Value Stream Mapping for Diagnostic Imaging Departments: From Referral to Final Report Without the Scan Backlog

Why value stream mapping fits diagnostic imaging

Diagnostic imaging is an information-and-material flow problem. The patient moves through preparation and scanning; the referral, clinical details, images and report move across people and systems. A scan can be completed promptly while its images wait in PACS or its report sits in a worklist.

Value Stream Mapping (VSM) makes that end-to-end flow visible. It records the steps, handoffs, processing time, waiting time and work in process (WIP), helping a team distinguish genuine capacity constraints from delays created by queues and batching.

The example below is a hypothetical mid-size hospital outpatient CT pathway, not a benchmark or a claim about a specific department. Its figures illustrate how to build a map and test improvement ideas. Real decisions should be based on locally observed data, stratified by urgency, modality, contrast requirements and operating hours. Published radiology improvement work has also used VSM and Lean methods to examine imaging delays, including CT turnaround and emergency CT workflows (RadioGraphics: CT Kaizen project; emergency CT workflow study).

Set the scope: referral received to final report available

Define the value stream boundary as GP referral received → final verified report available to the referrer. Ending the map when the scan is finished would miss image transfer, reporting queues, verification and distribution. The very steps that determine when the clinical question is answered.

For a useful baseline, capture timestamps from the referral system, RIS and PACS, then validate them with staff who perform the work. Record both process time (hands-on work) and lead time (elapsed time through work and waiting). The example assumes 62 referrals per working day, a 600-minute available production window and two CT scanners operating in parallel.

Takt time is available time divided by demand:

600 minutes ÷ 62 studies = 9.68 minutes per study.

This is the required system output interval, not the duration of an individual scan. A 15-minute scan slot can still support demand with parallel scanners; capacity should be checked at each resource, not inferred from takt alone.

Current state: where the 14.5 days accumulate

In this scenario, triage is reviewed in a once-daily batch at 10:30. Appointment matching and the time until the scheduled scan contribute substantial queues. The prep letter is sent nine days before the appointment. After the scan, reporting work is released in batches of 24 studies, creating another queue.

The WIP figures are indicative queue snapshots rather than a sum of every patient in the pathway. The largest visible constraint is the reporting queue: 310 unreported studies represent about five working days of demand at 62 studies per day. A team should confirm the constraint with queue age, hourly arrival and completion rates, and scan-to-final-report percentiles.

What the eight wastes look like in this pathway

Map each waste to an observable point in the process:

• Waiting: patients wait between triage and their scan; completed scans wait for reporting.
• Defects: incomplete referrals are returned to the GP for missing clinical details.
• Overproduction: duplicate imaging orders are placed before existing requests are checked.
• Transportation: physical films or discs move between departments instead of images being available electronically.
• Inventory: unreported studies accumulate in the radiologist worklist.
• Motion: radiologists walk to film rooms to retrieve or review material.
• Non-utilised talent: radiographers spend time chasing clerical information rather than applying imaging expertise.
• Overprocessing: referral details are manually re-keyed across systems.

Future state: create flow from referral through report

The future state should reduce queues without compromising clinical prioritisation, reporting quality or safety. In this example, the team redesigns referral completeness, triage cadence, patient flow and reporting release.

1. Standardise the referral. Use a template with mandatory clinical indication, requested examination, relevant history and referrer details. Send incomplete referrals through a defined exception route rather than repeated informal chasing.
2. Triage to a daily cadence. Review new referrals at planned intervals aligned to demand, with explicit urgent-priority rules and visibility of ageing requests.
3. Use a FIFO lane for scan-ready patients. Within appropriate clinical priority categories, make ready patients visible and sequence them consistently. FIFO must not override clinical urgency.
4. Pull reporting work through a supermarket. Release a controlled amount of scan-ready work to reporting based on capacity and priority. Make queue age visible and set escalation triggers for overdue studies.
5. Level the reporting mix. Balance routine and complex work across reporting sessions where clinically appropriate, rather than releasing large batches that create uneven demand.
6. Integrate electronic referrals. Reduce re-keying and make referral status, image availability and report distribution easier to track.

With these changes, the worked future-state target is 6.2 days from referral receipt to final report. The illustrative breakdown is 0.1 day for intake, 0.2 for triage, 0.1 for booking, 3.7 waiting for the scan, 0.1 for scan-day flow and 2.0 for the reporting queue and release. Hands-on process time falls from 40 to about 38 minutes, mainly through less re-keying and more consistent reporting work. Reporting WIP falls from 310 to a target of about 124 studies, two days of demand, while the remaining appointment queue is still a major improvement opportunity.

PCE: keep the arithmetic honest

Process Cycle Efficiency (PCE) is value-added time ÷ total lead time. For this example, count the 15-minute scan and 9-minute radiologist interpretation as value-adding clinical work: 24 minutes.

Using elapsed clock time:

• Current: 24 ÷ (14.5 × 1,440) = 0.11%
• Future: 23 ÷ (6.2 × 1,440) = 0.26%

The future value-added time is slightly lower because more consistent reporting is assumed to reduce interpretation time by one minute. Even a substantial lead-time reduction leaves PCE low because most elapsed time is still waiting. An illustrative 11% PCE would not be consistent with a 6.2-day end-to-end lead time and about 23 minutes of value-added work; it would require a different definition or denominator. Use the same operational definitions before and after. The Process Cycle Efficiency Calculator can help document the calculation.

Current versus future state

The turnaround-time measures and takt time answer different questions: scan-to-report is a service lead-time measure; takt is the required output rhythm from demand. Track both, alongside report quality, discrepancy rates, cancellations, overtime and urgent-study performance.

Sequence the kaizen over 90 days

Days 1–30: referral standard work and triage cadence

• Confirm scope, timestamp definitions and baseline performance.
• Pilot mandatory referral fields and a clear incomplete-referral pathway.
• Test scheduled triage reviews and monitor first-pass acceptance.

Days 31–60: FIFO flow and backlog burn-down

• Introduce a visible scan-ready FIFO lane with clinical-priority safeguards.
• Set a reporting WIP limit, ageing triggers and daily backlog review.
• Protect temporary reporting capacity to reduce accumulated backlog without compromising routine work.

Days 61–90: level reporting and electronic integration

• Pilot a balanced reporting mix and review whether work arrives at a sustainable pace.
• Reduce avoidable batching and handoffs; implement electronic referral integration where feasible.
• Audit results and update standard work, escalation rules and ownership.

Use run charts to see whether lead time, WIP and first-pass acceptance improve and stay improved. The project scope boundary calculator can help teams define a manageable improvement scope before starting.

Build the capability to improve imaging flow

VSM is most effective when the team can connect observation to disciplined measurement, root-cause analysis and sustainment. Develop those skills through Lean 6 Sigma Hub’s CSSC-accredited, self-paced online certification pathways, from White Belt to Master Black Belt. Courses use real DMAIC case studies, simulations, worked examples and dummy data so you can learn by doing.

Choose your Lean Six Sigma training and certification pathway, then apply the tools to a real process improvement opportunity.

Kaizen. Kai-Care. Kai-Done. Lean Six Sigma.