Watchdog Finds Higher Child Welfare Death Toll Than State Reported

Washington's independent child welfare watchdog told state overseers this week that more children died or nearly died last year than the agency running the system has acknowledged. The Office of the Family and Children's Ombuds reported that 35 children with a connection to the child welfare system died in 2025, and 49 more nearly died from abuse, neglect or other maltreatment. That is well above the 23 deaths and 34 near-deaths the Department of Children, Youth and Families has reported for the same year, according to Washington State Standard, which first reported the ombuds office's presentation to the state's Department of Children, Youth and Families Oversight Board on Thursday. Accidental drug ingestion drove most of the deaths, the report found, with most incidents occurring while children were in their parents' care and the majority of those who died younger than one year old. The gap between the two counts, and what it means for a state still arguing over how it separates families, is the story behind Thursday's numbers.

A Second Count, and a Higher Number

The Office of the Family and Children's Ombuds is not part of the Department of Children, Youth and Families. It is an independent state watchdog, created by the Legislature specifically to review how DCYF handles cases from outside the agency's own chain of command, and its fatality reviews use a broader set of criteria than the department applies to its own count. That structural difference is why the two 2025 tallies do not match: DCYF's own reporting put last year's toll at 23 deaths and 34 near-fatalities, while OFCO's review found 35 deaths and 49 near-fatalities, a combined 84 incidents against the department's 57. The Columbian, which also carried the report Thursday, noted the ombuds office reviews cases the department's own count does not capture, rather than disputing DCYF's methodology outright. Both counts describe the same underlying pattern: a child welfare system logging more fatalities and near-fatalities than officials have described in recent years, with the independent reviewer landing on the larger number. Multiple outlets covering the department's own data through 2025 already described the year as record-setting before OFCO's higher count arrived Thursday, which raises the stakes of whichever number the Legislature treats as authoritative when it writes next year's budget and policy. For a reader outside state government, the practical takeaway is that "how many children died in Washington's child welfare system" does not currently have one answer, and the answer that gets quoted in a legislative hearing or a budget request depends on which office's report the speaker chooses to cite.

How the State Watched This Number Climb All Year

Thursday's report was not the first warning. The ombuds office told the oversight board last November that 2025 was already on pace to set a record for deaths and near-fatalities tied to the child welfare system, based on data collected through the first three quarters of the year. That early accounting was incomplete by design: fatality and near-fatality reviews take months to finalize, so any in-year figure undercounts what eventually gets confirmed once investigators close their files. OFCO Director Patrick Dowd told the board this week that his office deliberately avoided suggesting things were improving as the year went on, even as monthly totals appeared to moderate. "We didn't want to give the impression that things are getting better, and are actually starting to decline, when in fact we had preliminary information for the first quarter of 2025 that might paint a very different picture," Dowd said, according to Washington State Standard's account of the presentation. The final number, arriving nine months after the year closed, bore that caution out: worse than the preliminary data suggested, not better. For a reader trying to judge how bad the child welfare system's 2025 truly was, the practical lesson is that any partial-year number the state releases should be read as a floor, not a forecast.

Fentanyl in the Home, and Children Too Young to Reach for Help

Both DCYF and OFCO's reporting point to the same driver behind last year's spike: accidental exposure to opioids, overwhelmingly fentanyl, inside homes where children were living with a parent rather than in foster care. State officials reporting on the trend through 2025 said opioids were present in roughly half of the year's fatality and near-fatality cases, a share that has held for more than a year of quarterly reporting. The ages skew young for a structural reason child welfare researchers have flagged for years: infants and toddlers cannot recognize a pill or residue as dangerous, cannot call for help, and are more likely to be in a caregiver's direct physical space when an overdose occurs nearby. That combination, a drug potent enough to kill in residue amounts inside a home where a parent's own crisis leaves a child unsupervised at the exact moment it matters, is why caseworkers and legislators alike have described these cases as different in kind from the abuse and neglect fatalities the system was originally built to catch. There is a partial counterpoint in the same report: through the first half of 2026, critical injuries tied to the child welfare system have dropped substantially, tracking a broader statewide decline in overdose rates. Whether that holds for a full year, or repeats last year's pattern of an optimistic early read followed by a worse final tally, is the open question the ombuds office is not yet willing to answer.

DCYF's Own Read on a Year It Called a Record Before It Ended

Department officials did not wait for Thursday's finalized figures to acknowledge 2025 was going badly. Speaking to the oversight board last November, when only partial-year data was available, DCYF assistant secretary Vickie Ybarra told members: "Although the numbers have moderated, thankfully, quite a bit since quarter one of 2025, we do anticipate that 2025 will be a record year." Natalie Green, the department's assistant secretary for child welfare, framed the department's approach around keeping families intact rather than defaulting to removal, telling the board the goal was supporting "families to be able to safely parent their children." As of that same November presentation, DCYF told lawmakers it wanted $4 million in additional funding for drug treatment programs serving families in the system, along with 37 new full-time caseworker and support positions. Those remarks and that funding request predate this week's finalized, higher count from the independent watchdog, and neither DCYF assistant secretary had addressed the specific gap between the department's own 57-incident total and OFCO's 84 as of Thursday's meeting. Neither Thursday's coverage nor the department's own November remarks disputed OFCO's authority to conduct its own count using its own criteria — the disagreement here is about which agency's methodology produces the more complete picture, not about whether either count is legitimate.

The Keeping Families Together Act Is Still the Fight Underneath This

The rising numbers have reopened a legislative argument that predates this year's report. Washington's 2021 Keeping Families Together Act raised the legal standard for separating a child from parents, aiming to stop poverty alone from being treated as grounds for removal and to reduce the trauma of family separation, a response to years of criticism that the state was taking children from struggling but non-abusive parents. When lawmakers debated the law's effects this past February, Rep. Travis Couture, R-Allyn, argued the current approach was failing the children it was meant to protect: "What we're doing right now, leaving them together and crossing our fingers and hoping for the best, is insane to me. This hurts me on almost a personal level. I have to fix this." Rep. Steve Bergquist, the Democrat who chairs the House Early Learning and Human Services Committee, pushed back against a full rewrite, saying he did not think the Legislature needed "wholesale changes this session" and preferred targeted fixes with time to show results before any larger overhaul. DCYF itself had not requested policy changes for the 2026 session as of that February hearing, according to Washington State Standard's reporting at the time. That position will be harder to hold if Thursday's higher, finalized numbers carry into the next legislative session, which typically opens in January, since a department asking for no new authority looks different against a watchdog report showing the death toll rose rather than fell.

What Happens Next, and Why the Count Itself Is the Fight

No new legislation was announced alongside Thursday's presentation, and the DCYF Oversight Board's next scheduled review of the data was not detailed in this week's reporting. But the pattern from the past year is likely to repeat: OFCO will keep publishing its own count on its own timeline, DCYF will keep reporting a lower figure drawn from narrower criteria, and lawmakers heading into the Legislature's next session will have to decide which number, if either, should drive policy. That decision is not a technicality. A lower DCYF-reported toll makes the case for Bergquist's targeted, patient approach easier to sustain; a higher OFCO-reported toll strengthens Couture's argument for structural change to the 2021 law. For a state that spent 2025 watching its preliminary child welfare statistics get revised upward every few months, Thursday's report is less a single data point than the closing entry in a year-long pattern of undercounting followed by correction. Whether the modest improvement in 2026's first-half numbers survives a similar revision once caseworkers finish this year's reviews is the number Washington families, caseworkers and lawmakers will be watching for next. If this year's preliminary count follows the same arc as last year's, the figure the state is citing as good news right now, a drop in critical injuries tied to declining statewide overdose rates, may itself look different by the time the ombuds office finalizes its 2026 report late next year. That is the pattern Thursday's report leaves behind: not a single verdict on how Washington's child welfare system is doing, but a reminder that the state will not know the real number until long after the year it describes has ended.