Table 3. Results From Nonlinear Regression Modeling of Different Types of Media Items and Weekly Referrals for Each Gender Service.

Variable	Gender service
	RCHGS		GIDS
	β̂ (95% CI)a	P value	β̂ (95% CI)a	P value
Peripherally related media items in week 1, referral wk
1b	−0.17 (−0.42 to −0.07)	.17	0.03 (−0.23 to 0.18)	.78
2c	−0.10 (−0.35 to 0.14)	.42	0.03 (−0.18 to 0.23)	.80
3d	0.10 (−0.15 to 0.35)	.44	0.01 (−0.20 to 0.22)	.91
TGD-focused media items in week 1, referral wk
1e	0.16 (0.04 to 0.28)	.007	0.22 (0.01 to 0.44)	.04
2f	0.23 (0.11 to 0.35)	<.001	−0.04 (−0.25 to 0.18)	.75
3g	0.10 (−0.02 to 0.22)	.10	0.14 (−0.08 to 0.36)	.22
Media items in week 1 that mentioned RCHGS/GIDS, referral wk
1h	−0.06 (−0.28 to 0.17)	.62	0.50 (0.11 to 0.88)	.01
2i	−0.07 (−0.29 to 0.16)	.57	0.57 (0.18 to 0.96)	.004
3j	−0.17 (−0.40 to 0.05)	.13	0.21 (−0.18 to 0.60)	.30

Abbreviations: GIDS, Gender Identity Development Service; RCHGS, Royal Children’s Hospital Gender Service; TGD, transgender and gender diverse.

^aAdjusted for time, which showed a significant association with referrals in each model.

^bReferral in week 1 = β₀ + β₁ × peripherally related media items in week 1 + β₂ × time.

^cReferral in week 2 = β₀ + β₁ × peripherally related media items in week 1 + β₂ × time.

^dReferral in week 3 = β₀ + β₁ × peripherally related media items in week 1 + β₂ × time.

^eReferral in week 1 = β₀ + β₁ × TGD-focused medial items in week 1 + β₂ × time.

^fReferral in week 2 = β₀ + β₁ × TGD-focused medial items in week 1 + β₂ × time.

^gReferral in week 3 = β₀ + β₁ × TGD-focused medial items in week 1 + β₂ × time.

^hReferral in week 1 = β₀ + β₁ × medial items in week 1 that mentioned RCHGS/GIDS + β₂ × time.

ⁱReferral in week 2 = β₀ + β₁ × medial items in week 1 that mentioned RCHGS/GIDS + β₂ × time.

^jReferral in week 3 = β₀ + β₁ × medial items in week 1 that mentioned RCHGS/GIDS + β₂ × time.