Escalation Pattern Analysis (EPA): Recognizing Escalating Danger in Intimate Partner Violence (IPV)

In the aftermath of an intimate partner homicide (IPH), the central question for every professional is whether this tragedy could have been prevented.

In many cases, the circumstances leading up to these deaths suggest that the danger did not appear suddenly or without warning. Victim-survivors may report a mounting sense of fear. Neighbors may notice repeated disturbances. Health and social care professionals might observe subtle but concerning changes. Police may be called to the same address multiple times. Each of these encounters, on its own, provides only a partial view of the situation. When incidents are considered in isolation, they may seem troubling but unrelated. However, when we step back and examine the pattern over time, a trajectory of escalating abuse may become visible.

Recognizing this distinction is critical. Escalating danger in intimate partner violence (IPV) does not always present as a sudden spike in physical assaults. The frequency or severity of assaults may appear unchanged, even as surveillance increases, threats become more explicit, boundaries are repeatedly violated, protective measures are undermined, or interventions that once seemed effective lose their impact. The risk is not always concealed; sometimes, it is dispersed across multiple incidents, behaviors, agencies, and points in time.

Escalation Pattern Analysis (EPA) was developed specifically to address this challenge. As my principal conceptual contribution to the field of IPV, EPA offers a structured framework for examining how abusive behavior evolves over time and for determining whether the overall trajectory is becoming more dangerous. EPA focuses on three dynamic and interacting dimensions: Frequency, Intensity, and Inhibitory Decline.

EPA is not designed to predict homicide. Its primary purpose is to support earlier recognition by identifying significant changes in the direction, rate, momentum, and convergence of abusive behaviors. The goal is to support practitioners in recognizing when intervention or intensified safeguarding may be required, before they are left retrospectively searching for warning signs that, viewed together, may have revealed an escalating trajectory.

While conventional risk assessment may principally ask, “How dangerous is this today?”, EPA adds a crucial longitudinal question:

“Is this becoming more dangerous?”

The full academic framework is openly available:

Kaiser, R. (2026). Defining Escalation Pattern Analysis (EPA): Behavioral Trajectories and Early Lethality Indicators in Intimate Partner Violence (IPV). Zenodo. https://doi.org/10.5281/zenodo.17827775

What Is Escalation Pattern Analysis (EPA)?

EPA is a trajectory-based early-warning framework designed to support professionals in recognizing when IPV is developing along an increasingly dangerous trajectory. Rather than treating abusive behaviors as isolated incidents, EPA requires us to examine how these behaviors develop and interact over time, and whether the overall pattern is moving toward greater danger. By focusing on patterns of escalation and convergence, EPA supports earlier recognition of meaningful behavioral change and anticipatory, preventive decision-making when the future remains uncertain, and the stakes are high.

EPA does not predict the future. Instead, it serves as a translational framework that complements established risk assessment tools and supports structured professional judgment, early intervention, and coordinated safeguarding across agencies. The core principle is clear: the trajectory matters. Controlling behaviors, threats, surveillance, stalking, economic abuse, physical violence, and other forms of coercion rarely occur in isolation. Their significance can change as they become more frequent, more severe or intrusive, migrate into new forms, or persist despite interventions and restraints that previously appeared effective.

This distinction is critical for practitioners because escalation is not always obvious or dramatic. Physical assaults may appear unchanged, yet surveillance can intensify, threats may become more explicit, incidents may cluster closer together, or protective orders may be repeatedly breached. Sometimes, a reduction in one form of abuse is deceptive, simply signaling that the behavior has shifted into a different form, context, target, or channel. The assaults may appear static. The trajectory does not.

Research and frontline experience have established the importance of coercive control, stalking, separation-related risk, escalating severity, non-fatal strangulation, explicit threats, and other indicators associated with serious or lethal harm. EPA brings these concerns into a structured, longitudinal framework that reflects how violence can unfold within real relationships.

Drawing on coercive control theory, feminist criminology, femicide trajectory research, danger assessment scholarship, and trajectory-based analysis, EPA conceptualizes IPV as a patterned and cumulative process in which the direction, rate, sequencing, and interaction of abusive behaviors may be as important as any single incident.

EPA examines escalation through three dynamic and interacting dimensions that together provide a structured way to interpret how a behavioral trajectory changes over time.

Frequency refers to changes in how often abusive, controlling, or threatening behaviors occur. This includes persistence, recurrence, acceleration, and the shortening of intervals between incidents, a pattern known as temporal compression. Frequency therefore captures not simply whether abuse continues, but whether its recurrence and timing indicate meaningful directional change.
Intensity captures changes in the severity, intrusiveness, coercive impact, or potential lethality of abusive behavior. This may include more explicit threats, increasingly invasive stalking or surveillance, sexual violence, non-fatal strangulation, weapon-related behavior, and other qualitative shifts that may carry substantially greater significance even when the overall number of incidents remains relatively stable.
Inhibitory Decline describes the progressive weakening of psychological, social, situational, and institutional restraints that previously limited abusive behavior. Because internal psychological states cannot be directly observed, EPA interprets Inhibitory Decline principally through observable and documentable behavioral indicators: increasingly overt boundary violations, disregard for consequences, circumvention of restrictions, repeated breaches following intervention, continuation despite sanctions, and other evidence that previously effective restraints are losing regulatory force.2

These dimensions do not operate independently, and serious concern does not require deterioration across all three. Particular attention should be given to convergence, where concerning changes occur simultaneously across multiple dimensions. Convergence does not predict homicide or assign a categorical risk level. It does, however, provide a strong reason to reconsider whether existing safeguarding measures remain proportionate to the trajectory now unfolding.

In practical terms, whereas conventional risk assessment may principally ask, “How dangerous is this today?”, EPA adds a crucial longitudinal question: “Is this becoming more dangerous?”

A Typical Escalation Trajectory

January: He shouts, slams doors, and punches the wall. He apologizes afterward and promises it will never happen again. She wants to believe him.
April: The behavior becomes more frequent. He grabs her by the throat during an argument. His remorse appears diminished. He begins monitoring her phone and questioning her movements. Controlling behavior is becoming more persistent and intrusive.
May, first week: The trajectory intensifies. He strangles her until she loses consciousness. The marks may fade quickly, but the significance of what has happened does not. He increasingly isolates her from friends and controls her access to money. The abuse is becoming more severe, intrusive, and potentially lethal.
May, second week: Evidence of Inhibitory Decline becomes increasingly apparent. Threats become explicit. He brandishes a weapon, threatens the children or pets, justifies the abuse as deserved, and demonstrates increasing disregard for boundaries and consequences.

Viewed separately, these incidents may be recorded as different manifestations of abuse occurring at different points in time. Viewed longitudinally, however, a much more concerning trajectory becomes visible: incidents are occurring closer together, behaviors are becoming more severe and intrusive, coercive control is broadening, and previously apparent restraints are weakening.

This is convergence: concerning change occurring simultaneously across multiple dimensions. It does not mean that homicide will occur, nor does EPA assign a categorical level of risk. It does mean that the trajectory has changed in ways that demand serious scrutiny of whether existing safeguarding measures remain adequate.

The assaults may appear as individual events. The trajectory tells us something more: the danger is changing.

EPA Operational Trial Framework

To support responsible development and evaluation, EPA is being examined through an EPA Operational Trial Framework. This practitioner-facing initiative invites professionals in law enforcement, health and social care, housing, and advocacy services to consider how a non-scoring, trajectory-based framework may improve recognition of escalation patterns alongside existing safeguarding processes. The trial is exploratory and evaluative, focusing on usability, interpretive value, and whether EPA supports earlier or clearer identification of escalation concerns that incident-focused approaches may not capture.

Why Pattern Recognition Matters

Recognizing patterns is fundamental to survival, especially in the context of IPV. While individual incidents can carry profound significance, danger may also become clearer as behaviors shift, intensify, or interact over time. Monitoring, accusations, restrictions, threats, stalking, and violence can take on different meaning when we examine how they evolve across the course of a relationship. For practitioners, understanding this progression can be critical to recognizing an increasingly dangerous trajectory and determining whether existing safeguarding measures remain adequate.

EPA brings this longitudinal perspective into practice. By focusing on behavioral trajectories rather than isolated incidents alone, EPA supports practitioners in identifying meaningful shifts that may indicate increasing danger, even in the absence of severe visible injury. This approach moves beyond exclusively static or incident-focused assessment toward a dynamic, time-sensitive understanding of escalation. For those working on the front line, the critical issue is not simply what has happened, but whether the pattern itself is changing.

Pattern and trajectory analysis have long been used in intelligence and behavioral threat-assessment contexts. EPA adapts these principles to IPV, integrating trajectory-based reasoning with structured professional judgment and survivor-centered interpretation. The key question for practitioners is not only what has already occurred, but what is changing now, how quickly it is changing, and what those changes may mean for the direction of danger.

The Potential of EPA

When applied with rigor and supported by ongoing empirical research, EPA has the potential to significantly strengthen how we respond to IPV. Its value lies not only in its conceptual foundation, but in its ability to translate complex and evolving patterns of abuse into information that practitioners and institutions can interpret, communicate, and act upon.

For victim-survivors, EPA provides a language for describing something they may already understand at an instinctive and experiential level: that the danger is changing. Their fear may not arise from one incident alone, but from accumulated changes in the perpetrator’s behavior, the increasing surveillance, the shortening intervals, the more explicit threats, the boundaries that are no longer respected, or the consequences that no longer appear to restrain him. EPA provides a framework through which these changes can be recognized as part of an evolving trajectory rather than dismissed as disconnected events.
For practitioners, EPA provides a structured approach to identifying and interpreting concerning changes in an abusive person’s behavior over time. By looking beyond individual incidents to direction, rate of change, momentum, migration, convergence, and declining restraint, professionals can develop a richer understanding of whether the trajectory itself is becoming more dangerous and whether the existing response remains sufficient.
For agencies and institutions, EPA offers a shared framework for integrating information that may otherwise remain fragmented across disciplines and agencies. Police may see one incident, healthcare professionals another, advocates another, and the victim-survivor may be the only person who sees the entire pattern. A common trajectory-based language can help professionals connect those fragments, communicate meaningful change, and critically reconsider whether safeguarding measures remain proportionate to the danger now developing.
For prevention, EPA represents a fundamental shift in emphasis: from asking only what has happened to asking what is changing. Prevention cannot begin only when severe harm has already occurred. It requires professionals to recognize meaningful deterioration while opportunities to intervene still exist. EPA is designed to support that earlier recognition and the anticipatory decision-making that may follow from it.

At its core, EPA is driven by a preventive ambition. It asks professionals to look beyond the incident in front of them and understand the trajectory developing behind it. The objective is not to predict with precision what an abusive person will do next. It is to recognize when the pattern is changing in ways that demand greater attention, stronger safeguarding, or a different response, before professionals are left looking backwards at warning signs that, viewed together, were telling a much more serious story.

The trajectory matters. Understanding it may save lives.

The full academic EPA framework is openly available:

Kaiser, R. (2026). Defining Escalation Pattern Analysis (EPA): Behavioral Trajectories and Early Lethality Indicators in Intimate Partner Violence (IPV). Zenodo. https://doi.org/10.5281/zenodo.18207745