Can decision making in general surgery be based on evidence? An empirical study of Cochrane Reviews

Original Communications Can decision making in general surgery be based on evidence? An empirical study of Cochrane Reviews Markus K. Diener, MD,a,b Robert F. Wolff, MD,b Erik von Elm, MD,b,c Nuh N. Rahbari, MD,a Chris Mavergames, PhD,b Hanns-Peter Knaebel, MD,d Christoph M. Seiler, MD,a and Gerd Antes, PhD,b Heidelberg, Freiburg, and Tuttlingen, Germany, and Bern, Switzerland

Background. This empirical study analyzes the current status of Cochrane Reviews (CRs) and their strength of recommendation for evidence-based decision making in the field of general surgery. Methods. Systematic literature search of the Cochrane Database of Systematic Reviews and the Cochrane Collaboration’s homepage to identify available CRs on surgical topics. Quantitative and qualitative characteristics, utilization, and formulated treatment recommendations were evaluated by 2 independent reviewers. Association of review characteristics with treatment recommendation was analyzed using univariate and multivariate logistic regression models. Results. Ninety-three CRs, including 1,403 primary studies and 246,473 patients, were identified. Mean number of included primary studies per CR was 15.1 (standard deviation [SD] 14.5) including 2,650 (SD 3,340) study patients. Two and a half (SD 8.3) nonrandomized trials were included per analyzed CR. Seventy-two (77%) CRs were published or updated in 2005 or later. Explicit treatment recommendations were given in 45 (48%). Presence of a treatment recommendation was associated with the number of included primary studies and the proportion of randomized studies. Utilization of surgical CRs remained low and showed large inter-country differences. The most surgical CRs were accessed in UK, USA, and Australia, followed by several Western and Eastern European countries. Conclusion. Only a minority of available CRs address surgical questions and their current usage is low. Instead of unsystematically increasing the number of surgical CRs it would be far more efficient to focus the review process on relevant surgical questions. Prioritization of CRs needs valid methods which should be developed by the scientific surgical community. (Surgery 2009;146:444-61.) From the Department of General, Visceral and Transplantation Surgery, University of Heidelberg,a Heidelberg; Institute of Medical Biometry and Medical Informatics, German Cochrane Centre, University Medical Centre,b Freiburg, Germany; Institute of Social and Preventive Medicine, University of Bern,c Bern, Switzerland; and Aesculap AG, Am Aesculap Platz,d Tuttlingen, Germany

THREE STEPS are necessary to base medical decision making on the best available evidence: summarizing all evidence on a given clinical question systematically, translating it into clinical guidance, and applying this guidance at the right time and place to the right patient.1 The large number of biomedical publications makes it more and more difficult for busy clinicians, researchers, and medical teachers to Accepted for publication February 20, 2009. Reprint requests: Markus K. Diener, MD, Department of General, Visceral and Transplantation Surgery, University of Heidelberg, Im Neuenheimer Feld 110, D-69120 Heidelberg, Germany. E-mail: [email protected]. 0039-6060/$ - see front matter Ó 2009 Mosby, Inc. All rights reserved. doi:10.1016/j.surg.2009.02.016

444 SURGERY

keep abreast of new developments. Also, it is often difficult to distinguish valid medical information from other information that is of poor quality.2 Hence, in clinical practice the steadily growing body of evidence is not used in a timely and efficient manner. In contrast, forms of care deemed effective by many clinicians are often not evaluated scientifically, or the existing evidence on their effectiveness is underused.3 The resulting discrepancy between knowledge and practice has become known as the ‘‘know-do gap.’’4 Systematic reviews are an effective means to synthesize the scientific evidence and make it available for clinical practice.5 Of course, the strength of their recommendations is limited by the quality of included primary studies.6,7 In addition, empirical studies suggest that not all

Surgery Volume 146, Number 3

systematic reviews are performed with sufficient rigor to give an unbiased account of current knowledge.8 The Cochrane Collaboration, an international not-for-profit scientific organization, prepares and provides high-quality systematic reviews about the effects of healthcare. The reviews are made available in a comprehensive electronic library, the Cochrane Library.9,10 Evidence-based methods have been used in several medical fields for more than 15 years. In surgery, however, they have been introduced only recently. Thus, the potential of evidence-based medicine has not yet been fully applied to this discipline. Many topical questions of clinical relevance to surgeons await a systematic and scientifically rigorous assessment.8,11 Often, the questions that were addressed by systematic reviews could not be answered and definitive treatment recommendation could not be given. Simply, the available evidence was not strong enough to confirm the current practice or to refute it. We aimed to analyze Cochrane Reviews (CRs) on general surgery topics to gain insight into their current strength of recommendation, timeliness, and utilization. METHODS Search strategy and selection criteria. We systematically searched the Cochrane Database of Systematic Reviews (Cochrane Library [www.theco chranelibrary.com]) and the website of the Cochrane Collaboration (www.cochrane.org) for surgical CRs. Searches were conducted in March 2008 on issue 1/2008 of the Cochrane Library (Fig 1). We used only Cochrane Reviews because these reviews are conducted within a common methodological framework with standards of quality ensured by the supervising Cochrane Review Groups. Two investigators (M.K.D., R.F.W.) independently searched both information sources for CRs on general surgery of one of the following organ systems or body regions: abdominal wall, thyroid gland, stomach, duodenum, small bowel, large bowel (including rectum and anus), and hepatopancreatico-biliary system. We included reviews related to conventional or laparoscopic surgical techniques if compared to another surgical or medical intervention. Reviews on medical therapies including neo-adjuvant, adjuvant, or supportive therapy were included if they were directly related to surgical procedures. We excluded reviews of pediatric surgery. After screening of identified titles, full articles of eligible reviews were retrieved electronically for detailed evaluation. Disagreement on eligibility between the 2 investigators was resolved by discussion with a third (G.A.).

Diener et al 445

Fig 1. Number of abstracts and articles identified and evaluated during the review process.

Data extraction. Two investigators (M.K.D., R.F.W.) independently extracted the following quantitative and qualitative data using standardized forms for data abstraction: field of research within general surgery; number of included studies per review and of included participants in studies; point estimate, confidence intervals (CI), and P values for main outcome measure; number of included randomized and non-randomized studies; type of comparison group; date of first publication and of most recent update in the Cochrane Library, and type of treatment recommendation given by review. To classify the field of research, the following predefined domains were used: (1) abdominal wall/ closure techniques/hernia diseases; (2) colorectal diseases; (3) hepato-biliary disorders; (4) pain/ palliative or supportive care; (5) upper gastrointestinal/pancreatic diseases; and (6) wound care. Main outcome measures were extracted as defined as ‘‘primary objectives’’ by the review authors. In case of multiple outcome measures two authors (M.K.D.; R.F.W.) selected the one deemed most important for surgical practice. In case of discrepant outcome extraction, the selection was discussed with a third reviewer (E.E.). The reviews’ recommendations for treatment were classified as either ‘‘Explicit treatment recommendation’’ or ‘‘Further evidence is needed.’’ For this, we examined the ‘‘implications for practice’’ formulated in the Authors’ conclusion section of the reviews and the quantitative and qualitative amount of evidence

446 Diener et al

available from each review. Two investigators (M.K.D., R.F.W.) appraised the pooled estimates of effectiveness, the amount of included evidence (number of studies, sample size of included primary studies), and the conclusion of the authors drawn from the results. Utilization of included reviews. The log files of the Cochrane Library‘s Homepage (www. thecochranelibrary.com), as provided by the publisher Wiley & Sons, were analyzed in order to calculate estimates of the utilization of included CRs. Included CRs were identified in the top 50 list of Cochrane Reviews accessed between 2005 and 2007. Moreover, the top 20 surgical CRs were identified using the usage count databases of 2007. On the basis of the top 20 surgical CRs, we calculated country rankings. Data analysis. We used univariate and multivariate multiple logistic regression models with presence of a treatment recommendation in the CR as the dependent variable and several characteristics such as review topic, design of primary studies, and number and timeliness of included studies as independent variables. For the purpose of this analysis, the number of studies and of patients included in the review was dichotomized and the reviews grouped using the median number of studies (or patients) of all included reviews. Variables were eliminated from the full model with a threshold of P > .20. We used Stata version 10 (Stata Corporation, Austin, TX) for regression analyses. Results were expressed as odds ratios (ORs) with 95% confidence intervals. Descriptive data are presented as means, corresponding standard deviations (SD), and percentages (%). Where appropriate, we calculated the k-statistics to monitor inter-observer agreement.12 RESULTS Characteristics of included systematic reviews. The search of the Cochrane Library (issue 1/2008) identified 649 potentially eligible review abstracts and protocols. We excluded 265 protocols and 7 withdrawn reviews. The full text of the remaining 377 reviews was evaluated in detail. Of these, 284 were not related to surgical techniques or procedures and were excluded. Finally, 93 reviews were included in the study (Fig 1; all references are listed in Table I).55-97 Of the included reviews, 24 (25.8%) were on hepato-biliary diseases, 21 (22.6%) on colorectal diseases, 20 (21.5%) on postoperative pain and palliative care following general surgery, 14 (15.1%) on wound care, 8 (8.6%) on surgical

Surgery September 2009

techniques for abdominal wall closure and hernia treatment, and 6 (6.5%) on upper gastrointestinal or pancreatic diseases. Details of the comparisons used in the reviews are shown in Tables I and II. Thirty-eight (40.8%) reviews compared different surgical techniques. Pharmacological therapies for diseases within the scope of general surgery were investigated in 23 (24.7%) reviews. Supportive procedures such as bowel preparation,13 wound management,14 antiseptic procedures,14-23 and patient management were evaluated in 20 (21.5%) reviews. Neo-adjuvant and adjuvant treatment regimens for cancer treatment were addressed by 6 (6.5%) reviews: colorectal,24 oesophageus,25,26 hepato-cellular,27,28 and inoperable pancreatic.29 Six (6.5%) reviews addressed surgical strategies for treatment of colorectal cancer. Original studies included in systematic reviews. The included reviews analyzed 1,403 primary studies with a total of 246,473 participants. Of those, 1,173 (84%) studies were randomized controlled trials (RCTs) and 230 (16%) were non-randomized studies (Tables I and II). The mean number of included studies per review was 15.1 (SD 14.5) and the mean number of participants was 2,650 (SD 3,340) (Table II). Eighty CRs (86%) included RCTs only, and 13 (14%) also considered available nonrandomized evidence. Eight CRs (9%) included more nonrandomized than randomized studies. Three reviews were published in 199830-32 and one in 1999.27 These early reviews evaluated pharmacological therapies related to general surgery. All other reviews were published in 2000 or later. Seventy-two of 93 (77%) included reviews were published or updated in 2005 or later. Strengths of treatment recommendation. An explicit treatment recommendation was given in 45 (48.4%) reviews. Forty-eight (52%) reviews stated a need for further evidence and did not include a firm recommendation (Tables I and II). When rating the treatment recommendations given in the reviews, the inter-observer agreement was 90% with a corresponding k-value of 0.74 (95% CI 0.60--0.87). Factors associated with treatment recommendation. In the univariate logistic regression models the probability of an explicit treatment recommendation increased if the topic was Colorectal disease (OR 2.5) or Hepato-biliary disease (OR 1.2) if compared to the reference category (Abdominal wall/hernia diseases/closure technique). It decreased if the topic was Upper GI & pancreatic diseases (OR 0.2), Wounds (0.3) or Pain, Palliative & Supportive Care (OR 0.9). Also, if the review was updated this

Diener et al 447

Surgery Volume 146, Number 3

Table I. Systematic overview of 93 Cochrane Reviews of surgical interventions in the field of general surgery

Topic Abdominal wall/hernia diseases /closure technique

Colorectal diseases

Title

No. of included studies (% non Comparison randomized)

No. of included patients

Summary effect of PE

First published

Implications for practice

(OR/RR/RD /HR/WMD; 95% CI; P value)

(Most recent update)

(k = 0.74)

Antibiotic prophylaxis for hernia repair55

Pharma

12 (0)

6,705

Infection rate 2004 (2007) OR 0.64; 0.48--0.85; NA

FEN

Healing by primary versus secondary intention after surgical treatment for pilonidal sinus56 Laparoscopic techniques versus open techniques for inguinal hernia repair57 Open Mesh versus non-Mesh for groin hernia repair58 Tissue adhesives for closure of surgical incisions59 Transabdominal pre-peritoneal (TAPP) versus totally extraperitoneal (TEP) laparoscopic techniques for inguinal hernia repair60 Transverse verses midline incisions for abdominal surgery61 Wound drains after incisional hernia repair62

Surg

18 (0)

1,573

Rate of surgical site infection RR 1.20; 0.55– 2.63; P = .7

2007 (2007)

ETR

Surg

41 (0)

7,161

Operation time WMD 14.81; 13.98–15.64; P < .01

2000 (2007)

ETR

Surg

22 (0)

5,129

Recurrence OR 0.37; 0.26-0.51; P < .01

2000 (2005)

ETR

Surg

8 (0)

630

2004 (2005)

FEN

Surg

10 (100)

Dehiscence RR 2.48; 0.65-9.40; P = .2 Hernia recurrence RR 2.59; 0.11-60.69; P = .6

2005 (2007)

FEN

Surg

12 (0)

2,445

ETR

Surg

1 (0)

24

Analgesic use 2005 (2006) WMD --22.16; --26.11 to --18.21; P < .01 Infection 2007 (2006) RR 0.11; 0.01– 1.86; P = .1

Antibiotics versus placebo for prevention of postoperative infection after appendectomy63

Pharma

45 (17.7)

19,746

9,576

Wound infection 2001 (2005) OR 0.33; 0.29-0.38; P < .01

FEN

ETR

(continued)

448 Diener et al

Surgery September 2009

Table I. (continued)

Topic

Title Curative surgery for obstruction from primary left colorectal carcinoma: Primary or staged resection?64 Early enteral nutrition within 24 h of colorectal surgery versus later commencement of feeding for postoperative complications65 Follow-up strategies for patients treated for non-metastatic colorectal cancer66 Heparins and mechanical methods for thromboprophylaxis in colorectal surgery67 Hepatic artery adjuvant chemotherapy for patients having resection or ablation of colorectal cancer metastatic to the liver24 Ileostomy or colostomy for temporary decompression of colorectal anastomosis68 Laparoscopic versus open surgery for suspected appendicitis69 Laparoscopic versus open total mesorectal excision for rectal cancer70 Mechanical bowel preparation for elective colorectal surgery13 Operative procedures for fissure in ano71

No. of included studies Comp(% non arison randomized)

No. of included patients

Summary effect of PE

First published

Implications for practice

(OR/RR/RD /HR/WMD; 95% CI; P value)

(Most recent update)

(k = 0.74)

Complications NA

2002 (2005)

FEN

Strat

0 (0)

0

Supp

13 (0)

1,173

Anastomotic leakage RR 0.69; 0.36--1.32; P = .3

2006 (2006)

ETR

Strat

5 (0)

1,342

ETR

Supp

19 (0)

5,091

5 year mortality 2002 (2006) OR 0.73; 0.59--0.91; P < .01 Thromboembolic 2001 (2007) events NA (multiple comparisons)

Adj

7 (0)

592

Survival HR 1.09; 0.89--1.34; P = .4

2004 (2006)

ETR

Surg

5 (0)

334

Mortality RD 0.02; 0.02–0.05; P = .3

2007 (2006)

FEN

Surg

54 (0)

4,953

ETR

Surg

48 (93.8)

4,224

Wound infection 2002 (2004) OR 0.45; 0.35--0.58; P < .01 Survival 2006 (2006) NA

Supp

9 (11.1)

1,592

2003 (2005)

ETR

Surg

24 (41.7)

3,475

Anastomotic leakage OR 2.03; 1.27-3.26; P < .01 Persistence of anal fissure OR 1.28; 0.66-2.48; P = .5

2001 (2005)

FEN

ETR

FEN

(continued)

Diener et al 449

Surgery Volume 146, Number 3

Table I. (continued)

Topic

Title

No. of included studies Comp(% non arison randomized)

Summary effect of PE

First published

Implications for practice

No. of included patients

(OR/RR/RD /HR/WMD; 95% CI; P value)

(Most recent update)

(k = 0.74)

12,127

Recurrence OR 1.68; 1.54--1.83; P < .01 Mortality OR 0.93; 0.87–1.00; P = .04

2006 (2007)

ETR

2007 (2007)

ETR

Clinical anastomotic dehiscence OR 1.55; 0.61--3.95; P = .4 Time to first flatus WMD 0.52; 0.46--0.57; P < .01 Quality of life NA

2004 (2004)

ETR

2005 (2007)

ETR

2004 (2007)

FEN

Relief of symptoms RR 2.50; 0.20--30.61; P = .5 Morbidity RR 0.72; 0.55--0.95; P = .02 Recurrent hemorrhoids OR 3.85; 1.47--10.01; P < .01 Overall anastomotic dehiscence OR 0.99; 0.71--1.40; P=1 Overall anastomotic leak OR 0.34; 0.14– 0.82; P = .02

2005 (2007)

ETR

2005 (2007)

ETR

2006 (2006)

ETR

2001 (2007)

FEN

2007 (2007)

ETR

Perioperative blood transfusions for the recurrence of colorectal cancer72 Pre-operative radiotherapy and curative surgery for the management of localized rectal carcinoma73 Prophylactic anastomotic drainage for colorectal surgery74

Supp

36 (77.8)

Strat

28 (0)

12,97

Surg

6 (0)

1,14

Prophylactic nasogastric decompression after abdominal surgery75

Supp

33 (0)

5,24

Quality of life after rectal resection for cancer, with or without permanent colostomy76 Rubber band ligation versus excisional haemorrhoidectomy for haemorrhoids77

Surg

20 (100)

Surg

3 (0)

332

Short term benefits for laparoscopic colorectal resection78

Surg

25 (0)

3,5

Stapled versus conventional surgery for hemorrhoids79

Surg

12 (0)

1,076

Stapled versus Surg handsewn methods for colorectal anastomosis surgery80

9 (0)

1,233

Stapled versus handsewn methods for ileocolic anastomoses81

6 (0)

955

Surg

2,682

(continued)

450 Diener et al

Surgery September 2009

Table I. (continued)

Topic Hepatobiliary

Title Bile acids for livertransplanted patients82 Cholecystectomy deferral in patients with endoscopic sphincterotomy83 Cholecystectomy versus no cholecystectomy in patients with silent gallstones84 Cyclosporin versus tacrolimus for liver transplanted patients85 Day-case versus overnight stay in laparoscopic cholecystectomy86 Early versus delayed laparoscopic cholecystectomy for acute cholecystitis87 Elective surgery for benign liver tumours88 Endoscopic balloon sphincter dilation (sphincteroplasty) versus sphincterotomy for common bile duct stones89 Interventions for paracetamol (acetaminophen) overdose90 Ischemic preconditioning for liver transplantation91 Laparoscopic versus open cholecystectomy for patients with symptomatic cholecystolithiasis92 Laparoscopic versus small-incision cholecystectomy for patients with symptomatic cholecystolithiasis93

No. of included studies Comp(% non arison randomized)

No. of included patients

Pharma

7 (0)

335

Surg

5 (0)

662

Surg

0 (0)

0

16 (0)

3,813

Strat

5 (0)

429

Strat

5 (0)

451

Surg

0 (0)

0

Surg

15 (0)

1,768

Pharma

59 (83)

6,809

Pharma

Summary effect of PE

First published

Implications for practice

(OR/RR/RD /HR/WMD; 95% CI; P value)

(Most recent update)

(k = 0.74)

Mortality RR 0.85; 0.53-1.36; P = .5 Mortality RR 1.78; 1.15-2.75; P = .01

2005 (2005)

FEN

2007 (2007)

ETR

NA (no trials available)

2007 (2006)

FEN

Mortality RR 0.85; 0.73– 0.99; P = .03

2006 (2006)

ETR

Surgery related morbidity RR 1.26; 0.54– 2.94; P = .6 Bile duct injury OR 0.63; 0.15– 2.70; P = .5

2008 (2007)

ETR

2006 (2007)

FEN

NA 2007 (2007) (no trials available) Successful stone 2006 (2007) removal RR 0.90; 0.84– 0.97; P < .01

FEN

Mortality NA (multiple comparisons)

2002 (2006)

ETR

Mortality OR 0.19; 0.01– 4.21; P = .3

2008 (2007)

FEN

ETR

Strat

3 (0)

162

Surg

38 (0)

2,338

Complications 2006 (2006) RD --0.04; --0.07 to --0.01; P = .01

ETR

Surg

13 (0)

2,337

Complications 2006 (2006) RD --0.01; --0.07-0.05; 0.8

ETR

(continued)

Diener et al 451

Surgery Volume 146, Number 3

Table I. (continued)

Topic

Title Methods of vascular occlusion for elective liver resections94 Neoadjuvant and adjuvant therapy for operable hepatocellular carcinoma27 Portosystemic shunts versus endoscopic therapy for variceal rebleeding in patients with cirrhosis95 Primary closure versus T-tube drainage after 2laparoscopic common bile duct stone exploration96 Radiofrequency thermal ablation versus other interventions for hepatocellular 2carcinoma28 Routine abdominal drainage for 2uncomplicated liver resection97 Routine abdominal drainage for uncomplicated open cholecystectomy98 Small-incision versus open cholecystectomy for patients with symptomatic cholecystolithiasis99 Sphincterotomy for biliary sphincter of Oddi dysfunction100 Surgical versus endoscopic treatment of bile duct stones101 TIPS versus paracentesis for cirrhotic patients with refractory ascites102

No. of included studies Comp(% non arison randomized)

No. of included patients

Summary effect of PE

First published

Implications for practice

(OR/RR/RD /HR/WMD; 95% CI; P value)

(Most recent update)

(k = 0.74)

Mortality OR 0.55; 0.11-2.61; P = .4 Survival NA

2007 (2007)

ETR

1999 (2004)

FEN

Surg

15 (0)

1,189

Adj

8 (0)

548

Surg

22 (0)

1,409

Rebleeding OR 0.42; 0.32-0.54; P < .01

2006 (2006)

ETR

Surg

1 (0)

55

Biliay complication OR 0.38; 0.09-1.64; P = .2

2007 (2006)

FEN

Adj

2 (0)

174

Event-free survival RR 0.48; 0.27– 0.85; P = .01

2002 (2004)

FEN

Surg

6 (0)

651

Mortality OR 1.17; 0.37-3.70; P = .8

2007 (2007)

FEN

Surg

28 (0)

3,659

Mortality OR 0.79; 0.21– 2.97; P = .7

2007 (2007)

ETR

Surg

7 (0)

571

Overall 2006 (2006) complications RD 0.00; –0.06– 0.07; P = .9

ETR

Surg

2 (0)

126

FEN

Surg

13 (0)

1,351

Symptomatic cure 2001 (2003) OR 9.08; 2.97-27.77; P < .01 Stone clearance 2006 (2006) OR 3.26; 2.13-4.97; P < .01

Surg

5 (0)

330

12-months ascites 2004 (2006) re-accumulation OR 0.15; 0.08-0.28; P < .01

ETR

ETR

(continued)

452 Diener et al

Surgery September 2009

Table I. (continued)

Topic Pain and palliative care

Title

No. of included studies Comp(% non arison randomized)

No. of included patients

Summary effect of PE

First published

Implications for practice

(OR/RR/RD /HR/WMD; 95% CI; P value)

(Most recent update)

(k = 0.74)

Epidural local anaesthetics versus opioid-based analgesic regimens for postoperative gastrointestinal paralysis, PONV and pain after abdominal surgery103 Patient controlled opioid analgesia versus conventional opioid analgesia for postoperative pain104 Perioperative ketamine for acute postoperative pain105

Pharma

23 (0)

1,023

Return to GI 2000 (2007) function WMD --37.24; --55.67 to --18.82; P < .01

FEN

Pharma

55 (0)

3,861

Pain scores 2006 (2006) WMD --7.97; --11.72 to --4.21; P < .01

ETR

Pharma

37 (0)

2,24

2006 (2007)

ETR

Preoperative fasting for adults to prevent perioperative complications34 Single dose dextropropoxyphene, alone and with paracetamol (acetaminophen), for postoperative pain31 Single dose dihydrocodeine for acute postoperative pain40 Single dose dipyrone for acute postoperative pain39 Single dose oral celecoxib for postoperative pain41 Single dose oral diclofenac for postoperative pain42 Single dose oral ibuprofen and diclofenac for postoperative pain32

Supp

22 (0)

2,295

Morphine consumption 24 hours WMD --15.96; --19.69 to --12.24; P < .01 NA (multiple comparisons)

2003 (2004)

ETR

Pharma

11 (0)

2,313

50% pain relief OR 3.28; 2.46-4.37; P < .01

1998 (2003)

FEN

Pharma

4 (0)

359

50% pain relief OR 2.05; 1.02– 4.11; P = .04

2000 (2004)

ETR

Pharma

15 (0)

1,46

2001 (2004)

FEN

Pharma

2 (0)

690

2003 (2004)

FEN

Pharma

7 (0)

945

2004 (2004)

ETR

Pharma

39 (0)

8,216

50% pain relief RR 2.32; 1.79-3.00, P < .01 50% pain relief RR 2.32; 1.62-3.34; P < .01 50% pain relief RR 2.89; 1.87-4.48; P < .01 50% pain relief RR 3.47; 3.03-3.98; P < .01

1998 (2004)

ETR

(continued)

Diener et al 453

Surgery Volume 146, Number 3

Table I. (continued)

Topic

Title Single dose oral indometacin for the treatment of acute postoperative pain43 Single dose oral lumiracoxib for postoperative pain106 Single dose oral naproxen and naproxen sodium for acute postoperative pain44 Single dose oral paracetamol (acetaminophen) for postoperative pain30 Single dose oral rofecoxib for postoperative pain45 Single dose oxycodone and oxycodone plus paracetamol (acetominophen) for acute postoperative pain46 Single dose paracetamol (acetaminophen), with and without codeine, for postoperative pain47 Single dose piroxicam for acute postoperative pain48 Surgery for the resolution of symptoms in malignant bowel obstruction in advanced gynaecological and gastrointestinal cancer107 Systemic prokinetic pharmacologic treatment for postoperative adynamic ileus following abdominal surgery in adults108

No. of included studies Comp(% non arison randomized)

No. of included patients

Summary effect of PE

First published

Implications for practice

(OR/RR/RD /HR/WMD; 95% CI; P value)

(Most recent update)

(k = 0.74)

Pharma

1 (0)

94

50% pain relief RR 1.29; 0.85-1.96; P = .2

2004 (2004)

FEN

Pharma

3 (0)

737

2007 (2007)

ETR

Pharma

10 (0)

2,279

Pain relief RR 4.80; 2.91-7.91; P < .01 50% pain relief RR 4.18; 2.93-5.97; P < .01

2004 (2004)

ETR

Pharma

47 (0)

4,186

50% pain relief OR 1.95; 1.27– 2.99; P < .0

2004 (2004)

ETR

Pharma

7 (0)

982

50% pain relief RR 5.12; 3.68-7.14; P < .01

2004 (2004)

FEN

Pharma

7 (0)

1,432

50% pain relief OR 2.24; 1.22-4.50; P = .02

2000 (2004)

FEN

Pharma

50 (0)

10,493

50% pain relief OR 1.99; 1.42-2.79; P < .01

1998 (2004)

ETR

Pharma

3 (0)

548

50% pain relief OR 5.39; 3.28-8.85; P < .01 Resolution of bowel obstruction NA

2000 (2004)

FEN

2000 (2004)

FEN

2008 (2007)

FEN

Surg

25 (100)

2,103

Supp

39 (0)

4,615

NA (multiple comparisons)

(continued)

454 Diener et al

Surgery September 2009

Table I. (continued)

Topic

Upper GI and pancreatic diseases

Wounds

Title

No. of included studies Comp(% non arison randomized)

Ursodeoxycholic acid Pharma and/or antibiotics for prevention of biliary stent occlusion109 Extended versus Surg limited lymph nodes dissection technique for adenocarcinoma of the stomach110 Laparoscopic repair Surg for perforated peptic ulcer disease111

5 (0)

15 (86.7)

2 (0)

No. of included patients 258

917

214

Palliative biliary stents for obstructing pancreatic carcinoma29

Adj

21 (100)

1,454

Preoperative chemotherapy for resectable thoracic esophageal cancer26 Preoperative radiotherapy for esophageal carcinoma25 Surgery for morbid obesity33 Disposable surgical face masks for preventing surgical wound infection in clean surgery15 Double gloving to reduce surgical cross-infection17 Dressings and topical agents for surgical wounds healing by secondary intention16 Preoperative bathing or showering with skin antiseptics to prevent surgical site infection19 Preoperative hair removal to reduce surgical site infection18

Adj

11 (0)

2,051

Adj

6 (0)

1,147

Surg

26 (11.5)

2,402

Supp

2 (0)

1,453

Supp

31 (0)

5,443*

Supp

13 (0)

611

Supp

6 (0)

10,007

Supp

11 (0)

5,238

Summary effect of PE

First published

Implications for practice

(OR/RR/RD /HR/WMD; 95% CI; P value)

(Most recent update)

(k = 0.74)

Duration of stent patency OR 0.58; 0.36--0.93; P = .02 5-year survival OR 0.92; 0.72-1.17; P = .5

2002 (2004)

FEN

2003 (2005)

FEN

Septic abdominal complications OR 0.66; 0.30-1.47; P = .3 Therapeutic success RR 1.0; 0.93--1.08; P=1 Survival HR 0.88; 0.75--1.04; P = .1 Survival OR 0.91; 0.80--1.02; P = .2 Weight loss NA Wound infection NA

2005 (2005)

FEN

2006 (2006)

ETR

2001 (2006)

FEN

2000 (2006)

FEN

2003 (2005)

FEN

2002 (2006)

FEN

Surgical site infection NA Wound healing NA (multiple comparisons) Surgical site infection OR 0.91; 0.80--1.04; P = .2 Wound infection RR 1.59; 0.77--3.27; P = .2

2002 (2006)

FEN

2004 (2005)

FEN

2006 (2007)

ETR

2006 (2006)

FEN

(continued)

Diener et al 455

Surgery Volume 146, Number 3

Table I. (continued)

Topic

Title Preoperative skin antiseptics for preventing surgical wound infections after clean surgery20 Removal of nail polish and finger rings to prevent surgical infection21 Surgical hand antisepsis to reduce surgical site infection112 Systemic antimicrobial prophylaxis for percutaneous endoscopic gastrostomy22 Therapeutic touch for healing acute wounds23 Topical silver for treating infected wounds113 Use of plastic adhesive drapes during surgery for preventing surgical site infection114 Water for wound cleansing14

No. of included studies Comp(% non arison randomized)

No. of included patients

Summary effect of PE

First published

Implications for practice

(OR/RR/RD /HR/WMD; 95% CI; P value)

(Most recent update)

(k = 0.74)

Supp

6 (0)

2,85

Wound infection NA (multiple comparisons)

2004 (2006)

FEN

Supp

1 (0)

102

Bacteria colonies NA

2002 (2005)

FEN

Supp

6 (0)

4,766

NA

2008 (2007)

FEN

Supp

10 (0)

1.100

2006 (2006)

ETR

Supp

4 (0)

121

2004 (2006)

FEN

Supp

3 (0)

847

Peristomal infection OR 0.32; 0.20–0.50; P < .01 Wound healing RR 1.03; 0.12-8.60; P = 1 NA (multiple comparisons)

2007(2006)

FEN

Supp

7 (0)

4,195

Surgical site infection RR 1.23; 1.02–1.48; P = .03 Wound healing RR 1.26; 0.18--8.66; P = .8 Reoperation

2008 (2007)

FEN

2002 (2007)

FEN

2007 (2007)

FEN

Supp

Wound drains following Surg thyroid surgery115

11 (45.5)

3,449

13 (0)

1,646

RR 2.12; 0.77–5.83; P = .1 *We calculated the number of surgeons based on the number of gloves. Adj, Neo- and adjuvant therapy comparision; ETR, explicit treatment recommendation; FEN, further evidence is needed; PE, primary endpoint; Pharma, pharmacological comparision; Strat, comparison of surgical strategies; Supp, comparison of supportive therapy; Surg, comparison of surgical interventions; NA, not applicable.

probability decreased (OR 0.6). The probability of an explicit treatment recommendation increased if more than the median number of original studies were included (OR 4.4), if more than half of included studies were RCTs (OR 2.2), and if overall more than the median number of patients (>1,492) were studied in the included original studies (OR 2.1). Only the association with number of original studies was statistically significant.

In the multivariate (adjusted) model 4 variables were retained. If the reviews included more original studies than the median of 10 studies the odds of an explicit treatment recommendation was about 7.7 times increased (OR 7.7; 95% CI 2.4-24.0; P = .001). It was 6.2 times increased if more than half of included studies were RCTs (OR 6.2; 95% CI 0.9--41.5; P = .06). However, the confidence intervals of these estimates were large (Table III).

456 Diener et al

Surgery September 2009

Table II. Miscellaneous characteristics of included Cochrane Reviews CR characteristics No. included CRs No. included studies RCTs Non-RCTs No. included patients Surgical procedures Pharmacological procedures Supportive procedures Adjuvant/neo-adjuvant therapy Surgical strategy

Overall (%) 93 1,403 1,173 (83.6) 230 (16.4) 246,473 38 23 20 6 6

Mean (SD)

Explicit treatment recommendation (%) 45 (48.4)

15.1 12.6 2.5 2,650

(14.5) (13.2) (8.3) (3,340)

(40.9) (24.7) (21.5) (6.5) (6.5)

20 12 8 2 3

(52.6) (52.2) (40) (33.3) (50)

Table III. Probability of explicit treatment recommendation according to review characteristics

Review topic Abdominal wall/hernia diseases/closure technique Colorectal diseases Hepato-biliary Pain, palliative, and supportive care Upper GI and pancreatic diseases Wounds Review including more studies than median (versus review with less)y More than half of included studies are RCTs (versus less) Updated review versus not updated review Review including more patients than median versus review with lessz

Adjusted OR in final model (95%CI)§

Unadjusted OR in univariate models

P value

1* 2.5 1.2 0.9 0.2 0.3 4.4

— .29 .84 .90 .22 .15 .001

4.0 1.7 0.7 0.2 0.2 7.6

2.2 0.6 2.1

.30 .19 .08

6.2 (0.9–41.5) 0.5 (0.2–1.4) —

1* (0.6–27.6) (0.3–11.6) (0.1–4.8) (0.0–3.2) (0.0–1.7) (2.4–24.0)

P value — .16 .57 .75 .26 .14 .001 .06 .19

*Reference category. yMedian number of included studies: n = 10. zMedian number of included patients: n = 1,432. §Adjusted for the 4 variables for which OR estimates are given.

Utilization of reviews. We identified three of the included CRs in the top-50 accessed CRs (log files of the Cochrane Library’s Homepage [www.the cochranelibrary.com] as provided by the publisher Wiley & Sons Co.) between 2005 and 2007.14,33,34 The ranking of the 20 most accessed surgical CRs during 2007 also revealed only 3 surgical CRs within the top 100 accessed CRs.14,33,34 The most accessed surgical CR was accessed 2,050 times worldwide in 2007 (Table IV).34 In comparison, the most accessed CR overall was accessed 7,518 times in 2007.35 The greatest number of the top 20 surgical CRs were accessed in UK, USA, and Australia, followed by several Western and Eastern European countries. Fewer surgical CRs were accessed in Asian and South American countries (Fig 2).

DISCUSSION Summary of findings. We identified 93 Cochrane Reviews relevant to the field of general surgery and analyzed them for their strength of recommendation, timeliness, and utilization. In the analysis of the included primary studies, we observed broad ranges of effect estimates and correlated confidence intervals. An explicit treatment recommendation was correlated with the number of included primary studies and the proportion of randomized (versus observational) studies. Included CRs were published or updated recently, ie, in 2005 or later, which implies high timeliness. However, the utilization of surgical CRs remains to be augmented. Moreover, surgical CRs are rarely accessed in some Asian,

Diener et al 457

Surgery Volume 146, Number 3

Table IV. Ranking of the 20 most accessed CRs in 2007 Title

Usage count

World rank

1. Preoperative fasting for adults to prevent perioperative complications34* 2. Topical silver for treating infected wounds113 3. Water for wound cleansing14y 4. Patient controlled opioid analgesia versus conventional opioid analgesia for postoperative pain104 5. Preoperative bathing or showering with skin antiseptics to prevent surgical site infection19 6. Preoperative hair removal to reduce surgical site infection18 7. Surgery for morbid obesity33z 8. Early versus delayed laparoscopic cholecystectomy for acute cholecystitis87 9. Interventions for paracetamol (acetaminophen) overdose90 10. Double gloving to reduce surgical cross-infection17 11. Laparoscopic versus open surgery for suspected appendicitis70 12. Preoperative skin antiseptics for preventing surgical wound infections after clean surgery20 13. Mechanical bowel preparation for elective colorectal surgery13 14. Antibiotics versus placebo for prevention of postoperative infection after appendectomy63 15. Laparoscopic techniques versus open techniques for inguinal hernia repair57 16. Dressings and topical agents for surgical wounds healing by secondary intention16 17. Surgical versus endoscopic treatment of bile duct stones101 18. Pre-operative radiotherapy and curative surgery for the management of localized rectal carcinoma73 19. Perioperative ketamine for acute postoperative pain105 20. Laparoscopic versus open total mesorectal excision for rectal cancer70

2,050 1,842 1,815 1,498 1,417 1,194 1,176 1,001 992 960 954 932 908 876 814 779 710 692 661 653

37 57 59 108 129 183 189 261 264 285 290 309 330 363 414 448 525 555 611 626

*Top 50 2005, 2006, 2007. yTop 50 2005, 2006. zTop 50 2007. CR, Cochrane Reviews; SD, standard deviation.

Fig 2. Requested top 20 surgical CRs in 2007.

eastern European, and South American countries, as reflected by our random country sample (Fig 2). Limitations and strengths. We restricted our analysis to systematic reviews conducted within the framework of the Cochrane Collaboration knowing that many other reviews are undertaken and published elsewhere. While this limits the generalizability of our findings, we obtained a more homogeneous sample of reviews with similar methodological standards that allowed an assessment of the ensuing treatment recommendations. We acknowledge that, despite the standards of quality maintained by the Cochrane Review Groups, the included reviews may differ in quality.36,37 A more detailed assessment of the

methodological quality of each review or of the primary studies was not the scope of this study. Further, we may have missed relevant Cochrane Reviews. Since our inclusion criteria were broad and search in the Cochrane Library extensive, we are confident that included reviews reflect those that would be of interest to surgeons. Because almost all included reviews were published or updated in the last 4 years, our sample likely reflects the current state of evidence in general surgery. Our judgements on authors’ conclusions and treatment recommendations were based on several factors and may have been subjective. However, we observed acceptable levels of agreement in all steps that were conducted by two investigators independently (ie, literature search, selection of studies, and extraction of data). About half of the reviews formulated explicit implications for practice. However, we cannot tell whether a firm treatment recommendation voiced in any of the reviews has affected current clinical decision making. In our empirical study, we strove to include and analyze all Cochrane Reviews with potential relevance for general surgery. In consequence, also comparisons of pharmacological therapies, adjuvant, and neo-adjuvant treatment regimens, and supportive care relevant to this field were included.

458 Diener et al

This was motivated by our own experience that the success of surgical oncology treatment depends today on multimodal treatment strategies. Our synopsis could well serve as a basis for evidence syntheses of closely related topics within general surgery that have been studied in individual reviews before.38 Of note, 13 included reviews30-32,39-49 compared the effectiveness of pharmaceutical interventions for treatment of postoperative pain, and 4 interventions for prevention of surgical site infection.15,18,19,20 Findings in context. In about half of included reviews a need for further evidence was claimed. This is not surprising given that shortcomings such as insufficient sample size in the primary studies translate into weaker conclusions if such research data are aggregated later.50 When Dixon et al8 studied meta-analyses on topics of general surgery published in peer-reviewed journals, these reviews generally were of low scientific quality. Interestingly, factors associated with overall low scientific quality included absence of any prior meta-analysis publications by authors and meta-analysis produced by surgical department members without external collaboration. In contrast, the quality of CRs is thought to be high as compared to nonCochrane systematic reviews, because they follow a distinct methodology developed by the Cochrane Collaboration over the last decade. Of note, CRs implement an a-priori protocol, which defines the aim and methods in detail. This protocol is peer-reviewed before the CR is started. Before publication the completed CR is peer-reviewed again and both its methodological quality and interpretation evaluated. Of 3,385 reviews available in the Cochrane Library (issue 1/2008), only 93 (3%) were included. This low proportion may be due to the lack of surgical RCTs that could serve as a primary data source for systematic reviews.11,50,51 Clearly, this lack of suitable studies may deter potential reviewers from choosing a topic in this area, although a systematic review could be instrumental in drawing attention to gaps in the available evidence. We identified several fields in general surgery that are underrepresented in the included reviews. Only few reviews addressed upper GI and pancreatic diseases. Of course, such areas may have been addressed in systematic reviews other than CRs. On average, CRs on general surgery were accessed less often than reviews of other topics, since only three of the identified CRs were found in the 50 most-accessed hit list of requested CRs via the Cochrane Library’s homepage (www.theco chranelibrary.com). The awareness and impact of

Surgery September 2009

surgical CRs needs to be augmented in many countries all over the world, since the majority of the identified articles were accessed from UK, USA, and Australia. However, this analysis fails to distinguish between accessed abstracts, protocols, and full CRs, since only the full articles were counted. Nevertheless, without drawing any firm conclusions, this data might serve as an estimate for ‘‘surgical activities’’ within the Cochrane Library stratified for a random country sample. Implications. The future of surgical clinical research may be, in large, pragmatic RCTs.52 To be scientifically and ethically sound, such trials should be based on systematic reviews of the already available evidence.53 Hence, systematic reviews play an important role in setting the scene for new trials that, in turn, will add to the accumulating evidence in subsequent research syntheses.54 This fundamental role of systematic reviews in the preparation of new experiments, not only in surgery, has to be recognized by funding organizations as well as the researchers themselves. In conclusion, given the relatively small number of existing Cochrane Reviews with focus on general surgery, a structured approach is needed to fill in existing ‘‘white spots’’ in the coverage of clinical questions. If topical questions can be answered by systematic reviews, this may enhance the relevance of evidence-based methods for surgical decision making. One way forward could be a prioritization of clinical questions of high relevance for general surgery, for instance by setting an agenda and creating specific funding schemes, with the aim to stimulate review activity carried out by the scientific surgical associations. Also, the accumulating evidence needs to be used in operating theaters and surgical wards. The ‘‘know-do gap’’ can only be closed if such evidence is applied at the bedside. The present synopsis of systematic reviews can help stimulate future research endeavors in surgery. It also draws attention to the already available research evidence from systematic reviews that can be implemented in surgical practice. REFERENCES 1. Haynes RB, Sackett DL, Gray JM, Cook DJ, Guyatt GH. Transferring evidence from research into practice: 1. The role of clinical care research evidence in clinical decisions. ACP J Club 1996;125:A14-6. 2. Horton R. Surgical research or comic opera: questions, but few answers. Lancet 1996;347:984-5. 3. Clarke M. The Cochrane Collaboration: providing and obtaining the best evidence about the effects of health care. Eval Health Prof 2002;25:8-11. 4. Antes G, Diener MK. The role of systematic reviews in evidence-based healthcare. Chin J Evidence-Based Med 2006; 6:467-70.

Surgery Volume 146, Number 3

5. Cook DJ, Mulrow CD, Haynes RB. Systematic reviews: synthesis of best evidence for clinical decisions. Ann Intern Med 1997;126:376-80. 6. Chalmers TC, Frank CS, Reitman D. Minimizing the three stages of publication bias. JAMA 1990;263:1392-5. 7. Flather MD, Farkouh ME, Pogue JM, Yusuf S. Strengths and limitations of meta-analysis: larger studies may be more reliable. Control Clin Trials 1997;18:568-79. 8. Dixon E, Hameed M, Sutherland F, Cook DJ, Doig C. Evaluating meta-analyses in the general surgical literature: a critical appraisal. Ann Surg 2005;241:450-9. 9. The Cochrane Collaboration. http://www.cochrane.org. 10. The Cochrane Library. http://www.thecochranelibrary.org. 11. Wente MN, Seiler CM, Uhl W, Buchler MW. Perspectives of evidence-based surgery. Dig Surg 2003;20:263-9. 12. McGinn T, Wyer PC, Newman TB, Keiz S, Leipzig R, For GG. Tips for learners of evidence-based medicine: 3. Measures of observer variability (kappa statistic). CMAJ 2004; 171:1369-73. 13. Guenaga K, Atallah AN, Castro AA, Matos DDM, Wille-Jorgensen P. Mechanical bowel preparation for elective colorectal surgery. Cochrane Database Syst Rev 2005;1:CD001544. 14. Fernandez R, Griffiths R, Ussia C. Water for wound cleansing. Cochrane Database Syst Rev 2002;4:CD003861. 15. Lipp A, Edwards P. Disposable surgical face masks for preventing surgical wound infection in clean surgery. Cochrane Database Syst Rev 2002;1:CD002929. 16. Vermeulen H, Ubbink D, Goossens A, DeVos R, Legemate D. Dressings and topical agents for surgical wounds healing by secondary intention. Cochrane Database Syst Rev 2004;1: CD003554. 17. Tanner J, Parkinson H. Double gloving to reduce surgical cross-infection. Cochrane Database Syst Rev 2006;3: CD003087. 18. Tanner J, Woodings D, Moncaster K. Preoperative hair removal to reduce surgical site infection. Cochrane Database Syst Rev 2006;3:CD004122. 19. Webster J, Osborne S. Preoperative bathing or showering with skin antiseptics to prevent surgical site infection. Cochrane Database Syst Rev 2007;2:CD004985. 20. Edwards PS, Lipp A, Holmes A. Preoperative skin antiseptics for preventing surgical wound infections after clean surgery. Cochrane Database Syst Rev 2004;3:CD003949. 21. Arrowsmith VA, Maunder JA, Sargent RJ, Taylor R. Removal of nail polish and finger rings to prevent surgical infection. Cochrane Database Syst Rev 2001;1:CD003325. 22. Lipp A, Lusardi G. Systemic antimicrobial prophylaxis for percutaneous endoscopic gastrostomy. Cochrane Database Syst Rev 2006;1:CD005571. 23. O’Mathuna DP, Ashford RL. Therapeutic touch for healing acute wounds. Cochrane Database Syst Rev 2003;4: CD002467. 24. Nelson R, Freels S. Hepatic artery adjuvant chemotherapy for patients having resection or ablation of colorectal cancer metastatic to the liver. Cochrane Database Syst Rev 2006;4:CD003770. 25. Arnott SJ, Duncan W, Gignoux M, Girling DJ, Hansen HS, Launois B, et al. Preoperative radiotherapy for esophageal carcinoma. Cochrane Database Syst Rev 2005;4:CD001799. 26. Malthaner RA, Collin S, Fenlon D. Preoperative chemotherapy for resectable thoracic esophageal cancer. Cochrane Database Syst Rev 2006;3:CD001556. 27. Chan EY, Chow PH, Tai BC, Machin D, Soo KC. Neoadjuvant and adjuvant therapy for operable hepatocellular carcinoma. Cochrane Database Syst Rev 1999;3:CD001199.

Diener et al 459

28. Galandi D, Antes G. Radiofrequency thermal ablation versus other interventions for hepatocellular carcinoma. Cochrane Database Syst Rev 2004;2:CD003046. 29. Moss AC, Morris E, MacMathuna P. Palliative biliary stents for obstructing pancreatic carcinoma. Cochrane Database Syst Rev 2006;2:CD004200. 30. Barden J, Edwards J, Moore A, McQuay H. Single dose oral paracetamol (acetaminophen) for postoperative pain. Cochrane Database Syst Rev 2004;1:CD004602. 31. Collins SL, Edwards JE, Moore RA, McQuay HJ. Single dose dextropropoxyphene, alone and with paracetamol (acetaminophen), for postoperative pain. Cochrane Database Syst Rev 1999;1:CD001440. 32. Collins SL, Moore RA, McQuay HJ, Wiffen PJ, Edwards JE. Single dose oral ibuprofen and diclofenac for postoperative pain. Cochrane Database Syst Rev 1999;1:CD001548. 33. Colquitt J, Clegg A, Loveman E, Royle P, Sidhu MK. Surgery for morbid obesity. Cochrane Database Syst Rev 2005;4:CD003641. 34. Brady M, Kinn S, Stuart P. Preoperative fasting for adults to prevent perioperative complications. Cochrane Database Syst Rev 2003;4:CD004423. 35. Gillespie LD, Gillespie WJ, Robertson MC, Lamb SE, Cumming RG, Rowe BH. Interventions for preventing falls in elderly people. Cochrane Database Syst Rev 2003;4: CD000340. 36. Olsen O, Middleton P, Ezzo J, Gotzsche PC, Hadhazy V, Herxheimer A, et al. Quality of Cochrane reviews: assessment of sample from 1998. BMJ 2001;323:829-32. 37. Shea B, Moher D, Graham I, Pham B, Tugwell P. A comparison of the quality of Cochrane reviews and systematic reviews published in paper-based journals. Eval Health Prof 2002;25:116-29. 38. Bialy L, Domino FJ, Chang AB, Thomson D, Becker L. The Cochrane Library and chronic cough in children: an umbrella review. Evidence-Based Child Health 2006;1:736-42. 39. Edwards JE, Meseguer F, Faura CC, Moore RA, McQuay HJ. Single dose dipyrone for acute postoperative pain. Cochrane Database Syst Rev 2001;1:CD003227. 40. Edwards JE, McQuay HJ, Moore RA. Single dose dihydrocodeine for acute postoperative pain. Cochrane Database Syst Rev 2000;2:CD002769. 41. Barden J, Edwards JE, McQuay HJ, Moore RA. Single dose oral celecoxib for postoperative pain. Cochrane Database Syst Rev 2003;1:CD004233. 42. Barden J, Edwards J, Moore RA, McQuay HJ. Single dose oral diclofenac for postoperative pain. Cochrane Database Syst Rev 2004;2:CD004768. 43. Mason L, Edwards J, Moore RA, McQuay HJ. Single dose oral indometacin for the treatment of acute postoperative pain. Cochrane Database Syst Rev 2004;4:CD004308. 44. Mason L, Edwards JE, Moore RA, McQuay HJ. Single dose oral naproxen and naproxen sodium for acute postoperative pain. Cochrane Database Syst Rev 2004;4:CD004234. 45. Barden J, Edwards J, Moore RA, McQuay HJ. Single dose oral rofecoxib for postoperative pain. Cochrane Database Syst Rev 2005;1:CD004604. 46. Edwards JE, Moore RA, McQuay HJ. Single dose oxycodone and oxycodone plus paracetamol (acetominophen) for acute postoperative pain. Cochrane Database Syst Rev 2000;2:CD002763. 47. Moore A, Collins S, Carroll D, McQuay HJ, Edwards J. Single dose paracetamol (acetaminophen), with and without codeine, for postoperative pain. Cochrane Database Syst Rev 1998;4:CD001547.

460 Diener et al

48. Edwards JE, Loke YK, Moore RA, McQuay HJ. Single dose piroxicam for acute postoperative pain. Cochrane Database Syst Rev 2000;2:CD002762. 49. Juni P, Altman DG, Egger M. Systematic reviews in health care: Assessing the quality of controlled clinical trials. BMJ 2001;323:42-6. 50. Howes N, Chagla L, Thorpe M, McCulloch P. Surgical practice is evidence based. Br J Surg 1997;84:1220-3. 51. Sauerland S, Seiler CM. Role of systematic reviews and meta-analysis in evidence-based medicine. World J Surg 2005;29:582-7. 52. Peto R, Baigent C. Trials: the next 50 years. Large scale randomised evidence of moderate benefits. BMJ 1998; 317:1170-1. 53. Young C, Horton R. Putting clinical trials into context. Lancet 2005;366:107-8. 54. Yusuf S. Meta-analysis of randomized trials: looking back and looking ahead. Control Clin Trials 1997;18:594-601. 55. Sanchez-Manuel FJ, Seco-Gil JL. Antibiotic prophylaxis for hernia repair. Cochrane Database Syst Rev 2004;4: CD003769. 56. McCallum I, King PM, Bruce J. Healing by primary versus secondary intention after surgical treatment for pilonidal sinus. Cochrane Database Syst Rev 2007;4:CD006213. 57. McCormack K, Scott NW, Go PMNY, Ross S, Grant AM. The EU-Hernia Trialists-Collaboration. Laparoscopic techniques versus open techniques for inguinal hernia repair. Cochrane Database Syst Rev 2003;1:CD001785. 58. Scott N, Go PMNY, Graham P, McCormack K, Ross S, Grant AM. Open mesh versus non-mesh for groin hernia repair. Cochrane Database Syst Rev 2001;3:CD002197. 59. Coulthard P, Worthington H, Esposito M, Van-der-Elst M, Van Waes OJF. Tissue adhesives for closure of surgical incisions. Cochrane Database Syst Rev 2002;3:CD004287. 60. Wake BL, McCormack K, Fraser C, Vale L, Perez J, Grant AM. Transabdominal pre-peritoneal (TAPP) vs totally extraperitoneal (TEP) laparoscopic techniques for inguinal hernia repair. Cochrane Database Syst Rev 2005;1: CD004703. 61. Brown SR, Goodfellow PB. Transverse verses midline incisions for abdominal surgery. Cochrane Database Syst Rev 2005;4:CD005199. 62. Gurusamy KS, Samraj K. Wound drains after incisional hernia repair. Cochrane Database Syst Rev 2007;1:CD005570. 63. Andersen BR, Kallehave FL, Andersen HK. Antibiotics versus placebo for prevention of postoperative infection after appendicectomy. Cochrane Database Syst Rev 2005;3: CD001439. 64. De-Salvo GL, Gava C, Lise M, Pucciarelli S. Curative surgery for obstruction from primary left colorectal carcinoma: Primary or staged resection? Cochrane Database Syst Rev 2004;2:CD002101. 65. Andersen HK, Lewis SJ, Thomas S. Early enteral nutrition within 24h of colorectal surgery versus later commencement of feeding for postoperative complications. Cochrane Database Syst Rev 2006;4:CD004080. 66. Jeffery M, Hickey BE, Hider PN. Follow-up strategies for patients treated for non-metastatic colorectal cancer. Cochrane Database Syst Rev 2007;1:CD002200. 67. Wille-Jorgensen JP, Rasmussen MS, Andersen BR, Borly L. Heparins and mechanical methods for thromboprophylaxis in colorectal surgery. Cochrane Database Syst Rev 2004;1:CD001217. 68. Guenaga KF, Lustosa SA, Saad SS, Saconato H, Matos D. Ileostomy or colostomy for temporary decompression of

Surgery September 2009

69.

70.

71. 72.

73.

74.

75.

76.

77.

78.

79.

80.

81.

82. 83.

84.

85.

86.

87.

88.

89.

colorectal anastomosis. Cochrane Database Syst Rev 2007; 1:CD004647. Sauerland S, Lefering R, Neugebauer EAM. Laparoscopic versus open surgery for suspected appendicitis. Cochrane Database Syst Rev 2004;4:CD001576. Breukink S, Pierie J, Wiggers T. Laparoscopic versus open total mesorectal excision for rectal cancer. Cochrane Database Syst Rev 2006;4:CD005200. Nelson R. Operative procedures for fissure in ano. Cochrane Database Syst Rev 2005;2:CD002199. Amato A, Pescatori M. Perioperative blood transfusions for the recurrence of colorectal cancer. Cochrane Database Syst Rev 2006;1:CD005033. Wong RK, Tandan V, De-Silva S, Figueredo A. Pre-operative radiotherapy and curative surgery for the management of localized rectal carcinoma. Cochrane Database Syst Rev 2007;2:CD002102. Jesus EC, Karliczek A, Matos D, Castro AA, Attallah AN. Prophylactic anastomotic drainage for colorectal surgery. Cochrane Database Syst Rev 2004;2:CD002100. Nelson R, Edwards S, Tse B. Prophylactic nasogastric decompression after abdominal surgery. Cochrane Database Syst Rev 2005;1:CD004929. Pachler J, Wille JP. Quality of life after rectal resection for cancer, with or without permanent colostomy. Cochrane Database Syst Rev 2005;2:CD004323. Shanmugam V, Thaha MA, Rabindranath KS, Campbell KL, Steele RJC, Loudon MA. Rubber band ligation versus excisional haemorrhoidectomy for haemorrhoids. Cochrane Database Syst Rev 2005;1:CD005034. Schwenk W, Haase O, Neudecker J, Mu¨ller JM. Short term benefits for laparoscopic colorectal resection. Cochrane Database Syst Rev 2005;2:CD003145. Jayaraman S, Colquhoun PHD, Malthaner RA. Stapled versus conventional surgery for hemorrhoids. Cochrane Database Syst Rev 2006;4:CD005393. Lustosa SAS, Matos D, Atallah AN, Castro AA. Stapled versus handsewn methods for colorectal anastomosis surgery. Cochrane Database Syst Rev 2001;3:CD003144. Choy PY, Bissett IP, Docherty JG, Parry BR, Merrie AE. Stapled versus handsewn methods for ileocolic anastomoses. Cochrane Database Syst Rev 2007;3:CD004320. Chen W, Gluud C. Bile acids for liver-transplanted patients. Cochrane Database Syst Rev 2005;3:CD005442. McAlister VC, Davenport E, Renouf E. Cholecystectomy deferral in patients with endoscopic sphincterotomy. Cochrane Database Syst Rev 2007;4:CD006233. Gurusamy KS, Samraj K. Cholecystectomy versus no cholecystectomy in patients with silent gallstones. Cochrane Database Syst Rev 2007;1:CD006230. Haddad EM, McAlister VC, Renouf E, Malthaner R, Kjaer MS, Gluud LL. Cyclosporin versus tacrolimus for liver transplanted patients. Cochrane Database Syst Rev 2006;4: CD005161. Gurusamy KS, Junnarkar S, Farouk M, Davidson BR. Daycase versus overnight stay in laparoscopic cholecystectomy. Cochrane Database Syst Rev 2008;1:CD006798. Gurusamy KS, Samraj K. Early versus delayed laparoscopic cholecystectomy for acute cholecystitis. Cochrane Database Syst Rev 2006;4:CD005440. Colli A, Fraquelli M, Massironi S, Colucci A, Paggi S, Conte D. Elective surgery for benign liver tumours. Cochrane Database Syst Rev 2007;1:CD005164. Weinberg BM, Shindy W, Lo S. Endoscopic balloon sphincter dilation (sphincteroplasty) versus sphincterotomy for

Surgery Volume 146, Number 3

90.

91.

92.

93.

94.

95.

96.

97.

98.

99.

100.

101.

102.

common bile duct stones. Cochrane Database Syst Rev 2006;4:CD004890. Brok J, Buckley N, Gluud C. Interventions for paracetamol (acetaminophen) overdose. Cochrane Database Syst Rev 2006;2:CD003328. Gurusamy KS, Kumar Y, Sharma D, Davidson BR. Ischaemic preconditioning for liver transplantation. Cochrane Database Syst Rev 2008;1:CD006315. Keus F, de Jong JAF, Gooszen HG, van Laarhoven CJHM. Laparoscopic versus open cholecystectomy for patients with symptomatic cholecystolithiasis. Cochrane Database Syst Rev 2006;4:CD006231. Keus F, de Jong JAF, Gooszen HG, van Laarhoven CJHM. Laparoscopic versus small-incision cholecystectomy for patients with symptomatic cholecystolithiasis. Cochrane Database Syst Rev 2006;4:CD006229. Gurusamy KS, Kumar Y, Sharma D, Davidson BR. Methods of vascular occlusion for elective liver resections. Cochrane Database Syst Rev 2007;4:CD006409. Khan S, Tudur SC, Williamson P, Williamson P, Sutton R. Portosystemic shunts versus endoscopic therapy for variceal rebleeding in patients with cirrhosis. Cochrane Database Syst Rev 2006;4:CD000553. Gurusamy KS, Samraj K. Primary closure versus T-tube drainage after laparoscopic common bile duct stone exploration. Cochrane Database Syst Rev 2007;1:CD005641. Gurusamy KS, Samraj K, Davidson BR. Routine abdominal drainage for uncomplicated liver resection. Cochrane Database Syst Rev 2007;3:CD006232. Gurusamy KS, Samraj K. Routine abdominal drainage for uncomplicated open cholecystectomy. Cochrane Database Syst Rev 2007;2:CD006003. Keus F, de-Jong JAF, Gooszen HG, Van-Laarhoven CJHM. Small-incision versus open cholecystectomy for patients with symptomatic cholecystolithiasis. Cochrane Database Syst Rev 2006;4:CD004788. Craig AG, Toouli J. Sphincterotomy for biliary sphincter of Oddi dysfunction. Cochrane Database Syst Rev 2001;1: CD001509. Martin DJ, Vernon DR, Toouli J. Surgical versus endoscopic treatment of bile duct stones. Cochrane Database Syst Rev 2006;2:CD003327. Saab S, Nieto JM, Lewis SK, Runyon BA. TIPS versus paracentesis for cirrhotic patients with refractory ascites. Cochrane Database Syst Rev 2006;4:CD004889.

Diener et al 461

103. Jørgensen H, Wetterslev J, Møiniche S, Dahl JB. Epidural local anaesthetics versus opioid-based analgesic regimens for postoperative gastrointestinal paralysis, PONV and pain after abdominal surgery. Cochrane Database Syst Rev 2001;1:CD001893. 104. Hudcova J, McNicol E, Quah C, Lau J, Carr DB. Patient controlled opioid analgesia versus conventional opioid analgesia for postoperative pain. Cochrane Database Syst Rev 2006;4:CD003348. 105. Bell RF, Dahl JB, Moore RA, Kalso E. Perioperative ketamine for acute postoperative pain. Cochrane Database Syst Rev 2006;1:CD004603. 106. Roy YM, Derry S, Moore RA. Single dose oral lumiracoxib for postoperative pain. Cochrane Database Syst Rev 2007;4: CD006865. 107. Feuer DJ, Broadley KE. Surgery for the resolution of symptoms in malignant bowel obstruction in advanced gynaecological and gastrointestinal cancer. Cochrane Database Syst Rev 2000;3:CD002764. 108. Traut U, Brugger L, Kunz R, Pauli-Magnus C, Haug K, Bucher HC, et al. Systemic prokinetic pharmacologic treatment for postoperative adynamic ileus following abdominal surgery in adults. Cochrane Database Syst Rev 2008;1: CD004930. 109. Galandi D, Schwarzer G, Bassler D, Allgaier HP. Ursodeoxycholic acid and/or antibiotics for prevention of biliary stent occlusion. Cochrane Database Syst Rev 2002;3:CD003043. 110. McCulloch P, Nita ME, Kazi H, Gama-Rodrigues J. Extended versus limited lymph nodes dissection technique for adenocarcinoma of the stomach. Cochrane Database Syst Rev 2003;4:CD001964. 111. Sanabria AE, Morales CH, Villegas MI. Laparoscopic repair for perforated peptic ulcer disease. Cochrane Database Syst Rev 2005;4:CD004778. 112. Tanner J, Swarbrook S, Stuart J. Surgical hand antisepsis to reduce surgical site infection. Cochrane Database Syst Rev 2008;1:CD004288. 113. Vermeulen H, van Hattem JM, Storm-Versloot MN Ubbink DT. Topical silver for treating infected wounds. Cochrane Database Syst Rev 2007;1:CD005486. 114. Webster J, Alghamdi AA. Use of plastic adhesive drapes during surgery for preventing surgical site infection. Cochrane Database Syst Rev 2007;4:CD006353. 115. Samraj K, Gurusamy KS. Wound drains following thyroid surgery. Cochrane Database Syst Rev 2007;4:CD006099.