Diversion already has a documented cost. The controls that prevent it are known.
- Almost every figure comes from a government record: a Department of Justice or DEA action, a state licensing board order, or a court filing.
- Two figures are not from that record: CDC outbreak data, and the program cost the board sets.
- 550 documented cases in 52 states and territories, 2017-2026.
DivertGuard registry: 723 documented cases drawn from U.S. Department of Justice and DEA releases, state licensing board orders and state health department records, of which 550 are dated 2017-2026. The registry counts documented cases; it does not estimate total prevalence.
$34.0M in penalties and 16 deaths are already on the record, and all of it is recent
- Situation: every organization handling controlled substances already has diversion policies.
- Complication: 550 documented cases in 2017-2026; $34.0M in penalties; 16 deaths.
- Exposure: hospitals, retail pharmacies, oxycodone and fentanyl.
- Detection: audits, tips and harmed patients, rarely the control itself.
- Answer: name one owner, fund the five controls, return in 90 days.
DivertGuard registry of 723 documented cases; all figures computed over cases dated 2017-2026 (550 cases); penalty figures from 74 penalty records in that window.
In the last 10 years the record holds 550 documented cases in 52 states
- 550 cases from court filings, settlements and board actions.
- Every case is a government record: a Department of Justice or DEA action, a state licensing board order, or a state health department filing.
- The last ten years are the only period with enough cases to compare fairly.
- Earlier years exist but are far thinner: 55 cases across the 27 years before 2010.
- The registry counts what was caught and published. Total prevalence is larger.
Every one of these cases is a program that did not catch it in time. Treat every figure here as a floor, not a total.
DivertGuard registry of 723 documented cases from DOJ and DEA releases, state licensing board orders and state health department records; 550 dated 2017-2026.
Individuals appear in more penalty records, but organizations pay 71% of the money
- $24.2M paid by organizations, 71% of all penalties in the window.
- $9.8M paid by individuals, the remaining 29%.
- Individuals appear more often. Organizations pay far more each time.
- The financial exposure sits with the organization.
When this goes wrong, the organization pays - not the individual.
Penalty records within the DivertGuard registry dated 2017-2026 (74 records with a stated financial penalty). A settlement is not a finding of liability.
16 of the 550 documented cases record a death or overdose, all of them recent
- 16 cases ended in a death, an overdose or a fatal outcome.
- Every one falls inside the 2017-2026 window.
- The penalties are the recoverable part of the cost. This part is not.
- None of these controls is proposed because a rule requires it.
DivertGuard registry; cases dated 2017-2026 stating a death, overdose or fatal outcome. All 16 documented fatal outcomes fall in this window.
Retail Pharmacy and Hospital / Health System together account for 62% of documented exposure
- Retail pharmacy appears in 201 cases; Hospital / health system in 190.
- A case can involve both, so together they cover 340 distinct cases - 62%.
- Both handle the highest volume of controlled substances.
- Both have the most hand-offs between staff, which is where diversion hides.
Start where the exposure already is. These are the two workflows to control first.
DivertGuard registry; share of documented exposure by care setting, cases dated 2017-2026. A case can happen in more than one setting, so the two shares overlap; the total below counts each case once, not twice.
Fentanyl and Oxycodone together account for 46% of substances named in the record
- fentanyl appears in 24% of cases, oxycodone in 24%.
- A case can name both, so together they cover 252 distinct cases, 46%.
- Both sit in ordinary clinical workflows.
- Dozens of other drugs appear as well, so the control is the process, not any one drug.
DivertGuard registry; share of named substances, cases dated 2017-2026. A case can name more than one substance, so the two shares overlap; the total below counts each case once, not twice.
Nursing roles are named in 53% of the cases that state a clinician role
- 261 of 497 cases that name a clinician role involve nursing.
- That is 53% of those, or 47% of all 550.
- The remaining cases do not state a role either way.
- Registered nurse (RN) is the single most-named role on its own.
Controls belong in the workflow, not on the people inside it.
DivertGuard registry; most specific clinician role stated in the public record, cases dated 2017-2026. Denominator is the 497 cases that state a role, not all 550.
Most documented cases came to light through an audit, a tip or a harmed patient
- Detection came from an audit, a tip or a patient harmed.
- Rarely from the control that was meant to catch it.
- Every route below sits downstream of the diversion.
- The control sits upstream of it.
A control on paper but not reconciled is a policy, not a control.
DivertGuard registry; route by which each case was detected and reached the public record, cases dated 2017-2026.
The typical documented case ran 2 years before anyone caught it
- Only 180 of 723 documented cases say how long the diversion ran.
- The mean ran even higher, at 28 months.
- 51% ran two years or more before anyone caught it.
- 32% ran three years or more - the largest single group.
Two years is the median. That is two years of drug leaves and patients exposed - the case for controls that fire early.
DivertGuard registry; period of diversion as stated in the case record text, not inferred from resolution dates. Of the 723 documented cases, 180 say how long the diversion ran (24.9% of the registry); every share on this slide is a share of those 180.
Every control in this deck protects a patient and a colleague
- PATIENT: 16 documented cases ended in a death or an overdose, all of them recent.
- PATIENT: Separately, CDC surveillance found 45 hepatitis C infections in one 2012 outbreak.
- PATIENT: A reconciled count catches the discrepancy while the drug is still accountable.
- STAFF: Weak controls put everyone with access under suspicion, including people who did nothing wrong.
- STAFF: The nursing figure describes access, not character.
- STAFF: A colleague with substance use disorder is also a patient.
A case here means a patient was harmed, a colleague was investigated, or both.
DivertGuard registry; cases dated 2017-2026 stating a death, overdose or fatal outcome. CDC outbreak data as published on drugdiversionprevention.com.
Diversion surfaces at five handling stages, each with a known control and record
- Procurement: two-person receipt checked against the order. Record: the signed receipt, reconciled to the invoice.
- Storage: scheduled reconciliation of counts, with access logged. Record: the count and access log.
- Prescribing: orders matched to charting; every variance worked weekly. Record: the variance log.
- Administration: dispensing patterns reviewed by person and by unit, with a named reviewer. Record: the review and closure log.
- Waste and disposal: witnessed wastage, with periodic testing of returned waste. Record: witness entries and test results.
Five stages. Five controls. Each one leaves the record that proves it ran.
Control set aligned to the five stages at which diversion surfaces in the DivertGuard registry.
The five controls are what make a program provable, not just written down
- Owner: the Chief Compliance Officer, with pharmacy and nursing leadership as co-leads.
- Days 1 to 30: inventory the controls at all five stages and list every one not actually run.
- Days 31 to 60: stand up the variance log, witnessed wastage and the confidential route.
- Days 61 to 90: run the first full cycle of pattern review and closure.
Each control leaves a record, and the record is what shows it was run - the difference between a policy and a program. Prevention cost is not measured here.
Recommendation informed by the DivertGuard registry, cases dated 2017-2026.
The government record behind every number in this deck.
Every documented case, sourced and searchable, lives at drugdiversionprevention.com.
Searchable by state, drug, care setting, and year.
Documented cases, and the controls that prevent them.
Figures come from a registry of 723 documented cases, computed over the 550 dated 2017-2026. The CDC figure and the program cost come from outside that registry.
Download the PDF