How to Build a Radiology Report Timeline and Find the Recommendation Nobody Followed (Free Tool)

By John Mahoney · December 29, 2026 · 7 min read

A delayed-diagnosis case is rarely about one radiology report. It is about a sequence of reports in which a finding appeared, was described, was recommended for follow-up, and then was not followed. The chest CT that mentioned a 7 mm nodule in the findings but not the impression. The "recommend follow-up in 6 to 12 months" that no one ordered. The third study, two years later, that uses the word "mass." Laying those reports on one timeline, finding by finding, is the first hour of work in any missed-finding case, and it decides what the expert is asked to look at. This post shows how to do it and where the free tool fits.

What an imaging history actually contains

Pull every report for the body region in question, not just the one named in the complaint. For each, you are reading four things:

• The study header. Date, modality, contrast, the indication the ordering clinician gave, and which prior studies the radiologist compared to. A report that says "no priors available" when three exist is itself a fact.
• Findings. The structure-by-structure description. This is where incidental findings live.
• Impression. The summary most clinicians read. A finding that made it into the findings but not the impression has a much lower chance of being acted on.
• Recommendations and communication. "Recommend follow-up CT in 3 months." "Findings discussed with Dr. X by telephone at 14:32." Or nothing.

Pathology and echo reports belong on the same timeline when they bear on the same question: a biopsy result that arrived after the imaging, an ejection fraction that fell between two echos.

What to look for

1. The finding's path across studies

Give each distinct finding a row and each study a column. Fill in what each report said about it, in the report's words: "7 mm nodule, right upper lobe, likely benign" becomes "9 mm, indeterminate" becomes "2.4 cm spiculated mass." Growth, change in description, and the point at which the word changes are what the expert will test against the images.

2. Recommendations with no documented follow-up

For every recommendation in every report, look for the order, the scheduled appointment, and the completed study. When none appears in the record you have, write "no follow-up found in record reviewed," not "no follow-up." The distinction matters later; the gap may be a records gap rather than a care gap.

3. Findings-vs-impression discrepancies

Compare the two sections of each report line by line. A finding present in one and absent from the other, or described with different severity, is a discrepancy. These are common and not always consequential, but they are the mechanism by which an abnormal study reads as normal to a busy clinician.

4. Communication of unexpected or urgent findings

The ACR Practice Parameter for Communication of Diagnostic Imaging Findings describes circumstances in which direct, non-routine communication with the treating clinician may be warranted, including findings that need urgent intervention and unexpected findings that could be seriously adverse to the patient's health. It also notes that the referring provider shares responsibility for obtaining and acting on results. Whether any particular report called for a phone call is an expert question; whether a call is documented is a records question you can answer now.

5. Preliminary reads and addenda

Overnight preliminary reads, resident reads and later addenda each carry their own time stamp. A finding added by addendum two days after the patient was discharged is a different fact from the same finding in the original report.

A worked example (hypothetical)

The facts below are invented for illustration. No real patient, radiologist or case is described.

A 61-year-old man is diagnosed with stage III lung cancer. His attorney obtains four years of imaging. Three chest studies predate the diagnosis: a CT angiogram for suspected pulmonary embolism, a follow-up chest CT fourteen months later for a cough, and an abdominal CT that included the lung bases.

Pasted into the tool, the finding-progression row for the right lower lobe reads: Study 1, findings: "6 mm solid nodule, right lower lobe; no prior for comparison"; impression: negative for pulmonary embolism, with no mention of the nodule. Study 2, findings: "nodule right lower lobe now 11 mm"; impression: "11 mm RLL nodule, recommend PET-CT or tissue sampling." Study 3, the abdominal CT, lung bases: "partially imaged RLL nodule, see dedicated chest imaging."

The recommendations box lists the Study 2 PET-CT recommendation with no order, appointment or result anywhere in the pasted record. The discrepancy box flags Study 1: a 6 mm nodule in the findings and nothing in the impression. The gaps list asks for the critical-result log and the ordering provider's acknowledgement for Study 2, and the images for all three.

For plaintiff counsel, that is a sequence with two candidate failures, fourteen months apart, by two different physicians. For defense counsel, it says which report will be the focus, that the Study 1 read may be defended on the size and the indication, and that the Study 2 follow-up is a question for the ordering clinician's office before it is a question for the radiologist. Either way the expert gets the images for all three studies, not just the last one.

What to request next

• The prior comparison studies each report names, and any earlier studies of the same region that were available but not compared.
• The images themselves, in DICOM form, for every study, so an expert can re-read them.
• The radiology department's critical-result or unexpected-finding communication log for each study.
• The ordering provider's acknowledgement of each result: inbox sign-off, result note or portal read receipt, with time stamps.
• Any preliminary, wet or resident read, the final report and the addendum history.
• Orders placed after each report, to show whether the recommended follow-up was ordered, scheduled, cancelled or completed elsewhere.

What the tool does not do

The tool organizes what the pasted reports say. It does not look at images, decide whether a finding should have been called, or judge whether a recommendation was appropriate. "No follow-up found" means no follow-up in the text you pasted. Every quote it produces is drawn from your text; verify each against the report before it goes into a chronology or an expert letter.

Frequently asked questions

What is a findings-vs-impression discrepancy in a radiology report?

The findings section of a report describes what the radiologist saw, structure by structure. The impression is the summary the ordering clinician usually reads. When a finding appears in the body of the report but not in the impression, or is described differently in the two sections, a clinician reading only the impression can miss it. In delayed-diagnosis cases that gap is often where the case starts.

What does the free Imaging Report Timeline actually do?

You paste radiology, pathology or echocardiogram report text from one or more studies. It returns a dated timeline, a table showing how each finding progressed across studies, the recommendations the reports made and whether the pasted record shows any documented follow-up, findings-vs-impression discrepancies, and the records to request next. Every item is quoted from the report. It does not re-read images or decide whether a read was reasonable.

Which records do I need beyond the radiology reports?

The actual images (DICOM) so an expert can re-read them, the prior comparison studies each report names, the department's critical-result communication log, the ordering provider's acknowledgement of each result, any preliminary or resident read with the addendum history, and the orders placed after each report showing whether the recommended follow-up was ordered, scheduled or cancelled.

Related: Radiology missed-diagnosis cases, why radiologists get sued and building a lab value timeline.