Written by: Fiona Sheahan
Edited by: Maggie Wang
Is AI the future for Assessing PTSD in Postpartum Women? This intriguing question was recently answered on April 11th of this year. Our understanding of this recent breakthrough in artificial intelligence can help us further explore the underrepresented mental impacts of pregnancy.
Post-traumatic stress disorder (PTSD) is a mental condition prompted by a distressing event. Individuals with PTSD may have previously experienced or witnessed the painful event. Symptoms of PTSD include flashbacks, nightmares, severe anxiety, and uncontrollable thoughts about the specific event. To diagnose PTSD, a patient’s doctor would perform a physical exam and a psychological evaluation using criteria from the Diagnostic and Statistical Manual of Mental Disorders (DSM-5) (see here). According to the DSM-5, exposure to the event along with at least one of the previously stated symptoms calls for a diagnosis of PTSD (see here).
According to Postpartum Support International, approximately 9% of postpartum women experience PTSD. This is specifically caused by either a real or perceived trauma during delivery or postpartum. Some examples of causes include an unplanned c-section, the use of a vacuum extractor to deliver a baby, a feeling of powerlessness, and a lack of support. Moreover, this condition has been historically associated with military combat or severe sexual assault. Symptoms of postpartum PTSD are similar to PTSD because it consists of thoughts of the past (such as re-experiencing childbirth), night terrors; avoidance of people, feelings, and places related to the event; feeling a sense of detachment, anxiety, and panic attacks (see here).
The latest study published by Massachusetts General Hospital found a version of OpenAI that offers “rich insights into maternal mental health following traumatic childbirth.” The researchers hoped to find a version of AI that would analyze patient narratives of childbirth and convert it for assessment of childbirth post-traumatic stress disorder (CB-PTSD) after a traumatic birth. This finding has the potential to be an efficient, inexpensive, and patient-friendly approach to detecting women at risk for CB-PTSD before the condition fully develops. From a collection of short narratives from 1,295 women who had recently given birth, researchers found that the OpenAI model of “Text-embeddings-ada-002” successfully converts these narratives into a numerical format. The team’s developed algorithm-trained machine further analyzed this format to identify postpartum PTSD. This cost-effective screening strategy allows the possibility of other mental health disorders to be assessed. Therefore, AI could subsequently improve patients’ results and transform health care in a positive way (see here). This progressive step in technology will raise awareness for underrepresented women of postpartum PTSD.
Postpartum PTSD is a neglected condition, especially in low-income countries such as Pakistan. After childbirth, women in these countries are expected to return to work. Cultural and social factors influence how women and their families regard PTSD in Pakistan. For example, studies show how Muslim women choose ‘faith-based’ healthcare services that mostly consist of traditional or spiritual healers. This causes women to distrust proper medical care during their pregnancy. Moreover, women who particularly live in rural areas and urban slums of Pakistan are treated as inferiors to their husbands. These women’s opinions are not welcomed in either personal or family matters. Consequently, this chain of neglect and lack of mental care access causes women to refrain from identifying their condition (see here). This calls for the need to improve societies in need of understanding the link between mental health and maternal care. For example, incorporating mental health services into covered insurance accompanied by referrals for mental health professionals would allow women to seek help. The dependable expectations of healthcare providers overlooking their responsibilities for a maternal patient are unrealistic if the necessary mental health resources are absent (see here). Addressing these challenges that women experience, especially in low-income countries, is crucial to overcoming the societal stigma and creating a safe environment.
The emergence of AI has strong potential to be an efficient and inexpensive solution for identifying postpartum PTSD. Perhaps, AI could eventually be accessible to vulnerable populations of low-income countries. Nonetheless, overcoming barriers, particularly in low-income countries, is crucial for women receiving the necessary and inclusive mental health support.
References:
Ahsan, A., Nadeem, A., Habib, A., Basaria, A. A. A., Tariq, R., & Raufi, N. (2023).
Post-traumatic stress disorder following childbirth: a neglected cause.
Frontiers in global women's health, 4, 1273519.
American Psychiatric Association. (2013). Diagnostic and statistical manual of mental disorders
Free Stock Photo. (n.d.). Pexels. https://www.pexels.com/
Postpartum Post-Traumatic Stress Disorder | Postpartum Support International (PSI). (2024,
January 12). Postpartum Support International (PSI).
Post-traumatic stress disorder (PTSD) - Symptoms and causes - Mayo Clinic. (2022, December
13). Mayo Clinic.
Massachusetts General Hospital. (2024, April 11). AI model can accurately
assess PTSD in postpartum women. ScienceDaily. Retrieved April 11, 2024 from
Omer, S., Zakar, R., Zakar, M.Z. et al. (2021) The influence of social and cultural practices on
maternal mortality: a qualitative study from South Punjab, Pakistan. Reprod
Health 18, 97. https://doi.org/10.1186/s12978-021-01151-6-01151-6
Comments