Table 1

Attributes assessed and their importance for determining the success of QI projects

AttributeWhat is assessed?Why assessed/hypothesis?
CountryWhich country was the study conducted inUnderstand the geographic spread of QI studies
Target population: Unit of InterventionWhat type of individuals did the QI study target? (1) Patients; (2) caregivers; (3) healthcare providers; (4) mother–infant dyad; (5) other domainsQI interventions are aimed at different groups of people
Clinical topic of QI improvementWhat types of clinical questions are being addressed by current QI projects? Eg, reduce CLABSI*, improve workflow, adapt an evidence-based protocolUnderstand which problems are being addressed and provide insights into which problems may not be frequently addressed.
Healthcare settingWas the study conducted in an academic/teaching hospital?Are quality improvement activities mostly implemented in an academic hospital setting, which have teaching and research mandates?
Multicentre studiesAre studies being conducted at multiple hospitals/communities/centres?2 9How often do published QI activities involve, or report on, collaboration between multiple hospitals?
Ethics approvalWhether a review board approved or waived ethics requirement?How often did QI studies receive an ethics waiver, and hence, be considered as a QI study by the review board?
Root cause analysis (Problem identification method)Did the study use root-cause analysis or related methods (Fishbone/Ishikawa diagram, Pareto charts, or process mapping) when identifying the specific QI problem to tackle in their local setting?11Root cause analysis, or the three related methods, is a technique for understanding the problem and adapting a locally tailored solution; it is emphasised in QI primers.
Stakeholder engagementWere healthcare providers, caregivers or healthcare leaders involved in the design and adaptation of the QI activities?QI projects succeed when clinical expertise is used, the activities are backed by the hospital management, and when parents or caregivers are involved in shaping the QI programme.11 14
Components of QI interventionsModalities of implementing interventions. (See table 2 for definitions)Develop an understanding of the approaches to implementing QI interventions, and how often they are used.
MeasuresWhat types of outcomes were evaluated and how often were they reported?
  1. Patient-important outcome measures—outcomes that directly affect a patient’s health and well-being.12 20

  2. Process of care measures—(1) indicators on the quality of medical care, (2) the success of the implementation process, (3) balancing measures—side effects resulting from the interventions.12

Statistical analysisWhat type of statistical methodology to assess the effectiveness of the intervention?
  1. Statistical significance—use of p values.

  2. Confidence intervals, including process control limits

  3. Statistical process control—must include use of control limits

  4. Adjustment for confounders/multivariable analysis

The use of statistical techniques reflects how authors intend to interpret their studies.
  1. The use of p values reflects the study’s tendency to view the QI activity as being successful or not based on statistical significance.19

  2. Use of confidence intervals reflects the study’s understanding that any changes in quantitative measures are estimates based on a sample in their local setting.19

  3. Use of statistical process control reflects the understanding that QI is a continuous process defined by events over time. It examines if the persisted over time following the intervention,).18

  4. Adjustment of confounders shows that the study sought to ascertain the ‘true’ association between the QI intervention and study outcomes.21

  • *CLABSI, Central line-associated bloodstream infections.

  • QI, quality improvement.