The present study was approved by the Institutional Review Board of Tokyo Women's Medical University (approval number 4416, July 11, 2017). The need for informed consent was waived by the Review Board due to the retrospective nature of the study. The protocol was registered at the UMIN Clinical Trials Registry (No: UMIN000038476).We retrospectively reviewed all medical records of patients with RA who underwent toe arthroplasty at our institute between January 2014 and May 2018.
The inclusion criteria were as follows: patients aged 20 years or older, patients with prior diagnosis of RA, patients who underwent elective toe arthroplasty, patients who were scheduled for their first postoperative day (POD) rehabilitation. We categorized patients into the two groups: possible group, those who were able to walk during rehabilitation on the first POD and impossible group, those who were unable to walk during rehabilitation on the first POD.
The primary outcome was odds ratio (OR) with a 95% confidence interval (CI) between various patient factors and the infeasibility of walking rehabilitation on the first POD. The secondary outcome measures were time until the first rescue analgesic requirement after the surgery and the incidence of PONV after each anesthesia method. The post hoc analysis for association between anesthesia methods and postoperative outcome was planned in advance.
Patients’ demographic data including age, sex, height, body weight, preoperative medication for RA were collected by reviewing the medical records. Intraoperative data such as duration of anesthesia and surgery, anesthesia methods, blood loss and fluid infusion, total amount of intraoperative fentanyl were collected from the electronic anesthesia record system Mirrel (FUKUDA DENSHI, Tokyo).
There was a wide variety of anesthesia modalities: GA or non-GA, type of regional anesthesia (spinal, epidural or combined spinal-epidural anesthesia), supplementation of peripheral nerve block (PNB) and continuous postoperative opioid.
The GA was managed using remifentanil, fentanyl, propofol, sevoflurane, and desflurane. The type and dose of the anesthesia medication depended on the decision of the anesthesiologists in charge. The continuous postoperative intravenous patient controlled analgesia (IVPCA) was managed by fentanyl (base 15 or 20 µg / hour, bolus of 15 or 20 µg, and lockout interval of 10 minutes). Epidural catheter was placed at the L2-L3, L3-L4, L4-L5, or L5/S1 intervertebral space. Patient-controlled epidural analgesia consisted of 0.25% levobupivacaine (base dose of 4 ml / hour, bolus of 4 ml, and lockout interval of 30 minutes). The method of single injection PNB was mainly a combination of popliteal sciatic nerve and femoral or subsartorial saphenous nerve block with 20–40 mL of 0.375-0.5% ropivacaine or 0.25% levobupivacaine using ultrasound-guided technique. Two patients managed by ankle block with 30 ml of 0.5% ropivacaine using ultrasound-guided technique. The combined spinal-epidural anesthesia was administered epidural anesthesia at the L2-L3, L3-L4, or L4-L5 intervertebral space and spinal anesthesia at the L3-L4 or the L4-L5 intervertebral space with 2.2-4 ml of 0.5% hyperbaric or plain bupivacaine. The spinal anesthesia was administered at the L3-L4 or the L4-L5 intervertebral space with 2.2-4 ml of 0.5% hyperbaric or plain bupivacaine.
Postoperative pain was managed by uniform clinical pathway: there was no regular administration of analgesics, and additional analgesics were administered by the nurse at the patient request. IVPCA with fentanyl and patient-controlled epidural analgesia bolus was not counted as the number of rescue analgesia.
Data are presented as mean (standard deviation), median [interquartile range], or number (%). Categorical variables were compared using the chi-squared test and numerical variables were compared using Student's t-test for the parametric data and Mann–Whitney U test for the non-parametric data.
The primary outcome was assessed using logistic regression analysis and the p value was calculated with the Benjamini-Hochberg method. We considered that patient factors such as physique, age, sex, daily use of corticosteroid, the number of biological disease modifying anti-rheumatic drugs ( DMARDs), surgical factors (duration of surgery and blood loss), anesthesia factors (duration of anesthesia, intraoperative fluid volume, anesthesia methods and intraoperative fentanyl dose), and postoperative factors (postoperative continuous opioid use, postoperative pain and PONV) hindered postoperative rehabilitation, and comprehensively examined them using logistic regression analysis.
Time to the first rescue analgesic requirement after the surgery was compared using the log-rank test and the Kaplan-Meier curve. The incidence of PONV was compared by analysis of variance. Data were analyzed using JMP® Pro 15.0.0 (SAS, Cary, NC, USA). A P value < 0.05 was considered statistically significant.