Abstract P217 – Table 1. Results of multivariable generalised linear mixed model regressions for the occurrence of blips and level of residual viraemia.
	Occurrence of blips		Level of residual viraemiaa
Variables	OR (95% CI)	p‐value	OR (95% CI)	p‐value
Time since study inclusion (per year increase)	0.87 (0.78 to 0.98)	0.02	0.88 (0.86 to 0.90)	<0.001
Female sex (vs male sex)	0.90 (0.61 to 1.31)	0.56	0.90 (0.77 to 1.05)	0.17
Age at start of follow‐up (per year increase)	1.02 (0.89 to 1.19)	0.76	1.00 (1.00 to 1.00)	0.36
cART anchor
INSTI	−b	−b	1	−
PI	−b	−b	1.29 (1.15 to 1.44)	<0.001
NNRTI	−b	−b	0.76 (0.68 to 0.85)	<0.001
Time since ART initiation (per year increase)	0.97 (0.94 to 1.00)	0.06	0.95 (0.94 to 0.96)	<0.001
Fiebig stage VI at ART initiation (vs stage I to V)	1.99 (0.66 to 6.05)	0.23	1.51 (1.04 to 2.20)	0.03
Lowest available CD4+ count (per 10 cells/mm3 increase)	1.00 (0.99 to 1.01)	0.56	0.99 (0.99 to 1.00)	<0.001
Zenith VL copies/mL
<10 000	1	−	1	−
10 000 to 99 999	2.03 (0.66 to 6.23)	0.22	2.10 (1.54 to 2.85)	<0.001
100 000 to 999 999	2.80 (0.90 to 8.76)	0.08	3.49 (2.59 to 4.70)	<0.001
≥1 000 000	5.16 (1.54 to 17.29)	0.01	5.40 (3.59 to 8.13)	<0.001
Residual viraemia copies/mL
RNA‐	1	−	−	−
<20 copies/mL	2.72 (2.03 to 3.67)	<0.001	−	−
20 to 49 copies/mL	5.00 (3.49 to 7.17)	<0.001	−	−

cART, combination antiretroviral therapy; INSTI, integrase strand transfer inhibitor; PI, protease inhibitor; VL, viral load.

^aRV, categorised as an ordinal outcome (RNA‐, <20 copies/mL and 20 to 49 copies/mL), was analysed using backward continuation ratio model ordinal regression. The odds of <20 copies/mL versus RNA‐ and 20 to 49 copies/mL versus <20 copies/mL and RNA‐ were investigated and since the proportional odds assumption was not violated, one OR was given per variable;

^bas the variable ‘cART anchor’ is in the causal pathway as a precursor for the association between residual viraemia and the occurrence of viral blips, it was not included in the regression.