whereas claims related to difficult intubation did not change (table 1).

Compared to pre-1990 claims,7the proportion of claims with substandard care decreased, and appropriate care increased in 1990 or later claims (P < 0.001; table 1). The proportion of claims with payment to the plaintiff decreased from 58% in pre-1990 claims to 42% in 1990 or later (P < 0.001; table 1). Although median payments tended to be higher in pre-1990 claims, the ranges of payments were very broad and were not significantly different between the two time periods (table 1).

The most common injuries leading to obstetric claims in 1990 or later were newborn death/brain damage (21%) and maternal nerve injury (21%; fig. 1). Claims for maternal minor injuries (e.g. , headache, back pain, pain during surgery, and emotional distress) made up a large proportion (28%) of obstetric claims.

Liability characteristics for selected maternal and newborn injury claims are shown in figure 2. A payment was made on behalf of the anesthesiologist in only 21% of claims for newborn death/brain damage (fig. 2). A payment was made on behalf of the anesthesiologist more frequently in claims for maternal death/brain damage than for newborn death/brain damage (P < 0.001, fig. 2). Payment amounts for newborn and maternal death/brain damage were similar. Payment amounts for maternal nerve injury and minor injuries were reduced compared to maternal death/brain damage (P < 0.001).

Most (71%) claims for newborn death/brain damage (n = 91) were associated with a nonreassuring fetal heart tracing and urgent or emergent Cesarean section (table 2). Anesthesia care was judged as having a possible contribution in only 22% of newborn death/brain damage claims (kappa = 0.84; P < 0.001). Other factors observed in newborn death/brain damage claims included a variety of maternal and fetal conditions known to be associated with newborn encephalopathy (table 2).

When compared to newborn death/brain damage claims with no anesthesia contribution, anesthesia delay (P = 0.001), poor communication (P = 0.007), and substandard anesthesia care (P < 0.001) occurred more frequently when anesthesia care may have contributed to the newborn outcome (fig. 3). The anesthesia delays in the 11 newborn death/brain damage claims ranged from 10 to 70 min, with a mean delay of 40 min (SD = 22 min). Delay due to anesthesia care was a factor in 11 (55%) of 20 anesthesia-related newborn death/brain damage claims (fig. 3). The anesthesiologist was not in the hospital for approximately half (n = 6) of these cases. Poor choice of anesthesia technique, with inappropriate prolonged attempts to administer regional rather than general anesthesia, probably contributed to some of the delays (n = 3). Poor communication between the obstetrician and anesthesiologist, primarily regarding the urgency of the Cesarean section, also contributed to nearly two thirds of the anesthesia delays (n = 7). Other anesthesia contributions to newborn death/brain damage included difficult intubation resulting in maternal hypoxia (n = 4) and profound hypotension after neuraxial block (n = 4). Care was judged to be substandard in 82% of claims in which anesthesia care may have contributed to newborn death/brain damage (fig. 3).

There was a significant association of payment by the anesthesiologist with possible anesthesia contribution to the newborn injury (P < 0.001; table 3). Payment on behalf of the anesthesiologist was also associated with the judgment of substandard anesthesia care (P < 0.001; table 3). A payment was made on behalf of the anesthesiologist in only 7 claims (10%) in which anesthesia care was judged as not contributing to the newborn death/brain damage.

Nerve injury (n = 89) was the leading maternal injury in obstetric claims from 1990 or later, most of which was temporary or nondisabling (80%). All but one nerve injury was associated with regional anesthesia, and nerve injury was more commonly associated with vaginal delivery than with Cesarean section (P = 0.001). Most cases of nerve injury (63%) could have been caused by the administration of regional anesthesia as evaluated by the on-site anesthesiologist reviewer, but no specific event leading to nerve injury could be identified in 18 claims (20%), and 12 (13%) were related to patient condition or delivery. Radiculopathy of a lumbar or sacral root accounted for the majority of injuries (table 4). Spinal cord injury resulted in paraplegia in 10 nerve injury claims (11%). When known, the etiology for spinal cord injuries was epidural hematoma (n = 4, only 1 of which had a coagulopathy), epidural abscess (n = 4), direct injection into the cord (n = 2), and anterior spinal artery syndrome (n = 1). The majority of nerve injury claims were judged to have appropriate care, and payment was made by the anesthesia providers in fewer than one third of claims for nerve injury (fig. 2).

Causes of maternal death (n = 47) and permanent brain damage (n = 22) with general (n = 28) or regional (n = 41) anesthesia are shown in table 5. The most common anesthetic causes of maternal death/brain damage in claims associated with general anesthesia were difficult intubation and maternal hemorrhage (table 5). The seven difficult intubation injuries occurred between 1991 and 1998, mostly upon induction (six of seven cases). These claims involved multiple intubation attempts leading to progressive difficulty with ventilation. In two of the claims, tracheal intubation was assessed preoperatively as being possibly difficult, with a backup plan to awaken the patient and perform fiberoptic intubation. However, progressive airway difficulties occurred while attempting to awaken the patient. In two claims, general anesthesia was induced after failed regional block.

Maternal hemorrhage was associated with an inability of the anesthesiologist to keep up with blood loss despite best efforts (n = 10) and with inadequate fluid replacement in which it should have been possible to resuscitate during surgical bleeding (n = 1). Causes of maternal hemorrhage included subcapsular hepatic bleeding in a preeclamptic patient, placenta previa (n = 2), placenta accreta/placenta percreta (n = 2), and uterine rupture (n = 1).

The most common anesthetic cause of maternal death/brain damage in regional anesthesia claims was high neuraxial block (n = 15; table 5). Twelve claims involved epidural anesthesia (ten accidental intrathecal injections, two high epidural blocks), and three claims involved subarachnoid blocks. Half (53%) were placed for vaginal delivery, and half (47%) were placed for Cesarean section. Most high blocks associated with epidural anesthesia occurred with dosing of the epidural after negative aspiration for cerebrospinal fluid and an uneventful test dose (generally involving a test dose of 45–80 mg of lidocaine). In two claims, epidurals were repeated at a different interspace due to a wet tap. A high subarachnoid block occurred in two patients despite usual doses of local anesthetic. In most (80%) of the cases of high neuraxial block, there was a delay in recognizing and/or treating the sequelae of the high block (n = 12). In two of these cases, inadequate monitoring contributed to the delay, and four of these patients had to be transferred to the operating room for resuscitation due to inadequate emergency airway equipment or drugs in the labor room. Care was judged as substandard in most (73%) of the claims involving high neuraxial block.

Neuraxial cardiac arrest (defined as the sudden onset of severe bradycardia or cardiac arrest during neuraxial block in the absence of a high block with relatively stable hemodynamics preceding the event)22occurred in two claims. Of note, there were no claims for maternal death/brain damage due to intravascular injection of local anesthetic. Other causes of maternal death/brain damage included maternal conditions, as shown in table 5.

Claims for maternal death/brain damage were judged as having a possible anesthesia contribution in 64% of claims (kappa = 0.85; P < 0.001). There was a significant association of payment by the anesthesiologist with possible anesthesia contribution (P < 0.001) and substandard anesthesia care (P < 0.007). However, in five claims, payment was made on behalf of the anesthesiologist despite the lack of possible anesthesia contribution to the injury. Factors cited as contributory to payment in these claims included illegible, inaccurate, and incomplete anesthetic records (such as a 30-min gap in the record at the time of the critical events), anesthesiologist with poor English language skills or who failed boards multiple times, and damaging comments from the obstetrician concerning the anesthesia care.

The major finding of this closed claims review is that, compared to pre-1990 obstetric claims, the proportion of maternal death and newborn death/brain damage decreased in obstetric claims from 1990 or later, whereas the proportion of claims for maternal nerve injury increased. Claims for newborn death/brain damage resulted in a lower frequency of payments on behalf of anesthesiologists (21%) than maternal death/brain damage claims (60%). Anesthesia-related claims for newborn death/brain damage were associated with anesthesia delay, poor communication, and substandard anesthesia care.

Analysis of data collected from the ASA Closed Claims project has a number of limitations that have been previously described.19,20,23The database does not contain claims on all adverse anesthetic events, and it lacks denominator data on the number of anesthetics performed annually. Only claims from participating liability insurance organizations are included. Consequently, risk for specific injuries and populations cannot be estimated. Due to the large number of variables examined, false positive findings may result. Other limitations include only modest interobserver agreement regarding appropriateness of care21and outcome bias in the judgment of standard of care.24Although the long statute of limitations for newborn injury may result in incomplete claims data for newborn brain damage, the year of injury is similar in maternal and newborn claims in this study. For the purposes of this analysis, we combined claims for death and brain damage because they represent a continuum of severity of injury with similar damaging events. In contrast, other analyses of patterns of injury have focused primarily upon maternal death and newborn brain injury. Despite these deficiencies, the database provides useful information on large numbers of rare adverse events and a snapshot of liability in the practice of anesthesiology.

Our finding of a decrease in proportion of claims for maternal death and increase in the proportion of nerve injury in the 1990 or later claims compared to earlier claims7may reflect changes in anesthesia practice in general8and in obstetric anesthesia practice specifically over the last three decades.9–11These surveys have shown a steady decrease in use of general anesthesia for elective Cesarean sections, along with an increase in use of epidural anesthesia for labor analgesia.9–11In addition, the decrease in the proportion of claims for maternal death mirrors those in the closed claims database at large,25and it may reflect changes in drugs, training, an emphasis on safety, changes in legal strategies, or other events. Decreases in respiratory events in the 1990 or later obstetric claims are probably associated with the use of respiratory system monitors in modern anesthesia practice,8the decrease in the use of general anesthesia in obstetric practice, and the enhanced awareness of the risk of aspiration of gastric contents in the obstetric patient.26Although claims for newborn death and brain damage still remain a frequent (21%) claim against anesthesiologists, the reduction in these claims over time is probably related to the recognition beginning in the mid-1980s that most cases of newborn brain injury are not related to birth asphyxia.5,12–18Current evidence suggests most cases of newborn brain damage are due to antenatal factors, and only a minority are related to intrapartum hypoxia, most of which are not preventable.5,12–18Increasing public awareness of these antenatal factors may further reduce anesthesiologists’ liability for these claims in the future.

Anesthesiologist exposure to newborn injury claims may occur because the anesthesiologist serves as a “deep pocket” (especially in the case of an underinsured obstetrician). Although anesthesiologists may be named on a lawsuit, our study found that anesthesiologist liability (e.g. , payments) for newborn death/brain damage was limited, most likely because the obstetrician is primarily responsible for fetal wellbeing.

Obstetric, not anesthetic, causes are more common for peripartum hypoxic brain injury.27Therefore, working with obstetricians to decrease the incidence of this newborn death and brain damage is probably the best way to reduce anesthesiologists’ liability. We found that anesthetic events were rare; other than an intraoperative catastrophe (e.g. , difficult intubation, high block, severe hypotension), more than half of anesthesia-related events involved an anesthesia delay that resulted from being away from the hospital or from a poor choice of anesthesia technique. A decision to start of Cesarean delivery interval for emergency Cesarean section within 30 min is the international standard for fetal compromise.8‡‡Although the scientific evidence to support the standard is weak,29,30failure to meet this target time may be part of the judgment of the substandard care.28The anesthesia care team should therefore ensure that the patient is ready for surgical incision within 30 min of the decision for Cesarean section; if exceeded, thorough documentation of the reason for delay in the medical record is essential.

Newborn outcome may be improved by better communication between obstetrician and anesthesiologist, particularly concerning the urgency of Cesarean section. In our review, more than a third of anesthesia-related newborn death and brain damage had poor communication between the obstetrician and anesthesiologist. In a 2004 Joint Commission Sentinel Event Alert,§§the leading preventable cause of neonatal death and brain damage in 47 cases was miscommunication between care providers. In many practices, anesthesiologists are on-call for obstetrics from home. This practice may be acceptable in low-risk cases; however, in high-risk patients (e.g. , patients desiring vaginal birth after Cesarean section who are at higher risk of uterine rupture), a delay in time to get to the hospital can prove fatal to both mother and baby. Failure to communicate the urgency of a Cesarean section may compound this issue and result in an inappropriate choice of anesthesia technique. Emphasis on improved communication between all providers caring for individual patients, especially those at high risk, may therefore help avoid poor outcomes.31 

The most common anesthetic causes of maternal death or permanent brain damage in the claims included difficult intubation, maternal hemorrhage, and high neuraxial block, consistent with anesthetic causes of death in published reports.32–35There were no claims related to intravascular injection of local anesthetics, consistent with changes in clinical practice in the mid-1980s with the withdrawal of 0.75% bupivacaine in obstetrics and increased use of test doses and fractionated administration of local anesthetics.26The use of a test dose and incremental dosing of epidural catheters is recommended for early detection of accidental intrathecal injection and prevention of high neuraxial block.36However, delays in recognizing and treating cardiorespiratory collapse secondary to a high block as well as inadequate resuscitation equipment continue to result in maternal injury. Even more cautious incremental dosing of epidurals may aid in early detection of an accidental intrathecal injection and prevent a high block. In the event of a high neuraxial block, all staff should be familiar with the location of the code cart equipped with appropriate equipment and drugs. All epidural carts should include airway equipment and emergency drugs to allow for immediate resuscitation.

Although Mhyre et al.  35did not find any cases of anesthesia-related death on induction of general anesthesia in their review of 15 anesthesia-related or anesthesia-contributing maternal deaths in the state of Michigan between 1985 and 2003, difficult intubation resulting in death/brain damage occurred upon induction of anesthesia in our review. These claims originate from injuries in 1991 to 1998 and largely predate widespread use of the laryngeal mask airway as a bridge in the management of difficult intubation. Although our numbers are small, the leading causes of nonanesthesia maternal death/brain damage also reflect those reported in the literature (e.g. , embolism, hemorrhage, and hypertensive disease of pregnancy).32,34 

Our finding that claims for maternal death/brain damage resulted in a higher frequency of payments on behalf of anesthesiologists reflects an increased likelihood of an anesthesia contribution to the maternal injury compared to newborn injury. In addition, it reflects the ASA policy that the anesthesiologist is primarily responsible for the mother and that personnel other than the surgical team should assume responsibility for resuscitation of the depressed newborn.‡‡Payments made on behalf of anesthesiologists for maternal death/brain damage were also associated with the judgment of substandard anesthesia care; however, outcome bias may have confounded this judgment.24 

Our finding that nerve injury is the most common injury associated with obstetrical anesthesia claims reflects the increased liability for nerve injury with regional anesthesia.37Although a regional block may have caused the nerve injury in nearly two thirds of the nerve injury claims in our review, nerve injury resulting from obstetric causes (e.g. , pregnancy, vaginal delivery, fetal position, maternal position during the second stage of labor) is more likely than nerve injury resulting from regional anesthesia.38A good knowledge of neuroanatomy together with specialist neurologic examination and appropriate investigation, such as electromyography, can aid in the accurate diagnosis of nerve injuries. Serious disabling spinal cord injuries resulting in paraplegia were noted in 11% of nerve injury claims, with causes consistent with those described in the literature.39–41 

Newborn death/brain damage has decreased, but it still remains a common injury leading to obstetric anesthesia malpractice claims. Maternal nerve injury and newborn death/brain damage were the most common complications in obstetric anesthesia malpractice claims from 1990 or later. Claims for newborn death/brain damage resulted in a lower frequency of payments on behalf of anesthesiologists than maternal death/brain damage claims. Anesthesia-related claims for newborn death/brain damage were associated with anesthesia delay, poor communication, and substandard anesthesia care in response to difficult intubation or block-related hypotension. Delays in diagnosis and resuscitation of high neuraxial block were preventable causes of maternal death/brain damage.

The authors thank John Campos, M.S., Research Consultant, and Linda Stephens, Ph.D., Research Scientist, for data management and also thank Lynn Akerlund, Research Coordinator, for secretarial assistance. They are members of the Closed Claims Project research staff in the Department of Anesthesiology at the University of Washington, Seattle, Washington. The authors also thank the closed claims reviewers from the American Society of Anesthesiologists and the following liability insurance companies who have given permission to be acknowledged: Anesthesia Service Medical Group, Inc., San Diego, California; Armed Forces Institute of Pathology, Silver Spring, Maryland; Ascension Health, St. Louis, Missouri; COPIC Insurance Company, Denver, Colorado; Department of Veterans Affairs, Washington, D.C.; The Doctors’ Company, Napa, California; ISMIE Mutual Insurance Company, Chicago, Illinois; MAG Mutual Insurance Company, Atlanta, Georgia; Medical Liability Mutual Insurance Company, New York, New York; Midwest Medical Insurance Company, Minneapolis, Minnesota; Mutual Insurance Company of Arizona, Phoenix, Arizona; NORCAL Mutual Insurance Company, San Francisco, California; Pennsylvania Medical Society Liability Insurance Company, Mechanicsburg, Pennsylvania; Physicians Insurance A Mutual Company, Seattle, Washington; Physicians Insurance Company of Wisconsin, Madison, Wisconsin; Preferred Physicians Medical Risk Retention Group, Shawnee Mission, Kansas; ProMutual (Medical Professional Mutual Insurance Company), Boston, Massachusetts; Risk Management Foundation, Cambridge, Massachusetts; State Volunteer Mutual Insurance Company, Brentwood, Tennessee; The University of Texas Medical System, Austin, Texas; Utah Medical Insurance Association, Salt Lake City, Utah.