Test properties of the NDSS case definition.

AHS: Alberta Health Services; CCHS: Canadian Community Health Survey; CLS: Calgary Laboratory Services; EMR: Electronic Medical Records; GP: General Practitioner; HEDIS: Health Plan Employer Data and Information Set; HMO: Health Maintenance Organization; Kappa: Kappa statistic; MCBS: Medicare Current Beneficiary Survey; MHSC: Manitoba Health Services Commission; MHHP: Manitoba Heart Health Project; MHSIP: Manitoba Health Services Insurance Plan; NDSS: National Diabetes Surveillance System; NPHS: National Population Health Survey; NPV: Negative predictive value; ODB: Ontario Drug Benefit; ODD: Ontario Diabetes Database; OHI: Ontario health Insurance Plan; Sens: Sensitivity; SHB: Saskatchewan health beneficiaries; Spec: Specificity; PPV: Positive predictive value.|Included in meta-analysis.◊Test estimates and 95% confidence intervals which were not reported in the original paper were manually calculated from available raw data.||95% confidence intervals were estimated based on a diabetes prevalence of 6.8% via the ODD and a sample size of 4691 from the NPHS cycle 3 1998/9.*Test properties were recalculated to designate self-reported diabetes from survey as the reference standard.The sample size of the ODB were not explicitly stated in the paper. Hence, the 95%CIs could not be calculated.-Demographic data was not reported in the study.∫the authors validated one physician claim or one hospitalization which is similar but not identical to the NDSS criteria.@test measures of the case algorithm closet to the NDSS criteria is displayed in the table.¢Test measures were calculated using the reference standard closest to current diabetes diagnosis criteria (i.e. CDA guidelines for diagnostic criteria of diabetes OR glycated hemoglobin ≥6.7%) was displayed in the table.△The sensitivity was calculated by projecting the case-control design (n = 145 696) to the entire Saskatchewan population where 625 994 individuals would not have met the NDSS criteria of which 3443 would have been identified as having diabetes based in medication alone.