. 2022 Dec 17;22:2372. doi: 10.1186/s12889-022-14530-w

Table 3.

The ratio of average U5MR between countries in high PMH group and those with low PMH group

		Data	Data	Data	Data	Change in data
		1990	2000	2010	2019	Δ_{1990 − 2019}
Correlation 1990		1.0837	1.0540	1.0627	1.0667	−0.0171
Correlation 1990		(0.0013)	(0.0009)	(0.0010)	(0.0012)	(0.0018)
Correlation 2000		1.1003	1.0680	1.0764	1.0804	−0.0199
Correlation 2000		(0.0015)	(0.0010)	(0.0012)	(0.0014)	(0.0020)
Correlation 2010		1.0552	1.0368	1.0421	1.0444	−0.0108
Correlation 2010		(0.0009)	(0.0006)	(0.0007)	(0.0008)	(0.0012)
Correlation 2019		1.0961	1.1007	1.0964	1.0470	−0.0138
Correlation 2019		(0.0006)	(0.0005)	(0.0005)	(0.0008)	(0.0012)
Change in correlation	Δ_{1990 − 2019}	−0.0230	−0.0130	− 0.0176	−0.0197	− 0.0368
Change in correlation	Δ_{1990 − 2019}	(0.0005)	(0.0003)	(0.0004)	(0.0004)	(0.0016)

The under-five mortality ratio is calculated as $\hat{r} = {\bar{\hat{U 5 M R}}}_{P M H G R P = 1} / {\bar{\hat{U 5 M R}}}_{P M H G R P = 0}$ , where $\bar{\hat{U 5 M R}}$ stands for the mean of estimated U5MR. And the baseline prediction of $\hat{U 5 M R}$ uses PMH status, education, environment, GDP per capita, health behavior, water and sanitation, urbanization, healthcare level, and HIV prevalence. The reported ratios in Table 3 are computed as the mean of $\hat{r}$ across 200 bootstrap replications, standard errors in parentheses. In each bootstrap replication, we randomly draw 70% of the total sample