369 posters,  63 sessions,  7 topics,  1978 authors 
ePostersLive® by SciGen® Technologies S.A. All rights reserved.

The effect of anesthetic technique on mortality and major morbidity following hip fracture fixation. A retrospective, propensity-score matched-pairs cohort study
Session: MP-05b
Fri, April 20, 8-9:30 am
Uris (Shubert Complex), 6th floor

Please note, medically challenging cases are removed three months after the meeting and scientific abstracts after three years.

Poster Presenter


No votes yet

The Effect of Anesthetic Technique on Mortality and Major Morbidity Following Hip Fracture Fixation.  A retrospective, propensity-score matched-pairs cohort study     Dr. Sarah Tierney, FRCPC; Dr. Laith Malhas, FRCA; Dr. Vincent Chan, FRCPC; Dr. Scott Beattie, FRCPC; Dr. Anahi Perlas, FRCPC


•With an aging population, hip fracture incidence continues to rise
•Mortality post hip fracture is 5-11% at 30-days, and 30% annually1,2
•This is unchanged since the mid 1990s, despite healthcare advances
•Spinal vs general anesthesia (GA) has shown physiologic benefits3
•Greater hemodynamic stability
•Lower incidence of respiratory events
•Less cognitive dysfunction in the immediate perioperative period 
•Improved perioperative analgesia
•Decreased blood loss
•These may be of particular importance, given the overall frailty and limited physiologic reserve of hip fracture patients
•Despite the improved physiologic benefits, no previous study has shown a mortality benefit to spinal anesthesia
•A matched-pairs cohort study of all patients who underwent hip fracture fixation at the Toronto Western Hospital, from 2003-2015, inclusive. Approval was obtained from the IRB.
•Patients were identified via electronic databases
•Bivariate tests were used to compare the unmatched cohorts for 29 variables known to affect mortality and morbidity associated with hip fractures
•These included age, gender, ASA, year of surgical procedure, significant pre-operative co-morbidities, baseline laboratory values, pre-operative cardiac medications, RCRI and Charlson’s co-morbidity index variables
•A propensity-score matched pairs analysis was then conducted using the results of the tests on the unmatched cohorts, to minimize the treatment selection bias on study outcomes
•An iterative process was undertaken to build a multi-variate logistic regression model to estimate an individual propensity score for spinal anesthesia
•The propensity-score for spinal anesthesia was then used to match patients who received spinal anesthesia with those who received general anesthesia, resulting in matched pairs of discrete individuals
•Primary outcome: Impact of spinal anesthesia on 90 day mortality
Secondary outcomes: 30 and 60-day mortality, PE, hospital length-of-stay, major blood loss and MACE (cardiac arrest and myocardial infarction)
•2591 patients underwent hip fracture fixation from 2003-2015, 43.8% had spinal anesthesia
•883 matched-pairs were identified (78% of the total spinal anesthesia cohort)
•In the unmatched cohort, patients who received spinal anesthesia were:
•More likely to be male, elderly, have a higher ASA score, and receive earlier hip fracture fixation
•Less likely to have metastatic cancer, and take ASA and CCBs
•Overall, 90 day mortality was significantly lower in patients who received spinal anesthesia vs GA patients (6.3% vs 8.8%)
•Absolute Reduction Risk = 2.3%
•Relative Risk Reduction = 26.1%
•Number Needed to Treat = 44 patients
•No significant difference was found for mortality at 30 or 60-days post-operatively
•Incidence of death due to cardiac events was higher in GA (30.3%) vs Spinal (19.7%)
Death due to respiratory events (32.1 vs 26.3%) and sepsis (14.3 vs 5.3%) was higher in Spinal vs GA
•A strong association was identified between spinal anesthesia and lower 90-day mortality post hip fracture repair
•Spinal anesthesia was also associated with decreased rates of pulmonary embolism, major blood loss and a shorter hospital length-of-stay
•Causes of death were varied. Overall, patients who received spinal anesthesia were less likely to die from a cardiac event, but more likely to die from a respiratory cause or sepsis
•However, this study was not powered for cause of death
•Retrospective, Single center
•Patients were excluded from the final analysis if a propensity score match pair was not identified, which may limit the generalizability of results

However, matching did allow for greater bias elimination

Enter Poster ID (e.gGoNextPreviousCurrent