Why English? | Reporting RECORD-PE Pharmacoepidemiology Through Drug Exposure, Risk Windows, Switching, Confounding and Sensitivity Analysis

ORIENTATION

What this guide owns

Current fact check

RECORD-PE is a 2018 extension for pharmacoepidemiological studies using routinely collected health data. The publication states that the 15-item checklist should be used alongside RECORD and STROBE.

Canonical boundary

This article owns reporting of routine-data pharmacoepidemiology with RECORD-PE. It does not offer medication advice, replace RECORD or STROBE, validate database algorithms, remove confounding or cover randomised medicine trials.

Worked example

Every chapter returns to a fictional observational study comparing two medicine strategies using linked prescribing, dispensing and hospital records. It is fictional, low-stakes and educational; it does not report a real school, clinical decision or guaranteed outcome.

Warm up with eduKate’s guides to explaining scientific evidence and asking better questions in class.

CHAPTER 1 OF 13

1. Identify the Pharmacoepidemiological Design

Back to contents

Why this move matters

Name the observational design, medicines, data sources, population, period and principal question in the title or abstract.

Translate records into study variables

Name the observational design, medicines, data sources, population, period and principal question in the title or abstract. Explain how codes, prescriptions, dispensings, dates and linkages became exposure, outcome, covariate and follow-up variables. Routine collection does not make these definitions self-evident.

Trace one medicine timeline

Use The report identifies a new-user comparative cohort study using linked records. Follow eligibility, washout, initiation, supplied days, gaps, switching, outcome window, censoring and sensitivity analyses with calendar dates and explicit assumptions.

Expose database boundaries

Describe why the data were created, what is missing, how linkage works and which algorithms were validated. A large database can still misclassify drug use or outcomes systematically.

Protect the distinction between reporting and causality

A design label cannot reveal every eligibility and timing decision. RECORD-PE improves transparency of pharmacoepidemiology using routine data; it does not remove confounding, selection bias or measurement error.

Student, parent and career transfer

Make one evidence card for “Identify the Pharmacoepidemiological Design”: source location, plain-language meaning, limit and next question. A parent or teacher can ask for one everyday explanation and one specialist explanation; both must preserve the same conditions. This habit supports pharmacoepidemiology, health-data science, pharmacy, regulatory research, epidemiology and evidence appraisal without promising a grade, admission, funding decision or job.

CHAPTER 2 OF 13

2. Describe Why Each Data Source Exists

Back to contents

Why this move matters

Explain clinical, billing, dispensing, registry or administrative purposes, coverage, coding practices and dates of availability.

Translate records into study variables

Explain clinical, billing, dispensing, registry or administrative purposes, coverage, coding practices and dates of availability. Explain how codes, prescriptions, dispensings, dates and linkages became exposure, outcome, covariate and follow-up variables. Routine collection does not make these definitions self-evident.

Trace one medicine timeline

Use Prescribing records show clinician orders while dispensing records show pharmacy supply. Follow eligibility, washout, initiation, supplied days, gaps, switching, outcome window, censoring and sensitivity analyses with calendar dates and explicit assumptions.

Expose database boundaries

Describe why the data were created, what is missing, how linkage works and which algorithms were validated. A large database can still misclassify drug use or outcomes systematically.

Protect the distinction between reporting and causality

Neither source proves ingestion, and their missingness can differ. RECORD-PE improves transparency of pharmacoepidemiology using routine data; it does not remove confounding, selection bias or measurement error.

Student, parent and career transfer

Make one evidence card for “Describe Why Each Data Source Exists”: source location, plain-language meaning, limit and next question. A parent or teacher can ask for one everyday explanation and one specialist explanation; both must preserve the same conditions. This habit supports pharmacoepidemiology, health-data science, pharmacy, regulatory research, epidemiology and evidence appraisal without promising a grade, admission, funding decision or job.

CHAPTER 3 OF 13

3. Show Linkage and Cohort Construction

Back to contents

Why this move matters

Report identifiers or linkage method, quality, exclusions, eligibility dates and a diagram of the study timeline where helpful.

Translate records into study variables

Report identifiers or linkage method, quality, exclusions, eligibility dates and a diagram of the study timeline where helpful. Explain how codes, prescriptions, dispensings, dates and linkages became exposure, outcome, covariate and follow-up variables. Routine collection does not make these definitions self-evident.

Trace one medicine timeline

Use The cohort requires linked prescribing and hospital coverage for a defined baseline period. Follow eligibility, washout, initiation, supplied days, gaps, switching, outcome window, censoring and sensitivity analyses with calendar dates and explicit assumptions.

Expose database boundaries

Describe why the data were created, what is missing, how linkage works and which algorithms were validated. A large database can still misclassify drug use or outcomes systematically.

Protect the distinction between reporting and causality

Linkage failure may select a different population. RECORD-PE improves transparency of pharmacoepidemiology using routine data; it does not remove confounding, selection bias or measurement error.

Student, parent and career transfer

Make one evidence card for “Show Linkage and Cohort Construction”: source location, plain-language meaning, limit and next question. A parent or teacher can ask for one everyday explanation and one specialist explanation; both must preserve the same conditions. This habit supports pharmacoepidemiology, health-data science, pharmacy, regulatory research, epidemiology and evidence appraisal without promising a grade, admission, funding decision or job.

CHAPTER 4 OF 13

4. Explain How Drug Code Lists Were Developed

Back to contents

Why this move matters

Name dictionaries, substance names, routes, classification levels, manual or automated searches, review and access to code lists.

Translate records into study variables

Name dictionaries, substance names, routes, classification levels, manual or automated searches, review and access to code lists. Explain how codes, prescriptions, dispensings, dates and linkages became exposure, outcome, covariate and follow-up variables. Routine collection does not make these definitions self-evident.

Trace one medicine timeline

Use The study reports the classification version and publishes the exact codes used for each medicine. Follow eligibility, washout, initiation, supplied days, gaps, switching, outcome window, censoring and sensitivity analyses with calendar dates and explicit assumptions.

Expose database boundaries

Describe why the data were created, what is missing, how linkage works and which algorithms were validated. A large database can still misclassify drug use or outcomes systematically.

Protect the distinction between reporting and causality

A code list can be complete in one database and incomplete in another. RECORD-PE improves transparency of pharmacoepidemiology using routine data; it does not remove confounding, selection bias or measurement error.

Student, parent and career transfer

Make one evidence card for “Explain How Drug Code Lists Were Developed”: source location, plain-language meaning, limit and next question. A parent or teacher can ask for one everyday explanation and one specialist explanation; both must preserve the same conditions. This habit supports pharmacoepidemiology, health-data science, pharmacy, regulatory research, epidemiology and evidence appraisal without promising a grade, admission, funding decision or job.

CHAPTER 5 OF 13

5. Define Initiation and Washout

Back to contents

Why this move matters

State whether participants are initiators, prevalent users or treatment-naive initiators and specify the baseline washout rule.

Translate records into study variables

State whether participants are initiators, prevalent users or treatment-naive initiators and specify the baseline washout rule. Explain how codes, prescriptions, dispensings, dates and linkages became exposure, outcome, covariate and follow-up variables. Routine collection does not make these definitions self-evident.

Trace one medicine timeline

Use New use requires twelve months without the target medicine in available records. Follow eligibility, washout, initiation, supplied days, gaps, switching, outcome window, censoring and sensitivity analyses with calendar dates and explicit assumptions.

Expose database boundaries

Describe why the data were created, what is missing, how linkage works and which algorithms were validated. A large database can still misclassify drug use or outcomes systematically.

Protect the distinction between reporting and causality

Absence in the database may reflect incomplete prior coverage. RECORD-PE improves transparency of pharmacoepidemiology using routine data; it does not remove confounding, selection bias or measurement error.

Student, parent and career transfer

Make one evidence card for “Define Initiation and Washout”: source location, plain-language meaning, limit and next question. A parent or teacher can ask for one everyday explanation and one specialist explanation; both must preserve the same conditions. This habit supports pharmacoepidemiology, health-data science, pharmacy, regulatory research, epidemiology and evidence appraisal without promising a grade, admission, funding decision or job.

CHAPTER 6 OF 13

6. Define Exposure Start, Duration and Grace Periods

Back to contents

Why this move matters

Explain prescription or dispensing dates, assumed daily dose, supplied days, stockpiling, refill gaps and grace periods.

Translate records into study variables

Explain prescription or dispensing dates, assumed daily dose, supplied days, stockpiling, refill gaps and grace periods. Explain how codes, prescriptions, dispensings, dates and linkages became exposure, outcome, covariate and follow-up variables. Routine collection does not make these definitions self-evident.

Trace one medicine timeline

Use Exposure begins at dispensing and continues through recorded supply plus a prespecified grace period. Follow eligibility, washout, initiation, supplied days, gaps, switching, outcome window, censoring and sensitivity analyses with calendar dates and explicit assumptions.

Expose database boundaries

Describe why the data were created, what is missing, how linkage works and which algorithms were validated. A large database can still misclassify drug use or outcomes systematically.

Protect the distinction between reporting and causality

Administrative supply is an imperfect measure of medicine taken. RECORD-PE improves transparency of pharmacoepidemiology using routine data; it does not remove confounding, selection bias or measurement error.

Student, parent and career transfer

Make one evidence card for “Define Exposure Start, Duration and Grace Periods”: source location, plain-language meaning, limit and next question. A parent or teacher can ask for one everyday explanation and one specialist explanation; both must preserve the same conditions. This habit supports pharmacoepidemiology, health-data science, pharmacy, regulatory research, epidemiology and evidence appraisal without promising a grade, admission, funding decision or job.

CHAPTER 7 OF 13

7. Choose and Justify Risk Windows

Back to contents

Why this move matters

Connect biological or clinical reasoning to latency, current, recent, former or cumulative exposure windows.

Translate records into study variables

Connect biological or clinical reasoning to latency, current, recent, former or cumulative exposure windows. Explain how codes, prescriptions, dispensings, dates and linkages became exposure, outcome, covariate and follow-up variables. Routine collection does not make these definitions self-evident.

Trace one medicine timeline

Use The primary window begins after a defined latency and ends after the final covered day. Follow eligibility, washout, initiation, supplied days, gaps, switching, outcome window, censoring and sensitivity analyses with calendar dates and explicit assumptions.

Expose database boundaries

Describe why the data were created, what is missing, how linkage works and which algorithms were validated. A large database can still misclassify drug use or outcomes systematically.

Protect the distinction between reporting and causality

A convenient window can misclassify time at risk. RECORD-PE improves transparency of pharmacoepidemiology using routine data; it does not remove confounding, selection bias or measurement error.

Student, parent and career transfer

Make one evidence card for “Choose and Justify Risk Windows”: source location, plain-language meaning, limit and next question. A parent or teacher can ask for one everyday explanation and one specialist explanation; both must preserve the same conditions. This habit supports pharmacoepidemiology, health-data science, pharmacy, regulatory research, epidemiology and evidence appraisal without promising a grade, admission, funding decision or job.

CHAPTER 8 OF 13

8. Report Switching, Add-on Treatment and Discontinuation

Back to contents

Why this move matters

Define treatment changes, overlaps, combination use, censoring and how follow-up continues after each event.

Translate records into study variables

Define treatment changes, overlaps, combination use, censoring and how follow-up continues after each event. Explain how codes, prescriptions, dispensings, dates and linkages became exposure, outcome, covariate and follow-up variables. Routine collection does not make these definitions self-evident.

Trace one medicine timeline

Use A switch requires supply of the alternative before the first medicine’s grace period ends. Follow eligibility, washout, initiation, supplied days, gaps, switching, outcome window, censoring and sensitivity analyses with calendar dates and explicit assumptions.

Expose database boundaries

Describe why the data were created, what is missing, how linkage works and which algorithms were validated. A large database can still misclassify drug use or outcomes systematically.

Protect the distinction between reporting and causality

Different switching rules can change both groups and person-time. RECORD-PE improves transparency of pharmacoepidemiology using routine data; it does not remove confounding, selection bias or measurement error.

Student, parent and career transfer

Make one evidence card for “Report Switching, Add-on Treatment and Discontinuation”: source location, plain-language meaning, limit and next question. A parent or teacher can ask for one everyday explanation and one specialist explanation; both must preserve the same conditions. This habit supports pharmacoepidemiology, health-data science, pharmacy, regulatory research, epidemiology and evidence appraisal without promising a grade, admission, funding decision or job.

CHAPTER 9 OF 13

9. Describe Persistence and Adherence Measures

Back to contents

Why this move matters

State measures, numerators, denominators, thresholds, permissible gaps and assumptions about availability and use.

Translate records into study variables

State measures, numerators, denominators, thresholds, permissible gaps and assumptions about availability and use. Explain how codes, prescriptions, dispensings, dates and linkages became exposure, outcome, covariate and follow-up variables. Routine collection does not make these definitions self-evident.

Trace one medicine timeline

Use Proportion of days covered is calculated over a fixed period with a published threshold. Follow eligibility, washout, initiation, supplied days, gaps, switching, outcome window, censoring and sensitivity analyses with calendar dates and explicit assumptions.

Expose database boundaries

Describe why the data were created, what is missing, how linkage works and which algorithms were validated. A large database can still misclassify drug use or outcomes systematically.

Protect the distinction between reporting and causality

High refill coverage does not establish correct ingestion. RECORD-PE improves transparency of pharmacoepidemiology using routine data; it does not remove confounding, selection bias or measurement error.

Student, parent and career transfer

Make one evidence card for “Describe Persistence and Adherence Measures”: source location, plain-language meaning, limit and next question. A parent or teacher can ask for one everyday explanation and one specialist explanation; both must preserve the same conditions. This habit supports pharmacoepidemiology, health-data science, pharmacy, regulatory research, epidemiology and evidence appraisal without promising a grade, admission, funding decision or job.

CHAPTER 10 OF 13

10. Define Outcomes and Validation Evidence

Back to contents

Why this move matters

Report codes, algorithms, dates, ascertainment windows and evidence for positive predictive value, sensitivity or other validation.

Translate records into study variables

Report codes, algorithms, dates, ascertainment windows and evidence for positive predictive value, sensitivity or other validation. Explain how codes, prescriptions, dispensings, dates and linkages became exposure, outcome, covariate and follow-up variables. Routine collection does not make these definitions self-evident.

Trace one medicine timeline

Use The hospital outcome algorithm is linked to its validation study and any local modification is stated. Follow eligibility, washout, initiation, supplied days, gaps, switching, outcome window, censoring and sensitivity analyses with calendar dates and explicit assumptions.

Expose database boundaries

Describe why the data were created, what is missing, how linkage works and which algorithms were validated. A large database can still misclassify drug use or outcomes systematically.

Protect the distinction between reporting and causality

Validation in another database or population may not transfer fully. RECORD-PE improves transparency of pharmacoepidemiology using routine data; it does not remove confounding, selection bias or measurement error.

Student, parent and career transfer

Make one evidence card for “Define Outcomes and Validation Evidence”: source location, plain-language meaning, limit and next question. A parent or teacher can ask for one everyday explanation and one specialist explanation; both must preserve the same conditions. This habit supports pharmacoepidemiology, health-data science, pharmacy, regulatory research, epidemiology and evidence appraisal without promising a grade, admission, funding decision or job.

CHAPTER 11 OF 13

11. Address Time-varying Confounding and Bias

Back to contents

Why this move matters

Describe measured confounders, timing, analytic methods and relevant biases such as healthy adherer or depletion of susceptibles.

Translate records into study variables

Describe measured confounders, timing, analytic methods and relevant biases such as healthy adherer or depletion of susceptibles. Explain how codes, prescriptions, dispensings, dates and linkages became exposure, outcome, covariate and follow-up variables. Routine collection does not make these definitions self-evident.

Trace one medicine timeline

Use Baseline health service use and time-varying co-medication are defined before modelling. Follow eligibility, washout, initiation, supplied days, gaps, switching, outcome window, censoring and sensitivity analyses with calendar dates and explicit assumptions.

Expose database boundaries

Describe why the data were created, what is missing, how linkage works and which algorithms were validated. A large database can still misclassify drug use or outcomes systematically.

Protect the distinction between reporting and causality

Adjustment cannot control unmeasured or poorly measured confounding. RECORD-PE improves transparency of pharmacoepidemiology using routine data; it does not remove confounding, selection bias or measurement error.

Student, parent and career transfer

Make one evidence card for “Address Time-varying Confounding and Bias”: source location, plain-language meaning, limit and next question. A parent or teacher can ask for one everyday explanation and one specialist explanation; both must preserve the same conditions. This habit supports pharmacoepidemiology, health-data science, pharmacy, regulatory research, epidemiology and evidence appraisal without promising a grade, admission, funding decision or job.

CHAPTER 12 OF 13

12. Report Sensitivity and Negative-Control Analyses

Back to contents

Why this move matters

Test alternative exposure windows, grace periods, outcome definitions, missing-data choices and other plausible assumptions.

Translate records into study variables

Test alternative exposure windows, grace periods, outcome definitions, missing-data choices and other plausible assumptions. Explain how codes, prescriptions, dispensings, dates and linkages became exposure, outcome, covariate and follow-up variables. Routine collection does not make these definitions self-evident.

Trace one medicine timeline

Use Results are recalculated with shorter and longer refill gaps and an alternative validated outcome. Follow eligibility, washout, initiation, supplied days, gaps, switching, outcome window, censoring and sensitivity analyses with calendar dates and explicit assumptions.

Expose database boundaries

Describe why the data were created, what is missing, how linkage works and which algorithms were validated. A large database can still misclassify drug use or outcomes systematically.

Protect the distinction between reporting and causality

Agreement across a few analyses does not prove absence of bias. RECORD-PE improves transparency of pharmacoepidemiology using routine data; it does not remove confounding, selection bias or measurement error.

Student, parent and career transfer

Make one evidence card for “Report Sensitivity and Negative-Control Analyses”: source location, plain-language meaning, limit and next question. A parent or teacher can ask for one everyday explanation and one specialist explanation; both must preserve the same conditions. This habit supports pharmacoepidemiology, health-data science, pharmacy, regulatory research, epidemiology and evidence appraisal without promising a grade, admission, funding decision or job.

CHAPTER 13 OF 13

13. Interpret Results as Records-Based Evidence

Back to contents

Why this move matters

State uncertainty, database limits, exposure misclassification, generalisability, funding and conflicts without turning association into personal treatment advice.

Translate records into study variables

State uncertainty, database limits, exposure misclassification, generalisability, funding and conflicts without turning association into personal treatment advice. Explain how codes, prescriptions, dispensings, dates and linkages became exposure, outcome, covariate and follow-up variables. Routine collection does not make these definitions self-evident.

Trace one medicine timeline

Use The conclusion refers to an association under the reported definitions and recommends further triangulation. Follow eligibility, washout, initiation, supplied days, gaps, switching, outcome window, censoring and sensitivity analyses with calendar dates and explicit assumptions.

Expose database boundaries

Describe why the data were created, what is missing, how linkage works and which algorithms were validated. A large database can still misclassify drug use or outcomes systematically.

Protect the distinction between reporting and causality

Large sample size can narrow random error while leaving systematic error. RECORD-PE improves transparency of pharmacoepidemiology using routine data; it does not remove confounding, selection bias or measurement error.

Student, parent and career transfer

Make one evidence card for “Interpret Results as Records-Based Evidence”: source location, plain-language meaning, limit and next question. A parent or teacher can ask for one everyday explanation and one specialist explanation; both must preserve the same conditions. This habit supports pharmacoepidemiology, health-data science, pharmacy, regulatory research, epidemiology and evidence appraisal without promising a grade, admission, funding decision or job.

PRACTICE LAB

Build a pharmacoepidemiology decision trail

Student route

1. Choose one original report inside this guide’s scope.
2. Copy its objective and underline actor, action, context, evidence and intended decision.
3. Draw the specific reasoning chain used in this article.
4. Mark each step reported, inferred, contested or unknown.
5. Rewrite the conclusion so it preserves the narrowest important limitation.

Parent or teacher route

Ask the learner to explain show how routine health records became a medicine-exposure timeline and an interpretable result twice: once to a classmate and once to a method specialist. Praise a precise correction and a well-stated unknown; both show disciplined attention to evidence.

Early-career route

Create a one-page record_pe audit with source link, date checked, scope, method logic, decisive wording, contrary evidence, limit and next action. A colleague should be able to continue without guessing.

Continue with eduKate’s guides to reading course and career information and giving useful feedback in group projects.

USEFUL NEXT READING

Connect this narrow owner to its neighbour

Read the ACCORD consensus-reporting guide for a related but different reader job. Return here when the question is specifically how to show how routine health records became a medicine-exposure timeline and an interpretable result. This boundary keeps one canonical owner per intent.

For English practice across research, school and career communication, visit the Singapore English Tuition Centre learning site.

FREQUENTLY ASKED QUESTIONS

Questions students and parents often ask

What does RECORD-PE cover?

Pharmacoepidemiological studies using routinely collected health data.

How many checklist items are there?

Fifteen, used with RECORD and STROBE.

Does a prescription prove exposure?

No. It records an order; dispensing and adherence require separate definitions.

What is a washout period?

A stated prior interval used to classify new or treatment-naive use.

Why report grace periods?

They determine when refill gaps count as continued exposure or discontinuation.

Can a large database remove confounding?

No. Size does not correct systematic bias or missing variables.