Background

Comparative studies and randomized controlled trials (RCTs) often use the P (probability) value to convey the statistical significance of their findings. P values are an imperfect measure, however, and are vulnerable to a small number of outcome reversals to alter statistical significance. The inclusion of a fragility index (FI) and fragility quotient (FQ) may aid in the interpretation of a study's statistical strength.

Purpose/hypothesis

The purpose of this study was to examine the statistical stability of studies comparing single-row to double-row rotator cuff repair. It was hypothesized that the findings of these studies would be vulnerable to a small number of outcome event reversals, often fewer than the number of patients lost to follow-up.

Study design

Systematic review; Level of evidence, 3.

Methods

We analyzed comparative studies and RCTs on primary single-row versus double-row rotator cuff repair that were published between 2000 and 2021 in 10 leading orthopaedic journals. Statistical significance was defined as a P < .05. The FI for each outcome was determined by the number of event reversals necessary to alter significance. The FQ was calculated by dividing the FI by the respective sample size.

Results

Of 4896 studies screened, 22 comparative studies, 10 of which were RCTs, were ultimately included for analysis. A total of 74 outcomes were examined. Overall, the median FI was 2 (interquartile range [IQR], 1-3), and the median FQ was 0.035 (IQR, 0.020-0.057). The mean FI was 2.55 ± 1.29, and the mean FQ was 0.043 ± 0.027. In 64% of outcomes, the FI was less than the number of patients lost to follow-up.) Additionally, 81% of significant outcomes needed just a single outcome reversal to lose their significance.

Conclusion

Over half of the studies currently used to guide clinical practice have a number of patients lost to follow-up greater than their FI. The results of these studies should be interpreted within the context of these limitations. Future analyses may benefit from the inclusion of the FI and the FQ in their statistical analyses.

Objective

Venous thromboembolism (VTE) is an important cause of postoperative morbidity and mortality. However, the reported incidence after major vascular surgery has ranged from as low as 1% to >10%. Furthermore, little is known about optimal chemoprophylaxis regimens or rates of postdischarge VTE in this population. In the present study, we aimed to better characterize the rates of in-hospital and postdischarge VTE after major vascular surgery, the role of chemoprophylaxis timing, and the association of VTE with mortality.

Methods

A single-center retrospective study of 1449 major vascular operations (2013-2020) was performed and included 189 endovascular abdominal aortic aneurysm repairs (13%), 169 thoracic endovascular aortic aneurysm repairs (12%), 318 open aortic operations (22%), 640 lower extremity bypasses (44%), and 133 femoral endarterectomies (9%). The baseline characteristics, anticoagulant and antiplatelet medications, and outcomes were abstracted from an electronic database with medical record auditing. Postoperative VTE (pulmonary embolism and deep vein thrombosis) within 90 days of surgery was classified by the location, symptoms, and treatment. A cut point analysis using Youden's index identified the most VTE discriminating timing of chemoprophylaxis (including therapeutic vs prophylactic anticoagulant and antiplatelet medications) and Caprini score. Multivariable logistic regression was used to test the association of VTE with chemoprophylaxis timing, Caprini score, and additional risk factors. Cox proportional hazard modeling was used to measure the association between VTE and mortality.

Results

The overall VTE incidence was 3.4% (65% deep vein thrombosis; 25% pulmonary embolism; 10% both), and 37% had occurred after discharge. The rate of symptomatic VTE was 2.4%, which was lowest for endovascular abdominal aortic aneurysm repair (0.0%) and highest for open aortic surgery (4.1%; P = .02). Those who had developed VTE had had a longer length of stay, higher rates of end-stage renal disease and prior VTE, and higher Caprini scores (8 vs 5 points; P < .01 for all). Those who had developed VTE were also more likely to have received ≥2 U of blood postoperatively, required an unplanned return to the operating room, had delayed chemoprophylaxis, anticoagulation, and/or antiplatelet initiation of >4 days postoperatively, and had increased 90-day mortality (P < .01 for all). A Caprini score of ≥7 (29% of patients) was associated with postdischarge VTE (2.6% vs 0.7%; P = .01), and chemoprophylaxis, anticoagulation, and antiplatelet timing of >4 days was associated with an increased adjusted odds of VTE (odds ratio, 2.4; 95% confidence interval, 1.1-4.9). Although no fatal VTEs were identified, VTE was an independent predictor of 90-day mortality (adjusted hazard ratio, 2.7; 95% confidence interval, 1.3-5.9).

Conclusions

These data have shown that patients undergoing major vascular surgery are particularly prone to the development of VTE, with frequent hypercoagulable comorbidities. The earlier initiation of chemoprophylaxis was associated with a reduced risk of VTE development. Furthermore, the postdischarge VTE rates might reach thresholds warranting postdischarge chemoprophylaxis, especially for patients with a Caprini score of ≥7.