What's New in Obstetric Anesthesia in 2009? An Update on Maternal Patient Safety

Abstract

Every year, the Society of Obstetric Anesthesia and Perinatology nominates 1 individual to survey the prior year's literature and to identify the most notable articles for the science and practice of obstetric anesthesiology. This article reviews the 2009 literature, focusing on the theme of maternal patient safety, and advancing the notion of the obstetric anesthesiologist as both a perioperative and a peripartum physician. Three specific topics are reviewed: complications of obstetric anesthesia, general obstetric complications, and quality and safety in peripartum care.

In Brief

Published ahead of print September 22, 2010

COMPLICATIONS OF ANESTHESIA

Multiple sources of data confirm that obstetric anesthesia is increasingly safe. Based on billing data recorded in the United States Nationwide Inpatient Sample between 1998 and 2005, the composite of severe anesthesia-related cardiac, pulmonary, and central nervous system complications decreased from 0.19 to 0.11 per 1000 delivery-related hospitalizations.¹ Risk was unchanged for vaginal deliveries (0.07 per 1000 for the years 1998–1999 and 0.06 per 1000 for the years 2004–2005), but decreased 70% for repeat cesarean deliveries (from 0.43 to 0.13 per 1000) and 60% for primary cesarean deliveries (from 0.72 to 0.29 per 1000). The International Classification of Diseases, Ninth Revision (ICD-9) codes queried in this analysis were relatively specific, and underestimated total risk.

Using a more comprehensive list of ICD-9 codes for anesthesia-related complications, Cheesman et al.² queried billing data from New York State between 2002 and 2005. The proportion of women who experienced at least 1 anesthesia-related complication was 0.46%. Many, but not all, of these cases likely resulted from unintentional dural puncture. For the most part, factors that predict more intensive anesthetic exposure were associated with anesthesia-related complications. In 1 notable exception, women who resided in rural zip codes had significantly increased odds of anesthetic complication compared with their urban peers, regardless of mode of delivery.

Although major complications of anesthesia are increasingly rare,¹ they are often preventable, and an overview of patient safety incidents can generate lessons to improve safety for future patients.³ Although reports from the American Society of Anesthesiology (ASA) Closed Claims Project Database cannot provide accurate event rates, the reports do reveal useful information about the trends in medical liability, and the mechanisms of injury judged to be sufficiently egregious to warrant a lawsuit. A report published in 2009 compared 426 claims for injuries in obstetric patients that occurred between 1990 and 2003 with 190 cases that occurred before 1990.⁴ The annual incidence of claims filed decreased compared with the prior cohort, and the proportion of claims filed for the most serious complications also decreased (including maternal death and newborn death or brain damage). Nevertheless, newborn death or brain damage, along with maternal nerve injury, remain the most common categories of injury-generating claims filed between 1990 and 2003.

Failure to adequately manage the airway continues to be among the leading anesthesia-related causes of adverse outcomes in obstetrics. Of 69 lawsuits for maternal death or permanent brain damage filed since 1990, 36 were for injuries attributed to an anesthetic complication. Failed gas exchange or aspiration of gastric contents contributed 10 of 11 events associated with general anesthesia and 3 of 25 events associated with neuraxial anesthesia. Difficult intubation resulting in maternal hypoxia led to 7 of these 13 events; an additional 4 cases of difficult intubation led to neonatal death or brain injury. For a number of these cases, injury resulted when multiple intubation attempts led to progressive difficulty in ventilation.

It is widely believed that obstetric patients have an increased risk of failed intubation. Recent debate centers on the question of whether this is true in contemporary anesthetic practice. Current evidence is insufficient to suggest otherwise, and ongoing precaution is warranted. A report from the Australian and New Zealand College Trials Group documented a failed intubation rate of 1:275 (95% confidence interval, 1:10,000, 1:110) for general anesthetics from a cohort of 1095 cesarean deliveries in 13 institutions between 2005 and 2007.⁵ One report from the United Kingdom (UK) suggests that failed intubation rates in obstetrics may be far lower than previously estimated (0 in 3430 cases)⁶; statistical heterogeneity, the practice environment, and patient population likely determine much of this variation.

Numerous factors contribute to the increased risk of difficult intubation in obstetrics, but perhaps the most significant factor is time pressure created by decreased functional residual capacity and increased oxygen consumption in pregnant women, both of which decrease tolerance to apnea in the setting of rapid sequence induction. McClelland et al.,⁷ using a computational simulation in which a series of hypothetical patients breathed 100% oxygen for 10 minutes before a period of apnea, estimated that the laboring woman with a body mass index of 50 kg/m² will begin to rapidly desaturate after just 90 seconds of apnea. Normal-weight pregnant women in labor may tolerate approximately 3.5 minutes of apnea before significant desaturation develops.

Cricoid pressure is recommended to reduce the risk of aspiration during general anesthesia, but some critics have argued that pressure over the cricoid ring may be ineffective if it displaces the esophagus laterally. To evaluate the changes in neck anatomy produced by cricoid pressure, a study enrolled 24 healthy volunteers who underwent neck magnetic resonance imaging with and without cricoid pressure in a variety of head positions.⁸ Approximately 2 to 4 kg of pressure over the cricoid ring effectively compresses the alimentary tract at the level of the postcricoid hypopharynx, not the esophagus. Anatomic compression is evident even when pressure displaces the crico-hypopharyngeal unit laterally from the vertebral body. It is not known whether the anatomic compression visualized in this study guarantees an effective barrier to regurgitation during the induction of anesthesia.⁹

A secondary analysis of a prospective observational study of 4891 cesarean deliveries conducted in Malawi between 1998 and 2000 found that the application of cricoid pressure did not seem to prevent the development of clinical aspiration syndrome. However, the data are retrospective, subject to hindsight bias, and the application of cricoid pressure by untrained and semitrained personnel may not represent clinical effectiveness in the developed world.¹⁰

The aspiration of gastric contents has become extremely rare because of modern safety precautions. Given the current rarity of aspiration, some obstetric providers recommend food in labor to promote maternal comfort, and claim that maternal food intake during labor improves birth outcomes. A 2009 prospective trial randomized >2400 women in labor to either a light diet of solid foods or to only clear liquids to test the hypothesis that food in labor would increase the probability of vaginal delivery.¹¹ This well-conducted trial showed no difference in the rate of spontaneous vaginal delivery, instrumental delivery, cesarean delivery, rates of vomiting, or duration of labor between groups. Although consumption of food in labor may promote maternal comfort, it does not improve delivery outcomes. This study was not powered to address the question of whether food during labor is safe.

Potentially lethal complications of neuraxial anesthesia include high neuraxial block, cardiac arrest, local anesthetic toxicity, infection, and drug-induced respiratory depression. To determine the relative frequency of major neuraxial block complications, The Royal College of Anesthetists conducted a national audit of practice in the UK between September 1, 2006 and August 31, 2007.¹² Detailed data were collected for all central neuraxial blocks performed during a 2-week period in September 2006, and the reported numbers were multiplied by 25 to approximate annual activity. Based on this estimate, there were just over 160,000 epidural, 133,000 spinal, and 25,000 combined spinal-epidural procedures performed in obstetric patients in the UK that year. Complications reported in obstetric patients in the 1-year study period included 1 epidural abscess, 1 case of block-associated meningitis, 2 central neuraxial nerve injuries, 6 wrong route errors in which dilute solutions of local anesthetics were administered IV with no evidence of patient harm, 1 high spinal anesthetic with apnea and cardiovascular collapse, 1 case of bilateral subdural hematomas, and no cases of spinal canal hematoma or spinal cord ischemia. All but 4 patients either experienced no harm or demonstrated rapid recovery. Of the 4 remaining patients, 3 had partial follow-up and most likely recovered by 6 months. Therefore, the article proposes an optimistic estimate of 1 in 320,000 and a pessimistic estimate of 1 in 80,000 for major complication with permanent injury in the obstetric population. Risk was significantly higher among non-obstetric patients.

High neuraxial block may be emerging as the most important major anesthetic complication of neuraxial anesthesia in the United States (US). According to the ASA Closed Claims report, high neuraxial blocks led to 15 lawsuits for maternal death or brain injury, 10 of which involved unrecognized intrathecal catheters intended for the epidural space. Most of the adverse events occurred when local anesthetic was injected via the “epidural” catheter after negative aspiration for cerebral spinal fluid and an uneventful test dose (generally involving 45–80 mg lidocaine).⁴ A previous report from the Doctors Company described 22 cardiac arrests in labor and delivery suites. Seven followed unrecognized intrathecal catheters intended for the epidural space, and another 2 followed high spinal anesthesia for cesarean delivery.¹³

Local anesthetic toxicity is extremely rare in contemporary obstetric anesthesia practice, but has been reported in the past decade (Table 1, Ref. ¹).^14,15 Lipid emulsion is increasingly recommended as a useful strategy to resuscitate patients with local anesthetic toxicity.¹⁶ However, the optimal resuscitation strategy is far from clear. A 2009 study using a rat model suggested that high-dose epinephrine may interfere with the efficacy of lipid emulsion.¹⁷ Thirty rats received bupivacaine 20 mg/kg to induce cardiac arrest. Ventilation with 100% oxygen and chest compressions were started immediately. At 3 minutes, the rats received a randomized rescue dose of saline, lipid, or lipid with 1 of 4 concentrations of epinephrine. All doses of epinephrine improved the initial rate-perfusion product, but by 10 minutes after asystole, high-dose epinephrine (10–25 μg/kg) worsened recovery. Applicability in humans is not guaranteed. However, the results of this rat model suggest that coadministration of lipid emulsion with low-dose epinephrine (1–2.5 μg/kg) may provide the greatest chance of early and sustained recovery.

Table 1: Website Citations

Finally, in 2009, the ASA issued practice guidelines for the prevention, detection, and management of respiratory depression associated with neuraxial opioid administration.¹⁸ Patients receiving neuraxial opioids should be monitored for adequacy of ventilation, oxygenation, and level of consciousness. Table 2 summarizes the recommended monitoring intervals for healthy patients. Increases in intensity, duration, or additional methods of monitoring are recommended for patients at increased risk for respiratory depression, including patients with morbid obesity, sleep apnea, those receiving opioids by multiple routes, and coadministration of other respiratory depressants.

Table 2: Recommended Monitoring Frequency for Healthy Patients Receiving Neuraxial Opioids

GENERAL OBSTETRIC MORBIDITY

Rates of severe obstetric morbidity may be increasing, particularly for deep venous thrombosis and postpartum hemorrhage. One study of severe obstetric morbidity in the US drew data from the Nationwide Inpatient Sample for >32 million delivery hospitalizations between 1998 and 2005.¹ The authors constructed a list of ICD-9 codes designating severe complications and selected those hospitalizations with 1 of the codes of interest plus a prolonged length of stay to identify >227,000 deliveries with severe obstetric morbidity. The rates of mechanical ventilation, adult respiratory distress syndrome, renal failure, shock, pulmonary embolism, and blood transfusion all increased between 1998 and 2005. The cesarean delivery rate increased from 21% to 30% over this time period. Adjustment for the increasing rate of cesarean delivery explained almost all increases in the estimated risk of mechanical ventilation, adult respiratory distress syndrome, renal failure, and shock, but only one-third of the increase in the risk of pulmonary embolism, and one-fifth of the increase in the risk of blood transfusion.

Among the 69 lawsuits for maternal death or permanent brain injury from the ASA Closed Claims Project data in obstetric patients, 27 claims were for injuries attributed to the complications of severe maternal disease.⁴ Epidemiologically, the obstetric anesthesiologist is far more likely to encounter a catastrophe associated with the complications of hemorrhage, amniotic fluid embolism, thromboembolism, hypertensive disorders of pregnancy, or chorioamnionitis, than a disaster attributed to an anesthetic complication. Errors of omission, commission, or inadequate documentation may increase liability for deaths that are not attributable primarily to an anesthetic complication.

A series of recent studies have documented increasing rates of postpartum hemorrhage in developed countries.^19–22 An international coalition of investigators pooled data from 12 countries and documented significant increases in the rates of clinical hemorrhage and blood transfusion in Australia, Canada, and the US.²² Much of the increase is attributed to postpartum uterine atony. Although abnormal placentation underlies some of the most severe cases of maternal hemorrhage,^4,23,24 it accounts for a relatively minor fraction of all cases.^19,20 Induction of labor has been associated with higher rates of severe postpartum hemorrhage than spontaneous labor, regardless of mode of delivery.²⁵

Pregnancy increases the risk of both venous thromboembolism and coagulopathy, and the placenta has a strong role in propagating hemostatic derangements.²⁶ The fetal syncytiotrophoblasts that envelop the chorionic villi and line the intervillous space express high levels of tissue factor,²⁷ readily stimulating the maternal extrinsic coagulation cascade. In addition, trophoblasts express a number of endothelial cell proteins (such as thrombomodulin and endothelial protein C receptor) that further modulate maternal coagulation and fibrinolysis. Endothelial protein C receptor is important for early pregnancy maintenance,²⁸ acting through protease-activated receptors to block placental apoptosis and thrombosis.²⁹ Later in pregnancy, maternal changes in systemic coagulation factors interact with fetal proteins on the maternal chorionic surface to promote coagulation, impair anticoagulation, and promote antifibrinolysis.^26,30 In addition, as the chorionic villi mature, the cytotrophoblasts fuse with the more superficial syncytiotrophoblasts, and this process mobilizes anionic and prothrombotic phospholipids (phosphatidylserine) to the cellular surface in contact with maternal blood.³¹

In 2009, the World Health Organization issued guidelines for the management of postpartum hemorrhage and retained placenta (Table 1, Ref. ²). These guidelines provide a general evidence-based stepwise algorithm for postpartum hemorrhage, beginning with (1) uterine massage and oxytocin, (2) ergometrine, (3) prostaglandin, and (4) tranexamic acid (if the bleeding is attributed in part to trauma). Recommended nonoperative and operative maneuvers include (1) bimanual uterine compression as a temporizing measure in the treatment of postpartum hemorrhage due to uterine atony after vaginal delivery in low-resource settings, (2) intrauterine balloon tamponade (but not uterine packing), (3) uterine artery embolization, (4) compression sutures, (5) vessel ligation, and (6) subtotal hysterectomy. Isotonic crystalloids are recommended in preference to colloids for fluid resuscitation. Health care facilities should adopt formal protocols for the management of postpartum hemorrhage. The states of New York (Table 1, Ref. ³), Illinois (Table 1, Ref. ⁴), and California (Table 1, Ref. ⁵) have recently deployed maternal hemorrhage protocols, all of which are designed to trigger clear action and to avoid treatment delays.

Increasing rates of severe obstetric complications in the US may be a function of increasing rates of risk factors for adverse obstetric outcomes. Preexisting medical conditions increased from 4.1% to 4.9% of all pregnancies between 1993–1997 and 2001–2005, based on data from the National Hospital Discharge Survey.²¹ The most significant increases were noted for chronic hypertension (1.5% to 1.9%) and asthma (0.7% to 1.3%). Increasing rates of maternal obesity and childbearing after age 35 years may explain some of the observed increases in obstetric morbidity and mortality.

Hypertensive disorders of pregnancy seem to underlie a significant portion of cases of severe obstetric morbidity. The overall prevalence of hypertensive disorders of pregnancy among delivery hospitalizations in the US increased significantly from 6.7% in 1998 to 8.1% in 2006, according to billing data from the Nationwide Inpatient Sample.³² These hospitalizations accounted for 57% of women who developed acute renal failure, and ≥27% of those who required mechanical ventilation, or developed pulmonary edema, puerperal cerebrovascular disorders, or respiratory distress syndrome. The highest rates of severe morbidity were concentrated among women with severe preeclampsia or eclampsia, with complication rates ranging from 12.7 to 184.5 per 10,000 deliveries for pulmonary embolism and disseminated intravascular coagulation syndrome, respectively. Severe preeclampsia or eclampsia also increased over this time period from 0.9% to 1.2% of delivery-related hospitalizations.

Finally, the dramatic increase in birth-related complications may be both a cause and a consequence of increasing cesarean deliveries. The US cesarean delivery rate has increased every year since 1996, most recently from 31.8% in 2007 to 32.3% in 2008.^33,34 Many experts are calling for strategies to reduce this rate, citing untenable societal cost and patient risk.³⁵

As the rates of cesarean delivery have increased, the corresponding rates of vaginal birth after cesarean delivery (VBAC) continue to decrease in the US. The policy of the American College of Obstetricians and Gynecologist recommending the immediate availability of a surgical team during a trial of labor after cesarean (TOLAC),^36,37 fear of uterine rupture, and obstetric liability^38,39 (Table 1, Ref. ⁶) have all served to decrease the rates of VBAC. In 2006, approximately 8% of pregnant women with a history of cesarean delivered vaginally.⁴⁰ Based on 2006 data from the Nationwide Inpatient Sample, 14.6% of all pregnant women had a previous cesarean delivery, 2.5% of all pregnant women attempted a TOLAC, and only 1.5% delivered by VBAC.⁴¹ TOLAC has become so rare that only 26% of cases of uterine rupture were diagnosed during a TOLAC in a recent report of all uterine ruptures diagnosed in hospitals participating in the Hospital Corporation of America network.⁴² In other words, in contemporary obstetric practice in the US, most uterine ruptures develop before the onset of labor or during labor in women without a uterine scar.

New data may begin to encourage more pregnant women with a history of cesarean delivery to plan a vaginal birth. For example, a retrospective cohort study in an academic center of 672 women with 1 prior cesarean delivery demonstrated a 74.2% success rate (VBAC after TOLAC) and suggested that for this population, TOLAC, as opposed to elective repeat cesarean delivery, reduced overall maternal and neonatal risk.⁴³ The aggregate cost for patients managed with planned cesarean delivery was 19% higher than for those managed by planned vaginal birth. Recent work from the National Institutes of Child Health and Human Development Maternal-Fetal Medicine Units Network suggests that when the predicted probability of vaginal birth exceeds 70%, total maternal and neonatal morbidity is not different when compared with elective cesarean delivery.⁴⁴ Variables known in the early antepartum period can be used to predict an individual woman's specific probability of achieving a VBAC.^45–47 In addition, sonographic measurement of the lower uterine segment thickness may be a useful tool to estimate the risk of uterine rupture.⁴⁸ Future work is needed to demonstrate whether the use of either a prediction model or sonographic measurement can reduce the chance of adverse outcomes related to TOLAC.

Regardless of trends in VBAC, overall increases in maternal age and the rates of obesity, multiple gestation, patient coexisting disease,²¹ and cesarean deliveries will continue to challenge anesthesiologists as we strive to provide high-quality anesthetic care that ensures both maternal comfort and maternal and perinatal safety.

QUALITY AND SAFETY IN PERIPARTUM CARE

Amnesty International published a report in March 2010 entitled “Deadly Delivery, The Maternal Health Care Crisis in the USA.”⁴⁹ The report tells the story of 12 women who died needlessly and tragically as the result of gaps in the US maternal health care system. Each year in the US approximately 600 women die,⁵⁰ 68,000 experience a severe obstetric morbidity,¹ and 1.7 million experience delivery-related complications.²¹ The US maternal mortality ratio (maternal deaths per 100,000 live births) is higher than that of 40 other countries.⁵¹ Amnesty International condemns this record as both a public health emergency and a human rights crisis.⁴⁹

On a national level, quality measurement in US maternal health care has lagged behind that of other medical specialties. Traditional measures by the Joint Commission and the Agency for Healthcare Research and Quality focused primarily on the mode of delivery and the rate of third- or fourth-degree perineal lacerations (Table 1, Refs. ⁷ and ⁸).⁵² The Surgical Care Improvement Project Core Measure Set excludes obstetric patients (Table 1, Ref. ⁹), and there is no comparable national partnership to deploy process measures for obstetric care. It is possible that measuring the quality of maternal health care has not received significant national federal attention because Medicare does not pay for it.⁵³ Nevertheless, childbirth is among the leading indications for hospital admissions (Table 1, Ref. ¹⁰), and cesarean delivery is the most common major surgery performed in the US.⁵⁴

In 2008, the National Quality Forum endorsed 17 process measures of maternal and perinatal health care (Table 1, Ref. ¹¹).⁵³ Measures relevant to the practice of anesthesiology include the rate of cesarean delivery for low-risk first births, the rate of prophylactic antibiotics administered within 1 hour before surgical incision or at the time of delivery, the proportion of patients undergoing cesarean delivery who receive either pneumatic compression devices or pharmacologic deep vein thrombosis prophylaxis before surgery, and the rate of elective delivery before 39 weeks of gestation. Many of the measures have been criticized because they will require medical record audit unless electronic medical records are adapted to capture the information routinely.

More than 100 quality and safety indicators related to the practice of anesthesia have been recently cataloged.⁵⁵ Of these, 3 indicators are specific for the practice of obstetric anesthesia: the proportion of patients who deliver within 30 minutes of request for immediate lower-segment cesarean delivery, the proportion of patients who receive general anesthesia for cesarean delivery, and the proportion of obstetric patients with neuraxial block who experience postdural puncture headache.⁵⁵

A decision-to-delivery interval of 30 minutes has served as a clinical benchmark in obstetrics for many years,⁵⁶ but notably was excluded from the 2008 list of indicators endorsed by the National Quality Forum. Particularly for nonreassuring fetal heart patterns, the evidence linking rapid delivery with improved neonatal outcomes is weak and conflicting.^57,58 Only a subset of neonates may actually benefit from rapid delivery. Focusing exclusively on cases of persistent fetal bradycardia (fetal heart rate <110 bpm for at least 3 minutes), 1 retrospective cohort attributed 16% of cases to irreversible causes, including placental abruption, cord prolapse, uterine rupture, preeclampsia, or failed instrumental delivery. For these irreversible cases, the uterine arterial pH declined at an average rate of 0.011 per minute of bradycardia-to-delivery interval, and infants born within 15 minutes of the decision to operate demonstrated the lowest risk of uterine arterial pH <7.0 (21% vs 40%).⁵⁷ In the remaining 84% of cases, the cause of fetal bradycardia was unknown or reversible, and there was no relationship between the bradycardia-to-delivery interval and the umbilical arterial pH at birth.

Team training can effectively reduce the decision-to-delivery interval for cases of true emergency.⁵⁹ In 2000, a maternity unit in Bristol, UK caring for 5000 to 6000 deliveries per year, began to require annual interdisciplinary team training for all labor and delivery staff, including a station focused on the management of umbilical cord prolapse.⁶⁰ Trend analysis shows that the median decision-to-delivery interval for actual cases of cord prolapse decreased from 25 minutes to 15 minutes after training sessions began.⁶⁰ The unit accomplished this without decreasing the use of spinal anesthesia and the authors point to greater efficiency in transporting the patient to the operating room as the most important factor in saving time. With a total sample of only 62 patients, neonatal outcome was not different between time periods, but the composite of neonatal intensive care unit admission, Apgar score <7, or stillbirth decreased from 41% to 18%.

The Adverse Outcome Index (AOI) is the proportion of deliveries complicated by 1 or more specified adverse events, including 6 maternal events (death, uterine rupture, return to the operating room, maternal intensive care admission, transfusion, third- to fourth-degree lacerations) and 4 neonatal events (death, Apgar score <7, neonatal intensive care admission, and birth trauma) (Table 1, Ref. ¹²). This index was developed by expert consensus, with guidance from the American Congress of Obstetricians and Gynecologists Patient Safety Committee, to focus measurement on indicators that are widely applicable, reliable, and easily measured.⁶¹ Although these indicators may be treated as a simple composite rate (the proportion of all births complicated by at least 1 of the indicators), a weighting system has also been developed to reflect severity.

The AOI was used to measure patient safety improvement between 2004 and 2006 at an academic center with 4650 deliveries per year.⁶² After review by 2 independent consultants in perinatal risk, unit leaders hired a patient safety nurse who established an anonymous event reporting system, tracked events, followed up with ongoing root cause analysis, and presented her findings to a patient safety committee. The unit clarified a series of clinical protocols, and introduced team training and electronic fetal monitoring certification for nursing staff. Over time, the composite AOI decreased from 3.25% to 1.75%, and staff perception of the overall patient safety climate improved by 66%.

A community hospital labor and delivery unit also reported that team training followed by focused quality improvement efforts improved the AOI from 7% to 4% over 4 years.⁶³ Among other specific interventions, the anesthesia group in this hospital deployed an in-service lecture for labor and delivery staff to review the delivery room nurse's role in assisting with general anesthesia.

Team training based on crisis resource management theory does not by itself improve safety as measured by the AOI.⁵⁹ Multidisciplinary simulations of obstetric emergencies do improve processes of care, regardless of whether training takes place in a simulation center or local hospitals, or whether training is informed by crisis resource management theory.⁶⁴ Based on a review of reports from successful obstetric emergency training programs, the most important components of effective training appear to include institution-level incentives to train, multi-professional training of all staff in the unit, teamwork training integrated with clinical teaching, and the use of high-fidelity simulation models.⁶⁵ High-fidelity simulation does not necessarily equate with advanced technology; for example, training using a patient-actor may provide better psychological and environmental fidelity than training with a computerized manikin.⁶⁶ Instruments to measure teamwork behavior in the operating room⁶⁷ and for obstetric simulation⁶⁸ were also published in 2009.

Finally, team training may be an effective strategy to engage teams of care providers in local, multidisciplinary, ongoing, proactive, and effective quality improvement initiatives, and these sustained efforts may then improve patient outcomes.^62,63 At the same time, centralized corporate, professional, or government-led organizations may contribute to patient safety by developing best practice guidelines, coordinating surveillance and analysis of both frequent and rare patient safety incidents, delivering education and simulation programs to local facilities, and supporting local quality improvement efforts.^69–73

ACKNOWLEDGMENTS

The author acknowledges with appreciation Linda Polley, Mary Lou Greenfield, Lawrence Tsen, John Sullivan, and the University of Michigan Health System Division of Obstetric Anesthesiology research assistants.