A medicine passes its quality tests. An aircraft component is replaced. A financial transaction receives approval.
Each event generates data. Yet regulated activities also require procedures, review records, explanations, and signatures. Why has this approach to evidence become so established?
Across these activities, a common need emerges: someone must be able to examine the work later and understand what happened, whether the applicable requirements were met, and who took responsibility.
Healthcare is one example of a much broader pattern.
| Industry or regulated activity | Established forms of evidence | Supporting source |
|---|---|---|
| Healthcare | Clinical notes, treatment records, examination results, and referral records | Japan’s Ministry of Health, Labour and Welfare: Guidelines on Providing Medical Information |
| Pharmaceutical manufacturing | Production and laboratory records, review records, metadata, and audit trails | FDA: Data Integrity and Compliance With Drug CGMP |
| Aviation maintenance | Descriptions of completed work, completion dates, and authorized sign-offs | 14 CFR § 43.9: Maintenance Records |
| Food production | Hazard analyses, HACCP plans, monitoring records, and corrective-action records | FDA: HACCP Principles & Application Guidelines |
| Public-company financial reporting | Evidence supporting management’s assessment of internal controls | SEC: Management’s Report on Internal Control Over Financial Reporting |
These sources describe different obligations and frameworks; they do not establish one universal documentation rule. Together, however, they show how widely evidence preservation is embedded in consequential work.
The role of documentation is particularly explicit in food safety. The HACCP principles published by the FDA include:
“Establish record-keeping and documentation procedures.”
The accompanying guidance includes the reasoning behind hazard identification and control measures among the records to retain. This means preserving both operational observations and the basis for the safety plan. — HACCP Principles & Application Guidelines, Principle 7.
Financial reporting follows a comparable logic. In its 2003 final rule, the SEC explains that:
“a company must maintain evidential matter, including documentation”
That evidence supports management’s assessment of whether internal controls are effective. A conclusion about the controls needs a basis that others can examine. — SEC Release No. 33-8238, discussion of methods of evaluation.
Why does this practice become so deeply established? These examples suggest four reinforcing reasons.
First, the final outcome cannot fully establish the quality of the process. A successful result may coexist with a process failure. An adverse outcome may occur despite a reasonable decision. Evaluating the work requires evidence of the conditions, information, and actions that preceded the result.
Second, responsibility crosses organizational boundaries. People change shifts, suppliers change, and reviewers arrive after the original team has moved on. Records preserve knowledge that would otherwise disappear with the people who held it.
Third, verification often happens after the fact. An inspector or auditor cannot observe every activity directly. Documentation makes selected aspects of that activity available for later scrutiny.
Fourth, recordkeeping becomes part of the work itself. Once training, review, approval, and inspection processes depend on particular evidence, producing that evidence becomes an operational requirement. Its persistence reflects both its practical value and its integration into oversight.
This brings us to a more precise question: what is missing from numbers and operational logs?
The FDA’s pharmaceutical data-integrity guidance offers a useful starting point:
“Metadata is the contextual information required to understand data.”
A measurement needs information such as its units, acquisition time, source, and associated material. The guidance explains that data and the metadata needed to reconstruct the activity should be retained together. — FDA Data Integrity Guidance, Question 1(b).
Even a contextualized measurement may leave the decision unexplained.
Consider a hypothetical pharmaceutical production run. A sensor records a temperature excursion. The system logs an alert, an acknowledgment, and a later batch-release action.
Those records establish a sequence. A reviewer may still need to know:
- Was the excursion a real process deviation or a sensor fault?
- Which acceptance criteria and procedure versions applied?
- What evidence supported the assessment of product impact?
- Why was release considered acceptable?
- Who reviewed the assessment and accepted responsibility?
An investigation record can connect those questions to the measurements and actions. Its value comes from preserving reasoning that the event log did not capture.
Approval also requires context. Aviation regulation makes the scope of a maintenance signature explicit:
“The signature constitutes the approval for return to service only for the work performed.”
That wording connects the signature to a defined body of work. It illustrates why a generic “approved” event can be ambiguous unless the system preserves what was approved and by whom. — 14 CFR § 43.9(a)(4).
This does not mean every explanation must live in a separate document. A structured electronic workflow can capture the applicable procedure, supporting evidence, decision rationale, exception handling, and approval scope. Logs can be designed to retain much richer context.
Documents also have weaknesses. A polished report can omit inconvenient evidence, contain copied text, or offer an explanation unsupported by the original records. Narrative needs to remain connected to trustworthy source data.
The design question is therefore: can a qualified reviewer reconstruct the work and evaluate the decision from the evidence retained?
Measurements contribute observations. Logs contribute recorded events. Procedures establish expectations. Assessments preserve reasoning. Approvals assign responsibility.
Document-based practices are deeply rooted in regulated industries because they help connect these elements across people, organizations, and time. Better digital systems can carry that same function—provided they preserve the context that makes evidence understandable.