The concept, theory, and attributes of high reliability organizations that operate highly complex and hazardous technologic systems, essentially without mistakes, was introduced to clinical obstetrics practice approximately 12 years ago. The original description was based on an empiric observation that differentiated perinatal units that produce more or less harm to patients, with professional liability claims as a proxy for perinatal injury. In the original description, attributes of low-risk, harm-free, highly reliable perinatal units were identified: (1) safety is declared as the hallmark of organizational culture and is understood to be the responsibility and duty of every team member. All clinicians are considered competent with an obligation to speak up if the question of safety arises. (2) Patient safety is considered a team rather than individual property. Team interaction is noted to be collegial rather than hierarchic. (3) Respectful communication is highly valued and rewarded. Extensive, transparent communication is used to orient, plan, update, and adjust to the unexpected. Routine debriefing is practiced for unusual or unexpected events. (4) Emergencies are rehearsed and unexpected events are anticipated. (5) Paradoxically, successful operations (ie, absence of maternal, fetal, and neonatal injury) are viewed as potentially dangerous.

As pointed out by Vaughn, all work groups and teams are susceptible to “normalization of deviance,” which is a degradation of professional, behavioral, and technical standards that increases the probability of a major accident or harm. Because large accidents or patient harm occur infrequently, clinical practice in the absence of professional standards and/or incorrect process (eg, work-a-rounds) may go undetected or deteriorate over long periods of time. For example, elective early term birth <39 completed weeks of pregnancy, a clear violation of professional guideline and evidence-based practice is often undertaken because in the experience of a single practitioner or obstetrics unit “we never had a problem.” Normalization of deviance (“slide to failure”) must be understood as unavoidable unless cumulative evidence and professional standards are actively used instead of individual experience, care processes are continuously monitored, and importantly, an absence of poor outcomes is not used as a indication that care is universally safe.

Since that time, the concept of “high reliability” has been refined and expanded and, in some instances, confused or misunderstood. Pronovost et al framed the question most directly: “Exactly what does reliability mean in health care today, and how do we know if we are reliable?” The purpose of this article is to (1) update and reinforce the basic definition/concept of high reliability as culture (“the way we work around here”), (2) draw an important distinction between high reliability as a qualitative behavioral descriptor as opposed to simply quantitative measurement, (3) describe organizational tactics and strategy for moving toward high reliability, and (4) illustrate increased obstetric safety and quality where elements of high reliability have been incorporated into the culture and operations of obstetrics units.