I85PREP
⌘ K
CPPS study materials
CPPS study resources

CPPS: Measuring and Monitoring: Choosing, Defining, and Interpreting Safety Data

CPPS · Work through your exam, one topic at a time.

All guide sections

Measuring and Monitoring: Choosing, Defining, and Interpreting Safety Data

This domain tests whether you can select a measure that actually reflects harm, define it operationally, and interpret variation honestly. Items contrast data sources, denominator choices, and statistical process control.

Core rules and concepts

  • Donabedian's frame: structure, process, outcome. Pair an outcome measure with the process measures that drive it and with balancing measures that catch unintended harm elsewhere.
  • Report rates, not counts: CLABSI per 1,000 central line days, CAUTI per 1,000 catheter days, falls per 1,000 patient days, harm per 100 admissions.
  • Know each source's bias. Voluntary reporting reflects culture more than incidence. Administrative data (AHRQ Patient Safety Indicators) is inexpensive but coding-dependent and under-detects. Observation captures behavior. Trigger-based chart review finds harm that reporting misses.
  • The IHI Global Trigger Tool reviews small random chart batches (commonly 20), using triggers such as naloxone, vitamin K, diphenhydramine, abrupt medication discontinuation, ICU transfer, or readmission. Reviewers assign NCC MERP harm categories E–I: E temporary harm requiring intervention, F prolonged hospitalization, G permanent harm, H near death, I death.
  • Separate common from special cause variation with run charts and control charts. Run chart rules: a shift of six or more points on one side of the median, a trend of five, or an astronomical point.
  • Define every measure operationally: numerator, denominator, exclusions, data source, and frequency.

Worked example

A unit reports 6 CLABSIs over 1,200 central line days: 6 ÷ 1,200 × 1,000 = 5.0 per 1,000 line days, up from 3.1. Before declaring an outbreak, check the denominator (did line days fall because fewer lines are used?), whether surveillance definitions or laboratory methods changed, and whether the shift is special cause on a control chart. Small numbers produce unstable rates, so use rolling windows rather than single-month comparisons.

Traps

  • Comparing counts or rates across units with different size, case mix, or device use.
  • Celebrating a fall in reported events after staff were disciplined.
  • Using administrative indicators alone to judge an organization's safety.
  • Reacting to common cause variation with a new project every month.