Table 1. Regression model and estimates of annual percent variation of oral health care coverage and primary care coverage and the difference among time variation according to Municipal Human Development Index (MHDI) from 2008 to 2015 in Brazil.

		Oral health care coverage	Primary care coverage
Year (time)		0.0265 (0.0234–0.0296)	0.0223 (0.0199–0.0244
MHDI
	Medium	26.98 (19.13–34.82)	16.21 (101.47–232.99)
	High + very high	39.29 (31.24–47.34)	178.20 (11.32–243.92)
Population size
	< 5 k
	5–9.9 k	-0.0904 (-0.1211 to -0.0596)	-0.0098 (-0.0270–0.0075)
	10–49.9 k	-0.2823 (-0.3108 to -0.2538)	-0.0891 (-0.1051 to -0.0732)
	> 50 k	-0.6926 (-0.7344 to -0.6507)	-0.3500 (-0.3735 to -0.3266)
Interaction MHDI * year
	Medium	-0.0134 (-0.0173 to -0.0095)	-0.0081 (-0.01112 to -0.0051)
	High + very high	-0.0195 (-0.0235 to -0.0155)	-0.0091 (-0.0122 to -0.0057)
	Intercept	-50.14 (-56.35 to -43.93)	-4039.95 (-445.90 to -364.01)
Random effects parameters
	Variance (standard error) (year)	0.0026 (0.0001)	0.0013 (0.0001)
	Variance (standard error) (_cons)	1057.8 (254.84)	5477.37 (191.96)
	Covariance (standard error) (year,_cons)	-5.25 (0.13)	-2.72 (0.10)
Marginal estimates – Annual variation of the indicator (Y variation / time variation - DY/DT)
	Low + very low MDHI	0.0265 (0.0248–0.0281)	0.0223 (0.01958–0.0243)
	Medium MDHI	0.0131 (0.0119–0.0145)	0.0139 (0.01196–0.0157)
	High + very high MDHI	0.0070 (0.0055–0.0084)	0.0131 (0.0111–0.0151)
Wald test – difference among slopes
	Low + very low x medium	0.0265 (p < 0.001)	0.0081 (p < 0.001)
	Low + very low x high + very high	0.0131 (p < 0.001)	0.0091 (p < 0.001)
	Medium x high + very high	0.0069 (p < 0.001)	0.0004 (p = 0.861)

^*The models were adjusted for the proportion of women, older adults, residents in rural areas, and fluoride presence in water supply.