GHOST VISITS: THE NATIONAL INVESTIGATION — ARTICLE 5 Ghost Visits

Falsified Records, Drive-By Checks, and the Paper Trail of Lies

Jul 08, 2026

In Article 1 of this Ghost Visits series, we established what federal law actually requires: monthly caseworker visits for at least 95% of children in foster care, and not just any visits — visits that must be well-planned, focused on issues relevant to case planning, and explicitly designed to ensure the child’s safety, permanency, and well-being. In Article 2, we followed the money through the federal child welfare system and discovered a financial architecture built on open-ended Title IV-E entitlements that creates powerful structural incentives for the volume of children removed and placed in out-of-home care while allocating almost nothing to the most basic safety function the law requires: the actual verification that those children are safe once they are in the system. In Article 3, we moved from the incentives to the compliance reality, documenting state after state failing to meet even the most basic quantitative thresholds for monthly visitation, with up to 140,000 children per month receiving no safety check at all under realistic assumptions drawn from documented state performance. In Article 4, we exposed the human infrastructure crisis that makes those thresholds impossible to meet: impossible caseloads far exceeding evidence-based standards, catastrophic turnover rates that destroy case continuity, and a workforce being systematically burned out and driven from the profession by the very conditions the system creates.

Now, in Article 5, we confront something worse than failure. We confront fraud.

This article documents a pattern of caseworkers who have been criminally prosecuted, civilly disciplined, and administratively sanctioned for manufacturing a paper trail of visits that never happened. These are not allegations, suspicions, or anonymous whistleblower claims. These are felony convictions, guilty pleas, and state disciplinary records spanning multiple jurisdictions and documented in court proceedings. The conclusion is unavoidable and it reframes everything that has come before in this series: the self-reported visit compliance data that states submit to the federal government — the data that forms the foundation of every official claim about child safety in foster care — cannot be taken at face value.

The failure rates documented in Article 3 are not the ceiling of this crisis. They are the floor.

Executive Summary: The Fraud Beneath the Numbers

The monthly caseworker visit is the single federally mandated safety check standing between a foster child and complete invisibility inside the system. When that visit does not occur but the record says it did, the child disappears from meaningful oversight. The state reports compliance. The federal government accepts the numbers. The court relies on the agency’s reports. And the child remains unseen.

Across multiple states, caseworkers have been criminally prosecuted for exactly this crime: manufacturing a paper trail of visits that never happened.

Ohio: Caseworker Ashley Steele was charged with 11 third-degree felony counts of Tampering with Records for claiming visits with clients that never occurred. She ultimately pleaded guilty to two counts of tampering and one count of attempted tampering.

Colorado: Caseworker Sandra Spraker was arrested on 99 charges — including 29 counts of forgery of government documents — after investigators found 32 fabricated claims of face-to-face meetings with families. She was sentenced to 60 days in jail.

Georgia: More than 800 CPS workers — roughly 42% of the state’s caseworker workforce — were disciplined over six years. About 14% of those disciplined were cited specifically for falsifying records or lying about checking on children — translating to approximately 112 workers caught falsifying records in a single state in six years.

Illinois: In the Mackenzi Felmlee death investigation — the 18-year-old who died weighing roughly 90 pounds after cycling through 10 caseworkers in five years — a search warrant application alleged that a caseworker was copy-pasting home-visit notes, and the last “visit” on record was a phone call with the foster mother reported as an in-person contact with the child.

These cases are not isolated. They are documented, proven, and criminally prosecuted. They represent the tip of an iceberg whose full size we cannot know because the detection model American child welfare currently relies on is accident — not systematic audit.

The conclusion is unavoidable and it reframes this entire series: self-reported visit compliance data cannot be taken at face value. The 95% mandate is not merely unmet. In significant portions of the system, it is actively falsified.

The Crime at the Heart of the System

Every safeguard in federal child welfare law ultimately depends on one event actually occurring: a trained professional standing in the same room as a foster child, looking at that child, speaking with that child privately when appropriate, observing the child’s physical condition and emotional state, assessing the safety of the home environment, and making a professional judgment about whether that child is safe.

Federal law — 42 U.S.C. § 622(b)(17) and § 624(f), as documented in detail in Article 1 of this series — requires that these visits happen monthly for at least 95% of children in care. It further requires that the visits be “well-planned and focused on issues relevant to case planning” and “designed to ensure the safety, permanency, and well-being of the child.” These are not suggestions. They are statutory mandates tied to federal funding.

A ghost visit is the betrayal of that requirement at its root: a visit that exists on paper and nowhere else in reality.

The caseworker documents a home visit that never took place. The caseworker documents a face-to-face contact that never occurred. The caseworker documents a safety assessment that was never conducted. The child was not seen. The home was not entered. The assessment was never made. But the official record says otherwise, and that record travels forward through every layer of the system.

Every downstream actor inherits the lie and acts on it:

The supervisor reviews the file, sees documented compliance, and signs off.

The state’s aggregate compliance statistics improve, shielding the agency from scrutiny.

The federal oversight apparatus accepts the numbers as evidence of state performance.

The juvenile court relies on agency reports that claim the child was seen, assessed, and found safe.

Subsequent caseworkers inherit a file that appears complete and current.

The child remains invisible to the system that removed them from their family in the name of protection.

Perhaps the most alarming finding of this investigation is that this pattern is not hypothetical, not speculative, and not limited to a few bad actors. It is documented, proven in court, and criminally prosecuted. What follows in the sections below is not speculation about what might be happening inside child welfare agencies across America. It is the record of what courts, prosecutors, and state disciplinary bodies have already established under oath, under penalty of perjury, and under the rules of evidence in multiple jurisdictions.

Ohio: Ashley Steele — 11 Felony Counts

Ashley Steele, 37, a Muskingum County Children Services caseworker in Ohio, was charged with 11 third-degree felony counts of Tampering with Records — one of the most serious criminal charges that can be brought against a public official who falsifies government documents.

Ten of those charges stemmed directly from falsifying information in caseworker reports — claiming she had conducted visits with clients when those visits never occurred. In multiple instances, Steele documented home visits, face-to-face contacts, and safety assessments that court records later established had no basis in reality. The children on her caseload were not seen. Their homes were not entered. Their safety was not assessed. But the official record said otherwise.

The eleventh charge alleged that she provided false documentation claiming a prior criminal conviction had been expunged or sealed — meaning she lied to get the job in the first place, then lied about doing the job after she got it. The lie that got her in the door was followed by lies about what she did once inside.

A review of her case files uncovered multiple instances in which Steele entered false information into official records, including claims that she met with clients when those meetings did not happen. These were not minor errors or omissions. They were affirmative fabrications entered into the permanent case record that would follow each child through the system.

The Plea and the Damage

Steele ultimately pleaded guilty to two counts of tampering with records and one count of attempted tampering. The plea deal reduced 11 felony charges to three — a significant reduction that reflects the practical realities of plea bargaining in cases involving voluminous documentary evidence and multiple victims.

But the damage to the children on her caseload is irreversible and incalculable.

Every falsified visit represents a child who was supposed to be seen by a trained professional and was not. Every fabricated case note represents a safety assessment that never occurred. Every time Steele documented a visit that did not happen, she created a false record that could have been used by supervisors to believe the child was safe, by courts to make placement decisions, by subsequent caseworkers to assume continuity of oversight, or by the system as a whole to claim compliance with federal mandates. The lie did not stay on the page. It traveled forward and shaped decisions that affected real children in real time.

The Question the Steele Case Forces

If a single caseworker can falsify ten or more visit reports without detection — until a separate investigation happens to reveal the fraud — how many other falsified reports exist in systems across the country that have never been audited?

The Steele case was not caught by any compliance monitoring system. It was not caught by any supervisor review process. It was not caught by any data validation or anomaly detection mechanism. It surfaced collaterally during an unrelated investigation into other matters.

That is the detection model American child welfare currently relies on: accident. The system does not systematically look for fraud. It waits for fraud to surface through some other channel, and then acts — if it acts at all.

Colorado: Sandra Spraker — 99 Charges, 60 Days

If the Steele case demonstrates that ghost visits can happen without detection, the case of Sandra Spraker demonstrates that they can happen systematically, at scale, inside a single caseload, for an extended period of time.

Sandra Spraker, a Larimer County Child Protection Services caseworker in Colorado, was arrested on 99 criminal charges — a staggering number that reflects sustained, systematic fraud rather than isolated lapses.

The Charge Breakdown

Charge Counts Classification

Attempt to Influence a Public Servant 10 Felony 4

Forgery of Government Documents 29 Felony 5

Embezzlement 1 Felony 5

First Degree Official Misconduct 29 Misdemeanor 2

Second Degree Forgery 30 Misdemeanor 1

Investigators found that Spraker never met with six families, though she made it appear as if she had provided services to them. The investigation uncovered 20 false claims of face-to-face meetings with families — then found 12 more. She also inflated mileage claims for travel to visits that never occurred, adding financial fraud to the safety fraud.

The Sentence

Spraker was sentenced to 60 days in jail in a work-release program, three years of supervised probation, and 300 hours of public service.

Sixty days — for 99 charges involving the safety of vulnerable children.

A prosecutor characterized the case as leaving “a scar on the justice system.” The sentence itself is evidence of the systemic undervaluing of child welfare oversight: the fraud was prosecuted, proven, and then priced at almost nothing.

What Spraker Proves

The Spraker case demonstrates that ghost visits are not isolated incidents or occasional oversights. They can be systematic patterns sustained over extended periods inside a single caseload. If one caseworker in one county can fabricate enough false documentation to generate 99 criminal charges, the question becomes unavoidable: how many other counties across America have a Sandra Spraker they haven’t caught yet?

Georgia: 42% of the Workforce

Individual prosecutions can be dismissed as bad apples. Georgia is the orchard.

An 11Alive investigation revealed that between January 2013 and August 2018, more than 800 DFCS employees assigned to Child Protective Services were disciplined for various violations — out of a total workforce of approximately 1,900 caseworkers. That is 42% of the entire CPS caseworker workforce in a single state over six years.

The Breakdown of Violations

Over 40% of those disciplined were cited for negligence or inefficiency.

About 14% were disciplined specifically for falsifying records or lying about checking on children.

That 14% translates to approximately 112 workers caught falsifying records in a single state in six years.

Not Bad Actors — Institutional Culture

The Georgia data is significant precisely because it is not one rogue caseworker. It is a systemic pattern across an entire state agency.

When 42% of a workforce has been disciplined — and 14% of those specifically for falsification — the problem is not individual bad actors. It is institutional culture.

Consider what the Georgia numbers imply nationally. If Georgia’s falsification rate among disciplined workers is applied to the broader child welfare workforce — and if Georgia is not uniquely dysfunctional (and the CFSR data documented in Article 3 suggests it is not) — then the national scope of visit record falsification may be vastly larger than the handful of criminal cases that have been prosecuted.

The system is not merely failing to meet the 95% mandate. In significant portions of the country, the mandate is being actively falsified, and the detection rate is so low that only the most egregious or unlucky offenders are caught.

Illinois: The Copy-Paste Case Notes

In the Mackenzi Felmlee death investigation — the case documented at length in Article 4, in which an 18-year-old died weighing roughly 90 pounds after cycling through 10 different caseworkers in five years — a search warrant application alleged that a caseworker was “copying and pasting notes for home visits, which does not appear to have correctly documented the visits.”

This practice — creating the appearance of individual, substantive visits through recycled, generic documentation — is a more insidious form of fraud than outright fabrication, because it is harder to detect.

A supervisor reviewing the file sees what appears to be a complete record of regular visits with substantive observations. Only a careful comparison of notes across multiple visits — looking for identical language, implausible consistency, or observations that do not match the child’s changing circumstances — would reveal the fraud.

The Last “Visit” That Wasn’t

The Illinois case also reveals a second falsification mechanism: misclassification.

Mackenzi’s last “contact” on record was a phone call with the foster mother — not a face-to-face visit with Mackenzi herself. Yet the case agency initially reported it as an in-person visit.

Whether intentional or the product of poor documentation practices, the last “visit” on record before a child’s death was fundamentally different from what actually occurred.

A phone call with a caregiver cannot see a child’s weight loss. It cannot see bruises. It cannot see the child at all.

When the final documented contact before a child’s death is a phone call misclassified as an in-person visit, the paper trail has completely detached from reality.

The Drive-By Visit Problem

Even when visits genuinely occur, advocates, child welfare researchers, and frontline workers themselves have identified a pervasive and well-documented pattern of “drive-by” visits — caseworkers who arrive at a foster home, conduct a brief and superficial check, sometimes lasting as little as 5 to 10 minutes, and leave without meaningfully assessing the child’s safety, well-being, emotional state, or living conditions.

These are not visits in any substantive sense. They are checkbox exercises. The caseworker drives to the location, makes contact, confirms the child is physically present, and departs. No private conversation with the child. No careful observation of the home environment. No assessment of whether the child’s needs are being met. No screening for new safety threats or changes in the child’s circumstances. Just enough presence to generate a “completed visit” entry in the case management system.

Drive-by visits technically count toward compliance statistics. A caseworker who drives to a home, knocks on the door, sees the child for five minutes, and leaves can document a “completed visit” and move on to the next name on the list. On paper, the 95% mandate is satisfied. In reality, the child has received no meaningful protection from the visit.

But such visits violate the spirit and the letter of the federal quality requirements, which mandate that visits be well-planned, focused on case planning, and designed to ensure the safety, permanency, and well-being of the child. The law does not simply require that a body show up. It requires that the visit accomplish something protective.

The Caseload Connection

The drive-by problem is compounded by the caseload crisis documented in Article 4.

A caseworker carrying 30 or more cases who must complete 30 or more visits per month is under enormous structural pressure to minimize time per visit. The incentive structure rewards quantity — completing the visit and moving to the next one — over quality — conducting a meaningful assessment that could actually identify danger or unmet needs.

The compliance metric asks one question and only one question: Did the visit happen?

It does not ask whether the visit was meaningful. It does not ask whether the child was spoken with privately. It does not ask whether the home environment was observed. It does not ask whether any new information was gathered or any new safety threat was identified. It does not ask whether the visit accomplished the protective purpose the law requires.

This is the quiet corollary to the fraud documented in Sections II through V: a system that only counts occurrences — without verifying quality or even existence — invites both the five-minute drive-by visit and the completely fabricated one, because on paper they look identical to a real safety assessment. Both produce the same checkbox. Both improve the compliance statistic. Both allow the system to claim it is meeting its obligations to children.

The difference is that the drive-by visit at least involves a physical presence. The fabricated visit involves nothing at all except a keystroke.

Four Ways to Fake a Visit

The cases documented above reveal at least four distinct methods by which ghost visits are manufactured:

Outright Fabrication: The caseworker documents a visit that never occurred at all. (Ashley Steele in Ohio; Sandra Spraker in Colorado for multiple families.)

Copy-Paste Recycling: The caseworker reuses generic language or previous visit notes across multiple visits, creating the appearance of individualized documentation without conducting individualized assessments. (Alleged in the Illinois Mackenzi Felmlee case.)

Misclassification: The caseworker conducts a phone call, a drive-by, or a contact with a caregiver and documents it as a face-to-face visit with the child. (The final “visit” in the Mackenzi Felmlee case.)

Drive-By Minimization: The caseworker conducts an extremely brief, superficial contact that technically satisfies the “visit occurred” checkbox but fails every substantive requirement of federal law. This is not criminal fraud in the same sense, but it produces the same result: a child who is not meaningfully seen.

All four methods produce the same outcome on paper: a record that says the child was seen when the child was not seen in any way that matters.

The Systemic Implication: The Data Cannot Be Trusted

The cases documented in this article are not the full extent of the problem. They are the cases that were caught.

The detection model is accident, not audit. Most falsification will never surface unless it collides with an unrelated investigation, a child’s death, or an unusually thorough supervisor.

When Georgia alone disciplined 112 workers for falsifying records or lying about checking on children in six years — and when that represents only the caught cases in one state — the national scope of visit record falsification is unknown and likely substantial.

The self-reported visit compliance data that states submit to the federal government — the data that underpins every claim about child safety in foster care — is built on a foundation of unverified, and in some documented cases actively falsified, caseworker entries.

The 95% mandate is not merely unmet in many jurisdictions. It is, in significant portions of the system, a number that has been manufactured.

What Honest Verification Would Look Like

If the system took the safety of foster children seriously, verification of monthly visits would not rely on self-reported caseworker entries. It would incorporate independent, auditable mechanisms:

Random in-person audits of a statistically significant sample of claimed visits.

Cross-referencing of caseworker GPS data, time-stamped photographs, or other objective indicators (with appropriate privacy protections).

Required collateral contacts — speaking with the child, the foster parent, teachers, or other adults who can independently confirm that a meaningful visit occurred.

Supervisor sign-off that includes review of whether the documented observations are consistent with the child’s known circumstances and prior records.

Public reporting of verification methodology and audit findings, not just raw compliance percentages.

None of these mechanisms exist at scale in American child welfare today.

Recommendations

The ghost visit crisis is not an inevitable feature of human services work. It is the direct, predictable result of a system that has created overwhelming structural incentives to produce the appearance of compliance on paper while providing almost no meaningful verification, accountability, or consequences for the quality — or even the existence — of the underlying work. The fraud documented in this article is not the product of a few bad actors. It is the product of a system that makes fraud both easy to commit and difficult to detect, then treats detection as a rare accident rather than a core function of oversight.

The following recommendations represent the minimum structural changes necessary to begin restoring integrity to the visitation mandate and to the data that supposedly measures whether children in foster care are safe:

Mandate independent verification of a statistically significant sample of claimed monthly visits in every state, conducted by personnel who are not part of the same agency that produced the original documentation. Verification must include direct observation, collateral contacts with the child and other adults who can independently confirm that a meaningful visit occurred, and public reporting of verification methodology, sample size, and findings — not just aggregate compliance percentages.

Criminalize and aggressively prosecute the falsification of child welfare records with penalties that reflect the gravity of the harm to children. A sentence of 60 days in jail for 99 charges involving the safety of vulnerable children sends a clear message that the system does not take this fraud seriously. Prosecutors and judges must treat the fabrication of safety assessments as the serious crime it is — a crime that can directly contribute to children remaining in dangerous situations undetected.

Tie a meaningful portion of federal Title IV-E and other child welfare funding to verified visit compliance, not self-reported numbers. States that cannot demonstrate through independent audit that visits actually occurred at the rates claimed should face predictable, escalating financial consequences. The current system rewards the appearance of compliance; federal funding must reward verified compliance.

Reduce average caseloads to levels that make meaningful visits possible, consistent with the Child Welfare League of America standards documented in Article 4. The drive-by visit and the fabricated visit are both symptoms of a system that demands quantity over quality and then expresses surprise when workers respond to impossible demands by cutting corners or manufacturing records. Workload reduction is not optional; it is a prerequisite for honest compliance.

Implement workload-weighted documentation requirements that reduce the crushing administrative burden that incentivizes recycling or fabricating notes simply to meet volume requirements. Documentation should serve child protection, not primarily serve federal reimbursement compliance or internal bureaucratic metrics.

Create robust whistleblower protections and meaningful incentives for caseworkers who report pressure to falsify records, who observe systematic falsification by colleagues or supervisors, or who are asked to sign off on documentation they know to be false. The current culture in many agencies punishes those who speak up; that culture must be inverted.

Require public, transparent dashboards that report not just claimed visit compliance rates, but verification audit results, falsification findings, methodology used to validate the numbers, and trends over time. Sunlight is the best disinfectant, and the current system operates largely in the dark.

Fund independent, rigorous research into the prevalence of ghost visits across jurisdictions, the effectiveness of different verification mechanisms, and the relationship between verification practices and child safety outcomes. The current evidence base is stronger on documenting individual cases than on understanding the systemic scope and testing solutions.

The children in America’s foster care system deserve caseworkers who actually see them — not caseworkers who document seeing them. They deserve records that reflect reality, not records that manufacture a reassuring fiction. They deserve a system that does not treat the fabrication of safety assessments as a minor administrative matter to be plea-bargained down to almost nothing.

Every child deserves to be seen.

Every visit that is claimed on paper must be real in the world.

The paper trail of lies ends when the system chooses to verify instead of trust, to audit instead of assume, and to protect children instead of protect the appearance of protection.

Project Milk Carton is a 501(c)(3) nonprofit dedicated to U.S. child welfare transparency, CPS and foster care accountability, human trafficking prevention, and OSINT-driven investigations.

Data current as of July 2026. Sources include State v. Steele (Ohio 2024), State v. Spraker (Colorado 2023), 11Alive investigation of Georgia DFCS discipline records (2013-2018), NPR Illinois reporting on the Mackenzi Felmlee case, state disciplinary records, and PMC CivicOps analysis. Full methodology and replication files available at projectmilkcarton.org.

Share this investigation. Demand that your state implement independent verification of visit compliance. Demand that falsification of child welfare records be treated as the serious crime it is. The children cannot wait.

If you enjoyed this work and feel encouraged by independent investigations like this, please consider becoming a paid subscriber to The Constitutional Republic on Substack or donate at We Built the Tools. Help Us Keep Them Running.

Project Milk Carton Inc. Your gift keeps our site alive, spotlighting high-risk children!

Your support helps us continue:

investigative reporting
public records research
educational transparency projects
community outreach
and future investigations like this series

Most importantly, please share this article.

The Constitutional Republic

Discussion about this post

Ready for more?