Quality, Safety, and Process Improvement in Informatics
Frameworks and methods
Quality and safety work uses the Donabedian model (structure, process, outcome), the IOM six aims (safe, effective, patient-centered, timely, efficient, equitable), and improvement methods such as PDSA, Lean, and Six Sigma. Safety analysis includes root cause analysis (RCA) after an event and failure mode and effects analysis (FMEA) before a process change. Just culture distinguishes human error, at-risk behavior, and reckless behavior. High-reliability organizations emphasize sensitivity to operations, reluctance to simplify, preoccupation with failure, resilience, and deference to expertise.
Measurement
Use run charts and control charts to separate common-cause from special-cause variation. Avoid tampering with normal variation. Measure before and after an intervention, include balancing measures, and watch for unintended consequences such as alert fatigue or new workarounds. Informatics safety includes usability, interoperability errors, copy-forward, wrong-patient orders, and downtime.
Traps
Common distractors confuse RCA with FMEA, blame individuals for system failures, or declare success from a single pre/post comparison without a control. Another trap is choosing a measure because it is easy rather than because it reflects the outcome.
Worked example / rule of thumb
A unit's falls rate rises for three consecutive months after a new EHR workflow. A control chart shows a special-cause signal. The team performs an RCA, finds a documentation field moved off-screen, and redesigns the screen; it then tracks falls and documentation time as a balancing measure. Rule of thumb: measure, understand variation, intervene, and re-measure with balancing measures.