Outcome of Spinal Versus General Anesthesia in Revision Total Hip Arthroplasty

A Propensity Score-Matched Cohort Analysis

Venkatsaiakhil Tirumala, MS; Georges Bounajem, MD; Christian Klemt, PhD; Stephen P. Maier, MD; Anand Padmanabha, MD; Young-Min Kwon, MD, PhD


J Am Acad Orthop Surg. 2021;29(13):e656-e666. 

In This Article


Matched Cohorts

We used propensity scores to match 548 total patients in two groups in a 1:1 ratio (274 in each group). Baseline demographic characteristics of the propensity-score-matched cohorts are in summarized in Table 1. The matched cohorts are statistically similar in age, gender, BMI, laterality, follow-up time, ASA, CCI, race, comorbidities, and preoperative inflammatory markers (all P > 0.05). Furthermore, all SMDs are below 0.25, indicating overall balance between the covariates in the two cohorts. The most prevalent race was Caucasian (spinal: 92.7%, general: 91.61%, P = 0.631), whereas the most prevalent comorbidity was vascular disease (spinal: 63.1%, general: 67.2%, P = 0.370). Procedure details are summarized in Table 2 The etiologies for revision THA were statistically similar for both matched cohorts, with the SMDs all being below 0.25, indicative of balanced cohorts and successful propensity score matching. Aseptic loosening was the most common etiology for revision THA (spinal: 32.1%, general: 34.0%, P = 0.687), followed by adverse local tissue reaction (spinal: 18.5%, general: 13.2%, P = 0.122). Most THA revisions in both cohorts included the revision of at least one nonmodular implant (spinal: 70.2%, general: 67.9%, P = 0.639), and most THAs were performed noncemented (spinal: 85.6%, general: 89.81%, P = 0.185). To control for intraoperative blood loss, intravenous tranexamic acid was administered in 72.8% and 72.1% of patients in the spinal and general anesthesia cohorts, respectively (P = 0.926). The number of different orthopaedic surgeons and the prevalence of their revision THAs in both groups were also matched and determined to be statistically similar.

Perioperative Outcomes

Patients with general anesthesia had a higher mean surgical time (spinal: 161.3, general: 174.8 min, P < 0.001), which was defined as the period between the first skin incision and skin closure. This factor was controlled for in subsequent multivariate regression analyses. With surgical time controlled, patients in the general anesthesia cohort had significantly higher intraoperative (spinal: 305.5, general: 402.6 mL, P < 0.001) and total perioperative blood loss (spinal: 1,684.2, general: 1808.2 mL, P < 0.001) and higher odds for requiring transfusions in the perioperative period (OR, 2.05, P = 0.030) (Table 3).

Although the various types of inhospital perioperative complications did not differ significantly between groups, we found that patients administered general anesthesia were more likely to experience two or more complications during their stay (spinal: 1.51%, general: 6.42%, OR, 4.51, P = 0.007). The most common inhospital complications in both groups were pulmonary embolism (spinal: 1.13%, general: 1.51%,P = 0.710), wound drainage (spinal: 3.02%, general: 3.40%,P = 0.803), urinary tract infection (spinal: 2.64%, general: 3.02%,P = 0.799), acute renal failure (spinal: 1.89%, general: 3.40%,P = 0.287), and acute myocardial infarction (spinal: 0.75%, general: 1.13%,P = 0.673).

Patients administered general anesthesia also had a longer LOS (spinal: 3.52, general: 4.27 days, P = 0.001) and were at higher odds for having an extended LOS (spinal: 3.77%, general: 8.68%, OR, 2.45, P = 0.021). In addition, general anesthesia was associated with higher rates of discharge to inpatient rehabilitation than spinal anesthesia (spinal: 3.77%, general: 10.94%, P = 0.003), representing an OR of 3.14.

Postdischarge Complications

Readmissions and common complications that occurred within 90 days postdischarge are summarized in Table 4. No significant differences were found for 30-, 60-, and 90-day readmission rates between the spinal and general anesthesia cohorts. The most prevalent minor complications were the development of urinary tract infection (spinal: 1.51%, general: 1.13%, OR, 0.75, P = 0.722) and superficial surgical site infections (spinal: 1.13%, general: 1.51%, OR, 1.34, P = 0.715). The most prevalent major complication was revision surgery (spinal: 4.91%, general: 5.28%, OR, 1.09, P = 0.847). Ninety-day mortality in the overall cohort was low (<1%) and was not associated with either group. The aggregate total minor (spinal: 4.52%, general: 4.90%, OR, 1.09, P = 0.843) and major complication rates (spinal: 16.23%, general: 14.72%, OR, 0.89, P = 0.654) were not statistically different between the matched cohorts, which held for all individual complications as well.


Re-revisions for aseptic modes of failure were similar between the spinal and general anesthesia cohorts (spinal: 11.70%, general: 15.09%, P = 0.242). The most common etiologies for aseptic re-revision were aseptic loosening (spinal: 4.91%, general: 6.42%, P = 0.461), followed by instability (spinal: 3.02%, general: 3.40%, P = 0.803). Patients administered general anesthesia were found to be at higher odds for developing periprosthetic joint infection (spinal: 3.77%, general: 7.92%, OR, 2.20, P = 0.045); however, no statistical differences were observed in the incidence rate of acute and chronic infections between both cohorts (Table 5).