I enjoyed reading Talukder et al’s1 retrospective comparison of outcomes after either endovascular or open surgical repair of the descending thoracic aorta using the TriNetX network. I am concerned that the authors have not provided enough methodological details to understand exactly how their patient cohorts were constructed. The authors provide the clinical codes used to define their outcomes, and state that the codes used to construct their cohorts are available in their attached Appendix. However, this file does not define any of the clinical codes used. Furthermore, the temporal relationship between the inclusion and exclusion criteria for each cohort is not well-defined.
These details can greatly influence the findings of TriNetX-based studies.2 For example, in the comparison of repair approaches for unruptured aneurysms, it is unclear if the exclusion of aneurysm rupture is set to before the index surgery or for the entirety of the patient's record. If the latter was performed, then this may underestimate the rate of complications because it would exclude patients with subsequent ruptures resulting from surgical complications such as endoleak.3
When an outcome analysis is run on TriNetX, the user can export a word doc summary of the study, which includes tables of the exact cohort designs used in the analysis (Table). I would encourage the authors of the study to provide these tables, allowing for readers to more deeply engage with their findings.
Table.
Example exported cohort design demonstrating endovascular repair patients without a prior rupture
| Group 1 | ||||
|---|---|---|---|---|
| Group 1A | ||||
| Must have | Any of | Procedure | UMLS:CPT:1006313 | Endovascular repair procedures of the descending thoracic aorta |
| Procedure | UMLS:CPT:1006314 | Endovascular repair of descending thoracic aorta (eg, aneurysm, pseudoaneurysm, dissection, penetrating ulcer, intramural hematoma, or traumatic disruption) | ||
| Cannot have | Procedure | UMLS:ICD10PCS:02QW0ZZ | Repair thoracic aorta, descending, open approach | |
| or | Diagnosis | UMLS:ICD10CM:I71.012 | Dissection of descending thoracic aorta | |
| Date constraint | The terms in this group occurred between January 1, 2000, and June 14, 2025 | |||
| Event relationship | Any instance of group 1B occurred on or before the first instance of group 1A | |||
| Group 1B | ||||
| Cannot have | Diagnosis | UMLS:ICD10CM:I71.13 | Aneurysm of the descending thoracic aorta, ruptured | |
| or | Diagnosis | UMLS:ICD10CM:I71.8 | Aortic aneurysm of unspecified site, ruptured | |
| or | Diagnosis | UMLS:ICD10CM:I71.51 | Supraceliac aneurysm of the thoracoabdominal aorta, ruptured | |
| or | Diagnosis | UMLS:ICD10CM:I71.52 | Paravisceral aneurysm of the thoracoabdominal aorta, ruptured | |
| or | Diagnosis | UMLS:ICD10CM:I71.11 | Aneurysm of the ascending aorta, ruptured | |
This design includes both the clinical codes used as inclusion/exclusion criterion, and the temporal relationships between those codes.
Funding
None.
Disclosures
None.
References
- 1.Talukder R., Montez K., Moin Z., et al. Comparing neurological complications between endovascular and open surgical repair of the descending thoracic aorta. J Vasc Surg Cases Innov Tech. 2026;12:102086. doi: 10.1016/j.jvscit.2025.102086. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Chen H.D., Wang J. Event timing inaccuracies in a retrospective study of low-dose aspirin and acute pancreatitis. Pancreatology. 2025 doi: 10.1016/j.pan.2025.11.006. [DOI] [PubMed] [Google Scholar]
- 3.Szmidt J., Galazka Z., Rowinski O., et al. Late aneurysm rupture after endovascular abdominal aneurysm repair. Interact Cardiovasc Thorac Surg. 2007;6:490–494. doi: 10.1510/icvts.2007.152447. [DOI] [PubMed] [Google Scholar]