Effects of Acupuncture-Point Stimulation on Perioperative Sleep Disorders: A Systematic Review with Meta-Analysis and Trial Sequential Analysis

Background Perioperative sleep disorders exert a severe adverse impact on postoperative recovery. Recently, some observational studies reported that acupuncture-point stimulation (APS) provided benefits for promoting perioperative sleep quality. However, the effects of APS on perioperative sleep disorders following general anesthesia have not been thoroughly assessed by any systematic study and meta-analysis. Therefore, we conducted this systematic review and meta-analysis to reveal the effects of APS on perioperative sleep disorders. Methods Eight databases (Chinese: CNKI, VIP, CBM, and Wanfang; English: PubMed, Embase, Web of Science, and Cochrane Library) were thoroughly searched to find randomized controlled trials (RCTs) that indicated a link between APS and the occurrence of perioperative sleep disorders. We applied RevMan 5.4 (Cochrane Collaboration) and Stata 16.0 (Stata Corp) to conduct our meta-analysis. In addition, the trial sequential analysis (TSA) tool was utilized to estimate the validity and reliability of the data. Results In this study, nine RCTs with 719 patients were conducted. Compared to the control group, APS significantly improved perioperative subjective sleep quality (SMD: −1.36; 95% CI: −1.71 to −1.01; P < 0.00001). Besides, it increased perioperative TST (preoperative period MD = 24.29, 95% CI: 6.4 to 42.18, P = 0.0008; postoperative period MD = 45.86, 95% CI: 30.00 to 61.71, P < 0.00001) and SE (preoperative MD = 3.62, 95% CI: 2.84 to 4.39, P < 0.00001; postoperative MD = 6.43, 95% CI: 0.95 to 11.73, P < 0.00001). The consequence of trial sequential analysis further confirmed the reliability of our meta-analysis results. Conclusion According to the currently available evidence, APS could effectively improve perioperative sleep quality and play an essential role in decreasing the incidence of perioperative sleep disorders.


Introduction
Sleep is vital for overall health as it infuences the quality of life and physical functioning [1].Good sleep quality is a sign of wellbeing, whereas poor sleep quality raises comorbidity, mortality, and medical expenses [2].Despite advances in operative and anesthetic procedures, sleep disorders continue to be a concern in the perioperative period.Perioperative sleep problems, as compared to household sleep disorders, have gained more attention recently, such as reduced nighttime sleep, fragmentation of sleep, and circadian rhythm disruptions at night after admission [3,C4].Te proportion of sleep disorders in hospitalized patients (64%-73.1%) is much higher than that in people at home (30.6%-41.2%)due to the surgical stress response, ward environment, and patients' underlying diseases [5][6][7].Perioperative sleep problems were noted in several observational studies to be important risk factors for poor recovery [8] and were associated with anxiety, altered pain perception, and postoperative cognitive dysfunction (such as delirium), further deteriorating the patient's physical state [1,6,7,9].Moreover, it was suggested that perioperative sleep disorders might lead to postoperative fatigue, episodic hypoxemia, cardiovascular disorders, metabolic impairment, and immune disorders [10,11].Enhancing perioperative sleep quality is probably related to improved health, surgery efect, and patient safety [9].Unfortunately, there are no specifc therapeutic strategies for perioperative sleep disorders.Even though it was reported that pharmaceutical therapies, such as benzodiazepines and other sedating agents, might improve perioperative sleep quality, the potentially addictive nature and risk of tolerance restricted their clinical application [9].
In complementary and alternative medicine, acupuncturepoint stimulation (APS) was employed to improve perioperative sleep quality [12].According to the theory of traditional Chinese medicine (TCM), APS refers to stimulating the body's acupoints with various methods, including acupressure (manual stimulation), acupuncture (sterile needle stimulation), laser acupuncture, electrical stimulation (transcutaneous electrical acupoint stimulation and electrical acupuncture), magnetic acupuncture (stimulation with special equipment), moxibustion therapy, and auricular point pressing, with the goal of achieving therapeutic efects such as sleep improvement and pain management [13][14][15][16][17][18].APS has been an integral part of traditional Chinese medicine treatment.Based on the traditional Chinese medicine meridian theory, the body's qi circulation is restored when target acupoints along meridians are stimulated [14].It was reported that APS might lower the risk of poor sleep quality due to its considerable analgesic and sedative efects [19,20].Recently, a signifcant number of RCTs were carried out on the efects of APS on perioperative sleep disorders [8,15].Besides, a previous review concluded that EA might work to alter neurotransmitter concentrations and reduce the levels of norepinephrine and dopamine so as to enhance the quality of postoperative sleep quality after general anesthesia [12].However, this study not only did not conduct a statistical assessment of the efciency of EA on inpatients' sleep quality during the postoperative period but also did not examine preoperative sleep quality thoroughly [12].
Consequently, we carried out a quantitative metaanalysis to evaluate the efects of APS on perioperative sleep disorders, including the preoperative period and the postoperative period, to provide research-based evidence for clinical practice.

Materials and Methods
Te PRISMA statement was implemented in the conception and presentation of our systematic review and meta-analysis, which was carried out to estimate the efciency of APS in improving perioperative sleep quality [21].Te metaanalysis was registered in the PROSPERO database (registration number: CRD42023387848).

Study Selection.
EndNote X9 was utilized to remove duplicate articles and organize the references.Two writers independently assessed the abstracts and titles for initial evaluation after removing duplicates.If one study was determined to be qualifed, we would download the whole text and carry out detailed screening.When the two authors came to an understanding, data extraction and analysis were performed on the qualifed research.A third reviewer solved disagreements during selection.

Data Extraction. Essential characteristics of the included
RCTs were independently extracted by diferent researchers using a designed data extraction framework.In the special data extraction table, primary writer, publication date, sample size, study design, surgery type, intervention, and reported outcome type (sleep quality assessment methods) were collected.If any material in the report needed to be clarifed, an attempt was made to consult the original study's author.Te collected data were considered for a systematic review and meta-analysis to assess the efects of APS on perioperative sleep disorders.

Assessment of the Risk of Bias.
Te Cochrane risk of bias assessment was used to evaluate the RCTs' methodological quality in seven distinct aspects, including (I) random sequence generation, (II) allocation concealment, (III) blinding of participants and personnel, (IV) blinding of outcome assessment, (V) incomplete outcome data, (VI) selective reporting, and (VII) other bias [22].In this method, risk levels were classifed as "high risk," "uncertain risk," and "low risk."Signifcantly, the APS operator could not be blindfolded due to the unique nature of APS therapy and only participants and result measurement personnel might be.Two researchers separately conducted the assessment; in the event of a dispute, a third evaluator was asked to mediate.International Journal of Clinical Practice 2.6.Data Synthesis and Analysis.Our meta-analysis was synthesized to quantitatively summarize the qualifed studies using ReviewManager software version 5.4.Te efciency of APS on perioperative sleep disorders was estimated by the following continuous outcomes reported as mean diferences (MDs) with 95% CIs: PSQI, AIS, or ISI, and TST and SE.Generally, MD with 95% confdence intervals (CIs) was deemed to be suitable for continuous outcomes.However, the standardized mean diference (SMD) should be computed when the same continuous result was examined using multiple assessment tools.Results were considered signifcant for P values under 0.05.Te chi 2 test and I 2 statistic were employed to assess study heterogeneity.If statistical homogeneity (P > 0.10; I 2 < 50%) was observed in each trial, the fxed-efect model was employed for data analysis.If there was statistical heterogeneity between trials (P < 0.10; I 2 > 50%), the source of heterogeneity was examined using a random-efects model in conjunction with a subgroup or sensitivity analysis.Sensitivity analyses were carried out for the included studies to determine which RCT impacted the overall results.Descriptive analysis was used if heterogeneity was too considerable for a meta-analysis.In addition, we performed the trial sequential analysis with TSA program version 0.9.5.10 beta to ascertain the statistical reliability of the data.

Literature Search.
A total of 419 studies were obtained from 8 databases: CNKI, Wanfang, VIP, CBM, Embase, PubMed, Cochrane Library, and Web of Science.128 duplicate studies were identifed and excluded using EndNote X9 software.After the titles and abstracts were screened, 234 irrelevant publications were eliminated and 28 full-text papers were evaluated to determine which study fulflled the qualifying requirements.Due to inconsistencies with the inclusion criteria, an additional 19 studies were removed after careful reading of the whole text.Ultimately, for this research, 9 trials [8,[15][16][17][18][23][24][25][26] were included.Figure 1 shows a fow diagram that summarizes the selected results. 1 presents the baseline data from the nine studies included in our research.Te publications selected were published from 2017 to 2022, with seven completed in the last 3 years.Tere were 359 patients in the APS cohorts and 360 patients in the control cohorts (non-APS and usual care).Te type of surgery included breast conserving surgery [15], elective video-assisted thoracoscopic surgery [8], thyroid surgery [16,24], lung cancer surgery [17], living kidney transplantation [23], spinal surgery [25], radical surgery for esophageal carcinoma [26], and cesarean section [18].
Of all the included studies, three assessed the outcome with a subjective scale PSQI only [16,17,24], one used two types of subjective scale to evaluate sleep quality, including AIS and PSQI [26], and one used the subjective scale ISI [23].Tree studies reported the efcacy of APS on SE [8,15,25], and three reported the efect on TST [15,18,25]. 2 and 3 illustrate the risk of bias for every qualifed research, as well as the overall risk of bias for all chosen trials.Te random number table approach was employed in eight studies, which were regarded as a low selection risk [8,[15][16][17][23][24][25][26].Te randomization strategy of the fnal one study [18] was not disclosed in detail; hence, it was considered to have an unknown selection risk.Two studies [8,15] assessed the low selection risk of selection bias; the rest seven studies [16][17][18][23][24][25][26] showed unclear risk for the lack of information on the selection bias.Most of the qualifed studies had a high risk of bias among the performance bias because of the specifcity of APS therapy.For detection bias, eight studies [8, 15-18, 23, 24, 26] showed low risk, and the remaining one [25] apprised unclear risk due to lack of relevant details.   5(a).Furthermore, a subgroup analysis was conducted based on the diferent Control groups, which included usual care and non-APS groups.When we excluded the results of the non-APS groups, no statistical heterogeneity was found in our data analysis (SMD = −1.79,95% CI: −2.09 to −1.48; P � 0.73, I 2 = 0%; 231 participants, Figure 5(b)).Te credibility of the synthesized data was further verifed by TSA with the Z-curve transcending not only the conventional boundary but also the trial sequential monitoring boundary, see Figure 6.

Efects of APS Terapy on Perioperative TST.
Tree studies [15,18,25] assessed the efects of APS therapy on perioperative TST, with one study [15] evaluating the TST during the preoperative period and two, [18,25] during the postoperative period.Te results suggested that, compared to the control group, TST was higher in the APS group during the preoperative period (MD � 24.29; 95% CI: 6.4 to 42.18; P � 0.0008) and the postoperative period (MD � 45.86; 95% CI: 30.00 to 61.71; P < 0.00001), see Figure 7.

Sensitivity Analysis.
A sensitivity analysis was performed to examine the impact of a single study on the total efect size.No matter which article was left out, the combined total result of the primary outcome was unafected, proving that the main conclusion in our meta-analysis was robust, see Figure 9.

Discussion
It was frequently reported that patients experienced various kinds and severity levels of sleep problems during the perioperative period, which might last for a long time after anesthesia and surgery [27].Patients undergoing various procedures were vulnerable to perioperative sleep disorders.
Obstructive sleep apnea (OSA), decreased total sleep time, fragmented sleep, abnormal circadian rhythms, and other issues were all included in perioperative sleep disorders [3].
Several risk factors could lead to the incidence of perioperative sleep disorders, such as age [28], pain [29], mental diseases [30], and surgery [31].Interestingly, some prior studies suggested that perioperative sleep disorders could contribute to altered pain perception and postoperative cognitive dysfunction and lead to poor recovery [3,32].It was suggested that patients in the intensive care unit following thoracic surgery were observed to have postoperative sleep problems, including 62% of patients within 6 months and 12% of patients at all follow-up visits experiencing poor sleep [33].However, the prevalence of sleep disturbance was not taken seriously by medical staf for a long time.For example, almost 25% of patients undergoing cardiac surgery sufered from sleep disturbance, with 80% of these patients remaining unidentifed before the operations [34].6 International Journal of Clinical Practice daytime napping, and nocturnal waking following surgery [36].However, the traditional pharmacological therapies were often limited due to the apparent adverse reactions.APS, a method of traditional Chinese medicine, was used as a new nonpharmacological therapy for perioperative sleep disorders recently, which was easy to learn and safe to apply [37].A large number of RCTs examined the efciency of APS on perioperative sleep disorders, and the results indicated that APS could signifcantly improve perioperative sleep quality [8].A prior meta-analysis also concluded that APS at various acupoints might signifcantly increase the body's neurotransmitter levels and regulate biological clock genes to enhance patients' sleep quality during the postoperative period [37].Despite the fact that its therapeutic efcacy was relatively defnite, it was not examined quantitatively and systematically in prior studies.
We conducted the meta-analysis to analyze and confrm the results by gathering as many publications as feasible, summarizing features of the trials, and synthesizing their data.Tis study dealt with whether APS therapies could improve perioperative sleep disorders.We chose this topic, given that APS was frequently employed to gain better Total (95% CI) Heterogeneity: tau 2 = 0.16; chi 2 = 21.12,df = 6 (P = 0.002); I 2 = 72% Heterogeneity: tau 2 = 0.05; chi 2 = 1.91, df = 1 (P = 0.17   Test for subgroup differences: chi 2 = 3.13, df = 1 (P = 0.08); I 2 = 68.0% Test for overall effect: Z = 6.01 (P < 0.00001) Test for overall effect: Z = 2.66 (P = 0.008)     International Journal of Clinical Practice perioperative sleep quality in observational studies in recent years [8,15].We detected nine trials with remarkable variations in the types of APS and acupuncture points, types of surgery, study sample sizes, and research qualities.Given these biases, we carried out a comprehensive summary, accompanied by a quantitative evaluation with subgroup analysis.Te pooled results (Figure 4) suggested that the APS therapies could contribute to improve perioperative sleep quality.Ten, we conducted a subgroup analysis since diferent types of subjective scales (PSQI, AIS, and ISI) were used to estimate the efciency of APS on perioperative sleep quality (Figure 5(a)) and identifed that the diferences of subjective sleep quality scales did not afect our statistical results.Nonetheless, it was still necessary to unify the assessment criteria of sleep disturbance based on the internal discrepancy among the subjective scales.For instance, the PSQI is used to evaluate the last month's sleep time, subjective sleep quality, sleep habits, sleep efciency, sleep interruptions, usage of sleeping drugs, and daytime dysfunction [27].As another subjective scale assessing sleep quality, the AIS is a self-rating scale that assesses sleep induction, nighttime awakenings, ultimate awakening, overall sleep duration, and sleep quality [27].Te advantages of our meta-analysis were as follows: Tis was the frst meta-analysis to estimate the efects of APS on sleep quality in all surgical patients during the whole perioperative period.Prior studies had certain restrictions.Luo's study merely examined the efciency of electroacupuncture on sleep quality during the postoperative period and did not conduct a quantitative meta-analysis [12].In addition, no meta-analysis was conducted to evaluate the efects of APS on perioperative objective sleep-related indicators.Terefore, our meta-analysis combined and evaluated the qualifed studies that provided objective sleep-related indicators (TST and SE), which might more accurately and objectively estimate the efciency of APS on perioperative sleep disorders.Meanwhile, we identifed two possible causes of heterogeneity, which were diferences in the type of subjective scales employed and in their arrangements of the control groups.More importantly, we also performed trial sequential analysis to validate the results of our study, which made our results more valuable than the prior studies.
Our meta-analysis included some limitations that should also be taken into account.First, only articles published in Chinese and English were considered for this meta-analysis, while the studies in other languages were excluded.Second, several qualifed RCTs did not precisely notify allocation concealment and reporting bias, which might cause a mild deviation in the meta-analysis results.Tird, small sample sizes and the lack of follow-up evaluation in the majority of the randomized trials might cause the advantages of APS on perioperative sleep disorders to be overstated.Only six studies were included in the analysis of the main outcome, the subjective sleep quality, so further studies with a large sample size and high quality are still required.

Conclusion
Taken together, our systematic review and meta-analysis suggested that APS could efectively improve perioperative sleep disorders, indicating that it was worthwhile to ofer such therapies in perioperative healthcare institutions.Although our results had certain clinical implications of APS for the treatment of perioperative sleep disorders, more studies are needed to ofer stronger evidence to further verify our conclusions in the future.International Journal of Clinical Practice 3.4.Outcome 3.4.1.Meta-Analysis Findings.Te combined results of the qualifed RCTs examining the efciency of APS on sleep disorders among perioperative inpatients were presented in

Figure 4 :
Figure 4: Forest plot assessing the efect of APS on perioperative subjective sleep quality.

Figure 7 :
Figure 7: Forest plot assessing the efect of APS on perioperative TST.

Figure 8 :
Figure 8: Forest plot assessing the efect of APS on perioperative SE.

Figure 9 :
Figure 9: Sensitivity analysis for the efect of APS on perioperative subjective sleep quality.
Criteria.Nine qualifed studies were published in Chinese and English between 2017 and 2022.Studies that matched the following criteria (PICOS) were eligible for our meta-analysis:

Table 1 :
Characteristics of the included studies.