Clearer wording in computed tomography reports can reduce uncertainty when adults are assessed for suspected acute appendicitis. An investigation published in Insights into Imaging compared original radiologist impressions, reassessed wording and structured schemes based on selected imaging findings. It reviewed 983 consecutive CT reports from a single academic medical centre in Thailand, with final diagnoses established through surgical, pathological and clinical outcomes. Reassessment of report language produced fewer indeterminate classifications and closer agreement with final diagnoses than either the original wording or the feature-based schemes. The findings support more consistent expressions of certainty while preserving radiologists’ overall judgement.
Reassessment Clarifies Diagnostic Certainty
Original impressions were placed into negative, indeterminate or positive categories according to the wording used. Clear statements excluding or supporting appendicitis were classified accordingly, while terms such as “possible”, “suspected”, “equivocal” or “cannot be excluded” were treated as indeterminate. Two radiologists then independently reassessed the reports initially placed in the uncertain category. They reviewed the impression and the full written report but did not reinterpret the CT images.
This reassessment reduced indeterminate reporting from 15.5% to 12%. The number of positive reports did not change, while some uncertain reports moved into the negative category. Agreement between the 2 reviewers was substantial, suggesting that the intended level of certainty could often be identified from the report as a whole even when the original impression used ambiguous wording.
Must Read: Lung Screening Reporting Lacks Consensus
Confirmed appendicitis became more common as the stated level of certainty increased. Very few negative reports were associated with appendicitis, while the condition was present in a substantial share of indeterminate cases and in most positive cases. After reassessment, the uncertain group contained a higher proportion of confirmed cases because some lower-risk reports had been reclassified as negative. This pattern shows the value of a distinct indeterminate category but also indicates that it should be used selectively. More precise language can narrow the uncertain group without altering image interpretation or adding another reading step.
Structured Rules Produce More Uncertain Results
The feature-based schemes assigned certainty using appendiceal diameter and signs of inflammation, including wall changes and fat stranding around the appendix. One scheme used a 6-mm diameter threshold and the other used 7 mm. Both divided findings into 4 levels, ranging from negative to positive, with possible and probable categories between them.
These schemes produced substantially more uncertain classifications than the radiologist impressions. When the 2 intermediate categories were combined, almost half of reports were indeterminate with the 6-mm scheme and more than a quarter with the 7-mm version. The 7-mm approach performed better than the 6-mm version in several comparisons, including specificity and agreement with the final diagnosis. Even so, reassessed radiologist impressions achieved the strongest overall performance.
The original and reassessed reports performed similarly when clearly positive findings were separated from all other categories. The main improvement after reassessment appeared when negative reports were compared with non-negative results. Reassessed wording increased accuracy and produced stronger agreement with the final diagnosis, although some performance differences were modest.
The findings indicate that a small set of predefined imaging features does not fully reproduce the way radiologists weigh the entire examination. Structured terminology may improve consistency and help organise certainty, but the tested schemes were less effective when used as stand-alone substitutes for integrated interpretation. The 7-mm threshold appeared more suitable than the lower threshold when a feature-based system was applied.
Imaging Patterns Influence Indeterminate Wording
Indeterminate reports commonly contained findings between clearly negative and positive patterns. The appendix was generally wider than in negative cases but narrower than in positive cases. Wall thickening, increased wall enhancement and fat stranding were also present at intermediate frequencies. These mixed patterns help explain why some examinations were difficult to classify with confidence.
Several imaging features were independently linked to the reporting category. The absence of fat stranding, wall enhancement and wall thickening favoured a negative rather than indeterminate impression. A larger appendiceal diameter favoured a positive report, while a smaller diameter favoured a negative one. The absence of another diagnosis on CT was also associated with a positive rather than uncertain conclusion. Other factors, including age, sex and the presence of an appendicolith, showed more limited associations with specific categories.
Requests for clinical correlation appeared much more often in indeterminate reports than in negative or positive reports. By contrast, radiologist subspecialty did not significantly affect the likelihood of uncertain wording. Reporting time and prior ultrasound were also not associated with the final certainty category.
The retrospective, single-centre design may limit wider applicability. Imaging findings were extracted from existing reports rather than through a new review of the images, and findings not mentioned were treated as absent. Different reference standards were used for operated and non-operated patients. Reassessment was limited to reports already classified as indeterminate, so prospective evaluation is needed before a simplified certainty framework can be adopted routinely.
More deliberate wording can reduce avoidable uncertainty in CT reporting for suspected adult appendicitis. Reassessment of written impressions produced fewer indeterminate classifications and stronger agreement with final diagnoses than the original wording or the tested feature-based schemes. A concise 3-level approach may help radiologists distinguish negative, uncertain and positive findings more consistently while retaining the contextual judgement that fixed rules did not fully capture. The results support careful use of equivocal terminology when the full report permits a clearer expression of diagnostic certainty.
Source: Insights into Imaging
Image Credit: iStock
References:
Siriphiphatcharoen P, Kaewlai R, Tongsai S et al. (2026) Indeterminate CT reporting of adult acute appendicitis: radiologist versus lexicon-based diagnostic certainty. Insights Imaging; 17, 189.
