The Order Set That Wouldn't Change
A hospital's informatics team built a standardized admission order set for community-acquired pneumonia based on national guidelines. After 18 months, adherence is 38%. Interviews reveal that physicians find the order set 'too long' and bypass it by free-texting orders. The informatics nurse who championed it has left, and no one owns the content. A physician group wants to add a new antibiotic to the order set; pharmacy objects because the drug is on the restricted formulary and requires infectious disease approval. The CMIO asks you to 'fix the order set' before the next accreditation survey. There is no formal change-control process for order sets; the last update was 14 months ago.
1. Diagnose why adherence is low and propose a redesign approach grounded in informatics theory.
Read the worked answer
Low adherence is a sociotechnical and cognitive-load problem, not a content problem alone. Diagnose across levels:
- Individual/cognitive: A long order set increases cognitive load and decision time; clinicians free-text to reduce effort. The design likely presents too many options without defaults or progressive disclosure.
- Workflow: The order set may not match the actual admission workflow (timing, sequence, who enters what).
- Organizational: No owner, no change control, and a departed champion mean the artifact has no stewardship and no feedback loop. The 38% adherence is a symptom of abandonment.
Theory-informed redesign:
- Apply human factors / cognitive load principles: shorten the set, set evidence-based defaults, hide rarely used options behind progressive disclosure, and order items by workflow sequence.
- Apply diffusion of innovations: identify physician champions and early adopters per service line to drive peer adoption.
- Apply the sociotechnical model: address people, workflow, and governance alongside the technology.
- Close the loop with usability testing on representative clinicians before release.
Do not simply 'fix the content' before establishing ownership and change control; otherwise the redesign will decay the same way.
2. How do you resolve the antibiotic dispute and establish change control before the survey?
Read the worked answer
Resolve the dispute through governance, not informatics authority. The informatics team does not adjudicate formulary policy; it operationalizes it.
- Convene the clinical content owners: the pharmacy and therapeutics committee (or formulary authority) and the infectious disease service. The order set should reflect their decision.
- If the drug is restricted, build it into the order set with a hard stop or approval prompt that routes to ID approval, rather than omitting it or allowing unrestricted use. This encodes the policy in the workflow.
- Document the decision and rationale.
Establish change control (minimum viable, before survey):
- Name an owner (clinical content steward) for the order set, with a backup.
- Define a review cadence (e.g., annual, plus triggered review on guideline change or safety event).
- Create a request-and-approval workflow: request → clinical review by content owners → informatics build → testing → release → communication.
- Version control and audit trail for every change.
- Post-release monitoring of adherence and safety signals, feeding back into the next review.
For the survey timeline, prioritize: assign ownership, document the current version and review date, and record the antibiotic decision. A functioning governance record matters more than a rushed content rewrite.