Part 2 historical-date applicability

Editorial approval scope: The team checked current source fidelity, scope boundaries, dates, arithmetic, reader usefulness, practical workflow, and general-information limitations.

Current rule checkpoint

The live 42 CFR 2.12 applicability rule is displayed by eCFR as current through August 20, 2026 and last amended August 13, 2026. It retains March 20, 1972 for substance-use-disorder information and May 13, 1974 for alcohol-use-disorder information. The HHS final-rule fact sheet confirms that compliance with the 2024 amendments was required by February 16, 2026. Use the current text and a record-specific history rather than an old policy summary.

The threshold depends on information type

42 CFR 2.12 states separate dates for SUD and alcohol-use-disorder information. Record the acquisition date, program, service, purpose, information type, treatment episode, maintenance history, and current repository.

Restrictions have related but distinct scopes

The record-use and disclosure restriction and the patient criminal-use restriction use closely related historical rules. Classify the proposed action and information under the applicable paragraph rather than assigning one broad legacy flag.

Migration can erase needed date evidence

Preserve original system, creation or acquisition date, episode links, program status, imports, scans, conversions, source metadata, and uncertainty. Route unsupported dates to qualified Part 2 counsel before release or use.

Separate the information from the record container

The dates apply to information obtained by a federally assisted program, not simply to a paper chart's cover date or the day a file was scanned. Classify the information type, date obtained, obtaining program, reason it was obtained, treatment episode, and present maintenance. A later transcription, import, correction, or index entry does not automatically replace the underlying acquisition date.

The rule also reaches older information when a covered program maintains it in connection with a treatment episode that continues past the pertinent threshold. Document the link between the older information and the ongoing episode. Do not infer continuity only because records now share one medical-record number.

Reconstruct provenance before deciding

Start with the original paper or source system when available. Compare admission and discharge records, service dates, author signatures, program rosters, conversion logs, accession dates, archive labels, billing history, and documented mergers. Record what is established, what is inferred, and what remains unknown.

For bulk migrations, create a decision table with data set, information category, earliest and latest supported acquisition dates, source program, episode evidence, federally assisted status, current custodian, classification, reviewer, and review date. Keep the evidence with the decision so a later user can understand why the flag exists.

Preserve uncertainty in modern systems

Avoid replacing an unknown historical date with a convenient system default. Store a precision level such as exact day, month, year, range, before threshold, after threshold, or unknown. Keep the original value and conversion notes. Restrict unresolved material while privacy and counsel determine the safest handling.

Test downstream tools after migration. Search, release, analytics, payer, legal, patient-access, retention, and deletion workflows should use the supported classification without exposing the protected fact through file names, queues, or error messages. Reassess after a repository merge, new source discovery, or episode correction.

Example

Twelve legacy data sets are reviewed. Nine have supported information type, date, program, episode, and maintenance evidence; three lack source dates. Classification completeness is 9 of 12.

Turn the review into a release decision

For each data set, record one outcome: supported as within the historical rule, supported as outside it, restricted pending evidence, or escalated for legal review. Name the permitted next action, system control, responsible owner, missing evidence, and reconsideration date. This prevents a research note from being mistaken for release approval.

Sample records from every outcome and compare the decision with the source evidence. Include records near each threshold and older information carried into a continuing episode. If the sample exposes a conversion default, mixed episode, or unsupported source, widen the review before relying on the classification.

Historical-date review checklist

• identify whether the information concerns SUD generally or alcohol use specifically;
• preserve the date obtained, source program, service, episode, and maintenance evidence;
• assess the March 20, 1972 or May 13, 1974 threshold that fits the information;
• document whether older information remains tied to an episode continuing past that date;
• retain original values, date precision, conversion history, reviewer, and unresolved questions; and
• test releases and other uses after migration or reclassification.

This historical rule does not establish that a program is federally assisted, that a proposed use or disclosure is allowed, or that another privacy law is inapplicable. Program status, current Part 2 provisions, HIPAA, state law, and the particular action require qualified review.