| Bias | Author's judgement | Support for judgement |
| Randomization |
Low |
Quote: The permuted block (four patients per block) randomization sequence, including stratification, was prepared by a statistician not involved in the trial, using SAS software, version 9.4 (SAS Institute). To minimize allocation bias, we performed allocation concealment with an interactive Web-based response system until randomization was finished on the system through a computer or phone. |
| Deviations from intervention |
Some concerns |
Comment: Unblinded study. A small number of participants crossed over (3 out of 100; 2 were assigned to the lopinavir–ritonavir arm but did not receive the intervention, 1 was assigned to the standard care arm but received lopinavir-ritonavir). Treatment non-receipt for 3 patients who died before receiving the treatment. A comparable proportion of participants received glucocorticoids in the two arms (33,7% vs 32,3%). Efficacy outcome data were analyzed by using modified intention-to-treat analysis. Safety outcome data (adverse events) were analyzed by using "naïve" per protocol/as-treated analysis (2 participants who did not receive lopinavir-ritonavir were excluded from the intervention group; one participant who was allocated to the control group but received lopinavir-ritonavir was analysed in the intervention group). |
| Missing outcome data |
Low |
Comment: 199 randomized, 199 analyzed for all outcomes except viral negative conversion (n=130 analyzed, excluding participants without detectable viral RNA at baseline; missingness is likely dependent on the true value of the outcome). Risk assessed to be "low" for the outcomes: Mortality. Time to death. Adverse events. Serious adverse events. WHO Clinical progression scale score 6 or above . WHO Clinical progression scale score 7 or above. Time to Clinical Improvement. Clinical improvement incidence. Risk assessed to be "high" for the outcomes: Viral negative conversion incidence. |
| Measurement of the outcome |
Some concerns |
Comment: The outcomes that concern clinical improvement/progression are observer-measured outcomes that involve clinical decision-making, and the assessment could theoretically be influence by knowledge of the intervention assignment, though we do not consider this likely in the context of the pandemic. The outcomes in question are not statistically significant. We consider that the outcome WHO clinical progression scale score 7 or above cannot be influenced by knowledge of the intervention assignment. Some adverse-events were patient-reported or involved assessor judgment. The remaining outcomes are observer-measured and do not involve clinical decision-making. Risk assessed to be "low" for the outcomes: Mortality. Time to death. Viral Negative Conversion incidence. WHO Clinical progression scale score 7 or above. Risk assessed to be "some concerns" for the outcomes: WHO Clinical progression scale score 6 or above. Time to Clinical Improvement. Clinical improvement incidence. Adverse events. Serious adverse events. |
| Selection of the reported results |
Some concerns |
Comment: The protocol was available (in Chinese with the final part presented in English regarding outcomes; the part that was in Chinese was examined by a native speaker). The statistical analysis plan was not available. The analysis reported in the paper follows the plan outlined in the protocol (i.e., reporting of both ITT and per-protocol analysis), with the exception of viral negative conversion incidence, for which the ITT analysis is not reported in the paper. Risk assessed to be "low" for the outcomes: Mortality. Time to death. Adverse events. Serious adverse events. WHO Clinical progression scale score 6 or above . WHO Clinical progression scale score 7 or above. Time to Clinical Improvement. Clinical improvement incidence. Risk assessed to be "some concerns" for the outcomes: Viral Negative Conversion incidence. |
| Overall risk of bias |
Some concerns |