The study by Roy et al. on the procedural outcomes of transradial access (TR) versus transfemoral (TF) access for diagnostic cerebral angiography by BMI provides initial data suggesting that TR access may offer advantages in reduced procedure time and length of stay (LOS), particularly for obese patients 1)
However, several critical issues in design, analysis, and generalizability undermine the robustness and clinical relevance of its findings.
### 1. Single-Center, Retrospective Design
-
The retrospective, single-center nature of the study limits generalizability, as it relies on existing records from a single institution. Procedural approaches, patient demographics, and access site expertise vary between institutions, potentially impacting the reported outcomes. A multi-center or randomized design could have yielded broader insights.
### 2. Questionable Clinical Impact of Results
-
While TR access is associated with reduced procedure time by 10-13 minutes across BMI groups, this difference is relatively small in the context of diagnostic cerebral angiography and may not be meaningful in clinical settings. Additionally, the slightly reduced LOS for obese patients (1.33 days) lacks sufficient exploration into the underlying causes, leaving it unclear if access site alone significantly affects recovery times.
### 3. Propensity Score Matching Limitations
-
Although propensity score matching was used to reduce selection bias, it cannot control for unmeasured confounding factors inherent in retrospective designs. Important factors, such as operator experience, anatomical variations, or other patient comorbidities, may still bias the results. A randomized trial or more sophisticated statistical methods could have better addressed these complexities.
### 4. BMI as a Central Focus
-
The emphasis on BMI as a primary stratification factor may be misplaced, as BMI alone is seldom the determining factor in choosing access sites for cerebral angiography. Other variables, such as comorbidities or vascular conditions, might have provided a more clinically relevant framework, and the choice of BMI subgroups without additional analysis on access site suitability limits the depth of the study’s insights.
### 5. Limited Outcome Measures and Follow-Up Data
-
The study reports immediate procedural outcomes but does not assess long-term safety or vascular health implications of TR versus TF access. Long-term follow-up would be essential to capture delayed complications or recurrent access challenges that might arise, especially given the propensity for repeated angiography in certain patients.
### 6. Statistical and Analytical Shortcomings
-
The study’s use of basic statistical tools like linear regression and chi-square tests may not capture complex interactions between BMI, access site, and procedural outcomes. More sophisticated modeling could have yielded deeper insights, particularly in understanding how various confounding factors contribute to access site efficacy and safety across different patient populations.
### 7. Limited Contribution to Existing Literature
-
While the study adds data on TR access safety across BMI groups, its findings do not substantially advance clinical practice due to the limited procedural time savings and lack of evidence that BMI should be a deciding factor in access site selection. The study reiterates existing knowledge on TR access benefits without significantly expanding on their implications in a specialized field like cerebral angiography.
### Conclusion In summary, while this study provides a preliminary comparison of TR versus TF access in diagnostic cerebral angiography, its methodological limitations, reliance on BMI as the main stratification factor, and narrow focus on short-term procedural outcomes reduce its impact. Future studies would benefit from multi-center, prospective designs with randomized control, long-term follow-up, and consideration of broader patient-specific factors to provide a more comprehensive understanding of the ideal access strategies for cerebral angiography across diverse populations.