Table 3

Critical decision points that need explicit reporting for a quality improvement study

Critical decision pointReporting requirements
Need for a stable baseline to draw inferencesRationale for number of data points, the time unit and sample size for each data point
Choice of outcome measuresThe same primary outcome before and during intervention implementation, including definition and data collection.
Balancing and intervention fidelity process measures report on safety and compliance with the intervention.
QI design featuresExplicit reporting on several design features:
  • Routinely collected (existing) data or newly collected (prospective) data.

  • Time horizon.

  • Controlled or uncontrolled design, type of control and for what type of confounding factor this controls (and what not).

For instance, the study design in the title could be: a controlled 1-year QI study with routinely collected data. Additional detail can be reported in the methods.
SPC analysisExplicit reporting on:
  • Type of chart and which control limits.

  • Which specific rules for special cause variation.

  • Calculation of new centre line and control limits.

  • Rationale for number of data points to establish a new stable process (sustainability).

  • Other interventions or secular trends that may affect the outcome measures.

  • QI, quality improvement; SPC, statistical process control.