Skip to content
Curious Context Labs

case study · Healthcare · Automation

Making sure no abnormal result goes unseen

How flagged results can be tracked until a clinician acknowledges them and the follow-up happens, with escalation when they stall.

business
Healthcare clinics
build
Automation
timeline
6 weeks
read
3 min

01 the business

Group of general practices

8 practices, 60,000 patients

Results arrive after patients leave, sometimes for locum or part-time clinicians who are not in. Most are handled well, but the practices have had near misses where an abnormal result sat unacknowledged for weeks.

02 before

The problem

Results come back after the patient has left the clinic or ward, sometimes after the requesting doctor has moved on. Abnormal results can sit unacknowledged, and the follow-up that was intended never happens. Missed results are a known cause of harm.

  • Results for absent clinicians are not always picked up.
  • Intended follow-ups are not tracked.
  • No daily view of outstanding results.

03 the build

What the AI does

  1. Step 1: Reads results as they are filed, using the lab's own abnormal flags and your rules for what needs action.
  2. Step 2: Checks that the requesting clinician, or their cover, has acknowledged each flagged result.
  3. Step 3: Reminds the clinician and escalates to a named deputy if a result stays unacknowledged past your time limit.
  4. Step 4: Tracks the follow-up the clinician records, such as a repeat test or a call to the patient, until it is done.
  5. Step 5: Lists open results for each team every day, oldest first.

paused: waiting for a person to approve

How it was put together

  • Read results as they were filed, using the lab's abnormal flags and the practices' rules.
  • Checked acknowledgement by the requester or their cover.
  • Reminded and escalated to a named deputy past agreed time limits.
  • Tracked recorded follow-ups until done, with a daily oldest-first list.

04 human in the loop

Where a person stays in charge

Clinicians interpret every result and decide what to do; the tool only uses the lab's flags and your rules and never interprets results itself.

  • Clinicians interpret results and decide every action.
  • The tool uses lab flags and practice rules only; it never interprets results.
  • Escalation rules are set by the clinical lead.

05 rollout

From kickoff to live in 6 weeks

  1. Weeks 1–2Agree rules, time limits and deputies with the clinical lead.
  2. Weeks 3–4Run in two practices, reviewing every escalation.
  3. Weeks 5–6Extend to all practices.

06 after

What changes

Flagged results do not go unseen when people change shifts or rotations, and every intended follow-up has an owner until it is done.

worked example · your numbers will differ

Assume 2,000 flagged results a month, with 1% currently waiting more than a week for acknowledgement.

2,000 × 1% = 20 results a month at risk. With escalation, every one has an owner within the time limit.

What to measure

  • Flagged results acknowledged within the time limit.
  • Open results older than the limit.
  • Follow-ups completed.
  • Near-miss reports about results.

07 evidence

What the research says

  1. Across 19 studies, between 6.8% and 62% of laboratory tests and between 1.0% and 35.7% of radiology tests were not followed up, with harms including missed cancer diagnoses.

    Journal of General Internal Medicine, Failure to follow-up test results for ambulatory patients: a systematic review (Callen et al.) (opens in a new tab) · 2012

  2. Around 1 in 10 patients is harmed in health care, and more than half of that harm is preventable.

    World Health Organization, Patient safety fact sheet (opens in a new tab) · 2023

08 lessons

Mistakes to avoid

  • Interpreting results. Leave that to clinicians.
  • Too many alerts. Use the lab's flags and clear rules.
  • No named deputies. Escalation needs a person.

09 questions

Common questions

How common are missed results?
A systematic review of 19 studies found between 6.8% and 62% of lab tests were not followed up in outpatient settings, with harms including missed cancer diagnoses.
Does it work in hospitals too?
Yes, including for results that return after discharge.

keep reading

All case studies ->

Tell us how your business runs. We'll tell you where AI fits.

A free 30-minute call. You talk, we ask questions, and you leave with a few concrete ideas — or an honest answer that AI isn't the right fix yet.

Working with teams around the world.