| Verified against primary record | |
| Report structure | Clinical referral, technique, findings, impression or conclusion, and advice, with the report checked and signed[1] |
|---|---|
| Prior studies | The reporter should review relevant prior studies and clinical information, so a report produced without them is a known-degraded product[2] |
| Identification of the reporter | Professional status and regulatory registration details should be clear on every written report[2] |
| Independently reported | |
| Whether images travel with the report | In the dominant cross-enterprise sharing design, no: the report is submitted as a document while the images are registered only as a reference[3] |
| Bands apply only to the rows beneath them. The two professional standards cited carry no publication or review date on the pages consulted, and a third edition of one of them exists which was not read. | |
Radiology reports are the radiologist’s written interpretation of an imaging study. For a patient treated abroad the report usually comes home and the images usually do not, and that asymmetry has an architectural cause rather than an administrative one.
What a report should contain
A European professional body sets out a five-part structure. The clinical referral section summarises the clinical problem, with a statement in the report where referral information was insufficient and a radiation justification where applicable. The technique section records the procedure, contrast and any reaction, technically suboptimal features and patient dose. The findings section describes systematically, with size, density or signal, anatomical location, image references, relevant negatives, incidental findings and comparison with prior studies including their dates. The impression gives a precise diagnosis or a ranked limited differential tied to the clinical question, and the advice section sets out next steps. Reports must be checked and signed, and for urgent findings the radiologist has a duty to inform the referrer, recording the date, time, person notified and method in the report. The same body presents structured reporting as standardised, modular and template-based and favours movement toward uniformity, without mandating it.[1]
A British college’s standards add four provisions that bear directly on foreign reports. The reporter should review relevant prior studies from the same and other modalities together with clinical information, laboratory results and pathology reports. Long descriptive reports should close with a short summary of key findings, interpretation and advice. Ad hoc reviews and second opinions should be recorded as supplementary reports or addenda, and where the interpretation changes significantly the referrer must be told. And professional status and regulatory registration details should be clear on every written report, with the individual who signs accountable for the content.[2]
Two of those are directly usable. The provision on second opinions assumes that re-reads happen and governs how they are recorded, as an addendum rather than a replacement. The provision on identification means that a foreign report which does not identify the reporter’s regulator fails a published standard, which is a checkable test a patient can apply.
Template libraries exist to support structure. A radiological society maintains a freely downloadable library of more than 200 best-practice report templates with standard headings and terms mapped to controlled vocabularies, developed from a consensus workshop and clinical subcommittees, with a template-exchange profile approved so templates can be moved between reporting systems regardless of vendor.[4]
Why the images stay behind
The imaging interchange standard makes a disc or a transfer technically readable. It specifies network communication protocols with the structure and meaning of commands and associated information, media storage services, a file format and a medical directory structure for accessing images and related data on interchange media, and the information that must accompany any implementation claiming conformance. It expressly does not specify implementation details, the full feature set of an integrated system, or any testing or validation procedure for assessing conformance.[5] So the standard guarantees a common format and guarantees nothing about whether a receiving hospital’s system is configured to ingest a particular disc, because conformance testing is outside its scope.
The sharing architecture explains the rest. In the dominant profile for sharing imaging across enterprises, the report is submitted to a document repository as a portable document or a text document and can be found and retrieved by any consumer, whereas the images are not transferred to the repository at all: the imaging source registers a manifest that references the image objects, and a consumer must query the registry, read identifiers from the manifest, then retrieve the images from the source.[3] The report is a document that can be handed over. The images exist as a pointer that resolves against a server at the originating institution, which is exactly what stops working once the patient has left the country and the referring relationship has ended.
The practical instruction that follows is to obtain the images themselves on physical or exported media before leaving, rather than relying on a link, a portal or an undertaking to send them later.
Re-reads differ, and that is normal
Interpretive disagreement is a baseline property of the work rather than a sign of poor practice. A review of error and discrepancy in radiology reports an estimated day-to-day error rate of 3 to 5 per cent of studies reported against a retrospective error rate averaging about 30 per cent, major disagreement between two observers reading emergency radiographs of 5 to 9 per cent, and major discrepancy rates on computed tomography of 26 per cent between observers and 32 per cent within the same observer; it argues that discrepancy is the better term where reasonable radiologists can differ, that discrepancy does not equate to negligence, and that retrospective judgement is subject to hindsight bias.[6]
The honest reading cuts against the obvious one. Baseline discrepancy between any two readers is high enough that a re-read on return will often differ for reasons unrelated to the foreign provider’s quality, so a differing second opinion is not in itself evidence of substandard care abroad. No study was located measuring discrepancy rates specifically in second reads of imaging performed abroad for medical travel patients, and figures circulating in trade publications for that question are not used here.
Imaging reports are themselves coded document types in the terminologies used to exchange them, which is what allows a report to be filed and found as a document in the architecture described above.[7] A third edition of the British standards was published in October 2025 and was not read for this entry, so the numbered standards cited here are those of the second edition.[8]
See also
- Operative Reports, the document that does reliably travel home
- Pathology Reports, where the material rather than the images is at issue
- Diagnostic Tourism, travel for imaging and assessment rather than treatment
- Personal Health Records, and the limits of what a patient can compel a provider to release
References
- European Society of Radiology. Good practice for radiological reporting: guidelines from the European Society of Radiology. Insights into Imaging 2011;2:93-96, published online 6 February 2011. Verified against primary record: the guidelines were opened and the five-part structure, the urgent-findings duty and the position on structured reporting read. The guidelines describe the report as part of the permanent health record and say nothing about ownership. Retrieved 10 October 2026.
- The Royal College of Radiologists. Standards for interpretation and reporting of imaging investigations, second edition, reference BFCR(18)1, March 2018. Verified against primary record: the standards were opened and the provisions on prior studies, summarising, second opinions recorded as addenda, and identification of the reporter read. Retrieved 10 October 2026.
- Integrating the Healthcare Enterprise. Cross-enterprise Document Sharing for Imaging (XDS-I.b). Final text profile; page last edited 26 June 2024. Verified against primary record: the profile page was opened and the submission of the report as a document, the registration of the images as a manifest, and the retrieval sequence read. Retrieved 10 October 2026.
- Radiological Society of North America. RSNA Radiology Reporting Initiative. Verified against primary record: the page was opened and the template count, the vocabulary mapping, the development history and the template-exchange arrangement read. Retrieved 10 October 2026.
- National Electrical Manufacturers Association, DICOM Standards Committee. DICOM PS3.1 Introduction and Overview, section 1.1 Scope. Current edition served as 2026d. Verified against primary record: the section was opened and the three areas of specification and the three express exclusions read. Retrieved 10 October 2026.
- Brady AP. Error and discrepancy in radiology: inevitable or avoidable? Insights into Imaging 2017;8(1):171-182, published online 7 December 2016. Verified against primary record: the article was opened and the reported error and discrepancy rates and the argument about terminology and hindsight bias read. The paper does not quantify the effect of double reading and does not study imaging performed abroad. Retrieved 10 October 2026.
- Regenstrief Institute. LOINC 18748-4, Diagnostic imaging study. Verified against primary record: the code record was opened and its name, class, status and release history read. The page does not state that this code is the document type for imaging reports, and no specific code assignment is claimed here. Retrieved 10 October 2026.
- The Royal College of Radiologists. Standards for interpretation and reporting of imaging investigations published. 6 October 2025. Verified against primary record: the news item was opened and the publication of a third edition and its stated scope read. The third edition itself was not opened. Retrieved 10 October 2026.
Sourcing note: the numbered standards cited are from the second edition of the British standards, dated March 2018; a third edition was published in October 2025 and was not read, so no standard is renumbered here. Neither professional standards document carries a publication or review date on the pages consulted beyond those given. Figures circulating in trade publications for discrepancy rates on imaging performed abroad were seen during this work and are not used, because the underlying studies were not opened and the publishers are not sources this site cites.