TITLE: Hydroxyethyl Starch versus Other Plasma Volume Expanders: A Review of the Clinical and Cost-Effectiveness, and Guidelines for Use

DATE: 23 May 2013

CONTEXT AND POLICY ISSUES

Fluid resuscitation is indicated for the management of hypovolemia (decreased blood plasma volume) and hypovolemic shock,1 and its ultimate objective is to restore organ perfusion and tissue oxygenation.2 Hypovolemia can be induced by a wide range of clinical conditions such as dehydration, burns, sepsis, malignancies, trauma, hemorrhage, and surgical anesthesia. Clinical signs of hypovolemia include blood pressure, urine output, mental status, and peripheral perfusion.1 There are two main types of fluids used for fluid resuscitation, colloids and crystalloids. Crystalloid solutions include normal saline and balanced fluids such as Ringer’s lactate.3 Human albumin preparations are natural colloids while dextran, gelatin, and starch products are synthetic colloids, also called synthetic plasma volume expanders.3

Hydroxyethyl starches are a common choice for fluid resuscitation.4 They are preferred over albumin because of their relatively lower price.5 These starches are supplied with different molecular weights ranging from 120 kDa to >450kDa. They are also characterized by the degree of substitution which is the molar substitution by hydroxyethyl groups. The degree of substitution ranges from 0.4 to 0.7. It is believed that the molecular weight and degree of substitution can affect patient outcomes.6

Four hydroxyethyl starches are available in Canada; these are Volvulen, Volulyte, Pentaspan and Hextend. Characteristics of these solutions are presented in Table 1.

Table 1. Characteristics of the Hydroxyethyl Starch Solutions Available in Canada Type of Molecular Degree of Maximum Solution Solvent/ medium Tonicity starch weight (kda) substitution dose Volvulen Tetrastarch 130 0.4 Isotonic electrolyte 6% 50 ml/kg/day Volulyte Tetrasarch 130 0.4 0.9% NaCl 6% 33 ml/kg/day Pentaspan Pentastarch 200-300 0.4-0.5 0.9% NaCl 10% 28 ml/kg/day Hextend Hexastarch 450-800 0.75 Lactate electrolyte 6% 20 ml/kg/day

Disclaimer: The Rapid Response Service is an information service for those involved in planning and providing health care in Canada. Rapid responses are based on a limited literature search and are not comprehensive, systematic reviews. The intent is to provide a list of sources of the best evidence on the topic that CADTH could identify using all reasonable efforts within the time allowed. Rapid responses should be considered along with other types of information and health care considerations. The information included in this response is not intended to replace professional medical advice, nor should it be construed as a recommendation for or against the use of a particular health technology. Readers are also cautioned that a lack of good quality evidence does not necessarily mean a lack of effectiveness particularly in the case of new and emerging health technologies, for which little information can be found, but which may in future prove to be effective. While CADTH has taken care in the preparation of the report to ensure that its contents are accurate, complete and up to date, CADTH does not make any guarantee to that effect. CADTH is not liable for any loss or damages resulting from use of the information in the report.

Copyright: This report contains CADTH copyright material and may contain material in which a third party owns copyright. This report may be used for the purposes of research or private study only. It may not be copied, posted on a web site, redistributed by email or stored on an electronic system without the prior written permission of CADTH or applicable copyright owner.

Links: This report may contain links to other information available on the websites of third parties on the Internet. CADTH does not have control over the content of such sites. Use of third party sites is governed by the owners’ own terms and conditions.

The use of hydroxyethyl starches for fluid resuscitation is undergoing a comprehensive review.7 Results from recent randomized controlled trials suggest that these starches may be associated with increased risk of bleeding, renal dysfunction and mortality.7 However, results from these RCTs were not consistent; some RCTs such as the CRYSTMAS trial reported that HES did not differ from normal saline in terms of adverse events,8 while the 6S and VISEP trials reported that HES was associated with increased mortality and kidney injury.9,10 On the other hand, CHEST trial evaluated almost 7000 ICU patients who were resuscitated with HES or saline. The trial concluded that HES, compared to saline, was not associated with a significant increase in mortality, but it increased the incidence of renal replacement therapy.11,12 The evaluation of HES was further complicated by the retraction from the peer-reviewed literature of the studies published by Boldt and co-workers.11,13. Given these issues, the appropriate use of HES for fluid resuscitation remains unclear. The objective of the current review is to evaluate the evidence surrounding the use of hydroxyethyl starch in intensive-care units.

RESEARCH QUESTIONS

1. What is the clinical effectiveness of hydroxyethyl starch compared with other plasma volume expanders for use in patients in the intensive care unit?

2. What are the guidelines regarding the use of hydroxyethyl starch as a volume expander for patients in the intensive care unit?

3. What is the cost-effectiveness of hydroxyethyl starch used as a volume expander for patients in the intensive care unit?

KEY FINDINGS

Twelve meta-analyses, two clinical guidelines, and one cost-effectiveness study were included in the review. The available evidence indicates the use of hydroxyethyl starch is associated with increased risk of mortality and renal dysfunction when used for fluid resuscitation in critically ill patients and patients with severe sepsis. The available evidence suggests that the use of hydroxyethyl starch in patients undergoing surgery may be associated with increased risk of bleeding and reoperation due to bleeding. One cost-effectiveness analysis showed that hydroxyethyl starches were less cost-effective, compared with albumin, when used for fluid resuscitation in sepsis patients.

Hydroxyethyl Starch versus Other Plasma Volume Expanders 2

METHODS

Literature Search Strategy

A limited literature search was conducted on key resources including PubMed, The Cochrane Library (2013, Issue 4), University of York Centre for Reviews and Dissemination (CRD) databases, Canadian and major international health technology agencies, as well as a focused Internet search. No filters were applied to limit the retrieval by study type. Where possible, retrieval was limited to the human population. The search was also limited to English language documents published between Jan 1, 2010 and Apr 24, 2013.

Selection Criteria and Methods

One reviewer screened citations and selected studies. In the first level of screening, titles and abstracts were reviewed for relevance. Full texts of any relevant titles/abstracts were retrieved, and assessed for inclusion. The final article selection was based on the inclusion criteria presented in Table 2.

Table 2 : Selection Criteria Patients who need fluid replacement treated in intensive care unit, Population and patients who could be considered as intensive care (e.g., trauma, post-surgery)

Intervention Hydroxyethyl starch (HES), a colloidal starch

Crystalloid solutions (e.g., saline, Ringer’s lactate) Comparator Albumin Q1: comparative clinical effectiveness (benefits, harms, safety) Outcomes Q2: evidence-based guidelines Q3: cost-effectiveness, comparative or not Systematic reviews, meta-analyses, health technology assessments Study Designs and cost-effectiveness studies

Exclusion Criteria

Systematic reviews and meta-analyses were excluded if they included primary trials conducted on healthy volunteer and of stable patients with chronic diseases. An additional exclusion criteria for systematic reviews was the inclusion of uncontrolled trials and observational studies. This was justified by the abundance of RCTs evaluating HES in different medical contexts; a scoping literature search detected more than 70 RCTs published between 2008 and 2013. Therefore, the inclusion of systematic reviews with uncontrolled trials or observational studies would have compromised the quality of evidence without additional value in terms of information quantity. Furthermore, reviews were excluded if they did not evaluate HES specifically i.e., reviews on colloids without specifying HES as a separate subgroup.

Hydroxyethyl Starch versus Other Plasma Volume Expanders 3

Systematic reviews or guidelines with a more recent update were be considered duplicates and were excluded; the most recent update will be considered the primary publication, and previous versions were used as secondary source of data.

Critical Appraisal of Individual Studies

Critical appraisal of the included studies was based on study design.

The Appraisal of Guidelines Research and Evaluation (AGREE) instrument14 was used to evaluate the quality of the included guideline. The methodological quality of the included systematic reviews and meta-analyses was evaluated using the “assessment of multiple systematic reviews” (AMSTAR).15 AMSTAR is an 11-item checklist that has been developed to ensure reliability and construct validity of systematic reviews. The methodological quality of the included cost-effectiveness study was assessed using the guidelines for appraisal of economic studies by Drummond et al.16

For the included studies a numeric score was not calculated. Instead, the strengths and limitations of the study were described.

SUMMARY OF EVIDENCE

Quantity of Research Available

A total of 713 potential citations were identified by searching the bibliographic database, with 683 citations being excluded during the title and abstract screening based on their irrelevance to the questions of interest. The full text documents of the remaining 30 articles were retrieved. Three additional articles were identified by grey literature and hand search. Of the 33 articles, 18 did not meet the inclusion criteria and were excluded; leaving 15 articles that reported 12 meta-analyses, two guidelines, and one cost-effectiveness study.

A PRISMA diagram demonstrating the study selection process is presented in Appendix 1.

Summary of Study Characteristics

Fifteen articles that addressed at least one of the review questions were included in this report. The use of HES for resuscitation of critically ill patients was evaluated in seven meta-analyses17- 23 and one guideline.24 Resuscitation with HES in patients with sepsis was assessed in two meta-analyses,25,26 one guideline,27 and one cost-effectiveness study.5 Finally, the use of HES in surgical settings was evaluated in two meta-analyses.6,28 Details on studies characteristics are tabulated in APPENDIX 2.

Critically ill patients (Table 3, Table 4)

The meta-analyses in this category included RCTs on patients treated in intensive care unit or emergency room; however, the medical condition for which fluid resuscitation was indicated was not systematically reported or considered in the analysis. One systematic review conducted subgroup analysis by diving trials’ population into sepsis and non-sepsis subgroups.23 The size of the analyzed population ranged from 1,61219 to almost 11,000 patients.17 Recent RCTs, with publication dates up to 2012, were included in three meta-analyses.17,18,20

Hydroxyethyl Starch versus Other Plasma Volume Expanders 4

Four meta-analyses evaluated non-specific formulation of HES.17-19,23 In these reviews, data from different HES formulations were pooled together. The molecular weights of the studied HES ranged from 120 kDa to 450 kDa; the degree of substitution ranged from 0.4 to 0.7.17-19,23 Two meta-analyses included RCTs that studies HES 130 kDa, 0.4 degree of substitution (130/0.4),20,21 and one meta-analysis evaluated hyperoncotic HES (10%) without specifying the molecular formulation.22 Comparators were crystalloids only,18 colloids only,19 or a both colloids and crystalloids which were pooled together as single comparator group.17,20-23 The two meta- analyses that evaluated HES 130/0.4 formulation included other HES formulations in the comparators group.20,21

The included meta-analyses focused on two safety outcomes mainly, mortality and renal failure. Five studies reported comparisons in the incidence death events,17-22 three studies reported on acute kidney injury or renal replacement therapy.17,22,23.

One guideline was identified on the management of fluid resuscitation for critically ill patients.24 The guideline development was based on a literature review of published systematic reviews and meta-analyses. The recommendations were supported by an overall assessment of benefits, risks and economic burden associated with the use of fluid resuscitation.

Sepsis Patients (Table 5, Table 6, and Table 7)

Patel et al.25 and Haase et al.26 conducted two meta-analyses of RCTs on patients with sepsis and RCTs that reported sepsis patients in separate subgroups. Patel et al. included six RCTs published between 2006 and 2012; patients in these RCTs (n= 3000) were diagnosed with severe sepsis or septic shock.25 Haase et al. included RCTs on sepsis patients without specifying the severity.26 Nine RCTs were included, of which six were included by Patel et al.

The two meta-analysis evaluated HES 130/0.40 and HES 130/0.42 at an isoncotic concentration.25,26 Comparators were both non-HES colloids and crystalloids. Patel et al. reported separate analyses for colloids, crystalloids, and both groups pooled together.25 They also provided subgroup analysis based on the origin of starch: maize or potato.25 Haase et al. reported results of studies using colloids and crystalloids as comparators pooled together.26

Different analyses were reported by the two meta-analyses, and they are considered complementary to each other.25,26 Patel et al. reported risk ratio of mortality at 28 days, 90 days and the overall mortality.25 Haase et al. reported all-cause mortality estimates within 28 days of follow-up and beyond 28 days of follow-up.26

One guideline on the management of sepsis patients was identified.27 In 2013, Dellinger et al. provided an update for the “Surviving Sepsis Campaign Guidelines for Management of Severe Sepsis and Septic Shock” last published in 2008.27 The guideline considered survival of sepsis patients and their length of stay in the ICU as the basis for the recommendations.27

The analysis conducted by Farrugla et al. evaluated the cost-effectiveness of using HES in sepsis as compared with albumin.5 The estimation was based on a non-systematic literature search of RCTs on sepsis patients. The incremental cost-effectiveness ratio was estimated using decision analysis model.

Hydroxyethyl Starch versus Other Plasma Volume Expanders 5

Surgery Patients (Table 8)

Martin et al. included 17 RCTs on patients undergoing elective surgeries.6 The included trials were published between 2000 and 2011, and they included 1230 patients. Navickis et al. and Shi et al. included 18 and 52 RCTs respectively; these RCTs were conducted on patients undergoing cardiac surgeries.

Martin’s meta-analysis compared the effects of HES 130/0.4 with a mix of crystalloids and colloids that included HES 200.6 Outcomes included acute renal failure, creatinine levels and length of stay. The meta-analyses by Navickis and Shi compared a pool of HES formulations with albumin,4,29 gelatin,29 and crystalloids.29 Outcomes reported in these two studies included mortality, reoperation, and blood loss.4,29

Summary of Critical Appraisal

Details on studies appraisal are tabulated in APPENDIX 3.

Critically ill patients

The seven meta-analyses evaluating the use of HES for resuscitation in critically ill patients shared major strengths and limitations.17-23 They were based on systematic review of literature that was conducted by at least two investigators for each meta-analysis. The quality of the included RCTs was evaluated in all meta-analyses.

Limitations of the meta-analyses might affect the external validity of their results. In six meta- analyses the medical condition, for which fluid resuscitation was needed, was not reported or considered in the interpretation of results.17-22 The intervention in four meta-analyses was considered as a pool composed of different HES formulations.17-19,23 These analyses might mask the differences between HES formulations on patients outcomes. Three meta-analyses used the same concept of pooling for the comparator group;20,21,23 the results of crystalloids and colloid fluids were pooled together as one comparator.

The guideline published by Reinhart et al. was based on published systematic reviews and meta-analyses.24 The quality of evidence was incorporated in the recommendations. One potential shortcoming is that the guideline validation method was not reported, and it could not be evaluated. Examples of methods for guidelines validated are the peer review process and the collection of feedback from stakeholders, end users, or the target group of patients.

Sepsis Patients

The meta-analyses by Patel et al. and Haase et al. were based on systematic review of the literature and evidence evaluation.25,26 Furthermore, these two meta-analyses evaluated the effects of the length of follow-up and the HES formulation on patients’ outcomes. They also reported the fluid dose and its total volume; however, this was not factored in the analyses. One potential limitation was detected for Patel’s meta-analysis; for trials that had more than one comparator arm, the analysis considered one arm only without further explanation on how this arm was selected.25

Hydroxyethyl Starch versus Other Plasma Volume Expanders 6

One potential limitation could be detected in Patel’s meta-analysis.25 The analysis included four RCTs that had more than one comparator arm; however, only one arm was selected for the analyses. The selection criteria for the comparator arm from these four trials was not specified.25

The guideline published by Dellinger et al. was based on several systematic reviews and previous meta-analyses.27 The quality of evidence was incorporated in the recommendations. There were no major limitations to report for this guideline.

Surgery Patients

The three meta-analyses were based on systematic reviews of literature; two of them did not report if the quality of the included RCTs was evaluated.4,6 The medical condition for which fluid resuscitation was indicated was not considered in two meta-analyses.4,29 Martin’s meta-analysis reported inappropriate comparisons in which the intervention and comparator groups were compared in terms of baseline values and in terms of extreme post-intervention values separately.6 The analysis did not report the comparison of change from baseline to extreme values between the intervention and comparator groups.

Summary of Findings

Details on studies appraisal are tabulated in APPENDIX 4.

Reports on Critically ill patients (Table 9 and Table 10)

Three types of analysis were identified in the literature on critically ill patients. The first type grouped different HES formulation as one intervention, the second considered HES 130/0.4 as an independent intervention, and the third situation considered hyperoncotic HES formulation as an independent intervention.

Various HES formulations

Different HES formulations were analyzed together as one group of intervention. This was the case in four included meta-analyses.17-19,23 The comparator group in these meta-analyses was a mix of colloids and crytalloids,17,23 colloids only,19 or crystalloids only.18

Results of mortality showed that HES was associated with a statistically significant increased risk of death when compared with crystalloids only; the relative risk (RR) was 1.1 (95% confidence interval [CI] 1.02 to 1.19).18 This risk became statistically significant when HES was compared with either albumin or gelatin.19 The comparison with both colloids and crystalloids showed that HES was associated with a statistically increased risk of death only when trials published by Boldt’s et al. were removed from the analysis; RR 1.09 (95% CI 1.02 to 1.17).17

Renal replacement therapy was reported in two meta-analyses.17,23 The most recent one was published in 2013 by Zarychanski et al., and it included RCTs from 2012.17 Zarychanski reported that HES was associated with an increased risk of renal replacement therapy; RR 1.32 (95% CI 1.15 to 1.50).17 The second meta-analysis was published in 2010 by Dart et al.; its most recent RCTs were published in 2008.23 Dart reported that HES was not associated with statistically increased risk of renal replacement therapy; however, the subgroup analysis showed that HES increased this risk in sepsis patients; RR 1.59 (95% CI 1.20 to 2.10).23 The non-sepsis subgroup did not show a risk difference.23

Hydroxyethyl Starch versus Other Plasma Volume Expanders 7

Acute renal failure was also reported by Zarychanski and Dart.17,23 Zarychanski showed that HES was associated with a statistically increased risk of acute renal failure when compared with colloids and crystalloids; RR 1.27 (95% CI 1.09 to 1.47). Dart reported that HES was not associated with an increased risk of acute renal failure according to the RIFLE criteria;23 however, it was shown that, when kidney failure was evaluated as per authors’ definition, HES increased the risk of kidney failure; RR 1.50 (95% CI 1.20 to 1.87).23 Subgroup analysis showed that this risk was derived mainly for the sepsis patients subgroup; RR 1.55 (95% CI 1.22 to 1.96).23 The non-sepsis subgroup did not show a significant increase of risk.23

Zarychanski et al. reported that HES, compared to colloids and crystalloids, was associated with an increased rate of patients who needed red blood cell (RBC) transfusion; RR was 1.42 (95% CI 1.15, 1.75).17 They also reported that HES was associated with lower urine output; standardized mean difference (SMD) -0.15 (95% CI -0.19 to -0.1).17 On the other hand, the meta-analysis did not show statistical differences in terms of length of stay in intensive care unit or overall stay, major hemorrhage, or volume of transfused RBC.

HES 130/0.4

HES 130/0.4 alone was compared with both colloids and crystalloids which were pooled together.20,21 Both meta-analyses by Gattas21 and Weidermann20 did not show statistically significant increased risk of mortality associated with HES. Worth mentioning that the direction of the relative risk changed from favoring HES in Gattas review21 to favoring comparators in Wiedermann analysis.20 This might be explained be the fact that Wiedermann excluded all trials published by Boldt et al. and included two additional trials; FIRST and CRYSTMAS trials. Both analyses did not include the CHEST trial which was published in November 2012 after the meta-analyses were published.

Hyperoncotic HES

Hyperoncotic HES (10% concentration) was compared with both colloids and crystalloids which were pooled together.22 In this 2010 meta-analysis, Wiedermann reported that HES was associated with increased risk of mortality and acute kidney injury; the associated relative risks were 1.41 (95% CI 1.01 to 1.96) and 1.92 (95% CI 1.31 to 2.81) respectively.22

The guidelines published by Reinhart et al. on the management of fluid resuscitation in critically ill patients recommended against the use of HES with molecular weight of ≥200 kDa and/or degree of substitution >0.4 in patients with sepsis.24 The guideline also recommended that the use of HES 130/0.4 should be restricted to a clinical research context for severe sepsis patients and other ICU patients with increased risk of acute kidney injury or bleeding.24 These were strong recommendations based on moderate quality evidence.

Reports on Sepsis Patients (Table 11, Table 12, Table 13)

Meta-analyses by Patel et al.25 and Haase et al.26 both evaluated safety outcomes associated with the use of HES 130/0.40-0.42 compared with other resuscitation fluids used in patients with sepsis.25,26 Both meta-analyses showed that the use of HES was associated with an increased risk of mortality at a follow-up longer than 28 day or at 90-day follow-up; RR were 1.11 (95% CI 1.01 to 1.22)26 and 1.13 (95% CI 1.02 to 1.25)25 respectively. This risk was not statistically significant at follow-ups below 28 days.25,26

Hydroxyethyl Starch versus Other Plasma Volume Expanders 8

HES 130/0.40-0.42 was also associated with an increased risk of renal replacement therapy;25,26 evaluation at anytime was associated with RR of 1.36 (95% CI 1.08 to 1.72),26 and evaluation at 90 days gave a RR of 1.41 (1.08 to 1.84).25

The guidelines published by Dellinger et al. on the management of sepsis patients reported a strong recommendation against the use of HES for fluid resuscitation in patients with severe sepsis or septic shock.27

Cost-effectiveness analysis published by Farrugia et al. showed that the use of HES for fluid resuscitation in sepsis patients was associated with unfavorable incremental cost-effectiveness ratio when compared with albumin.5

Reports on Surgery Patients (Table 14)

Two systematic reviews reported that the use of HES was not associated with increased risk of adverse effects when used in surgery settings.6,29 Shi’s meta-analysis included 12 RCTs published by Boldt et al.;29 these trials had a high risk of data falsification.11,13 The meta-analysis by Navickis et al. reported that the use of HES for cardiopulmonary bypass patients was associated with increased risk of bleeding; SMD was 0.33 (95% CI 0.18 to 0.48), and reoperation due to bleeding; RR 2.24 (95% CI 1.14 to 4.40).4 This meta-analysis did not show statistically significant differences between HES and albumin in terms of fluid balance, heart rate, length of stay in the ICU or mortality.4

Limitations

The included trials evaluated trials conducted on patients requiring fluid resuscitation; however, the medical conditions that cause patients to need resuscitation were not systematically considered in the analyses. The study of patients with sepsis and patients undergoing major surgeries provided clearer information about the safety of HES solutions in these two patients groups. However, conditions such as burns and trauma were not evaluated separately; therefore, the generalizability of findings to these conditions remains unclear.

Furthermore, most of the included trials did not consider factors associated with fluid administration that could affect patients’ outcomes. These factors include the dose of fluid, the daily and total volume administered, and the strategy of fluid administration.

CONCLUSIONS AND IMPLICATIONS FOR DECISION OR POLICY MAKING

This report aimed to evaluate effectiveness of hydroxyethyl starches used for fluid resuscitation. The clinical guidelines and cost-effectiveness of these fluids were also reviewed. A total of twelve meta-analyses, two guidelines and one cost-effectiveness study were retrieved.

With respect to the effectiveness of hydroxyethyl starches, the included reports focused mainly on safety outcomes without a full assessment of efficacy. Therefore, conclusions should be made on the safety of starches rather than effectiveness. The use of hydroxyethyl starches was shown to be associated with increased risk of mortality and need for renal replacement therapy when used for fluid resuscitation of patients with severe sepsis and critically ill patients. This risk was accrued after four weeks of fluid administration. This risk could not be established in patients undergoing major surgeries, however the available evidence suggests that the use of hydroxyethyl starch in patients undergoing surgery may be associated with increased risk of

Hydroxyethyl Starch versus Other Plasma Volume Expanders 9

bleeding and reoperation due to bleeding. The included guidelines recommended against the use hydroxyethyl starches in patients with sepsis and other patients treated in intensive care settings who present an increased risk of renal failure.

The cost-effectiveness analysis showed that hydroxyethyl starches were not cost-effective, compared with albumin, when used for fluid resuscitation in sepsis patients.

PREPARED BY: Canadian Agency for Drugs and Technologies in Health Tel: 1-866-898-8439 www.cadth.ca

Hydroxyethyl Starch versus Other Plasma Volume Expanders 10

REFERENCES

1. Mandel J, Palevsky PM. Treatment of severe hypovolemia or hypovolemic shock in adults. 2013 [cited 2013 May 23]. In: UpToDate [Internet]. Version 22.0. Waltham (MA): UpToDate; 1992 - . Available from: www.uptodate.com Subscription required.

2. Marik PE, Monnet X, Teboul JL. Hemodynamic parameters to guide fluid therapy. Annals of Intensive Care. 2011;1:1.

3. Vincent JL. Fluid resuscitation: colloids vs crystalloids. Acta Clin Belg Suppl. 2007;(2):408-11.

4. Navickis RJ, Haynes GR, Wilkes MM. Effect of hydroxyethyl starch on bleeding after cardiopulmonary bypass: a meta-analysis of randomized trials. J Thorac Cardiovasc Surg. 2012 Jul;144(1):223-30.

5. Farrugia A, Martin G, Bult M. Colloids for sepsis: effectiveness and cost issues. Annual Update in Intensive Care and Emergency Medicine. 2013;515-26.

6. Martin C, Jacob M, Vicaut E, Guidet B, Van AH, Kurz A. Effect of waxy maize-derived hydroxyethyl starch 130/0.4 on renal function in surgical patients. Anesthesiology. 2013 Feb;118(2):387-94.

7. Hartog CS, Kohl M, Reinhart K. A systematic review of third-generation hydroxyethyl starch (HES 130/0.4) in resuscitation: safety not adequately addressed. Anesth Analg. 2011 Mar;112(3):635-45.

8. Guidet B, Martinet O, Boulain T, Philippart F, Poussel JF, Maizel J, et al. Assessment of hemodynamic efficacy and safety of 6% hydroxyethylstarch 130/0.4 vs. 0.9% NaCl fluid replacement in patients with severe sepsis: The CRYSTMAS study. Crit Care [Internet]. 2012 May 24 [cited 2013 May 23];16(3):R94. Available from: http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3580640

9. Perner A, Haase N, Guttormsen AB, Tenhunen J, Klemenzson G, Aneman A, et al. Hydroxyethyl starch 130/0.42 versus Ringer's acetate in severe sepsis. N Engl J Med [Internet]. 2012 Jul 12 [cited 2013 Apr 26];367(2):124-34. Available from: http://www.nejm.org/doi/pdf/10.1056/NEJMoa1204242

10. Brunkhorst FM, Engel C, Bloos F, Meier-Hellmann A, Ragaller M, Weiler N, et al. Intensive insulin therapy and pentastarch resuscitation in severe sepsis. N Engl J Med. 2008 Jan 10;358(2):125-39.

11. Shafer SL. Shadow of doubt. Anesth Analg. 2011 Mar;112(3):498-500.

12. Myburgh J, Li Q, Heritier S, Dan A, Glass P, Crystalloid Versus Hydroxyethyl Starch Trial (CHEST) Management Committee. Statistical analysis plan for the Crystalloid Versus Hydroxyethyl Starch Trial (CHEST). Crit Care Resusc. 2012 Mar;14(1):44-52.

13. Wise J. Boldt: the great pretender. BMJ. 2013;346:f1738.

Hydroxyethyl Starch versus Other Plasma Volume Expanders 11

14. The AGREE Collaboration. Appraisal of guidelines for research and evaluation (AGREE) instrument [Internet]. London: The AGREE Research Trust; 2001 Sep. [cited 2013 May 23]. Available from: http://www.agreetrust.org/?o=1085

15. Shea BJ, Grimshaw JM, Wells GA, Boers M, Andersson N, Hamel C, et al. Development of AMSTAR: a measurement tool to assess the methodological quality of systematic reviews. BMC Med Res Methodol [Internet]. 2007 [cited 2013 Mar 20];7:10. Available from: http://www.ncbi.nlm.nih.gov/pmc/articles/PMC1810543/pdf/1471-2288-7-10.pdf

16. Drummond MF, Jefferson TO. Guidelines for authors and peer reviewers of economic submissions to the BMJ. BMJ [Internet]. 1996 Aug 3 [cited 2013 Mar 4];313(7052):275- 83. Available from: http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2351717/pdf/bmj00553-0039.pdf

17. Zarychanski R, Abou-Setta AM, Turgeon AF, Houston BL, McIntyre L, Marshall JC, et al. Association of hydroxyethyl starch administration with mortality and acute kidney injury in critically ill patients requiring volume resuscitation: a systematic review and meta- analysis. JAMA. 2013 Feb 20;309(7):678-88.

18. Perel P, Roberts I, Ker K. Colloids versus crystalloids for fluid resuscitation in critically ill patients. Cochrane Database Syst Rev [Internet]. 2013 [cited 2013 Apr 26];2:CD000567. Available from: http://onlinelibrary.wiley.com/doi/10.1002/14651858.CD000567.pub6/pdf Subscription required.

19. Bunn F, Trivedi D. Colloid solutions for fluid resuscitation. Cochrane Database Syst Rev [Internet]. 2012 [cited 2013 Apr 26];7:CD001319. Available from: http://onlinelibrary.wiley.com/doi/10.1002/14651858.CD001319.pub5/pdf Subscription required.

20. Wiedermann CJ, Joannidis M. Mortality after hydroxyethyl starch 130/0.4 infusion: an updated meta-analysis of randomized trials. Swiss Med Wkly. 2012;142:w13656.

21. Gattas DJ, Dan A, Myburgh J, Billot L, Lo S, Finfer S, et al. Fluid resuscitation with 6% hydroxyethyl starch (130/0.4) in acutely ill patients: an updated systematic review and meta-analysis. Anesth Analg. 2012 Jan;114(1):159-69.

22. Wiedermann CJ, Dunzendorfer S, Gaioni LU, Zaraca F, Joannidis M. Hyperoncotic colloids and acute kidney injury: a meta-analysis of randomized trials. Crit Care [Internet]. 2010 [cited 2013 May 23];14(5):R191. Available from: http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3219298

23. Dart AB, Mutter TC, Ruth CA, Taback SP. Hydroxyethyl starch (HES) versus other fluid therapies: effects on kidney function. Cochrane Database Syst Rev. 2010;(1):CD007594.

24. Reinhart K, Perner A, Sprung CL, Jaeschke R, Schortgen F, Johan Groeneveld AB, et al. Consensus statement of the ESICM task force on colloid volume therapy in critically ill patients. Intensive Care Med. 2012 Mar;38(3):368-83.

Hydroxyethyl Starch versus Other Plasma Volume Expanders 12

25. Patel A, Waheed U, Brett SJ. Randomised trials of 6 % tetrastarch (hydroxyethyl starch 130/0.4 or 0.42) for severe sepsis reporting mortality: systematic review and meta- analysis. Intensive Care Med. 2013 May;39(5):811-22.

26. Haase N, Perner A, Hennings LI, Siegemund M, Lauridsen B, Wetterslev M, et al. Hydroxyethyl starch 130/0.38-0.45 versus crystalloid or albumin in patients with sepsis: systematic review with meta-analysis and trial sequential analysis. BMJ [Internet]. 2013 [cited 2013 May 23];346:f839. Available from: http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3573769

27. Dellinger RP, Levy MM, Rhodes A, Annane D, Gerlach H, Opal SM, et al. Surviving Sepsis Campaign: international guidelines for management of severe sepsis and septic shock, 2012. Intensive Care Med. 2013 Feb;39(2):165-228.

28. Van der LP, James M, Mythen M, Weiskopf RB. Safety of modern starches used during surgery. Anesth Analg. 2013 Jan;116(1):35-48.

29. Shi XY, Zou Z, He XY, Xu HT, Yuan HB, Liu H. Hydroxyethyl starch for cardiovascular surgery: a systematic review of randomized controlled trials. Eur J Clin Pharmacol. 2011 Aug;67(8):767-82.

30. Cramer JA, Bradley-Kennedy C, Scalera A. Treatment persistence and compliance with medications for chronic obstructive pulmonary disease. Can Respir J [Internet]. 2007 Jan [cited 2013 Mar 26];14(1):25-9. Available from: http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2690446/pdf/crj14025.pdf

31. Bland JM, Altman DG. The use of transformation when comparing two means. BMJ [Internet]. 1996 May 4 [cited 2013 May 23];312(7039):1153. Available from: http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2350653

Hydroxyethyl Starch versus Other Plasma Volume Expanders 13

APPENDIX 1: SELECTION OF INCLUDED STUDIES

713 citations identified from electronic literature search and screened

683 citations excluded

30 potentially relevant articles retrieved for scrutiny (full text, if available)

3 potentially relevant

reports retrieved from other sources (grey literature, hand search)

33 potentially relevant reports

18 reports excluded: Population not of interest (1) Intervention not of interest (8) Outcomes not of interest (1) Study design not of interest (3) Duplicates (5)

15 reports included in review

Hydroxyethyl Starch versus Other Plasma Volume Expanders 14

APPENDIX 2: Characteristics of the Included Systematic Reviews and Guideline

1. Critically Ill Patients

Table 3. Characteristics of the Included Systematic Reviews on Critically Ill Patients (Seven Meta-analyses) Objectives/Scope Type of primary Population/ Intervention Comparator Outcomes Notes studies Medical context 1/7. Zarychanski et al. 201317 – Meta-analysis on Critically ill Patients To evaluate the RCTs only The review included • HES Colloids: • Mortality Trials conducted by association of HES • A total of 38 trials, trials on critically ill o 450: 3 trials • Albumin: 12 trials • AKI Boldt et al. were use with mortality were included in adults in emergency o 300-500: 1 trial • Gelatin: 8 trials • RIFLE criteria used as sensitivity and acute kidney the review; of or ICU settings. o 200-300: 20 • Plasma: 1 trial • Urine output analysis criteria injury when used for which, 10 trials trials • ICU LoS resuscitation of were included in The medical o 130/0.4: 11 Crystalloids: • Overall LoS critically-ill patients the meta-analysis condition for which trials • 18 Ringer’s solution: • RBC transfusion of Perel et al. patients were 130/0.42: 1 trial 7 trials o o patients • Trials were treatment in the ICU 120/0.7: 1 trials • o 0.9% saline: 8 o volume published was not considered o Unknown: 1 trials The duration of between 1982 in the analysis trial • 4% saline: 1 trial follow-up or the and 2012 • Hartman solution: time of outcome • A total of 10,880 1 trial assessment were patients not considered in contributed to the analysis mortality analysis 2/7. Perel et al. 201318 – Meta-analysis on Critically ill Patients To evaluate the RCTs only The review included Colloids: Crystalloids: • Mortality The review effects on mortality • A total of 56 trials, trials on critically ill • Dextran 70 • Isotonic excluded trials on of using colloids as were included in patients who • HES • Hypertonic The duration of patients undergoing compred to the review; of require volume o 450: 2 trials follow-up or the elective surgical crystalloids when which, 25 trials replacement. o 200: 2 trials time of outcome patients used for evaluated HES o 139/0.4: 1 trial assessment were resuscitation in vs. crystalloids The medical o 130/0.4: 6 trials not considered in The review critically ill patients • Trials were condition for which o 130/0.42: 1 trial the analysis excluded studies published patients were o Pentastarch: 3 with cross-over between 1982 treatment in the ICU trials design was not considered and 2012 o Hetastarch: 1 trial • HES trials in the analysis o Haes steril: 1 trial Trials conducted by included total of o Unknown: 9 trials Boldt et al. were 9174 patients. • Modified gelatine used as sensitivity • Albumin analysis criteria • PPF

Hydroxyethyl Starch versus Other Plasma Volume Expanders 15

Table 3. Characteristics of the Included Systematic Reviews on Critically Ill Patients (Seven Meta-analyses) Objectives/Scope Type of primary Population/ Intervention Comparator Outcomes Notes studies Medical context AKI= acute kidney failure; HES= hydroxyethyl starch; ICU= intensive care unit; LoS= length of stay; PPF= plasma protein fraction; RBC= red blood cells 3/7. Bunn et al. 201219 – Meta-analysis on Critically ill patients To compare the RCTs only The review included Colloids: Colloids: • Mortality The review effects of different • A total of 86 trials, trials on patients • Albumin or • Albumin or excluded trials that colloid solution on were included in requiring volume plasma protein plasma protein The duration of used colloids during patients’ outcome the review; of replacement or fraction fraction follow-up or the plasma exchange which, 53 trials maintenance of • Gelatin • Gelatin time of outcome or for priming compared HES colloid osmotic • Dextran • Dextran assessment were extracorporeal with other colloids pressure. • HES • HES not considered in circuits. • Trials were the analysis published The medical The protocol between 1983 condition requiring reported that trials and 2010 volume expansion conducted by Boldt • A total of patients was not considered et al. will be contributed to in the analysis evaluated in mortality analysis sensitivity analysis. However, this analysis was not included in the report.

4/7. Gattas et al. 201221 – Meta-analysis on Critically ill Patients To update the RCTs only The review included HES 130/0.4 Colloids: • Mortality Retracted effects of HES • A total of 36 trials, trials on patients • Albumin: 8 trials publications by 130/0.4 on mortality were included in who are critically ill The type of solvent • Gelatin: 10 The duration of Boldt. et al. were of critically ill the review; or undergoing major was not reported • HES (other follow-up or the included as patients • Trials were surgery, and who formulations): 14 time of outcome sensitivity analysis published are requiring fluid The strategy used assessment were between 2000 resuscitation. for fluid Crystalloids: not considered in and 2010 administration was • Ringers’ solution: the analysis • A total of 1582 Resuscitation was not reported 7 trials patients needed for sepsis, • Normal saline: 1 contributed to cardiovascular trial mortality analysis surgery, or hypovolemia

HES= hydroxyethyl starch

Hydroxyethyl Starch versus Other Plasma Volume Expanders 16

Table 3. Characteristics of the Included Systematic Reviews on Critically Ill Patients (Seven Meta-analyses) Objectives/Scope Type of primary Population/ Intervention Comparator Outcomes Notes studies Medical context

5/7. Wiedermann et al. 201220 – Meta-analysis on Critically ill Patients To provide an RCTs only The review included HES 130/0.4 Colloids • Mortality The review update to Gattas et • A total of 13 trials, trials on patients excluded all trials/ 21 al. were included in who are critically ill The type of solvent Crystalloids The duration of publications by the meta- or undergoing major was not reported follow-up or the Boldt et al. analysis; surgery, and who time of outcome are requiring fluid The strategy used assessment were The review included • Trials were resuscitation. for fluid not considered in FIRST and published administration was the analysis CRYSMAS trials; between 2003 not reported these two trials and 2012 were not included in the meta-analysis 30 • A total of 1131 by Gatta et al. patients contributed to mortality analysis

6/7. Dart et al. 201023 – Meta-analysis on Critically ill Patients To evaluate the RCTs and quasi- The review included HES – any Colloids • Need for renal The review effects of HES on RCTs trials on patients formulation replacement excluded cross-over kidney functions • A total of 34 trials, who received fluids Crystalloids therapy trials and trials compared to other were included in for the treatment of • RIFLE criteria conducted on resuscitation fluids the review; effective • Creatinine healthy volunteers in different patient intravascular clearance populations • Trials were volume depletion • Glomerular Medical conditions published filtration rate requiring fluid between 1982 • Serum creatinine resuscitation were and 2008 classified as sepsis Outcomes had to and non-sepsis • A total of 2607 be assessed at indications; this was patients least 24 hours after used as subgroup contributed to treatment analysis kidney failure analysis

Hydroxyethyl Starch versus Other Plasma Volume Expanders 17

Table 3. Characteristics of the Included Systematic Reviews on Critically Ill Patients (Seven Meta-analyses) Objectives/Scope Type of primary Population/ Intervention Comparator Outcomes Notes studies Medical context HES= hydroxyethyl starch

7/7. Wiedermann et al. 201022 – Meta-analysis on Critically ill Patients To evaluate the RCTs The patient Hyperoncotic Crystalloids • Acute kidney Cross-over studies effect hyperoncotic • A total of 11 trials, population was not colloids: injury were not included in colloids on the were included in clearly defined in • 20% or 10% Hypo-oncotic • Mortality the review incidence of acute the review; 4 trials the report. Albumin colloids (4% or 5% kidney injury compared • 10% HES albumin) hyperoncotic HES The included trials with other fluids used fluid The volume and Trials that used iso- • Trials were resuscitation for strategy of fluid oncotic (6%) HES published surgery (abdominal administration were were excluded between 1981 or cardiac), sepsis not considered and 2008 and hypovolemia • A total of 696 patients were included in the HES trials HES= hydroxyethyl starch

Hydroxyethyl Starch versus Other Plasma Volume Expanders 18

Table 4. Characteristics of the Included Guidelines on the Management of Critically Ill Patients Objectives Methodology Intervention and Evidence collection, Recommendation Intended users/ Major Outcomes Evidence Quality Guideline Practice Selection and Formulation and Target population Considered and Strength Validation Considered Synthesis Evaluation Reinhart et al. 201224 – Guidelines for the Management of Fluid Resuscitation in Critically ill Patients Intended users Fluid resuscitation Benefits (efficacy) Systematic literature Four classes of The Not reported were not clearly for critically ill Risk and burden search for published evidence were recommendations defined. However, patients Cost systematic reviews, used: were developed this can be meta-analyses and A: RCT or through a voting deducted from the health technology upgraded process to achieve reported that this assessment. observational data consensus. can any medical staff who is in B: Downgraded Two grades for the charge or contribute RCT or upgraded strength of in the management observational data recommendations of fluid resuscitation were used: of critically ill C: Well-done Grade 1: Strong patients observational Grade 2: weak studies or down- graded RCT The grades were attributed for D: Down-graded individual outcomes observational or of interest and not RCT data for individual studies

Hydroxyethyl Starch versus Other Plasma Volume Expanders 19

2. Sepsis Patients Table 5. Characteristics of the Included Systematic Reviews on Sepsis Patients Objectives/Scope Type of primary Population/ Intervention Comparator Outcomes Notes studies Medical context 1/2. Patel et al. 201325 – Meta-analysis To evaluate the RCTs only Patients diagnosed HES 130/0.40 – 5 Colloids (non-HES) • 90-day mortality • For trials that had safety of 6% • A total of 6 trials with severe sepsis trials • Albumin – 2 trials • Overall mortality more than one hydroxyethyl starch • Trials were treated in HES 130/0.42 – 1 at final follow-up comparator arms, when used in the published critical/intensive trial Crystalloids • 28-day mortality only one arm was treatment of severe between 2006 care unit setting. • 0.9% saline – 3 • RRT at any time pooled in the sepsis patients and 2012 Fluid therapy was Subgroup analysis trials • AKI analysis. Full • Trials were administered for was based on: • Ringer’s acetate – • Allogeneic explanation for included if they volume expansion • Starch origin: 1 trial transfusion selecting the reported mortality after randomization. o Potato or support comparator arm outcome at 90 o Waxy maize • Adverse events was not provided days, 28 days, or Trials were included • C2/C6 another follow-up if they included hydroxylation ratio: time point severe sepsis o 6:1 exclusively or if they o 9:1 presented data for sepsis patients in subgroup analysis. 2/2. Haase et al. 201326 – Meta-analysis To evaluate the RCTs only Patients diagnosed HES 130/0.38-0.45 Colloids (non-HES) • Overall mortality effects of HES • A total of 9 trials, with sepsis. • HES 130/0.40 – 6 • Albumin – 2 trials • RRT at the end of 130/0.38-0.45 6 of which were trials follow-up patients’ adverse included by Patel A total of 3456 • HES 130/0.42 – 1 Crystalloids • RRT at any time 25 outcomes et al. patients were trial • 0.9% saline – 4 • AKI • Trials were included trials • RBC transfusion published The intervention • Ringer’s acetate – (number of between 2006 Trials were included period ranged from 3 trials patients and total and 2012 if they included 24 hours to >90 volume) sepsis patients days • Blood loss exclusively or if they (number of presented data for patients and total sepsis patients in volume) subgroup analysis • Number of patients having serious adverse events AKI= acute kidney injury; RRT = renal replacement therapy

Hydroxyethyl Starch versus Other Plasma Volume Expanders 20

Table 6. Characteristics of the Included Guidelines on Sepsis Patients Objectives Methodology Intervention and Evidence collection, Recommendations Intended users/ Major Outcomes Evidence Quality Guideline Practice Selection and Formulate and Target population Considered and Strength Validation Considered Synthesis Evaluation Dellinger et al. 201327 – Guidelines on Sepsis Management Nurses and To update the Survival of patients Systematic literature Four classes of The Peer review advanced practice “Surviving Sepsis and length of stay searches and evidence were recommendations process nurses, emergency Campaign in ICU reviews. used: were developed medical Guidelines for Several systematic A: RCT using the Delphi paramedics, health Management of reviews of the and nominal group care providers, Severe Sepsis and literature were B: Downgraded technique to hospitals, physician Septic Shock” last performed to cover RCT or upgraded achieve consensus. and physician published in 2008. different areas of observational assistants/ sepsis management. studies Two grades for the Developers reviewed strength of Adult and pediatric published meta- C: Well-done recommendations patients with severe analyses and observational were used: sepsis or septic systematic reviews as studies Grade 1: Strong shock well. Grade 2: weak D: Case series or expert opinion ICU= intensive care unit;

Table 7. Characteristics of the Included Cost-Effectiveness Analysis on Sepsis Patients Objectives Intervention and Comparator Source of data and Assumptions Method of anlaysis Farrugia et al. 20135 – Cost-effectiveness analysis To evaluate the HES versus Albumin Non-systematic literature search Decision analysis evidence and cost- • The analysis assumed the willingness to pay $10,000 model effectiveness of colloids incremental life-year gain from fluid therapy. used in sepsis • Post-discharge mortality rate of 23% ICU= intensive care unit;

Hydroxyethyl Starch versus Other Plasma Volume Expanders 21

3. Surgery Patients

Table 8. Characteristics of the Included Systematic Reviews on Surgery Patients Objectives/Scope Type of primary Population/ Intervention Comparator Outcomes Notes studies Medical context 1/3. Martin et al. 20136 – Meta-anlaysis To evaluate renal RCTs only Patients undergoing HES 130/0.4 Colloids: • Creatinine The review safety of the third • A total of 17 trials elective surgery. • Gelatin solution – clearance; excluded trials on generation HES • Trials were • A total of 1230 5 trials estimated and kidney transplant preparation in published patients The type of solvent • Human albumin calculated surgery, sepsis or surgery between 2000 was not reported – 3 trials (average of 2 trauma patients. and 2011 The included • HES 200 – 4 days after the The review • The length of surgeries were: The strategy used trials surgery) excluded all trials trials was not • Cardiovascular – for fluid • Acute renal published by Boldt reported 11 trials administration was Crystalloids: failure et al. • Abdominal and not reported • Ringer solution – Two studies had 2 liver – 5 trials 4 trials different • Spine surgery – 1 • Isotonic saline – comparators; the trial 2 trials review pooled data One trial used from the different “standard of care” comparator arms as comparator and compared it to HES 130/0.40 2/3. Navickis et al. 20124 – Meta-analysis To evaluate the RCTs only Adults undergoing HES: Albumin • Cumulative blood The review effects of HES on • A total of 18 trials cardiopulmonary • HES 450/0.7 loss during the excluded trials if bleeding after CPB • Trials were bypass surgery. • HES 200/0.5 first 24 hours patients were surgery published • A total of 970 • HES 130/0.7 after CPB. crossed-over between 1982 patients • HES 130/0.4 • Reoperation for between the and 2008 bleeding and investigated • The length of Indication for colloid The solvent was blood product interventions. trials was not use: saline in all trials transfusion Trials conducted by reported • Volume expansion except one that during the first 24 Boldt et al. were – 9 trials evaluated one arm postoperative excluded due to the • Pump priming – 5 with HES 450/0.7 in hours disclosed scientific trials balanced electrolyte • hemodynamics misconduct. • Both volume • fluid balance expansion and The strategy used • ventilator time The review pump priming – 4 for fluid • intensive care excluded trials if trials administration was unit stay they did not report not reported • mortality blood loss data

Hydroxyethyl Starch versus Other Plasma Volume Expanders 22

Table 8. Characteristics of the Included Systematic Reviews on Surgery Patients Objectives/Scope Type of primary Population/ Intervention Comparator Outcomes Notes studies Medical context CPB= cardiopulmonary bypass; FFP= fresh-frozen plasma; HES= hydroxyethyl starch; RBC= red blood cells; 3/3. Shi et al. 201129 – Meta-analysis To evaluate the RCTs only Patients undergoing HES: Colloids: • Blood loss. The review did not safety of HES • A total of 52 trials cardiovascular • HES 450 • Albumin • Blood transfusion specify the preparations when • Trials were surgery. • HES 400 • Gelatin (volume and exclusion of cross- used in published • A total of 3,234 • HES 264 number of over trials. cardiovascular between 1985 patients • HES 200 Crystalloids: patients) The review included surgery and 2010 • HES 130 • Not described • Reoperation 12 trials conducted • The length of Indication for colloid • HES 120 • Mortality by Boldt et al. trials was not use: Blood-based • Renal function reported • Volume expansion Neither molecular products: • Complications The review – 26 trials substitution nor the • Fresh-frozen excluded trials if • Pump priming – type of solvent were plasma they did not report 15 trials reported blood loss data • Both volume The strategy used expansion and for fluid pump priming – administration was 11 trials not reported CPB= cardiopulmonary bypass; FFP= fresh-frozen plasma; HES= hydroxyethyl starch; RBC= red blood cells;

Hydroxyethyl Starch versus Other Plasma Volume Expanders 23

APPENDIX 3: Critical Appraisal of the Included Systematic Reviews and Guideline

1. Critically Ill Patients

Strengths Limitations Perel et al. 201318 – Meta-analysis on Critically-ill Patientsa Zarychanski et al. 201317 – Meta-analysis on Critically ill Patientsa Bunn et al. 201219 – Meta-analysis on Critically ill patientsa • literature search and data extraction • The review did not evaluate the impact of patients’ medical were conducted by two investigators condition on patients’ outcome. Patients were treated in the • Review outcomes and data to be ICU for different medical conditions, but they were considered extracted were predefined a priori equal in the analysis. • The quality of the included studies • The review pooled different formulations of HES, and did was evaluated not evaluate the effect of these formulations on patients’ • Trials conducted by Boldt et al. were outcomes used as sensitivity analysis criteria • The review did not consider the impact of the solvent used with HES nor the strategy used for fluid administration. These factors might affect patients’ outcomes. • The follow-up period for patients’ outcome assessment were not evaluated. Gattas et al. 201221 – Meta-analysis on Critically ill Patientsa Wiedermann et al. 201220 – Meta-analysis on Critically ill Patientsa • literature search and data extraction • The review did not evaluate the impact of patients’ medical were conducted by two investigators condition on patients’ outcome. Patients were resuscitated for • The quality of the included studies different medical conditions, but they were considered equal was evaluated. in the analysis. • Retracted publications by Boldt. et • The review pooled different formulations of HES with other al. were included as sensitivity analysis colloids or crystalloids, and did not evaluate the effect of these comparators on patients’ outcomes • The review did not consider the impact of the solvent used with HES nor the strategy used for fluid administration. These factors might affect patients’ outcomes. • The follow-up period for patients’ outcome assessment were not evaluated. Dart et al. 201023 – Meta-analysis on Critically ill Patients • literature search and data extraction • The volume and strategy of fluid administration were not were conducted by two investigators taken into consideration in the analysis or results • Review outcomes and data to be interpretation extracted were predefined a priori • The review pooled all HES formulations. The different • The quality of the included studies comparators were pooled as well; these included colloids and was evaluated. crystalloids fluids. This pragmatic analysis might mask • The medical indication for fluid existing differences between the compared fluids in terms of resuscitation was used as subgroups patients’ outcomes. analysis. However, these were only classified to main groups of sepsis and non-sepsis indications. Wiedermann et al. 201022 – Meta-analysis on Critically ill Patients • literature search and data extraction • The patient population was clearly defined in the review. were conducted by all investigators • The review did not considered information on patients’ • Review outcomes and data to be medical condition, fluid administration management, and the extracted were predefined a priori different HES formulations. These factors might affect patients • The quality of the included studies outcomes was evaluated.

Hydroxyethyl Starch versus Other Plasma Volume Expanders 24

Strengths Limitations a Studies were grouped together because of the similarity of their strengths and limitations Reinhart et al. 201224 – Guidelines for the Management of Fluid Resuscitation in Critically ill Patients • The development was based on • Methods used for guideline validation were not reported literature review of published systematic reviews and meta-analyses • The guideline was developed by individuals experienced in intensive and critical care, anesthesiology, and epidemiology • The quality of evidence was evaluated and used to support the strength of recommendations • The source of evidence was cited.

Hydroxyethyl Starch versus Other Plasma Volume Expanders 25

2. Sepsis Patients

Strengths Limitations Patel et al. 201325 – Meta-analysis on Patients with Severe Spesis • literature search and data extraction were conducted by • For trials that had more than one two investigators comparator arms, only one arm was • Review outcomes and data to be extracted were pooled in the analysis. Full predefined a priori explanation for selecting the • The quality of the included studies was evaluated. comparator arm was not provided. • The length of follow-up period was considered in the analysis • The review evaluated starch characteristics such as the origin and C2/C6 hydroxylation ratio; these factors were believed to affect patients outcomes • The dose and total volume of fluid administration were reported. However, this information was not taken into consideration in the analysis or results interpretation

Haase et al. 201326 – Meta-analysis on Patients with Sepsis • literature search and data extraction were conducted by • No major limitations two investigators • Review outcomes and data to be extracted were predefined a priori • The quality of the included studies was evaluated. • The length of follow-up period was considered in the analysis • The dose and total volume of fluid administration were reported. However, this information was not taken into consideration in the analysis or results interpretation Dellinger et al. 201327 – Guidelines on Sepsis Management • The development was based on several systematic • No major limitations reviews of the literature and previous meta-analysis • The guideline was developed by individuals experienced in intensive and critical care, emergency medicine, microbiology and infectious diseases, and sepsis management • The quality of evidence was evaluated and used to support the strength of recommendations • The source of evidence was cited, but it was not listed in full. Therefore, the inclusion of results from the most recent RCTs could not be verified. Farrugia et al. 20135 – Cost-effectiveness analysis • The analysis used decision analysis model • The analysis was not based on a systematic review of the literature • Costs and burden attributed renal replacement therapy was not reported. • The time horizon of costs and benefits were not reported • Statistical testing for cost- effectiveness the estimates was not reported

Hydroxyethyl Starch versus Other Plasma Volume Expanders 26

3. Surgery Patients

Strengths Limitations Navickis et al. 20124 – MA on Cardiac Surgery Patients • literature search and data extraction were • The quality of the included trials was not conducted by two investigators evaluated or considered in the interpretation of • Review outcomes and data to be extracted results were predefined a priori • The review included trials that used colloids for • The review investigators contacted trials different reasons/ indications; however, the authors for additional data systematic review did not evaluate the effect of each indication on the measured outcomes. • The review did not report the length of the included trials and the follow-up periods. • The review pooled the results of two arms from one trial. One arm evaluated HES 450/0.7 in balanced electrolyte; the other arm used HES 450/0.7 in saline. Pooling decision was justified by the equivalence of the mean chest tube draining. However, the mean chest tube drain was not justified or defended as an appropriate surrogate for the clinical outcomes. • The volume and strategy of fluid administration were not taken into consideration in the analysis or results interpretation • The review reported the means and standard deviations of several outcomes. These measures showed signs of severe skewness.31 The use of skewed data in the meta-analysis without adjustment is a methodological flaw. Shi et al. 201129 – MA on Cardiac Surgery Patients • literature search and data extraction were • The review included trials that used colloids for conducted by two investigators different reasons/ indications; however, the • Review outcomes and data to be extracted systematic review did not evaluate the effect of were predefined a priori each indication on the measured outcomes. • The quality of the included studies was • The review did not report the length of the evaluated. This evaluation was used a included trials and the follow-up periods. sensitivity analysis during the result • The volume and strategy of fluid administration interpretation. were not taken into consideration in the analysis or results interpretation • The molecular substitution and the carrier solvents were not reported in the review. These factors might affect the fluid behavior and patients’ outcomes. • The review included 12 trials that present high risk of data falsification. • The review reported the means and standard deviations of several outcomes. These measures showed signs of severe skewness.31 The use of skewed data in the meta-analysis without adjustment is a methodological flaw.

Hydroxyethyl Starch versus Other Plasma Volume Expanders 27

Strengths Limitations Martin et al. 20136 – Meta-analysis on Surgery Patients • literature search and data extraction were • The quality of the included trials was not conducted by all authors evaluated or considered in the interpretation of • Review outcomes and data to be extracted results were predefined a priori • The review did not report the length of the • The review excluded all trials published by included trials and the follow-up periods. Boldt et al. to avoid the potential of biased data. • The volume and strategy of fluid administration were not taken into consideration in the analysis or results interpretation • The review compared HES 130/0.4 with other colloids including a hydroxyethyl starch and crystalloids. One trial had “standard of care” as comparator. The pooling of all these comparators might bias the results toward the null difference because of the inclusion a hydroxyethyl starch (HES 200). This starch has the potential to mimic the behavior of HES 130/0.4 due to the similar. Furthermore, the generalizability of these comparators was compromised by the ambiguous standard of care that was used as comparator in one trial. • The comparisons between HES 130/0.4 and the comparators in terms of serum creatinine, calculated creatinine, and urea were made for the value registered at baseline and for the most extreme values. The analysis did not compare the change between baseline and extreme values. This analysis makes the differences between groups at baseline, and can be considered as cross-sectional comparison rather than prospective randomized cohort comparison.

Hydroxyethyl Starch versus Other Plasma Volume Expanders 28

APPENDIX 4: Main Study Findings and Authors’ Conclusions

1. Critically Ill Patients

Table 9. Summary of Findings from Studies on Critically Ill Patients (Seven Meta-analyses) Main Study Findings Conclusions 1/7. Zarychanski et al. 201317 – Meta-analysis on Critically ill Patients • A total of 38 trials, were included in the review; of which, 10 trials were included in the The authors concluded that meta-analysis of Perel et al.18 the use of HES for • A total of 10,880 patients contributed to mortality analysis trials evaluated HES vs. resuscitating critically ill crystalloids. patients did not improve • Different HES formulations ranged in molecular weight from 450 to 120 kD; these patients’ survival. When Boldt formulations were pooled together. No subgroup analysis was performed. et al.’s trials were excluded, • HES was compared to different colloids and crystalloids; these comparators were HES was associated with pooled with no subgroup analysis increased risk of mortality and • The length of follow-up was not considered. acute kidney injury. • The strategy of fluid administration was not considered in the analysis. The total volume and rate of fluid administration were not evaluated. Reviewer’s comments: • • Fluids were used for resuscitation of patients with sepsis, burns, hypovolemia, or The medical condition, for trauma. These groups were analyzed collectively. which patients were resuscitated, was not Outcome/ No. of No. of HES vs. Comparators subgroup trials patients MDa or RR (95% CI) considered in the outcome analysis. Mortality – all studies 35 10880 RR: 1.07 (1.00, 1.14) • The analysis did not • Boldt’s studies removed 28 10290 RR: 1.09 (1.02, 1.17) consider the length of trials • 7 590 RR: 0.91 (0.74, 1.12) Boldt’s studies alone and follow-up periods. Renal replacement therapy 10 9258 RR: 1.32 (1.15, 1.50) The strategy and volume of RIFLE criteria fluid administration was not • Risk 4 8111 RR: 0.89 (0.77, 1.03) considered in the analysis. • Injury 4 8111 RR: 0.93 (0.81, 1.06) These factors might affect • Failure, acute 5 8725 RR: 1.27 (1.09, 1.47) patients’ outcomes. • Failure, chronic 1 196 RR: 0.69 (0.23, 2.09) • Loss 3 1038 RR: 0.87 (0.34, 2.22) • End-stage 3 1038 OR: 1.00 (0.06, 15.9) Urine output, mL 11 6909 SMDb: -0.15 (-0.19, -0.1) ICU LoS, days 5 7016 MD: 0.35 (-0.20, 0.90) Overall LoS, days 6 7337 MD: 0.32 (-0.45, 0.09) Major hemorrhage 1 800 RR: 1.52 (0.94, 2.47) RBC transfusion, patients RR: 1.42 (1.15, 1.75) RBC transfusion, volume MD: 46.3 (-610.8, 703.4) a the difference was based on (HES – comparator); positive MD favors comparator b positive SMD favors HES CI= confidence interval; HES= hydroxyethyl starch; MD= mean difference; SMD= standardized mean difference 2/7. Perel et al. 201318 – Meta-analysis on Critically Ill Patients • A total of 56 trials, were included in the review; of which, 25 trials evaluated HES vs. The authors concluded that crystalloids. HES trials included total of 9174 patients the use of HES might • The length of follow-up was not considered. increase mortality. • The strategy of fluid administration was not considered in the analysis. The total volume and rate of fluid administration were not evaluated. Reviewer’s comments: • Fluids were used for resuscitation of patients with trauma, burns or surgery. These Same as per comments on groups were analyzed collectively. the review by Zarychanski et al.17 No. of No. of HES vs. Crystalloids Mortality trials patients RR (95% CI) All studies included 25 9174 1.10 (1.02, 1.19) Boldt et al.’s studies 21 8976 Not changed excluded

Hydroxyethyl Starch versus Other Plasma Volume Expanders 29

Table 9. Summary of Findings from Studies on Critically Ill Patients (Seven Meta-analyses) Main Study Findings Conclusions HES= hydroxyethyl starch 3/7. Bunn et al. 201219 – Meta-analysis on Critically ill patients • A total of 86 trials were included; 53 trials compared HES with other colloids The authors concluded that • there was no evidence that • Different HES formulations ranged in molecular weight from 450 to 120 kD; these one colloid is more effective formulations were pooled together. No subgroup analysis was performed. of safe than any other. • The length of follow-up was not considered. • The strategy of fluid administration was not considered in the analysis. The total Reviewer’s comments: volume and rate of fluid administration were not evaluated. • The medical condition, for • Fluids were used for volume expansion; the medical conditions requiring volume which patients were expansion was not considered in the analysis resuscitated, was not considered in the outcome Comparisons – Relative Risk (95% CI) analysis. • Outcome Albumin or plasma protien The analysis did not Gelatin vs. HES vs. HES consider the length of trials 31 trials/ n=1719 22 trials/ n= 1612 and follow-up periods. Mortality 1.06 (0.86, 1.31) 1.02 (0.84, 1.26) • The strategy and volume of HES= hydroxyethyl starch fluid administration was not considered in the analysis. These factors might affect patients’ outcomes. 4/7. Gattas et al. 201221 – Meta-analysis on Critically ill Patients • A total of 36 trials, were included in the review; 16 of which contributed to the mortality The authors concluded that analysis the included trials were of • Trials were published between 2000 and 2010 poor quality and reported too • A total of 1582 patients contributed to mortality analysis few events to reliably • The length of follow-up was not considered. evaluate the benefits/ risk of • The strategy of fluid administration was not considered in the analysis. The total HES 130/0.4 volume and rate of fluid administration were not evaluated. • Fluids resuscitation was indicated for different medical conditions; these conditions Reviewer’s comments: were not considered in the analysis Same as per comments on Bunn et al.19 Outcome/ No. of No. of HES vs. Comparators Subgroup trials patients RR (95% CI) Mortality – all studies 25 1582 RR: 0.92 (0.63, 1.34) • Retracted Boldt’s 16 1063 RR: 0.95 (0.64, 1.42) studies removed • Retracted Boldt’s 11 519 RR: 0.73 (0.24, 2.16) studies alone CI= confidence interval; HES= hydroxyethyl starch

5/7. Wiedermann et al. 201220 – Meta-analysis on Critically ill Patients • A total of 13 trials, were included in the review; trials were published between 2000 and Authors concluded that the 2012. available evidence suggested o The review included the FIRST and CRYSMAS trials; these two trials were not a trend toward higher included in the meta-analysis by Gatta et al.30 mortality among HES 130/0.4 o The review excluded all publications by Boldt et al. recipients • A total of 1131 patients contributed to mortality analysis • The length of follow-up was not considered. • The strategy of fluid administration was not considered in the analysis. The total Reviewer’s comments: Same as per comments on volume and rate of fluid administration were not evaluated. 19 • Fluids resuscitation was indicated for different medical conditions; these conditions Bunn et al. were not considered in the analysis Outcome/ No. of No. of HES vs. Comparators subgroup trials patients RR (95% CI) Mortality 13 1131 RR: 1.25 (0.98, 1.58) CI= confidence interval; HES= hydroxyethyl starch

Hydroxyethyl Starch versus Other Plasma Volume Expanders 30

Table 9. Summary of Findings from Studies on Critically Ill Patients (Seven Meta-analyses) Main Study Findings Conclusions

6/7. Dart et al. 201023 – Meta-analysis on Critically ill Patients • A total of 34 trials; these were published between 1982 and 2008 Authors concluded that the • A total of 2607 patients contributed to kidney failure analysis potential for increased risk of • The length of follow-up was not considered. acute kidney failure should be • The strategy of fluid administration was not considered in the analysis. The total considered when using HES volume and rate of fluid administration were not evaluated. fluids for resuscitation, • The review evaluated the effect of the medical conditions requiring fluid resuscitation; particularly in septic patients. these were classified as sepsis/ non-sepsis and were used in subgroup analysis

Outcome/ No. of No. of HES vs. Comparators Subgroup trials patients MD or RR (95% CI) Renal replacement therapy 12 1236 RR: 1.38 (0.89, 2.16) • Non-sepsis 8 487 RR: 0.44 (0.14, 1.38) • Sepsis 3 702 RR: 1.59 (1.20, 2.10) • Deceased organ donor 1 47 RR: 6.67 (0.92, 48.45) RIFLE criteria 1. Risk or worse 4 325 RR: 1.21 (0.81, 1.80) • Non-sepsis 2 185 RR: 0.88 (0.27, 2.85) • Sepsis 2 140 RR: 1.28 (0.81, 2.02) 2. Injury or worse 4 325 RR: 1.34 (0.83, 2.15) • Non-sepsis 2 185 RR: 0.81 (0.12, 5.40) • Sepsis 2 140 RR: 1.39 (0.84, 2.30) 3. Failure 4 325 RR: 1.33 (0.75, 2.36) • Non-sepsis 2 185 RR: 0.49 (0.07, 3.73) • Sepsis 2 140 RR: 1.45 (0.80, 2.64) Kidney failure 8 1199 RR: 1.50 (1.20, 1.87) • Non-sepsis 5 367 RR: 1.13 (0.57, 2.25) • Sepsis 4 832 RR: 1.55 (1.22, 1.96) Creatinine clearance 3 199 MD: 2.33 (-6.01, 10.67) Creatinine – 1 day 15 1084 MD: -2.29 (-6.64, 2.07) postoperative (by fluid) • HES vs. albumin 8 646 MD: -2.82 (-8.38, 2.74) • HES vs. gelatine 6 418 MD: -3.28 (-10.88, 4.31) • HES vs. crystalloids 1 20 MD: 19.0 (-3.86, 41.86) Creatinine – 1 day 15 1084 MD: -2.29 (-6.64, 2.07) postoperative (by patient) • Non-sepsis 14 914 MD: -2.06 (-6.58, 2.47) • Sepsis 2 170 MD: -5.73 (-21.95, 10.49) 7/7. Wiedermann et al. 201022 – Meta-analysis on Critically ill Patients • A total of 11 trials, were included in the review; 4 trials compared hyperoncotic HES Authors concluded that the with other fluids available evidence suggested • Trials were published between 1981 and 2008 that hyperoncotic HES might • A total of 696 patients were included in the HES trials increase the incidence of acute kidney injury and death. Hyperoncotic HES vs. Outcome/ No. of No. of Comparators Subgroup trials patients OR (95% CI) Acute kidney injury 4 696 OR: 1.92 (1.31, 2.81) Mortality 4 696 OR: 1.41 (1.01, 1.96) HES= hydroxyethyl starch; OR= odd ratio

Hydroxyethyl Starch versus Other Plasma Volume Expanders 31

Table 10. Summary of Findings from Guidelines on the Management of Critically Ill Patients Main Study Findings Conclusions Reinhart et al. 201224 – Guidelines for the Management of Fluid Resuscitation in Critically ill Patients Relevant Recommendations: Authors concluded that colloids were not • The use of HES with molecular weight of ≥200 kDa and/or degree of substitution recommended to be used in >0.4 is not recommended in patients with sepsis (grade 1B): Strong patients with head injury, and recommendation based on downgraded RCT or upgraded observational studies they suggested not use hyperoncotic solutions for • The use of HES is not recommended in ICU patients with increased risk of acute fluid resuscitation. kidney injury (grade 1C): Strong recommendation based well-done observational studies or down-graded RCT

• The guideline recommended that the use of HES 130/0.4 should to be restricted to clinical research context for severe sepsis patients and other ICU patients with increased risk of acute kidney injury or bleeding (grade 1C): Strong recommendation based well-done observational studies or down-graded RCT

• The use of HES or gelatin in organ donors is not recommended outside the context of clinical trials (grade 1C): Strong recommendation based well-done observational studies or down-graded RCT

Hydroxyethyl Starch versus Other Plasma Volume Expanders 32

2. Sepsis Patients

Table 11. Summary of Findings from Studies on Patients with Sepsis (Two Meta-analyses) Main Study Findings Conclusions 1/2. Patel et al. 201325 – Meta-analysis on Severe Sepsis Patients • A total of 6 trials and 3033 sepsis patients were reviewed. Of which, 1937 patients The authors concluded that were reported in one trial; these patients were a mix of severe sepsis and septic shock the overall and 90-day patients mortality were significantly • The length of follow-up was <6 days in two trials, 28 days in one trial, and 90 days in higher in severe sepsis three trials patients receiving 6% • The median hydroxyethyl starch dose exposure was 37.4 ml/kg (range 30-43 ml/kg) or tetrastarch 130 kDa solutions 14 ml/kg/day (range 8-37 ml/kg/day) as part of the initial fluid • Fluid administration ranged from 1 to 5 days; the median was 3.5 days. resuscitation. HES 130/0.4 vs. Comparators – Relative Risk (95% CI) Outcome/ Colloids & subgroups Colloids Crystalloids crystalloids combined 3 trials/ n= 2913 90-day mortality 1.13 (1.02, 1.25) Maize-derived HES 2 trials/ n= 2115 No trials Same as previous (C2/C6 ratio 9:1) 1.09 (0.94, 1.26) Potato-derived HES 1 trial/ n= 798 (C2/C6 ratio 6:1) 1.17 (1.01, 1.36) 1 trial/ n= 56 Not reported 3 trialsa/ n= 1049a 28-day mortality a 1.73 (0.55, 5.47) separately 1.10 (0.94, 1.30) 2 trials/ n= 76 4 trials/ n= 2937 6 trials/ n= 3013 Overall mortality 1.53 (0.66, 3.53) 1.12 (1.01, 1.24) 1.13 (1.02, 1.25) Maize-derived HES 2 trials/ n= 76 3 trials/ n= 2139 5 trials/ n= 2215 (C2/C6 ratio 9:1) 1.53 (0.66, 3.53) Not reported 1.09 (0.94, 1.25) Potato-derived HES 1 trial/ n= 798 No trials Same as previous (C2/C6 ratio 6:1) 1.17 (1.01, 1.36)

2 trialsa/ n= 994a Renal replacement No trials a Same as previous 1.41 (1.08, 1.84) Allogeneic 2 trials/ n= 972 No trials Same as previous transfusion 1.21 (1.08, 1.36) a CHEST trial was not included in the analysis because it reported results for all patients enrolled in to the sepsis and non-sepsis groups. 2/2. Haase et al. 201326 – Meta-analysis on Patients with Sepsis • A total of 9 trials were included; 6 trials were already included by Patel et al.25 The authors concluded that • A total of 3456 sepsis patients were included in these trials. HES 130/0.38-0.45, • The length of intervention ranged was <6 days in 6 trials, up to 90 days in two trials, compared with crystalloid or and for one trial this information was not reported albumin, increased the use of • The total HES dose ranged from 2104 ml to 6383 ml RRT and transfusion with red • The review pooled the results of all comparators; these included albumin, Ringer’s blood cells. solution, and isotonic saline Outcome/ No. of No. of HES vs. Comparators Subgroups trials patients MD or RR (95% CI) All-cause mortality 8 3414 RR: 1.04 (0.89, 1.22) <28-day follow-up 4 258 RR: 0.63 (0.35, 1.15) >28-day follow-up 4 3156 RR: 1.11 (1.01, 1.22) RRT at any time 5 NR RR: 1.36 (1.08, 1.72) Acute kidney injury 4 997 RR: 1.18 (0.99, 1.40) RBC transfusion (patients) RR: 1.29 (1.13, 1.48) 3 973 RBC transfusion (volume) MD: 65 ml (-20, 149) Bleeding (patients) RR: 1.34 (0.81, 2.21) 2 993 Bleeding (volume) MD: 26 ml (-89, 140) HES= hydroxyethyl starch; MD= mean difference; RBC= red blood cells; RR= relative risk; RRT= renal replacement therapy Table 12. Summary of Findings from Guidelines on Patients with Sepsis

Hydroxyethyl Starch versus Other Plasma Volume Expanders 33

Main Study Findings Conclusions Dellinger et al. 201327 – Guidelines on Sepsis Management Relevant Recommendations: The authors concluded that there were strong agreement Fluid Therapy of Severe Sepsis among a large cohort of • Crystalloids as the initial fluid of choice in the resuscitation of severe sepsis and international experts septic shock (grade 1B): Strong recommendation based on downgraded RCT or regarding many strong upgraded observational studies recommendations for the best care of patients with severe • The guideline recommended against the use of hydroxyethyl starches for fluid sepsis resuscitation of severe sepsis and septic shock (grade 1B): Strong recommendation based on the results of the VISEP, CRYSTMAS, 6S, and CHEST trials. The results of CRYSTAL trial were not considered.

• Albumin in the fluid resuscitation of severe sepsis and septic shock when patients require substantial amounts of crystalloids (grade 2C): Weak recommendation based on well-done observational studies

Table 13. Summary of Findings from Cost-effectiveness Analysis on Patients with Sepsis Main Study Findings Conclusions Farrugia et al. 20135 – Cost-effectiveness analysis • The analysis considered 15 RCTs that reported renal replacement therapy, and it Authors concluded that the considered four studies for mortality use of albumin in septic patients, when colloid fluid therapy is indicated, shows superior cost-effectiveness to Outcome HES Albumin HES when all medical costs Life-years gained -0.69 0.84 incurred from therapy Incremental life-year -1.53 associated adverse events Total medical costs $48,488 $20,403 are considered. Incremental costs $28,085 ICER Dominant ICER= incremental cost-effectiveness ratio

Hydroxyethyl Starch versus Other Plasma Volume Expanders 34

3. Surgery Patients

Table 14. Summary of Findings from Studies on Patients Undergoing Surgery (Three Meta-analyses) Main Study Findings Conclusions 1/3. Navickis et al. 20124 – MA on Cardiac Surgery Patients • A total of 18 trials were included. The authors of the systematic • A median of 48 patients were included per trial; the total number was 970 patients review concluded that HES • The length of follow-up was not provided was found to increase • Indication for colloid use: postoperative bleeding, reoperation due to bleeding, o Volume expansion – 9 trials; and blood product transfusion o Pump priming – 5 trials; when used for patients o Both volume expansion and pump priming – 4 trials • The review pooled the results of two arms from one trial. One arm evaluated HES undergoing cardiopulmonary 450/0.7 in balanced electrolyte; the other arm used HES 450/0.7 in saline. Pooling bypass. The authors decision was justified by the equivalence of the mean chest tube draining. recommended that HES • The strategy of fluid administration was not reported. The total volume and rate of fluid should be avoided or used administration were not reported with cautious.

Outcome Comparisons Reviewer’s comments: HES 450/0.7 vs. HES 200/0.5 vs. HESa vs. Albumin The main objective of the Albumin Albumin systematic review was to Efficacy Outcomes evaluate bleeding after CPB surgery. Fluid bal. (mL): 2 trials 4 trials 6 trials b The reported results showed MD (95% CI) 212 (-361, 785) -105 (-621, 411) 37 (-347, 420) inconsistent signs of efficacy Heart rate (bt/min): No data was 5 trials 6 trials b and safety issues. The results MD (95% CI) presented -2.3 (-5.5, 0.8) -3.2 (6.0, -0.5) were either favoring albumin Cardiac index: 3 trials 5 trials 8 trials c over HES or showing no MD (95% CI) -0.06 (-0.38, 0.25) -0.14 (-0.26, -0.03) -0.17 (-0.24, -0.11) statistical difference between MAP (mmHg): No data was 4 trials 4 trials c the two interventions. MD (95% CI) presented -2.9 (-6.3, 0.4) -2.9 (-6.3, 0.4) CVP (mmHg): No data was 5 trials 5 trials c Results should be interpreted MD (95% CI) presented -0.97 (-1.82, -0.12) -0.97 (-1.82, -0.12) in light of the fact that the Safety Outcomes systematic review excluded Bleeding (vol): 9 trials 6 trials 15 trials trials that did not report blood SMD (95% CI)b 0.36 (0.17, 0.55) 0.285 (0.04, 0.53) 0.333 (0.18, 0.48) loss data; therefore the Reoperations (n): 4 trials/ 281 patients 3 trials/ 117 patients 7 trials/ 398 patients efficacy results can’t be RR (95% CI) 2.13 (0.84, 5.38) 2.38 (0.89, 6.38) 2.24 (1.14, 4.40) considered as comprehensive RBC transf. (vol.): 8 trials 5 trials 13 trials of all trials conducted in this SMD (95% CI)b 0.340 (0.14, 0.54) 0.181 (-0.09, 0.45) 0.284 (0.12, 0.45) type of population.

FFP transf. (vol.): 4 trials 3 trials 7 trials b The rates of bleeding events SMD (95% CI) 0.256 (-0.006, 0.52) 0.446 (0.006, 0.89) 0.306 (0.08, 0.53) or blood product Platelets (vol.): 2 trials 2 trials 4 trials b administration was not SMD (95% CI) 0.325 (0.02, 0.63) 0.224 (-0.29, 0.74) 0.298 (0.03, 0.56) reported or compared. This Ventilator time (h): 3 trials 2 trials 5 trials kind of analysis gives more MD (95% CI)b 0.79 (-0.21, 1.8) 0.86 (-2.74, 4.46) 0.80 (0.17, 1.76) insights at the patient’s level. ICU stay (d): 5 trials 2 trials 7 trials MD (95% CI)b -0.04 (-0.17, 0.09) 0.51 (0.05, 0.96) 0.00 (-0.12, 0.12) Mortality (n): 2 trials No data was 3 trials RR (95% CI) 0.71 (0.16, 3.12) presented 0.99 (0.27, 3.57) a both HES 450/0.7 and HES 200/0.5 were pooled against albumin b the difference was based on (HES – albumin); positive SMD favors albumin c the difference was based on (HES – albumin); negative SMD favors albumin bal.= balance; bt/min= beats/minute; CI= confidence interval; CVP= central venous pressure; d= days; h= hours; HES= hydroxyethyl starch; MAP= mean arterial pressure; SMD= standardized mean difference; vol.= volume

Hydroxyethyl Starch versus Other Plasma Volume Expanders 35

Table 14. Summary of Findings from Studies on Patients Undergoing Surgery (Three Meta-analyses) Main Study Findings Conclusions 2/3. Shi et al. 201129 – MA on Cardiac Surgery Patients • A total of 52 trials were included. Nine trials were included in the MA of Navickis et al. The authors concluded that 20124 the administration of HES • The total number was 3,234 patients preparations did not cause • The length of follow-up was not provided more renal damage, • Indication for colloid use: complications, reoperation or mortality than control o Volume expansion – 26 trials; solutions. Perioperative o Pump priming – 15 trials; administration of HES o Both volume expansion and pump priming – 11 trials • The review review included 12 trials conducted by Boldt et al.; these trials have high preparations was risk of falsification comparatively safe in • The strategy of fluid administration was not reported. The total volume and rate of fluid cardiovascular surgery. administration was not reported Reviewer’s Comments: The reported results do not Blood loss; Results presented in terms of SMDa (95% CI) and RRb (95% CI) permit the conclusion of Comparator absolute safety HES HES preparation preparations, for the following Albumin Gelatin Crystalloid c c reasons: HES 450 7 trials 1 trial 5 trials • The inclusion of Boldt et al. SMD (95% CI) 0.47 (0.26, 0.68) 2.07 (1.16, 2.98) 0.22 (-0.10, 0.55) trials RR (95% CI) 1.77 (1.15, 2.73) 1.50 (0.29, 7.73) Not assessable • The results of blood loss HES 400 1 trial 1 trial SMD (95% CI) 0.87 (0.11, 1.62) 0.29 (-0.52, 1.09) were not consistent across RR (95% CI) Not reported Not reported the comparators and the different HES preparations. HES 264 3 trials No data was No data was • SMD (95% CI) 0.18 (-0.11, 0.47) presented for this presented for this It was not clear from the RR (95% CI) Not reported comparison comparison review how blood loss assessment was done in HES 250 No data was 2 trials the included trials. SMD (95% CI) presented for this 0.21 (-0.01, 0.44) Information on when these RR (95% CI) comparison Not reported c c c bleedings occurred and HES 200 7 trials 7 trials 7 trials their cause were not SMD (95% CI) -0.01 (-0.29, 0.28) 0.26 (0.02, 0.49) -0.20 (-0.46, 0.06) reported or discussed. RR (95% CI) 1.23 (0.68, 2.23) 1.4 (0.65, 3.00) 0.58 (0.23, 1.49) c c • The review did not report HES 130 7 trials 10 trials 3 trials on how or when the renal SMD (95% CI) -0.61 (-0.82, -0.40) -2% (-16, 12) -0.19 (-0.45, 0.08) function was assessed. RR (95% CI) 0.77 (0.62, 0.94) 1.03 (0.86, 1.24) 0.67 (0.13, 3.44) Different measured are 1 trial No data was 1 trial HES 120 available for this SMD (95% CI) 0.68 (-0.09, 1.44) presented for this 0.19 (-0.53, 0.91) assessment, and some of RR (95% CI) Not reported comparison Not reported these measures are time- Reoperation: Relative risk (95% CI) dependent. Comparator • The collective evaluation HES preparation Albumin Gelatin Crystalloid (pooling) of all HES HES 450 2.04 (0.52, 8.07) Not assessable 0.36 (0.01, 9.19) preparations against all HES 200 Not assessable 2.25 (0.48, 10.53) 0.32 (0.03, 3.18) comparators is a practical HES 130 0.33 (0.01, 7.89) 1.16 (0.32, 4.16) Not assessable flaw. This is because of the different behaviour of the Renal Function; various HES preparations HES (all preparations) vs. All Comparators as well as the used 17 trials; RR (95 CI) : 0.77 (0.35, 1.70) comparators. Furthermore, Mortality: this flaw was added to the HES (all preparations) vs. All Comparators fact that the volume and 27 trials; of which 15 trials reported no death events strategy used to administer RR (95% CI): 0.59 (0.32, 1.09) the fluids were not reported a difference was based on (HES – Control); positive SMD favors control solution or discussed. Therefore, b Relative risk the exact interventions c Trials from Boldt et al. were included in this analysis evaluated in these trials HES= hydroxyethyl starch; RR= relative risk; SMD= standardized mean difference; can’t be generalised.

Hydroxyethyl Starch versus Other Plasma Volume Expanders 36

Table 14. Summary of Findings from Studies on Patients Undergoing Surgery (Three Meta-analyses) Main Study Findings Conclusions 3/3. Martin et al. 20136 – Meta-analysis on Surgery Patients • A total of 17 trials were included. The authors concluded that • The total number of patients was 1230 there was no verifiable • The length of follow-up was not provided association between the • The included surgeries were: administration of HES 4 130/0.40 and changes of o Cardiovascular – 11 trials/ one trial was included in Navickis et al. , and eight trials were included in Shi et al.29 serum creatinine, calculated creatinine clearance, or the o Abdominal and liver – 5 trials incidence of acute renal o Spine surgery – 1 trial • The review pooled the results of all comparators; these included gelatin, albumin, HES failure 200, Ringer’s solution, isotonic saline and “standard of care”. • The strategy of fluid administration was not reported. The total volume and rate of fluid Reviewer’s comments: administration was not reported. The results of the meta- analysis are limited by the • The review calculated an average effect size calculated from weighted averages (+d) based on the variance of unbiased effect sizes for each trial. This method was following: • described by Hedges and Olkin. However, this effect size was calculated for the The analysis did not baseline values and for the extreme values, but it is was not calculated for the compare the changes from difference between groups in change form baseline until the registration of the extreme baseline between HES values. 130/0.4 and comparators. • The analysis pooled all Outcome/ No. of No. of HES 130/0.4 vs. Comparators a kinds of comparators Time of assessment trials patients +d or RD (95% CI) including one HES Serum creatinine preparation. The analysis • Baseline 13 NR +d: -0.021 (-0.261, 0.219) did not take into • Extreme 14 NR +d: +0.068 (-0.227, 0.362) consideration the class Calculated creatinine effect of starch fluids • Baseline +d: +0.302 (-0.098, 0.703) • The analysis did not NR 344 • Extreme +d: +0.783 (-0.229, 1.795) consider the length of Urea trials and follow-up • Baseline +d: -0.068 (-0.371, 0.236) periods. Renal failure is a NR 390 delayed adverse event • Extreme +d: -0.148 (-1.077, 0.782) which could be missed in the included trials ICU length of stay NR 723 +d: +0.113 (-0.172, 0.398) • The strategy and volume Hospital length of stay NR 940 +d: +0.212 (-0.035, 0.46) of fluid administration was Acute renal failure 8 NR RD: 0.000298 (-0.18, 0.019) not considered in the NR 531 -0.003 (-0.028, 0.022) Renal replacement therapy RD: analysis. These factors a the effect size was based on the difference HES 130/0.4 – comparators; negative sings could affect patients’ favors HES 130/0.4 outcomes. RD= risk difference

Hydroxyethyl Starch versus Other Plasma Volume Expanders 37