Langenbecks Arch Surg (2015) 400:567–576 DOI 10.1007/s00423-015-1316-9
SYSTEMATIC REVIEW AND META-ANALYSIS
The efficacy of adhesiolysis on chronic abdominal pain: a systematic review Jonas Gerner-Rasmussen 1 & Jakob Burcharth 1 & Ismail Gögenur 2
Received: 5 February 2015 / Accepted: 10 June 2015 / Published online: 20 June 2015 # Springer-Verlag Berlin Heidelberg 2015
Abstract Introduction Abdominal adhesions are a frequent reason for chronic abdominal pain. The purpose of this systematic review was to investigate the evidence of performing laparoscopic adhesiolysis as a treatment for patients with chronic abdominal pain. Methods Medline, Embase, and The Cochrane Central Register of Controlled Trials were searched for trials performing lysis of adhesions on patients suffering from chronic abdominal pain. Clinical studies on patients being treated for chronic abdominal pain with surgical adhesiolysis were included. The main outcome of the study was the postoperative assessment of symptoms. The Newcastle-Ottawa scale was used for bias assessment of non-randomized studies while the Jadad score was used for the randomized controlled trials. Results A total of 25 studies were identified evaluating the efficacy of adhesiolysis in 1281 patients suffering from chronic abdominal pain. A total of 22 trials were identified as caseseries and included no control group. Three studies were identified as randomized controlled trials (RCT). A benefit of the intervention varied from 16 to 88 % in the non-randomized studies, with the majority reporting pain relief in more than 50 % of their patients. However, analysis indicated a high risk of bias in most of the studies. The RCTs also showed variance, with one study showing benefit, one study showing benefit only in a subgroup in which the patients had dense and
* Jonas Gerner-Rasmussen [email protected] 1
Department of Surgery, Herlev Hospital, University of Copenhagen, Herlev Ringvej 75, 2730 Herlev, Denmark
2
Department of Surgery, Roskilde and Koege Hospitals, University of Copenhagen, 4600 Koege, Denmark
vascularized adhesions, and one study showed no difference between the intervention and control group. Conclusion The identified studies showed promising but preliminary results of laparoscopic adhesiolysis as a treatment of chronic abdominal pain. The evidence for laparoscopic adhesiolysis is not sufficient to make definitive conclusions. Keywords Adhesions . Abdominal pain . Chronic pain . Bowel symptoms . Adhesiolysis . Laparoscopy . Laparotomy
Introduction Chronic abdominal pain is a difficult issue for physicians and surgeons with regard to diagnostics and treatment. Patients suffering from chronic abdominal pain are frequently admitted to surgical clinics and are often subjected to multiple diagnostic tests with no conclusive outcome [1–3]. Laparoscopy allows surgeons to visualize pathologic changes, that otherwise would not have been discovered with conventional diagnostic imagery and physical examination. Through laparoscopy in patients with chronic abdominal pain, adhesions have often been the sole finding, giving cause to believe that adhesions are a possible source of the chronic pain [4, 5]. A meta-analysis investigating reasons for chronic abdominal pain found that adhesions were the most likely cause of pain in 57 % of the patients [6]; however, controversies still exist regarding the correlation between the presence of adhesions and chronic abdominal pain. One study found an equal prevalence of chronic abdominal pain independent of the presence of adhesions (30–31 % in both target and control groups) [7]. It has thus been speculated that adhesiolysis would be inadequate in the treatment of chronic abdominal pain [8], partly due to the above-mentioned and partly due to a high degree of adhesion reformation [9]. It has been shown that
568
laparoscopy is less probable to cause postoperative adhesions [10–12], making it a relevant procedure for adhesiolysis. The purpose of this systematic review was to investigate if laparoscopic adhesiolysis was useful in reducing symptoms in patients with chronic abdominal pain.
Methods The reporting of this systematic review was performed according to the Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA) statement [13]. We performed study selection, assessment of eligibility criteria, data extraction in accordance with a predefined protocol registered at the International Prospective Register of Systematic Reviews (PROSPERO) (PROSPERO ID: CRD42014009095) as recommended by the methodology guidelines from the Cochrane Collaboration [14]. Eligibility criteria The selection of studies was performed using PICO(S) (Fig. 1). All English written observational (case-control, cohort, register-based) studies that fulfilled the PICO(S) were eligible for inclusion. There were no limitations regarding the date of publication of the included studies. Only studies retrievable in full-text were subject to possible inclusion. For a study to be suitable for the qualitative synthesis, it had to contain information on patients being treated for chronic abdominal pain with surgical adhesiolysis. In addition, the study had to describe a defined follow-up period exceeding the primary hospital stay, in which there had to be an assessment of the patients’ pain progress. Any study not meeting the abovementioned requirements was not eligible for inclusion.
Langenbecks Arch Surg (2015) 400:567–576
Study selection Records were screened by two authors (JB, JGR), and the records deemed conceivably relevant were obtained in fulltext and evaluated in detail. The full-text records were scrutinized by the two authors (JB, JGR) independently, who also performed data extraction. All discrepancies in the abovementioned were settled by discussion and by conference with the senior author (IG). The process of the selection of studies is depicted in Fig. 2. Data collection process and data items A standard data-extraction form was developed for data collection. Information from the included studies was systematically extracted and handled in a customized Microsoft Excel (2011 for Mac, Version 14.0.0, © 2010 Microsoft Corporation) spread sheet. Extraction variables consisted of demographic baseline variables and study characteristics and included age, gender, mean duration of symptoms, study design, method of data collection, mode of surgery, number of subjects, duration of follow-up (months), and patient dropout rate. The main outcome of interest was any kind of assessment of the patients’ experience of postoperative pain in the followup period. This included the subjective report of improvement or non-improvement of symptoms, change in visual analogue scale (VAS), the McGill pain score, the medical outcome study score in short form (MOS SF-36) or the disability score. When possible, we prioritized data from validated pain assessment scales. Alternatively, the mean change from baseline and corresponding standard deviation was extracted. When necessary, we approximated means and measures of dispersion from figures in the reports. We divided the included studies into groups, according to their surgical approach, making them available for a subgroup analysis.
Information sources and search strategy Risk of bias assessment To identify all trials relating chronic abdominal pain to adhesiolysis, we searched the following bibliographic databases: MEDLINE via PubMed from 1950, EMBASE via OVID from 1980, and The Cochrane Central Register of Controlled Trials (CENTRAL) from 1898. Reference lists of the retrieved reviews and selected studies were searched for other relevant publications/reports as described [15]. The search strategy from MEDLINE was (((((((((((Bowel obstruction) OR ileus) OR intestinal obstruction) OR small bowel obstruction) OR SBO) OR post-operative) OR post operative) OR adhesion) OR adhesions)) AND (((laparoscopy) OR laparotomy) OR adhesiolysis)) AND (((((((resolution of symptoms) OR analgesic consumption) OR analgesics consumption) OR abdominal pain) OR pelvic pain) OR visceral pain) OR chronic pain). The literature search was adapted to EMBASE and CENTRAL from the MEDLINE search strategy.
For the non-randomized trials, risk of bias (ROB) on study level was investigated as suggested by the Cochrane Collaboration [16] by the use of Newcastle-Ottawa scale (NOS) [17]. The NOS uses a Bstar-based^ rating system and scores the studies from 0 to 9, where 9 indicates the lowest degree of bias and 0 indicates the highest degree of bias on three levels (selection, comparability, and outcome assessment). Studies scoring 1–3 were considered to be at high risk of bias, while 4–6 indicated moderate risk of bias and 7–9 indicated low risk of bias. In the NOS outcome assessment, an acceptable dropout rate as well as a minimum follow-up period must be predefined [17]. We considered an acceptable dropout rate to be less than 10 % and 12 months as the adequate duration of follow-up. Regarding the randomized trials, ROB was assessed by the Jadad score [18], which ranges from 0 (bad)
Langenbecks Arch Surg (2015) 400:567–576
569
Fig. 1 PICOS
to 5 (good) and evaluates methodological elements such as randomization, sequence generation, method of blinding, and dropouts. In addition to the NOS and Jadad score, we supplemented the risk of bias evaluation of the included studies by examining each study’s effort to exclude other causes for the chronic abdominal pain than adhesions. Ideally, this was perceived as the attempt to exclude psychological and psychiatric factors, through scrutiny of medical records to exclude psychiatric diagnosis and with the aid of a recognized diagnostic interview form to screen for psychological illnesses. Also, we found it pertinent for the authors to have investigated the possibility of other intra-abdominal pathologies, by a thorough collection of medical history, physical examination, and diagnostic imagery. Finally, we extracted the reported mean duration of symptoms before the intervention.
Results Study selection and study characteristics This review included a total of 25 studies out of the 5615 records found at the original search. The process of inclusion
is depicted in Fig. 2. Of the included studies, three were identified as randomized controlled trials (RCTs) [19–21] and the remainder was case-series without controls. The total patient population of the included studies was n=1281, with 151 (12 %) males and 1128 (88 %) females. The age of the patients ranged from 17 to 65 years, but age was not described in four studies [22–25]. All patients underwent adhesiolysis for chronic abdominal pain, and 90.3 % of the population had previously undergone surgery. The majority of the patients in the included trials (87 %) underwent laparoscopic adhesiolysis, whereas only three of the included studies involved adhesiolysis through laparotomy [19, 22, 26]. We further divided the laparoscopic studies into groups according to their surgical approach used to achieve adhesiolysis. In this manner, we categorized two studies using adhesion barriers after adhesiolysis [25, 27], four studies using CO2 laser when lysing adhesions [23, 24, 28, 29], and 11 studies using the conventional technique with sharp dissection (scissors along with electrocoagulation to achieve hemostasis) [1, 20, 21, 30–37]. A total of five studies could not be classified, as they did not describe their surgical method [38–42]. In general, chronic pain was summarized in three different ways. The majority of the studies assessed pain by simple
570
Langenbecks Arch Surg (2015) 400:567–576
Fig. 2 Study selection process
dichotomy, where the investigator asked the patients whether the pain had improved or stayed the same [1, 19, 22–29, 33–35, 39–42]. Other studies distinguished between improvement and complete relief [20, 30–32, 36], while only eight studies assessed pain through recognized pain scales such as the visual analogue scale [20, 21, 27, 37] or the McGill pain scale [19, 35, 38, 40] which ranges from 0 to 45 and considers different aspects of subjective pain experience. In addition to the above-mentioned, only two studies investigated the patients consumption of analgesics [20, 25]. Each study’s characteristics are specified in Table 1.
Risk of bias The Newcastle-Ottawa scale was used to investigate the risk of bias of the non-randomized studies. It should be noted that of the possible nine stars given in the NOS assessment, three stars attributed by the NOS are awarded on merits regarding the comparability and characteristics of the control group. Since none of the included observational studies had reported control groups, they achieved a NOS score of no higher than 6, classifying them all to a moderate to high risk of bias. For the studies evaluated by NOS, the median score was 4.5 (3–6)
Langenbecks Arch Surg (2015) 400:567–576 Table 1
571
Study results
Adhesiolysis with a special technique Adhesion barrier—laparoscopy Author Publication date Khaitan et al. 2003
Number 19
Steege et al.
25
Laser—laparoscopy Author Daniell et al. Fayez et al. Nezhat et al. Sutton et al. Adhesiolysis only Laparotomy Author Chan et al. Peters et al.
Saravelos et al.
1991
Keltz et al.
7.96±1.96 3.92±3.56 (p
SYSTEMATIC REVIEW AND META-ANALYSIS
The efficacy of adhesiolysis on chronic abdominal pain: a systematic review Jonas Gerner-Rasmussen 1 & Jakob Burcharth 1 & Ismail Gögenur 2
Received: 5 February 2015 / Accepted: 10 June 2015 / Published online: 20 June 2015 # Springer-Verlag Berlin Heidelberg 2015
Abstract Introduction Abdominal adhesions are a frequent reason for chronic abdominal pain. The purpose of this systematic review was to investigate the evidence of performing laparoscopic adhesiolysis as a treatment for patients with chronic abdominal pain. Methods Medline, Embase, and The Cochrane Central Register of Controlled Trials were searched for trials performing lysis of adhesions on patients suffering from chronic abdominal pain. Clinical studies on patients being treated for chronic abdominal pain with surgical adhesiolysis were included. The main outcome of the study was the postoperative assessment of symptoms. The Newcastle-Ottawa scale was used for bias assessment of non-randomized studies while the Jadad score was used for the randomized controlled trials. Results A total of 25 studies were identified evaluating the efficacy of adhesiolysis in 1281 patients suffering from chronic abdominal pain. A total of 22 trials were identified as caseseries and included no control group. Three studies were identified as randomized controlled trials (RCT). A benefit of the intervention varied from 16 to 88 % in the non-randomized studies, with the majority reporting pain relief in more than 50 % of their patients. However, analysis indicated a high risk of bias in most of the studies. The RCTs also showed variance, with one study showing benefit, one study showing benefit only in a subgroup in which the patients had dense and
* Jonas Gerner-Rasmussen [email protected] 1
Department of Surgery, Herlev Hospital, University of Copenhagen, Herlev Ringvej 75, 2730 Herlev, Denmark
2
Department of Surgery, Roskilde and Koege Hospitals, University of Copenhagen, 4600 Koege, Denmark
vascularized adhesions, and one study showed no difference between the intervention and control group. Conclusion The identified studies showed promising but preliminary results of laparoscopic adhesiolysis as a treatment of chronic abdominal pain. The evidence for laparoscopic adhesiolysis is not sufficient to make definitive conclusions. Keywords Adhesions . Abdominal pain . Chronic pain . Bowel symptoms . Adhesiolysis . Laparoscopy . Laparotomy
Introduction Chronic abdominal pain is a difficult issue for physicians and surgeons with regard to diagnostics and treatment. Patients suffering from chronic abdominal pain are frequently admitted to surgical clinics and are often subjected to multiple diagnostic tests with no conclusive outcome [1–3]. Laparoscopy allows surgeons to visualize pathologic changes, that otherwise would not have been discovered with conventional diagnostic imagery and physical examination. Through laparoscopy in patients with chronic abdominal pain, adhesions have often been the sole finding, giving cause to believe that adhesions are a possible source of the chronic pain [4, 5]. A meta-analysis investigating reasons for chronic abdominal pain found that adhesions were the most likely cause of pain in 57 % of the patients [6]; however, controversies still exist regarding the correlation between the presence of adhesions and chronic abdominal pain. One study found an equal prevalence of chronic abdominal pain independent of the presence of adhesions (30–31 % in both target and control groups) [7]. It has thus been speculated that adhesiolysis would be inadequate in the treatment of chronic abdominal pain [8], partly due to the above-mentioned and partly due to a high degree of adhesion reformation [9]. It has been shown that
568
laparoscopy is less probable to cause postoperative adhesions [10–12], making it a relevant procedure for adhesiolysis. The purpose of this systematic review was to investigate if laparoscopic adhesiolysis was useful in reducing symptoms in patients with chronic abdominal pain.
Methods The reporting of this systematic review was performed according to the Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA) statement [13]. We performed study selection, assessment of eligibility criteria, data extraction in accordance with a predefined protocol registered at the International Prospective Register of Systematic Reviews (PROSPERO) (PROSPERO ID: CRD42014009095) as recommended by the methodology guidelines from the Cochrane Collaboration [14]. Eligibility criteria The selection of studies was performed using PICO(S) (Fig. 1). All English written observational (case-control, cohort, register-based) studies that fulfilled the PICO(S) were eligible for inclusion. There were no limitations regarding the date of publication of the included studies. Only studies retrievable in full-text were subject to possible inclusion. For a study to be suitable for the qualitative synthesis, it had to contain information on patients being treated for chronic abdominal pain with surgical adhesiolysis. In addition, the study had to describe a defined follow-up period exceeding the primary hospital stay, in which there had to be an assessment of the patients’ pain progress. Any study not meeting the abovementioned requirements was not eligible for inclusion.
Langenbecks Arch Surg (2015) 400:567–576
Study selection Records were screened by two authors (JB, JGR), and the records deemed conceivably relevant were obtained in fulltext and evaluated in detail. The full-text records were scrutinized by the two authors (JB, JGR) independently, who also performed data extraction. All discrepancies in the abovementioned were settled by discussion and by conference with the senior author (IG). The process of the selection of studies is depicted in Fig. 2. Data collection process and data items A standard data-extraction form was developed for data collection. Information from the included studies was systematically extracted and handled in a customized Microsoft Excel (2011 for Mac, Version 14.0.0, © 2010 Microsoft Corporation) spread sheet. Extraction variables consisted of demographic baseline variables and study characteristics and included age, gender, mean duration of symptoms, study design, method of data collection, mode of surgery, number of subjects, duration of follow-up (months), and patient dropout rate. The main outcome of interest was any kind of assessment of the patients’ experience of postoperative pain in the followup period. This included the subjective report of improvement or non-improvement of symptoms, change in visual analogue scale (VAS), the McGill pain score, the medical outcome study score in short form (MOS SF-36) or the disability score. When possible, we prioritized data from validated pain assessment scales. Alternatively, the mean change from baseline and corresponding standard deviation was extracted. When necessary, we approximated means and measures of dispersion from figures in the reports. We divided the included studies into groups, according to their surgical approach, making them available for a subgroup analysis.
Information sources and search strategy Risk of bias assessment To identify all trials relating chronic abdominal pain to adhesiolysis, we searched the following bibliographic databases: MEDLINE via PubMed from 1950, EMBASE via OVID from 1980, and The Cochrane Central Register of Controlled Trials (CENTRAL) from 1898. Reference lists of the retrieved reviews and selected studies were searched for other relevant publications/reports as described [15]. The search strategy from MEDLINE was (((((((((((Bowel obstruction) OR ileus) OR intestinal obstruction) OR small bowel obstruction) OR SBO) OR post-operative) OR post operative) OR adhesion) OR adhesions)) AND (((laparoscopy) OR laparotomy) OR adhesiolysis)) AND (((((((resolution of symptoms) OR analgesic consumption) OR analgesics consumption) OR abdominal pain) OR pelvic pain) OR visceral pain) OR chronic pain). The literature search was adapted to EMBASE and CENTRAL from the MEDLINE search strategy.
For the non-randomized trials, risk of bias (ROB) on study level was investigated as suggested by the Cochrane Collaboration [16] by the use of Newcastle-Ottawa scale (NOS) [17]. The NOS uses a Bstar-based^ rating system and scores the studies from 0 to 9, where 9 indicates the lowest degree of bias and 0 indicates the highest degree of bias on three levels (selection, comparability, and outcome assessment). Studies scoring 1–3 were considered to be at high risk of bias, while 4–6 indicated moderate risk of bias and 7–9 indicated low risk of bias. In the NOS outcome assessment, an acceptable dropout rate as well as a minimum follow-up period must be predefined [17]. We considered an acceptable dropout rate to be less than 10 % and 12 months as the adequate duration of follow-up. Regarding the randomized trials, ROB was assessed by the Jadad score [18], which ranges from 0 (bad)
Langenbecks Arch Surg (2015) 400:567–576
569
Fig. 1 PICOS
to 5 (good) and evaluates methodological elements such as randomization, sequence generation, method of blinding, and dropouts. In addition to the NOS and Jadad score, we supplemented the risk of bias evaluation of the included studies by examining each study’s effort to exclude other causes for the chronic abdominal pain than adhesions. Ideally, this was perceived as the attempt to exclude psychological and psychiatric factors, through scrutiny of medical records to exclude psychiatric diagnosis and with the aid of a recognized diagnostic interview form to screen for psychological illnesses. Also, we found it pertinent for the authors to have investigated the possibility of other intra-abdominal pathologies, by a thorough collection of medical history, physical examination, and diagnostic imagery. Finally, we extracted the reported mean duration of symptoms before the intervention.
Results Study selection and study characteristics This review included a total of 25 studies out of the 5615 records found at the original search. The process of inclusion
is depicted in Fig. 2. Of the included studies, three were identified as randomized controlled trials (RCTs) [19–21] and the remainder was case-series without controls. The total patient population of the included studies was n=1281, with 151 (12 %) males and 1128 (88 %) females. The age of the patients ranged from 17 to 65 years, but age was not described in four studies [22–25]. All patients underwent adhesiolysis for chronic abdominal pain, and 90.3 % of the population had previously undergone surgery. The majority of the patients in the included trials (87 %) underwent laparoscopic adhesiolysis, whereas only three of the included studies involved adhesiolysis through laparotomy [19, 22, 26]. We further divided the laparoscopic studies into groups according to their surgical approach used to achieve adhesiolysis. In this manner, we categorized two studies using adhesion barriers after adhesiolysis [25, 27], four studies using CO2 laser when lysing adhesions [23, 24, 28, 29], and 11 studies using the conventional technique with sharp dissection (scissors along with electrocoagulation to achieve hemostasis) [1, 20, 21, 30–37]. A total of five studies could not be classified, as they did not describe their surgical method [38–42]. In general, chronic pain was summarized in three different ways. The majority of the studies assessed pain by simple
570
Langenbecks Arch Surg (2015) 400:567–576
Fig. 2 Study selection process
dichotomy, where the investigator asked the patients whether the pain had improved or stayed the same [1, 19, 22–29, 33–35, 39–42]. Other studies distinguished between improvement and complete relief [20, 30–32, 36], while only eight studies assessed pain through recognized pain scales such as the visual analogue scale [20, 21, 27, 37] or the McGill pain scale [19, 35, 38, 40] which ranges from 0 to 45 and considers different aspects of subjective pain experience. In addition to the above-mentioned, only two studies investigated the patients consumption of analgesics [20, 25]. Each study’s characteristics are specified in Table 1.
Risk of bias The Newcastle-Ottawa scale was used to investigate the risk of bias of the non-randomized studies. It should be noted that of the possible nine stars given in the NOS assessment, three stars attributed by the NOS are awarded on merits regarding the comparability and characteristics of the control group. Since none of the included observational studies had reported control groups, they achieved a NOS score of no higher than 6, classifying them all to a moderate to high risk of bias. For the studies evaluated by NOS, the median score was 4.5 (3–6)
Langenbecks Arch Surg (2015) 400:567–576 Table 1
571
Study results
Adhesiolysis with a special technique Adhesion barrier—laparoscopy Author Publication date Khaitan et al. 2003
Number 19
Steege et al.
25
Laser—laparoscopy Author Daniell et al. Fayez et al. Nezhat et al. Sutton et al. Adhesiolysis only Laparotomy Author Chan et al. Peters et al.
Saravelos et al.
1991
Keltz et al.
7.96±1.96 3.92±3.56 (p
