Improving type 2 diabetes detection among at-risk individuals – comparing the effectiveness of active opportunistic screening using spot capillary-HbA1c testing and venous HbA1c testing: a cluster randomized controlled trial

Table 3 Results of the two-level mixed-effects logistic regression model for the uptake rate of OGTT

	*OGTT Uptake Rate*
	OR [95% CI]	p
*Fixed Effects*
Intercept	0.05 [0.02–0.14]	0.001**
Intervention Condition (POC-cHbA1c)	1.74 [1.00–3.02]	0.049*
Gender (Female)	1.15 [0.83–1.59]	0.406
Age	1.02 [1.01–1.04]	0.002*
First-Degree Relative with Diabetes (Yes)	1.59 [1.13–2.24]	0.008*
History of Gestational Diabetes (Yes)	1.34 [0.53–3.39]	0.540
Hypertension (Yes)	0.68 [0.47–0.98]	0.037*
Impaired Fasting Glucose (Yes)	0.94 [0.42–2.08]	0.874
Impaired Glucose Tolerance (Yes)	0.69 [0.06–7.46]	0.762
Hyperlipidemia (Yes)	1.73 [1.18–2.54]	0.005*
Obesity (Yes)	2.09 [1.42–3.07]	0.001**
*Random Effects*	Coef
Estimated Variance of the Intercept	0.05 [0.00–1.46]	0.555
ICC	0.016

OROdds ratio, CI Confidence interval, Coef. Coefficient, OGTT Oral glucose tolerance test, POC-cHbA1c Point-of-care capillary glycated hemoglobin, ICC Intra-class correlation
^*p < 0.05
**p < 0.001
Table 3 presents the results of a two-level mixed-effects logistic regression model for the uptake rate of OGTT, adjusting for patient characteristics and accounting for clustering at the clinic level. An unconditional means model was run to assess the need for multilevel modelling. Substantial evidence of clustering at the clinic level justified the use of multilevel logistic regression
The fixed effects include the intervention condition and patient characteristics, with random intercepts for clinics to account for within-clinic correlation
Odds Ratios (OR) represent the change in odds of OGTT uptake per unit change in predictor, with 95% confidence intervals (CI) provided for each OR

ISSN: 1741-7015