Copyright © 2002 R.K. Piper and Kevin M. Fasana
The Phase II Evaluation of the Coordinated Response to Domestic Violence in Omaha was conducted through funding provided by the Domestic Violence Coordinating Council of Greater Omaha (DVCC).
Many individuals contributed valuable input to the study. Special thanks go to Dr. Cassia C. Spohn of the Criminal Justice Department, University of Nebraska at Omaha who provided early insights into the study problem and approach.
Special thanks also go to the Executive Committee of the DVCC: Patricia Lenaghan, Chair; Donald Carey, Vice-Chair; Frank Goodroe, Treasurer; Mary Ann Borgeson, Secretary and Joan Skogstrom, Director.
We also thank the members of the Evaluation Committee of the DVCC: Anne Camp, Chair, Omaha Community Foundation; Mary Larsen and Nancy Livingston, YWCA; Chris Carlson, Catholic Charities; Joan Macdonald, Omaha/Douglas County Victim Assistance Unit; Teri Knight, Protection Order Office; Lisa Jackson, Douglas County Sheriff; Ron Broich and Maureen Rohlfing, Nebraska State Probation; Sue Michalski, DVCC; Shelly Stratman and Chantel Wellner, Douglas County Attorney's Office; Charlie Venditte, OPD Domestic Violence Squad; and Larry Lavelle, Douglas County Communications Department (911).
Finally, at ISED, Tom Martin reviewed initial drafts of this document and provided helpful comments and suggestions. Ms. Deveri Johnson provided valuable production support.
This report documents the findings from the Phase II evaluation of the coordinated response to domestic violence in Omaha and Douglas County. The evaluation consists of two studies: 1) a "high-risk" case outcome study and 2) a performance management and quality control study. These studies were designed to meet the following research objectives:
To test the transferability of key domestic violence research and mortality review findings from other jurisdictions in the U.S. to the Omaha/ Douglas County area;
To determine to what degree indicators or predictors of "high-risk" (serious-injury or lethal) domestic violence cases could be identified locally and potentially used to reduce the incidence of future domestic injury or death;
To determine whether a case review process could be developed which could be used in the future by the DVCC and the agencies of the coordinated response for on-going quality control and performance management;
To document existing data sources, to explore whether the agencies of the coordinated response might be able to employ newer technology to capture and use such information more effectively.
The major findings of the Phase II studies are as follows:
An extensive case review process, involving all agencies of the coordinated response, was developed, implemented and successfully tested on fifteen (15) domestic homicide and serious-injury cases.
Key findings about indicators/predictors from domestic violence mortality reviews and research conducted in other jurisdictions in the U.S. are transferable to Omaha/Douglas County.
Indicators or predictors (risk factors) of "high-risk" cases in Omaha/Douglas County were successfully identified.
16 of 25 indicators tested were present in a majority of the cases reviewed.
9 of 10 major indicators identified in the Florida Mortality Review Project (321 cases reviewed) were also present in the majority of the Omaha cases.
Information/data for 14 "high-risk" indicators is currently stored in electronic/computerized databases by the agencies of the coordinated response.
Based on the study findings, ISED recommends that the DVCC and the evaluation committee pursue the following initiatives, to prevent and reduce the incidence of "high-risk" domestic violence and to improve the effectiveness of the coordinated response:
Development of a WEB-based management and information system (MIS) to connect the multiple data bases in the coordinated response and for use in:
Identification of high-risk cases
Prevention and intervention efforts
Annual reporting, quality control and outcome evaluation
Development of a domestic violence risk factor or "dangerousness" scale to identify high-risk cases (from among the approximately 450-480 domestic violence incidents reported to police and 225-240 cases prosecuted monthly).
Development of improved prevention and intervention strategies for both victims and offenders/suspects in "high-risk" cases.
Continued development and regular use of the case review process (not limited to serious-injury and lethal cases) as part of the ongoing efforts of the DVCC to develop mechanisms for quality control, performance monitoring and outcome evaluation.
The first phase of the evaluation of the official Omaha and Douglas County criminal justice system response to domestic violence (hereafter referred to as the coordinated response [1] ), conducted by the Institute for Social and Economic Development (ISED), consisted of process and data studies. [2] The Phase I report documented the major accomplishments made by the six components [3] of the coordinated response under the auspices of the Domestic Violence Coordinating Council of Greater Omaha (DVCC), as well as remaining data and implementation barriers. The report included recommendations for further improvements in the coordinated response and for future research and evaluation.
Phase II of the evaluation builds upon our initial research findings and is intended to foster the implementation of several of the most important recommendations made in the Phase I report. Specifically, the Phase II project focuses on two key elements or methods which were identified as necessary for the development of a formal quality control function for the coordinated response: (1) domestic violence case monitoring, and (2) the generation of aggregate statistical snapshots. Both are necessary to ensure that case processes are consistent with established domestic violence policies, protocols and procedures and that desired program performance and outcomes are achieved.
This report documents the findings from the Phase II evaluation of the coordinated response to domestic violence in Omaha. The purpose of the Phase II evaluation is twofold. First it documents and analyses the data obtained in a review of "high-risk" domestic violence cases that led to serious injury or loss of life. Second, based on information obtained and lessons learned during the case review process, it provides recommendations for improving existing information/data collection, utilization and reporting systems. Such improvements will be necessary to better inform domestic violence policy development, program performance management and reporting, and quality control for the coordinated response.
The remainder of this introduction is organized as follows: Section A reviews the literature of domestic violence research related to "high-risk" cases, indicators of fatal and serious-injury domestic violence and potential models for prevention/intervention. Section B describes the study design and methodology used for the Phase II evaluation and Section C concludes this chapter with the organization of the remainder of this report.
In this review, we focus on domestic violence homicide case reviews and other studies that investigate indicators or predictors of death or serious injury. We also examine studies of risk factors relevant to domestic violence prevention and intervention efforts and models or strategies that many research findings suggest could be effective in "high-risk" cases. (The literature cited in reference to prevention and intervention efforts is presented separately in Appendix D).
In the mid-1980's and throughout the 1990's, considerable academic and applied research specifically examined domestic violence cases and official records. This research was undertaken to better understand and identify risk-factors that might be important in reducing the incidence of lethal and serious-injury domestic violence.
Many of the researchers suggested that breaking down the various components of an abusive relationship and identifying "situational antecedents," allowed for a determination of factors most likely associated with, and therefore predictive of, lethal and serious-injury domestic situations ( Johnson, Li, Websdale, 1998 ). Such situational antecedents (for example, prior history of domestic violence, emotional or mental difficulties, and acknowledged conflict in the relationship, etc.) also included criminal justice system responses and demographic characteristics.
According to Websdale et al, the most important components and antecedents of abusive relationships, identified by the researchers in cases of intimate partner killings, include:
Escalating domestic violence and increasing entrapment of battered women ( Stark and Flitcraft, 1996 );
Separation/estrangement/divorce of the parties ( Daly and Wilson, 1998 );
Obsessive possessiveness or morbid jealousy on the part of the abusive partner ( Daly and Wilson, 1998 ; Easteal 1993 );
Threats to commit intimate partner homicide, suicide or both ( Hart, 1988 );
Prior agency involvement, particularly with law enforcement ( Browne, 1987 );
Issuance of protection or restraining orders against one of the parties, nearly always male ( Buteau et al, 1993 );
Depression on the part of the abuser ( West, 1967 ; Lester, 1992 );
Prior criminal history of violent behavior on the part of the abusive man ( Klein, 1993 ; Fagan, Stewart and Jensen, 1983 ; Bureau of Justice Statistics, 1998 ).
Research by Campbell ( 1986 , 1995 ) summarizes the key risk factors identified by the majority of experts in the field and notes that early "dangerousness assessments" were intended to help battered women determine their own levels of risk, rather than providing absolute definitions of risk. Finally, Hart ( 1988 ) identifies basically the same list of key indicators and points out, quite importantly, that when a number of these factors are present in combination, risk is considerably elevated.
Over the past decade, twenty-seven (27) jurisdictions in twenty-one (21) states (not including the current effort in Omaha/Douglas County) have initiated adult domestic violence death reviews employing varying methodologies ( Websdale, Sheeran and Johnson, 1998 ). Among the most notable completed reviews are the Charan Investigation of homicide-suicides in San Francisco (1990); the Philadelphia Women's Death Review (1995); the Santa Clara County, California Death Review (1997); the Kentucky Attorney General's Task Force (1993) and the New York Commission on Domestic Violence Fatalities (1997). [4]
The most comprehensive of the studies, however, is the Florida Mortality Review Project, which examined 321 domestic violence cases that resulted in domestic fatality in 1994.
A review of the findings showed that many factors were indeed strongly associated with cases that ended in the death of a domestic partner. Among the most important findings, the study revealed that domestic fatalities with female victims were usually final events in long-standing abusive relationships.
The identification of this and other specific predictors in the study, led to the development of ten (10) "red flags" to serve as bases for intervention and prevention of domestic fatalities by law enforcement, the courts and victim advocates. "Red flags" identified with domestic fatalities in Florida, included:
Prior history of domestic violence in the relationship
Prior criminal histories of both victims and perpetrators (perpetrator chronic and extensive)
Histories of mental illness or medical treatment of mental health problems
Morbid/unhealthy beliefs and behaviors by perpetrator including: obsessive-possessive, total control over a victim, extreme jealousy or stalking
Threats to kill
Victim attempts to separate or break away
Prior police calls to the residence
Drug or alcohol use prior to the incident
Perpetrator has a history of violent crimes (including weapons charges)
Victim has or had a protection order against the perpetrator.
The risk indicators identified in the Florida study generally concur with the major findings from the case reviews completed in other jurisdictions and from academic research ( Websdale, 1998 ). To determine whether these identified risk-factors were present and equally-relevant in the Omaha area, or whether additional indicators might also exist, it became evident that a local domestic homicide and serious-injury case review was needed.
In July 2000, the DVCC requested assistance from ISED to conduct a case outcome study to identify indicators or predictors of the most serious cases of domestic violence in the Omaha area, and also to improve measurement of the local incidence of domestic violence and the performance of the coordinated response. In response, ISED developed an evaluation plan to accomplish these tasks and this report documents the results of our efforts and the study findings.
The Phase II evaluation of the coordinated response consists of two studies: (1) a case outcome study and (2) a performance management and quality control study. Both studies were conducted while working closely and in active participation with the newly-established DVCC evaluation committee. [5]
A research protocol was developed by ISED for use by the evaluation committee during the case reviews. The protocol, which was modified and updated during the study, guided the research process and included: a general research strategy, a description of the case review process, a listing of indicators or predictors identified in previous case reviews in other jurisdictions, a listing of potential data sources, a data collection instrument (template), a description of the case selection methodology, and rules to ensure data confidentiality and security. (The data collection instrument and confidentiality/security agreement are shown in Appendix A).
While noting that the broader intent of the original ISED recommendation for conducting domestic violence case reviews was to provide a mechanism for quality control monitoring for the coordinated response, a modified focus necessarily guided the development of the research strategy. This study is intended to meet the following research objectives:
To test the transferability of key domestic violence research and mortality review findings from other jurisdictions in the U.S. to the Omaha/Douglas County area
To determine to what degree indicators or predictors of "high-risk" domestic violence cases could be identified locally and potentially used to reduce the incidence of future domestic injury or death
To determine whether a case review process could be developed which could be used in the future by the DVCC and the agencies of the coordinated response for on-going quality control and performance management
To document existing data sources, to explore whether the agencies of the coordinated response might be able to employ newer technology to capture and use such information more effectively.
To meet these objectives, the research design called for an examination of a sample of the most serious incidents of domestic violence in Omaha, that is, those that ended in either fatality or serious injury. The case reviews focused on to what degree indicators or predictors identified in previous research were present in those cases. More specifically, the research was intended to help answer questions about whether indicators or predictors of "high-risk" domestic violence cases [6] could be identified locally and potentially used to reduce the incidence of future domestic injury or death.
This study was also intended to capture, as part of the performance management and quality control study, the input and feedback of the local professionals (law enforcement, prosecutors, advocates, etc.) of the coordinated response. In particular, information was to be gathered about the case review process itself, the applicability and usefulness of potential indicators and the existence and accessibility of data and official records. Moreover, members of the evaluation committee were asked to be active participants in many aspects of the research process, including: operationalization (definition) of key variables, data collection, analyses of findings and making recommendations for policy and program improvements, as well as future research.
The study design required members of the evaluation committee to be active participants in the case review process. First, committee members were asked to review the initial research protocol developed by ISED and sign written agreements to participate in the study, collect specific information from official records and case files and maintain the confidentiality of individuals and any information gathered or discussed. (See Appendix A. DVCC Evaluation Committee Purpose and Memorandum of Understanding, June 20, 2001).
For each case selected for review, evaluation committee members (acting as representatives from each agency of the coordinated response) were to locate and collect specifically identified information/data from the case files and official agency records. Once gathered, the information was to be brought back to the committee and shared with the entire group during regularly scheduled meetings.
Accordingly, the group as whole was to review the information gathered (relating to each indicator being tested [7] ) to see if it met the criteria to be counted as "present" in the case under review. These decisions were to be subsequently recorded by researchers on a data collection instrument (template) developed for use in reviewing each case (see Section 5. The Data Collection Instrument).
Members of the committee and researchers agreed that the overriding priority for the research effort was to improve the coordinated response system. This was to be accomplished, if possible, by enabling the members of the coordinated response to identify "high-risk" cases and implement changes or interventions to address them more adequately. It clearly was not the intention of the committee to use the study or its findings "to point fingers of blame" in the cases being reviewed.
A secondary goal of the case reviews, however, was to note important "lessons learned" in how to review domestic violence cases. This effort was to include documenting, particularly for use in future case reviews (which may be more directly related to performance management and quality control in the coordinated response), instances where important established policies and protocols may not have been, or are not being, followed.
The evaluation committee reviewed a list of 20 potential indicators, initially identified in the research protocol prepared by ISED, which were drawn from the findings of previous research and case reviews in other jurisdictions. [8] Each potential indicator was reviewed in general terms of appropriateness, potential usefulness and applicability in the Omaha area.
Based upon committee input and feedback, some of the initially-proposed indicators were modified and refined in operational definition (in some cases adding additional sub-categories or qualifiers), some were dropped (mostly due to perceptions of redundancy), and some indicators were also added for investigation.
Overall, this process of revision and refinement of indicators continued throughout the test-case phase of the project (see Section 6. Sample Selection for the Case Reviews), and to a limited extent into later data collection and analyses stages. In all, a total of 25 major "high-risk" indictors or predictors were identified for investigation and were grouped into six (6) general categories which included: calls for assistance, reported incidents, criminal history, protection/restraining orders &probation/parole, health treatment &assessment and miscellaneous/other indicators.
Additionally, eight (8) of the major indicators were further broken down into sub-categories. For example, criminal history included sub-categories for prior arrest charges, prior arrests, charges filed for prosecution, convictions, and sentencing outcomes. Other key variables, such as demographic information, were also included. In all, data were to be collected for 54 variables.
The committee also identified the data sources and agencies that would be responsible for collecting specific information/data for each indicator. This identification and tracking of the data sources was considered critical for a number of reasons.
First, the documentation of existing data sources would be necessary for any subsequent case reviews. Second, as it was anticipated that many of the indicators would likely have multiple data sources, it was important to document each source to be able to determine those that might be more reliable and consistent providers of information, and to see if comparisons of sources yielded different results. Finally, a documentation of the data sources was important to explore whether the agencies of the coordinated response might be able to employ newly-developed or existing technology to capture and use such information in an automated manner.
ISED designed the initial data collection instrument (template) with multiple purposes in mind. First it was intended primarily to organize and simplify the collection of needed data and information. Second, the instrument was created in an electronic spreadsheet form to allow for the efficient manipulation of data and tabulation of aggregate results.
Finally, the research instrument was also intended to provide a straightforward means for evaluation committee members to tie the list of key indicators being investigated with the data sources that would need to be consulted. That is, the initial template was to provide a working document for the committee to use to finalize the list of indicators and also to identify the data sources and agencies that would be responsible for gathering the needed information. The final version of the research instrument is shown in Appendix B.
Initially, three (3) test cases consisting of a homicide case, a homicide/suicide case, and a serious-injury case were selected for review by the Douglas County Attorney's office and researchers. The purpose of conducting test cases was three-fold: 1) testing the review process, including the refinement of key variables and operational definitions, 2) verifying the existence, accessibility and reliability of data sources, and 3) refining the data collection instrument and tabulation spreadsheet.
Representatives from each agency reviewed their files and data sources for each test case and brought the information back to the group for discussion and review. After necessary changes had been made in case review procedures and the research instrument, twelve (12) additional cases were randomly selected for review.
A list of domestic violence homicide and homicide/suicide cases provided by the DVCC and a combined list of serious-injury cases [9] provided by the OPD-Domestic Violence Investigation Squad and the Douglas County Attorney's Office were used as the sampling frames for the study. [10] The stratified sample of 15 cases selected for review includes 7 homicide cases, 3 homicide/suicide cases, and 5 serious-injury cases (including the test cases). [11]
Only cases that occurred after the coordinated response was established in 1997 and that had been officially "closed" by either the death of both partners or a judicial sentence, were included in the sample. [12] Due to the exploratory nature and budgetary limitations of the research project, no attempt was made to draw a sample that would be statistically representative of all domestic violence homicides and serious-injury cases in Omaha.
Maintaining the confidentiality and security of information was of primary importance to members of the evaluation committee from the outset. While ISED provided the initial rules and guidelines in the research protocol, these were strengthened and expanded by the committee to cover a broader range of issues and concerns in this area.
Committee members were asked to sign and strictly adhere to a comprehensive confidentiality agreement (see Appendix A.) in which they agreed to keep any information or discussion of cases under review strictly confidential. To address additional concerns, the committee was also subsequently also divided into two groups: 1) those with official access to criminal justice case records (the criminal justice review group) and 2) those without such access (the victim advocate group).
The remainder of this report is organized as follows: Chapter II discusses case review findings and analysis regarding the indicators of "high-risk" domestic violence cases, data sources, and the case review process. To complete the report, Chapter III presents ISED's conclusions and recommendations for future development efforts and further investigation.
In this chapter, we report on findings and analysis of a case review study conducted as part of the coordinated response to domestic violence in the greater Omaha area. We begin with a presentation of major findings regarding indicators of "high-risk" domestic violence cases.
In the next section, we discuss the further classification of indicators based on the findings, consensual judgment of the evaluation committee and the availability of electronic data sources. Our conclusions are presented in Chapter III, which also includes recommendations for further development efforts and investigation to improve the coordinated response.
A total of 25 "high-risk" indicators were investigated as to whether or not they were associated with or "present in" the 15 lethal and serious-injury domestic violence cases selected for review. These indicators were grouped into six general categories: 1) Calls for assistance, 2) Reported Incidents, 3) Criminal History, 4) Protection Orders/Probation &Parole, 5) Health Treatment and Assessment, and 6) Other Indicators. See Table II.1.
Five (5) of the 25 major indicators were also further broken down into more-specific sub-categories in order to gather more-detailed information. For example, indicator 12. Multiple Protection/Restraining Orders, also includes sub-categories containing additional information about a) the number of cases where only one protection/restraining order had been issued, and b) the number of probation and parole violations.
In addition, information/data was also gathered for other key variables that may be important in identifying "high-risk" cases, such as demographic information and criminal justice system outcomes. Altogether data was collected for 54 different variables.
The information gathered for each indicator was also grouped by type of case, that is whether it was a homicide, homicide/suicide or serious-injury case. Table II.1 shows the categories, major indicators and a summary of the results obtained.
Table 1. Table II.1. Indicators/Predictors of Serious-Injury and Lethal Domestic Violence In Omaha/Douglas County
| Indicator (Risk Factor) | Homicide (N=7) | Homicide/Suicide (N=3) | Serious Injury (N=5) | Total (N=15) N % |
|---|---|---|---|---|
| Calls For Assistance | ||||
| 1. Multiple calls to CJ Authorities | 6 | 3 | 4 | 13 86.7% |
| 2. Prior history of DV in Relationship | 5 | 2 | 5 | 12 80.0% |
| 3. Multiple calls to Advocates | 2 | 0 | 0 | 2 13.3% |
| Reported Incidents | ||||
| 4.Multiple incidents Reported to CJ | 4 | 2 | 3 | 9 60.0% |
| 5. Multiple incidents Reported to Advocates | 2 | 1 | 3 | 6 40.0% |
| Criminal History | ||||
| 6. Chronic/Extensive(offender/suspect) | 4 | 3 | 4 | 11 73.3% |
| 7.Violent crimes (including weapon charges)[offender/suspect] | 5 | 3 | 4 | 12 80.0% |
| 8.DV and Abuse(offender/suspect) | 3 | 2 | 3 | 8 53.3% |
| 9.Drug Use/Abuse(offender/suspect) | 3 | 1 | 4 | 8 53.3% |
| 10. Arrest Warrants(offender/suspect) | 6 | 2 | 4 | 12 80.0% |
| 11. Chronic/Extensive(victim/injured) | 0 | 0 | 2 | 2 13.3% |
| Protection Orders/Probation &Parole | ||||
| 12. Multiple Protection/Restraining Orders(2 +) [vs. offender/suspect] | 3 | 0 | 2 | 5 33.3% |
| a) Multiple Protection/Restraining Orders(1+) [vs. offender/suspect] | 4 | 1 | 3 | 8 53.3% |
| b) Protection/Restrain Order Violations (vs. offender/suspect) | 2 | 1 | 1 | 4 26.7% |
| 13. History of Probation or Parole (offender/suspect) | 5 | 1 | 5 | 11 73.3% |
| a) Probation/Parole Violation (offender/suspect) | 2 | 1 | 3 | 6 40.0% |
| Health Treatment/Assessment | ||||
| 14. Medical/Emergency Room Treatment for DV(victim/injured) | 0 | 0 | 3 | 3 20.0% |
| 15. Mental Health Treatment, Assess,Counsel (victim/injured) | 1 | 0 | 1 | 2 13.3% |
| a) Mental Health Treatment, Assess, Counsel (offender/suspect) | 2 | 0 | 1 | 3 20.0% |
| b) DV Mental Health Treat,Assess, Counsel (victim/injured) | 0 | 0 | 0 | 0 0.0% |
| c) DV or batterer's mental health treatment/assess/counseling (offender/suspect) | 0 | 0 | 0 | 0 0.0% |
| 16. Lethality Assessments | n/a | n/a | n/a | n/a |
| 17. Morbid/unhealthy behavior(obsess/possess, jealousy, stalking, betrayal, control) [offender/suspect] | 6 | 2 | 5 | 13 86.7% |
| Other Indicators | ||||
| 18. Attempts by victim/injured to leave, separate, relocate | 4 | 2 | 5 | 11 73.3% |
| 19. Threats by offender/suspect to kill or injure the victim | 4 | 2 | 5 | 11 73.3% |
| a) Threats by offender/suspect to kill or injure others | 3 | 1 | 4 | 8 53.3% |
| b) Threats by offender/suspect to kill or injure themselves | 1 | 0 | 1 | 2 13.3% |
| 20.Pattern of abuse in relationship(s):persistent, escalating, strangulation, weapons, drugs/alcohol | 4 | 2 | 5 | 11 73.3% |
| 21. Victim believes her/himself to be in danger | 4 | 2 | 5 | 11 73.3% |
| a) Victim believes something is morbidly strange, very wrong, or frightening | 4 | 2 | 5 | 11 73.3% |
| 22. Family/friends believe victim is in danger, something strange, etc. | 4 | 1 | 4 | 9 60.0% |
| 23. Changes in offender/ suspect employment, or long-term unemployment | 1 | 0 | 0 | 1 6.7% |
| 24. Children involved or witness abuse/injury | 3 | 1 | 2 | 6 40.0% |
| 25. Animal Abuse | 0 | 0 | 0 | 0 0.0% |
Overall, 16 of the 25 major indicators/predictors investigated were associated with or "present in" a majority of the 15 cases. [13] Of these, five (5) were found in over 80% or more of the cases, five (5) were in the 70%-79.9% range, two (2) were in the 60%-69.9% range and four (4) were in the 50%-59.9% range of the cases reviewed. Significantly, 9 of the 10 major indicators identified in the Florida Mortality Review Project (321 cases examined) were also present in the majority of the Omaha/Douglas County cases reviewed (see Chapter I.A.2 and Table II.2 below). [14]
The associations of indicators with the high-risk cases reviewed ranged from a high of 86.7% (13/15) for multiple calls to criminal justice authorities (911 calls) and for morbid/unhealthy offender/suspect behavior to 0.0% for a history of mental health treatment, assessment or counseling for both the victim and offender/suspect, changes in employment status &long-term unemployment, and animal abuse reported. [15] Table II.2 on the following page presents a rank ordering of the association (shown in percentage) of each indicator/predictor with the "high-risk" cases reviewed.
Table 2. Table II.2:Rank Order of Indicators/Predictors of Serious-Injury and Lethal Domestic Violence In Omaha/Douglas County
| 1 | Multiple calls for assistance to criminal justice authorities (911) | 86.7%*(e) |
| Morbid/unhealthy offender behavior: obsessive/possessive,jealousy, stalking, betrayal, control | 86.7* | |
| 2 | Prior history of domestic violence in relationship | 80.0*(e) |
| History of violent crimes (including weapons) | 80.0*(e) | |
| Arrest warrants issued for offender | 80.0(e) | |
| 3 | Chronic/extensive criminal history (offender) | 73.3*(e) |
| Offender/suspect history of probation or parole | 73.3(e) | |
| Attempts by victim to leave, separate or relocate | 73.3* | |
| Threats by offender/suspect to kill or injure victim | 73.3* | |
| Victim believes her/himself to be in danger or something is morbidly strange, very odd or frightening | 73.3 | |
| 4 | Evidence of a pattern of abuse in relationship(s): persistent, escalating, strangulation, weapons or drugs/alcohol | 60.0(e) |
| Multiple DV incidents reported to criminal justice authorities | 60.0*(e) | |
| Family members, friends or others believe victim in danger or something morbidly strange or wrong | 60.0 | |
| 5 | Prior history of domestic violence crime (offender/suspect) | 53.3*(e) |
| Prior history of drug/alcohol crime (offender/suspect) | 53.3*(e) | |
| Threats by offender/suspect to kill or injure others | 53.3 | |
| Protection/restraining order(s) against offender | 53.3*(e) | |
| 6 | Multiple DV incidents reported to advocates | 40.0(e) |
| Probation or parole violations by offender/suspect | 40.0(e) | |
| Children involved or witness abuse, injury or violence | 40.0 | |
| 7 | Protection/restraining order violations by offender/suspect | 26.7(e) |
| 8 | Medical/emergency room treatment for DV (victim) | 20.0 |
| Mental health treatment, assessment or counseling (offender) | 20.0* | |
| 9 | Multiple calls for assistance to advocates, crisis lines, shelters | 13.3 |
| Mental health treatment, assessment or counseling (victim) | 13.3 | |
| Threats by offender/suspect to kill or injure themselves | 13.3 | |
| 10 | Domestic violence mental health treatment, assessment or counseling (victim) | 0.0 |
| Domestic violence or batterer's mental health treatment, Assessment or counseling (offender) | 0.0 | |
| Changes in offender employment or long-term unemployment | 0.0 | |
| Animal abuse reported | 0.0 | |
| Lethality Assessments (conducted and/or high scores) | n/a |
*=Florida Mortality Review Project Indicator (e)=Data available in Omaha/Douglas County in electronic form
In addition to the situational antecedent and criminal justice system response information reported above, demographic information may also be useful in identifying "high-risk" cases. Indeed, some demographic variables could be considered risk-factors in and of themselves in domestic abuse relationships, or could be useful in further refining the major risk factors/indicators which have been identified.
Gender. In our study, 14 of 15 (93.3%) offenders/suspects were male and 14 of 15 victims/injured were female. The inclusion in the sample of one domestic homicide case in which a female killed a male, was considered controversial by some members of the evaluation committee. As the female was later acquitted of manslaughter charges in the case due to self-defense, some felt that including this case (with her labeled as an "offender" and the male, who had previous incidents of domestic abuse reported against him, labeled as a "victim") was inappropriate.
The research team, however, contended that if the primary purpose of the research was to prevent domestic killing and serious injury to either participant in the relationship, such cases should be included. In addition, many in the field contend that such self-defense killings may, in many instances, be highly likely and predictable. After reviewing this case, labels were changed from "offender" to "offender/suspect" and from "victim" to "victim/injured."
Age. The average offender/suspect age was 34.1 years and victims/injured averaged 29.5 years. Interestingly, average ages of offenders/suspects and victims/injured varied according to the type of incident: homicide/suicides were the oldest (offender/suspect 40.3 years, victim/injured 39.0 years); followed by homicides (offender/suspect 34.3 years, victim/injured 28.1 years); and serious-injury cases (offender/suspect 30.0 years, victim/injured 25.1 years).
Geographic Location. The geographic location of the victim/injured residence was east of 72nd Street in 86.0% of the cases (66.0% north of and 20% south of Dodge St.) and west of 72nd St. in 14.0% of the cases. As other demographic variables such as income, education and employment levels (data for which is currently unavailable for domestic violence partners) are known to vary according to geographic area in Omaha/Douglas County (e.g., generally lower levels of each for those living east of 72nd St.), geographic location may be an important risk-factor.
Additional information was also gathered for criminal justice system responses, beyond that for the major indicators/predictors being investigated. For example, offenders/suspects averaged a total 5.1 felony arrest charges in their criminal histories (76 felony arrest charges for the offender/suspects in the 15 cases reviewed) and 12.6 misdemeanor arrest charges, while victim/suspects averaged 0.4 felony arrest and 2.1 misdemeanor arrest charges. [16]
Information was also gathered for total arrests, [17] charges filed by prosecutors, convictions and sentencing (including probation). As shown in Table II.3, offender/suspects had at least one prior felony conviction in 10 (66.7%) of the 15 cases, had received a least one prior jail sentence for a felony in 7 (46.7%) of the cases, and had been sentenced to probation for a felony (possibly in addition to a fine or jail sentence) in 4 (26.7%) of the cases. [18]
Table 3. Table II.3: Offender/Suspect Prior Felony Convictions and Sentences*
| Felony Convictions (Off/Susp) | Homicide | Homicide/Suicide | Serious Injury |
|---|---|---|---|
| Case #1 | 0 | 4 (4 Jail) | 2 (1 Prob/1Jail) |
| Case #2 | 0 | 0 | 5 (5 Jail) |
| Case #3 | 0 | 1 (1 Probation) | 1 (1 Probation) |
| Case #4 | 0 | 1 (1 Jail) | |
| Case #5 | 3 (3 Jail) | 4 (4 Jail) | |
| Case #6 | 1 (1 Probation) | n/a | |
| Case #7 | 1 (1 Jail) |
*Does not include convictions or sentences for the presenting offenses (i.e., for the homicide or felony assault cases) being reviewed.
After reviewing the findings of association with high-risk cases, the evaluation committee considered each indicator in an attempt to arrive at a group consensus about their importance and usefulness in identifying high-risk cases. Based on study findings, professional experience and consideration of data availability issues, indicators were classified as "good indicators" if the group agreed they would be important and useful, or "questionable indicators" if there was not a consensus or additional information was needed.
The researchers and evaluation committee also reviewed 16 data sources to re-confirm whether information for each indicator was stored and could potentially be retrieved in electronic form (automated data files), rather than in paper files. The 12 major data sources used in the study (e=available in electronic format) include: The Douglas County Mainframe (e), Douglas County Communication 911 Database (e) , the OPD Domestic Violence Tracking System (e), Nebraska State Probation NPMIS (e), OPD/Douglas County Victim Assistance Unit Program (e), the OPD Domestic Violence Uniform Crime Report (UCR) and Domestic Violence Supplemental Form, Douglas County Prosecutor files, YWCA files (e), Catholic Charities files, and Douglas County Protection Order Office files.
Based primarily on the availability of information in electronic format, each indicator was further classified as "available now" or "not available now." Overall, information was available for 14 of the indicators/predictors in electronic/computerized format. Table II.4 shows the three classifications of indicators. Indicators within each classification are ranked in order of their association with high-risk cases.
Table 4. Table II.4: Classification and Rank Order of Indicators/Predictors of Serious-Injury and Lethal Domestic Violence in Omaha/Douglas County
| Good Indicator, Available Now | ||
| 1 | Multiple calls for assistance to criminal justice authorities (911) | 86.7%*(e) |
| 2 | Prior history of domestic violence in relationship | 80.0*(e) |
| History of violent crimes (including weapons) | 80.0*(e) | |
| 3 | Chronic/extensive criminal history (offender) | 73.3*(e) |
| Offender/suspect history of probation or parole | 73.3(e) | |
| 4 | Evidence of a pattern of abuse in relationship(s): persistent, escalating, strangulation, weapons or drugs/alcohol | 60.0(e) |
| Multiple DV incidents reported to criminal justice authorities | 60.0*(e) | |
| 5 | Prior history of domestic violence crime (offender/suspect) | 53.3*(e) |
| Prior history of drug/alcohol crime (offender/suspect) | 53.3*(e) | |
| Protection/restraining order(s) against offender | 53.3*(e) | |
| 6 | Multiple incidents reported to advocates | 40.0(e) |
| Probation or parole violations by offender/suspect | 40.0(e) | |
| 7 | Protection/restraining order violations by offender/suspect | 26.7(e) |
| Good Indicator, Not Now Available | ||
| 1 | Morbid/unhealthy offender behavior: obsessive/possessive, Jealousy, stalking, betrayal, control | 86.7* |
| 2 | Attempts by victim to leave, separate or relocate | 73.3* |
| Threats by offender/suspect to kill or injure victim | 73.3* | |
| Victim believes themself to be in danger or something Is morbidly strange, very odd or frightening | 73.3 | |
| 3 | Family members, friends or others believe victim in danger or Something morbidly strange or wrong | 60.0 |
| 4 | Children involved or witness abuse, injury or violence | 40.0 |
| Questionable Indicators | ||
| 1 | Arrest warrants issued for offender | 80.0(e) |
| 2 | Medical/emergency room treatment for DV (victim) | 20.0 |
| Mental health treatment, assessment or counseling (offender) | 20.0* | |
| 3 | Multiple calls for assistance to advocates, crisis lines, shelters | 13.3 |
| Mental health treatment, assessment or counseling (victim) | 13.3 | |
| Threats by offender/suspect to kill or injure themselves | 13.3 | |
| 4 | Domestic violence mental health treatment, assessment or counseling (victim) | 0.0 |
| Domestic violence or batterer's mental health treatment, Assessment or counseling (offender) | 0.0 | |
| Changes in offender employment or long-term unemployment | 0.0 | |
| Animal abuse reported | 0.0 | |
| Lethality Assessments (conducted and/or high scores) | n/a |
*=Florida Mortality Review Project Indicator (e)=Data available in Omaha/Douglas County in electronic form
As shown in the table, even though some indicators had relatively low association scores with high-risk cases, such as protection order violations by offender/suspect (26.7%), they were classified as "good indicators" by the committee based on findings from other jurisdictions and professional experience. Some indicators, such as mental health or emergency room treatment and employment, however, were classified as "questionable indicators" at this time, due to uncertainties about their predictive ability or importance, and/or based on the local findings and data availability issues.
In the next chapter, we present our conclusions and recommendations for further development and investigation, to prevent and reduce the incidence of "high-risk" domestic violence and to improve the effectiveness of the coordinated response.
The purpose of the coordinated response is to more effectively process domestic violence cases, serve the needs of victims, and prevent/reduce the incidence of domestic violence. The findings and data source documentation provided in this study have important implications for these purposes in four (4) areas: 1) indicators/predictors of "high-risk" cases, 2) data sources and information technology, 3) prevention/intervention strategies, and 4) the case review process.
Risk factors associated with serious-injury and lethal domestic violence cases were successfully identified and tested in the study. Sixteen (16) of 25 indicators tested were present in a majority of the 15 cases reviewed in Omaha/Douglas County. Significantly, 9 of 10 predictors identified in the Florida Mortaility Review Study (which examined 321 lethal DV cases) were also present in the majority of Omaha cases.
These findings suggest that the key domestic violence research and mortality review findings from other jurisdictions in the U.S., regarding indicators or predictors of high-risk domestic violence cases (see Chapter I.), are highly transferable to the Omaha area. The results also add to the existing evidence that numerous key indicators/predictors of "high-risk" domestic violence (e.g., situational antecedents in abusive relationships and criminal justice system responses such as multiple police calls and protection orders) have a high degree of generalizability across urban jurisdictions throughout the U.S.
More specifically, the findings support the contention of many researchers that abusive relationships may be broken down into "situational antecedents" (such as prior criminal history, emotional or mental difficulties, and acknowledged conflict in the relationship, etc.) that allow for a determination of factors (or combinations of factors) most likely associated with, and therefore predictive of, lethal and serious-injury domestic violence.
However, certain cautions or caveats regarding the use of indicators or predictors of domestic violence lethality and serious-injury (and associated assessment instruments) are required. Websdale notes, for example, that "?[results obtained by] working backward from domestic homicides is problematic because it assumes that certain permutations, combinations, and intensities of antecedents, culminate in or indeed cause death." He cautions that association or correlation is not proof of causation.
One of the biggest problems with lethality assessment instruments (and their application), according to Websdale, is that "lethality indicators" are characteristics of many domestic violence relationships, the vast majority of which do not end in death. Despite these difficulties and limitations in predicting lethality, Websdale concludes such instruments are powerful "dangerousness indicators," which can be tremendously useful in identifying cases that are "more likely" to result in serious-injury or death and in developing and targeting more effective intervention strategies.
Based on a review of the literature, the case review study findings and conclusions, ISED makes the following recommendations:
The evaluation committee of the DVCC should build on the study findings regarding indicators/predictors of "high-risk" domestic violence cases, through an initiative to develop a domestic violence risk-factor or "dangerousness scale."
Such a scale is needed to identify "high-risk" domestic violence cases from among the approximately 450-480 domestic violence incidents which are reported to law enforcement and 225-240 cases which are prosecuted monthly in Omaha/Douglas County.
The domestic violence "dangerousness scale" should be developed in conjunction with an investigation of similar domestic violence assessment instruments that have already been developed and implemented in other jurisdictions in the U.S. (An example of one such instrument, developed by the Colorado State Court Administrator's Office, Office of Probation Services, is shown in Appendix C).
Multiple data systems/sources for indicators/predictors of domestic violence and other variables (such as demographic information) were successfully identified during the case review process. Currently, data are available in these systems in automated/computerized format for 14 of the domestic violence indicators (13 of which were classified as "good indicators"), as well as for demographic variables, such as age, race, gender, and location of victim residence.
These findings have promising implications for the use of new technology in the coordinated response. More specifically, technology that was developed during the 1990's to make use of capacities inherent in the world-wide web and WEB-based management and information systems (MIS), [19] could potentially improve the effectiveness of the coordinated response and the DVCC in three areas: 1) operations/interventions, 2) quality control and performance monitoring/reporting, and 3) outcome evaluation.
Operations/Interventions. The problem of critical information residing in multiple data bases (as is the case in the coordinated response) and the need to connect those sources/systems is not a new one. The primary reason for creating the DVCC/OPD-Domestic Violence Tracking System (DVTS) was to improve the sharing of information and improve communication and coordination in the operations (interventions) of the coordinated response (see Phase I report).
The DVTS has largely been under-utilized, however, among the agencies of the coordinated response. While the Omaha Police Department (which developed and maintains the computer program and database) regularly uses and enters data into the system, [20] information is not being entered or used by the other agencies of the coordinated response (victim advocates, prosecutors, protection order office, and probation) as originally intended. [21]
Utilization of the DVTS breaks down primarily due to the system requirement for agency personnel to re-enter data into the DVTS from other official records or data systems. Limitations on available personnel, time and resources, and training and turnover issues have all contributed to the relatively ineffective use of the DVTS in coordinated response operations and interventions.
Quality Control and Program Performance Monitoring/Reporting. As noted earlier, the case review process included the collection of data from all agencies in the coordinated response ranging from initial complaint (911 calls) to final case disposition (sentencing and post-adjudication supervision [probation or parole]). Just as a WEB-based MIS might be used to more-effectively improve communications and operations (interventions) in the coordinated response, it could also potentially be used to automate and greatly improve quality control and program performance monitoring and reporting in the coordinated response.
Reports could automatically be generated by each agency to track their agency's activities, personnel, and outcomes related to domestic violence interventions. Reports could also be generated on weekly, monthly, quarterly, or annual bases, that would describe the number of domestic violence 911 calls, incidents, investigations, arrests, advocate contacts and services, protection orders, prosecutions, adjudication outcomes (convictions and sentences) and post-adjudication supervision contacts.
Such information could be used by each agency in the coordinated response to track, document and improve their performance in the coordinated response. In addition, such "statistical snapshot" information (some of which the DVCC currently collects from each agency monthly and forwards to the Nebraska Crime Commission as required) could be provided to the DVCC in automated format, over the internet, by each agency in the coordinated response.
Beyond this, such statistical information could also be used by the DVCC for a variety of other purposes, the most important of which may be to prepare an Annual Domestic Violence Report (as recommended by ISED in the Phase I report). Having such "real-time" information [22] available in computerized format in a WEB-based MIS, it could readily be used by the DVCC, agency staff and researchers to make annual comparisons, to identify and document trends in domestic violence and to monitor the activities/performance of the coordinated response over time.
Outcome Evaluation. Just as a WEB-based MIS might be used to more-effectively improve quality control and program performance monitoring and reporting in the coordinated response, so might it also be used as a tool by researchers to conduct outcome evaluation of the effectiveness of the coordinated response. In particular, the establishment of such a WEB-based MIS could potentially overcome several major obstacles that have hindered or prevented attempts to conduct domestic violence research and outcome evaluations in the past. [23]
Specifically, two obstacles that have been identified by ISED and other researchers in Omaha/Douglas County are: 1) the difficulty of linking domestic violence victims and offenders in official records, and 2) the difficulty of identifying domestic violence crimes in the Douglas County criminal history and other data bases. [24] Based on our observations and the findings obtained in the case review process of data systems/sources, we conclude that a WEB-based MIS could potentially be used to overcome existing obstacles and more efficiently conduct domestic violence research/evaluation efforts.
Based on the study findings and additional observations, ISED makes the following recommendations:
The evaluation committee of the DVCC should investigate the development of a WEB-based management and information system (MIS) to connect the agencies and multiple data bases of the coordinated response. Such a system has the potential to greatly improve the effectiveness of the coordinated response in the following areas:
Criminal justice system operations and domestic violence interventions
Quality control, performance monitoring and reporting
Evaluation of offender recidivism/behavior, victim safety and other program outcomes
A risk-factor or "dangerousness scale" developed to identify "high-risk" cases (see section B. above) should be integrated into a WEB-based information system that can access multiple databases.
The primary objective of the case review study was to identify indicators or predictors of "high-risk" cases. A corollary objective of the evaluation committee was to also investigate and develop interventions for these cases, to prevent or reduce the incidence of domestic homicide or serious injury. To facilitate further discussion and development in this area, additional literature was reviewed by ISED and is included in Appendix D.
This review reveals that many findings regarding domestic violence risk-factors, from studies of domestic violence arrests, protection/restraining orders, victim feedback, chronic offenders, court reviews, and batterer programs are strikingly consistent with those found in the literature of domestic violence lethality studies and case reviews. In particular, these findings and the supporting evidence from our local case review study, support the contention of many professionals and academics, that "high-risk" domestic violence offenders are not specialists in domestic violence crime, as is [has been] commonly believed in the field.
Rather, such "high-risk" offenders are likely to have extensive, chronic and violent criminal histories that cover a wide range of criminal activities including, non-domestic assaults, drug/alcohol involvement or abuse and weapons violations. We believe these findings have important implications for the development of interventions in Omaha/Douglas County, for both victims and offender/suspects, which can prevent and reduce the incidence of lethal and serious-injury domestic violence.
Based on the study findings and conclusions and our review of the literature, ISED makes the following recommendations:
The evaluation committee of the DVCC should continue to investigate interventions for "high-risk" domestic violence cases, as follows:
A coordinated, multi-agency intervention strategy for "high-risk" domestic violence offenders/suspects should be developed and officially incorporated into existing policies and protocols for the coordinated response.
A coordinated, multi-agency intervention strategy for "high-risk" domestic violence victims should be developed and officially incorporated into existing policies and protocols for the coordinated response.
Interventions for "high-risk" offenders and victims should, if possible, build upon strategies, models and programs that have been successfully implemented, tested and evaluated in comparable U.S. cities and jurisdictions.
An extensive case review process, involving all agencies of the coordinated response serving as members of the DVCC evaluation committee (assisted by ISED), was developed, implemented and successfully tested. The primary objectives of this case review process (to identify indicators or predictors of "high-risk" DV cases and data sources in Omaha/Douglas County) were research-based and successfully completed, to a great extent, due to the oversight and diligent completion of research tasks by the members of the evaluation committee.
In large measure, this study tested whether member agencies of the coordinated response could work together effectively to oversee and complete the steps and tasks necessary in a traditional research project: identification of a study question(s), development of a research methodology, data collection, data manipulation, data analysis and formulation of conclusions based on findings. Based on these criteria, our conclusion is that the "evaluation committee format" is a very effective mechanism for conducting domestic violence case reviews for research purposes.
The case review process also broke new ground and met secondary objectives of the project in two areas identified by ISED in the Phase I report. [25] First, the case review study successfully provided researchers and the DVCC with highly-detailed and important insights about criminal justice and advocate processes in the coordinated response. Specifically, information was successfully collected along each step of the coordinated response process flow (documented by ISED in the Phase I report) consistent with how a typical domestic case might unfold. These critical data collection points are shown in Appendix E.
The case review process also successfully identified instances where important established policies and protocols may not have been, or are not being followed in the coordinated response. During the study, researchers noted, for example, that victim advocates from the YWCA indicated that they were only being notified of incidents by law enforcement in about 25% of the cases, whereas established coordinated response protocol calls for notification of advocates in 100% of the cases. This observation was supported by the case review findings that advocates had received calls for assistance in only 2 of the 15 cases reviewed (13.3%). [26]
Based on the serious-injury and lethal domestic violence case study findings, literature review and conclusions presented in this report, ISED makes the following recommendations:
Case reviews (not limited to serious-injury and lethal cases) which build on the institutional knowledge and experience gained in this study, should be conducted regularly as part of the coordinated response. Such a case review process should be part of the ongoing efforts of the DVCC to develop mechanisms for quality control, performance monitoring and outcome evaluation.
In future case reviews, process, protocol and performance benchmark criteria (based on existing policies and procedures in the coordinated response) should be established and incorporated into the review process.
Evaluation Committee Memorandum of Understanding (Chapter I) (This link will open MS Word document.)
Research Instrument (Template)(Chapter I) (This link will open a spreadsheet in MS Excel.)
Literature Review (Development of Interventions) (Chapter III)
Along with the literature cited in Chapter I., additional domestic violence research, not limited to fatality cases, has also focused specifically on risk-factors or characteristics which are strongly associated with serious-injury domestic violence. Risk-factors which have been identified for such intimate-partner violence, from a perspective of health and development ( Moffitt, Caspi, Silva, 1998 ), include: conviction for other types of crime, especially violent crime ; a background of family adversity, leaving school early, and juvenile aggression; a variety of mental illnesses; drug abuse; long-term unemployment; and motherhood and cohabitation at a young age.
Similarly, research focusing on domestic protection and restraining orders and factors associated with their effectiveness, reported findings identifying high-risk domestic offenders and victims ( Keilitz, Hannaford, Efkeman, 1998 ). Data from interviews and criminal court records showed that: most petitioners had suffered physical abuse and more than half of the cases were severe, and about 25% had endured abuse for more than five (5) years; the longer the women experienced abuse, the more intense the abusive behavior became; the majority of abusers had criminal records and had been arrested for violent crimes other than domestic violence; victims whose partners have a history of violent crime are more likely than other victims to be revictimized after receiving protection orders.
In another study of the effectiveness of restraining orders ( Harrell and Smith, 1998 ), investigators found that more than half of victims had been physically injured during the incident that led them to seek the court order. The majority of women (60%) who obtained orders reported that the orders were violated in the year after they were issued. Almost one-third reported that the violations involved severe violence.
Perhaps even more significantly, Harrell and Smith found that the severity and persistence of prior domestic abuse were significantly related to the severity of subsequent abuse. In other words, the more severe and persistent the history of abuse in a relationship, the greater the likelihood of even more severe (and perhaps fatal) abuse in the future.
It should also be noted that in this study, the severity of an incident described in any one particular complaint did not, however, predict the subsequent number or specific types of abuse (e.g., verbal threats, physical violence, violence using a weapon, etc.) in the following year. Interestingly, researchers also found that men who voiced strong objections to restraining orders were three to four times as likely to violate the orders.
Other significant predictors of abuse for court-restrained male batterers (the average record length was 13 complaints) were age, prior criminal history and court-ordered "no-contact" provisions ( Klein 1996 ). Klein also reported that younger abusers re-abuse more than older abusers, those with criminal histories abuse more than those without, and those with longer prior records abuse regardless of whether or not the orders specify "no contact" with the victims. Variables that did not predict new abuse included: an arrest at the time of the order, the violent or non-violent nature of the precipitating order and the existence of a prior restraining order.
In addition, in a study of cases in a pro-active domestic violence court in Quincy, Massachusetts ( Buzawa et al, 1997 ) found that prior to the presenting offense, almost 70% of victims interviewed felt they were going to be seriously injured in the future by their partner. After the offense, almost one-fourth (22%) felt they needed medical attention as a result of the incident.
Perhaps particularly relevant to the situation in Omaha/Douglas County, [27] despite being labeled as misdemeanors, 71% of incidents in that jurisdiction involved physical violence, 10% of victims experienced a serious injury (including broken bones, broken noses, internal injuries, lacerations and loss of consciousness) and another 27% experienced moderate injuries such as bruises, swellings and joint soreness. In addition, in 16% of incidents, a weapon other than hands or feet (firearm, knife, blunt object, motor vehicle, etc) was used by the offenders.
Researchers also identified other predictors including: child abuse (victims had 6 to 10 times the rate of child physical abuse and 36% were victims of child sexual abuse) and prior violent adult relationships (36%). Almost 75% had called police on a prior occasion about the same offender and less than half of the victims were living with the offender at the time of the incident (many had moved away from a violent partner/husband) although about 75% of the victimizations occurred in their homes.
As for offenders, a restraining order history proved to be a reliable predictor for other criminal activity and future violence. Offenders who had an active restraining order out at the time of the incident were almost twice as likely to re-offend against the same victim with one year. Additionally, those with multiple victims had a significantly greater number of restraining orders and a greater number of prior drug/alcohol charges. Offenders who had restraining orders taken out against them in the past by more than one female were 13 times more likely to re-offend against a new, different victim.
Another study in Massachusetts of 18,000 men subject to restraining orders ( Jones 1994 ) found that about 75% had some sort of prior criminal record, including: over half (51%) had criminal histories for a violent crime, over a quarter (27%) for driving under the influence, more than 20% for a drug offense. A second study ( Soloman and Thomas, 1997 ) showed that domestic violence offenders were not "specialists" in domestic violence crime, as commonly thought, and just as likely as non-domestic offenders to have committed non-domestic offenses in the previous five years (almost half of each group studied had been arraigned for non-domestic offenses and the two groups had statistically equal proportions of high-rate offenders).
Further, in research ( Buzawa and Buzawa 1996 ) based on victim interviews, Klein concluded that, " batterer's as a group were demonstrably dangerous. Fifty percent of them had prior criminal records of which the victim was aware (and still others of course, had hidden criminal records)." Similarly, an examination of reports from 270 women in intervention programs ( Fagan et al, 1983 ) found that almost 50% of spouse abusers had been arrested previously for other violent crimes and those who had been arrested for violence against strangers were more frequently and severely violent at home.
Finally, researcher David M. Kennedy noted that, contrary to popularly held views, domestic violence offenders who are known to the criminal justice system, "tend to have robust offending histories and that taken as a whole, the available evidence does suggest that much of the most serious domestic violence is committed by classic high-rate offenders." While the majority of Kennedy's research focused on violent juvenile behavior and related intervention strategies in Boston, he also suggested that similar approaches (for offenders with extensive criminal histories) be developed and applied for other types of crimes, particularly domestic violence ( Kennedy, Pulling Levers: Chronic Offenders, High-Crime Settings, and a Theory of Prevention, 1997 ).
In summary, Kennedy poses a question in his 1997 study, that we feel has great relevance for the evaluation committee, the DVCC and the coordinated response in Omaha/Douglas County: "Why, given the clear self-identification of certain persistently criminal individuals and groups-repeat violent offenders, violent youth gangs, chronic domestic abusers, drug addicts-can we not prevent or control their misbehavior?"
Coordinated Response Process Flow (Chapter III) (This link will open a spreadsheet in MS Excel.)
Bureau of Justice Statistics, U.S. Department of Justice. 1998. Violence by Intimates: Analysis of Data on Crimes by Current or Former Spouses, Boyfriends and Girlfriends. March, 1998.
Buteau, Jacques, Alain Lesage, Margaret Kiely. 1993. "Homicide Followed by Suicide." Canadian Journal of Psychiatry 38: 552-556.
Buzawa, Eve S., Gerald T. Hotaling, Andrew Klein and James Byrne. 1997. Response to Domestic Violence in a Pro-active Court (Quincy, Mass): Summary Findings. Quincy, MA.
Buzawa, Eve S. and Carl G. Buzawa. 1996. Do Arrests and Restraining Orders Work? Washington DC: U.S. Department of Justice, National Institute of Justice.
Campbell, Jacquelyn C. 1986. "Assessment of Risk of Homicide for Battered Women." Advances in Nursing Science 8: 36-51.
Campbell, Jacquelyn C. 1995. Assessing Dangerousness: Violence by Sexual Offenders, Batterers and Child Abusers. Sage. Thousand Oaks, CA.
Easteal, Patricia W. 1993. Killing the Beloved: Homicide Between Adult Sexual Intimates. Canberra: Australian Institute of Criminology.
Fagan, J. D. and K. Jansen. 1983. "Violent Men or Violent Husbands? Background Factors and Situational Correlates." Current Family Violence Research, 49-67.
Harrell, Adel and Barbara Smith. July, 1998. Effects of Restraining Orders on Domestic Violence Victims: Selected Findings and Implications. Washington DC: U.S. Department of Justice, National Institute of Justice.
Hart, Barbara. 1988. "Beyond the Duty to Warn: A Therapist's Duty to Protect Battered Women and Children." In Yllo, K. and M. Bograd (eds.) Feminist Perspectives on Wife Abuse. Newbury Park, CA: Sage.
Johnson, Byron, De Li and Neil Websdale. July, 1998. Florida Mortality Review Project: Executive Summary.M. Legal Interventions in Family Violence: Research Findings and Policy Implications. Washington, DC: U.S. Department of Justice, National Institute of Justice.
Keilitz, Susan, Paula Hannaford and Hillery S. Efkeman. 1998. The Effectiveness of Civil Protection Orders: Selected Findings and Implications. Washington DC: U.S. Department of Justice, National Institute of Justice.
Kennedy, David M. 1997. "Chronic Offenders, High-Crime Settings and a Theory of Prevention." Valparaiso University Law Review, 31: 449-484.
Klein, Andrew R. 1996. "Re-abuse in a Population of Court-Restrained Male Batterers: Why Restraining Orders Don't Work." National Institute of Justice Journal, 3: 52-54.
Klein, Andrew R. 1993. Spousal /Partner Assault: A Protocol for the Sentencing and Supervision of Offenders. Swampscott, MA: Production Specialties.
Stark, Evan, and Anne Flitcraft. 1996. Women at Risk: Domestic Violence and Women's Health. London: Sage.
Trone, Jennifer. 1999. Calculating Intimate Danger: Mosaic and the Emerging Practice of Risk Assessment. New York: Vera Institute of Justice.
[1] The term "coordinated response" in this report refers only to the criminal justice system elements of the response to domestic violence (and not to other functions such as education and public awareness initiatives, provision of shelter, medical and mental health treatment, etc., which are also community "responses" to the problem).
[2] See "The Evaluation of the Coordinated Response to Domestic Violence in Omaha (Phase I)," April, 2000 by R.K. Piper and Thomas J. Martin.
[3] These components include law enforcement, prosecution, victim advocacy, protection orders, probation and an integrated domestic violence tracking system.
[4] The document "Reviewing Domestic Violence Fatalities: Summarizing National Developments, " by Neil Websdale, Maureen Sheeran and Byron Johnson, 1998, summarizes the case reviews, key study findings, policy and program recommendations and contact persons in each jurisdiction. This document and other information related to domestic violence case reviews is available at the U.S. Department of Justice web-site ( http://www.vaw.umn.edu ).
[5] The evaluation committee is composed of representatives/staff of the agencies of the coordinated response as follows: Douglas County Computer Aided Dispatch (911-CAD), Omaha Police Department (Domestic Violence Investigation Squad and Victim Assistance Unit), Douglas County Sheriff's Department, Protection Order Office, Victim Advocates (YWCA, Catholic Charities and DVCC) Douglas County and U.S. Attorney's Office(s) and the Nebraska/Douglas County Probation Department. The committee, the establishment of which was recommended to the DVCC by ISED in the Phase I report, guides and participates in evaluation and research activities.
[6] "High-risk" cases are operationally defined in this study as those domestic violence cases with the greatest likelihood or probability of additional, future domestic violence, which would result in serious injury or death to one or both of the domestic partners.
[7] Information about other key demographic variables, such as age, gender, race, relationship status and geographic location of the victim residence was also collected.
[8] 8. The initial list and operational definitions of indicators relied heavily on the results of the Florida Mortality Review Project, 1994, which generally concurred with the major findings of previous research and case reviews conducted in other jurisdictions in the U.S. ( Websdale, 1998 ).
[9] Serious-injury cases were operationally defined as those cases that resulted in significant physical injury to the victim that could be charged as a 1st or 2nd degree felony crime in Nebraska.
[10] The list of Omaha/Douglas County Domestic Violence Homicides 1997-present included 22 cases. The combined list of Omaha/Douglas County Felony Domestic Violence Assaults (in which the victim sustained serious physical injuries) for the same time-period included 42 cases.
[11] It had been the original intent of the researchers to include 5 homicide, 5 homicide/suicide and 5 serious-injury cases in the sample. However, the sample frame contained only 3 homicide/suicide cases, all of which were included in the sample.
[12] Originally, researchers had intended that after the three test cases were completed, two groups of six cases each beginning with homicide/suicides and progressing to homicides and serious-injury cases, would be reviewed by the evaluation committee. During the review of the first group of six cases, it was learned that one case was not closed, however, and it was dropped and later replaced. This meant that groups of five and then seven cases were reviewed.
[13] Data was not available for indicator 16. Lethality Assessment Scores due to unresolved confidentiality issues surrounding pre-sentence investigations, which were the primary data source.
[14] Two (2) of the 10 major indicators identified in the Florida Mortality Review Project (see Chapter I) were divided into or include sub-categories in our study. ?Prior police calls to the residence? is broken down into ?multiple 911 calls? and ?multiple incidents reported,? and ?history of violent crimes? includes a separate sub-category ?history of domestic violence crimes? in our study. As a result, 12 indicators are identified with an (*) in Table II.2.
[15] As with Lethality Assessment Scores, access to reliable and adequate data sources was an issue for numerous indicators, especially those related to medical or mental health and employment. While police reports, protection order applications and other sources provided some information, existing medical and mental health records were not accessible (as was the case with pre-sentence investigation reports), due to unresolved confidentiality issues.
[16] In addition to extensive, chronic and violent criminal histories for perpetrators, a criminal history (misdemeanor or felony) for victims was also an indicator (?red flag?) in the Florida Mortality Review Study.
[17] Total arrests differs from total arrest charges in that each actual physical arrest may be made on the basis of multiple arrest charges. The actual charges for which an offender/suspect is prosecuted (prosecution charges) may also be different than those for which he/she was arrested.
[18] The totals for arrests does not include the presenting offense for the cases reviewed (i.e., for the homicide or felony assault).
[19] Two examples of such WEB-based management and information systems are currently being developed, tested and implemented in the criminal justice area in Omaha/Douglas County and the State of Nebraska. They are The Nebraska Criminal Justice Information System (N-CJIS) developed by the Nebraska Crime Commission and The Project Impact MIS developed by the City of Omaha/Douglas County through the U.S. Department of Justice, Juvenile Accountability Incentive Block Grant program.
[20] OPD enters data into the system on a daily basis for every domestic violence incident for which a police report was generated. Information includes reports of incidences of domestic violence, as well as victim and associate (offender/ suspect and witness) identification and demographic data. OPD also uses this system to assign cases requiring additional investigation to officers from the Domestic Violence Investigation Squad.
[21] The system remains inaccessible or limited (e.g., only one computer terminal) for some agencies of the coordinated response. When originally established in 1997, some data regarding contacts and services was entered by victim advocates at the YWCA (in a specially devised comments section) but this has essentially stopped due to personnel turnover and training issues. Advocates, prosecutors, and probation consult the database at times, but do not enter data as was originally intended to track cases from initial complaint, through adjudication to final case disposition (see Phase I report).
[22] In such a system, information may often be downloaded on a daily or hourly basis from multiple data bases, depending on the "urgency" of data needs.
[23] In 1999, Omaha was selected as one of four research sites for a national study, funded by the National Institute of Justice, of "victimless prosecution" of domestic violence crimes in the U.S. Researcher Rob Davis informed ISED that they discontinued/abandoned planned research into offender recidivism and victim safety outcomes after preliminary data studies. The data studies revealed that in order to identify domestic violence cases and collect the needed information, researchers would be required to sort through an extraordinary number of paper files, non-computerized official records and multiple data bases within numerous agencies (e.g., police reports, prosecutor files, criminal histories, court records and docket sheets, etc). The complexity of the data collection process in Omaha/Douglas County led them to conclude that such research was not feasible within the "normal" time and resource parameters typically budgeted for such research.
[24] Such databases reflect the "statutory reality" that, in Nebraska, state law does not make a distinction between domestic violence crimes/assaults and any other form of misdemeanor or felony crime/assault. Records of OPD incident reports (Uniform Crime Reports) and arrest bookings at the city and county jails are maintained in the Douglas County mainframe. Supplemental codes have been created to identify "domestic violence" crimes in these databases that, if reliable, would be important in future efforts to develop an integrated domestic violence MIS."
[25] In Phase I, ISED recommended that case reviews be conducted for two reasons: 1) to determine whether criminal justice system processing (from initial complaint to final case disposition) is consistent with procedural standards and protocols that have been established for domestic violence cases, and 2) to gain valuable insights into the capabilities and limitations of current data systems for tracking criminal justice system processes and outcomes.
[26] Indicator 3. Multiple Calls for Assistance to Advocates was defined as two or more calls from any source, including advocate notification by law enforcement. In the 15 cases reviewed, advocates were contacted/notified of the incidents by law enforcement in only these two cases, according to advocate records.
[27] In the Phase I report, we noted that only a small proportion of criminal court domestic cases were prosecuted as felonies (in October, 1999, for example, this amounted to only 15 of 223 [6.7%] cases). Douglas County prosecutors contend that this is largely due to a ?high thresholds? for charging crimes and obtaining convictions for ?felony assaults? under Nebraska State Law. Such statutes provide guidelines to ?categorize? assaults as felonies or misdemeanors, based on factors such as ?permanence? and ?severity,? and are subject to widely-varying judicial interpretation, according to prosecutors.
