Corrigendum to ‘Oral Azvudine for hospitalised patients with COVID-19 and pre-existing conditions: a retrospective cohort study.’ [EClinicalMedicine 59 (2023) 101981]

We would like to issue a correction for a mistaken phrase ‘time from symptom onset to treatment exposure’ in our published study and replace it with ‘time from symptom onset to admission’ as we stated in Data Source and Supplementary Fig. 1. We provide the detailed instructions for the modifications and the revised version of Fig. 3.

• (1)Six modifications (replacing “time from symptom onset to treatment exposure” with “time from symptom onset to admission”) are needed which appeared in Summary/Methods, Baseline covariates, Statistical analysis, Results, Table 1 and Supplementary Table 1.
• (2)Similarly, “About 87.3% patients were treated with Azvudine above 5 days of the symptom's onset” in Results should be changed to “About 87.3% patients were admitted above 5 days of the symptom's onset”.
• (3)“Onset to exposure” in Fig. 3 should be changed to “Onset to admission”

Fig. 3.

The effectiveness of Azvudine in reducing the risk of composite disease progression outcome and risk of all-cause death by subgroups of selected baseline characteristics. Abbreviation, COPD, chronic obstructive pulmonary disease. Ellipsis (…) means that the model does not converge due to few outcomes; Horizontal lines indicate the ranges of the 95% CIs and the vertical dash lines indicate the hazard ratio of 1.

Table 1.

Baseline characteristics of the participants.

Baseline characteristics	Before matching			After 1:1 propensity-score matching
	Azvudine (n = 245)	Controls (n = 722)	SMD	Azvudine (n = 245)	Matched controls (n = 245)	SMD
Age (years), mean (SD)	69.13 (13.4)	66.89 (14.8)	0.159	69.13 (13.4)	69.25 (14.0)	0.022
Sex, n (%)			0.079			0.025
Men	154 (62.9)	426 (59.0)		154 (62.9)	157 (64.1)
Women	91 (37.1)	296 (41.0)		91 (37.1)	88 (35.9)
COVID-19 vaccination status			0.010			0.016
Vaccinated	118 (48.2)	344 (47.6)		118 (48.2)	120 (49.0)
Unknown	127 (51.8)	378 (52.4)		127 (51.8)	125 (51.0)
Time from symptom onset to admission, n (%)			0.334			0.012
0–5 days	214 (87.3)	537 (74.4)		214 (87.3)	215 (87.8)
>5 days	31 (12.7)	185 (25.6)		31 (12.7)	30 (12.2)
Severity at admission, n (%)			0.189			0.060
Non-severe	88 (35.9)	196 (27.1)		88 (35.9)	81 (33.1)
Severe	157 (64.1)	526 (72.9)		157 (64.1)	164 (66.9)
Concomitant treatments initiated at admission, n (%)
Systemic steroid	64 (26.1)	152 (21.1)	0.119	64 (26.1)	56 (22.9)	0.076
Antibiotics	130 (53.1)	265 (36.7)	0.333	130 (53.1)	138 (56.3)	0.065
Immunomodulators	25 (10.2)	57 (7.9)	0.080	25 (10.2)	22 (9.0)	0.042

SMD, Standard mean difference.

Supplementary Table 1.

Propensity-score model conditional on baseline covariates.

Covariates	Coefficients	Standard Errors	P values
Intercept	−2.185619	0.522340	2.86e-05
Age (years)	0.004412	0.005592	0.430118
Gender (female vs. male)	−0.077771	0.157773	0.622062
Severity at admission (severe vs. non-severe)	−0.146920	0.166242	0.376818
COVID-19 vaccination status (vaccinated vs. unknown)	0.149249	0.154748	0.334811
Time from symptom onset to admission (0–5 vs. >5 days)	0.725780	0.218937	0.000916
Systemic steroid (yes vs. no)	0.035365	0.181723	0.845697
Antibiotics (yes vs. no)	0.544309	0.155723	0.000473
Immunomodulators (yes vs. no)	0.550108	0.194569	0.004694

The phrase “time from symptom onset to treatment exposure” initially used in the paper refers to the duration from the onset of COVID-19 symptoms to the period of treatment exposure, while the patients in the control group did not take any antiviral agents during the observation period. In fact, the variable we used for propensity-score matching in the study is the time from symptom onset to admission, but we mistakenly used the expression ‘time from symptom set to treatment exposure’. We want to correct this error in the article and it did not change the results, direction, or conclusions of our study.