Application of quantile mixed-effects model in modeling CD4 count from HIV-infected patients in KwaZulu-Natal South Africa

Table 3 Parameter estimates for CAPRISA 002 AI study data across several quantiles

Parameter	\({\widehat{Q}}_{0.05}\) (SE)	\({\widehat{Q}}_{0.25}\) (SE)	\({\widehat{Q}}_{0.5}\) (SE)	\({\widehat{Q}}_{0.75}\) (SE)	\({\widehat{Q}}_{0.85}\) (SE)	\({\widehat{Q}}_{0.95}\) (SE)
Intercept	19.996 (1.161)*	22.171 (1.403)*	24.628 (1.464)*	26.595(1.419)*	27.972 (1.420)*	31.381 (1.397)*
Time	0.063 (0.015)*	0.069 (0.013)*	0.056 (0.013)*	0.046 (0.013)*	0.041 (0.013)*	0.034 (0.015)*
SQRT of time	− 0.866 (0.142)*	− 0.871 (0.129)*	− 0.695 (0.117)*	− 0.593 (0.119)*	− 0.581 (0.124)*	− 0.385 (0.162)*
Baseline BMI	0.056 (0.021)*	0.078 (0.024)*	0.082 (0.026)*	0.112 (0.032)*	0.131 (0.033)*	0.145 (0.030)*
Log of baseline VL	− 0.564 (0.078)*	− 0.568 (0.103)*	− 0.641 (0.096)*	− 0.713 (0.093)*	− 0.714 (0.089)*	− 0.739 (0.084)*
Post HAART initiation	1.683 (0.054)*	2.125 (0.073)*	2.560 (0.088)*	3.021 (0.096)*	3.114(0.097)*	2.287 (0.089)*
Age	0.021 (0.025)	0.029 (0.029)	0.029 (0.031)	0.029 (0.032)	0.026 (0.032)	0.013 (0.030)
Log-lik	− 18,454.68	− 17,169.85	− 16,828.96	− 17,344.63	− 17,952.50	− 19,088.77
AIC	36,937.36	34,367.69	33,685.92	34,717.25	35,933	38,205.55

^*Significance at 5% level. See, Additional file 1, for more significant test results and confidence intervals

ISSN: 1471-2334