Plan an Indicator Definition Change

Evidence and applicability: use only supplied evidence for source values, historical facts, completed checks and approvals. Cite source labels and row, page or section references. Distinguish observations, source claims, calculations, assumptions and proposals. Preserve conflicting evidence and mark missing evidence unknown or not evidenced. Do not invent values, authority, permissions, sign-off or agreement requirements. Apply only supplied applicable obligations. Keep recommendations conditional on unresolved checks and the responsible owner's decisions. Descriptive change alone does not establish program impact.

Use only materials supplied in this conversation. Ask me for the actual dated old and new definitions, relevant form/codebook extracts, effective dates, reasons for change, historical raw/summary availability, baseline/target versions, reporting obligations and known decision route. Do not treat a file name as its contents. Use safe extracts; avoid personal records. If essential inputs are absent, provide only a clearly labeled transition outline and targeted requests, not findings about an unseen indicator.

Plan the transition between the supplied old and new indicator definitions without manufacturing a continuous trend.

1. Inventory each dated definition and instrument version, effective date, reason for change, source coverage and available historical records. Separate planned changes from changes already implemented. Ask for missing old/new definitions and forms before completing a crosswalk. Record the indicator construct, unit, numerator, denominator, eligibility, population, geography, recall/reference period, collection mode, frequency, coding and missing-data rules for both versions. Flag simultaneous population or method changes as well as wording changes. Matching names do not establish comparability.
2. Produce a version-to-version crosswalk with the old field/rule, new field/rule, nature of change, source reference, likely comparability effect and unresolved evidence. Distinguish a label-only change from a changed construct, threshold, denominator or coverage. State whether a proposed mapping is exact, lossy, conditional or infeasible, and why. Do not call a mapping validated without actual validation evidence.
3. Assess historical recoding separately for each period/source. Require the necessary original variables, metadata, permissible use and reproducible transformation before proposing a recalculation. Preserve original records and published version-specific totals; retain a derived version with transformation rules, source version and checks. Summary totals alone may not support recoding. Do not reverse-engineer missing records, impute old values to complete a trend, or silently redefine past results. Show supported recalculations only if sufficient supplied data and rules allow them; otherwise give a feasibility plan.
4. Compare feasible transition options: retain separately labeled series with a break; recode supported history; or run a justified overlap/bridge measurement. For a bridge proposal, state the question it can answer, comparable units/periods and mode, paired or parallel measurement logic, feasibility, burden, permissions to confirm and checks of the mapping across relevant groups. Distinguish a proposed bridge from completed evidence. Do not invent a conversion factor, assume overlap proves equivalence, or recommend collecting sensitive data solely to make a series continuous. If a bridge is not feasible or does not establish a defensible mapping, retain the break.
5. Keep baseline, target and milestone values attached to their original definition/version and date. Show any proposed rebasing or target revision separately with its evidence, rationale and decision status. Check actual reporting/change-control obligations and approval route if provided; otherwise mark them unknown and identify who must clarify them without inventing a role. A recalculation does not approve a revised target or authorize replacing a published figure.
6. Deliver the crosswalk, period-by-period recoding feasibility table, options and recommended transition with stated conditions, and change-control record. Include version labels/effective dates, original versus derived series, documented checks, unresolved items, actual or proposed responsible roles, decision status and reporting/chart wording that makes any series break visible. Do not report a cross-version change or trend as comparable until its basis is evidenced. Explain which claims remain unsupported and what specific evidence could resolve them.