Post-traumatic stress disorder (PTSD) as a distinct psychiatric syndrome originated in the work of psychiatrists and therapists engaged in supporting veterans of war. The term ‘shell-shock’ was coined by Myers, a British military psychiatrist, in 1915 to describe the emotional disturbance of combat personnel from World War I. He also noted the close similarity between persistent distress, or ‘war neurosis’, and ‘hysteria’ which was at that time a well-accepted medical phenomenon.

Following World War II, a major revision of the descriptions of mental disorders occurred. Medical practitioners affiliated with the armed forces developed new diagnostic classifications, and mental disorders were included in the World Health Organization’s International Statistical Classification of Diseases, Injuries and Causes of Death (ICD-6) (World Health Organization 1992) for the first time. Among these disorders, however, post-traumatic stress syndromes were still understood to be acute situational maladjustments (Brett 1996; Jones and Wessely 2007).

In 1980 the third edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM III) included post-traumatic stress disorder as a diagnostic category for the first time (American Psychiatric Association 1980). Modern research into trauma was precipitated by the needs of the large population of people who experienced psychological damage while in military service during the Vietnam War. However, since its first description in psychiatric textbooks the definition of PTSD has been revised repeatedly, making it one of the more controversial diagnoses in modern psychiatry (Rosen et al. 2008). Debates about the definition of traumatic events have been criticized either for trivialising the experience of those exposed to extreme and shocking events (McHugh and Treisman 2007) or for pathologising normal reactions to overwhelming experiences (Rosen et al. 2008). Debate continues, but there seems to be broad agreement that the unremitting distress that some people experience following traumatic exposure can be associated with high levels of functional disability and considerable economic and social costs.

The DSM III described a traumatic stressor as an ‘extreme event that was outside the range of usual human experience’. This early conceptualisation was based on the belief that traumatic events were rare; however increasing use of the new diagnostic category revealed that trauma responses were not uncommon and arose in response to events that were not unusual. Key to the traumatic nature of trigger events was that they overwhelmed the ordinary adaptive and coping mechanisms used by people in everyday life (Herman 1992). The subsequent fully revised edition of the diagnostic manual, the DSM IV (American Psychiatric Association 1994), and its successor, DSM IV-TR (American Psychiatric Association 2000), reflected this increased awareness and provided a different system for categorising an event as traumatic. It focused on the characteristics of the event and on the individual’s reaction to the event. However, subsequent research demonstrated that the subjective appraisal of the event as horrifying or leading to helplessness reduced the specificity of the diagnosis. Fewer than one third of events identified as traumatic in subjective appraisals met the DSM IV definition of a traumatic event. This led to overdiagnosis (Boals and Hathaway 2010), and in DSM 5 (American Psychiatric Association 2013), this criterion has been removed (Friedman et al. 2011). The DSM IV, DSM IV-TR and DSM 5 provide extensive lists of the types of events that may lead to the development of post-traumatic stress disorder. In its definition of a traumatic stressor, the DSM IV-TR is careful to state that its lists include possible traumatic events, but that trauma is not limited to these experiences and that diverse experiences can precipitate PTSD. In DSM 5 medical events that qualify as potentially traumatic are sudden and catastrophic events like becoming conscious during surgery. Childbirth is not listed as a potential traumatic event in either DSM IV-TR or DSM 5.

Childbirth is within normal human experience, but individual birth experiences are highly variable. Women can perceive risks to life or of serious injury when giving birth, or there can be actual injury to the woman or her baby. These are often accompanied by intense fear, helplessness and horror. As such, it is postulated that the experience of traumatic childbirth is potentially to meet recent DSM IV and current DSM 5 diagnostic criteria for a stressor that could lead to post-traumatic stress disorder.

Often credited with being the first to publish their observations of a syndrome of postnatal post-traumatic stress disorder (PTSD), Bydlowski and Raoul-Duval (1978) described ‘la nervose traumatique post-obstetricale’ (post-obstetric traumatic neurosis) among women who had experienced long and difficult labours, instrumentally assisted births or the birth of a dead or injured infant. Women experiencing this syndrome were identified, during the last trimester of a subsequent pregnancy, with symptoms typical of post-traumatic stress. Following this identification, greater research attention was focused on the potential for women to develop stress reactions, such as PTSD, following childbirth.

There were early case descriptions of trauma reactions among women who had experienced caesarean surgery with inadequate anaesthesia (Ballard et al. 1995; Fones 1996) or prolonged labours with severe pain or insufficient support (Ballard et al. 1995). Descriptions also emerged of births which were considered to be normal by maternity staff but where the woman’s subjective experience was characterized by intense feelings of lack of agency (Allen 1998). Moleman et al. (1992) reported a small case series of women who experienced recurrent intrusive thoughts and images, numbing, dissociation and other trauma-related symptoms following complicated births and prior experiences of spontaneous abortion or infertility. Subsequent consequences included nightmares, intrusive memories, difficulty bonding with the baby, fear of sexual intimacy and avoidance of future childbearing (Ballard et al. 1995; Fones 1996; Goldbeck-Wood 1996). Requests for elective caesarean section during a subsequent pregnancy were considered to be a form of avoidance following traumatising birth experiences (Ryding 1991, 1993). Subsequent research examining childbirth stress has followed the DSM IV diagnostic revisions in 1994 and has conceptualised such reactions as post-traumatic stress. Over the past 15 years, a growing body of research has sought to establish the nature of childbirth-related PTSD and attempted to establish its prevalence. This literature has found that symptoms of hyperarousal, re-experiencing and avoidance of reminders of the birth are prevalent among new mothers and that up to one third of women experience at least some symptoms of PTSD following childbirth (Olde et al. 2006).

Post-traumatic stress disorder (PTSD) is a diagnosable clinical condition which requires all eight DSM 5 diagnostic criteria (American Psychiatric Association 2013) to be met. In order to diagnose PTSD following any event, it must be established that the event itself meets criteria for being traumatic (Criterion A). Applying this to childbirth, a woman’s birth experience must have involved an actual or threatened injury to herself or her baby. There then have to be symptoms of intrusive or distressing memories (Criterion B), persistent avoidance of memories or external reminders of the events (Criterion C), negative alterations in cognitions or mood beginning or worsening after the event (Criterion D), marked changes in arousal or emotional reactivity after the event (Criterion E) that have persisted for more than 1 month (Criterion F) and they must have led to disrupted functioning in important domains of the woman’s everyday life (Criterion G), and finally, not be attributable to substance use or medication (Criterion H). In order to determine prevalence accurately, investigations need to have used standardized psychometric instruments that assess all criteria, but few studies have done so. Most were published prior to the release of DSM 5 and so used criteria from earlier editions.

Few studies have used diagnostic interviews to assess Criterion A (Verreault et al. 2012). Most used self-report questionnaires, in particular the Traumatic Event Scale (Wijma et al. 1997; Soderquist et al. 2002), and asserted that Criterion A had been met if a woman endorsed items stating that ‘the childbirth was a trying experience’ and ‘during the childbirth I felt anxious/helpless/terrified’. There is potential therefore for a woman who experienced anxiety and described her birth as ‘trying’ to be identified as having met Criterion A, while a woman who perceived her baby’s life to be at risk may not be detected. Symptoms must have been present for at least 1 month, although many studies assume this to be the case when assessments are carried out at more than 4 weeks postpartum; the duration of symptoms was rarely asked about directly. Few studies examined functional impact of the symptoms, and without this a diagnosis cannot be made. Further, few studies screened for PTSD symptoms during pregnancy and therefore cannot claim with certainty that postnatal symptoms are new and a consequence of birth, rather than being a continuation of symptoms that predated birth.