Run charts and SPC: look beyond two monthly numbers

Why this matters

A lower value this month than last month may not establish improvement. IHI describes run charts as data plotted over time to reveal patterns and assess change. ^[1] A time series adds context that a two-point comparison misses, but its meaning still depends on reliable definitions and collection.

Evidence and important boundaries

NHS England uses SPC to understand variation and identify possible process change. ^[2] Making Data Count explains limitations of two-point and traffic-light reporting and cautions against unsupported causal attribution. ^[3] Common-cause variation belongs to the routine process; special-cause signals warrant investigation. A signal alone does not prove its cause.

Suggested local application

ICORP suggests agreeing the measure, numerator, denominator and collection interval first. Annotate workflow changes and data-source changes. Ask an appropriately skilled analyst to choose the chart type. If counting methods change, make that break explicit rather than presenting the entire sequence as directly comparable.

What to monitor

Investigate data quality and context when a signal appears. A stable but unacceptable process may require redesign rather than pressure on individuals for routine fluctuation. Use the chosen tool’s signal rules and process limits; do not invent thresholds. Report uncertainty when information is insufficient.