Note: The risk of bias by domain corresponds to the highest risk of bias among outcomes by domain.
The overall risk of bias corresponds to the overall highest risk of bias assessed among outcomes.
| Bias | Author's judgement | Support for judgement |
| Confounding |
Serious |
The analysis adjusted for age, sex, socioeconomic status, and geographic location. In addition to these, the authors controlled for contact frequency, which occurred post-intervention and could have been affected by vaccination status. Uncontrolled confounding is likely, for example due to ethnicity, health-seeking behaviour, specific populations, comorbidities, calendar time, covid symptoms at time of potential vaccination, hospitalization and need for health care. |
| Selection of participants into the study |
Serious |
We suspect the study is at serious risk of selection bias. First, participants were vaccinated at different, unknown time points before start of follow-up (which was the last contact with a confirmed COVID case) and the effect of vaccination is likely to change over time (it is expected to build up and then wane), so vaccine effect may have been different for different participants during the follow-up period. Second, the reason for selection into the study (being a close contact to a confirmed COVID case) is likely to influenced by vaccination status and associated with the outcome (COVID illness) and there was no mention of any methods to adjust for this. |
| Clasification of interventions |
Moderate |
Vaccination status was based on interviewing individuals and reviewing vaccination records. There is potential for recall bias/differential misclassification, but this is considered unlikely in the pandemic context. |
| Deviations from intervention |
Low |
No concerns in this domain - the study was observational. |
| Missing outcome data |
Low |
Nearly all of those identified were included in the analysis. |
| Measurement of the outcome |
Moderate |
The assessment of symptomatic infection involves clinical judgement and could have been influenced by knowledge of the intervention received, but we think it an important biase is unlikely because participants were mandatorily quarantined in centralized managed facilities in the context of a pandemic. |
| Selection of the reported results |
Moderate |
There is no evidence of a protocol, or and analysis plan or and study registry and we have concerns that results could have been selected for reporting because of the findings. |
| Overall risk of bias |
Serious |
|
| Overall comment | We have concerns about uncontrolled confounding (due to the observational nature of the data) and about selection bias due to how participants were selected. |