What Happened

Tara Severino, a registered nurse in the ICU at the West Haven VA Medical Center in West Haven, Connecticut, diverted fentanyl, hydromorphone, and oxycodone intended for critically ill and dying veterans between January and July 2023. In one documented pattern, she administered partial doses to patients and kept the remainder for herself, then falsified records to make it appear the drugs had been properly disposed.

Severino pleaded guilty to obtaining controlled substances by fraud and was sentenced to two years of probation with six months of home confinement. The case was prosecuted by the U.S. Attorney's Office for the District of Connecticut following an investigation into the diversion from VA patients.

The case is a particularly stark example of diversion from dying patients, and it shows how partial-dose administration combined with falsified disposal records can defeat a documentation-only control system.

What Went Wrong

The scheme exploited the gap between documented and actual disposal:

  • Partial-dose administration placed the nurse in sole control of the remainder, and nothing required a second person to witness its disposal.
  • Falsified disposal records were accepted without independent verification, so charted waste was never reconciled against witnessed waste or actual container contents.
  • No reconciliation linked the quantity dispensed to the quantity documented as administered plus the quantity documented as wasted — the core closed-loop check was missing.
  • Waste rates were apparently not benchmarked per clinician, so a nurse with a persistently elevated or irregular waste pattern generated no alert over seven months.
  • Trust in the ICU workflow substituted for verification: the vulnerability of the patient population made the deception more harmful, but no control compensated for the heightened risk.

How It Could Have Been Prevented

  • Require witnessed waste for all partial doses, with the witness's identity recorded and any unwitnessed waste escalated as an exception.
  • Reconcile dispensed minus administered minus witnessed waste to zero for every shift, resolving any gap before the shift closes.
  • Benchmark waste rates per clinician and alert when an individual's waste volume or pattern diverges from unit peers.
  • Conduct random audits of administration and waste records against patient records and pharmacy dispensing data.
  • Apply tamper-evident controls and periodic integrity checks to medications held at the point of care.

Related Guidance

Sources