Intrathecal versus local infiltration analgesia for pain control in total joint arthroplasty

Background The purpose of this meta-analysis was to assess the efficacy of intrathecal morphine (ITM) analgesia and local infiltration analgesia (LIA) for pain control in total joint arthroplasty (TJA). Methods Embase, PubMed, the Cochrane Library, and Web of Science were systematically searched for randomized controlled trials (RCTs). All RCTs were comparing intrathecal analgesia and local infiltration analgesia in TJA. Primary outcomes were the visual analog scale (VAS) score with rest or mobilization up to 72 h. Secondary outcomes were the total morphine consumption, length of hospital stay, and morphine-related complications. Results Compared with the intrathecal analgesia group, the LIA group was associated with a reduction in VAS score with rest up to 72 h. Moreover, LIA was associated with a decrease in VAS score with mobilization at 6 h, 12 h, 48 h, and 72 h. Moreover, LIA significantly reduced total morphine consumption (weighted mean difference (WMD)  = − 15.37, 95% CI − 22.64 to − 8.83, P  = 0.000), length of hospital stay (WMD  =  − 1.39, 95% CI − 1.67 to − 1.11, P  = 0.000), and morphine-related complications (nausea and pruritus). Conclusions Local infiltration provided superior analgesia and morphine-sparing effects within the first 72 h compared with ITM following TJA.


Introduction
Total joint arthroplasty (TJA) mainly includes total knee arthroplasty (TKA) and total hip arthroplasty (THA). TJA is considered an effective surgical method for the treatment of end-stage osteoarthritis (OA) [1][2][3]. However, TJA is associated with severe pain after surgery. Effective pain control is crucial for early ambulation and good functional outcomes [4]. Early ambulation can lead to accelerated rehabilitation and a shortened length of hospital stay, which are the essential elements of a fasttrack recovery program. Moreover, effective pain control after TJA could not only increase patient satisfaction but also decrease the economic costs caused by the length of hospital stay [5,6].
Several methods (femoral nerve block [7], intrathecal morphine (ITM) [8], local infiltration analgesia [9], nonsteroidal anti-inflammatory drugs, oral opiates, and gabapentinoids [10]) have been applied in clinical practice for pain control in hip and knee surgeries. Femoral nerve blocks may increase the occurrence of falls and thus have limited clinical use [11]. Oral opiates and gabapentinoids may be associated with complications such as nausea, vomiting, and somnolence [12]. Local infiltration analgesia (LIA) and intrathecal analgesia are two common analgesia methods for pain control in TJA patients. However, it is still inconclusive as to which is preferable for pain relief in TJA. McCarthy et al. [13] concluded that LIA conferred superior analgesia compared with intrathecal morphine at 24 and 48 h following TKA. While Rikalainen-Salmi et al. reported that LIA might only enable early mobilization after THA, it was not associated with less nausea than ITM.
Therefore, we performed a meta-analysis of randomized controlled trials (RCTs) to evaluate the efficiency and safety of LIA and ITM for pain control in TJA.

Methods
This meta-analysis was based on the recommendations of the Cochrane Handbook for Systematic Reviews of Interventions and was written in accordance with the PRISMA checklist (Preferred Reporting Items for Systematic Reviews and Meta-Analyses).

Literature search
The following electronic databases were independently and extensively searched by two investigators from their inception through April 2019: Embase, PubMed, the Cochrane Library, and Web of Science. The search keywords were centered on the terms "local infiltration analgesia," "intrathecal analgesia," "total knee arthroplasty," and "total hip arthroplasty," which were adjusted to each database as necessary. In addition, the bibliographies of the included studies and dissertations were searched for additional publications. The search language was restricted to English. As all analyses were on previously published studies, ethical approval was not necessary.

Inclusion and exclusion criteria
The PRISMA guidelines were followed for the inclusion of studies in the meta-analysis. The detailed description of the inclusion criteria is as follows: (1) trials had to be properly randomized; (2) no additional agents or interventions confounded the comparison; (3) the patients in the trials were given a bolus dose via local injection; and (4) with respect to trials with several intervention groups, the eligibility of each individual group was evaluated, and only those qualified were included. Early studies published as a series of articles from the same institution or author that contained significant overlapping data were excluded for fear of multiple publication bias. Additionally, case reports, editorials, experimental studies, conference articles, commentaries, and other studies that failed to provide detailed results were excluded.

Data collection
After duplicates were removed and the study selection process was completed, the titles and abstracts were  We attempted to contact the study authors for supplementary information when there were insufficient or missing data in the articles.

Quality assessment
Two reviewers independently assessed the risk of bias in the RCTs using the Cochrane Collaboration's tool [14].

Characteristics of the trials
The general characteristics of the included studies are shown in Table 1. The publication years ranged from 2007 to 2019. The sample size of the included studies ranged from 15 to 61. The mean age of the included patients ranged from 61 to 71. Three studies compared intrathecal analgesia with LIA in THA patients, and the remaining studies were in TKA patients. All of the included studies were RCTs. All of the included studies performed spinal anesthesia during surgery.

Risk of bias assessment
The risk of bias summary and risk of bias graph can be seen in Figs. 2 and 3, respectively. Three studies had an unclear risk of bias for random sequence generation, and five studies had an unclear risk of bias for allocation concealment. Six studies had an unclear risk of bias for blinding of participants and personnel. There was a low risk of bias other bias.
A random-effects model was used because significant heterogeneity was found among the studies (I 2 = 98.8%, P = 0.000). The pooled results demonstrated that the LIA group was associated with a reduction in the VAS score with rest at 6 h compared with the ITM group (WMD =

Total morphine consumption
A total of 7 studies with 439 participants reported the total morphine consumption. Compared with the ITM group, the LIA group was associated with a reduction in total morphine consumption (WMD = − 15.37, 95% CI − 22.64 to − 8.83, P = 0.000; Fig. 6).

Length of hospital stay
A total of 7 studies with 332 participants reported the length of hospital stay. Compared with the ITM group, Fig. 6 Forest plots of the included studies comparing the total morphine consumption the LIA group was associated with a reduction in the length of hospital stay (WMD = − 1.39, 95% CI − 1.67 to − 1.11, P = 0.000; Fig. 7).

Occurrence of nausea
A total of 8 studies with 469 participants reported the occurrence of nausea. Compared with the ITM group, the LIA group was associated with a reduction in the occurrence of nausea (RR = 0.45, 95% CI 0.33 to 0.60, P = 0.000; Fig. 8).

Occurrence of pruritus
A total of 7 RCTs with 530 participants reported the occurrence of pruritus. Compared with the ITM group, the LIA group was associated with a reduction in the occurrence of pruritus (RR = 0.30, 95% CI 0.19 to 0.47, P = 0.000; Fig. 9).

Occurrence of respiratory depression
A total of 7 RCTs with 460 participants reported the occurrence of respiratory depression. There was no significant difference between the LIA and ITM groups in terms of the occurrence of respiratory depression (RR = 0.73, 95% CI 0.47 to 1.13, P = 0.162; Fig. 10).

Main findings
Our meta-analysis found that (1) compared with ITM, LIA significantly reduced pain scores with rest or mobilization at 6 h, 12 h, 24 h, 48 h, and 72 h and (2) LIA further reduced the total morphine consumption, length of hospital stay, and morphine-related complications.

Comparison with other meta-analyses
Only one relevant meta-analysis has been published [25]. Differences between our meta-analysis and the previous one should be noted. First, the previous meta-analysis included only four trials and 242 patients. In comparison, our current meta-analysis included 11 trials with 382 patients. With the added statistical power of at least 440 cases, our current meta-analysis is the latest and most comprehensive meta-analysis and generally concurs and further reinforces the results from previous meta-analyses. Second, we performed a subgroup analysis and sensitivity analysis. The results from these analyses further confirm our conclusions. Third, we also This meta-analysis demonstrated that LIA conferred better analgesia with rest up to 72 h. Moreover, LIA was associated with a reduction in morphine consumption compared with the ITM group. Yin et al. [26] conducted a meta-analysis and found that LIA can be used for patients undergoing THA because of its ability to reduce pain scores and analgesic consumption without any additional adverse events.
The results found that LIA could significantly reduce pain scores with mobilization up to 72 h. Pain with mobilization is more important than pain with rest. LIA is not only effective for pain control but also facilitates patients mobilizing early and returning to normal physiological functions quickly. Jia et al. [25] conducted a metaanalysis of LIA and ITM for total knee and hip arthroplasty. In this meta-analysis, LIA was only superior to ITM for pain control within the first 24 h. A major shortcoming of their meta-analysis was that they only included 4 RCTs with hip and knee surgeries. Lalmand et al. [27] revealed that LIA and ITM have equivalent analgesic effects in elective cesarean delivery. Karlsen et al. [28] conducted a review and found that the pain control efficacy of LIA is equivalent to that of other protocols.
Currently, opioids are commonly administered for pain control after TJA. However, morphine was associated with many insupportable complications, such as nausea and vomiting. Thus, total morphine consumption was also measured as the degree of pain control. We found that LIA was associated with a reduction in morphine consumption compared with ITM. These results were in accordance with the reduction in pain.
We measured morphine-related complications (nausea, pruritus, and respiratory depression) between the LIA and ITM groups. The results found that LIA was associated with a reduction in the occurrence of nausea and pruritus. However, there was no significant difference between the LIA and ITM groups in terms of the occurrence of respiratory depression. Kuchálik et al. [29] revealed that LIA was a safe technique for THA during the long-term follow-up (2 years).
There were a total of 5 limitations in the current meta-analysis. (1) Economic costs and functional outcomes for the LIA and ITM groups were not compared due to insufficient data, and future studies should focus on the economic costs and functional outcomes of LIA and ITM. (2) We included TKA and THA patients, and thus, there was high heterogeneity between the included groups. (3) The dose of anesthetics was different in the included studies, and more studies should be focused on the optimal dose and anesthetic drugs for anesthesia. (4) Long-term

Conclusion
Local infiltration provided superior analgesia and morphinesparing effects within the first 72 h compared with ITM following TJA. There were fewer adverse effects in the local infiltration anesthesia groups. However, it should be noted that these conclusions are based on a limited number of studies and patients. Future studies can focus on the economic costs, functional outcomes, and incidence of adverse events to provide more comprehensive results.