RD161 - Evaluation of the Child Protective Services Differential Response System - December 7, 2004
Executive Summary: As directed by House Bill 1360 (2000), the Department of Social Services (Department) implemented a Child Protective Services Differential Response System (DRS) on May 1, 2002. The Department also was directed to evaluate and report on DRS by submitting annual reports to the House Committee on Health, Welfare and Institutions and the Senate Committee on Rehabilitation and Social Services. The Department has entered into an interagency agreement with Virginia Tech to assist in evaluation of the Differential Response System. This is the fifth annual report on the status of the Department’s implementation of DRS. The Differential Response System provides two different response options to reports of suspected child abuse and neglect. 1. The Investigation response track is the traditional Child Protective Services (CPS) response. If the local agency determines that abuse or neglect did occur, a disposition of “founded” is made, and the name(s) of the caretaker(s) responsible for the abuse or neglect is placed in the state’s Central Registry. Local departments offer services, when needed, to reduce the risk of abuse or neglect. 2. The Family Assessment response track is for valid CPS reports where there is no allegation that is required to be investigated or immediate concern for child safety. A family assessment identifies family strengths and service needs. Local departments offer services, when needed, to reduce the risk of abuse or neglect. No disposition is made and no names are entered into the Central Registry. Virginia’s Online Automated Services Information System (OASIS) is a primary source of data for the evaluation. Most data in this report are from referrals accepted by local agencies from January through December 2003. For some variables, DRS is compared to a baseline period from fiscal year 2000 to fiscal year 2002. Data from the Department’s CPS Referrals and Findings Reports (based on OASIS) are used for those comparisons. The other primary data source is a review of 446 referrals conducted by a very experienced, retired CPS supervisor. The case reviews focus on families with more than one referral during 2003. Families with a referral in January and a subsequent referral any time during the remainder of the year were included in the review. Outcomes from Analysis of OASIS Data Last year’s evaluation report to the 2004 General Assembly presented some preliminary outcomes based on six months of data for the period July through December 2002, just after the introduction of DRS across the state. This year’s report is based on a full twelve month period from January through December 2003. The outcomes reported here can be viewed as representative of DRS outcomes that are likely to be typical over the next few years. Outcome analyses are based on 28,987 valid referrals for suspected abuse and neglect received from January through December 2003. Track Assignment DRS offers local agencies the flexibility of assigning most referrals to either the investigation or the assessment track. Local agencies are using that flexibility. From January through December 2003, 61 percent of referrals were assigned to the assessment track. Track assignment varied among the three Department Service Areas. (*1) The majority of referrals in the Northern (68 percent) and Western (66 percent) Service Areas were placed in the assessment track. In the Eastern Service Area, fewer than half (46 percent) were assigned to that track. Local agencies varied in their use of the assessment track. Many placed more than 60 percent of their referrals in the assessment track, but others used assessments for 20 percent or fewer of their referrals. A number of factors can influence track assignment. When investigation is not mandated, the choice of the family assessment track is predicated on the ability of the agency to work with the family and community service providers to develop strategies to prevent abuse or neglect and provide services if needed. If the information from the person making the complaint suggests that the agency will need to take a more authoritative approach with the family, the complaint should be placed in the investigation track. In addition, a local agency may investigate any referral. There are no circumstances under which an assessment is mandated. With the exception of allegations of sexual abuse which must be investigated, the two tracks are quite similar in the kinds of abuse or neglect assigned to them. In both tracks, neglect (*2) was the most frequent allegation. Forty-five percent of the investigations and 60 percent of the assessments had allegations of neglect. The second most frequent allegation was physical abuse, found in 37 percent of the referrals in each track. Small percentages of both investigations and assessments involved medical neglect or emotional abuse. The majority of referrals with each type of alleged abuse or neglect, other than sexual abuse, were placed in the assessment track. When more than one type of abuse/neglect was alleged, use of the investigation track increased, from 38 percent in referrals with one kind of abuse or neglect to 59 percent in referrals with three or more kinds. A referral that is initially treated as a family assessment may be changed to an investigation if the local agency discovers a serious safety issue or circumstances that mandate investigation. Two percent of referrals originally put in the family assessment track were later changed to investigations. This low rate of reassignment suggests that errors in track assignment are rare. A review of 2002 cases that had been reassigned showed that the reassignments were appropriate and generally resulted from new information discovered by the local agency. The addition of the family assessment track naturally meant that there were fewer investigations under DRS than in the preceding years. There were 27,795 investigations in State Fiscal Year (SFY) 2000 and 25,570 in SFY 2001, the last two years before DRS implementation. In SFY2004, under DRS, there were 12,506. While the number of investigations and the number of founded investigations decreased, the percentage of investigations that were founded increased from 23 percent during the two baseline years to 36 percent in SFY04. Services CPS workers identified 32 percent of families in the assessment track as having one or more service needs. (*3) The percentage of families needing services varied depending on the type of abuse or neglect. Service needs were most often identified in cases involving emotional abuse (44 percent), followed by physical abuse (39 percent), medical neglect (35 percent), and physical neglect (28 percent). In terms of the risk assessment made at the conclusion of the family assessment, 92 percent of high risk and 82 percent of moderate risk families were determined to have service needs. The three most frequently needed services were counseling, parent education, and substance abuse evaluation or treatment. CPS workers enter the status of service receipt at the time they complete data entry for an assessment. At that point, 26 percent of all families in the assessment track either had received or were expected to receive one or more services. Among those needing services, 81 percent received or were expected to receive services. Fourteen percent of families declined at least one service, and six percent needed at least one service that was not available. Among families receiving or expected to receive services, community resources provided 39 percent of the services; DSS provided or purchased 33 percent of services; and the families obtained 28 percent of the services on their own. Sometimes the local agency asks the Juvenile and Domestic Relations Court to order the family to accept a service. Court orders can be sought in both assessments and investigations. Two percent of all families in the assessment track and six percent of families with service needs were the subject of a court order to accept services. The most frequent court-ordered services were counseling, substance abuse evaluation or treatment, and parent education. Comparable data are not available for investigations. If a child is placed in foster care, or if the agency determines that the family needs child protective services beyond the 45 to 60 day assessment or investigation period, the agency opens a foster care or “ongoing CPS” case. Seventeen percent of all referrals resulted in ongoing CPS or foster care services. Cases opened for further services varied by disposition: founded investigations, 59 percent; assessments, 12 percent; and unfounded investigations, 9 percent. Three percent of all CPS referrals in 2003 led to a child being placed in foster care. As would be expected, founded investigations had the highest foster care rate. Twelve percent of founded investigations resulted in a foster care placement. Children in one percent of unfounded investigations and two percent of assessments were also placed in foster care. Since foster care data include any placements within 90 days of the disposition, these placements may have occurred after completion of the unfounded investigation or assessment. Case Reviews Case reviews are helpful in understanding the operations of DRS because there are many details of the case that are not captured by the OASIS data used for the statistical analyses presented in this report. The case reviewer could see other OASIS screens that provided a fuller picture of the issues present in the referral and how the agency responded. The reviewer also applied her judgment, as a highly experienced CPS supervisor, to several issues that provide additional insight into local agency practices and performance. The case reviews focus on families with more than one CPS referral in 2003. The reason for this selection was that these were the families most likely to have a substantial abuse or neglect problem and it would be possible to see how their CPS referrals were handled. Specifically, the reviews covered 194 families with an initial referral in January 2003 and a subsequent referral any time during the remainder of the year, a total of 446 referrals. One hundred fifty families had two referrals, 31 had three, and 13 had four or more. The reviewer encountered some problems with incomplete documentation of referrals in OASIS. In over a third of the referrals, inadequate documentation was an issue. This problem should improve with the changes in OASIS documentation introduced in July 2004. The pattern of track assignments in the cases reviewed was similar to that found in all referrals in 2003. The reviewer, however, was able to view the details of the referral and determine the basis for the track assignments. Thirty-nine percent of investigations were mandated by code or CPS policy and 61 percent were a matter of agency judgment. All assessments are ultimately matters of agency judgment since no referrals are mandated for family assessment, but the reviewer found that in 53 percent of assessment track referrals the child was clearly safe and there was no reason to consider an investigation, while the remainder of assessment track referrals were based on agency judgment about the best track to choose for the situation. The reviewer disagreed with track assignment in 22 percent of the referrals. The primary reason for disagreement was that the reviewer believed that a complaint that was investigated should have been treated as a family assessment. In those cases, she believed the local agency had an opportunity to handle a referral in a way that might have led to a better outcome for the family. In particular, in many unfounded investigations, it seemed that even though abuse or neglect were not substantiated, the family could have benefited from services, and the less threatening family assessment approach might have served the child and family better. Case review data suggest a problem in some agencies in ensuring a timely response to reports of abuse or neglect. In 18 percent of the referrals reviewed, the first meaningful contact took place more than five days after receiving the referral, with no evidence of any earlier attempt to contact the family. Available data suggested that this problem may be concentrated in specific agencies. Forty-three percent of all the reviewed referrals were investigations and 57 percent were assessments. The percent of investigations that were founded grew from 36 percent in the families’ first referrals, to 51 percent in the second, and 57 percent in third referrals. Forty-two percent of assessments and 43 percent of investigations had a service plan or other evidence that the worker had tried to assure services. Services plans were naturally most frequent in founded investigations (72 percent) and assessments with identified child abuse or neglect needs (83 percent). One of the purposes of DRS, and particularly the family assessment track, was to try to engage families in a less threatening way and to involve them in identifying needs and planning for services to meet those needs. In referrals with services plans, families were included in planning for services in 95 percent of assessments and 67 percent of investigations. (It is possible that a higher proportion of families in investigations may have been included in service planning. Because OASIS did not have the same service screens for investigations as for family assessments conducted during this time period, sometimes the case reviewer could not tell whether the family was included.) The reviewer tried to determine whether families with service needs received services. Data for the families’ first and second referrals show that services were definitely provided for half to two-thirds of families in founded investigations or family assessments with identified child abuse or neglect needs. Other families may also have received services, but the documentation was unclear. In both the assessments and the investigations, the percent definitely receiving services was higher in the second referral than in the first, suggesting that a repeat complaint either makes more clear the need for services or results in a greater attempt to provide services. Families in assessment cases with identified needs related to child abuse or neglect received services as frequently, or more frequently, than families in founded investigations. These data may provide evidence that the less threatening assessment approach is successfully engaging families and meeting their service needs. Ninety-three of the 194 families (48 percent) whose cases were reviewed received at least one service. Twenty-nine percent of assessments with identified abuse or neglect needs and 28 percent of founded investigations had cases opened for CPS services. The case reviewer also found that, in her judgment, 31 families (16 percent) had a least one service need that the worker did not identify and for which no services were received. Counseling and substance abuse evaluation were the two most frequent unidentified service needs. The reviewer looked at whether the referrals had any court orders associated with them. Court orders were issued most often in founded investigations but also in unfounded investigations and assessments. Court orders most frequently involved a petition for removal of a child from the home (37 percent), followed by criminal proceedings (22 percent), and a protective order for abuse or neglect (14 percent). The reviewer found evidence of possible substance abuse in 43 percent of the families. In 42 percent of referrals with a possible substance abuse issue, there was clear evidence that the problem was addressed by the worker. Twenty percent of the families had evidence of domestic abuse. That problem was clearly addressed 31 percent of the time. Conclusion DRS outcomes reported this year are generally similar to those reported last year. Local agencies are placing about sixty percent of referrals in the family assessment track, but there is great variation among agencies in track assignment. About a third of families have identified services needs and most of them appear to have received at least some services. The most frequent reason for a family not receiving needed services is that they declined to do so. The case reviews focused on families with more than one referral in 2003. The reviewer found some problems such as gaps in documentation, occasional failure to respond in a timely manner, and sometimes a failure to identify service needs. In founded investigations and assessments with identified abuse or neglect needs, there was usually a service plan or other evidence of the worker’s attempt to provide services. Families were usually included in service planning, particularly in assessments. About half the families received services – a higher proportion than for all families with referrals in 2003. As the number of referrals for a family increased there was increased use of the investigation track, a higher percentage of founded investigations, and more frequent provision of services. Outcome of Recommendations from the 2003 DRS Evaluation Several recommendations were made last year based on the evaluation report and other information. The recommendations and action taken in the past year are reported here. 1. The Department should review a random sample of cases to more accurately determine whether track assignments are being made in a consistent manner, whether risk assessments and safety plans are identifying the most relevant risk factors, and whether needed services to prevent maltreatment are being provided. Rather than reviewing a random sample of all referrals, the Department reviewed records of families with more than one referral for suspected abuse and neglect. These records were chosen because these families are more likely to have a substantial problem with abuse or neglect, and it would be possible to see how their CPS referrals were handled. The findings of the case review are summarized in this report. 2. The Department should require all CPS staff to attend the “Engaging Families” course, currently under development by VISSTA, and ensure that the course is readily available statewide. The “Engaging Families” course was developed by the Virginia Institute for Social Services Training Activities, Virginia Commonwealth University (VISSTA/VCU) and published in August 2004. Course attendance has been defined as a benchmark for measuring progress in the Department’s Program Improvement Plan to address deficits identified through the Child and Family Services Review. 3. The Department should implement a structured decision making model in order to improve the decision making skills of staff and to obtain more consistency and uniformity in safety and risk assessments made by social workers across the state. The Department contracted with Children’s Resource Center to develop and implement a structured decision making model for Virginia. In November 2004, 30 localities began piloting policies and automated tools for the model. Policy groups began meeting to discuss possible changes to policy and procedure for statewide implementation. Subcommittees are working on evaluation, training, automation, and outreach so that the findings of the pilots can be applied to statewide implementation. 4. The Department should give recognition to those local departments of social services that have implemented a community-based prevention and protection system and publish this information so that other localities may replicate their efforts. Regional forums with the theme of “Protecting the Generations: The Community Challenge of Family Violence” presented an overview of the broad scope of family violence, facilitated communications, and promoted collaboration across service providing agencies. Participants shared effective community collaboration models, identified barriers to service, and shared successful collaborative strategies. The forums, held in Fredericksburg, Roanoke, and Chesapeake in the fall of 2003, were the cooperative effort of the following partner agencies: Office of the Attorney General; Virginians Against Domestic Violence; Virginians Aligned Against Sexual Assault; Virginia Coalition for the Prevention of Elder Abuse; Virginia Departments of Aging, Criminal Justice Services, Health, Housing and Community Development, Mental Health Mental Retardation and Substance Abuse Services, and Social Services; and VISSTA/VCU. While this clearly remains a very important issue for the success of the CPS differential response system in Virginia, much of the focus on community collaboration was diverted to other child welfare-related issues this past year. These include implementation of the structured decision making pilot and development of the Program Improvement Plan. DRS Recommendations for 2005 • The Department should work toward more consistency in decision-making for assigning reports of suspected child abuse and neglect to the family assessment track. • The Department should provide consultation and technical assistance to local agencies who are not responding to reports of suspected abuse and neglect in a timely manner. • The Department should reinforce the importance of documenting service needs and service provision now that the OASIS has been enhanced to record these functions in a more consistent manner for investigations and on-going services as well as family assessments. • Local departments of social services that are piloting the structured decision making model should educate community stakeholders about using level of risk to establish priorities for providing services. • The Department should use "A Blue Ribbon Plan to Prevent Child Abuse and Neglect in Virginia", a strategic plan being developed by a broad-based steering committee under the leadership of the Department, as a tool to enhance community collaboration with local agencies in implementing DRS. __________________________________ (*1) A list of local agencies by Service Area can be found in Appendix B. (*2) Neglect includes both deprivation of necessities, such as inadequate food, clothing, shelter, or hygiene and inadequate supervision. (*3) OASIS did not provide parallel data on service needs of families in the investigation track during the period covered by these data. |