Article Text


Acupuncture for recovery after surgery in patients undergoing colorectal cancer resection: a systematic review and meta-analysis
  1. Kun Hyung Kim1,
  2. Dae Hun Kim2,
  3. Hee Young Kim3,
  4. Gyung Mo Son4
  1. 1School of Korean Medicine, Pusan National University, Yangsan, South Korea
  2. 2Department of Acupuncture & Moxibustion, Korean Medicine Hospital, Pusan National University, Yangsan, South Korea
  3. 3Department of Anesthesia and Pain Medicine, Pusan National University Yangsan Hospital, Yangsan, South Korea
  4. 4Department of Surgery, Pusan National University Yangsan Hospital, Yangsan, South Korea
  1. Correspondence to Professor Gyung Mo Son, Department of Surgery, Pusan National University Yangsan Hospital, Yangsan, 50612, South Korea; skm171{at}


Objective To assess the efficacy/effectiveness and safety of acupuncture in patients recovering from colorectal cancer resection.

Methods We systematically searched four English language databases (Medline, Embase, CENTRAL (Cochrane Central Register of Controlled Trials), and AMED (Allied and Complementary Medicine Database)) and one Chinese database (CAJ, China Academic Journals). Randomised trials of acupuncture compared with usual/routine care, sham interventions or active comparators in patients undergoing colorectal cancer resection were eligible for inclusion. Postoperative symptoms and quality of life (QoL) were the primary outcomes for the review.

Results Of 1225 screened hits, seven randomised trials with 540 participants were included. High or uncertain risk of bias and significant heterogeneity were observed. All outcomes were measured before discharge, and no trial explicitly reported post-discharge outcomes. The response to acupuncture in terms of postoperative symptoms was inconsistent across trials. QoL was not measured in the included studies. For certain outcomes reflecting physiological recovery, favourable effects of acupuncture were observed compared with sham acupuncture, namely time to first flatus (n=207, three studies; mean difference (MD) −7.48 h, 95% CI −14.58 to −0.39 h, I2=0%) and time to first defaecation (n=149, two studies; MD −18.04 h, 95% CI −31.90 to −4.19 h, I2=0%). Two studies reported there were no acupuncture-related adverse events, whereas the remaining studies did not consider adverse events.

Conclusions We found low-to moderate-quality evidence for the efficacy and safety of acupuncture for recovery after surgery in colorectal cancer patients. Future trials with adequate allocation concealment, blinding of outcome assessors, and measurement of post-discharge outcomes including QoL or functional recovery are warranted.

Trial registration number CRD42014015537.


Statistics from


Colorectal cancer is one of the most prevalent gastrointestinal neoplasms, and surgical resection is the standard treatment. After surgery, patients can experience various symptoms and diminished quality of life (QoL).1 Recovery after colorectal cancer resection is a complex process encompassing multiple dimensions, including biological and physiological variables, symptoms, physical, emotional, social and economic function, health perception, and overall QoL.2 To address these multiple aspects of the patient's perioperative trajectory, multidisciplinary, coordinated and standardised care plans that integrate evidence-based interventions have been emphasised.3 ,4 Although acupuncture is not typically included as a component of care,4 previous research suggests that acupuncture can reduce postoperative pain, nausea/vomiting and anaesthesia-related complications.5–8 Whether or not and, if so, how acupuncture contributes to the multi-dimensional aspects of recovery after colorectal cancer resection is an important issue to be addressed. To map current evidence and guide further clinical research, this review assessed the efficacy/effectiveness and safety of acupuncture during recovery from colorectal cancer resection.


The review protocol and the search strategy were registered in PROSPERO (International Prospective Register of Systematic Reviews) and are openly accessible (registration no. CRD42014015537).

Search strategy and eligibility criteria

Five databases were searched from their inception to June 2015: Medline, Embase, the Cochrane Central Register of Controlled Trials (CENTRAL), the Allied and Complementary Medicine Database (AMED), and the China Academic Journals (CAJ) database. We also searched the International Clinical Trials Registry Platform (ICTRP) of the WHO and for any ongoing studies. The search terms included terms related to acupuncture (eg, acupuncture, needling or meridian) and the condition (eg, colorectal cancer). We also searched the reference lists of relevant publications (eg, included clinical trials or registered trial information). Clinical experts were also contacted and asked to inform us of any potentially eligible studies. One review author screened the titles and abstracts of the identified hits for potentially eligible studies. Randomised controlled trials (RCTs) comparing needle acupuncture with usual/routine care, sham acupuncture or other active comparators were eligible without any language restrictions. Another review author supervised the selection process. Disagreements regarding the eligibility of a study were resolved by discussion between the two authors.

Outcomes of interest

Based on a recently published systematic review of the multidimensional aspects of recovery after surgery,2 we considered early postoperative symptoms, such as pain and postoperative nausea/vomiting (PONV), as well as post-discharge QoL, as primary outcomes. Secondary outcomes included physiological recovery, functional status, perceived general health, medication use, postoperative complications, and other adverse events.

Data collection

One review author extracted data using a pre-defined data collection form and another review author verified the data extraction process. Disagreements were resolved by discussion. Additional information was sought by contacting the authors of the included trials.

Risk of bias assessment

Two review authors assessed the risk of bias independently using the Cochrane risk of bias tool for the following seven domains: (1) random sequence generation; (2) concealment of allocation; (3) blinding of participants and personnel; (4) blinding of outcome assessors; (5) incomplete outcome data; (6) selective reporting; (7) and other sources of bias.9 This assessment was performed at the study level and not at the outcome level.

Statistical analyses

Where appropriate, we used RevMan V.5.3 software (Nordic Cochrane Centre, Copenhagen, Denmark) for meta-analysis of the data using the random effects model. Heterogeneity across studies was quantified using I2 scores.10 If there was considerable heterogeneity (I2 value ≥50%)10 among the studies, we qualitatively synthesised the data. Risk ratios (RR), mean difference (MD) or standardised mean difference (SMD) with 95% CIs were used to present the results of dichotomous and continuous outcomes, respectively. Trials with missing values for quantitative synthesis were excluded from meta-analysis. Although we planned to perform pre-specified sub-group and/or sensitivity analyses to investigate potential sources of heterogeneity, we were not able to do so due to the small number of trials. Similarly, the number of included studies was too small to be able to reliably perform a funnel plot to examine for potential publication bias.


Search results and characteristics of the included studies

Of 1225 initial hits, seven studies including 540 participants were deemed eligible (figure 1 and table 1).11–17 Registry information was identified for two additional trials, one of which was ongoing and one completed but unpublished (see online supplementary file 1). The studies were conducted in China,11 ,12 ,15–17 Hong Kong13 and the USA14 and were reported in English13 ,14 ,16 ,17 or Chinese.11 ,12 ,15 Sample sizes varied between 30 and 165 (median 70). The style of acupuncture stimulation included manual acupuncture (MA) alone,11 ,14 ,16 MA plus electroacupuncture (EA),12 ,13 ,17 and MA plus heat application (warming the needle using the moxibustion technique).15 The types of comparators included sham acupuncture,13 ,14 ,17 usual/routine care12 ,13 ,15 ,16 or a fast-track recovery programme11 without acupuncture (see online supplementary files 2 and 3).

Table 1

Summary characteristics of the included RCTs

Figure 1

Flow diagram of trial selection. AMED, Allied and Complementary Medicine Database; CAJ, China Academic Journals database; CENTRAL, Cochrane Central Register of Controlled Trials.

Types of outcomes

All outcomes were measured before hospital discharge. Five trials measured early postoperative symptoms including pain,11 ,13 ,14 ,16 nausea/vomiting,11 ,14 ,16 ,17 abdominal distension,11 ,16 and sleep disturbance.16 All trials assessed physiological recovery, although the definition varied among studies. The return of bowel function was measured based on the following aspects or their combination: passage of gas11–17; passage of stool11–13 ,15 ,17; return of bowel sounds17; tolerance of oral intake13 ,14; and nasogastric tube reinsertion rates.14 Postoperative complications, such as prolonged ileus, postoperative infection or acute gastric dilatation, were measured in three studies.12 ,13 ,16 Two studies measured functional status, including length of hospital stay13 ,17 and independent mobilisation.13 Other outcomes included patient-perceived well-being,16 use of analgesics,13 ,14 ,17 global recovery rate,11 and adverse events related to acupuncture.16

Risk of bias assessment

The random sequence generation was regarded as adequate in all included studies; however, allocation concealment was sometimes inadequate13 or unclear.11 ,12 ,15 ,16 Three sham-controlled trials attempted to blind both participants and outcome assessors.13 ,14 ,17 Five studies that analysed all randomised patients,11–13 or had loss to follow-up or exclusions that were deemed unlikely to affect the study results,14 ,16 ,17 were rated as having a low risk of attrition bias. Only one study reported all pre-specified outcomes and was rated as having a low risk of selective reporting bias.13 There were no other apparent sources of bias (figure 2; see online supplementary file 4).

Figure 2

Risk of bias among included trials (presented using Cochrane risk of bias assessment).

Primary outcomes

Postoperative pain

Four studies reported postoperative pain but varied widely with respect to their methods of pain measurement.11 ,13 ,14 ,16 Compared with sham or usual care alone, there was no overall benefit of acupuncture14 ,16 or modest benefit only on postoperative day (POD) 2 and 3 (figure 3).13 No additional benefit of acupuncture on abdominal pain was observed when it was used as an adjunct to fast-track recovery at POD 1 (RR 0.89, 95% CI 0.47 to 1.67) or POD 5 (RR 1.00, 95% CI 0.07 to 14.55).11

Figure 3

Forest plots demonstrating effect of acupuncture on postoperative pain and nausea/vomiting. POD, postoperative day.

Postoperative nausea/vomiting

Four studies reported rates of PONV.11 ,14 ,16 ,17 Two studies reported on postoperative nausea only,14 ,16 and the other two studies provided information separately for nausea and vomiting events.11 ,17 The outcome of post-operative nausea could not be meta-analysed due to variability in the measurement methods. One study reported similar numbers of patients experiencing an emetic episode (18% of the acupuncture group vs 10% of the placebo acupuncture group) and no significant difference in mean nausea scores from POD 1–3 when measured on a scale of 0 to 10 (MD −0.10, 95% CI −1.00 to 0.80).14 Similarly, acupuncture as an adjunct to usual care did not significantly impact nausea scores on a 0–10 scale (MD 0.46, 95% CI −0.12 to 1.04).16 The risk of nausea in patients receiving acupuncture as an adjunct to a fast-track recovery programme was higher on POD 1 (RR 1.75, 95% CI 0.64 to 4.75) but lower on POD 5 (RR 0.44, 95% CI 0.17 to 1.13).11 Similar results were observed in a study that compared verum EA at ST36 with sham EA (RR 0.35, 95% CI 0.08 to 1.53).17 However, rates of postoperative vomiting in the acupuncture group did not differ from the fast-track only or sham EA groups, respectively, in these two studies (figure 3).11 ,17

Other symptoms

There were no differences between acupuncture and usual care in terms of sleep disturbance (n=85; MD −0.07, 95% CI −0.90 to 0.76)16 or abdominal distension (n=85; MD 0.22, 95% CI −0.34 to 0.78).16 Similarly, acupuncture as an adjunct to a fast-track recovery programme did not influence abdominal distension on POD 1 (n=30; RR 1.00, 95% CI 0.65 to 1.54)11 or POD 5 (RR 0.67, 95% CI 0.13 to 3.44).11 No study measured post-discharge QoL.

Secondary outcomes

All included trials reported at least one secondary outcome (table 2). Meta-analysis revealed that, compared to sham acupuncture, some markers of physiological recovery were improved in the acupuncture group, namely time to first flatus (n=207, three studies; MD −7.48 h, 95% CI −14.58 to −0.39 h, I2=0%) and time to first defaecation (n=149, two studies; MD −18.04 h, 95% CI −31.90 to −4.19 h, I2=0%). Acupuncture was marginally superior to sham acupuncture in terms of time taken for patients to be able to walk independently. Due to a high degree of heterogeneity among the studies, the use of medication compared with sham acupuncture and physiological recovery outcomes compared with usual care alone were not meta-analysed.

Table 2

Effect estimates of secondary outcomes

Postoperative complications

Three studies reported postoperative complications. One study assessed the number of patients with prolonged postoperative ileus (PPOI), which occurred in 21/36 individuals in the acupuncture group and 25/40 in the usual care group.16 PPOI resolved in all patients by POD 7. Two studies measured a set of complications including PPOI as a composite outcome.11 ,13 The total number of complications was 6/55, 5/55, and 10/55 in the acupuncture, sham acupuncture, and usual care (no acupuncture) groups, respectively,13 and 7/31 in the acupuncture plus fast-track recovery programme group versus 3/29 in the fast-track recovery only group.11

Safety of acupuncture

Two studies stated that there were no adverse events related to acupuncture,13 ,16 whereas five studies did not report whether adverse events related to acupuncture occurred or not.11 ,12 ,14 ,15 ,17


Summary of findings

Seven trials with 540 participants were included in this review. The significant heterogeneity and high/unclear risk of bias for the domains of allocation concealment, blinding of patients and/or outcome assessors and selective reporting mandate careful interpretation of the study results. Postoperative symptoms and short-term markers of physiological recovery within 3 weeks of surgery were frequently reported, whereas functional aspects or QoL were rarely measured. The safety of acupuncture remains unclear due to the incomplete reporting of information regarding the safety/harms of treatment.

Overall completeness and applicability of evidence

Previous evidence has primarily focused on the role of acupuncture in the management of symptoms such as postoperative pain or nausea/vomiting.6 ,18–20 Decreased pain intensity may result in the reduced use of opioid drugs, which is consequently associated with a reduction in opioid-related side effects such as nausea, dizziness, sedation, pruritus, and urinary retention.21 Experimental studies have suggested that acupuncture can modulate gastrointestinal mobility via excitation of the vagus nerve,22 opioidergic/serotonergic pathways23–25 or spinal/supraspinal reflexes,22 all of which may be relevant to the early phase of postoperative recovery after colorectal surgery. The findings of this review are consistent with existing clinical and experimental evidence and highlight the need for further well-designed RCTs to support the case for an expanded role of acupuncture in recovery after colorectal cancer surgery.

The studies included in this review did not address long-term post-discharge functional recovery or patient-reported outcomes and QoL. Patients have clearly not fully recovered at the time of hosptial discharge after colorectal surgery,26 ,27 yet most studies that evaluate interventions to assist recovery after surgery concentrate on reductions in the length of hospital stay and short-term changes in basic organ function.2 However, recovery is not confined to the restoration of physiological and basic organ function but also includes the ability to perform regular activities, which, from the patients’ perspective, typically takes weeks to months after surgery.3 ,28 Assessments of successful recovery after surgery should therefore be based on both early in-hospital phase biological/physiological outcomes and longer term patient-reported functional improvements and QoL post-discharge.2 ,3 As these relatively more complex aspects of recovery were not well addressed by the findings of this review, future trials focusing on multi-dimensional outcomes reflecting both early and long-term trajectories of recovery from surgery are warranted.

In all included studies, acupuncture was delivered postoperatively, although the exact timing of the intervention was not always clearly reported. In two meta-analyses of acupuncture for PONV, it was suggested that differences in the timing of acupuncture (ie, pre-, intra- or postoperative) were associated with the observed heterogeneity.6 ,18 There is no conclusive evidence for the optimal timing of acupuncture for improved postoperative outcomes. Previous studies in which acupuncture was delivered after the induction of anaesthesia observed decreased efficacy for PONV and a lesser reduction in analgesic use.8 The effects of acupuncture administered under anaesthesia have been questioned because general anaesthesia can block the sensory stimulation of acupuncture and potentially interfere with the induction of the necessary physiological and clinical effects to prevent PONV.29 ,30 A recent RCT observed favourable effects of intra-/postoperative acupuncture on postoperative analgesia over sham acupuncture in patients undergoing prostatectomy.31 Another trial reported favourable analgesic effects of both perioperative (pre-/intra-/postoperative) acupuncture and pre-/postoperative acupuncture over sham acupuncture, but no additional benefit of intraoperative acupuncture in inguinal hernia surgery with abdominal wall mesh reconstruction.32 Potential benefits of inserting superficial indwelling acupuncture needles before surgery have been proposed, including reduced preoperative anxiety, attenuated use of anaesthetic agents during surgery, and decreased postoperative pain intensity and PONV.21 However, these results should be interpreted with caution due to the paucity of well-designed trials directly comparing the effects of different timings of acupuncture.8 Future trials should explore the potential impact of the timing of acupuncture on postoperative outcomes and report the method and timing of acupuncture delivery in detail.33

The trials included in this review used different methods to measure the same outcomes. For example, the methods of reporting pain varied and included average pain over POD 1–3 as well as pain levels on each individual POD. Differences in the types of surgery, acupuncture stimulation methods (manual or electrical), and comparators among the studies are also possible contributors to the observed heterogeneity.

Potential bias and limitations of the review

The population of interest in this review was patients undergoing colorectal cancer resection, thus the findings of the review may not be applicable to other types of surgery. One major limitation is the exclusion of trials that reported relevant outcomes but were not eligible due to differences in the study population. The rationale to limit the scope of our review was that colorectal cancer patients have specific symptom clusters and illness trajectories, and we aimed to cover broad clinical outcomes that are particularly relevant to colorectal cancer patients. For example, coping with a stoma or persistent chronic pain in the abdominal or lower pelvic region are specific issues that affect QoL after surgery in colorectal cancer patients but not patients undergoing other types of operation, such as tonsillectomy.34 Moreover, the magnitude of the effect on length of hospital stay may have different clinical implications for different types of surgery or in different contexts (eg, ambulatory surgery settings). For example, the Enhanced Recovery After Surgery (ERAS) Society provides separate guidelines for perioperative care for different types of surgery.35 Until sufficient evidence accumulates, limiting the population of interest to maintain clinical relevance remains a reasonable approach for under-investigated topics such as that addressed by the present review. A further limitation of this review is that it is based on relatively few RCTs, of low to moderate quality only, and we were unable to examine publication bias and sources of heterogeneity due to the small number of included studies.

Implications for practice

Low to moderate quality evidence supports the use of acupuncture during recovery after colorectal cancer surgery. Acupuncture is considered a relatively safe intervention when performed by a qualified practitioner but is not a risk-free procedure.36 The safety of acupuncture in patients undergoing colorectal cancer surgery remains unclear due to incomplete reporting within the primary studies. Local needling points of the abdomen may promote local wound healing mediated by increased local muscle blood supply,37 despite the potential risk of irritating the incision area or inducing local tissue infection, which may delay discharge and subsequent chemotherapy.38 In our review, all acupuncture points used in the included studies were peripheral (located in the extremities) with the single exception of ST25, which was used in one study.14 Optimal combination of acupuncture points for safe and effective practice in the early postoperative phase needs to be further investigated. There was also a lack of evidence regarding whether acupuncture is beneficial for post-discharge management.

Implications for research

In future trials, concealment of allocation, blinding of outcome assessors, transparent reporting of participant flows, analysis according to the intention-to-treat principle and prospective trial registration39 are recommended to reduce the risk of bias. Post-discharge outcomes including long-term functional improvements and QoL following surgery are necessary to evaluate the sustained benefits of acupuncture. Transparency in the methods used to define and measure physiological outcomes is necessary and adverse events related to acupuncture treatment should be clearly reported. Sources of heterogeneity should be investigated in future research to identify factors that are important for improved recovery outcomes. Currently, acupuncture is not incorporated into the ERAS guidelines for perioperative multimodal care in elective colonic surgery.4 The potential role of acupuncture as one of a range of multimodal interventions or as a supporting treatment for long-term care post-discharge, as suggested by previous studies,40 ,41 would be of interest in future research.


There is low to moderate quality evidence that acupuncture may be safe and beneficial for short-term symptom reduction and physiological recovery after colorectal cancer surgery. The paucity of primary studies, high risk of bias, and incomplete reporting of the included trials are the main limitations of this review. Future well-designed RCTs are required to evaluate the role of acupuncture in the recovery of patients from colorectal cancer surgery.


View Abstract


  • Contributors KHK conceived and designed the study. KHK and DHK searched the literature and extracted the information. KHK, DHK and GMS analysed the data. KHK wrote the draft and the final manuscript. All authors read and criticised the draft. All authors read and approved the final manuscript. KHK and KMS coordinated the study.

  • Funding This work was supported by a grant to the Korean Medical Science Research Center for Healthy Aging from the National Research Foundation of Korean Government (grant no. 2014R1A5A2009936) and the Korea Institute of Oriental Medicine (grant no. K16123). The funders had no role in the conduct of this study.

  • Competing interests None declared.

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Data sharing statement More information for the review can be obtained by contacting the first author or corresponding author.

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.