Share This Article
An outside review of 327 Arkansas child abuse investigations found children who were never seen by investigators, safety assessments that were never completed and cases closed against the state’s own recommendation, according to documents obtained by the Mountain Home Observer.
The review, conducted in November 2025 by the nonprofit Evident Change at the request of the Arkansas Division of Children and Family Services, examined investigations involving families that had been reported to the state’s child abuse hotline five or more times over the past five years without ever receiving ongoing services. Every one of the 327 cases had been closed.
The findings paint a picture of an agency where investigators routinely failed to follow basic policy requirements. In 27% of cases, investigators did not see the alleged child victim within the required time frame. In 18% of those delayed cases, investigators never made face-to-face contact with the child at all. Eleven percent of investigations had no documented safety assessment of any kind, and 72% showed no use of the agency’s own practice model for engaging families.
The report was completed in January 2026 but has received limited public attention. DCFS Director Tiffany Wright discussed portions of it in a radio interview on KTLO Monday morning, focusing on system improvements and workforce challenges. She did not address the report’s most significant findings.
The Observer is publishing the full report, executive summary, area addendum and legislative presentation alongside this article so that readers can review the data for themselves.
The review traces directly to an October 2025 joint hearing of the Arkansas House and Senate State Agencies and Government Affairs Committee, where legislators publicly confronted Wright over the agency’s handling of a Baxter County child abuse case in which 29 hotline reports over 11 years were deemed unsubstantiated while a boy was allegedly locked in a bathroom, starved and restrained by his guardians.
At that hearing, Rep. Jack Rose pressed Wright on whether DHS had searched its own database for other children in similar situations. Wright told the committee she had not.
“So in almost one year’s time, you discovered that there were 29 reports for one name, and it hasn’t occurred to anybody to export the data,” Rose said. “I just thought of this sitting here.”
Rose made three specific requests of DCFS leadership: search the database immediately for other children with high numbers of unsubstantiated reports, conduct one-on-one reviews with the employees who handled the Barnett/Wright case and report findings back to the committee.
DHS eventually ran the search. It identified 341 investigations meeting the criteria. Evident Change removed 14 that had incorrectly been included because the families had actually received services, leaving 327 for review.
The findings confirmed that the failures in the Baxter County case were not isolated.
Of the 327 investigations, 16% were closed despite the agency’s own structured risk assessment recommending the case be opened for ongoing services. Documentation rarely explained why the recommendation was not followed. An additional 13% did not have risk assessments or safety assessments recorded at all, making it impossible to determine whether a service recommendation had been generated.
Only 28% of investigations had a safety assessment documented at closure, a step required by DCFS policy. A majority of cases, 79%, had only one safety assessment on file when policy requires at minimum two: one at the start and one at the end of the investigation. Less than half, 47%, had a safety assessment completed within the first five days.
Out of every safety assessment that was completed across all 327 investigations, only one identified a safety threat. The rest were marked “safe.” But when Evident Change reviewers read the actual case narratives, they found unaddressed safety concerns in 13% of investigations at the time of closure. Examples included unaddressed access to firearms, untreated medical conditions, undisclosed sexual abuse allegations involving siblings and cases where the family was never located or seen.
The disconnect between formal assessments and what reviewers found in the case files suggests the state’s safety assessment tool is being used to document closure rather than to evaluate danger.
Risk assessments showed a similar pattern. Only 29% included scoring for family characteristics, a category that captures conditions such as substance abuse, domestic violence and mental health concerns. The remaining 71% of risk scores were based solely on the family’s history of contact with the agency. In other words, investigators were scoring paperwork, not families.
Interviews with children, siblings and caregivers fell short of policy requirements across the sample. Only 59% of siblings received timely interviews. Only 60% of caregivers were interviewed at the initial face-to-face contact. In 37% of investigations, reviewers found that not all relevant collateral contacts, such as school officials, mental health professionals and law enforcement, had been contacted.
The report noted specific examples of missed opportunities without identifying families. In one case, a child had written a note found by a teacher mentioning sadness about a stepfather and having access to a gun. Investigators never asked the child about the stepfather’s arrest or the weapon. In another, a mother asked for help getting a child tested for autism, and no documentation showed the agency followed up.
Safety-Organized Practice, the engagement framework DCFS adopted in 2022 as part of its “At One Table” practice model, was absent from 72% of investigations reviewed. Wright had described the model to legislators at the October hearing as a key reform, telling the committee the agency had been “rolling out structured decision making” and working toward “more engagement, both with the family, but also relying on safeties of support and networks of support.”
The review also identified 20 families that appeared more than once in the 327-case sample, meaning they had been investigated, closed and then investigated again during the same review period. Evident Change found that approximately 30% of those re-referrals might have been avoided if investigators had provided services during the first investigation. Examples included families needing parenting support, domestic violence intervention and thorough interviews that never took place.
Eight investigations were so concerning that Evident Change escalated them to DCFS during the review itself, before the report was finished. These cases typically involved a lack of any contact with the children, limiting the agency’s ability to assess safety, or incomplete assessment of allegations.
A legislative presentation prepared by DCFS and obtained by the Observer reveals what happened next. After reviewing those eight cases, DCFS leadership took personnel action: one employee was terminated, one was demoted, one received a written warning and four received counseling statements.
That discipline marks the first documented personnel action connected to the systemic failures identified since the Barnett/Wright case became public in November 2024. At the October 2025 hearing, Wright told the legislature that no employees had been disciplined.
“There has not been disciplinary action for staff,” Wright said when Rose asked directly.
Rose responded: “In your letter, you say, we will do everything in our power to prevent this from ever happening again. The 29 reports, nothing was done. Nobody’s punished. Does that sound like you’re doing everything in your power to prevent it from happening again?”
The discipline that did occur came only after an outside contractor flagged the cases. It did not result from internal review by DCFS leadership. And it did not address Rose’s second demand: one-on-one reviews with the specific employees who handled the 29 Barnett/Wright reports over 11 years.
There is no mention of those individual employee reviews in the CQI report, the area addendum, the executive summary or the legislative presentation. Rose’s request appears to have gone unaddressed.
His third demand, that DCFS report back to the State Agencies committee, has also not been fulfilled in a public hearing. Sen. Scott Flippo of Bull Shoals, who chaired the October hearing, told the committee he intended to hold “at least one more state agencies meeting between now and the end of the year.” That hearing never took place. State Rep. Stetson Painter told the Observer earlier this year that a second hearing is expected later in 2026 after the legislature completes work on the state budget.
The committee also voted unanimously in October to refer the closed cases to the Child Maltreatment Investigations Oversight Committee, a closed-door legislative body chaired by Sen. Alan Clark. No public report has been issued from that referral.
The CQI report itself does not reference the legislative hearing, Rose’s demands, the Barnett/Wright case or the circumstances that led to the review. The background section states that DCFS “requested” the review as part of “current statewide priorities for continuous quality improvement.” The framing presents the review as a proactive quality initiative rather than a response to a public confrontation in which legislators demanded answers the agency could not provide.
In her KTLO interview Monday, Wright focused on improvements DCFS has made since the review. She described an IT system enhancement that now prevents investigators from closing a case without documenting a review of the family’s history. She described a new requirement for leadership review of cases involving families with 10 or more prior reports. She confirmed the eight escalated cases and said the agency “took further action on each of those eight cases.”
She did not describe the nature of that action. The legislative presentation specifies the personnel outcomes.
Wright also disclosed that DCFS currently has 137 vacant budgeted positions and that frontline worker turnover stands at 53%, down from 60% to 70% in prior years.
Those numbers provide operational context for the findings in the CQI report. If the agency is running with 137 unfilled positions and losing more than half its investigation workforce every year, the gaps in timely contact, safety assessments and documentation identified in the review have a structural explanation. The report itself does not address staffing levels or caseloads as contributing factors.
The legislative presentation outlines additional steps DCFS has taken or plans to take. The agency has amended its Supervisor Review Tool to require area director or program administrator sign-off when a family has 10 or more prior reports. It has established a centralized email for prosecuting attorneys to request records. It has begun a second round of investigation training and is developing a new staff training program to launch in July 2026 for new hires. DCFS Director Wright completed travel to all 10 service areas to discuss the review findings, and each area was required to submit an improvement plan by the end of February 2026. A statewide supervisor convening is scheduled for April 2026.
The Evident Change report breaks findings down by the agency’s 10 service areas. Baxter County falls in Area 5, which had 27 investigations in the sample. Area 5 showed 74% timely face-to-face contact with alleged victims, slightly above the statewide average, but only 59% had a safety assessment completed within five days and only 15% had documented use of Safety-Organized Practice.
Area 7 had the worst performance on several measures, with only 38% of investigations including timely face-to-face contact with the alleged victim and the highest proportion of missing safety assessments. Area 1, which includes Northwest Arkansas, had the largest share of the sample at 74 investigations and the highest proportion of high or very high risk scores at closure.
Evident Change acknowledged limitations. The review was conducted in three weeks on what the report calls “an expedited timeline.” Small sample sizes in most areas, ranging from five to 74 investigations, limited the ability to draw statistically meaningful comparisons. The review was also constrained by the same documentation gaps it was designed to identify. In many cases, reviewers could not determine whether practices occurred because they were not recorded.
The report recommends expanding future reviews to include families new to the system, families investigated by the Crimes Against Children Division rather than DCFS and families that do receive ongoing services. It also recommends quarterly feedback sessions with staff and supervisors, longitudinal tracking of the 327 families to monitor re-involvement and implementation of a supervisory case-reading process for ongoing quality review.
The Barnett/Wright criminal case, which prompted the legislative hearing that led to this review, remains active. Daniel Wright, 41, and Jaclyn Barnett, 42, face 112 counts each of first-degree false imprisonment, endangering the welfare of a minor and permitting child abuse. Kidnapping charges have been added. A pretrial hearing was scheduled for today before Judge Andrew S. Bailey in Baxter County Circuit Court, with a jury trial set for May 18.
The internal investigation report on the 29 unsubstantiated calls in that case, which former DHS Director Kristi Putnam personally promised to Baxter County Prosecutor David Ethredge before leaving the agency, has never been released. DHS later called that promise “a mistake” and cited confidentiality statutes to block public disclosure.
No DHS employee has been publicly identified as having been disciplined in connection with the agency’s failure to act on 29 abuse reports spanning more than a decade in that case.
The full Evident Change report, executive summary, area addendum and DCFS legislative presentation are available for download on the Mountain Home Observer’s website.
