Obstetric caregivers are plagued with lawsuits alleging negligence for suboptimal outcomes. Some of those claims are unjustified, but many have merit. We are obligated to create systems designed to minimize the potential for errors that harm our patients. A variety of safety initiatives have been shown to improve patient outcomes in several centers in the United States, but it has been difficult to document the expected association between those results and reduced liability premiums. Furthermore, some individuals and institutions have been reluctant to adopt safety tools such as electronic fetal monitoring certification for all staff working on their Labor and Delivery floor, protocols for managing common clinical scenarios, simulation drills for dealing with uncommon dangerous events, and pre-procedure checklists because of the paucity of evidence based data documenting the effectiveness of those approaches. It is time to move forward with these and other safety initiatives in a serious national attempt to eliminate all preventable adverse patient outcomes in our specialty.
Litigation against obstetricians is a major problem for our specialty. The most recent American College of Obstetricians and Gynecologists (ACOG) Professional Liability Survey indicates that 90.5% of the respondents had been sued at least once during their lifetime, and the average number of claims among those that had been sued was 2.7. Data from that survey further indicates that the paid indemnities in cases of “neurologically impaired” infants averaged $1,055,222 for the respondents. Aside from the emotional turmoil that accompanies the formal allegation of negligence these startling numbers have led to an enormous rise in the cost of malpractice insurance, increasing numbers of physicians giving up, or limiting the scope of their obstetric practices relatively early in their careers, and significant changes in the traditional patient-physician relationship. Despite a massive increase in the number of cesarean deliveries performed in the United States the incidence of cerebral palsy in infants delivered at term has remained stable over the past 4 decades, and numerous studies have indicated that the majority of cases of hypoxic ischemic encephalopathy, the precursor to long-term neurologic impairment caused by intrapartum hypoxia, are not related to identifiable events that occurred during labor and delivery. Furthermore, electronic fetal heart rate (FHR) monitoring, the only tool currently available for assessing fetal well-being during labor, is an extremely poor predictor of the eventual development of cerebral palsy. All of this, coupled with widely acknowledged weaknesses in the current tort review system have led many obstetricians to believe that they are powerless to win against a deck that is unfairly stacked against them.
However, some facts are important to consider. The first is that although it is unreasonable to expect that all patient outcomes will be optimal, it is clear that good outcomes will be associated with fewer damages, and therefore less grounds for suits, than those that are not as good. The second is that there definitely are poor outcomes associated with substandard medical care in general, and obstetric care in particular. The report “To Err Is Human” has famously stated that between 44,000 and 98,000 patients in the United States die each year as a result of iatrogenic medical injury. In a recent article, Abuhamad and Grobman cite 4 peer-reviewed publications, all of which showed that a substantial number of poor obstetric outcomes in the series reviewed were deemed to be preventable. Many of those adverse events were thought to be due to communication issues, poor teamwork, and other systems problems rather than individual human error, but the bottom line is that they were preventable. Similarly, Jonsson et al reported a series of 161 neonates >34 weeks with metabolic acidosis at birth of more than 28,000 consecutive deliveries in 2 Swedish university hospitals. When reviewers who were blinded to the outcome evaluated the final 2 hours of electronic fetal monitoring before delivery it was believed that 40-50% of the cases of acidosis and related neonatal morbidity could potentially have been prevented.
We all want our patients to have the best possible outcome for obvious reasons. Physicians, midwives, nurses, and other paramedical caregivers choose careers in obstetrics because they want to maximize the outcome for pregnant women and their offspring. Furthermore, this is the best possible way to eliminate, or at least dramatically reduce claims of obstetric malpractice. Yet our track record is far from perfect in this arena, and the studies mentioned above indicate that at least some of the adverse outcomes experienced by our patients come from suboptimal care.
The safety literature continuously stresses the fact that all human beings frequently make mistakes and the way to consistently achieve results of the highest quality is to create systems that either prevent those errors from occurring, or catch them before they produce irrevocable harm. We know that extraordinary safety records have been achieved by inherently dangerous industries such as commercial aviation and producers of nuclear power. We have also heard again and again that the practice of human medicine is infinitely more complicated than any of those industries that have demonstrated their ability to consistently avoid serious accidents while performing potentially life-threatening activities, but the actions taken by the American Society of Anesthesia have shown that similar results can be obtained if the members of a medical specialty commit themselves to doing so. The Anesthesia Patient Safety Foundation was founded in 1985, and following a national review of closed claims against anesthesiologists, that group made a series of recommendations that were subsequently adopted throughout the United States. Those actions resulted in a 10- to 20-fold decrease in mortality and catastrophic morbidity for healthy patients receiving routine anesthetics, and within 10 years the anesthesiologist’s liability payouts had decreased by a proportionate percentage.
Numerous strategies have been proposed and used in various combinations in attempts to improve safety for obstetric patients. These include adherence to principles of crew resource management and smoothly integrated teamwork, development of quality assurance review groups for departments and practices to track a variety of safe practice metrics and examine adverse events, utilization of simulation to practice individual skills, and team coordination for unanticipated emergencies, development of electronic medical records to provide usable databases for monitoring performance, reminders to order specific studies in given clinical situations and provision of “hard-stops” for erroneously prescribing unacceptable medications, credentialing for electronic FHR monitoring, and utilization of practice protocols and checklists.
In their superb review article, Abuhamad and Grobman synopsize reported data on the effectiveness of error reporting, crew resource management, and simulation training in Obstetrics. They point out that although virtually all of the cited studies of clinical outcomes showed positive results they were all longitudinal analyses, “which highlight the lack of randomized trials in the evaluation of safety initiatives.” They also state that although logic would suggest that utilization of the approaches mentioned above would “improve the professional liability climate through the occurrence of fewer adverse events and the related liability, it remains important to confirm this supposition with empiric data.” Unfortunately, very little information documenting this association has been published. Both Clark et al and Pettker et al have described the results of multifaceted safety programs introduced in their institutions that led to improved patient outcomes, and preliminary data suggest that their malpractice rates have been significantly reduced. In a study of almost 60 counties in California over a 10-year period, it was found that a decrease of 10 adverse events in a given year was associated with a concomitant decreased of 3.7 malpractice claims. These reports notwithstanding, hard data demonstrating the anticipated relationship between improved patient outcomes and diminished malpractice litigation remains very sparse.
One of the most contentious suggestions for improving safety on an obstetric unit is the introduction of clinical protocols and checklists. The objective of the former is to develop a uniform approach to common problems that will apply to the overwhelming majority of patients who develop that condition. The latter is a list of steps that must be completed whenever a particular process or procedure is initiated.
Protocols are guidelines and are not intended to be definitive statements of best management for all patients. In many cases, alternative approaches would be equally effective. The intent of the protocols, however, is to provide process standardization within a unit because it is reasonable to believe that this will lead to an improvement in clinical outcomes for the entire service. It should always be possible for a clinician to disregard one or more aspects of a protocol for an individual patient, but if this is done an explanation should be recorded in the chart explaining the reason for doing so in that particular case. This approach accomplishes two objectives. The first is that a uniform standard of care is established for the institution in which a clinician works for managing particular problems, and adherence to that approach eliminates the need to “reinvent the wheel” every time the problem arises. The second is that whenever a patient is treated “off protocol” the explanation for why that was done will reveal the thinking behind the actions that were taken. In the event of a lawsuit this may prove to be very useful in defense of those actions. Finally, all protocols should be reviewed and revised periodically to incorporate improvements derived from scientifically valid data.
Checklists are designed to assure that whenever a procedure is performed the steps necessary to maximize its safety are always performed. The classic example of this approach is the steps taken by cockpit crews before every flight, and its effectiveness in medical procedures has been demonstrated by the virtual elimination of infectious complications following central line placement. A preprocedure verification checklist is now required by the Joint Commission for all operative and other invasive procedures that expose patients to more than minimal risk. A recently published prospective, controlled study from the Netherlands tested the impact of a comprehensive, multidisciplinary safety checklist on the outcomes of general surgical patients in 6 hospitals with high standard of care profiles. When a baseline period of 3 months was compared with a similar period after implementation of the checklist, the total number of complications per 100 patients decreased from 27.3 to 16.7, the proportion of patients with one or more complications declined from 15.4% to 10.6%, and the inhospital mortality fell from 1.5% to 0.8%. During the same periods, these outcomes were unchanged in 5 comparable hospitals that had not used the checklists. In an accompanying editorial, Dr John Birkmeyer states that the results documented in this report and a similar large international study supported by the World Health Organization indicate that “checklists could avert tens of thousands of surgical deaths and hundreds of thousands of serious complications every year in the United States” and concludes that “checklists seem to have crossed the threshold from good idea to standard of care.”
Obstetric caregivers, however, often object to the notion of using both protocols and checklists for several different reasons. One is that clinicians like to function as thoughtful individuals and believe in their capacity to make sound clinical decisions when rendering care to their patients. The notion of practicing “cookbook” medicine may be abhorrent to people who take great pride in their diagnostic and therapeutic skills. This has led to the wry observation that trying to channel the activities of a group of clinicians is like attempting to herd cats. But would you like to be a passenger in planes that were being piloted by crews with a similar mindset?
Another important reason why many clinicians object to following protocols is because they believe that there is not enough evidence-based data to justify a specific proposed approach over one or more alternatives. These individuals may argue, for example, that they have given oxytocin the same way throughout their career and have never had a problem, so why should they alter that approach to adhere to a different regimen? They might also say that unless there was definitive evidence that a different regimen was proven to be superior to the one they have used for years there was no reason to abandon an approach they have been using successfully. There certainly is merit to that argument, and they might very well be correct in saying that there are multiple, equally safe ways to accomplish a clinical objective. However, the standardization of approaches to dealing with common problems has benefit to the service because it establishes a uniform methodology that reduces the potential for individual human errors to occur. In other words, refusal to adapt a protocol because it has not been irrefutably shown to be optimal interferes with the creation of a safety net for the entire service. Furthermore, if the components of the protocol fall within nationally accepted standards, adherence to that set of guidelines will guarantee that the individual has practiced within the standard of care at his/her institution, even if a different protocol is being followed at another hospital in the same community.
Finally, what about the argument that with the possible exception of checklists none of the above recommendations for improving safety on a service have definitively been shown to improve patient outcomes and reduce liability claims in a prospective randomized study? Although there is no argument that this type of study provides the gold standard for establishing scientific validity, it is not always necessary to obtain this type of unassailable proof of concept before adopting alterations in behavior that are highly likely to improve outcomes. To my knowledge, the principles underlying crew resource management have never been subjected to this type of analysis, but acceptance of that concept has become an integral part of “standard operating procedure” in the commercial aviation industry and is widely credited for much of the dramatic improvement in safety that has occurred since its inception. It has also been said that no one has, or is ever likely to perform a randomized prospective study to demonstrate the effectiveness of parachutes!
To paraphrase Albert Einstein, it is not reasonable to keep repeating the same behavior and expect to get different results. What we are currently doing as a profession is leading to too many preventable suboptimal outcomes, and we all need to get serious about instigating alterations in our institutional systems to change that. We should not be afraid to adopt well constructed, scientifically valid clinical protocols just because alternative approaches may ultimately be proven to be equally effective, or fail to support FHR credentialing for everyone who works on our labor and delivery floors because we assume that all of them already know how to read tracings and use the same standardized nomenclature. One size does not fit all, and each facility in the United States that offers obstetric care will have to adapt its own menu of approaches to improve safety within its doors. What is not acceptable is to keep doing what we are doing, hope that we will never have a potentially avoidable bad outcome and wait for tort reform to solve the problems imposed on us by an inherently unjust legal system.