Pain levels and patient comfort after lower limb arthroplasty comparing i.v. patient-controlled analgesia, continuous peripheral nerve block and neuraxial analgesia: a retrospective cohort analysis of clinical routine data

Abstract

Background: Insufficient pain control after lower limb arthroplasty results in delayed recovery and increased risk for pain chronicization. The ideal kind of analgesia is still discussed controversially. We conducted a retrospective analysis of single-center routine data from a German university hospital, including patients receiving either total hip (THA) or knee arthroplasty (TKA).

Methods: All patients received general anesthesia. Patients undergoing THA received either continuous epidural ropivacaine infusion (0.133%, Epi) or patient-controlled analgesia (PCA) with the Wurzburg Pain Drip (tramadol, metamizole and droperidol, WPD) or with piritramide (Pir). After TKA, patients received either continuous femoral nerve block (ropivacaine 0.2%, PNB) or Pir.

Results: The analyzed cohort comprised 769 cases. Use of WPD after THA (n = 333) resulted in significantly reduced Numeric Rating Scale (NRS) values at rest, compared to Epi (n = 48) and Pir (n = 72) (.75 [IQR 1.14] vs. 1.17 [1.5], p = .02 vs. 1.47 [1.33], p < .0001) as well as maximum NRS scores (2.4 [1.7] vs. 3.29 [1.94], p < .001 vs. 3.32 [1.76], p < .0001). Positive feedback during follow-up visits was significantly increased in patients with a WPD PCA (p < .0001), while negative feedback (senso-motoric weakness/technical problems/nausea/dizziness/constipation) was particularly increased in Epi patients and lowest in those with WPD (p < .0001). After TKA, Pir (n = 131) resulted in significantly reduced NRS values at rest, compared to PNB (n = 185) (1.4 [1.4] vs. 1.6 [1.68], p = .02). Positive feedback was increased in patients with a Pir PCA in comparison with PNB (p = .04), while negative feedback was increased in PNB patients (p = .04). Overall, WPD presented with the lowest rate of any complications (8.7%), followed by Pir (20.2%), PNB (27.6%) and Epi (31.3%) (p < .001).

Conclusions: In the assessed population, the use of a WPD PCA after THA offered better pain control and patient comfort in comparison with continuous epidural or piritramide-based analgesia. After TKA, the use of a Pir PCA provided superior analgesia and a lower complication rate compared to continuous PNB.

Keywords: Hip arthroplasty; Knee arthroplasty; Patient-controlled analgesia; Peripheral nerve block.

Fig. 1

Retrospective study design and patient flowchart. APS: acute pain service, Epi: epidural analgesia, PCA: i.v. patient-controlled analgesia, PNB: continuous peripheral nerve block

Fig. 2

Postoperative analgesia following total hip arthroplasty. Protocols from the in-house acute pain service from 2016 and 2018 were retrospectively evaluated. Data from patients after total hip arthroplasty with either epidural (Epi, n = 48) or i.v. patient-controlled analgesia with the Wurzburg Pain Drip (WPD, n = 333) or piritramide (Pir, n = 72), respectively, are given. The figure shows median average Numeric Rating Scale (NRS) values at rest and maximum NRS values (A, upper panels), median number of PCA bolus requests and the median ratio between requested and administered boluses (A, lower panels). In B, positive and negative feedback during regular follow-up visits (expressed as median relative frequency, upper panels) and median duration and frequency of patient visitations (lower panels) are shown. Data are visualized as violin diagrams with median and interquartile range (25–75), indicated by the dashed lines. Kruskal–Wallis test or Mann–Whitney test was used for comparison. *p < .05, ***p < .005

Fig. 3

Postoperative analgesia following total knee arthroplasty. Protocols from the in-house acute pain service from 2016 and 2018 were retrospectively evaluated. Data from patients after total knee arthroplasty with either continuous peripheral nerve block (PNB, n = 185) or i.v. patient-controlled analgesia with piritramide (Pir, n = 131), respectively, are given. The figure shows median average Numeric Rating Scale (NRS) values at rest and maximum NRS values (A, upper panels), median number of PCA bolus requests and the median ratio between requested and administered boluses (A, lower panels). In B, positive and negative feedback during regular follow-up visits (expressed as median relative frequency, upper panels) and median duration and frequency of patient visitations (lower panels) are shown. Data are visualized as violin diagrams with median and interquartile range (25–75), indicated by the dashed lines. Mann–Whitney test was used for comparison. *p < .05, ***p < .005, ns: not significant