In 1984, the first Children’s Advocacy Center (CAC), the National Children’s Advocacy Center (NCAC), was founded (Faller & Palusci, 2007; Jensen, Jacobson, Unrau, & Robinson, 1996). The original goals of the CAC focused on lessening the re-victimization of children and increasing prosecution rates of child sexual abuse (CSA) cases. The founders of the NCAC also incorporated multidisciplinary teams (MDTs) as a response to what was seen as a decentralized and uncoordinated investigative response to CSA (Faller & Palusci, 2007; Jensen et al., 1996). This vision has since expanded into an international movement.

The Children’s Justice and Assistance Act of 1986 has encouraged state representatives to develop interagency coordination between Child Protective Services (CPS) and law enforcement regarding the investigation of crimes against children (Jensen et al., 1996). The federal government, in collaboration with state governments, has since taken an active role in promoting an MDT approach to investigating child abuse. A majority of states now require law enforcement and CPS to work together as part of an MDT when investigating criminal cases of child abuse (Newman, Dannefelser, & Pendleton, 2005). Additionally, the Children’s Justice Act Grants require states receiving monetary assistance for child abuse to establish a multidisciplinary task force (Lalayants & Epstein, 2005). The way in which states choose to implement these policies varies widely (Cross, Finkelhor, & Ormrod, 2005). The lack of consistency has made systematic evaluation of CACs difficult. However, one relatively consistent theme is the integration of MDTs into local CACs.

CACs have evolved into a cornerstone of CSA investigations. The coordination of an MDT is one of the core components of every CAC (Chandler, 2006; Lalayants & Epstein, 2005). Additionally, CACs seek to foster coordination between law enforcement and CPS to assist in better case outcomes. The nature of child abuse investigations requires multiple agencies to be involved in order to meet the needs of the child and family (Chandler, 2000; Cross et al., 2005; Jones, Cross, Walsh, & Simone, 2005; Lalayants & Epstein, 2005). MDTs provide an opportunity to coordinate multiple agencies, leading to a more effective approach (Bonach, Mabry, & Potts-Henry, 2010; Simone, Cross, Jones, & Walsh, 2005; Walsh, Cross, Jones, & Simone, 2007).

While the majority of research supports the use of MDTs in CSA investigations (Cross et al., 2005; Lalayants & Epstein, 2005) argued that there is no empirical evidence that these teams decrease fragmentation. Since that time some studies, including Wolfteich and Loggins (2007), have given evidence supportive of decreased fragmentation among MDT members. Cases processed through a CAC or Child Protection Team were compared to cases processed using traditional investigative techniques over a 5 year period. They found that the programs with MDTs were associated with higher rates of substantiation when compared to traditional investigative techniques, without MDTs. These findings support the earlier research of Tjaden and Anhalt (1994) who found that those cases with joint investigations involving both law enforcement and CPS had more criminal prosecutions and more guilty pleas than those without joint investigations and Smith, Witte, and Fricker-Elhai (2006) who found that 80% of cases substantiated though an MDT at a CAC were referred for prosecution. However, only 42.8% of substantiated cases not served through an MDT had prosecution referrals.

Faller and Henry (2000) examined 323 criminal records for CSA and found that the MDT approach was successful in facilitating collaboration between CPS and law enforcement. In addition, they asserted that the MDT approach could assist in the investigator’s ability to substantiate reports of CSA; however, Faller and Henry did not examine the extent of MDT collaboration for the records they examined. Additionally, Cross, Jones, Walsh, Simone, and Kolbo (2007) conducted comparison research on four large CACs with comparison jurisdictions without CACs. They found that joint child protection and law enforcement investigations were more common in jurisdictions with CACs.

Not only does the use of an MDT have the potential to improve investigations, it can have positive effects on investigators through improved training, skills, and decreased secondary or vicarious trauma (Lalayants & Epstein, 2005; Sheppard & Zangrillo, 1996). Powell, Wright, and Hughes-Scholes (2011) found that interagency discussions, frequently held at MDT meetings referred to as case staffings, can help identify best practices while assisting in understanding the role of other team members. This supported previous research by Cross et al. (2005) who found that coordination of law enforcement and child protective agencies can decrease friction between the two agencies.

Society has an expectation that child abuse offenders will be prosecuted because children are recognized as among the most innocent and vulnerable in the population. CACs were originally created to assist with the investigation and prosecution of child abuse cases, however, according to Hagborg, Strömwall, and Tidefors (2012) a high prosecution rate is not always desirable, based upon the truthfulness of the allegations and the individual needs of each family. A cost benefit analysis is often calculated in deciding to prosecute (Cross, Walsh, Simone, & Jones, 2003). There are some cases were prosecution may do more harm than good to the child victim. For example, if there is a low chance of conviction because of sparse evidence, a child’s testimony may not result in a conviction. When the offender is the primary provider for the family, the financial cost to the family may also be considered (Walsh, Jones, Cross, & Lippert, 2010).

Researchers have attempted to explain what variables are associated with cases being accepted or rejected for prosecution. However, with the exception of a few key variables, little has been agreed upon. Walsh et al. (2010) found four types of evidence that predicted the likelihood of charges being filed for CSA cases: victim disclosure, corroborating evidence, offender confession, and additional allegations against the offender. Having a corroborating witness was the single piece of evidence that had the greatest impact on the decision to charge. As the types of supporting evidence available increased so did the likelihood of charging the alleged perpetrator. According to Walsh et al. (2010) evidence is often available to support victim statements. This is often cursory information that can lend credibility to the victim while not indicating an actual criminal act. Stroud, Martens, and Barker (2000) compared criminal justice outcomes of CSA cases. They found that 32% of cases dropped involve children 4 years of age and under, indicating young age is a significant impediment to referring cases for prosecution. In a later study by Hagborg et al. (2012) a difference in the quality of interview based upon age was found, however, they did not find that age had a significant effect on the decision to prosecute.

Researchers who attempt to empirically evaluate CACs face a number of challenges. Perhaps the most substantive challenge is obtaining accurate case information from different agencies including law enforcement, CACs, CPS, and prosecution offices. No national clearinghouse exists to house data on CACs. While record retention guidelines do exist, each independent CAC is ultimately responsible for how records are kept.

Although increased prosecution rates were one of the reasons for the development of CACs, whether they do, in fact, affect rates has received limited attention from academics (Walsh, Lippert, Cross, Maurice, & Davison, 2008). There is a noticeable lack of research on how investigative techniques affect prosecution rates. Most of the research conducted on CACs is descriptive in nature or has focused on how the criminal justice and child protective systems can lessen the revictimization of child victims (Wolfteich & Loggins, 2007). The existing research regarding the overall effectiveness of CACs is limited (Cross et al., 2007).

In 2005, Newman, Dannenfelser, and Pendleton found that controlled studies on the effectiveness of MDTs had not been conducted. An impediment found to conducting this type of research was a lack of outcome data measuring effectiveness. Because of this lack of information available and the decentralized nature of CACs little outcome research is available (Jones et al., 2005). This has also led to disparate results for the few studies that have been conducted (Cross et al., 2007; Faller & Palusci, 2007; Jones et al., 2005). Moreover, Cross, Helton, and Chauncy (2012) found that law enforcement participation in child protective service investigations has rarely been studied. Therefore, it is difficult to measure the effectiveness of the MDT approach to child abuse investigations. Additionally, most of the research to date has assumed coordination if a case was processed through a CAC. Research measuring if actual coordination occurred is needed.

Little empirical research exists focusing on the relationship between the CAC and prosecutorial decisions. Descriptive research is the most common type of research conducted involving CACs. Research evaluating CAC effectiveness as an organization is much less common (Wolfteich & Loggins, 2007). While descriptive research is useful, it does little to justify these programs and the amount of federal dollars allocated to fund them. Research examining the role CACs play in prosecutorial decisions will add to the extant research.

This research sought to determine to what extent, if any, does the involvement of the core components of the MDT correlate with the prosecutor’s decision to prosecute. The effects of the involvement of the different MDT members was also examined. Additionally, this research examined to what extent, if any, does case coordination between law enforcement and CPS correlate with a prosecutor’s decision to accept or reject a case of CSA. There is little evidence regarding the utility of programs that foster coordinated investigations from a prosecutorial perspective. The existing research regarding the overall effectiveness of CACs and MDTs is limited (Cross et al., 2007).

Methodology

Texas has the largest network of CACs in the United States. This network includes the Dallas CAC, which has been the subject of several national CAC studies. Most of the evaluation research involving CACs has focused on large metropolitan areas such as Dallas. While this research is useful, it does not represent CACs that serve rural or medium-sized populations, which may differ from large metropolitan areas. For example, many CACs located in large metropolitan areas are frequently collocated with law enforcement, CPS, and prosecutors. Collocation is not possible for most medium-sized or rural CACs because of limited resources. Several different agencies including a unit with the Dallas Police Department, Dallas CPS, prosecutors, and medical staff are collocated with the Dallas CAC. Additionally, the financial and personnel resources available in less populated areas differs significantly from large metropolitan areas. The Dallas CAC serves, on average, over 5000 children yearly and has approximately $20 million in assets. The official area of service for the Dallas CAC is Dallas County, Texas with a population of over 2.5 million. CACs are intentionally decentralized to allow for community specific responses. Therefore, variation between all CACs is expected within the framework of what are considered best practices. It is reasonable to assume that less populated areas will differ significantly from the large CACs used in the existing literature.

Study Site

This research was conducted at one Texas CAC that offers services to multiple counties with an approximate total population of 410,000. These counties vary in size and demographics when compared to previous studies conducted on CACs. The CAC serves approximately 700 children yearly and has an annual revenue of approximately $700,000. Two counties with a total of 12 law enforcement agencies serving populations from under 1000 to over 140,000 are officially serviced by the CAC. However, several smaller agencies in the area, while not officially serviced, do use the services of the CAC. The county considered urban for this study has a population of approximately 335,000 over 1088 square miles. The county considered rural for this study has a population of approximately 75,000 over 1057 square miles. No agencies are collocated with the CAC. Only one of the 12 agencies houses a crimes against children unit. All other agencies house investigators who work a variety of cases including CSA. All jurisdictions served by the CAC use a prosecutor screening process for felony cases. All felony cases are referred to the prosecutor’s office to determine whether charges will be accepted before a warrant is issued. This is in contrast to other jurisdictions where a police officer is able to independently seek a felony warrant from a judge. A total of five prosecutor offices prosecute CSA in the service area. The screening process sometimes varies between a designated prosecutor screening all cases or a rotation of prosecutors.

The CAC employees an executive director, one office manager, two forensic interviewers, two therapists, one MDT coordinator, and one family advocate. The CAC has a medical room onsite where non-acute forensic exams can be conducted. However, the CAC does not employee a forensic nurse, all medical exams are conducted by staff from the local hospital. The CAC only conducts forensic interviews at the request of CPS or law enforcement. Either CPS or law enforcement must be present before a forensic interview is conducted. If both entities are involved it is preferred that both are present at the CAC and observe the interview via closed circuit television.

The CAC also coordinates the MDTs for the different jurisdictions. The MDTs meet at a minimum of once a month. All required representatives of the MDT are invited to attend every meeting. It should be noted that generally one or two representatives from the prosecutor offices are present. However, these prosecutors are usually not the individuals determining whether to accept a case for prosecution. The prosecutors attending the MDT meetings frequently serve as advisors for cases that are still in the investigative stages. Each meeting has an agenda with both new and old cases. All new cases, those that were referred since the last meeting, are actively discussed in detail. All involved parties are called on to give an update to the case. Cases that have been previously staffed are kept on a re-staff agenda and are discussed on an as needed basis until final law enforcement and CPS decisions are available.

Data Collection

All cases referred to the CAC for CSA where both CPS and law enforcement were involved from 2010 to 2013 were considered for inclusion in this study; however, approximately 32% of the cases were included in the final analyses. The researcher found that all cases through 2013 had final CPS and law enforcement dispositions. The researcher was granted access to all available data by the Executive Director of the CAC. Additionally, this study was approved by the Institutional Review Board at Texas State University. All information was deidentified prior to analysis. A report providing the aggregate number of cases from 2010 to 2013 indicated 1732 forensic interviews were conducted during this time with both CPS and law enforcement involvement. It should be noted that not all cases processed through a CAC have both CPS and law enforcement involvement. This study was limited to those cases that had both CPS and law enforcement involvement. The CAC does provide services for other alleged offenses as well as witnesses to violent crimes. These cases were excluded. The CAC provides what are referred to as “courtesy interviews” which were also excluded. These forensic interviews are conducted with children who are in the CAC service area, but whose alleged offense occurred elsewhere. These cases are forwarded to the appropriate jurisdiction and not followed by the interviewing CAC. Particularly, cases involving allegations of physical abuse and neglect were excluded. While these cases do warrant an MDT response, the investigative nature is inherently different than CSA. Approximately 75% of the cases processed through the study site are related to sexual abuse allegations. Physical evidence, the need for a forensic interview, family cooperation, and prosecution acceptance rates differ for physical abuse when compared to sexual abuse. Additionally, multi-victim cases were excluded prior to the analysis. This is consistent with prior studies (Cross et al., 2005; Cross, Whitcomb, & De Vos, 1995; Hagborg et al., 2012; Walsh et al., 2010). Multi-victim cases will inherently violate the regression assumption that all cases are independent. Multi-victim cases are defined as cases involving multiple known victims with one alleged perpetrator. These cases are frequently more complex, contain more evidence, and are more likely to be prosecuted than single victim cases. While multi-victim cases, along with physical abuse cases, warrant an MDT response, they, like physical abuse cases, differ significantly from most CSA cases. These cases were removed prior to the analysis. An unexpected category of cases that were removed prior to analysis were those cases that were never presented to the prosecution because of the decision of the law enforcement agency charged with investigating the case. The reasons these cases were never presented for prosecution is beyond the scope of this research. A total of 1174 cases were excluded from analysis. Cases that were never presented for prosecution or originated in a different jurisdiction accounted for 612 of the excluded cases. Multi-victim cases, child witness interviews, and physical abuse interviews accounted for 512 of the excluded cases. Cases with adult victims and alleged perpetrators under the age of 10 accounted for 33 of the excluded cases. There were a total of 17 incomplete cases. These cases were also excluded from the analysis because the files were largely incomplete. A total of 558 cases were included in the analyses.

There are a number of data sources for this project. While all data are housed at the CAC, there is no standard system for tracking all needed data. The computer system established by CACs of Texas, Case Tracking, was used for some of the data. However, not all data are captured in that system. Case specific information was collected by the researcher from individual case files and minutes from MDT meetings. Only the initial staffings were considered for inclusion in this study. The initial staffing represents the most likely time that cases are discussed in detail. A quasi-case staffing is conducted at the time of the initial forensic interview. The forensic interviewer is briefed about the case prior to the interview. At the conclusion of the interview a discussion ensues involving the forensic interviewer, CPS, and law enforcement.

Analysis

Every MDT has core components consisting of law enforcement, CPS, prosecution, mental health, medical, family advocacy, and CAC staff. The researcher first addressed the question: to what extent, if any, does the presence of the core components of the MDT at the meetings correlate with the prosecutor’s decision to accept a case? This question yielded two logistic models. The first model addressed whether the involvement of individual MDT members at case staffings is correlated with the likelihood of prosecution. Each component was examined individually. The second model addressed whether having more MDT disciplines present at meetings is correlated with the likelihood of prosecution.

The researcher then addressed the question: To what extent, if any, does case coordination between law enforcement and CPS correlate with the prosecutor’s decision to accept or reject the case? The MDT meeting is the only official time and place that all members of the MDT have the opportunity to collaborate on CSA investigations. All parties, not just investigators attend the meetings. Conversely, the investigators involved in the case have the opportunity to collaborate at both the forensic interview and the MDT meeting. Case coordination was measured at two different points in time: the initial forensic interview and the subsequent case staffing or MDT meeting. This was measured by whether CPS was present at the forensic interview, whether law enforcement was present at the forensic interview, whether CPS was present at the initial case staffing, and whether law enforcement was present at the initial case staffing. If both parties were present at both the initial staffing and the forensic interview the case was labeled as “full coordination.” If both parties were present at either the forensic interview or the initial case staffing the case was labeled as “partial coordination.” If both parties were not present at either the initial case staffing or the forensic interview the case was labeled as “no coordination.”

The research questions yielded three models evaluating individual MDT participation, aggregate MDT participation, and coordination between CPS and law enforcement. All three models included the following control variables: victim age, victim race (white, non-white), alleged perpetrator race (white, non-white), alleged perpetrator age, alleged perpetrator relationship to victim (relative, non-relative), whether the alleged perpetrator lived in the same home as the victim (in home, not in home), and the size of the county of the alleged offense (urban, rural). Stranger was initially included as a relationship variable. However, the stranger cases (n = 5) were excluded due to lack of variability. Please see below for a detailed explanation of the decision to exclude the five stranger cases. The following case specific control variables were used in all equations: whether there was a reported prior outcry, whether there was a child witness, and the disposition of CPS regarding the case. A reported prior outcry was defined as the child making an intentional outcry to an individual prior to the forensic interview. This differs from cases where children are considered at risk for concerning sexual behaviors or a child making an “accidental disclosure” such as a family member finding a diary entry describing abuse. CPS has three different dispositions in which to close a case: reason to believe (RTB), unable to determine (UTD), or ruled out (RO). RTB indicates that, based on the preponderance of the evidence standard, the abuse occurred. UTD indicates that the investigation did not adequately prove that the abuse did or did not occur. RO indicates that, based on the preponderance of the evidence standard, the abuse did not occur.

For the alleged perpetrator relationship to victim variable, the category of “stranger” posed several unexpected problems during analysis. Only five of the 558 cases had a relationship of “stranger”. Additionally, all of the five cases were accepted for prosecution and all had sexual abuse outcries at both the forensic interview and SANE. Because there was no variability among the “stranger” cases the decision was made to run all analyses twice, both with and without the five “stranger” cases. The pattern of findings were consistent across both analyses for all research questions for all variables other than alleged perpetrator relationship to victim. For alleged perpetrator relationship, removing “stranger” allowed the researcher to accurately analyze the significance of familial relationship. To show this familiar relationship, reported analyses exclude the five “stranger” cases. The final number of cases included in the study was 553.

Results

The first model, illustrated in Table 1, examined whether having more MDT members at case staffings was correlated with an increase in the odds that the prosecutor would accept the case. The second model, illustrated in Table 2, examined how these odds changed based on the individual effects of each MDT member at case staffings. In the first model, the Hosmer and Lemeshow’s goodness of fit test was significantly improved over the control models by adding the number of MDT participants. Additionally, the chi square value increased by 9.173 in this model and it is statistically significant indicating that this model is a better fit to the data than the models without this predictor variable. The odds ratio for the MDT Amount variable is 1.305 and is statistically significant, that there is a difference in cases that are accepted for prosecution and cases that are rejected in reference to this variable.

Table 1 Number of MDT participants
Full size table
Table 2 Individual MDT member contribution
Full size table

In the second model that addressed the individual MDT members, when all seven MDT members were included as variables, the model was not statistically significant. It was determined that the Forensic Interviewer, Therapist, and Family Advocate were present for almost all of the case staffings. Because there was little variability regarding the presence of these three MDT members, another model was created omitting these member variables. When these variables were omitted, the model became statistically significant, illustrated in Table 2. The Hosmer and Lemeshow’s goodness of fit test is, again, significantly improved with this model. Additionally, the chi square value increased by 13.051 for this model and is statistically indicating that this model is a better fit to the data than the models without these predictor variables.

Of all the possible MDT members, the only member that had a significant odds ratio was the prosecutor at 1.80 indicating that prosecutor presence was correlated with an increase of an 80% likelihood that a case was accepted for prosecution.

The second model compared no coordination to partial and full coordination with CPS and law enforcement. Table 3 contains the results of this analysis. The Hosmer and Lemeshow’s goodness of fit test indicates that this model is a better fit to the data than the model without the predictor variables. However, the Chi square value is 2.893 and is not statistically significant. Therefore, including the amount of official CPS and law enforcement coordination in the model did not improve the ability to predict whether or not the prosecutor would accept the case.

Table 3 CPS law enforcement coordination
Full size table

Discussion

This research revealed an increase in MDT member involvement was correlated with an increase in the likelihood of prosecution. Conversely, when the MDT members were examined separately only one out of the seven members yielded a statistically significant correlation with the increase in odds of having the case accepted for prosecution. When all seven disciplines that comprise the MDT (law enforcement, prosecution, CPS, forensic interviewing, victim advocate, therapist, and medical staff) were assessed independently in the model, the prosecutor was the only member of the core make-up of the MDT that was correlated with increased odds of the case being accepted for prosecution. Prosecutorial involvement in cases appears crucial to the acceptance of CSA cases. While it could be argued that this finding appears to negate the objectivity that CACs claim, it can also be argued that this finding allows prosecutors the opportunity to ensure that necessary steps in the investigative process are taken.

While the models examining the MDT approach indicate only one out of seven were significant, they also indicated that the more MDT members that are present, the more likely a case is to be accepted for prosecution. These findings do support previous studies (Bonach et al., 2010; Simone et al., 2005; Walsh et al., 2007), bolstering the widely held belief that the team approach to investigating child abuse leads to different outcomes than the traditional methods of investigation that include little to no coordination. Additionally, while the individual prosecutor attending the MDT meetings was usually not the decision maker, their expertise and contribution may play a factor in the strength of an investigation. This possibility is supported by the earlier research of Tjaden and Anhalt (1994) and Wolfteich and Loggins (2007) who found that a collaborative investigation leads to greater substantiation rates.

Another widely held belief is that a joint investigation between law enforcement and CPS yields better case outcomes. The phrase “better case outcomes” is ambiguous, yet usually refers to an increase in prosecution or substantiation rates (Bonach et al., 2010; Chandler, 2000; Faller & Palusci, 2007). This research aimed to help determine whether case coordination between different agencies can alter prosecutorial decisions without the assertion that higher prosecution rates is the equivalent of “better case outcomes”. This belief was not supported by the research, as degree of case coordination between the two agencies was not a statistically significant predictor. It must be noted that it is impossible to gather a complete picture of case coordination without examining case reports. This research solely relied on coordination at the only two official points in time law enforcement and CPS staff cases. Additionally, it is possible that individuals were present but did not actively participate. An attempt to negate this limitation was made by only examining the initial case staffing, where members are actively called on to discuss the investigation. Though the MDT is the only official case coordination strategy, it is reasonable to believe that additional, undocumented coordination does occur. This unofficial coordination may impact case outcomes.

In all models the CPS disposition was statistically significant. CPS cases that are ruled RTB have a greater likelihood of prosecution than cases ruled UTD or RO. This is not surprising because the ruling of CPS can lend to the strength of the case as a whole. While CPS has a lower threshold of proof, their findings can be indicative of strength of the entire case. This research produced seemingly contrary results than those produced by Tjaden and Theonnes (1992) and Cross et al. (2003) who assert that CPS dispositions do not appear to have an effect on the prosecution of child abuse cases. Both prior studies examined CPS cases in general, not just cases that were processed through a CAC. The results of this study support the view that CPS decisions may have an impact on prosecutorial decisions of cases of CSA. These findings underscore the importance of ensuring that CPS investigators are competent and properly trained.

While states differ in CPS employment requirements, elements of the social work field are found in most CPS job descriptions. Additionally, social worker education and training vary widely between states. Social workers, both those employed by CPS and those providing mental health services, are an integral part of the CAC model. The decentralized nature of social worker licensure and CPS employment complicates the evaluation of licensed social workers regarding MDTs. However, the importance of education and on the job training cannot be overemphasized. Social workers can provide expert testimony and are expected to conduct competent, thorough investigations regarding complicated CSA cases. A primary goal of CACs is to foster communication and cooperation between agencies (Faller & Palusci, 2007; Jensen et al., 1996). Social workers should use this platform to share their expertise with others and strengthen their own MDTs. Every professional in an MDT has specific knowledge related to his or her area of expertise (Powell et al., 2011). However, social workers are in a unique position in that they have formal training and education regarding collaboration. Social workers should be given the opportunity to assist in MDT formation and provide assistance in collaboration efforts. Providing social workers this platform can help inform decisions, create MDTs where none exist, and strengthen existing MDTs.

Limitations

A limitation of this research was the relatively small sample size (n = 553). While the roughly 30% of cases examined provide a picture of the data as a whole, its interpretation and generalizability are limited. While the majority of cases removed from the analysis were expected, 245 cases were removed because the case was never presented to the prosecutor. Had those cases been presented to the prosecutor the analyses would have included roughly 46% of the cases. It is likely that the cases not presented to the prosecutor were significantly different from those presented. For example, descriptive statistics revealed that only 57.9% of the cases with female perpetrators were ever presented to the prosecutor to review. Additionally, only 48.9% of the cases with no prior outcry were presented to the prosecutor to review. Future research is needed to examine the qualitative differences between the cases presented to prosecutors and those closed by law enforcement agencies before a prosecutor examines them. This limitation is similar to that observed in other studies involving CACs and MDTs. The number of cases varies widely between studies. For example, studies using the National Survey of Child and Adolescent Well-Being include over 5000 cases (Cross et al., 2005, 2012). However, most studies examining individual files have fewer than 500 total cases (Faller & Henry, 2000; Jensen et al., 1996; Smith et al., 2006; Walsh et al., 2010; Wolfteich & Loggins, 2007). The need for research comparing CACs and prosecution rates, with larger samples, from differing regions is apparent.

While this research helps in determining if CAC factors predict prosecutorial decisions, it did not consider most qualitative differences between cases such as other evidence or confessions, which were unavailable for this research. Additionally, there are many factors that affect prosecutorial decision making that are beyond the scope of this research including budgetary constraints, political variables, quality of investigation, and caregiver cooperation. This research examined cases organizationally, not qualitatively. Given the data the researcher had access to it is likely that the case specific control variables accounted for some of the same variance in prosecutorial decisions that other case specific information would have accounted for, such as the presence of offender confessions; however, without that information, it is impossible to be certain. Additional qualitative research is needed to further explain what specific factors predict prosecutorial decisions.

While CPS dispositions were included in this research, additional CPS involvement is beyond the scope of this research. It is important to realize that although prosecution may not occur, CPS intervention may still proceed. CPS intervention can include offering a family services. Additionally, if warranted CPS can also initiate civil proceedings that can result in the child being removed and eventual termination of parental rights. Additional research is needed to further examine CPS involvement in cases and how that involvement may affect prosecutorial decisions.

In addition to the limitations for secondary research, the data used for this research only represents cases processed through one CAC. Therefore, the findings may not be generalizable across CACs; however, any findings can be compared to other research in the area of CAC utility and prosecution of CSA.

Conclusion

The multidisciplinary team approach to child abuse is central to the development and purpose of CACs. The belief that the MDT approach is more effective than traditional approaches is widely held among child abuse professionals. However, the MDT approach was a common sense solution to a perceived problem in CSA investigations. A dearth of empirical research still exists more than 30 years after the creation of the first CAC. The research, to date, examining MDTs has assumed coordination between law enforcement and CPS if a case was processed through a CAC. Additionally, the existing research has considered higher prosecution or substantiation rates as evidence of MDT success. Depending on a number of factors previously discussed an increase in these rates should not, alone, indicate success.

This research provides a starting point to fill the dearth of information in reference to the effectiveness of MDTs. While the MDT philosophy encompasses different goals at different CACs, the overarching theme is that a coordinated response to child abuse cases is better than one that is uncoordinated. If this assertion is correct, one would expect to see a statistically significant difference in cases that are processed with an MDT and those that are not when other crucial factors are taken into account. Only one member, prosecution, was individually statistically significant; however, when looked at as a whole, the MDT approach does appear to produce different results. It is possible that the team approach is what most relates to prosecutorial decisions, more so than any individual member.

Because CACs are now common place in child abuse cases, it is imperative to empirically examine the core components of this international movement. Evaluating CACs is inherently difficult because of the decentralized nature of the CAC structure. The research to date has focused on major metropolitan areas, which by the nature of CACs, is not representative of less populated areas. Additionally, this research should serve as a beginning in examining the utility of CACs and the MDTs they facilitate. Future research should include unofficial coordination and the individual differences in MDT members. Learning how MDTs function is the first step in implementing evidence based procedures for facilitation.