Quality Improvement and Patient Safety
Core concepts
Know the IOM (now NAM) six aims - safe, effective, patient-centered, timely, efficient, equitable - and the STEEEP mnemonic. Understand Donabedian's structure–process–outcome framework and the Model for Improvement (Plan-Do-Study-Act). Be fluent in measurement: structure, process, and outcome measures; balancing measures; statistical process control; run charts; and the difference between common-cause and special-cause variation. Know error taxonomy (active vs. latent errors, slips vs. mistakes), the Swiss cheese model, and high-reliability organization principles. Understand root cause analysis (RCA) for sentinel events and failure mode and effects analysis (FMEA) for prospective risk assessment.
Traps
RCA is retrospective; FMEA is prospective. Donabedian's structure measures resources (staffing, equipment); process measures what is done; outcome measures results. A frequent trap is labeling a measure as an outcome when it is actually a process (e.g., "percentage of patients receiving beta-blockers" is a process measure, not an outcome). Also distinguish quality assurance (retrospective, punitive) from quality improvement (prospective, systems-focused).
Worked example / rule of thumb
A hospital sees an increase in central line infections. First step: determine whether the variation is special cause (a shift or trend on the control chart) versus common cause. If special cause, investigate the specific event; if common cause, redesign the system. Rule of thumb: RCA for what already happened; FMEA for what could happen; PDSA to test a change.
Remember: most errors are system failures, not individual failures - a central tenet tested repeatedly.