The Statistical Fragility of Tranexamic Acid Use in the Orthopaedic Surgery Literature: A Systematic Review of Randomized Controlled Trials.

INTRODUCTION: Randomized controlled trials (RCTs) represent the highest level of evidence in orthopaedic surgery literature, although the robustness of statistical findings in these trials may be unreliable. We used the fragility index (FI), reverse fragility index (rFI), and fragility quotient (FQ) to evaluate the statistical stability of outcomes reported in RCTs that assess the use of tranexamic acid (TXA) across orthopaedic subspecialties. METHODS: PubMed, EMBASE, and MEDLINE were queried for RCTs (2010-present) reporting dichotomous outcomes with study groups stratified by TXA administration. The FI and rFI were defined as the number of outcome event reversals needed to alter the significance level of significant and nonsignificant outcomes, respectively. FQ was determined by dividing the FI or rFI by sample size. Subgroup analyses were conducted based on orthopaedic subspecialty. RESULTS: Six hundred five RCTs were screened with 108 studies included for analysis comprising 192 total outcomes. The median FI of the 192 outcomes was 4 (IQR 2 to 5) with an associated FQ of 0.03 (IQR 0.019 to 0.050). 45 outcomes were reported as statistically significant with a median FI of 1 (IQR 1 to 5) and associated FQ of 0.02 (IQR 0.011 to 0.034). 147 outcomes were reported as nonsignificant with a median rFI of 4 (IQR 3 to 5) and associated FQ of 0.04 (IQR 0.023 to 0.051). The adult reconstruction, trauma, and spine subspecialties had a median FI of 4. Sports had a median FI of 3. Shoulder and elbow and foot and ankle had median FIs of 6. DISCUSSION: Statistical outcomes reported in RCTs on the use of TXA in orthopaedic surgery are fragile. Reversal of a few outcomes is sufficient to alter statistical significance. We recommend reporting FI, rFI, and FQ metrics to aid in interpreting the outcomes reported in comparative trials.

The Fragility of Statistical Findings in the Femoral Neck Fracture Literature: A Systematic Review of Randomized Controlled Trials.

OBJECTIVES: Randomized controlled trials (RCTs) in the femoral neck fracture literature frequently report P -values for outcomes, which have substantial implications in guiding surgical management. This study used the fragility index (FI), reverse fragility index (rFI), and fragility quotient (FQ) to assess the statistical stability of outcomes reported in RCTs evaluating the management and treatment of femoral neck fractures. DESIGN: PubMed, Embase, and MEDLINE were queried for RCTs (January 1, 2010 to February 28, 2023). SETTING: RCTs that evaluated surgical management or treatment of femoral neck fractures were included. STUDY SELECTION CRITERIA: RCTs with 2 treatment arms reporting categorical dichotomous outcomes were included. Non-RCT studies, RCTs with greater than 2 treatment arms, and RCTs without a femoral neck fracture cohort were excluded. OUTCOME MEASURES AND COMPARISONS: The FI and rFI were calculated as the number of outcome event reversals required to alter statistical significance for significant ( P < 0.05) and nonsignificant ( P ≥ 0.05) outcomes, respectively. The FQ was calculated by dividing the FI by the sample size for the study. RESULTS: Nine hundred eighty-five articles were screened, with 71 studies included for analysis. The median FI across a total of 197 outcomes was 4 [interquartile range (IQR) 2-5] with an associated FQ of 0.033 (IQR 0.017-0.060). Forty-seven outcomes were statistically significant with a median FI of 2 (IQR 1-4) and associated FQ of 0.02 (IQR 0.014-0.043). One hundred fifty outcomes were statistically nonsignificant with a median rFI of 4 (IQR 3-5) and associated FQ of 0.037 (IQR 0.019-0.065). CONCLUSIONS: Statistical findings in femoral neck fracture RCTs are fragile, with reversal of a median 4 outcomes altering significance of study findings. The authors thus recommend standardized reporting of P -values with FI and FQ metrics to aid in interpreting the robustness of outcomes in femoral neck fracture RCTs. LEVEL OF EVIDENCE: Therapeutic Level III. See Instructions for Authors for a complete description of levels of evidence.

Statistical Outcomes Guiding Periprosthetic Joint Infection Prevention and Revision Are Fragile: A Systematic Review of Randomized Controlled Trials.

BACKGROUND: Dichotomous outcomes are frequently reported in orthopaedic research and have substantial clinical implications. This study utilizes the fragility index (FI) and fragility quotient (FQ) metrics to determine the statistical stability of outcomes reported in total joint arthroplasty randomized controlled trials (RCTs) relating to periprosthetic joint infection (PJI). METHODS: The RCTs that reported dichotomous data related to PJI published between January 1, 2010, and December 31, 2022, were evaluated. The FI and reverse FI (RFI) were defined as the number of outcome event reversals required to reverse the significance of significant and nonsignificant outcomes, respectively. The FQ was determined by dividing the FI or RFI by the respective sample size. There were 108 RCTs screened, and 17 studies included for analysis. RESULTS: A total of 58 outcome events were identified, with a median FI of 4 (interquartile range [IQR] 2 to 5) and associated FQ of 0.0417 (IQR 0.0145 to 0.0602). The 13 statistically significant outcomes had a median FI of 1 (IQR 1 to 2) and FQ of 0.00935 (IQR 0.00629 to 0.01410). The 45 nonsignificant outcomes had a median RFI of 4 (IQR 3 to 5) and FQ of 0.05 (IQR 0.0361 to 0.0723). The number of patients lost to follow-up was greater than or equal to the FI in 46.6% of outcomes. CONCLUSIONS: Statistical outcomes in RCTs analyzing PJI are fragile and may lack statistical integrity. We recommend a comprehensive fragility analysis, with the reporting of FI and FQ metrics, to aid in the interpretation of outcomes in the total joint arthroplasty literature.

Randomized Controlled Trials Comparing Bone-Patellar Tendon-Bone Versus Hamstring Tendon Autografts in Anterior Cruciate Ligament Reconstruction Surgery Are Statistically Fragile: A Systematic Review.

PURPOSE: To assess the statistical fragility of recently published randomized controlled trials (RCTs) comparing the use of hamstring tendon autograft with bone-patellar tendon-bone autograft for anterior cruciate ligament (ACL) reconstruction. METHODS: The PubMed, Embase, and MEDLINE databases were queried for RCTs published since 2010 comparing autograft type (bone-patellar tendon-bone vs hamstring tendon) in ACL reconstruction surgery. The fragility index (FI) and reverse FI (rFI) were determined for significant and nonsignificant outcomes, respectively, as the number of outcome reversals required to change statistical significance. The fragility quotient (FQ) and reverse FQ, representing fragility as a proportion of the study population, were calculated by dividing the FI and rFI, respectively, by the sample size. RESULTS: We identified 19 RCTs reporting 55 total dichotomous outcomes. The median FI of the 55 total outcomes was 5 (interquartile range [IQR], 4-7), meaning a median of 5 outcome event reversals would alter the outcomes' significance. Five outcomes were reported as statistically significant with a median FI of 4 (IQR, 2-6), meaning a median of 4 outcome event reversals would change outcomes to be nonsignificant. Fifty outcomes were reported as nonsignificant with a median rFI of 5 (IQR, 4-7), meaning a median of 5 outcome event reversals would change outcomes to be significant. The FQ and reverse FQ for significant and nonsignificant outcomes were 0.025 (IQR, 0.018-0.045) and 0.082 (IQR, 0.041-0.106), respectively. For 61.8% of outcomes, patients lost to follow-up exceeded the corresponding FI or rFI. CONCLUSIONS: There is substantial statistical fragility in recent RCTs on autograft choice in ACL reconstruction surgery given that altering a few outcome events is sufficient to reverse study findings. For over half of outcomes, maintaining patients lost to follow-up may have been sufficient to reverse study conclusions. CLINICAL RELEVANCE: We recommend co-reporting FIs and P values to provide a more comprehensive representation of a study's conclusions when conducting an RCT.

The Fragility of Tourniquet Use in Total Knee Arthroplasty: A Systematic Review of Randomized Controlled Trials.

BACKGROUND: Physicians utilize P-values to interpret clinical trial data and guide patient-care decisions. Fragility analysis assesses the stability of statistical findings in relation to outcome event reversals. This study assessed the statistical fragility of recent randomized controlled trials (RCTs) investigating tourniquet use in total knee arthroplasty (TKA). METHODS: We queried PubMed, EMBASE, and MEDLINE for RCTs comparing outcomes in TKA based on tourniquet use. Fragility index (FI) and reverse fragility index (reverse FI) were calculated - for significant and nonsignificant outcomes, respectively - as the number of outcome reversals required to change statistical significance. The fragility quotient (FQ) was calculated by dividing the FI or reverse FI by the sample size. Median overall FI and FQ were calculated for all included outcomes, and sub-analyses were performed by reported significance. The literature search yielded 23 studies reporting 91 total dichotomous outcomes. RESULTS: Overall median FI was 4 with an interquartile range (IQR) of 3 to 6. Overall median FQ was 0.0476 (IQR 0.0291 to 0.0867). A total of 11 outcomes were statistically significant with a median FI and FQ of 2 (IQR 1.5 to 5) and 0.0200 (IQR 0.0148 to 0.0484), respectively. There were 80 outcomes that were nonsignificant with a median reverse FI of 4 (IQR 3 to 6). Loss to follow-up was greater than the median FI in 17.6% of outcomes. CONCLUSION: Altering a small number of outcomes is often sufficient to reverse findings in RCTs evaluating tourniquet use in TKA. We recommend including fragility analyses to increase reliability in the interpretation of study conclusions.