TABLE 1—

Logistic Regression Model of Self-Reported Arrests for Syringe Possession Among Municipal Police After SHIELD Training: Tijuana, Mexico, 2015–2018

Variable	OR (95% CI)
Unadjusted Model	Adjusted Model
Time (visit)	0.86 (0.83, 0.88)	0.87 (0.85, 0.90)
Female sex (vs male)	0.49 (0.32, 0.75)	0.56 (0.35, 0.87)
Age, y	1.02 (0.99, 1.03)
Less than high school education (vs more)	1.11 (0.97, 1.26)
Work experience (years)	0.89 (0.82, 0.97)
Correct knowledge of syringe laws (vs incorrect)	0.51 (0.44, 0.59)	0.63 (0.44, 0.89)
Positive attitudes toward PWID (vs negative)	1.33 (1.14, 1.55)
Patrol assignment (vs administrative duty)	3.01 (2.02, 4.48)	6.71 (3.83, 11.76)
High drug use precinct (vs low)	1.90 (1.57, 2.31)	2.42 (1.56, 3.77)

Note. CI = confidence interval; OR = odds ratio; PWID = people who inject drugs; SHIELD = Safety and Health Integration in the Enforcement of Laws on Drugs program. Generalized estimating equations with an exchangeable correlation structure were used to conduct logistic regression modeling. Arrests for syringe possession refer to the previous 6 months. The sample size was 693 (3523 observations).