First 90 Days of a Drug Diversion Program
A sequenced plan for a hospital starting a program: what to do in each two-week block, who owns it, what it produces, and what goes wrong if it is skipped.
The plan has 6 blocks and 18 steps over 12 weeks. Each step names one action, the role that owns it, the artifact it leaves behind, and the failure it prevents. Where a step needs a time window or a threshold, you set it from your own policy and your own data. This page gives no benchmark, because a number from another hospital will not carry over to your cabinets, counting rules and drug mix.
Jump to: Weeks 1-2 · Weeks 3-4 · Weeks 5-6 · Weeks 7-8 · Weeks 9-10 · Weeks 11-12. The working files are on the toolkit page.
How to use it. Owners are roles. Substitute the people in your organization, and merge roles where one person holds both. The artifacts are your evidence that the step happened, and the “if skipped” column is the reason to do it when a week is tight. This is a planning guide, not legal or regulatory advice; check your obligations with your compliance officer and counsel.
Weeks 1-2: Baseline and scope
Know what you control, where the risk sits and who is accountable before you measure anything.
| Action | Owner | Artifact it produces | If it is skipped | Tool or guide |
|---|---|---|---|---|
| Name the program lead and the sponsor, and write down who can authorise an investigation and who is told when one opens. | Executive sponsor with the compliance officer | One-page charter with named roles and the notification list | The first real alert has no owner, and notification happens by whoever hears first. | Policy templates |
| Inventory where controlled substances are stored, dispensed, administered and wasted, and the system that records each step. | Pharmacy manager | Location and system inventory (unit, storage, system of record) | A storage area or a record system nobody listed becomes the blind spot a later review finds. | Controlled substance recordkeeping checklist |
| Run the self-assessment and a unit risk review, then rank the units you will cover first. | Diversion program lead | Completed assessment and a ranked unit list with the reason for each rank | Effort goes to the loudest unit instead of the riskiest one, and there is no starting point to show progress against. | Program self-assessment Risk assessment by unit |
Weeks 3-4: Access and ADC configuration review
Make sure the cabinets and the people who use them are set up the way policy assumes.
| Action | Owner | Artifact it produces | If it is skipped | Tool or guide |
|---|---|---|---|---|
| Compare the list of active cabinet users with current HR records, and remove or suspend access for people who have left or changed role. | Pharmacy manager with the ADC administrator and HR | Reconciled user list with each removal dated | Former staff keep working credentials, and any later alert on one of them cannot be tied to a person who should have had access. | Policy templates |
| Export the cabinet configuration for override-capable drugs, override reasons, discrepancy-resolution rules and witness requirements, and compare it with policy. | ADC administrator | Configuration export with a list of differences from policy and an owner for each | The policy says one thing and the cabinet allows another, so review rules are built on behavior that is not actually enforced. | ADC override review worksheet |
| Decide the override review window and who reviews, and start reviewing a sample of overrides by hand. | Pharmacy manager | Written review window and a first filled worksheet | Overrides accumulate unreviewed, and the first time anyone looks is during an investigation. | ADC override review worksheet |
Weeks 5-6: Data feeds and the first indicator
Get cabinet, order and administration data into one place and turn one indicator on.
| Action | Owner | Artifact it produces | If it is skipped | Tool or guide |
|---|---|---|---|---|
| Identify the data sources (cabinet transactions, orders, administration records, waste, staffing), who owns each, and how each can be exported. | ADC administrator with informatics | Data source list with owner, export method and refresh interval | An indicator is built on a feed that quietly omits a cabinet or a unit, and the gap is read as a clean result. | Data sources for diversion KPIs |
| Reconcile transaction counts in your reporting copy against the source system for a test period. | Informatics analyst | Reconciliation note stating what matched and what did not | Every later number rests on a copy that was never checked against its source. | Analytics dashboard |
| Pick one indicator, run it silently against your own history, and set the line for review from your own baseline. | Diversion program lead | Indicator definition, baseline and written threshold, with the first list of who it would have flagged | A threshold copied from elsewhere floods the analyst with alerts or never fires, and the team stops trusting it. | Calibrating a threshold Discrepancy log |
Weeks 7-8: Documentation standards and training
Write down what good documentation is, and teach it to the people who produce it.
| Action | Owner | Artifact it produces | If it is skipped | Tool or guide |
|---|---|---|---|---|
| Adopt or revise the waste, override and discrepancy-resolution policies, so each states what is recorded, by whom, and within what window. | Pharmacy manager with nursing leadership and compliance | Approved policies, with the waste and override logs attached as the expected record | Staff document differently by unit, so audits find variation that is noise instead of risk. | Policy templates Waste documentation log |
| Train pharmacy, nursing and anesthesia staff on the policies and on the signs that should be reported, and record attendance. | Nurse educator with the diversion program lead | Training content, attendance list and a short competency check | Staff who never learned the standard are measured against it, and a report that should have been made is not. | Training materials Red flags |
| Agree how a concern is raised, who receives it, and what happens to the person who raises it. | Compliance officer with HR | Reporting route written into policy and communicated to staff | Concerns stay informal, are told to the wrong person, or are not raised at all. | Communication tips |
Weeks 9-10: First audit cycle
Run the full loop once, on real data, from alert to a closed record.
| Action | Owner | Artifact it produces | If it is skipped | Tool or guide |
|---|---|---|---|---|
| Run the cycle count and discrepancy process for the priority units, and log every mismatch to a root cause. | Pharmacy manager | Filled discrepancy and cycle-count log with an owner on each open row | Discrepancies are resolved verbally, so repeat patterns by person or unit cannot be seen. | Discrepancy and cycle-count log |
| Pull a sample of waste events and trace each to its witness and to the cabinet record, and review the period's overrides. | Diversion analyst | Waste log sample and override worksheet, each with a disposition | The policy is never tested against practice, and witness gaps stay invisible. | Waste documentation log ADC override review worksheet |
| Take one alert through the investigation process, even if it ends as explained, to test the case file and the notification steps. | Diversion program lead | A completed case file and an entry in the case index | The first live case is also the first time the template, the notifications and the evidence hold are tried. | Investigation case file Investigation playbook |
Weeks 11-12: Scorecard to leadership and threshold review
Show leadership what the program found and measured, and fix what the first cycle taught you.
| Action | Owner | Artifact it produces | If it is skipped | Tool or guide |
|---|---|---|---|---|
| Fill in the monthly scorecard from the first cycle, with the count next to every rate, and write what each number led to. | Diversion program lead | Completed scorecard and a one-page summary for leadership | Leadership sees activity without outcomes, and the program has no standing agenda item to protect its resources. | Monthly program scorecard KPIs and scorecard |
| Review each threshold against what it produced: alerts raised, alerts explained, alerts that became cases. Change the ones that did not work and record why. | Diversion program lead with the pharmacy manager | Threshold change log with the reason and date for each change | Thresholds fixed in week 6 persist unexamined, and a rule that never fires or always fires keeps consuming time. | Calibrating a threshold |
| Set the next quarter's priorities from the open items, and fix the meeting and report calendar. | Executive sponsor with the compliance officer | Next-quarter plan with owners and a recurring review date | The program runs on the energy of its launch and stops when the launch team moves on. | Case index |
After day 90
At the end of week 12 you should be able to show a named owner, a scoped inventory, a reconciled user list, one indicator with a calibrated threshold, written documentation standards with trained staff, one full audit cycle and a first scorecard. Whatever is not finished goes onto the next quarter’s plan with an owner and a date. From here the monthly rhythm on the KPIs page takes over, and each new case adds a row to the case index.
Questions
Why does the plan start with scope instead of software?
Because every later step depends on knowing where controlled substances are stored, dispensed, administered and wasted, and which system records each step. A tool bought or switched on before that inventory is built on a guess about what it covers.
Can a smaller pharmacy or hospital shorten the plan?
Yes. The order matters more than the pace: scope, then access and configuration, then data, then documentation and training, then an audit cycle, then reporting. A small site may finish a phase in days, and a large system may need to run phases in parallel by unit. Keep each artifact even when the phase is short.
Which indicator should we start with?
One that your data can support today and that someone will act on. Override review is a common first choice because it uses a cabinet report the pharmacy already has, but the right first indicator depends on your cabinet fields and your risk assessment. Run it silently against history before you set a line.
What if a step shows a problem we cannot fix in 90 days?
Record it as a program risk with an owner and a review date, and carry on. The plan produces evidence of what is true, not a guarantee that everything is fixed. A known and tracked gap is better than an unknown one.
Next: download the files from the toolkit, then set your first baseline with the program self-assessment.
Last reviewed: September 2026 · Content is educational, not legal advice.