1. Median improvements in process adherence across included studies.

Dichotomous outcomes (number of intervention vs. control comparisons)	Median absolute improvement (Interquartile range)
	Using median outcome from each study	Using best outcome from each study
All process outcomes (N = 32)	4.2% (0.8% to 18.8%)	5.6% (2.0% to 19.2%)
Prescription of medications (N = 21)	3.30% (0.5% to 10.6%)	6.2% (3.0% to 28.0%)
Prescription of recommended vaccines (N = 6)	3.8% (0.5% to 6.6%)	4.8% (0.5% to 7.8%)
Test ordering (N = 13)	3.8% (0.4% to 16.30%)	9.6% (0.6% to 24.0%)
Elements of recommended documentation (N = 3)	0.0% (‐1.0% to 1.3%)	2.0% (2.0% to 4.0%)
Other process outcomes (N = 7)	1.0% (0.8% to 8.5%)	4.0% (0.8% to 8.5%)

The Table shows average improvements (expressed as the median and interquartile range) across included comparisons for different types of process outcomes. All process outcomes were defined so that higher values always represent an improvement. For example, data from a study aimed at reducing the percentage of patients receiving inappropriate medications would be captured as the complementary percentage of patients receiving appropriate medications, so that an increase in process adherence would represent an improvement.

Most studies reported multiple endpoints but did not specify a primary outcome. For the main analyses, we used the median improvement from each study (that is the median change in adherence to a target guideline or process of care across all such changes reported for the study) as the single representative outcome for that study. We then calculated the median improvements across all included studies for different types of process measures, as shown in the middle column of the table. The column to the far right presents the same results when we used the best improvement from each study as its representative outcome.