Assess Baseline Options After Implementation Starts

Evidence and applicability: use only supplied evidence for historical facts, values, quotations, completed checks and approvals. Cite source labels and row, page or section references. Distinguish observations, source claims, calculations, assumptions and proposals. Keep missing evidence unknown or not evidenced, not proof an activity never occurred. Preserve conflicts and their provenance. Do not invent data, dates, permissions, authority, agreement requirements or sign-off. Apply only supplied applicable obligations. A recommendation is conditional on unresolved checks and the responsible owner's decisions; it does not authorize collection, reuse or publication. Descriptive change alone does not establish program impact.

Use only material supplied in this conversation. Ask me for the program and exposure timeline, indicator definitions, intended decisions/comparisons, actual dated candidate records and instruments, historical coverage and permissions, available resources and reporting requirements. Prefer safe extracts. If the essential timeline or source contents are missing, retain the gap and provide only a labeled options framework and focused requests before making indicator-specific findings.

Assess the options after preintervention measurement was missed, indicator by indicator. Do not complete a baseline template with invented history.

1. Build a timeline separating administrative program launch, actual activities and first relevant exposure for each population/site/cohort, source recording dates, reference periods, and proposed measurement dates. Ask for actual dates and exposure evidence; do not infer that all units were untreated before an official launch or first survey. Where exposure timing is uncertain, retain that uncertainty. Define the decision, indicator construct, unit, population, denominator, location, reference period and intended later comparison before assessing sources.
2. Inventory the actual candidate evidence with source, custodian if known, collection and reference dates, definition/instrument version, coverage, processing history and permissible use status. Separate contemporaneous pre-exposure observations, later-extracted records about a documented pre-exposure period, post-start measurements, and retrospective reports made after exposure. An extraction date after launch does not automatically invalidate genuinely historical records, but their underlying dates, content and provenance must be checked. A prelaunch date alone does not establish relevant population coverage or lack of prior exposure. A source title alone cannot establish its contents.
3. For each indicator, assess whether authentic historical records can support the intended preintervention value. Check construct, population, denominator, geographic and seasonal coverage, measurement mode, missingness, completeness and comparability with future measurement. Distinguish direct historical measurement from a proxy. If reconstruction is feasible, give a reproducible derivation with sources and assumptions, calculate only from sufficient supplied data, and label it a reconstructed historical value with its uncertainty and limits. Preserve original records. Never backdate new collection, infer unknown values from targets or endline results, assume a missing baseline is zero, or create synthetic participant history.
4. Assess retrospective recall or other reconstruction only where it answers the question with acceptable burden, feasibility and safeguards. Specify the reference period, anchoring evidence, validation opportunities actually available, and recall, survival, selection and response-shift risks. Retrospective self-report is not automatically equivalent to contemporaneous observation, and a recall question cannot validate itself. Mark proposed interviews or verification as plans, not completed checks. Do not force a survey or reconstruction when records, a narrower question, or acknowledging the gap is more defensible.
5. Consider a prospective reference measurement from now, clearly dated and labeled post-start, and a narrower descriptive or process question. State what prior exposure it may incorporate and what future comparison it could support. For genuinely not-yet-exposed units, assess whether a pre-exposure measurement is feasible for those units without relabeling them a baseline for everyone; account for selection, spillovers, timing and cohort differences. Do not delay needed services to manufacture a baseline. Neither a delayed measurement nor an untreated comparison by itself establishes a causal counterfactual; causal questions require their own credible design and assumptions.
6. Deliver a per-indicator options matrix: evidence classification, supported historical value if calculable, reconstruction/recall feasibility, prospective alternative, comparability limits, unsupported claims, missing evidence and recommended option with conditions. Keep preintervention state explicitly unmeasured where it cannot be supported. Provide a practical next-measurement plan with roles and dates only if supplied or clearly proposed, source/definition versioning, and targeted checks. Keep changes to baselines, targets, reporting labels or evaluation claims as proposed decisions subject to the actual applicable route. Supply honest reporting wording that distinguishes historical observations, reconstructed values, retrospective reports and post-start reference values without claiming validation or causal impact.