When radiology departments face pressure to improve critical findings turnaround, the default answer is to add capacity. Hire another radiologist. Extend coverage hours. Contract a teleradiology group for overnight reads. These are legitimate options, but they are also expensive, slow to implement, and often unavailable given the national shortage of radiologists in subspecialty areas.
There is a second category of interventions that gets less attention: changing what the existing radiologists see first.
Time-to-Read Versus Throughput
Throughput and time-to-read are related but not the same metric. Throughput measures how many studies a department reads per unit time. Time-to-read measures how long it takes a specific study to reach a radiologist after it becomes available for read. For routine studies, the distinction matters less. For critical findings, it matters a great deal.
A high-throughput department can still have poor time-to-read on critical cases if those cases arrive in the middle of a large queue and wait their turn. The radiologists are working quickly, but the urgent study is still at position 22 in a list of 40. Throughput optimizations that make the entire list shorter will eventually reach that study, but they do not specifically accelerate it relative to the cases above it.
Improving critical findings time-to-read without adding staff requires directly addressing what determines position in the queue. That means the worklist ordering logic, not the number of readers.
What Worklist Ordering Logic Actually Controls
Most RIS systems offer some version of priority-based worklist sorting. The challenge is that priority assignment happens at the order level, before the scan, based on the ordering clinician's judgment about urgency. This works when urgency is apparent at the time of ordering. It fails in several common scenarios:
A patient presents with a cough and mild dyspnea. The ordering physician is not thinking PE. The chest CT is ordered routine. The scan shows a large central pulmonary embolism. The study enters the worklist at routine priority.
A patient is scanned overnight as part of a trauma evaluation. The ordering is stat, but by the time the study arrives in the overnight teleradiology queue alongside 35 other studies, the relative priority within that queue is still arrival time.
A screening CT for a long-term smoker shows a Lung-RADS 3 nodule that should be communicated and tracked. The ordering was routine, appropriate for a screening context. The study enters the queue at routine priority with no signal that this read requires particular follow-up attention.
In each case, the clinical significance of the finding is not represented in the worklist position because the information about the finding did not exist at the time the worklist position was assigned.
Scan-Side Triage Changes the Information Available
The insight behind scan-side triage is that the information about what a study contains becomes available after acquisition and before read, creating a window in which worklist position can be updated before the radiologist opens their session.
ImageAssist operates in this window. After a chest CT is acquired and transmitted via DICOM, the triage system evaluates the study and updates the worklist priority in the RIS before the radiologist's next read. For the PE scenario above, a study that arrived without a stat flag can be repositioned to the front of the queue based on the imaging characteristics, giving the radiologist the right first study without manual intervention from the ordering team or radiology staff.
This is not a replacement for stat ordering when the ordering clinician knows a case is urgent. It is a safety net for cases where urgency is present in the imaging but was not known at the time of ordering. The cases where this matters most, the unexpected finding in a routine-priority study, are exactly the cases where manual workflows are least likely to catch the urgency before it delays read time.
What "Without Adding Staff" Actually Means
The claim in this article's title deserves a precise interpretation. Scan-side triage does not increase the number of studies a radiologist can read per hour. It does not reduce the total volume of reads required. What it does is change the sequence in which the existing volume reaches existing radiologists, specifically improving the position of studies most likely to contain critical findings.
The practical effect is that a radiologist working a 200-study worklist will reach the studies that matter most sooner, without anyone deciding in real time to manually flag or move those studies. The department's existing capacity, the same radiologists, the same reading software, the same available hours, produces better critical-findings turnaround because the sequence is better informed.
We want to be specific about what this does not do. It does not guarantee that every critical finding reaches a radiologist within a fixed time window. It does not replace critical findings protocols for communication after a finding is made. It does not work for finding types outside the scope of the current model. The product page describes the specific finding categories ImageAssist currently handles.
Measuring Whether It Is Working
For a department evaluating whether scan-side triage is improving critical findings time-to-read, the measurement framework should be specific to the finding categories being targeted. Average time-to-read across all studies will not change much; the improvement is concentrated in the priority-flagged cases. The right metric is time from study availability to radiologist read for studies that contained one of the tracked finding types, compared to the same metric before deployment.
This requires linking back-end radiology audit data (when was the study read) with finding type data (did the final report include a critical finding) and comparing the pre-deployment baseline. Most radiology information systems retain the data needed for this analysis; it is a matter of extracting and comparing it.
During our pilot deployments, we worked with departments to establish this baseline before go-live and track the delta after. The comparison is the most honest way to evaluate whether the prioritization is having the intended effect. If your department is thinking about how to structure this measurement, the clinical team can walk through the methodology.
The Staffing Conversation Does Not Go Away
Improving worklist order is not a substitute for appropriate staffing. High-volume departments with chronic radiologist shortages need both better ordering logic and sufficient reader capacity. A well-ordered list of 500 studies still requires enough radiologist hours to read 500 studies.
The argument here is more specific: that for departments where the bottleneck is not total capacity but specifically time-to-read on critical findings, and where manual intervention to flag urgent studies is unreliable or slow, changing the ordering logic is a faster and less expensive intervention than adding staff. It also addresses a different root cause. Staffing more radiologists speeds up all reads; better triage logic specifically accelerates the reads that matter most. Both may be needed, but they address different problems.
Read more about how worklist order connects to patient safety outcomes in the related article on worklist order as a patient safety issue.
