The Five-Minute Lab Result: Decomposing a Mandate
A ministry mandate requires that all laboratory results be available to any participating clinician within 5 minutes of validation across an 11-hospital health information exchange. Two facts complicate delivery. A 300-bed hospital's chemistry middleware releases results in 60-minute batches; the vendor quotes €180,000 and seven months to enable event-based push. Two other hospitals send 30% of their tests to an outsourced laboratory with a four-hour turnaround, and those results enter the exchange already late. The emergency department clinical lead insists five minutes is a patient-safety requirement for stat troponin and potassium. The programme board, facing a fixed go-live, wants to declare the requirement satisfied by narrowing scope to inpatient validated results. In week 9 the laboratory accreditation body writes to say it requires an immutable audit trail of every result distribution, which the current exchange design does not capture.
1. Is the five-minute figure a requirement or a solution constraint? Decompose it into something the programme can actually verify.
Read the worked answer
The number as written conflates two different clocks. The mandate says 'within 5 minutes of validation', which measures distribution latency. The ED lead means specimen-to-clinician time, which includes the 60-minute batch. A system can meet the mandate while the ED still waits 65 minutes - and the ministry will eventually measure the thing that matters to clinicians.
Decompose into two requirement families:
- Distribution latency, source system to clinician view, by result class. This is technical, owned by the exchange, and verifiable with timestamps at three points.
- Turnaround time, specimen collection to clinician view, by result class. This is clinical and owned by laboratories; the exchange can only report it.
Class the results rather than the patients. Stat troponin and potassium from ED; routine inpatient chemistry; send-out and reference tests. Each class gets its own target and its own owner. Send-outs cannot meet five minutes because the clock starts at an external laboratory - that is a property of the supply chain, not of the software.
Write acceptance criteria per class, with the measurement point named, so 'met' is not a matter of interpretation.
2. Map the stakeholders and identify the conflict that the board's proposed scope narrowing does not resolve.
Read the worked answer
Stakeholders: ministry (mandate and political exposure), ED clinicians (the safety claim), laboratory directors (who own the batching middleware and its budget), the outsourced laboratory provider, the accreditation body (audit trail), patients, exchange operations, and the privacy office - absent from the current list and needed for distribution audit.
What narrowing achieves. Restricting to inpatient validated results removes outpatients and send-outs from the claim. It does not touch the 60-minute batch, because stat troponin and potassium are frequently on inpatients. The ED lead's requirement survives the narrowing intact.
The unresolved conflicts:
- The board is trading a scope reduction for a claim that still reads as 'all results', which is a reporting risk to the ministry.
- The batched hospital has no funded incentive to change; €180,000 and seven months are not in anyone's plan.
- The accreditation requirement arrived late, is non-negotiable, and has design and cost impact that is not in the baseline - it needs formal entry, not a verbal note.
Missing voice. Clinicians in low-bandwidth sites consume results too; latency targets set without them may be unachievable where they practise.
3. Recommend a course of action to the programme board, including what you would tell the ministry.
Read the worked answer
Options considered:
- (a) Narrow scope, publish the limitation, proceed. Cheapest, but leaves the ED safety case unaddressed.
- (b) Fund the vendor interface everywhere. Correct but seven months, and the send-out labs remain impossible.
- (c) Certify by result class, publish latency per class, fund fixes where clinical risk is highest.
Recommendation: (c) plus targeted funding. Certify distribution latency per class; publish a per-class latency table to clinicians so they know what they can rely on; fund event-based push at the two sites handling the highest-volume stat pathways. Include the accreditation audit trail as a mandatory scope item with its own acceptance test.
What to tell the ministry. Say plainly that the mandate as worded is achievable for distribution latency but not for send-out tests, because the clock starts at an external laboratory. Propose revised wording that separates the two clocks and states targets per class. Do not silently redefine 'all results' - that becomes an audit finding.
Record it. An ADR capturing the interpretation, the evidence, the residual risk, and the owning executive. The board should accept risk explicitly, not by ambiguity.