## The Ohio Family Violence Prevention Project

Welcome to the online database of the Ohio Family Violence Prevention Project. These data are designed to help local stakeholders and policy makers understand and communicate the scope of family violence in their community. On this site, you can access data on child maltreatment, intimate partner violence and elder abuse for:

• ##### Ohio Alcohol, Drug, and Mental Health (ADAMH) Boards (Multi-County)

We offer two formats for viewing the data. These may be most appropriate for grant writing and public education, but may also be beneficial for program planning and evaluation:

• ##### 1-2 page profiles, in .pdf format that are ready for printing

These data are available for free and are not copyrighted. We only ask that you use the following citation when presenting these findings: Ohio Colleges of Medicine Government Resource Center. The Scope of Family Violence in Ohio: Ohio Family Violence Prevention Project http://grcapps.osu.edu/OFVPP/ ; 2014.

Connect with the Ohio Colleges of Medicine Government Resource Center:

### Some notes about these figures:

Sometimes multiple reports are associated with a single victim.

Many allegations of child maltreatment are unsubstantiated for lack of evidence.

Do not compare figures to data published previously. The methods and definitions have changed.

*Rounded conservative estimate based on the 95% lower confidence limit of survey data.

**Figure only includes shelters in Ohio. Victims may use shelters elsewhere.

### Some notes about these figures:

*Rounded, conservative estimate, based on the 95% lower confidence limit of survey data.

**Figures for domestic violence shelters only include those located in Ohio. Some residents may use shelters elsewhere. Figure may include a small number of victims 65+.

***Includes incidents recorded through the Ohio Incident Based Reporting System.

### Some notes about these figures:

*Rounded, conservative estimate, based on the 95% lower confidence limit of survey data.

**Figure omits self-neglect

***Estimate interpolated from state and national data.

ED = Emergency Department

## The Ohio Family Violence Prevention Project

### Some notes about these figures:

Sometimes multiple reports are associated with a single victim.

Many allegations of child maltreatment are unsubstantiated for lack of evidence.

Do not compare data from county to county, because local agencies often use different procedures for entering data.

Do not compare figures to data published previously. The methods and definitions have changed.

*Rounded conservative estimate based on the 95% lower confidence limit of survey data.

**Figure only includes shelters in this area. ''n/a'' indicates there are no shelters in this area.

### Some notes about these figures:

*Rounded, conservative estimate, based on the 95% lower confidence limit of survey data.

***Includes incidents recorded through the Ohio Incident Based Reporting System. ''n/a'' indicates there was insufficient data to compute a figure for this area.

Do not compare figures to data published previously. The methods and definitions have changed.

### Some notes about these figures:

*Rounded, conservative estimate, based on the 95% lower confidence limit of survey data.

**Figure omits self-neglect. Tallies by report type (e.g., physical abuse) were only available at the state level, so we estimated the number of non-self-neglect reports by applying state-level proportions to county-level total reports. Sometimes multiple reports are associated with a single victim.

***Estimate interpolated from state and national data.

**** ''n/a'' means there are too few seniors to calculate a reliable estimate.

## The Ohio Family Violence Prevention Project

In Ohio, Area Agencies on Aging are regional planning and service areas used for addressing issues related to aging. This page presents profiles for each Area Agency on Aging in our state. For a map defining these areas, please visit: http://www.ohioaging.org/pages/area%20agencies.

### Some notes about these figures:

Sometimes multiple reports are associated with a single victim.

Many allegations of child maltreatment are unsubstantiated for lack of evidence.

Do not compare data from county to county, because local agencies often use different procedures for entering data.

Do not compare figures to data published previously. The methods and definitions have changed.

*Rounded conservative estimate based on the 95% lower confidence limit of survey data.

**Figure only includes shelters in this area. ''n/a'' indicates there are no shelters in this area.

### Some notes about these figures:

*Rounded, conservative estimate, based on the 95% lower confidence limit of survey data.

***Includes incidents recorded through the Ohio Incident Based Reporting System. ''n/a'' indicates there was insufficient data to compute a figure for this area.

Do not compare figures to data published previously. The methods and definitions have changed.

### Some notes about these figures:

*Rounded, conservative estimate, based on the 95% lower confidence limit of survey data.

**Figure omits self-neglect. Tallies by report type (e.g., physical abuse) were only available at the state level, so we estimated the number of non-self-neglect reports by applying state-level proportions to county-level total reports. Sometimes multiple reports are associated with a single victim.

***Estimate interpolated from state and national data

**** ''n/a'' means there are too few seniors to calculate a reliable estimate.

## The Ohio Family Violence Prevention Project

In Ohio, ADAMH Boards are local planning and service areas used for addressing issues related to alcohol, drug addiction, and mental health. This page presents profiles of child maltreatment, intimate partner violence and elder abuse for ADAMH Boards that extend across multiple counties. For a map of Ohio’s ADAMH Boards, please visit: http://www.hrs.org/boards.html.

### Some notes about these figures:

Sometimes multiple reports are associated with a single victim.

Many allegations of child maltreatment are unsubstantiated for lack of evidence.

Do not compare data from county to county, because local agencies often use different procedures for entering data.

Do not compare figures to data published previously. The methods and definitions have changed.

*Rounded conservative estimate based on the 95% lower confidence limit of survey data.

**Figure only includes shelters in this area. ''n/a'' indicates there are no shelters in this area.

### Some notes about these figures:

*Rounded, conservative estimate, based on the 95% lower confidence limit of survey data.

***Includes incidents recorded through the Ohio Incident Based Reporting System. ''n/a'' indicates there was insufficient data to compute a figure for this area.

Do not compare figures to data published previously. The methods and definitions have changed.

### Some notes about these figures:

*Rounded, conservative estimate, based on the 95% lower confidence limit of survey data.

**Figure omits self-neglect. Tallies by report type (e.g., physical abuse) were only available at the state level, so we estimated the number of non-self-neglect reports by applying state-level proportions to county-level total reports. Sometimes multiple reports are associated with a single victim.

***Estimate interpolated from state and national data.

**** ''n/a'' means there are too few seniors to calculate a reliable estimate.

## The Ohio Family Violence Prevention Project

#### What is Family Violence?

Our conceptual definition recognizes that family violence occurs in the context of a trust relationship and involves a pattern of behavior over time. The consequences of family violence are especially harmful and complex due to these characteristics.

The data on the county profiles help describe the scope of common types of family violence in our county in a single year. Specifically, we focus on:

• Child Abuse and Neglect: When a family member or caretaker neglects basic needs or inflicts physical, sexual and/or emotional abuse. Neglect is the most common type of child maltreatment, followed by physical, sexual and emotional abuse.
• Intimate Partner Violence: When physical, sexual, and/or emotional violence occurs in the context of a current or former relationship. The most serious injuries and adverse consequences of intimate partner violence are disproportionately experienced by women. A perpetrator often abuses power in order to control his partner.
• Elder Abuse and Neglect: When a family member or caretaker neglects basic needs, financially exploits an elder, or inflicts physical, sexual, and/or emotional abuse. Neglect is the most common type of elder abuse reported to adult protective services, followed by financial exploitation and then emotional, physical, and sexual abuse. Self-neglect is an important related issue. However, because it does not require interpersonal interaction, it is beyond the scope of our work. Consistent with our focus on elders, we exclude victims under 60 years old.

These three types are not the only important kinds of family violence. We focus on them because of the lack of research on other types (e.g., neglect of non-elderly disabled adults).

#### How Can I Use These Data?

These profiles are designed to help local stakeholders and policy makers understand and communicate the scope of family violence in their community. It is designed to help with grant writing and public education, but may also be appropriate for program planning and evaluation.

Many of our figures are estimates and conservative ones at that (see; 'How Accurate are These Data?'). The following statements are some examples of appropriate ways to quote findings:

• ''...A recent study found that each year in Allen County, 485 children are victims of a substantiated/indicated case of child abuse or neglect.''
• ''...A recent study estimated that each year in Scioto County, at least 530 women (ages 18-64) are physically assaulted by an intimate partner.''
• ''...A recent study estimated that each year in Butler County, at least 2,760 seniors (60+ years old) in the community are abused, neglected or finically exploited, but only 210 reports are filed with adult protective services.''

Data from the county profile may also be useful to highlight specific themes, including:

• Each year, many victims of family violence never come to the attention of authorities. ''According to a recent study, at least ___ elders are abused, neglected, or exploited each year in our county, yet the local adult protective service only received ____ reports.''
• Family violence places a tremendous burden on social service agencies. ''Since 2007-2009, the number of petitions for civil protection orders jumped from __ to __.''
• Certain types of family violence are every bit as common as other, more widely recognized threats to health and well-being. ''One recent study estimated that each year in our county, at least ___ teenage girls experience dating violence, compared to ___ who are injured in motor vehicle crashes.''

#### What Should I Know About the Data?

For several findings, the numbers we present refer to ''reports,'' ''petitions,'' ''incidents,'' or ''cases.'' These totals should not be confused with the number of individuals. For example, a single child who experiences both physical abuse and neglect, would merit two reports to children’s services. In contrast, a single petition for a civil protection order may seek protection for multiple individuals (e.g., a mother and her children) as protected parties. We try to report counts of unduplicated persons whenever possible, yet often these data are unavailable. Whenever possible, we took the annual mean across multiple years (e.g., 2009, 2010, 2011 and 2012) to provide more stable counts.

In some instances, our figures may not match up with numbers from another source. Usually, such discrepancies can be explained by carefully reading the relevant section of this document. If, after reviewing the material, you think you have found an error, please contact the Ohio Family Violence Prevention Project so we can investigate it and make corrections.

The profiles are designed to help policy-makers, professionals and other stakeholders understand and communicate the scope of family violence in their community. They are primarily designed to help with grant writing and public education, but may also be appropriate for program planning or evaluation.

For Medicaid Managed Care Regions, please see below:

Child Maltreatment in West Medicaid MC Region
Intimate Partner Violence in West Medicaid MC Region
Elder Abuse in West Medicaid MC Region
Child Maltreatment in Central/Southeast Medicaid MC Region
Intimate Partner Violence in Central/Southeast Medicaid MC Region
Elder Abuse in Central/Southeast Medicaid MC Region
Child Maltreatment in Northeast Medicaid MC Region
Intimate Partner Violence in Northeast MC Medicaid Region
Elder Abuse in Northeast MC Medicaid Region

#### How Accurate are these Data?

These data represent the best available figures for family violence in Ohio. They are based on a thorough review of current research and have been reviewed by independent researchers and practitioners. In general, they are also internally consistent. Assuming that the most family violence incidents do not come to the attention of authorities, our figures from service agencies (e.g., reports to adult protective services) suggest that our estimates of the true extent of family violence are not wildly off the mark.

Nonetheless, our prevalence estimates are only an approximation. Measuring the true scope of family violence is difficult, as many victims are isolated and may be unwilling or unable to report their experience to trained professionals, let alone to researchers. Similarly, perpetrators have little incentive to report behaviors that are socially undesirable and illegal.

Another challenge is that our conceptual definition of family violence often includes certain patterns of behavior for which data are unavailable or where findings are too disparate to summarize. Most experts agree, for example, that emotional abuse is a serious, common type of intimate partner violence, yet it is difficult to measure on a survey. Moreover, emotional abuse is often not illegal, so police and other social services have limited authority to intervene. As a result, service reports are especially likely to undercount individuals experiencing emotional, but not other types of abuse.

It is difficult to quantify the extent to which these and other phenomena may result in under-counting or over-counting the number of victims. Overall, we believe these estimates are much more likely to under-count than over-count the true number of victims each year.

To account for the uncertainty of survey estimates, we rely on multiple studies whenever possible to ensure that our estimates do not reflect the bias of any one particular study’s methodology. Also, whenever our estimates were based on survey data, we used the 95% lower confidence level estimate. (This is akin to the lower end of the “margin of error” often reported in political polls.) If, for example, a prevalence estimate is 4% ± 1%, we use 3% for calculating the estimated number of victims.

For many of our estimates, we have relied on assumptions that are difficult or impossible to test empirically. Our estimates of teen dating violence, for example, are based on youth enrolled in high school. Thousands of 15-19 year olds across the state do not attend high school, and it is unclear if the prevalence of dating violence in that group resembles that of their peers who are enrolled in school. Our final estimate assumes the prevalence in each group about the same. For each data source, we try to state these assumptions explicitly and describe their potential effect on our estimates. As more information and better research become available, we look forward to updating our estimates.

Finally, we recognize that data based on agency reports inevitably include errors. While we have tried to identify and remedy all such mistakes, readers may find inaccuracies in a county profile. If you think you find an error, please contact the authors so we can investigate and correct it.

#### Can I Compare my County to Other Counties?

For most data sources used in this study, county comparisons are inappropriate. Most apparent county-level differences in the family violence county profiles are due to four related factors:

• Underlying Prevalence -while family violence is a serious concern in all communities, many studies have found that its underlying prevalence can differ by region.
• Demographic Characteristics -regional differences in family violence are often associated with characteristics like the population’s age distribution and poverty level.
• Organizational Capacity -agencies with more staff and better community relations may elicit more reports of abuse and neglect.
• Reporting Procedures -agencies that record every report of family violence may appear to have more family violence than agencies with more selective criteria for recording reports.

Usually readers are tempted to compare counties in order to examine whether (or highlight that) their location has a greater underlying prevalence of family violence. For most data sets, however, differences across counties are more due to organizational capacity and reporting procedures than underlying prevalence. For this reason, we do not recommend individual county-level comparisons for most family violence indicators.

There are, however, two family violence indicators that we believe are appropriate for making some limited county comparisons. New petitions for civil protection orders and reports of abuse, neglect and exploitation in long term care facilities are recorded in a relatively similar manner across Ohio. To limit the degree to which apparent differences are due to counties’ demographic differences, we created four groups of counties for making more appropriate comparisons. These groups are presented in Table 1.

Table 1. Groups of Ohio counties

Major metro core counties include the center cities of Ohio's six largest metropolitan areas: Akron, Cincinnati, Cleveland, Columbus, Dayton, and Toledo. Adjacent metro counties refer to those that are immediately adjacent to the metro core counties and are included in the corresponding metropolitan area. Moreover, they tend to have population growth and/or median household income that are well above average. Appalachian counties include those that are officially classified as such by the Appalachian Regional Commission. Non-metro counties include those that did not meet the criteria of the other groups. They include non-Appalachian rural counties (e.g., Wyandot) as well as those with smaller cities outside of major metropolitan areas (e.g., Allen).

By choosing indicators with similar reporting procedures and then grouping counties to reduce differences in demographic characteristics, county-level variation is more likely due to the two remaining factors: underlying prevalence and organizational capacity. Further research will be necessary to try and disentangle the relative contribution of each of these factors to county-level differences in these indicators.

Our estimates of the underlying prevalence of family violence are mostly based on interpolation from national or statewide data. As such, our estimates are largely based on each county’s population. Within each county, these estimates are useful for highlighting how the scope of family violence compares with other important threats to health and well-being. Between counties, however, they merely reflect differences in each county’s population.

#### Do these Data Prove that Family Violence is Increasing?

No. Within a county, some data sources may suggest family violence is increasing; others suggest it is stable or is even decreasing. Because most family violence indicators are based on agency reports, changes over time are probably mostly due to changes in victims’ ability to access services. Even if we cannot be sure whether the underlying prevalence of family violence is changing, these data can help demonstrate changes in how agencies address the problem.

#### My County's Rate Looks Like it has Changed Over Time. So Why State that the Rates are ''About the Same?''

When comparing quantitative data, apparent differences are often just due to chance. When comparing quantitative data, apparent differences are often unremarkable. Let us say one county had 86 petitions for domestic violence civil protection orders (DVCPOs) in 2011 but only 80 in 2012. Clearly, the number decreased, but is the decrease noteworthy? After all, it is unreasonable to expect that there will be exactly 86 petitions for DVCPOs each year. In this study, we employ a subjective criterion to identify differences that are noteworthy. Based on some statistical assumptions (although not a statistical test), we calculate a confidence interval that is bounded by upper and lower confidence limits. In Adams County from 2010-2012, 370 petitions for civil protection orders (DVCPOs) were filed, or a mean of (370/3=) 123 per year. In addition, using the formulas below, we calculate a lower and an upper confidence limit (rounded to 0 decimal places).

$$[ \# DVCPOs] - 1.96 * \sqrt{ \# DVCPOs} = 370 - 1.96* \sqrt{370} = 111 = \text{lower confidence limit}$$

$$[ \# DVCPOs] + 1.96 * \sqrt{ \# DVCPOs} = 370 + 1.96* \sqrt{370} = 136 = \text{upper confidence limit}$$

So, the lower confidence limit of the annual mean is (332/3=) 111 and the upper confidence limit is (408/3=) 136. We can use these upper and lower limits to calculate rates and assess changes over time or across regions. As a subjective, conservative criterion, we treated as noteworthy differences where the confidence intervals did not overlap.

Rates in less populous counties are often based on few cases, so the confidence intervals are quite large, making it difficult to conclude that the differences are noteworthy. As a result, sometimes a county’s rate may appear to be quite different from a reference rate, but we still conclude that the rates are ''about the same.''