New 26-Item Questionnaire Aims to Standardize Violence Screening in Psychiatric Care

People living with severe mental illness are disproportionately affected by interpersonal violence, both as victims and as perpetrators, yet psychiatric services rarely measure this systematically. A research team led by Paula Kittelmann and Georg Schomerus at the University of Leipzig, working with colleagues from Ulm University and the anti-stigma organization Vorurteilsfrei e.V., has now published exploratory findings on a purpose-built screening instrument, the Interpersonal Violence Questionnaire, or IVIQ. The study, published in BMC Psychiatry, tested the tool in 479 adult psychiatric inpatients and outpatients across six German hospitals and offers one of the first structured attempts to capture violence involvement in this population with a standardized self-report checklist.

The rationale behind the IVIQ reflects a long-standing gap in psychiatric care. International studies have repeatedly documented that people with severe mental illness experience violent victimization at rates several times higher than the general population, while a smaller but clinically meaningful proportion also report perpetrating violence. Despite this dual burden, routine psychiatric intake examinations rarely include structured questions about violence, and research instruments designed specifically for this population have been scarce. Existing tools often focus on a single direction of violence, either victimization or perpetration, or are too lengthy and burdensome for everyday clinical use. The IVIQ was developed to address both shortcomings by covering physical violence, sexual violence, and threats in a single 26-item checklist, asking separately about victimization and perpetration and about two time frames, the past twelve months and the entire lifetime.

The choice of a self-report checklist format carries both advantages and trade-offs. Structured checklists can be administered quickly, require little interviewer training, and make it easier to compare results across patients, sites, and time points. At the same time, self-reported violence is inherently sensitive material. Respondents may underreport experiences that carry shame, fear of consequences, or painful memories, and they may also interpret items differently depending on their cultural background and personal understanding of what counts as violence. The exploratory design of the present study explicitly acknowledged these limitations, positioning the IVIQ as a starting point for instrument development rather than a finished diagnostic standard.

The sample of 479 adult psychiatric inpatients and outpatients recruited at six German hospitals represents a broad cross-section of routine mental health care rather than a narrowly defined diagnostic group. This breadth matters for a screening instrument intended for routine assessment, because it provides preliminary evidence on how the tool performs across the heterogeneous mix of conditions encountered in everyday psychiatric practice. The study was prospectively registered in the German Clinical Trials Register and the WHO International Clinical Trials Registry Platform under DRKS00032041, a step that supports transparency in how the investigation was planned and conducted. Ethical approval was obtained from the ethics committees at Leipzig University, the State Chamber of Physicians of Saxony, and Ulm University, and all participants provided written informed consent before taking part.

Among the most notable descriptive findings was the pattern of prevalence across the individual items. Lifetime sexual victimization emerged as one of the most frequently reported experiences, alongside threats of violence and being pushed within the past twelve months. This pattern is consistent with a wider body of literature showing that sexual victimization is strikingly common among people with severe mental illness, often beginning early in life and frequently recurring. It also highlights that less dramatic forms of physical aggression, such as pushing, and non-physical forms, such as threats, are highly prevalent in this population and deserve attention alongside severe physical assault.

The strong floor effects observed in the IVIQ sum scores deserve careful interpretation. A floor effect means that a large proportion of respondents scored at or near the lowest possible value, indicating no or minimal reported violence involvement. In a general population sample this might suggest limited usefulness, but in a clinical screening context it is not necessarily a flaw. Most psychiatric patients, like most people in the general population, have not experienced every form of violence asked about, so low average scores are expected. The practical implication is that sum scores alone may obscure important information: a patient scoring zero overall may still have endorsed a single highly consequential item, such as recent sexual victimization. This is why the authors suggest that different scoring approaches, including both cumulative burden scores and item-level or domain-specific evaluation, may be needed to capture the full range of violence involvement.

The frequent co-occurrence of victimization and perpetration reports aligns with a well-established finding in violence research. Across many studies and populations, individuals who report being victimized are more likely to also report perpetrating violence, and vice versa. This overlap has been described in terms of bidirectional violence and reflects complex dynamics within relationships, households, and social networks. For clinical practice, the message is that asking only about victimization or only about perpetration provides an incomplete picture. A patient who discloses aggressive behavior toward others may simultaneously be experiencing violence from others, and a patient presenting primarily as a victim may also have engaged in aggression. The IVIQ’s dual coverage makes such overlap visible in a way that single-direction instruments cannot.

The low to moderate inter-item correlations observed in the descriptive analyses suggest that the individual IVIQ items capture partly distinct phenomena rather than a single uniform construct of violence. This is consistent with the conceptual rationale of the instrument, which treats physical violence, sexual violence, and threats as different forms of interpersonal harm that may have different antecedents, contexts, and consequences. For researchers, low inter-item associations mean that aggregate scores should be interpreted cautiously and that item-level or domain-level analysis may be more informative. For clinicians, it underscores that a patient’s violence history is rarely uniform and that specific forms of involvement may require specific follow-up questions and support.

Temporal stability, assessed by readministering the instrument, varied considerably across items. High-prevalence behaviors showed more consistent reporting over time, whereas low-prevalence behaviors showed lower stability. This pattern is common in self-report instruments covering rare events, because small numbers of endorsements can shift substantially between assessments by chance alone. It may also reflect genuine changes in exposure over time, recall difficulties, or differences in how respondents interpret items on repeated occasions. The finding suggests that the IVIQ may be more reliable for tracking frequently occurring forms of violence than for rare but severe events, and that repeated administration should be interpreted with this variability in mind.

One of the most consequential findings concerns the relationship between social desirability and reporting. Using the Kurzskala Soziale Erwünschtheit-Gamma, a German scale for assessing the social desirability trait, the researchers found that higher social desirability predicted lower reporting of perpetration within the past twelve months. In other words, respondents who were more inclined to present themselves in a socially favorable light were less likely to disclose having perpetrated violence. This has direct implications for how violence screening should be implemented in clinical settings. Perpetration is likely to be systematically underreported, particularly in face-to-face contexts or when patients fear legal, social, or therapeutic consequences. Sensitive framing, clear assurances of confidentiality, and possibly self-administered formats may help reduce this bias. The finding also reinforces a broader principle in violence research: self-report data on sensitive behaviors should always be interpreted as a lower bound rather than an exact count.

The authors’ call for future versions to broaden the assessment of psychological violence points to an important conceptual limitation of the current instrument. Psychological violence, including coercive control, humiliation, intimidation, and systematic isolation, is increasingly recognized as a major component of interpersonal violence with substantial mental health consequences. Yet it is notoriously difficult to measure with brief checklist items, because it often involves patterns of behavior rather than discrete incidents. Expanding the IVIQ in this direction would align it with contemporary definitions of intimate partner violence and abuse that give central weight to non-physical forms of harm.

The involvement of Vorurteilsfrei e.V., a Leipzig-based anti-stigma organization, in the research team reflects a growing recognition that violence research involving psychiatric patients must be conducted with attention to stigma and power dynamics. People with mental illness frequently report being perceived as dangerous, a stereotype that can discourage honest disclosure of either victimization or perpetration. Trauma-informed assessment practices, which the authors explicitly emphasize, involve asking about violence in a way that prioritizes safety, choice, and control for the respondent, avoids re-traumatization, and provides pathways to support when disclosure occurs. Implementing a screening tool without such safeguards risks harming the very patients it is meant to help.

For clinical practice, the exploratory findings suggest several practical takeaways. A brief structured checklist can feasibly be administered in psychiatric settings and can surface violence experiences that would otherwise remain unaddressed. Prevalence patterns in the sample confirm that threats, pushing, and sexual victimization are common enough to warrant routine inquiry. The co-occurrence of victimization and perpetration means that both directions should be assessed. Floor effects and low inter-item correlations caution against relying on a single total score. Variability in temporal stability suggests caution when interpreting change over time, particularly for rare events. And the influence of social desirability, especially on perpetration reporting, argues for sensitive implementation and for treating self-reported perpetration as a probable underestimate.

For research, the study provides a foundation on which further instrument development can build. Future work will need to refine scoring approaches, expand coverage of psychological violence, examine the tool’s performance in more narrowly defined diagnostic groups, and evaluate whether structured violence screening leads to improved clinical outcomes such as earlier identification of victimization, better-targeted safety planning, and reduced revictimization. Validation against interview-based assessments and informant reports would also help clarify the accuracy of the self-report format. As it stands, the IVIQ represents an early but meaningful step toward making violence involvement a routine, measurable dimension of psychiatric assessment rather than an invisible burden carried silently by many patients.

Subject of Research: Validation of a self-report questionnaire for assessing violence victimization and perpetration in psychiatric patients

Article Title: Assessing interpersonal violence involvement in psychiatric patients: exploratory findings on the Interpersonal Violence Questionnaire (IVIQ)

Article References: Kittelmann, P., Helmert, C., Hojka, N., Karp, J., Pouwels, M., Scheuermann, A., Steglich, J., Krumm, S., & Schomerus, G. (2026). Assessing interpersonal violence involvement in psychiatric patients: exploratory findings on the Interpersonal Violence Questionnaire (IVIQ). BMC Psychiatry, 26(1), Article 715. https://doi.org/10.1186/s12888-026-08597-7

Image Credits: AI Generated

DOI: 10.1186/s12888-026-08597-7

Keywords: interpersonal violence, psychiatric patients, IVIQ, victimization, perpetration, self-report questionnaire, severe mental illness, trauma-informed care, social desirability, psychiatric assessment, screening tools, BMC Psychiatry