In their Letter to the Editor, Riggs et al1 describe the concept of immortal time bias as a possible limitation of our retrospective cohort study evaluating the efficacy of tixagevimab-cilgavimab pre-exposure prophylaxis in solid organ transplant recipients (SOTRs).2 Immortal time bias can occur when study participants cannot experience the outcome during a period of follow-up time (Fig. 1 A). A frequently used example is evaluating the efficacy of inhaled glucocorticoids in reducing readmission and death rates after hospitalization for acute exacerbation of chronic obstructive pulmonary disease.3 If follow-up for the control group is started on the day of discharge and follow-up for the treatment group is started on the day of obtaining the prescription, the time between discharge and obtaining the prescription in the inhaled glucocorticoid group would be classified as immortal time, because an event occurring during that time would not be included in a fixed-time analysis. This is an important concept because it can alter the findings in favor of the intervention group.
Figure 1.
Immortal time bias in observational cohort studies. (A) Time-fixed analysis showing immortal time that could potentially impact the study findings. (B) Prescription time-distribution matching strategy to reduce the risk of immortal time bias.
In their Letter, Riggs et al1 proposed that our findings could be subject to immortal time bias for the following 2 reasons: (1) infections were more common in the early period of follow-up, and (2) early infections in the study could result in exclusion from receiving tixagevimab-cilgavimab and therefore count as an “additional” breakthrough event in the control group, favoring a protective effect in the intervention group. Immortal time bias is unlikely to have significantly affected our main finding of lower risk of breakthrough severe acute respiratory syndrome-coronavirus-2 (SARS-CoV-2) infection in SOTRs who received tixagevimab-cilgavimab compared with the vaccine–matched control group for several reasons. First, we used a prescription time-distribution matching strategy (Fig. 1B), which has been shown to reduce the risk of immortal time bias.4 This strategy involved starting the follow-up of each vaccine–matched control on the same day that their counterpart in the intervention group received tixagevimab-cilgavimab. A description of when the follow-up for each participant in the control group is present in the methods, although we acknowledge that the specific phrase “prescription time-distribution matching strategy” was not used and its association with lower risk of immortal time bias was not discussed. Using this methodology, any SARS-CoV-2 infection event that occurred in the control group before the matched start date would not count as a breakthrough infection event and therefore would not impact the estimated protective effect of tixagevimab-cilgavimab pre-exposure prophylaxis. In addition, because previous SARS-CoV-2 infection is associated with a lower future risk of breakthrough SARS-CoV-2 infections,5 and there was a higher proportion of SOTRs with a history of SARS-CoV-2 infection in the control group, this would bias our findings toward a lower risk of infection in the control group and therefore underestimate the protective effect of tixagevimab-cilgavimab pre-exposure prophylaxis. Furthermore, review of the 42 participants in the control group with previous SARS-CoV-2 infection showed that the timing of previous infection would have precluded receiving tixagevimab-cilgavimab for only 3 individuals of the control cohort (n = 222) by their matched initiation date; however, we acknowledge that this was not described in the original manuscript. Finally, this matching strategy reduces bias from the varying incidence of SARS-CoV-2 infection over time. Based on these reasons, we believe that it is unlikely that immortal time bias would have had a significant impact on our findings.
Funding
There was no specific funding for this letter.
Disclosure
The authors of this manuscript have no conflicts of interest to disclose as described by the American Journal of Transplantation.
Declaration of interests
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Acknowledgments
None
References
- 1.Riggs J., Charya A., Eberlein M., Reed R., Saharia K. Tixagevimab/cilgavimab pre-exposure prophylaxis and breakthrough infection risk in vaccinated solid organ transplant recipients: concern for immortal time bias effect. Am J Transpl. 2022 doi: 10.1016/j.ajt.2022.11.021. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Al Jurdi A., Morena L., Cote M., Bethea E., Azzi J., Riella L.V. Tixagevimab/cilgavimab pre-exposure prophylaxis is associated with lower breakthrough infection risk in vaccinated solid organ transplant recipients during the Omicron wave. Am J Transplant. 2022;22(12):3130–3136. doi: 10.1111/ajt.17128. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Suissa S. Effectiveness of inhaled corticosteroids in chronic obstructive pulmonary disease: immortal time bias in observational studies. Am J Respir Crit Care Med. 2003;168(1):49–53. doi: 10.1164/rccm.200210-1231OC. [DOI] [PubMed] [Google Scholar]
- 4.Zhou Z., Rahme E., Abrahamowicz M., Pilote L. Survival bias associated with time-to-treatment initiation in drug effectiveness evaluation: a comparison of methods. Am J Epidemiol. 2005;162(10):1016–1023. doi: 10.1093/aje/kwi307. [DOI] [PubMed] [Google Scholar]
- 5.Altarawneh H.N., Chemaitelly H., Hasan M.R., et al. Protection against the Omicron variant from previous SARS-CoV-2 infection. N Engl J Med. 2022;386(13):1288–1290. doi: 10.1056/NEJMc2200133. [DOI] [PMC free article] [PubMed] [Google Scholar]