BMJ Open · For peer review only PMX SR protocol (BMJ open) 7 98 improved in patients treated with PMX-HP11.In a multi-center Italian trial, Cruz et al. 99 showed that PMX-HP improved

For peer review only

Polymyxin B-immobilized Hemoperfusion and Mortality in

Critically Ill Patients with Sepsis/Septic Shock: Protocol for

a Systematic Review and Meta-Analysis

Journal: BMJ Open

Manuscript ID bmjopen-2016-012908

Article Type: Protocol

Date Submitted by the Author: 31-May-2016

Complete List of Authors: Fujii, Tomoko; Kyoto University Graduate School of Medicine, Epidemiology and Preventive Service Ganeko, Riki; Kyoto Daigaku Igakubu Fuzoku Byoin, Department of

Surgery Kataoka, Yuki; Hyogo Prefectural Amagasaki General Medical Center, Department of Respiratory Medicine Featherstone, Robin; University of Alberta, Pediatrics; University of Alberta, Alberta Research Centre for Health Evidence Bagshaw, Sean; University of Alberta, Canada Furukawa, Toshi; Kyoto University, Graduate School of Medicine and School of Public Health

<b>Primary Subject Heading</b>:

Intensive care

Secondary Subject Heading: Infectious diseases

Keywords: sepsis, Adult intensive & critical care < INTENSIVE & CRITICAL CARE,

polymyxin, hemoperfusion, meta-analysis, systematic review

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

BMJ Open on M

ay 9, 2021 by guest. Protected by copyright.

http://bmjopen.bm

j.com/

BM

J Open: first published as 10.1136/bm

jopen-2016-012908 on 21 Novem

ber 2016. Dow

nloaded from

http://bmjopen.bmj.com/


PMX SR protocol (BMJ open)

1

Polymyxin B-immobilized Hemoperfusion and Mortality in 1

Critically Ill Patients with Sepsis/Septic Shock: Protocol for a 2

Systematic Review and Meta-Analysis 3

4

Tomoko Fujii, MD1,2; Riki Ganeko, MD3; Yuki Kataoka, MD, MPH4; Robin Featherstone, 5

MLIS5; Sean M Bagshaw, MD, MSc6 Toshi A. Furukawa, MD, PhD7 6

7

Affiliation and addresses of the authors: 8

1Department of Epidemiology and Preventive Service, Kyoto University Graduate 9

School of Medicine, [email protected] 10

2Research Fellow of Japan Society for the Promotion of Science 11

3Department of Surgery, Kyoto University Hospital, [email protected] 12

4Department of Respiratory Medicine, Hyogo Prefectural Amagasaki General Medical 13

Center, [email protected] 14

5Alberta Research Center for Health Evidence (ARCHE), Department of Pediatrics, 15

University of Alberta, [email protected] 16

6Division of Critical Care Medicine, Faculty of Medicine and Dentistry, University of 17

Alberta, [email protected] 18

7Department of Health Promotion and Human Behavior, Kyoto University Graduate 19


Correspondence to: Toshi A. Furukawa, MD PhD. Department of Health Promotion 21

and Human Behavior, Kyoto University Graduate School of Medicine, Yoshida 22

Konoe-cho, Sakyo-Ku, Kyoto 606-8501, Japan. Mail to: [email protected] 23

Phone: +81-75-753-9491, Fax: +81-75-753-4641 24

Page 1 of 28


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on May 9, 2021 by guest. P

rotected by copyright.http://bm

jopen.bmj.com

/B

MJ O

pen: first published as 10.1136/bmjopen-2016-012908 on 21 N

ovember 2016. D

ownloaded from




2

25

Word count: Abstract 275 words, main manuscript (including Introduction, 26

Methods and analysis, Ethics and Dissemination, and Discussion) 3,522 words. 27

28

Page 2 of 28


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960



jopen.bmj.com

/B

MJ O


ovember 2016. D

ownloaded from




3

ABSTRACT 29

Introduction: Polymyxin-B immobilized hemoperfusion (PMX-HP) is a promising 30

adjuvant strategy for the treatment of sepsis or septic shock. PMX-HP therapy works by 31

clearing circulating endotoxin through binding to polymyxin-immobilized fibers during 32

hemoperfusion. Clinical trials have shown PMX-HP therapy is associated with improved 33

hemodynamic profile, oxygenation, and survival. However, clear inferences from prior 34

studies have been largely inconclusive due to limitations in study design (e.g., small, 35

unblinded) and generalizability. We therefore plan to perform an up to date systematic 36

review and evidence synthesis to describe the efficacy, safety and effectiveness of 37

PMX-HP for adult patients with sepsis or septic shock. 38

Methods and analysis: We will search the following databases from 1946 to 2016 39

MEDLINE (Ovid), EMBASE (Ovid), Cochrane Central Register of Controlled Trials 40

(CENTRAL), Health Technology Assessment Database (HTA), Cumulative Index to 41

Nursing and Allied Health Literature (CINAHL), Pubmed, and “Igaku Chuo Zasshi” 42

(ICHUSHI) for randomized controlled trials of PMX-HP in critically ill patients with sepsis 43

or septic shock. There will be no language restrictions in the electronic search for 44

studies. Two reviewers will extract data and appraise the quality of each study 45

independently. The primary outcome will be the pooled risk ratio of 28-day all-cause 46

mortality. Serious adverse events and changes in organ dysfunction scores will also be 47

evaluated. The secondary outcomes will be 90-day all-cause mortality, changes in 48

hemodynamic profile and endotoxin levels, and health services use. 49

Ethics and dissemination: Our systematic review will synthesize the literature on the 50

value of PMX-HP as an adjuvant therapy in sepsis/septic shock to improve 51

patient-centered, physiological and health services outcomes. Research ethics is not 52

required for this review. 53

Page 3 of 28


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960



jopen.bmj.com

/B

MJ O


ovember 2016. D

ownloaded from




4

Registration: PROSPERO--NIHR Prospective Register of Systematic Reviews 54

(CRD42016038356)55

Page 4 of 28


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960



jopen.bmj.com

/B

MJ O


ovember 2016. D

ownloaded from




5

ARTICLE SUMMARY 56

Strengths and limitations of this study: 57

� This is a protocol for a systematic review and evidence synthesis of randomized 58

trials evaluating the impact of PMX-HP on outcomes in sepsis/septic shock. 59

� We have assembled a strong inter-professional team to perform this systematic 60

review that includes clinical content experts, methodology experts and a research 61

librarian experienced in the performance of high-quality systematic reviews. 62

� We have developed a rigorous peer-reviewed high-yield literature search strategy 63

to identify relevant randomized trial for inclusion. 64

� The search results may yield numerous small clinical trials with significant clinical 65

and statistical heterogeneity that may present challenges for interpretation. 66

� This systematic review is an important initial step in the assessment of PMX-HP 67

technology that will inform clinical practice guidelines for recommendations on its 68

utilization as an adjuvant therapy of sepsis/septic shock in the clinical practice 69

guidelines. 70

71

Page 5 of 28


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960



jopen.bmj.com

/B

MJ O


ovember 2016. D

ownloaded from




6

INTRODUCTION 72

Description of the condition 73

Sepsis remains desperately fatal and is a major cause of mortality in the intensive care 74

unit (ICU) worldwide1,2. Many efforts have been made to improve prognosis in this 75

condition but large multi-centered trials of various innovative therapies have failed to 76

demonstrate consistent benefit3. As basic elements of sepsis treatment, including 77

definitive antibiotic therapies, adequate fluids and vasopressors, have not changed for 78

decades4,5, new effective therapies are desperately needed. 79

Description of the intervention 80

Experimental animal models of peritonitis have been used to understand the 81

pathophysiological mechanisms involved in sepsis and septic shock. Among these 82

mechanisms, endotoxin, one of the components on the outer membrane of 83

gram-negative bacteria, is recognized as a potent mediator of the host responses in 84

sepsis. A study measuring endotoxin levels in patients with septic shock found high 85

levels of endotoxin activity were associated with worse clinical outcomes6. 86

Polymyxin (PMX) is a cyclic cationic polypeptide antibiotic with high affinity for endotoxin. 87

A novel strategy whereby PMX is bound and immobilized to polystyrene fibers in a 88

hemoperfusion device was developed in Japan7,8. 89

How the intervention might work 90

The suggested mechanism of action of PMX hemoperfusion (PMX-HP) is to remove 91

circulating endotoxin by adsorption, which limits the host response to infection and the 92

progression of the biological cascade of sepsis. 93

Why it is important to do this review 94

The role of endotoxin in sepsis is well established in the literature9,10. Some clinical trials 95

have shown PMX-HP can improve the physiological profile of patients with abdominal 96

sepsis11,12. In a small pilot study, cardiac index and oxygen delivery index were 97

Page 6 of 28


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960



jopen.bmj.com

/B

MJ O


ovember 2016. D

ownloaded from




7

improved in patients treated with PMX-HP11. In a multi-center Italian trial, Cruz et al. 98

showed that PMX-HP improved blood pressure and decreased dose requirement for 99

vasopressor support12. However, it has not to be determined whether they offer any 100

clinically important benefits and actually improve patient outcomes. Individual studies 101

tend to focus on surrogate outcomes or be underpowered to detect effects on clinically 102

important outcomes13. A number of additional studies have been completed evaluating 103

the efficacy, safety and effectiveness of PMX-HP, including a large multi-centre RCT14,15. 104

Accordingly, we propose to perform an up to date systematic review and evidence 105

synthesis evaluating the impact of PMX-HP as an adjuvant therapy for critically ill 106

patients with sepsis or septic shock on clinical outcomes and health services utilization. 107

108

Objectives 109

To assess efficacy, safety and effectiveness of PMX-HP for adult patients with sepsis or 110

septic shock. 111

112

METHODS AND ANALYSIS 113

Study design 114

We will perform a systematic review and meta-analysis, using the guidelines from the 115

Cochrane Collaboration and Centre for Reviews and Dissemination, and described 116

according to the PRISMA-P guideline (see online supplementary Appendix 1). 117

118

Study registration 119

The systematic review protocol has been registered with the PROSPERO International 120

Prospective Register of Systematic Reviews (http://www.crd.york.ac.uk/prospero), and 121

will be reported using Preferred Reporting Items for systematic review and 122

meta-analysis protocols16. Registration number CRD42016038356. 123

Page 7 of 28


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960



jopen.bmj.com

/B

MJ O


ovember 2016. D

ownloaded from




8

124

Criteria for considering studies for this review 125

Types of studies 126

All relevant randomized controlled trials that investigate the effect of PMX-HP for 127

patients with sepsis or septic shock will be included. We will also include 128

cluster-randomized trials if the intra-cluster correlation coefficient is reported. We will 129

exclude crossover studies because of the nature of sepsis requiring intensive treatment 130

within a short period. We will also exclude quasi-randomized studies. We will include 131

studies reported as full text, those published as abstract only, and unpublished data. 132

Types of participants 133

Adults aged 18 years or older with sepsis or septic shock will be included. Studies that 134

included a minority (less than 10%) of patients under 18, or studies with median or 135

mean of age of patients over 20 will be included. 136

The diagnosis of sepsis or septic shock will be based on the diagnosis with 137

documented or clinically suspected systemic infection. Septic shock will be classically 138

defined as a hypotension resistant to fluid administration requiring norepinephrine or 139

other vasopressor17. Patients who developed sepsis or septic shock after elective 140

surgery will also be included. 141

Participants of any gender and ethnicity and who are treated in any setting will be 142

included. 143

Types of interventions 144

The interventions will be the use of the PMX-HP fiber column for the treatment of 145

sepsis or septic shock. Comparison will be performed between the patients who receive 146

standard treatment plus PMX-HP (PMX-HP group) and the patients with standard 147

treatment only or sham hemoperfusion (Control group). 148

Types of outcome measures 149

Page 8 of 28


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960



jopen.bmj.com

/B

MJ O


ovember 2016. D

ownloaded from




9

Primary outcomes 150

1. 28-day all-cause mortality 151

2. Serious adverse events defined as fever, hypotension, and massive bleeding 152

3. Changes in organ dysfunction scores (Changes in Sequential Organ Failure 153

Assessment (SOFA) score over 72hours after the treatment) 154

Secondary outcomes 155


2. Changes in mean arterial blood pressure over 72hours after the treatment 157

3. Changes in endotoxin levels over 72hours after the treatment 158

4. Duration of vasopressor therapy or vasopressor-free days 159

5. ICU length of stay 160

6. Cost related to health services 161

Search methods for identification of studies 162

PROSPERO was searched for any registered systematic reviews on this topic (26 May 163

2016). 164

Electronic searches 165

The search strategy will be developed in consultation with a research librarian at the 166

Alberta Research Centre for Health Evidence (ARCHE) at the University of Alberta and 167

will be peer reviewed by another research librarian18. Draft of search strategy is 168

available in supplementary materials (Appendix 2).We will obtain all relevant studies 169

irrespective of language or publication status (published, unpublished, in press and in 170

progress) from the following databases: MEDLINE (Ovid), EMBASE (Ovid), Cochrane 171

Central Register of Controlled Trials (CENTRAL), Health Technology Assessment 172

Database (HTA), Cumulative Index to Nursing and Allied Health Literature (CINAHL), 173

Pubmed, and “Igaku Chuo Zasshi (ICHUSHI)”. ICHUSHI is a bibliographic database 174

that was established in 1903 and is being updated by the Japan Medical Abstracts 175

Page 9 of 28


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960



jopen.bmj.com

/B

MJ O


ovember 2016. D

ownloaded from




10

Society, a non-profit and non-governmental body 176

(http://www.jamas.or.jp/about/english.htm). ICHUSHI contains bibliographic citations 177

and abstracts from more than 2,500 biomedical journals and other serial publications 178

published in Japan. We will modify the MEDLINE search strategy for searching all the 179

other databases. For ongoing trials, we will search National Institute of Health Clinical 180

Trials Register (https://clinicaltrials.gov/), and World Health Organization International 181

Clinical Trials Registry Platform (http://www.who.int/ictrp/en/), and University hospital 182

Medical Information Network Clinical Trials Registry (http://www.umin.ac.jp/ctr/). 183

Searching other resources 184

We will hand search citations from all included studies. We will contact authors of the 185

identified trials if necessary to inquire about unpublished studies. We will contact 186

experts in the field and selected industry that develop or license PMX-HP to identify 187

additional unpublished and on-going trials. 188

Data collection and analysis 189

Selection of studies 190

Two review authors (T.F. and R.G.) will independently screen titles and abstracts of all 191

studies we identified by the search and code them as ‘retrieve’ (eligible or potentially 192

eligible/unclear) or ‘do not retrieve’. We will retrieve their full texts and identify studies for 193

inclusion and record reasons for exclusion of the ineligible studies. We will resolve any 194

disagreement through discussion or, if required, we will consult a third person (Y.K.). 195

Agreement between the two review authors in determining study eligibility will be 196

reported as percentage agreement and weighted kappa. We will identify and exclude 197

duplicates of the same study so that each study rather than each report is the unit of 198

interest in the review. 199

Data extraction and management 200

Page 10 of 28


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960



jopen.bmj.com

/B

MJ O


ovember 2016. D

ownloaded from




11

We will use a standardized data extraction form for study characteristics and outcome 201

data that has been piloted on at least one study in the review. Two review authors (T.F. 202

and R.G.) will extract data from the included studies independently. Agreement between 203

the two reviewers concerning the primary outcomes will be reported as percentage 204

agreement and intra-class correlation coefficient. Agreement between the two reviewers 205

in the risk of bias items will be reported as percentage agreement and weighted kappa. 206

Any disagreement will be resolved through discussion, or discussed with a third person 207

(Y.K.) if necessary. We will abstract the following information: study characteristics, 208

patient characteristics, sample size, interventions, comparators, potential biases in the 209

conduct of the trial, outcomes including adverse events, methods of statistical analysis, 210

clinical trial registration, and funding support. 211

Literature search results will be uploaded to Rayyan Software (https://rayyan.qcri.org/), 212

an internet based software program that facilitates collaboration among reviewers to 213

select with titles and abstracts. Candidate full text articles will be uploaded to Dropbox, 214

an internet based storage. Microsoft Excel sheet based on the inclusion and exclusion 215

criteria will be used for full text screening. Prior to the formal screening process, a 216

calibration exercise will be undertaken to pilot and refine the screening sheet. 217

Analyses will be performed using Review Manager Software (V.5.3, Cochrane 218

Collaboration, http://tech.cochrane.org/RevMan) and GRADEpro GDT, 219

http://gradepro.org/. 220

Assessment of risk of bias in included studies 221

At least two reviewers will independently assess the risk of bias of the included studies 222

using the tool described in the Cochrane Handbook for Systematic Reviews of 223

Interventions19. The domains listed in Table 1 will be assessed. 224

Page 11 of 28


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960



jopen.bmj.com

/B

MJ O


ovember 2016. D

ownloaded from




12

The risk of bias, in each domain and overall, will be assessed as shown in Table 2. 225

When inadequate details of randomization and other characteristics of trials are 226

provided, the risk of bias will be classified as unclear unless further information can be 227

obtained by contacting the authors. If the assessors disagree, the final rating will be 228

made by discussion or with the involvement of another member of the review group, if 229

necessary. Treatment fidelity will be assessed as possible sources of performance bias. 230

A study will be classified as being at low risk of attrition bias when data for all 231

randomized patients are available. In the case of dropouts, a study may still be 232

assessed as being at low risk of attrition bias when: 233

・ missing outcome data are few and balanced in numbers across intervention 234

groups, with similar reasons for missing data across groups; 235

・ reasons for missing outcome data are unlikely to be related to true outcome; 236

・ missing data have been imputed using appropriate methods. 237

Whenever possible, study protocols will be retrieved in order to assess the risk of 238

reporting bias. A study will be considered to be at low risk of reporting bias when the 239

study protocol is available and all of the study’s pre-specified outcomes that are of 240

interest in the review have been reported in the pre-specified way. When the study 241

protocol is not available, the study will be classified as being at unclear risk of reporting 242

bias unless reported information is sufficient to make a judgment. 243

Measures of treatment effect 244

Dichotomous outcomes 245

As the measure of treatment effect for dichotomous outcomes, we will use the odds 246

ratio (OR) and its 95% confidence interval (CI) because of its favorable mathematical 247

properties. Graphical displays for meta-analysis performed on ratio scales will use a log 248

scale. 249

Page 12 of 28


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960



jopen.bmj.com

/B

MJ O


ovember 2016. D

ownloaded from




13

Continuous outcomes 250

The unit of measurement to be used in the included studies will be identical. Data will 251

therefore be pooled by calculating the mean difference (MD) with 95% CI. 252

Unit of analysis issues 253

1. Cross-over trials 254

Clinical trials with a cross-over design will be excluded in this review. 255

2. Cluster-randomized trials 256

Cluster randomized trials will be included in this review as long as proper adjustment for 257

the intra-cluster correlation is conducted19. We will reduce the size of each trial to its 258

effective sample size. A common design effect will be assumed across intervention 259

groups. For dichotomous data both the number of participants and the number 260

experiencing the event should be divided by the same design effect. For continuous 261

data only the sample size needs to be reduced; means and standard deviations should 262

remain unchanged. 263

3. Multiple intervention groups 264

Where multiple trial arms are reported in a single trial, arms will be combined as long as 265

they can be regarded as subtypes of the same hemoperfusion therapy. When arms 266

cannot be regarded as if in each of them a different subtype of the same intervention is 267

administered; we will include only the relevant arms. In such a case, a single pair-wise 268

comparison will be used to avoid a unit of analysis error. For dichotomous outcomes, 269

data from a different dosage of the same relevant active intervention arms will be 270

summed into a single arm for comparison. For continuous outcomes, means and 271

standard deviations will be combined19. 272

Dealing with missing data 273

We will try to contact the study authors for all relevant missing data. 274

1. Missing participants 275

Page 13 of 28


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960



jopen.bmj.com

/B

MJ O


ovember 2016. D

ownloaded from




14


The proportion of treatment failure will be calculated using an intention-to-treat (ITT) 277

following the principle ‘once randomized always analyzed’. To this end, all randomized 278

patients for whom outcome data are not available will be assumed as treatment failures. 279

Any assumptions and imputations to handle missing data will be clearly stated, and the 280

effect of imputation will be explored by sensitivity analyses. 281


An available cases analysis will be performed for patients with a final assessment 283

presented by the original authors. 284

2. Missing statistics 285

When only P values or standard error (SE) values are reported, we will calculate 286

standard deviations (SDs) according to Altman20. If none of these values are available, 287

and in the absence of supplementary data after requests to the authors, the SDs will be 288

calculated from CIs, t values or P values19; or they will be imputed from other studies in 289

the meta-analysis according to a validated method21. We will examine the validity of 290

these imputations in a sensitivity analysis. Where actual P values obtained from t tests 291

are reported, the corresponding t value will be obtained from a table of the t distribution. 292

Assessment of heterogeneity 293

We will assess overall heterogeneity by visual inspection of the forest plots. We will 294

assess statistical heterogeneity using the I2 statistic (on a scale of 0% to 100%) and Chi2 295

test. I2 values above 50% will be considered to represent substantial statistical 296

heterogeneity and to be explored further. However the importance of the observed I2 297

depends on the magnitude and direction of treatment effects and the strength of 298

evidence for heterogeneity19. Since the Chi2 test has low power when studies have 299

small sample size or are few in number, we will use P value of 0.10 to determine 300

statistical significance. To provide an indication of the spread of true investigation effects, 301

Page 14 of 28


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960



jopen.bmj.com

/B

MJ O


ovember 2016. D

ownloaded from




15

we will also report between-study variance in a random-effects meta-analysis using Tau2. 302

For absolute measures of effect, an approximate 95% range of underlying effects can 303

be obtained by creating an interval from 1.96×Tau below the random-effects pooled 304

estimate, to 1.96×Tau above it. 305

Assessment of reporting biases 306

We will undertake comprehensive searches of multiple sources (including trial 307

registries), increasing efforts to identify the unpublished materials and including 308

non-English language publications as far as possible, to minimize the impact of 309

reporting biases. We will also try to identify outcome reporting bias in trials by recording 310

all trial outcomes, planned and reported, and noting where outcomes were missing. 311

When we find evidence of missing outcomes, we will attempt to obtain any available 312

data directly from the authors. Where this is not possible, and the missing data are 313

thought to introduce serious bias, the impact of including such studies in the overall 314

assessment of results will be explored by a sensitivity analysis. Funnel plots will be 315

constructed, and visual inspection will be performed to investigate the asymmetry. 316

When ten or more trials are pooled for primary outcomes, tests for funnel plot 317

asymmetry will be used to investigate the potential influence of reporting biases and 318

small-study effects. 319

Data synthesis 320

Meta-analyses will be performed should a sufficient number of studies are found that 321

share study design and measurement of comparators; otherwise, we will describe the 322

results from each studies. Given the likelihood of differing underlining disorders in the 323

population of interest, such as abdominal sepsis and others, a random-effects model will 324

be used in all analyses22. 325

Subgroup analysis and investigation of heterogeneity 326

Page 15 of 28


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960



jopen.bmj.com

/B

MJ O


ovember 2016. D

ownloaded from




16

We expect considerable heterogeneity for the primary outcomes, and we will use 327

subgroup and meta-regression analyses as exploratory tools to explain them. Exploring 328

sources of heterogeneity may result in false positive conclusions through multiple 329

analyses. Because these analyses lack power, they are more likely to result in false 330

negative results. Giving thought to these limitations, we will perform the following a priori 331

subgroup analyses for the participant group and intervention if sufficient detail is present 332

in the eligible studies. 333

1. Participants 334

・ Abdominal sepsis versus sepsis with other etiologies 335

・ Patients with high-level endotoxin activity defined as endotoxin activity assay 336

≥0.6 versus low-level endotoxin activity 337

・ Surgical patients versus medical patients 338

・ Sepsis versus septic shock 339

We will use the formal statistical test for heterogeneity across subgroup based on 340

random-effects model to test for subgroup interaction. 341

2. Intervention 342

・ Single session versus multiple sessions 343

Sensitivity analysis 344

To determine the sensitivity of the findings to the way in which we have conducted the 345

analysis, we will perform sensitivity analyses in the following areas. 346

1. Risk of bias: We will include only trials with low risk of bias in allocation 347

concealment. 348

2. Imputed missing data: We will exclude trials for which missing data are imputed. 349

3. Statistical method of data synthesis: a random-effects model versus fixed-effect 350

model. 351

352

Page 16 of 28


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960



jopen.bmj.com

/B

MJ O


ovember 2016. D

ownloaded from




17

Summary of findings tables 353

We will assess the quality of the body of evidence using the Grading of 354

Recommendations Assessment, Development and Evaluation (GRADE)23. We will 355

present the results of the review in the ‘Summary of findings’ tables including the 356

following outcomes: 28-day all-cause mortality, serious adverse events, changes in 357

organ dysfunction scores, 90-day all-cause mortality, changes in mean arterial blood 358

pressure over 72hours after the treatment, duration of vasopressor therapy or 359

vasopressor-free days, and cost related to health services. 360

Ethics and dissemination 361

The study will evaluate the clinical effect of PMX-HP for patients with sepsis/septic 362

shock on the available published and unpublished clinical trial data. As no primary data 363

is collected, formal ethical approval is not required. The study will be disseminated by 364

peer-reviewed publication and conference presentation. When the protocol needs 365

amendments, the date, the change, and the rationale will be documented in the revised 366

protocol paper. 367

368

DISCUSSION 369

The purpose of our planned systematic review is to determine efficacy, safety, and 370

effectiveness of PMX-HP for adult patients with sepsis or septic shock. The key strength 371

of this protocol is its comprehensive search for relevant studies, including on-going 372

trials14 and unpublished data. No language restriction will be placed and thorough 373

search in the databases in the country where PMX-HP was developed will enable data 374

comprehensive review to update our knowledge. 375

There are several expected limitations for this review. First, we defined sepsis as 376

documented or clinically suspected systemic infection and septic shock as 377

sepsis-induced hypotension persisting despite fluid resuscitation, in line with the 378

Page 17 of 28


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960



jopen.bmj.com

/B

MJ O


ovember 2016. D

ownloaded from




18

well-known definition17. A new definition for sepsis was proposed recently24, and future 379

research on sepsis might use this new definition. To address the issue of variation in 380

participants in interpreting and applying the results, we plan to perform a subgroup 381

analysis for the participants according to endotoxin activity levels. Second, the results of 382

this review will depend on the quality of the studies we identify, which might be low as in 383

the previous review13. We will, therefore, perform sensitivity analyses to see whether the 384

findings will change depending on the risk of bias of each study. Finally, small studies 385

with high heterogeneity could make it difficult to interpret the obtained results. 386

However, we expect the present review to provide the most comprehensive and 387

up-to-date critical summary of available evidence regarding the use of PMX-HP for 388

sepsis or septic shock in the ICU and will help guide treatment recommendations of 389

sepsis or septic shock in the clinical practice guidelines5. 390

391

Acknowledgements 392

We thank Dr. Toshiyuki Suwa at Osaka University for his assistance with search 393

strategy design in Japanese databases, and Tara Landry at the University of Alberta for 394

her peer reviewing the MEDLINE search strategy. 395

396

Funding and competing interests 397

TF is supported by Japan Society for the Promotion of Science. SMB is supported by a 398

Canada Research Chair in Critical Care Nephrology and is a Steering Committee 399

member of the EUPHRATES trial. TAF has received lecture fees from Eli Lilly, Janssen, 400

Meiji, MSD, Otsuka, Pfizer and Tanabe-Mitsubishi, and consultancy fees from Sekisui 401

Chemicals and Takeda Science Foundation. He has received royalties from Igaku-Shoin 402

and Nihon Bunka Kagaku-sha publishers. He has received grant or research support 403

from the Japanese Ministry of Education, Science, and Technology, the Japanese 404

Page 18 of 28


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960



jopen.bmj.com

/B

MJ O


ovember 2016. D

ownloaded from




19

Ministry of Health, Labour and Welfare, the Japan Society for the Promotion of Science, 405

the Japan Foundation for Neuroscience and Mental Health, Mochida and 406

Tanabe-Mitsubishi. He is diplomate of the Academy of Cognitive Therapy. This review 407

will be supported in part by JSPS KAKENHI (Grant-in-Aid for Scientific Research) Grant 408

Number 26670314 to TAF. All the other authors declare that they have no competing 409

interests. None of these funding organizations or any commercial organization have 410

contributed to the study design; collection, management, analysis and interpretation of 411

data; writing of the report or the decision to submit the report for publication. Only the 412

named authors have ultimate authority over these activities. 413

Authors’ Contributions 414

TAF is the guarantor of the protocol. TF and SMB conceived the project. TF and RG 415

worked equally on the initial architecture for the review with methodological input from 416

YK, RF, and SMB. TAF helped draft the article and revise it critically for important 417

intellectual content. All authors gave final approval of the version to be published. 418

419

Page 19 of 28


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960



jopen.bmj.com

/B

MJ O


ovember 2016. D

ownloaded from




20

References 420

1. Kaukonen KM, Bailey M, Suzuki S, Pilcher D, Bellomo R. Mortality related to severe 421

sepsis and septic shock among critically ill patients in Australia and New Zealand, 422

2000-2012. JAMA. 2014;311:1308-16. 423

2. Gaieski DF, Edwards JM, Kallan MJ, Carr BG. Benchmarking the incidence and 424

mortality of severe sepsis in the United States. Crit Care Med. 2013; 41: 1167-74. 425

3. Suffredini AF, Munford RS. Novel therapies for septic shock over the past 4 decades. 426

JAMA. 2011;306:194-9. 427

4. Dellinger RP, Carlet JM, Masur H, Gerlach H, Calandra T, Cohen J, Gea-Banacloche 428

J, Keh D, Marshall JC, Parker MM, Ramsay G, Zimmerman JL, Vincent JL, Levy MM; 429

Surviving Sepsis Campaign Management Guidelines Committee. Surviving Sepsis 430

Campaign guidelines for management of severe sepsis and septic shock. Crit Care 431

Med. 2004;32:858-73. 432

5. Dellinger RP, Levy MM, Rhodes A, Annane D, Gerlach H, Opal SM, Sevransky JE, 433

Sprung CL, Douglas IS, Jaeschke R, Osborn TM, Nunnally ME, Townsend SR, 434

Reinhart K, Kleinpell RM, Angus DC, Deutschman CS, Machado FR, Rubenfeld GD, 435

Webb SA, Beale RJ, Vincent JL, Moreno R; Surviving Sepsis Campaign Guidelines 436

Committee including the Pediatric Subgroup. Surviving sepsis campaign: 437

international guidelines for management of severe sepsis and septic shock: 2012. 438

Crit Care Med. 2013; 41:580-637. 439

6. Marshall JC, Foster D, Vincent JL, Cook DJ, Cohen J, Dellinger RP, Opal S, Abraham 440

E, Brett SJ, Smith T, Mehta S, Derzko A, Romaschin A; MEDIC study. Diagnostic and 441

prognostic implications of endotoxemia in critical illness: results of the MEDIC study. J 442

Infect Dis. 2004;190:527-34. 443

Page 20 of 28


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960



jopen.bmj.com

/B

MJ O


ovember 2016. D

ownloaded from




21

7. Shoji H, Tani T, Hanasawa K, Kodama M. Extracorporeal endotoxin removal by 444

polymyxin B immobilized fiber cartridge: designing and antiendotoxin efficacy in the 445

clinical application. Ther Apher. 1998;2:3-12. 446

8. Shoji H. Extracorporeal endotoxin removal for the treatment of sepsis: endotoxin 447

adsorption cartridge (Toraymyxin). Ther Apher Dial. 2003;7:108-14. 448

9. Danner RL, Elin RJ, Hosseini JM, Wesley RA, Reilly JM, Parillo JE. Endotoxemia in 449

human septic shock. Chest. 1991;99:169-75. 450

10. Casey LC, Balk RA, Bone RC. Plasma cytokine and endotoxin levels 451

correlate with survival in patients with the sepsis syndrome. Ann Intern Med. 452

1993;119:771-8. 453

11. Vincent JL, Laterre PF, Cohen J, Burchardi H, Bruining H, Lerma FA, Wittebole X, 454

De Backer D, Brett S, Marzo D, Nakamura H, John S. A pilot-controlled study of a 455

polymyxin B-immobilized hemoperfusion cartridge in patients with severe sepsis 456

secondary to intra-abdominal infection. Shock. 2005;23:400-5. 457

12. Cruz DN, Antonelli M, Fumagalli R, Foltran F, Brienza N, Donati A, Malcangi V, 458

Petrini F, Volta G, Bobbio Pallavicini FM, Rottoli F, Giunta F, Ronco C. Early use of 459

polymyxin B hemoperfusion in abdominal septic shock: the EUPHAS randomized 460

controlled trial. JAMA. 2009;301:2445-52. 461

13. Cruz DN, Perazella MA, Bellomo R, de Cal M, Polanco N, Corradi V, Lentini P, 462

Nalesso F, Ueno T, Ranieri VM, Ronco C. Effectiveness of polymyxin B-immobilized 463

fiber column in sepsis: a systematic review. Crit Care. 2007;11:R47. 464

14. Klein DJ, Foster D, Schorr CA, Kazempour K, Walker PM, Dellinger RP. The 465

EUPHRATES trial (Evaluating the Use of Polymyxin B Hemoperfusion in a 466

Randomized controlled trial of Adults Treated for Endotoxemia and Septic shock): 467

study protocol for a randomized controlled trial. Trials. 2014;15:218. 468

Page 21 of 28


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960



jopen.bmj.com

/B

MJ O


ovember 2016. D

ownloaded from




22

15. Payen DM, Guilhot J, Launey Y, Lukaszewicz AC, Kaaki M, Veber B, Pottecher J, 469

Joannes-Boyau O, Martin-Lefevre L, Jabaudon M, Mimoz O, Coudroy R, Ferrandière 470

M, Kipnis E, Vela C, Chevallier S, Mallat J, Robert R; ABDOMIX Group. Early use of 471

polymyxin B hemoperfusion in patients with septic shock due to peritonitis: a 472

multicenter randomized control trial. Intensive Care Med. 2015;41:975-84 473

16. Shamseer L, Moher D, Clarke M, Ghersi D, Liberati A, Petticrew M, Shekelle P, 474

Stewart LA; PRISMA-P Group. Preferred reporting items for systematic review and 475

meta-analysis protocols (PRISMA-P) 2015: elaboration and explanation. BMJ. 476

2015;349:g7647. 477

17. Levy MM, Fink MP, Marshall JC, Abraham E, Angus D, Cook D, Cohen J, Opal SM, 478

Vincent JL, Ramsay G, International Sepsis Definitions Conference. 2001 479

SCCM/ESICM/ACCP/ ATS/SIS international sepsis definitions conference. Intensive 480

Care Med. 2003;29:530–538. 481

18. Sampson M, McGowana J, Cogob E, Grimshaw J, Moherd D, Lefebvre C. An 482

evidence-based practice guideline for the peer review of electronic search strategies. 483

J Clin Epidemiol. 2009; 62: 944-952. 484

19. Higgins JPT, Green S (editors). Cochrane Handbook for Systematic Reviews of 485

Interventions. Chichester (UK): John Wiley & Sons, 2008. 486

20. Altman DG, Bland JM. Detecting skewness from summary information. BMJ. 1996; 487

313: 1200. 488

21. Furukawa TA, Barbui C, Cipriani A, Brambilla P, Watanabe N. Imputing missing 489

standard deviations in meta-analyses can provide accurate results. J Clin Epidemiol. 490

2006; 59: 7-10. 491

22. DerSimonian R, Laird N. Meta-analysis in clinical trials. Control Clin Trials. 1986; 7: 492

177-88. 493

Page 22 of 28


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960



jopen.bmj.com

/B

MJ O


ovember 2016. D

ownloaded from




23

23. .Guyatt G, Oxman AD, Akl EA, Kunz R, Vist G, Brozek J, Norris S, Falck-Ytter Y, 494

Glasziou P, DeBeer H, Jaeschke R, Rind D, Meerpohl J, Dahm P, Schünemann HJ. 495

GRADE guidelines: 1. Introduction-GRADE evidence profiles and summary of 496

findings tables. J Clin Epidemiol. 2011;64:383-94. 497

24. Singer M, Deutschman CS, Seymour CW, Shankar-Hari M, Annane D, Bauer M, 498

Bellomo R, Bernard GR, Chiche JD, Coopersmith CM, Hotchkiss RS, Levy MM, 499

Marshall JC, Martin GS, Opal SM, Rubenfeld GD, van der Poll T, Vincent JL, Angus 500

DC. The Third International Consensus Definitions for Sepsis and Septic Shock 501

(Sepsis-3). JAMA. 2016;315:801-10. 502

503

Page 23 of 28


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960



jopen.bmj.com

/B

MJ O


ovember 2016. D

ownloaded from




24

Tables 504

Table 1. Domains for assessment of the risk of bias. 505

1. Random sequence generation (selection bias).

2. Allocation concealment (selection bias).

3. Blinding of participants and personnel (performance bias).

4. Blinding of outcome assessment (assessment bias).

5. Incomplete outcome data reporting (attrition bias).

6. Selective outcome reporting (reporting bias).

7. Treatment fidelity (performance bias).

8. Other biases (Sample size estimation, Co-intervention imbalance, Sponsorship

bias).

506

507

Table 2. Assessment of the risk of bias. 508

The risk of bias Explanation

Low risk Plausible bias unlikely to seriously alter the results

Unclear risk Insufficient information to determine whether the risk of bias is low or high

High risk Plausible bias that seriously weakens confidence in the results

509

510 511

Page 24 of 28


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960



jopen.bmj.com

/B

MJ O


ovember 2016. D

ownloaded from



PRISMA-P (Preferred Reporting Items for Systematic review and Meta-Analysis Protocols) 2015 checklist: recommended items to

address in a systematic review protocol*

Section and topic Item No Checklist item

ADMINISTRATIVE INFORMATION

Title:

Identification 1a Identify the report as a protocol of a systematic review

See page 1, line 2-3

Update 1b If the protocol is for an update of a previous systematic review, identify as such

NOT APPLICABLE

Registration 2 If registered, provide the name of the registry (such as PROSPERO) and registration number


Authors:

Contact 3a Provide name, institutional affiliation, e-mail address of all protocol authors; provide physical mailing address of

corresponding author


Contributions 3b Describe contributions of protocol authors and identify the guarantor of the review


Amendments 4 If the protocol represents an amendment of a previously completed or published protocol, identify as such and list changes;

otherwise, state plan for documenting important protocol amendments


Support:

Sources 5a Indicate sources of financial or other support for the review


Sponsor 5b Provide name for the review funder and/or sponsor


Role of sponsor or funder 5c Describe roles of funder(s), sponsor(s), and/or institution(s), if any, in developing the protocol


INTRODUCTION

Rationale 6 Describe the rationale for the review in the context of what is already known


Objectives 7 Provide an explicit statement of the question(s) the review will address with reference to participants, interventions,

comparators, and outcomes (PICO)

Page 25 of 28


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

on May 9, 2021 by guest. Protected by copyright. http://bmjopen.bmj.com/ BMJ Open: first published as 10.1136/bmjopen-2016-012908 on 21 November 2016. Downloaded from




METHODS

Eligibility criteria 8 Specify the study characteristics (such as PICO, study design, setting, time frame) and report characteristics (such as years

considered, language, publication status) to be used as criteria for eligibility for the review


Information sources 9 Describe all intended information sources (such as electronic databases, contact with study authors, trial registers or other

grey literature sources) with planned dates of coverage


Search strategy 10 Present draft of search strategy to be used for at least one electronic database, including planned limits, such that it could be

repeated

See Appendix 2

Study records:

Data management 11a Describe the mechanism(s) that will be used to manage records and data throughout the review


Selection process 11b State the process that will be used for selecting studies (such as two independent reviewers) through each phase of the

review (that is, screening, eligibility and inclusion in meta-analysis)


Data collection process 11c Describe planned method of extracting data from reports (such as piloting forms, done independently, in duplicate), any

processes for obtaining and confirming data from investigators


Data items 12 List and define all variables for which data will be sought (such as PICO items, funding sources), any pre-planned data

assumptions and simplifications


Outcomes and prioritization 13 List and define all outcomes for which data will be sought, including prioritization of main and additional outcomes, with

rationale


Risk of bias in individual studies 14 Describe anticipated methods for assessing risk of bias of individual studies, including whether this will be done at the

outcome or study level, or both; state how this information will be used in data synthesis


Page 26 of 28


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960




Data synthesis 15a Describe criteria under which study data will be quantitatively synthesised

See page 15, 320-325

15b If data are appropriate for quantitative synthesis, describe planned summary measures, methods of handling data and

methods of combining data from studies, including any planned exploration of consistency (such as I2, Kendall’s τ)


15c Describe any proposed additional analyses (such as sensitivity or subgroup analyses, meta-regression)


15d If quantitative synthesis is not appropriate, describe the type of summary planned

See page 15, 322-323

Meta-bias(es) 16 Specify any planned assessment of meta-bias(es) (such as publication bias across studies, selective reporting within studies)


Confidence in cumulative evidence 17 Describe how the strength of the body of evidence will be assessed (such as GRADE)


* It is strongly recommended that this checklist be read in conjunction with the PRISMA-P Explanation and Elaboration (cite when available) for important

clarification on the items. Amendments to a review protocol should be tracked and dated. The copyright for PRISMA-P (including checklist) is held by the

PRISMA-P Group and is distributed under a Creative Commons Attribution Licence 4.0.

From: Shamseer L, Moher D, Clarke M, Ghersi D, Liberati A, Petticrew M, Shekelle P, Stewart L, PRISMA-P Group. Preferred reporting items for systematic review and

meta-analysis protocols (PRISMA-P) 2015: elaboration and explanation. BMJ. 2015 Jan 2;349(jan02 1):g7647.

Page 27 of 28


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960




Supplementary File: Sample Database Search Strategy

Database: Epub Ahead of Print, In-Process & Other Non-Indexed Citations, Ovid MEDLINE(R) Daily and Ovid MEDLINE(R) 1946 to Present 1. exp Sepsis/ 2. Sepsis-Associated Encephalopathy/ 3. Systemic Inflammatory Response Syndrome/ 4. Vasoplegia/ 5. bacill?emia*.tw,kf. 6. bacter* shock.tw,kf. 7. bacter?emia*.tw,kf. 8. (blood adj2 poison*).tw,kf. 9. candid?emia*.tw,kf. 10. endotox?emia*.tw,kf. 11. endotoxi* shock.tw,kf. 12. fung?emia*.tw,kf. 13. parasit?emia*.tw,kf. 14. (py?emia* or pyohemia*).tw,kf. 15. sepsis.tw,kf. 16. septic.tw,kf. 17. septic?emia*.tw,kf. 18. SIRS.tw,kf. 19. systemic inflammatory response syndrome*.tw,kf. 20. toxic shock.tw,kf. 21. vasopl?egia*.tw,kf. 22. vi?emia*.tw,kf. 23. or/1-22 24. Polymyxin B/ 25. aerosporin.tw,kf,nm. 26. PMX*.tw,kf,nm. 27. polymyxin*.tw,kf,nm. or 1404-26-8.rn. 28. Poly RX.tw,kf,nm. 29. toraymyxin*.tw,kf,nm. 30. or/24-29

31. Blood Component Removal/ 32. Endotoxins/bl [Blood] 33. exp Hemofiltration/ 34. Hemoperfusion/ 35. apheres?s.tw,kf. 36. blood component removal*.tw,kf. 37. DHP-PMX.tw,kf. 38. (endotoxin* adj3 (a?sor* or eliminat* or remov*)).tw,kf. 39. h?emadsor*.tw,kf. 40. (h?emo-filtrat* or h?emofiltrat*).tw,kf. 41. (h?emo-diafiltrat* or h?emodiafiltrat*).tw,kf. 42. (h?emo-dialysis or h?emodialysis).tw,kf. 43. (h?emo-perfus* or h?emoperfus*).tw,kf. 44. (h?emo-sor* or h?emosor*).tw,kf. 45. PMX-?HP*.tw,kf. 46. pheres?s.tw,kf. 47. or/31-46 48. and/30,47 49. and/23,48 50. controlled clinical trial.pt. 51. randomized controlled trial.pt. 52. drug therapy.fs. 53. groups.ab. 54. placebo.ab. 55. random*.ab. 56. trial.ab. 57. or/50-56 58. exp animals/ not humans.sh. 59. 57 not 58 60. and/49,59

Page 28 of 28


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960



jopen.bmj.com

/B

MJ O


ovember 2016. D

ownloaded from

Tomoko Fujii

Appendix 2. Sample search strategy for Ovid MEDLINE(R)

Tomoko Fujii



Polymyxin B-immobilized Hemoperfusion and Mortality in

Critically Ill Patients with Sepsis/Septic Shock: Protocol for

a Systematic Review and Meta-Analysis

Journal: BMJ Open

Manuscript ID bmjopen-2016-012908.R1

Article Type: Protocol

Date Submitted by the Author: 05-Sep-2016

Complete List of Authors: Fujii, Tomoko; Kyoto University Graduate School of Medicine, Epidemiology and Preventive Service Ganeko, Riki; Kyoto Daigaku Igakubu Fuzoku Byoin, Department of

Surgery Kataoka, Yuki; Hyogo Prefectural Amagasaki General Medical Center, Department of Respiratory Medicine Featherstone, Robin; University of Alberta, Pediatrics; University of Alberta, Alberta Research Centre for Health Evidence Bagshaw, Sean; University of Alberta, Canada Furukawa, Toshi; Kyoto University, Graduate School of Medicine and School of Public Health

<b>Primary Subject Heading</b>:

Intensive care

Secondary Subject Heading: Infectious diseases

Keywords: sepsis, Adult intensive & critical care < INTENSIVE & CRITICAL CARE,

polymyxin, hemoperfusion, meta-analysis, systematic review


BMJ Open on M

ay 9, 2021 by guest. Protected by copyright.

http://bmjopen.bm

j.com/

BM

J Open: first published as 10.1136/bm

jopen-2016-012908 on 21 Novem

ber 2016. Dow

nloaded from




1

Polymyxin B-immobilized Hemoperfusion and Mortality in 1

Critically Ill Patients with Sepsis/Septic Shock: Protocol for a 2

Systematic Review and Meta-Analysis 3

4

Tomoko Fujii, MD1,2; Riki Ganeko, MD3; Yuki Kataoka, MD, MPH4; Robin Featherstone, 5

MLIS5; Sean M Bagshaw, MD, MSc6 Toshi A. Furukawa, MD, PhD7 6

7

Affiliation and addresses of the authors: 8

1Department of Epidemiology and Preventive Service, Kyoto University Graduate 9


2Research Fellow of Japan Society for the Promotion of Science 11

3Department of Surgery, Kyoto University Hospital, [email protected] 12

4Department of Respiratory Medicine, Hyogo Prefectural Amagasaki General Medical 13

Center, [email protected] 14

5Alberta Research Center for Health Evidence (ARCHE), Department of Pediatrics, 15

University of Alberta, [email protected] 16

6Department of Critical Care Medicine, Faculty of Medicine and Dentistry, University of 17

Alberta, [email protected] 18

7Department of Health Promotion and Human Behavior, Kyoto University Graduate 19


21

Correspondence to: Toshi A. Furukawa, MD PhD. Department of Health Promotion 22

and Human Behavior, Kyoto University Graduate School of Medicine, Yoshida 23

Konoe-cho, Sakyo-Ku, Kyoto 606-8501, Japan. Mail to: [email protected] 24

Page 1 of 31


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960



jopen.bmj.com

/B

MJ O


ovember 2016. D

ownloaded from




2

Phone: +81-75-753-9491, Fax: +81-75-753-4641 25

26

Word count: Abstract 284 words, main manuscript (including Introduction, 27

Methods and analysis, Ethics and Dissemination, and Discussion) 3,633 words. 28

29

Page 2 of 31


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960



jopen.bmj.com

/B

MJ O


ovember 2016. D

ownloaded from




3

ABSTRACT 30

Introduction: Polymyxin-B immobilized hemoperfusion (PMX-HP) is a promising 31

adjuvant strategy for the treatment of sepsis and septic shock. PMX-HP therapy works 32

by clearing circulating endotoxin through binding to polymyxin-immobilized fibers during 33

hemoperfusion. Small clinical trials have shown PMX-HP therapy is associated with 34

improved hemodynamic profile, oxygenation, and survival. However, clear inferences 35

have been largely inconclusive due to limitations in study design (e.g., small, unblinded) 36

and generalizability. We therefore propose to perform an up to date systematic review 37

and evidence synthesis to describe the efficacy, safety and effectiveness of PMX-HP for 38

adult patients with sepsis or septic shock. 39

Methods and analysis: We will search the following databases from 1946 to 2016 40

MEDLINE (Ovid), EMBASE (Ovid), Cochrane Central Register of Controlled Trials 41

(CENTRAL), Health Technology Assessment Database (HTA), Cumulative Index to 42

Nursing and Allied Health Literature (CINAHL), Pubmed, and “Igaku Chuo Zasshi” 43

(ICHUSHI) for randomized controlled trials of PMX-HP in critically ill patients with sepsis 44

or septic shock. There will be no language restrictions in the electronic search for 45

studies. Two reviewers will extract data and appraise the quality of each study 46

independently. The primary outcome will be the pooled risk ratio of 28-day all-cause 47

mortality. Serious adverse events and changes in organ dysfunction scores will also be 48

evaluated. The secondary outcomes will be 90-day all-cause mortality, changes in 49

hemodynamic profile and endotoxin levels, and health services use. 50

Ethics and dissemination: Our systematic review will synthesize the evidence on use 51

of the PMX-HP as an adjuvant therapy in sepsis/septic shock to improve 52

patient-centered, physiological and health services outcomes. Research ethics is not 53

required for this review. The study will be disseminated by peer-reviewed publication 54

and conference presentation. 55

Page 3 of 31


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960



jopen.bmj.com

/B

MJ O


ovember 2016. D

ownloaded from




4

Registration: PROSPERO--NIHR Prospective Register of Systematic Reviews 56

(CRD42016038356) 57

58

Page 4 of 31


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960



jopen.bmj.com

/B

MJ O


ovember 2016. D

ownloaded from




5

ARTICLE SUMMARY 59

Strengths and limitations of this study: 60

� This is a protocol for a systematic review and evidence synthesis of randomized 61

trials evaluating the impact of PMX-HP on outcomes in sepsis/septic shock. 62

� We have assembled a strong inter-professional team to perform this systematic 63

review that includes clinical content experts, methodology experts and a research 64

librarian experienced in the performance of high-quality systematic reviews. 65

� We have developed a rigorous peer-reviewed high-yield literature search strategy 66

to identify relevant randomized trial for inclusion. 67

� The search results may yield numerous small clinical trials with significant clinical 68

and statistical heterogeneity that may present challenges for interpretation. 69

� This systematic review is an important initial step in the assessment of PMX-HP 70

technology that will inform clinical practice guidelines for recommendations on its 71

utilization as an adjuvant therapy of sepsis/septic shock in the clinical practice 72

guidelines. 73

74

Page 5 of 31


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960



jopen.bmj.com

/B

MJ O


ovember 2016. D

ownloaded from




6

INTRODUCTION 75

Description of the condition 76

Sepsis remains desperately fatal and is a major cause of mortality in the intensive care 77

unit (ICU) settings worldwide1,2. Many efforts have been made to improve prognosis in 78

this condition but large multi-centered trials of various innovative therapies have failed 79

to demonstrate consistent benefit3. As basic elements of sepsis treatment, including 80

definitive antibiotic therapy, adequate fluids and vasopressors, have not changed for 81

decades4,5, new effective therapies are desperately needed. 82

83

Description of the intervention 84

Experimental animal models of peritonitis have been used to understand the 85

pathophysiological mechanisms involved in sepsis and septic shock. Among these 86

mechanisms, endotoxin, one of the components on the outer membrane of 87

gram-negative bacteria, is recognized as a potent mediator of the host response in 88

sepsis. A study measuring endotoxin levels in patients with septic shock found high 89

levels of endotoxin activity were associated with worse clinical outcomes6. 90

Polymyxin (PMX) is a cyclic cationic polypeptide antibiotic with high affinity for endotoxin. 91

A novel strategy whereby PMX is bound and immobilized to polystyrene fibers in a 92

hemoperfusion device was developed in Japan7,8. 93

94

How the intervention might work 95

The suggested mechanism of action of PMX hemoperfusion (PMX-HP) is to remove 96

circulating endotoxin by adsorption, which limits and disrupts the maladaptive host 97

response to infection and the progression of the biological cascade of sepsis. 98

99

Why it is important to do this review 100

Page 6 of 31


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960



jopen.bmj.com

/B

MJ O


ovember 2016. D

ownloaded from




7

The role of endotoxin in sepsis is well established in the literature9,10. Some clinical trials 101

have shown PMX-HP can improve the physiological profile of patients with abdominal 102

sepsis11,12. In a small pilot study, cardiac index and oxygen delivery index were 103

improved in patients treated with PMX-HP11. In a multi-center Italian trial, Cruz et al. 104

showed that PMX-HP improved blood pressure and decreased dose requirement for 105

vasopressor support12. However, it has not to be determined whether they offer any 106

clinically important benefits and actually translate into improved outcomes for patients. 107

Individual studies tend to focus on surrogate outcomes or be underpowered to detect 108

effects on clinically important outcomes13. A number of additional studies have been 109

completed evaluating the efficacy, safety and effectiveness of PMX-HP, including a 110

larger multi-centre RCT14,15. Accordingly, we propose to perform an up to date 111

systematic review and evidence synthesis evaluating the impact of PMX-HP as an 112

adjuvant therapy for critically ill patients with sepsis or septic shock on clinical outcomes 113

and health services utilization. 114

115

Objectives 116

To assess efficacy, safety and effectiveness of PMX-HP for adult patients with sepsis or 117

septic shock. 118

119

METHODS AND ANALYSIS 120

Study design 121

We will perform a systematic review and meta-analysis, using the guidelines from the 122

Cochrane Collaboration and Centre for Reviews and Dissemination, and described 123

according to the PRISMA-P guideline (see online supplementary Appendix 1). 124

125

Study registration 126

Page 7 of 31


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960



jopen.bmj.com

/B

MJ O


ovember 2016. D

ownloaded from




8

The systematic review protocol has been registered with the PROSPERO International 127

Prospective Register of Systematic Reviews (http://www.crd.york.ac.uk/prospero), and 128

will be reported using Preferred Reporting Items for systematic review and 129

meta-analysis protocols16. Registration number CRD42016038356. 130

131

Criteria for considering studies for this review 132

Types of studies 133

All relevant randomized controlled trials that investigate the effect of PMX-HP for 134

patients with sepsis or septic shock will be included. We will also include 135

cluster-randomized trials if the intra-cluster correlation coefficient is reported. We will 136

exclude crossover studies because of the nature of sepsis requiring intensive treatment 137

within a short period. We will also exclude quasi-randomized studies. We will include 138

studies reported as full text, those published as abstract only, and unpublished data. 139

140

Types of participants 141

Adults aged 18 years or older with sepsis or septic shock will be included. Studies that 142

included a minority (less than 10%) of patients under 18, or studies with median or 143

mean of age of patients over 20 will be included. 144

The diagnosis of sepsis or septic shock will be based on the diagnosis with 145

documented or clinically suspected systemic infection. Septic shock will be classically 146

defined as a hypotension resistant to fluid administration requiring norepinephrine or 147

other vasopressor17. Patients who developed sepsis or septic shock after surgery will 148

also be included. 149

Participants of any gender and ethnicity and who are treated in any setting will be 150

included. 151

152

Page 8 of 31


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960



jopen.bmj.com

/B

MJ O


ovember 2016. D

ownloaded from




9

Types of interventions 153

The interventions will be the use of the PMX-HP fiber column for the treatment of 154

sepsis or septic shock. Comparison will be performed between the patients who receive 155

standard treatment plus PMX-HP (PMX-HP group) and the patients with standard 156

treatment only or sham hemoperfusion (Control group). 157

158

Types of outcome measures 159

Primary outcomes: 160


2. Serious adverse events, defined as hypotension or massive bleeding 162

3. Changes in organ dysfunction scores (Changes in Sequential Organ Failure 163

Assessment (SOFA) score over 72 hours after the treatment) 164

165

Secondary outcomes: 166


2. Changes in mean arterial blood pressure over 72 hours after the treatment 168

3. Changes in endotoxin levels over 72 hours after the treatment 169

4. Duration of vasopressor therapy or vasopressor-free days 170

5. Duration of RRT 171

6. ICU length of stay 172

7. Cost related to health services 173

174

Search methods for identification of studies 175

PROSPERO was searched for any registered systematic reviews on this topic (26 May 176

2016). 177

Electronic searches 178

Page 9 of 31


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960



jopen.bmj.com

/B

MJ O


ovember 2016. D

ownloaded from




10

The search strategy was developed in consultation with a research librarian at the 179

Alberta Research Centre for Health Evidence (ARCHE) at the University of Alberta and 180

underwent peer review by another research librarian18. Draft of search strategy is 181

available in supplementary materials (Appendix 2). We will obtain all relevant studies 182

irrespective of language or publication status (published, unpublished, in press and in 183

progress) from the following databases: MEDLINE (Ovid), EMBASE (Ovid), Cochrane 184

Library, Health Technology Assessment Database (HTA), Cumulative Index to Nursing 185

and Allied Health Literature (CINAHL), Pubmed, and “Igaku Chuo Zasshi (ICHUSHI)”. 186

ICHUSHI is a bibliographic database that was established in 1903 and is being updated 187

by the Japan Medical Abstracts Society, a non-profit and non-governmental body 188

(http://www.jamas.or.jp/about/english.htm). ICHUSHI contains bibliographic citations 189

and abstracts from more than 2,500 biomedical journals and other serial publications 190

published in Japan. We will modify the MEDLINE search strategy for searching all the 191

other databases. For ongoing trials, we will search National Institute of Health Clinical 192

Trials Register (https://clinicaltrials.gov/), and World Health Organization International 193

Clinical Trials Registry Platform (http://www.who.int/ictrp/en/), and University Hospital 194

Medical Information Network Clinical Trials Registry (http://www.umin.ac.jp/ctr/). 195

196

Searching other resources 197

We will hand search citations from all included studies. We will contact authors of the 198

identified trials if necessary to inquire about unpublished studies. We will contact 199

experts in the field and selected industries that develop or license PMX-HP to identify 200

additional unpublished and on-going trials. If necessary, we will also search conference 201

proceedings and manufacturer websites. 202

203

Data collection and analysis 204

Page 10 of 31


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960



jopen.bmj.com

/B

MJ O


ovember 2016. D

ownloaded from




11

Selection of studies 205

Two review authors (T.F. and R.G.) will independently screen titles and abstracts of all 206

studies we identified by the search and code them as ‘retrieve’ (eligible or potentially 207

eligible/unclear) or ‘do not retrieve’. We will retrieve their full texts and identify studies for 208

inclusion and record reasons for exclusion of the ineligible studies. We will resolve any 209

disagreement through discussion or, if required, we will consult a third person (Y.K. or 210

S.M.B.). Agreement between the two review authors in determining study eligibility will 211

be reported as percentage agreement and weighted kappa. We will identify and exclude 212

duplicates of the same study so that each study rather than each report is the unit of 213

interest in the review. 214

215

Data extraction and management 216

We will use a standardized data extraction form for study characteristics and outcome 217

data that has been piloted on at least one study in the review. Two review authors (T.F. 218

and R.G.) will extract data from the included studies independently. Agreement between 219

the two reviewers concerning the primary outcomes will be reported as percentage 220

agreement and intra-class correlation coefficient. Agreement between the two reviewers 221

in the risk of bias items will be reported as percentage agreement and weighted kappa. 222

Any disagreement will be resolved through discussion, or discussed with a third person 223

(Y.K. or S.M.B.) if necessary. We will abstract the following information: study 224

characteristics, patient characteristics, sample size, interventions, comparators, 225

potential biases in the conduct of the trial, outcomes including adverse events, methods 226

of statistical analysis, clinical trial registration, and funding support. 227

Literature search results will be uploaded to Rayyan Software (https://rayyan.qcri.org/), 228

an internet based software program that facilitates collaboration among reviewers to 229

Page 11 of 31


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960



jopen.bmj.com

/B

MJ O


ovember 2016. D

ownloaded from




12

select with titles and abstracts. Prior to the formal screening process, a calibration 230

exercise will be undertaken to pilot and refine the screening sheet. Analyses will be 231

performed using Review Manager Software (V.5.3, Cochrane Collaboration, 232

http://tech.cochrane.org/RevMan) and GRADEpro GDT, http://gradepro.org/. 233

234

Assessment of risk of bias in included studies 235

At least two reviewers will independently assess the risk of bias of the included studies 236

using the tool described in the Cochrane Handbook for Systematic Reviews of 237

Interventions19. The domains listed in Table 1 will be assessed. 238

The risk of bias, in each domain and overall, will be assessed as shown in Table 2. 239

When inadequate details of randomization and other characteristics of trials are 240

provided, the risk of bias will be classified as unclear unless further information can be 241

obtained by contacting the authors. If the assessors disagree, the final rating will be 242

made by discussion or with the involvement of another member of the review group, if 243

necessary. A study will be classified as being at low risk of attrition bias when data for all 244

randomized patients are available. In the case of dropouts, a study may still be 245

assessed as being at low risk of attrition bias when: 246

・ missing outcome data are few and balanced in numbers across intervention groups, 247

with similar reasons for missing data across groups; 248

・ reasons for missing outcome data are unlikely to be related to true outcome; 249

・ missing data have been imputed using appropriate methods. 250

Whenever possible, study protocols will be retrieved in order to assess the risk of 251

reporting bias. A study will be considered to be at low risk of reporting bias when the 252

study protocol is available and all of the study’s pre-specified outcomes that are of 253

interest in the review have been reported in the pre-specified way. When the study 254

Page 12 of 31


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960



jopen.bmj.com

/B

MJ O


ovember 2016. D

ownloaded from




13

protocol is not available, the study will be classified as being at unclear risk of reporting 255

bias unless reported information is sufficient to make a judgment. 256

257

Measures of treatment effect 258


As the measure of treatment effect for dichotomous outcomes, we will use the odds 260

ratio (OR) and its 95% confidence interval (CI) because of its favorable mathematical 261

properties. Graphical displays for meta-analysis performed on ratio scales will use a log 262

scale. 263

264


The unit of measurement to be used in the included studies will be identical. Data will 266

therefore be pooled by calculating the mean difference (MD) with 95% CI. 267

268

Unit of analysis issues 269

1. Cross-over trials 270

Clinical trials with a cross-over design will be excluded in this review. 271

2. Cluster-randomized trials 272

Cluster randomized trials will be included in this review as long as proper adjustment for 273

the intra-cluster correlation is conducted19. We will reduce the size of each trial to its 274

effective sample size. A common design effect will be assumed across intervention 275

groups. For dichotomous data both the number of participants and the number 276

experiencing the event should be divided by the same design effect. For continuous 277

data only the sample size needs to be reduced; means and standard deviations should 278

remain unchanged. 279

Page 13 of 31


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960



jopen.bmj.com

/B

MJ O


ovember 2016. D

ownloaded from




14

3. Multiple intervention groups 280

Where multiple trial arms are reported in a single trial, arms will be combined as long as 281

they can be regarded as subtypes of the same hemoperfusion therapy. When arms 282

cannot be regarded as if in each of them a different subtype of the same intervention is 283

administered; we will include only the relevant arms. In such a case, a single pair-wise 284

comparison will be used to avoid a unit of analysis error. For dichotomous outcomes, 285

data from a different dosage of the same relevant active intervention arms will be 286

summed into a single arm for comparison. For continuous outcomes, means and 287

standard deviations will be combined19. 288

289

Dealing with missing data 290

We will try to contact the study authors for all relevant missing data. 291

1. Missing participants 292

293


The proportion of treatment failure will be calculated using an intention-to-treat (ITT) 295

following the principle ‘once randomized always analyzed’. To this end, all randomized 296

patients for whom outcome data are not available will be assumed as treatment failures. 297

Any assumptions and imputations to handle missing data will be clearly stated, and the 298

effect of imputation will be explored by sensitivity analyses. 299

300


An available cases analysis will be performed for patients with a final assessment 302

presented by the original authors. 303

2. Missing statistics 304

Page 14 of 31


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960



jopen.bmj.com

/B

MJ O


ovember 2016. D

ownloaded from




15

When only P values or standard error (SE) values are reported, we will calculate 305

standard deviations (SDs) according to Altman20. If none of these values are available, 306

and in the absence of supplementary data after requests to the authors, the SDs will be 307

calculated from CIs, t values or P values19; or they will be imputed from other studies in 308

the meta-analysis according to a validated method21. We will examine the validity of 309

these imputations in a sensitivity analysis. Where actual P values obtained from t tests 310

are reported, the corresponding t value will be obtained from a table of the t distribution. 311

312

Assessment of heterogeneity 313

We will assess overall heterogeneity by visual inspection of the forest plots. We will 314

assess statistical heterogeneity using the I2 statistic (on a scale of 0% to 100%) and Chi2 315

test. I2 values above 50% will be considered to represent substantial statistical 316

heterogeneity and to be explored further. However, the importance of the observed I2 317

depends on the magnitude and direction of treatment effects and the strength of 318

evidence for heterogeneity19. Since the Chi2 test has low power when studies have 319

small sample size or are few in number, we will use P value of 0.10 to determine 320

statistical significance. To provide an indication of the spread of true investigation effects, 321

we will also report between-study variance in a random-effects meta-analysis using Tau2. 322

For absolute measures of effect, an approximate 95% range of underlying effects can 323

be obtained by creating an interval from 1.96×Tau below the random-effects pooled 324

estimate, to 1.96×Tau above it. 325

326

Assessment of reporting biases 327

We will undertake comprehensive searches of multiple sources (including trial 328

registries), increasing efforts to identify the unpublished materials and including 329

Page 15 of 31


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960



jopen.bmj.com

/B

MJ O


ovember 2016. D

ownloaded from




16

non-English language publications as far as possible, to minimize the impact of 330

reporting biases. We will also try to identify outcome reporting bias in trials by recording 331

all trial outcomes, planned and reported, and noting where outcomes were missing. 332

When we find evidence of missing outcomes, we will attempt to obtain any available 333

data directly from the authors. Where this is not possible, and the missing data are 334

thought to introduce serious bias, the impact of including such studies in the overall 335

assessment of results will be explored by a sensitivity analysis. Funnel plots will be 336

constructed, and visual inspection will be performed to investigate the asymmetry. 337

When ten or more trials are pooled for primary outcomes, tests for funnel plot 338

asymmetry will be used to investigate the potential influence of reporting biases and 339

small-study effects. 340

341

Data synthesis 342

Meta-analyses will be performed should a sufficient number of studies are found that 343

share study design and measurement of comparators; otherwise, we will describe the 344

results from each studies. Given the likelihood of differing underlining disorders in the 345

population of interest, such as abdominal sepsis and others, a random-effects model will 346

be used in all analyses22. 347

348

Subgroup analysis and investigation of heterogeneity 349

We expect considerable heterogeneity for the primary outcomes, and we will use 350

subgroup and meta-regression analyses as exploratory tools to explain them. Exploring 351

sources of heterogeneity may result in false positive conclusions through multiple 352

analyses. Because these analyses lack power, they are more likely to result in false 353

negative results. Giving thought to these limitations, we will perform the following a priori 354

Page 16 of 31


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960



jopen.bmj.com

/B

MJ O


ovember 2016. D

ownloaded from




17

subgroup analyses for the participant group and intervention if sufficient detail is present 355

in the eligible studies. 356

357

1. Participants 358

・ Abdominal sepsis versus sepsis with other etiologies 359

・ Culture positive sepsis versus others 360

・ Confirmed gram negative sepsis versus others 361

・ Patients with high-level endotoxin activity defined as endotoxin activity assay 362

≥0.6 versus low-level endotoxin activity 363

・ Surgical patients versus medical patients 364

・ Sepsis versus septic shock 365

・ Patients with acute kidney injury (AKI) versus others 366

We will use the formal statistical test for heterogeneity across subgroup based on 367

random-effects model to test for subgroup interaction. 368

369

2. Intervention 370

・ Single session versus two sessions versus more than two sessions 371

・ Less than two hours versus two hours versus longer than two hours 372

373

Sensitivity analysis 374

To determine the sensitivity of the findings to the way in which we have conducted the 375

analysis, we will perform sensitivity analyses in the following areas. 376

1. Risk of bias: We will include only trials with low risk of bias in allocation 377

concealment. 378

2. Imputed missing data: We will exclude trials for which missing data are imputed. 379

Page 17 of 31


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960



jopen.bmj.com

/B

MJ O


ovember 2016. D

ownloaded from




18

3. Statistical method of data synthesis: a random-effects model versus fixed-effect 380

model. 381

382

Summary of findings tables 383

We will assess the quality of the body of evidence using the Grading of 384

Recommendations Assessment, Development and Evaluation (GRADE)23. We will 385

present the results of the review in the ‘Summary of Findings’ tables including the 386

following outcomes: 28-day all-cause mortality, serious adverse events, changes in 387

organ dysfunction scores, 90-day all-cause mortality, changes in mean arterial blood 388

pressure over 72 hours after the treatment, duration of vasopressor therapy or 389

vasopressor-free days, and cost related to health services. 390

391

Ethics and dissemination 392

The study will evaluate the clinical effect of PMX-HP for patients with sepsis/septic 393

shock from available published and unpublished clinical trial data. As no primary data is 394

collected, formal health research ethics approval is not required. The study will be 395

disseminated by peer-reviewed publication and conference presentation. If our protocol 396

needs to be amended, the date, details of the change, and the rationale will be 397

documented in the revised protocol and updated on PROSPERO. 398

399

DISCUSSION 400

The role of endotoxin in sepsis is well established in the literature9,10. PMX-HP was 401

developed to remove and clear circulating endotoxin2,3, which leads to decreases in 402

inflammatory cytokines and mediators. PMX-HP has also been reported to adsorb 403

activated neutrophils24 and monocytes25 in septic patients. A variety of small open-label 404

clinical trials have been published with generally promising results13. Nevertheless, data 405

Page 18 of 31


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960



jopen.bmj.com

/B

MJ O


ovember 2016. D

ownloaded from




19

from previous studies should be considered as inconclusive, as those trials inherit high 406

risk of bias, i.e. underpowered or unblinded. The purpose of our planned systematic 407

review is to determine efficacy, safety, and effectiveness of PMX-HP for adult patients 408

with sepsis or septic shock. The key strength of this protocol is its comprehensive 409

search for relevant studies, including on-going trials14 and unpublished data. No 410

language restriction will be placed and thorough search in the databases in the country 411

where PMX-HP was developed will enable data comprehensive review to update our 412

knowledge. 413

There are several expected limitations for this review. First, we defined sepsis as 414

documented or clinically suspected systemic infection and septic shock as 415

sepsis-induced hypotension persisting despite fluid resuscitation, in line with the 416

well-known definition17. A new definition for sepsis was proposed recently26, and future 417

research on sepsis might use this new definition. To address the issue of variation in 418

participants in interpreting and applying the results, we propose to perform a subgroup 419

analyses for the participants according to endotoxin activity levels. Second, the results 420

of this review will depend on the quality of the studies we identify, which might be low as 421

in the previous review13. We will, therefore, perform sensitivity analyses to see whether 422

the findings will change depending on the risk of bias of each study. Finally, small 423

studies with high heterogeneity could make it difficult to interpret the obtained results. 424

However, we expect the present review to provide the most comprehensive and up 425

to date critical summary of available evidence regarding the use of PMX-HP for sepsis 426

or septic shock in the ICU and will help guide treatment recommendations of sepsis or 427

septic shock in the clinical practice guidelines5. 428

429

Acknowledgements 430

Page 19 of 31


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960



jopen.bmj.com

/B

MJ O


ovember 2016. D

ownloaded from




20

We thank Dr. Toshiyuki Suwa at Osaka University for his assistance with search 431

strategy design in Japanese databases, and Tara Landry at the University of Alberta for 432

her peer reviewing the MEDLINE search strategy. 433

434

Funding and competing interests 435

TF is supported by Japan Society for the Promotion of Science. SMB is supported by a 436

Canada Research Chair in Critical Care Nephrology and is a Steering Committee 437

member of the EUPHRATES trial. TAF has received lecture fees from Eli Lilly, Janssen, 438

Meiji, MSD, Otsuka, Pfizer and Tanabe-Mitsubishi, and consultancy fees from Sekisui 439

Chemicals and Takeda Science Foundation. He has received royalties from Igaku-Shoin 440

and Nihon Bunka Kagaku-sha publishers. He has received grant or research support 441

from the Japanese Ministry of Education, Science, and Technology, the Japanese 442

Ministry of Health, Labour and Welfare, the Japan Society for the Promotion of Science, 443

the Japan Foundation for Neuroscience and Mental Health, Mochida and 444

Tanabe-Mitsubishi. He is diplomate of the Academy of Cognitive Therapy. This review 445

will be supported in part by JSPS KAKENHI (Grant-in-Aid for Scientific Research) Grant 446

Number 26670314 to TAF. All the other authors declare that they have no competing 447

interests. None of these funding organizations or any commercial organization have 448

contributed to the study design; collection, management, analysis and interpretation of 449

data; writing of the report or the decision to submit the report for publication. Only the 450

named authors have ultimate authority over these activities. 451

452

Authors’ Contributions 453

TAF is the guarantor of the protocol. TF and SMB conceived the project. TF and RG 454

worked equally on the initial architecture for the review with methodological input from 455

Page 20 of 31


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960



jopen.bmj.com

/B

MJ O


ovember 2016. D

ownloaded from




21

YK, RF, and SMB. TAF helped draft the article and revise it critically for important 456

intellectual content. All authors gave final approval of the version to be published. 457

458

Page 21 of 31


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960



jopen.bmj.com

/B

MJ O


ovember 2016. D

ownloaded from




22

References 459

1. Kaukonen KM, Bailey M, Suzuki S, Pilcher D, Bellomo R. Mortality related to severe 460

sepsis and septic shock among critically ill patients in Australia and New Zealand, 461

2000-2012. JAMA. 2014;311:1308-16. 462

2. Gaieski DF, Edwards JM, Kallan MJ, Carr BG. Benchmarking the incidence and 463

mortality of severe sepsis in the United States. Crit Care Med. 2013; 41: 1167-74. 464

3. Suffredini AF, Munford RS. Novel therapies for septic shock over the past 4 decades. 465

JAMA. 2011;306:194-9. 466

4. Dellinger RP, Carlet JM, Masur H, Gerlach H, Calandra T, Cohen J, Gea-Banacloche 467

J, Keh D, Marshall JC, Parker MM, Ramsay G, Zimmerman JL, Vincent JL, Levy MM; 468

Surviving Sepsis Campaign Management Guidelines Committee. Surviving Sepsis 469

Campaign guidelines for management of severe sepsis and septic shock. Crit Care 470

Med. 2004;32:858-73. 471

5. Dellinger RP, Levy MM, Rhodes A, Annane D, Gerlach H, Opal SM, Sevransky JE, 472

Sprung CL, Douglas IS, Jaeschke R, Osborn TM, Nunnally ME, Townsend SR, 473

Reinhart K, Kleinpell RM, Angus DC, Deutschman CS, Machado FR, Rubenfeld GD, 474

Webb SA, Beale RJ, Vincent JL, Moreno R; Surviving Sepsis Campaign Guidelines 475

Committee including the Pediatric Subgroup. Surviving sepsis campaign: 476

international guidelines for management of severe sepsis and septic shock: 2012. 477

Crit Care Med. 2013; 41:580-637. 478

6. Marshall JC, Foster D, Vincent JL, Cook DJ, Cohen J, Dellinger RP, Opal S, Abraham 479

E, Brett SJ, Smith T, Mehta S, Derzko A, Romaschin A; MEDIC study. Diagnostic and 480

prognostic implications of endotoxemia in critical illness: results of the MEDIC study. J 481

Infect Dis. 2004;190:527-34. 482

Page 22 of 31


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960



jopen.bmj.com

/B

MJ O


ovember 2016. D

ownloaded from




23

7. Shoji H, Tani T, Hanasawa K, Kodama M. Extracorporeal endotoxin removal by 483

polymyxin B immobilized fiber cartridge: designing and antiendotoxin efficacy in the 484

clinical application. Ther Apher. 1998;2:3-12. 485

8. Shoji H. Extracorporeal endotoxin removal for the treatment of sepsis: endotoxin 486

adsorption cartridge (Toraymyxin). Ther Apher Dial. 2003;7:108-14. 487

9. Danner RL, Elin RJ, Hosseini JM, Wesley RA, Reilly JM, Parillo JE. Endotoxemia in 488

human septic shock. Chest. 1991;99:169-75. 489

10. Casey LC, Balk RA, Bone RC. Plasma cytokine and endotoxin levels 490

correlate with survival in patients with the sepsis syndrome. Ann Intern Med. 491

1993;119:771-8. 492

11. Vincent JL, Laterre PF, Cohen J, Burchardi H, Bruining H, Lerma FA, Wittebole X, 493

De Backer D, Brett S, Marzo D, Nakamura H, John S. A pilot-controlled study of a 494

polymyxin B-immobilized hemoperfusion cartridge in patients with severe sepsis 495

secondary to intra-abdominal infection. Shock. 2005;23:400-5. 496

12. Cruz DN, Antonelli M, Fumagalli R, Foltran F, Brienza N, Donati A, Malcangi V, 497

Petrini F, Volta G, Bobbio Pallavicini FM, Rottoli F, Giunta F, Ronco C. Early use of 498

polymyxin B hemoperfusion in abdominal septic shock: the EUPHAS randomized 499

controlled trial. JAMA. 2009;301:2445-52. 500

13. Cruz DN, Perazella MA, Bellomo R, de Cal M, Polanco N, Corradi V, Lentini P, 501

Nalesso F, Ueno T, Ranieri VM, Ronco C. Effectiveness of polymyxin B-immobilized 502

fiber column in sepsis: a systematic review. Crit Care. 2007;11:R47. 503

14. Klein DJ, Foster D, Schorr CA, Kazempour K, Walker PM, Dellinger RP. The 504

EUPHRATES trial (Evaluating the Use of Polymyxin B Hemoperfusion in a 505

Randomized controlled trial of Adults Treated for Endotoxemia and Septic shock): 506

study protocol for a randomized controlled trial. Trials. 2014;15:218. 507

Page 23 of 31


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960



jopen.bmj.com

/B

MJ O


ovember 2016. D

ownloaded from




24

15. Payen DM, Guilhot J, Launey Y, Lukaszewicz AC, Kaaki M, Veber B, Pottecher J, 508

Joannes-Boyau O, Martin-Lefevre L, Jabaudon M, Mimoz O, Coudroy R, Ferrandière 509

M, Kipnis E, Vela C, Chevallier S, Mallat J, Robert R; ABDOMIX Group. Early use of 510

polymyxin B hemoperfusion in patients with septic shock due to peritonitis: a 511

multicenter randomized control trial. Intensive Care Med. 2015;41:975-84 512

16. Shamseer L, Moher D, Clarke M, Ghersi D, Liberati A, Petticrew M, Shekelle P, 513

Stewart LA; PRISMA-P Group. Preferred reporting items for systematic review and 514

meta-analysis protocols (PRISMA-P) 2015: elaboration and explanation. BMJ. 515

2015;349:g7647. 516

17. Levy MM, Fink MP, Marshall JC, Abraham E, Angus D, Cook D, Cohen J, Opal SM, 517

Vincent JL, Ramsay G, International Sepsis Definitions Conference. 2001 518

SCCM/ESICM/ACCP/ ATS/SIS international sepsis definitions conference. Intensive 519

Care Med. 2003;29:530–538. 520

18. Sampson M, McGowana J, Cogob E, Grimshaw J, Moherd D, Lefebvre C. An 521

evidence-based practice guideline for the peer review of electronic search strategies. 522

J Clin Epidemiol. 2009; 62: 944-952. 523

19. Higgins JPT, Green S (editors). Cochrane Handbook for Systematic Reviews of 524

Interventions. Chichester (UK): John Wiley & Sons, 2008. 525

20. Altman DG, Bland JM. Detecting skewness from summary information. BMJ. 1996; 526

313: 1200. 527

21. Furukawa TA, Barbui C, Cipriani A, Brambilla P, Watanabe N. Imputing missing 528

standard deviations in meta-analyses can provide accurate results. J Clin Epidemiol. 529

2006; 59: 7-10. 530

22. DerSimonian R, Laird N. Meta-analysis in clinical trials. Control Clin Trials. 1986; 7: 531

177-88. 532

Page 24 of 31


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960



jopen.bmj.com

/B

MJ O


ovember 2016. D

ownloaded from




25

23. Guyatt G, Oxman AD, Akl EA, Kunz R, Vist G, Brozek J, Norris S, Falck-Ytter Y, 533

Glasziou P, DeBeer H, Jaeschke R, Rind D, Meerpohl J, Dahm P, Schünemann HJ. 534

GRADE guidelines: 1. Introduction-GRADE evidence profiles and summary of 535

findings tables. J Clin Epidemiol. 2011;64:383-94. 536

24. Kumagai T, Takeyama N, Yabuki T, Harada M, Miki Y, Kanou H, Inoue S, Nakagawa 537

T, Noguchi H: Apheresis of activated leukocytes with an immobilized polymyxin B 538

filter in patients with septic shock. Shock 2010;34:461-466. 539

25. Tsujimoto H, Ono S, Hiraki S, Majima T, Kawarabayashi N, Sugasawa H, Kinoshita 540

M, Hiraide H, Mochizuki H: Hemoperfusion with polymyxin B-immobilized fibers 541

reduced the number of CD16+CD14+ monocytes in patients with septic shock. J 542

Endotoxin Res 2004;10:229-237. 543

26. Singer M, Deutschman CS, Seymour CW, Shankar-Hari M, Annane D, Bauer M, 544

Bellomo R, Bernard GR, Chiche JD, Coopersmith CM, Hotchkiss RS, Levy MM, 545

Marshall JC, Martin GS, Opal SM, Rubenfeld GD, van der Poll T, Vincent JL, Angus 546

DC. The Third International Consensus Definitions for Sepsis and Septic Shock 547

(Sepsis-3). JAMA. 2016;315:801-10. 548

Page 25 of 31


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960



jopen.bmj.com

/B

MJ O


ovember 2016. D

ownloaded from




26

Tables

Table 1. Domains for assessment of the risk of bias.

1. Random sequence generation (selection bias).

2. Allocation concealment (selection bias).

3. Blinding of participants and personnel (performance bias).

4. Blinding of outcome assessment (assessment bias).

5. Incomplete outcome data reporting (attrition bias).

6. Selective outcome reporting (reporting bias).

7. Treatment fidelity (performance bias).

8. Other biases (Sample size estimation, Co-intervention imbalance, Sponsorship

bias).

Page 26 of 31


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960



jopen.bmj.com

/B

MJ O


ovember 2016. D

ownloaded from




27

Table 2. Assessment of the risk of bias.

The risk of bias Explanation

Low risk Plausible bias unlikely to seriously alter the results

Unclear risk Insufficient information to determine whether the risk of bias is low or high

High risk Plausible bias that seriously weakens confidence in the results

Page 27 of 31


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960



jopen.bmj.com

/B

MJ O


ovember 2016. D

ownloaded from



PRISMA-P (Preferred Reporting Items for Systematic review and Meta-Analysis Protocols) 2015 checklist: recommended items to address in a systematic review protocol* Section and topic Item No Checklist item

ADMINISTRATIVE INFORMATION Title:

Identification 1a Identify the report as a protocol of a systematic review See page 1, line 2-3

Update 1b If the protocol is for an update of a previous systematic review, identify as such NOT APPLICABLE

Registration 2 If registered, provide the name of the registry (such as PROSPERO) and registration number See page 7, line 119-123

Authors: Contact 3a Provide name, institutional affiliation, e-mail address of all protocol authors; provide physical mailing address of

corresponding author See page 1, line 8-24

Contributions 3b Describe contributions of protocol authors and identify the guarantor of the review See page 19, line 414-418

Amendments 4 If the protocol represents an amendment of a previously completed or published protocol, identify as such and list changes; otherwise, state plan for documenting important protocol amendments See page 17, line 365-367

Support: Sources 5a Indicate sources of financial or other support for the review

See page 18, line 397-413 Sponsor 5b Provide name for the review funder and/or sponsor

See page 18, line 397-413 Role of sponsor or funder 5c Describe roles of funder(s), sponsor(s), and/or institution(s), if any, in developing the protocol


INTRODUCTION Rationale 6 Describe the rationale for the review in the context of what is already known

See page 6, line 94-107 Objectives 7 Provide an explicit statement of the question(s) the review will address with reference to participants, interventions,

comparators, and outcomes (PICO)

Page 28 of 31


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960



jopen.bmj.com

/B

MJ O


ovember 2016. D

ownloaded from




METHODS Eligibility criteria 8 Specify the study characteristics (such as PICO, study design, setting, time frame) and report characteristics (such as years

considered, language, publication status) to be used as criteria for eligibility for the review See page 8, line 125-161

Information sources 9 Describe all intended information sources (such as electronic databases, contact with study authors, trial registers or other grey literature sources) with planned dates of coverage See page 9, line 162-188

Search strategy 10 Present draft of search strategy to be used for at least one electronic database, including planned limits, such that it could be repeated See Appendix 2

Study records: Data management 11a Describe the mechanism(s) that will be used to manage records and data throughout the review

See page 11, line 212-220 Selection process 11b State the process that will be used for selecting studies (such as two independent reviewers) through each phase of the

review (that is, screening, eligibility and inclusion in meta-analysis) See page 10, line 190-199

Data collection process 11c Describe planned method of extracting data from reports (such as piloting forms, done independently, in duplicate), any processes for obtaining and confirming data from investigators See page 10, line 200-208

Data items 12 List and define all variables for which data will be sought (such as PICO items, funding sources), any pre-planned data assumptions and simplifications See page 11, line 208-211

Outcomes and prioritization 13 List and define all outcomes for which data will be sought, including prioritization of main and additional outcomes, with rationale See page 8, line 149-161

Risk of bias in individual studies 14 Describe anticipated methods for assessing risk of bias of individual studies, including whether this will be done at the outcome or study level, or both; state how this information will be used in data synthesis See page 11, line 221-243

Page 29 of 31


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960



jopen.bmj.com

/B

MJ O


ovember 2016. D

ownloaded from



15a Describe criteria under which study data will be quantitatively synthesised

See page 15, 320-325 15b If data are appropriate for quantitative synthesis, describe planned summary measures, methods of handling data and

methods of combining data from studies, including any planned exploration of consistency (such as I2, Kendall’s τ) See page 12, line 244-305

15c Describe any proposed additional analyses (such as sensitivity or subgroup analyses, meta-regression) See page 15, line 326-351

Data synthesis

15d If quantitative synthesis is not appropriate, describe the type of summary planned See page 15, 322-323

Meta-bias(es) 16 Specify any planned assessment of meta-bias(es) (such as publication bias across studies, selective reporting within studies) See page 15, line 306-319

Confidence in cumulative evidence 17 Describe how the strength of the body of evidence will be assessed (such as GRADE) See page 17, line 353-355

* It is strongly recommended that this checklist be read in conjunction with the PRISMA-P Explanation and Elaboration (cite when available) for important clarification on the items. Amendments to a review protocol should be tracked and dated. The copyright for PRISMA-P (including checklist) is held by the PRISMA-P Group and is distributed under a Creative Commons Attribution Licence 4.0. From: Shamseer L, Moher D, Clarke M, Ghersi D, Liberati A, Petticrew M, Shekelle P, Stewart L, PRISMA-P Group. Preferred reporting items for systematic review and meta-analysis protocols (PRISMA-P) 2015: elaboration and explanation. BMJ. 2015 Jan 2;349(jan02 1):g7647.

Page 30 of 31


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960



jopen.bmj.com

/B

MJ O


ovember 2016. D

ownloaded from



Supplementary File: Sample Database Search Strategy

Database: Epub Ahead of Print, In-Process & Other Non-Indexed Citations, Ovid MEDLINE(R) Daily and Ovid MEDLINE(R) 1946 to Present 1. exp Sepsis/ 2. Sepsis-Associated Encephalopathy/ 3. Systemic Inflammatory Response Syndrome/ 4. Vasoplegia/ 5. bacill?emia*.tw,kf. 6. bacter* shock.tw,kf. 7. bacter?emia*.tw,kf. 8. (blood adj2 poison*).tw,kf. 9. candid?emia*.tw,kf. 10. endotox?emia*.tw,kf. 11. endotoxi* shock.tw,kf. 12. fung?emia*.tw,kf. 13. parasit?emia*.tw,kf. 14. (py?emia* or pyohemia*).tw,kf. 15. sepsis.tw,kf. 16. septic.tw,kf. 17. septic?emia*.tw,kf. 18. SIRS.tw,kf. 19. systemic inflammatory response syndrome*.tw,kf. 20. toxic shock.tw,kf. 21. vasopl?egia*.tw,kf. 22. vi?emia*.tw,kf. 23. or/1-22 24. Polymyxin B/ 25. aerosporin.tw,kf,nm. 26. PMX*.tw,kf,nm. 27. polymyxin*.tw,kf,nm. or 1404-26-8.rn. 28. Poly RX.tw,kf,nm. 29. toraymyxin*.tw,kf,nm. 30. or/24-29

31. Blood Component Removal/ 32. Endotoxins/bl [Blood] 33. exp Hemofiltration/ 34. Hemoperfusion/ 35. apheres?s.tw,kf. 36. blood component removal*.tw,kf. 37. DHP-PMX.tw,kf. 38. (endotoxin* adj3 (a?sor* or eliminat* or remov*)).tw,kf. 39. h?emadsor*.tw,kf. 40. (h?emo-filtrat* or h?emofiltrat*).tw,kf. 41. (h?emo-diafiltrat* or h?emodiafiltrat*).tw,kf. 42. (h?emo-dialysis or h?emodialysis).tw,kf. 43. (h?emo-perfus* or h?emoperfus*).tw,kf. 44. (h?emo-sor* or h?emosor*).tw,kf. 45. PMX-?HP*.tw,kf. 46. pheres?s.tw,kf. 47. or/31-46 48. and/30,47 49. and/23,48 50. controlled clinical trial.pt. 51. randomized controlled trial.pt. 52. drug therapy.fs. 53. groups.ab. 54. placebo.ab. 55. random*.ab. 56. trial.ab. 57. or/50-56 58. exp animals/ not humans.sh. 59. 57 not 58 60. and/49,59

Page 31 of 31


BMJ Open

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960



jopen.bmj.com

/B

MJ O


ovember 2016. D

ownloaded from

Tomoko Fujii

Appendix 2. Sample search strategy for Ovid MEDLINE(R)

Tomoko Fujii


Documents

BMJ Open · For peer review only PMX SR protocol (BMJ open) 7 98 improved in patients treated with PMX-HP11.In a multi-center Italian trial, Cruz et al. 99 showed that PMX-HP improved