Table A5: EPOC Risk of Bias Assessment—Observational Studies With Historical Controls.

Study	Independent of Other Changes	Intervention Effect Prespecified	Intervention Affected Data Collection	Incomplete Data Addressed	Free from Selective Outcome Reporting	Other Sources of Bias
Solberg et al (24); Sperl-Hillen et al (25)	No Authors reported the implementation of a diabetes care program during the same period	Yes	No	Unclear Sample sizes varied across time (cohorts differ) and the authors did not discuss use of services outside of the system	Yes	Yes
Gladstone et al (29)	Unclear Authors did not account for other changes occurring in the practice	Yes	No	Unclear Authors did not report on missing data and excluded people who were not seen after implementation	Yes	Yes
Cherniack et al (18)	Unclear Authors did not discuss other initiatives that may have been underway	No There was a change in clinic structure part way through the intervention	Unclear Data collection was based on the appointment system and may have changed with implementation	Unclear Authors did not report rates by patient and did not report missing data rates	Yes	Yes

Abbreviations: EPOC, Effective Practice and Organization of Care.