KDKirsten A. Dickins

Kirsten A. Dickins

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article

Previous Organizations / Institutes

The organizations / institutes from the scholar's other articles in the past ten years


Kirsten A. Dickins has a total of 49 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Trauma Exposure Prevalence and PTSD Symptoms in Women Who Are Incarcerated: Implications for Nurse Practitioner–Led Clinical Support and Intervention

    by Kirsten Dickins, Mary Clare Houlihan, Amanda LaMonica-Weier, Samuel Chen, Horace Nowell, Niranjan Karnik, Philip Held

  • Open Access

    Automated Safety Testing and Reporting Application for Conversational Safety Monitoring of Generative AI Tools for Mental Health: Development and Validation Study

    by Daniel Szoke, Ilana Hutzler, Jerry Liu, Samantha Addante, Zuhaib Akhtar, Dale L Smith, Kirsten Dickins, Charles Small, Sarah Pridgen, Philip Held

    Background: Artificial intelligence (AI)–based conversational tools are rapidly expanding within mental health care as a means of increasing access and scalability. At the same time, these systems introduce distinct safety risks arising from both user disclosures (eg, self-harm ideation) and inappropriate or inadequate AI responses. Objective: This study aimed to develop and evaluate the Automated Safety Testing and Reporting Application (ASTRA), an external system intended to identify clinically relevant risk behaviors across entire AI-mediated mental health conversations. Methods: ASTRA was tested on a dataset of 100 synthetic therapeutic conversations written by licensed clinicians to reflect risk behaviors and harmful responses between users and AI tools. Conversations varied in length and included both subtle and overt risk behavior examples across 8 predefined categories. Human coder consensus ratings served as the reference standard. ASTRA’s classifications were evaluated across 2 prompt iterations using standard diagnostic performance metrics and agreement statistics. Results: ASTRA demonstrated consistently high concordance with expert human ratings across all categories. Accuracy exceeded 0.90 for all risk behavior categories examined, with specificity uniformly high and sensitivity varying by category (range 0.55-1.00). Agreement beyond chance was substantial to almost perfect between ASTRA and human raters (κ=0.65-1.00). Detection of user self-harm indicators was particularly accurate, even in conversations where risk was expressed subtly. Conclusions: In this initial validation study, ASTRA reliably identified multiple forms of mental health–related risk behaviors at the conversation level. These findings support the feasibility of independent safety monitoring systems as a complement to AI tools used in mental health contexts and underscore the need for further evaluation using larger and real-world datasets.


  • Open Access

    Narrative exposure therapy for post-traumatic stress symptoms in women who are incarcerated: an open pilot study

    by Kirsten A. Dickins, Mary Clare Houlihan, Sara Neely, Dale L. Smith, Horace Nowell, Niranjan Karnik, Philip Held

    1. The population of women who are incarcerated (WI) in the United States (US) is increasing at a significantly greater rate than the male incarcerated population and women in other economically de...


  • Open Access

    Improving post-traumatic stress symptoms in homeless-experienced women using narrative exposure therapy: a single-arm, open pilot study

    by Kirsten A. Dickins, Mary Clare Houlihan, Sara Neely, Dale L. Smith, Horace Nowell, Niranjan Karnik, Philip Held

    Background: Homeless-experienced women face disproportionately high rates of trauma exposure and posttraumatic stress disorder (PTSD) compared to the general population. Despite this, there is little research into the effectiveness of accessible, trauma-focused interventions for this population.Objective: This pilot study sought to determine the feasibility and preliminary effects of Narrative Exposure Therapy (NET) on PTSD and co-occurring mental health symptoms in homeless-experienced women.Methods: We conducted a single-arm, open pilot study of NET with 53 women in three urban women's homelessness organisations. Feasibility was assessed via intervention and study protocol completion rates. We used linear mixed effects models and effect size metrics to analyse pre - to post-NET changes and strength in effects in primary (PTSD, PCL-5) and secondary outcomes variables (depressive symptoms, PHQ-9; anxiety, GAD-7; somatisation, SSS-8) in all participants who completed NET.Results: Fifty-three women began the NET protocol; 43 completed at least one outcome measurement, and 36 of these women completed the full protocol (81% treatment completion rate, 68% study protocol completion rate). We observed significant symptom reductions in all measured symptom domains one-week post-NET. Participants showed large changes in PTSD symptoms (PCL-5 change: 20.8 points, p < .001, d = 0.9) and medium changes in depressive symptoms (PHQ-9 change: 3.1 points, d = 0.57), anxiety (GAD-7 change: 3.8 points, d = 0.7), and somatisation (SSS-8 change: 3.7 points, d = 0.49).Conclusions: NET may be a feasible and clinically pragmatic intervention for reducing PTSD and co-occurring mental health symptoms in HEW. A fully powered, randomised trial to evaluate NET in this population is necessary to definitively determine intervention effects.

  • Open Access

    Automated Safety Testing and Reporting Application for Conversational Safety Monitoring of Generative AI Tools for Mental Health: Development and Validation Study (Preprint)

    by Daniel Szoke, Ilana Hutzler, Jerry Liu, Samantha Addante, Zuhaib Akhtar, Dale L Smith, Kirsten Dickins, Charles Small, Sarah Pridgen, Philip Held

    BACKGROUND Artificial intelligence (AI)–based conversational tools are rapidly expanding within mental health care as a means of increasing access and scalability. At the same time, these systems introduce distinct safety risks arising from both user disclosures (eg, self-harm ideation) and inappropriate or inadequate AI responses. OBJECTIVE This study aimed to develop and evaluate the Automated Safety Testing and Reporting Application (ASTRA), an external system intended to identify clinically relevant risk behaviors across entire AI-mediated mental health conversations. METHODS ASTRA was tested on a dataset of 100 synthetic therapeutic conversations written by licensed clinicians to reflect risk behaviors and harmful responses between users and AI tools. Conversations varied in length and included both subtle and overt risk behavior examples across 8 predefined categories. Human coder consensus ratings served as the reference standard. ASTRA’s classifications were evaluated across 2 prompt iterations using standard diagnostic performance metrics and agreement statistics. RESULTS ASTRA demonstrated consistently high concordance with expert human ratings across all categories. Accuracy exceeded 0.90 for all risk behavior categories examined, with specificity uniformly high and sensitivity varying by category (range 0.55-1.00). Agreement beyond chance was substantial to almost perfect between ASTRA and human raters (κ=0.65-1.00). Detection of user self-harm indicators was particularly accurate, even in conversations where risk was expressed subtly. CONCLUSIONS In this initial validation study, ASTRA reliably identified multiple forms of mental health–related risk behaviors at the conversation level. These findings support the feasibility of independent safety monitoring systems as a complement to AI tools used in mental health contexts and underscore the need for further evaluation using larger and real-world datasets.

  • Open Access

    Mortality by Age, Gender, and Race and Ethnicity in People Experiencing Homelessness in Boston, Massachusetts

    by Danielle R. Fine, Kirsten A. Dickins, Logan D. Adams, Nora K. Horick, Natalia Critchley, Katherine Hart, Jessie M. Gaeta, Elizabeth Lewis, Sara E. Looby, Travis P. Baggett

    Homelessness in the US has been growing over the past several years. Between 2019 and 2020, the number of people experiencing homelessness (PEH) increased by 2% with an estimated 580 445 individuals experiencing homelessness on a single night in 2020.1 The number of PEH in the US has continued to increase during the COVID-19 pandemic2 and is projected to continue increasing amid a national eviction and housing crisis.3


  • Publishing Your Literature Review

    by Susan W. Buchholz, Kirsten A. Dickins

    After completing the literature synthesis—arguably the most intellectually challenging component of the review—one is now ready to prepare to share the knowledge that one has generated. Preparing for dissemination can be arduous but is nonetheless essential to sharing our findings and making an important contribution to advancing the science. This chapter provides guidance for navigating through creating our first (and subsequent) drafts of our literature review and synthesis manuscript. It also provides important information regarding submitting our manuscript to a journal or other venue for publication. There are numerous publication standards available to provide guidance on ensuring that your manuscript exemplifies the rigor needed to excel during the review process for publication. The chapter discusses how nurses and other healthcare and health policy professionals can compose a literature review and synthesis manuscript for publication, including tips for dissemination.

  • Synthesizing the Literature

    by Susan W. Buchholz, Kirsten A. Dickins

    Throughout the synthesis process, the data have been extracted from each included study will be uniquely combined through the ascertainment of patterns, directions, and themes. Beginning by revisiting the review protocol will allow us to re-center on the original synthesis plan. Then, it will be useful to reflect on the separate findings of each of the included data sources. One will then create a narrative synthesis based on your reflections of the findings—the patterns, directions, and themes that you have identified. Depending on the type of review, one will complete a quantitative and/or qualitative data synthesis. The quantitative data synthesis uses effect sizes of individual studies to develop the meta-analysis. Different methods are used to synthesize qualitative data. Optimally, the synthesis that you generate will bring to light novel interpretations and new ideas that guide clinical practice, research, and healthcare policy.