Evolving tools for narrative analysis present an opportunity to identify common structure in stories that are socially important to tell, such as stories of survival from domestic abuse. A greater structural understanding of such stories could lead to stronger protections against de-anonymization, as well as future tools to help survivors navigate the complex trade-offs inherent in trying to tell their stories safely. In this work we explore narrative patterns within a small set of domestic violence stories, identifying many similarities. We then propose a method to assess the safety of sharing a story based on a distance feature vector.