How to organize years of blood test results

To organize years of blood test results, keep two linked layers: an archive of the original reports and a working table with one row per result. Give every source file an ID, copy dates and values exactly as reported, preserve units and reference ranges, verify each row against the source, and create smaller views only after the master table is trustworthy.

This is a common problem, not a personal filing failure. In a 2024 national survey, ASTP/ONC reported that 59% of individuals had multiple online medical records or portals, while only 7% reported using an organizing app to combine records from different portals.

The useful goal is not a perfect spreadsheet or a self-diagnosis system. It is a record where you can answer three questions: Where did this result come from? Did I copy it correctly? What do I want a qualified clinician to review?

The six-step workflow

Use this order:

1. Collect: save original portal downloads, PDFs, scans, and screenshots.
2. Inventory: assign each source an ID and record its organization, date coverage, and missing pieces.
3. Extract: create one row for each reported test result.
4. Preserve: keep the exact test name, value, unit, range, flag, collection date, and source.
5. Verify: compare every completed row with the original report before relying on it.
6. Use: filter the verified table into a focused timeline or question list for a visit.

The source archive is the evidence. The table is a working index. Keeping those roles separate lets you correct or reorganize the table without losing the report behind it.

Start with a 30-minute first pass

Do not begin by extracting every result you have ever received. Prove the workflow on a small set first.

• First 10 minutes: collect five to ten recent or frequently referenced lab reports into one inbox.
• Next 10 minutes: assign source IDs and note each file's organization, collection-date coverage, and format.
• Final 10 minutes: choose one repeated test, copy its rows, and verify each row against the report.

Stop after one useful mini-timeline. You now have a working method and can add older records in later sessions. If a report is missing, record the gap instead of postponing the entire project.

Build a source archive before a result table

Start with records you already have. ONC's Blue Button guidance explains that health information may be held by doctors, hospitals, laboratories, pharmacies, insurers, and other organizations, and that portals may provide a way to view or download it.

Look for:

• Portal lab-result pages and downloadable reports
• Laboratory PDFs from commercial, hospital, or specialty labs
• Visit notes that state why a test was ordered
• Clinician messages about follow-up or repeat testing
• Scans or photographs of older paper reports
• Existing spreadsheets whose rows still need source verification

With limited exceptions, HIPAA gives people a right to access information in designated record sets held by covered providers and health plans. HHS specifically includes clinical laboratory reports among the broad types of information that may be accessed. That does not mean every organization or missing document is covered, immediately available, or already present in a portal.

Keep the downloaded original. If you want a clearer filename, rename a working copy or record a friendly name in the inventory. A stable source ID such as LAB-001 is more important than a perfect folder taxonomy.

Inventory the sources before extracting values

Create one inventory row per file, portal page, or paper report. At minimum, record:

• Source ID: a unique label such as LAB-001
• Organization: the lab, hospital, clinic, or portal
• Collection-date coverage: one date or a date range, with uncertainty labeled
• Format: PDF, portal page, screenshot, scan, paper, or structured export
• Location: the filename or folder where the source can be found
• Extraction status: not started, partial, extracted, or verified
• Missing-record note: the period, panel, or report you still need to request

This inventory answers “What do I actually have?” before the result table creates an illusion of completeness. A missing period should remain visible as a gap, not silently disappear between two dates.

Use one row per reported result

A report can contain many tests. Put each result on its own row so you can filter one marker over time while still tracing the row back to its source.

Use these columns:

• Source ID: links the row to the inventory and original report
• Collection date: when the specimen was collected, with precision noted
• Exact test name: copied from the report rather than renamed from memory
• Value and unit: stored in separate columns but copied together
• Reference range: exactly as shown on that report
• Flag: high, low, abnormal, normal, or blank as reported
• Lab or provider: the reporting source
• Source file: the exact filename or portal location
• Context note: documented facts such as fasting status or a medication change
• Question for review: one thing you want a clinician to consider
• Verified against source: no until you complete the verification pass

Download the blood-test results CSV template. It includes these fields and one placeholder row that can be removed after you understand the structure.

Preserve reported facts before normalizing anything

You can standardize date formatting and filenames. Do not overwrite the reported value, unit, test name, range, or flag with a preferred version.

MedlinePlus explains that laboratories may use different testing methods and reference ranges, results may use different units, and the applicable reference range is the one on the reader's own report. MedlinePlus also cautions that an in-range value is not a guarantee of health and an out-of-range value may or may not indicate a health problem.

For each row:

• Keep the value and unit attached to each other.
• Keep the reference range and flag from the same report.
• Preserve less-than or greater-than signs, text results, and qualifiers.
• Label uncertain dates instead of inventing a day.
• Record a corrected result as a correction tied to the source, not as a silent replacement.
• Ask whether two results are comparable instead of assuming that similar names mean identical tests.

For deeper cross-lab questions, use how to compare Quest and Labcorp results over time.

Verify every extracted row

Extraction is not finished when the table is filled. It is finished when the row has been checked against the source.

Open the report and compare:

1. Collection date
2. Exact test name
3. Value, including decimal places or inequality signs
4. Unit
5. Reference range
6. Flag or report comment
7. Source ID and filename

Mark verified_against_source as yes only after that pass. If a value is unclear, leave it unresolved and add a note. A blank or uncertain field is more honest than a confident transcription you cannot support.

AI or optical character recognition can help create a draft table, but it should not be the verification step. Compare extracted rows with the original report yourself. The related guide on what to give ChatGPT before asking about lab results shows how to keep later AI work bounded and source-aware.

Build views from the master table

Do not create separate, disconnected spreadsheets for every question. Keep one verified master table and filter it into temporary views.

Useful views include:

• One repeated marker over a selected date range
• All results from one encounter or ordering clinician
• Rows that still need source verification
• Records from a lab whose units or ranges changed
• A short list of abnormal flags to ask about, without interpreting them
• Missing periods or reports that need follow-up

For a marker-specific example, see how to track ferritin, ApoB, and A1c over time. For an appointment-ready document built from the record, use how to prepare a health summary for a new doctor.

Keep context factual

Useful context notes may include:

• Fasting status when the report or clinician records it
• A medication or supplement start, stop, or dose change
• A documented illness, procedure, pregnancy, transfusion, or other event
• The stated reason the test was ordered
• The appointment or clinician associated with the result

Write “medication started two weeks before collection” rather than “medication caused the change.” A timeline preserves sequence. It does not establish diagnosis, cause, or treatment effect.

Add a privacy checkpoint before choosing a tool

Blood-test reports contain sensitive information. Before uploading them to any record, spreadsheet, storage, or AI service, decide what the tool actually needs and review how it handles your data.

ONC's current consumer guidance recommends checking an app's privacy policy, how it stores and protects information, what control the user has, how records can be shared, and whether the developer may use or sell information.

That review is a checkpoint, not a guarantee. A privacy policy does not by itself prove that a tool is secure, compliant, or appropriate for your records. Avoid placing identifiable health information in a tool until you understand and accept its terms and data practices.

Questions people ask about organizing lab results

Should I combine results from different laboratories?

Keep them in one source-linked master table, but do not erase the lab, unit, range, method notes, or exact test name. Ask a clinician whether particular results are comparable before drawing conclusions from a combined view.

Should I convert every result to the same unit?

Preserve the original reported value and unit. If a qualified professional wants a converted view, store it in separate clearly labeled fields and keep the source facts unchanged.

What if I cannot find an old report?

Add the missing period or test to the source inventory. Continue with verified records you do have, then use the missing-record list to guide portal searches or formal requests.

Can an AI tool build the table for me?

It can help draft rows from files you are permitted and comfortable sharing, but every row still needs comparison with the original report. Do not use an AI-generated table as a diagnosis or as proof that the archive is complete.

What Libby helps with

Libby supports the record-organization layer: keeping source files, copied lab facts, timelines, context notes, and open questions together. It can help replace “I know I saw this result somewhere” with a source-linked record you can revisit.

Libby does not turn a flag into a diagnosis, decide whether two tests are clinically comparable, or tell you what treatment to choose. Use the organized record to prepare better questions and verify important interpretation with a qualified clinician.

Safety boundaries

Organizing blood-test results can make information easier to retrieve and review. It does not prove that the archive is complete, diagnose a condition, rule out a concern, explain why a value changed, or determine treatment.

Keep these boundaries clear:

• Do not detach a result from its unit, range, source, or collection date.
• Do not treat an abnormal flag as a diagnosis.
• Do not treat an unflagged result as proof that everything is fine.
• Do not infer cause from events that appear before or after a result.
• Do not change medications, supplements, diet, testing, or treatment based only on a table, app, or AI answer.
• Discuss interpretation and medical decisions with a qualified clinician.

If a symptom or result may be urgent, contact a clinician, urgent care, emergency services, or your local emergency number instead of waiting to finish the archive.

References

• ASTP/ONC: Individuals' Access and Use of Patient Portals and Smartphone Health Apps, 2024
• ONC: Blue Button
• HHS: What health information individuals have a right to access
• MedlinePlus: How to Understand Your Lab Results
• ONC: Use It