Retrospective chart review and data abstraction
A retrospective chart review collects research data from existing medical records, using a written abstraction protocol, to answer a question the records were not written for.
Fit
When it matters, and when it does not
It matters when
- The information you need is written in notes, results or orders, but is not available as reliable coded data.
- A prospective study would be too slow or impractical for the question.
- You can name the design the charts will supply, such as a retrospective cohort, a case series or a cross-sectional estimate.
It matters less when
- The key variable is recorded only now and then, such as symptoms noted only when abnormal. Collect it prospectively instead. Read: Prospective cohort study
- The data already exist as reliable coded fields in a registry or database. Use them directly. Read: Registry and routinely collected data
- Documentation practice changed during the study period in a way linked to the exposure or outcome. Comparisons across that change will be biased.
Before you start
What to check in your study
- The design the charts supplyA cohort, case-control study, case series or cross-sectional estimate. The design decides what the study can conclude.
- The source population, and how charts are foundUse something that does not depend on the outcome, such as an admission date, a procedure code or a clinic list.
- A written definition for every variableWhere in the chart to look, and what to record when it is absent, contradictory or unclear.
- A sample for a second abstractorA random subset of charts, abstracted again independently, to measure agreement on key variables.
- A missing-data plan that starts from an assumptionDecide why chart entries may be missing before choosing a method. The assumption picks the method, not the share missing.
- The variables you need were recorded, and recorded the same way for everyone.How to check: Read a sample of charts first. If documentation differs by exposure or outcome, the comparison is biased.
- Every abstractor applies each definition the same way.How to check: Train abstractors, record who abstracted each chart, and report agreement from the second abstractor's sample.
- The protocol is fixed before abstraction begins.How to check: Pilot the form on charts outside the study sample, then date the protocol and log every later change.
- Missing entries follow the assumption you named.How to check: Tests are ordered, and findings written down, for patients who looked different. So missing completely at random (MCAR) is rarely plausible. Missing at random (MAR) means the gaps depend only on recorded factors, such as age, ward and severity. Under MAR, multiple imputation can use those factors, but it is not automatically better than analyzing complete records. If a finding goes unwritten because it was normal or never looked for, missing not at random (MNAR) is plausible. Then add a tipping-point analysis, which shifts the imputed values until the conclusion changes.
Plain language
How it works
Chart review is a way to collect data, not a design. First fix the question and the design the charts will supply. Then write a protocol: each variable's definition, where to find it, and what to record when it is missing. Pilot the form, then read every eligible chart. A second abstractor rereads a random sample to measure agreement.
Worked example
A worked example
- Setting
- Charts of 300 adults who had hip fracture surgery at one hospital over two years.
- Question
- How often was delirium recorded in the week after surgery?
- Approach
- Delirium is a documented diagnosis or a positive screening result, with a written rule for notes that describe confusion without either. A second abstractor rereads a random sample.
- Result
- Delirium met the definition in 66 of 300 charts, 22 in 100.
- What it means
- This counts delirium that someone screened for or wrote down. Delirium that went unrecorded is missing, so the true share may be higher.
In practice
Recent studies PubMed search · not vetted
Watch out
Common mistakes
The one to remember
Charts were written to deliver care, not to answer the study question. A finding that was not documented is not the same as a finding that was absent, abstractors who know the hypothesis or the outcome can shift what they record, and a definition revised after reading some charts makes the data depend on the data.
- Calling the study a chart review and stopping there.Instead: Name the design the charts supply. A cohort, a case-control study and a case series support different conclusions.
- Finding charts by searching for the outcome.Instead: Find them by a rule that ignores the outcome, and report how many were screened, excluded and abstracted, with reasons.
- Leaving out charts that do not record the exposure.Instead: Those left may differ from those dropped, which is selection bias. Keep them, and use multiple imputation under a stated assumption.
- Letting one abstractor read every chart with no check.Instead: Have a second abstractor reread a random sample, and report agreement with a statistic suited to each variable.
Inside Mallard
How Mallard uses this
Plans do not link to this concept yet. A plan that uses chart data links to the design it names, such as a retrospective cohort.
concept.chart_review, the id a plan cue records.How to report it
RECORD is the guideline for studies using routinely collected health data, and it extends STROBE rather than replacing it. RECORD asks for the codes and algorithms used to classify exposures, outcomes and confounders, the data cleaning and linkage methods, and how far the investigators could reach the database population the study was drawn from, all of which an abstraction from charts must state to be reproducible. A reporting guideline is not a design manual: use it to make the study transparent, not as a substitute for careful design. Check the target journal's own requirements as well, which this page cannot know.
Related Learn pages
Related designs and methods
Evidence
Sources & further reading
- A methodology for conducting retrospective chart review research in child and adolescent psychiatry (opens in a new tab) · J Can Acad Child Adolesc Psychiatry
- The retrospective chart review: important methodological considerations (opens in a new tab) · J Educ Eval Health Prof · doi 10.3352/jeehp.2013.10.12
- Looking through the retrospectoscope: reducing bias in emergency medicine chart review studies (opens in a new tab) · Ann Emerg Med · doi 10.1016/j.annemergmed.2014.03.025
- Advanced statistics: understanding medical record review (MRR) studies (opens in a new tab) · Acad Emerg Med
- The REporting of studies Conducted using Observational Routinely-collected health Data (RECORD) statement (opens in a new tab) · PLoS Med · doi 10.1371/journal.pmed.1001885
- Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies (opens in a new tab) · BMJ · doi 10.1136/bmj.39335.541782.AD