Description of the condition
Patients requiring intensive care are defined as critically ill patients (Thomas 2006). These patients are susceptible to anaemia and are vulnerable to its adverse consequences. The prevalence of anaemia among patients admitted to the intensive care unit (ICU) is about 67% (Corwin 2004); anaemia is frequently treated with allogeneic red blood cell (RBC) transfusions (Corwin 2006).
Allogeneic RBC transfusion has been found to be a risk factor associated with increased risk of morbidity and mortality in critical care patients (Hopewell 2013; Marik 2008); therefore safe substitution represents an urgent need. It is noteworthy that this evidence comes only from observational studies and is prone to confounding by indication. Accordingly, anaemia might well constitute a therapeutic target but also a prognostic factor or even an adaptive beneficial response (Zarychanski 2008). Reduced endogenous production of the hematopoietic growth factor erythropoietin is observed in most ICU patients; consequently, administration of erythropoiesis-stimulating agents (ESAs) has been suggested as a therapeutic option (Zarychanski 2007). The application of ESAs has been further extended in acute cardiovascular and neuronal disorders because the biological role of human erythropoietin and its receptor in tissue outside of the hematopoietic system has been recognized (Maiese 2005).
Description of the intervention
A synthetic form of human erythropoietin, which is produced by recombinant DNA technology, became available in the 1980s (recombinant human erythropoietin (rHuEPO)). Several currently available ESAs such as epoetin-alfa, epoetin-beta, epoetin-omega, epoetin-delta and darbepoetin-alfa are administered as intravenous or subcutaneous injections. These agents are analogues of recombinant erythropoietin with the same amino acid sequence, but their glycosylation pattern varies because of type- and host cell–specific differences in the production process. Darbepoetin-alfa is a novel erythropoiesis-stimulating protein that carries two additional glycosylation sites and produces a longer half-life and increased potency (Goldsmith 2010). ESAs are among the most widely manufactured recombinant bio-similar proteins in the world (Walsh 2010).
ESAs are approved for treatment of anaemia caused by end-stage renal disease, anaemia associated with human immunodeficiency virus infection and anaemia that occurs with non-myeloid cancers in which anaemia is due to concomitantly administered chemotherapy used to reduce the number of transfusions in patients scheduled for major surgery, except cardiac or vascular surgery (Fishbane 2010). Administration of ESAs to critically ill patients with none of the aforementioned conditions is outside the license of these agents and can be considered as an off-label indication.
Vekeman and co-workers conducted two retrospective analyses of data from participants admitted to more than 500 hospitals across the United States and treated with ESAs from July 2002 to March 2005. These investigators identified a total of 72,903 hospitalized patients (epoetin-alfa: 66,804; darbepoetin-alfa: 6099) treated with ESAs in the ICU setting (Vekeman 2006); a total of 25,645 hospitalized patients with cancer (epoetin-alfa: 22,873; darbepoetin-alfa: 2772) and a total of 66,822 hospitalized patients with chronic kidney disease (CKD) (epoetin-alfa: 60,079; darbepoetin-alfa: 6743) (Vekeman 2007). This study, which was conducted to evaluate erythropoiesis-stimulating protein use and clinical outcomes in hospitalized patients (ASSESS), showed that ESAs were used among a heterogeneous group of study participants with anaemia in the ICU at 19 sites across the United States in 2005 (Brophy 2008). On the basis of the overall cumulative dose per ICU stay, the costs of treatment of critically ill patients with erythropoietin-alfa and darbepoetin were estimated at approximately US$576 and US$841, respectively (Vekeman 2006).
How the intervention might work
Erythropoietin (EPO) is an endogenous glycoprotein hormone that is mainly produced in the kidneys under hypoxic conditions. EPO is well known for its role in production and differentiation of erythroid progenitor cells. However, a variety of non-haematopoietic activities of EPO have been identified (Chateauvieux 2011). Anaemia, which is common in critically ill patients, may be associated with a blunted EPO response (Corwin 2004; DeAngelo 2005; Hobisch-Hagen 2001; Krafte-Jacobs 1994; Rogiers 1997) and subsequently with inadequate stimulation of bone marrow erythropoiesis. Therefore, exogenous EPO therapy could result in higher haemoglobin concentrations and less exposure to RBC transfusion.
The results of a growing list of controlled trials show that ESA treatment might be not as safe as was first assumed. Systematic reviews of on-label use of ESAs have raised concerns about safety and increasing mortality (Bohlius 2009; Bohlius 2012; Phrommintikul 2007; Tonelli 2009). A meta-analysis of 27 randomized, controlled trials (RCTs) involving 10,452 participants with CKD concluded that targeting higher haemoglobin concentration increases risks for fatal and non-fatal stroke, hypertension and vascular access thrombosis compared with targeting lower haemoglobin concentration (Palmer 2010). A meta-analysis that included 52 RCTs (n = 12,006) found that ESAs increased the risk of thrombotic events (risk ratio (RR) 1.69, 95% confidence interval (CI) 1.27 to 2.24) and serious adverse events (RR 1.16, 95% CI 1.08 to 1.25) in participants with cancer-related anaemia. However, ESAs improved some disease-specific measures of quality of life and decreased the use of blood transfusions. Therefore, routine use of ESAs in patients with cancer-related anaemia as an alternative to blood transfusion is not recommended (Tonelli 2009). A randomized, double-blind, placebo-controlled, multicentre trial (N = 1460) in anaemic critically ill participants demonstrated an increased incidence of thrombotic vascular events among participants in the epoetin-alfa group as compared with those in the placebo group (16.5% vs 11.5%; hazard ratio 1.41; 95% CI 1.06 to 1.86) (Corwin 2007). Moreover, the US Food and Drug Administration (FDA) restricted the use and prescribing of ESAs under a risk management programme known as a risk evaluation and mitigation strategy (REMS) following studies showing that ESAs can increase the risk of tumour growth and can shorten survival in patients with cancer (FDA 2010).
Why it is important to do this review
Biological and pharmacological effects of drugs largely vary across several medical conditions. Beneficial and adverse effects of ESAs have been studied extensively in certain clinical settings (Aher 2012; Bohlius 2012; Cody 2005; Lambin 2009; Martí-Carvajal 2013; Ngo 2010; Tonia 2012), but we cannot necessarily extrapolate study results from one population to another. On the other hand, ESAs are frequently used beyond their label indications in critically ill patients (Vekeman 2006; Vekeman 2007), although not only drug cost but also economic burden resulting from adverse drug reactions may substantially impact health systems. Therefore, in the absence of an unequivocal benefit, potential harm in this vulnerable population needs to be systematically assessed and analysed and the findings made readily available for evidence-based practice. We recently published a safety review in this field, but given that eight studies have been published since 2010 (Mesgarpour 2013b) and several studies are ongoing (Mesgarpour 2013a), a regularly updated safety review within The Cochrane Library is an obvious measure.
With a focus on safety, we aim to assess the effects of ESAs (ESAs alone or in combination) compared with placebo, no treatment or a different active treatment regimen when administered off-label to critically ill patients. We will further describe and explore heterogeneity and will assess the influence of bias on the robustness of our effect estimates.
Criteria for considering studies for this review
Types of studies
We will include RCTs and controlled observational studies (cohort or case-control) investigating the effects of ESAs given for the treatment of any kind of critical illness unless its indication was approved by the European Medicine Agency (EMA) or the FDA (see Background and Types of interventions).
Types of participants
We will consider studies with acutely and critically ill adults and children. We will exclude studies looking at neonates and infants (one month to one year of age) because erythropoietin-beta is approved by the EMA for preventing anaemia in premature babies. We will exclude animal experiments. An expert in intensive care medicine will assess the setting of “critical illness” in the absence of a stringent definition of the condition.
Types of interventions
The intervention is the scheduled systemic administration of ESAs versus placebo, no treatment or any alternative active treatment. We consider administration of ESAs as on-label if used to treat anaemia due to CKD, anaemia due to chemotherapy in patients with cancer or anaemia due to zidovudine in HIV-infected patients. Some discrepancy has been noted between the EMA and the FDA regarding the indication for ESAs to reduce allogeneic RBC transfusions in patients undergoing surgery; therefore we consider this indication as off-label when a study was done in a critically ill population.
Types of outcome measures
- We will identify adverse effects of ESAs in a broad sweep scope, as suggested by Loke 2007.
- We will report all adverse events as reported by study authors and will specifically include any adverse event, any serious adverse event, thrombotic events (such as venous thromboembolism, deep venous thrombosis, pulmonary embolism, stroke, myocardial infarction, stent thrombosis and intracardiac thrombotic masses), bleeding, cardiovascular events including hypertension, infection and sepsis, respiratory events, neurological events, renal events, gastrointestinal events and musculoskeletal events. We will accept authors' definitions of outcome events.
- We will determine total mortality (in the ICU, in the hospital and three weeks after discharge). If mortality was assessed at several time points in a study, we will use data from the closest follow-up time to 30 days.
Search methods for identification of studies
We will search for eligible studies in the following databases: Ovid SP MEDLINE (from 1948 to date); Ovid SP EMBASE (from 1988 to date); Ovid SP PASCAL (from 1984 to date); Ovid SP All EBM Reviews; Ovid SP International Pharmaceutical Abstracts (from 1966 to date); Ovid SP PsycINFO (from 1806 to date); CINAHL (from 1980); BIOSIS Previews (from 2000 to date); Science Citation Index Expanded (from 1900 to date); Conference Proceedings Citation Index–Science (from 1995 to date) and TOXLINE (from 1965 to date) (BM). All EBM reviews consist of seven resources, including Cochrane Database of Systematic Reviews (CDSR), ACP Journal Club, Database of Abstracts of Reviews of Effects (DARE), Cochrane Central Register of Controlled Trials (CENTRAL), Cochrane Methodology Register (CMR), Health Technology Assessment (HTA) and NHS Economic Evaluation Database (NHSEED).
We will search for key words that describe the condition or the intervention. We will expand the common names of ESAs to trade names and chemical synonyms (Mesgarpour 2013a) to enhance the sensitivity of the search. To increase the objectivity of our search, we will analyse the contents of our non-Cochrane systematic review on ESAs in critically ill patients and its 48 included studies (Mesgarpour 2013b) by using the AntConc freeware concordance program (http://www.antlab.sci.waseda.ac.jp/) (BM). We will not combine the search results with methodological search filters (RCT or observational studies) to obtain a highly sensitive search. The full detailed search strategies for MEDLINE and EMBASE are given in Appendix 1 and Appendix 2, respectively. We will adopt our MEDLINE search strategy for searching all other databases.
Searching other resources
We will search current controlled trials and unpublished studies via the Internet (www.controlledtrials.com) by using the multiple database search option (metaRegister of Controlled Trials) (BM). We will also contact four main manufacturers of ESAs (Amgen, Roche, Janssen-Cilag, Ortho Biotech) (BM). Further, we will undertake backward and forward track citations of relevant studies identified from the initial searches using SciVerse Scopus (BM).
Data collection and analysis
Selection of studies
We will export all search results into the bibliographic software EndNote X5 and will remove duplicates (BM). Two review authors (BM and BH or DR) will independently screen the studies for exclusion, first by title and second by abstract, using a screening algorithm with inclusion and exclusion criteria. For all possibly relevant studies, we will retrieve full papers, and two independent review authors will assess them for inclusion and exclusion criteria. We will resolve any differences in opinion by consensus or by involving a third author (HH). We will record the reasons for exclusion when we exclude an abstract or a full paper.
Data extraction and management
Two review authors (BM and BHH or DR) will summarize data independently onto a structured extraction form based on study design (Appendix 3 and Appendix 4). The forms will be compared, and discrepancies in data extraction will be resolved by discussion or if necessary by a third review author (HH). Data will then be added to RevMan 5.2 (BM and DR).
Data extracted will include study characteristics, verification of eligibility, study design, study population, screening and baseline characteristics of participants, details on intervention and control, outcome measures, adverse events and mortality.
Assessment of risk of bias in included studies
Two review authors (BM and BHH or DR) will independently evaluate the risk of bias of included studies. We will employ The Cochrane Collaboration tool (Higgins 2011) for assessing risk of bias in RCTs. We will abstract whether adequate methods are used to generate a random sequence, whether allocation before assignment is concealed, whether the involved clinical staff are blinded to the intervention, whether the assessor of the outcome is blinded to the intervention, whether the outcome description is adequate, whether the outcome reporting is selective, whether the data were analysed by intention-to-treat and whether the study has received funding support from companies manufacturing ESAs. For assessing risk of bias in observational studies, we will use the Newcastle-Ottawa Scale (NOS) (Wells 2009). For cohort studies, we will abstract whether the exposed cohort was representative, what method was used to select the non-exposed cohort, which method of exposure ascertainment was used, whether the outcome of interest was not present at the start of the studies, whether cohorts were comparable on the basis of the design or analysis, what method of outcome assessment was used and whether follow-up was adequate and was long enough for outcomes to occur. For case-control studies, we will abstract whether the case definition was adequate, whether the cases were representative, what details on selection and definitions of controls should be noted, whether cases and controls were comparable on the basis of the design and analysis, which method of exposure ascertainment was used, whether the same method was used for ascertainment of cases and controls and whether the non-response rate was the same for both groups.
In addition, two review authors (BM and BHH or DR) will independently assess the quality of harms assessment and reporting in included studies using the McMaster Quality Assessment Scale of Harms (McHarm) (Santaguida 2008). We will abstract whether the harms were predefined using standardized or precise definitions, whether serious and severe events were precisely defined, whether the number of deaths in each study group was specified or whether the reason(s) for not specifying them were given, whether the mode of harms collection was specified as active or passive, whether the study specified who collected the harms, whether the study specified the training or background of who ascertained the harms and the timing and frequency of collection of the harms, whether the study used standard scale(s) or checklist(s) for harms collection, whether the study authors specified if the harms reported encompassed all events collected or a selected sample, whether the number of participants that withdrew or were lost to follow-up was specified for each study group, whether the total number of participants affected by harms was specified for each study arm, whether the number of each type of harmful event was specified for each study group and whether the types of analyses undertaken for harms data were specified. We will also create a McHarm graph by using RevMan 5.2. The results will be compared and disagreements will be resolved by discussion or, if necessary, by a third review author (HH). A risk of bias table will be completed for each eligible study and outcome in RevMan 5.2 and will be supplemented by the risk of bias graph for RCTs as well as observational studies.
Measures of treatment effect
For dichotomous data presented as absolute counts or relative frequencies, we will present results as risk ratios with 95% confidence intervals.
Unit of analysis issues
Generally, the unit of analysis will be a single participant. If required and sufficient data are available, we will incorporate properly analysed cross-over and cluster-randomized trials into the meta-analyses using the generic inverse variance method in RevMan 5.2 (Higgins 2011). We will check for unit of analysis errors if cross-over trials or cluster-randomized trials are included and yield appropriate solutions. For included studies with more than two treatment groups, we will group together all experimental groups if clinically sensible and will compare them collectively with the control group (Higgins 2011) or perform pair-wise comparisons against the control but refraining from a meta-analysis across all arms in this situation.
Dealing with missing data
We will contact authors of study reports to request information on missing data, if applicable. We will analyse data as available, and no statistical models for data imputation will be employed.
Assessment of heterogeneity
Data synthesis will be deemed appropriate if clinical heterogeneity and methodological heterogeneity were negligible. We will assess clinical heterogeneity by judging the comparability of populations, conditions and treatments. We will assess methodological heterogeneity by classifying study design types. We will assess statistical heterogeneity in each meta-analysis using the I
Assessment of reporting biases
We will use RevMan 5.2 and the R package meta (Schwarzer 2012) to create funnel plots of standard errors versus effect estimates to assess reporting bias and small-study effects if 10 or more studies are available for each outcome. Asymmetry will be assessed by visual inspection and will be formally tested using the arcsine test proposed by Rucker 2008 for data from studies for which valid n/N data are available, and the Egger test (Egger 1997) when effect estimates with their standard errors are available. We will consider P values less than 0.05 as statistically significant.
Trials with zero reported events in either arm are excluded from the analysis; such trials therefore do not contribute information. A zero-cell correction will be performed when zero events are observed in one of the trial arms. Assuming considerable heterogeneity, we will use random-effects models to calculate summary estimates in a meta-analysis. If no heterogeneity is observed, we will use fixed-effect meta-analysis. Ideally, observational studies and randomized studies should not be different if confounding is handled appropriately. However, confounding could not be excluded in the observational studies. Therefore, we will consider this a relevant source of methodological heterogeneity. To combine data from RCTs and observational studies, we will fit a three-level hierarchical Bayesian model (Prevost 2000; Schmitz 2013; Sutton 2008). This approach allows for between-study variability (in the same way as a classical random-effects model does), together with between-design variability. In this way, the overall estimate makes use of all available information (Higgins 2012). These analyses will be preformed within RevMan 5.2 and openBUGs using R version 2.15.2 and the BRugs package. For very rare events, we will use the Peto odds ratio method (Higgins 2011).
Subgroup analysis and investigation of heterogeneity
We will perform subgroup analyses according to type and dosage of ESAs. Further, we will attempt subgroup analysis according to baseline anaemia or indication as a non-haematopoietic drug. We will use the test for subgroup differences within RevMan 5.2 (Higgins 2011) to formally test for subgroup differences.
We will assess the robustness of our estimates by comparing the effects from models that included all studies (possibly biased but higher precision due to the utilization of all individuals) versus the effects from models that excluded studies with high risk of bias or low quality (potentially lower risk of bias but also lower precision due to the exclusion of studies).
We will compare the estimates of fixed-effect and random-effects meta-analyses to assess our assumptions on heterogeneity.
Acknowledging that observational evidence is of a different nature, a sensitivity analysis down-weighting this in the synthesis will be carried out. This will be done explicitly using a specified parameter that represents the weight given to observational evidence modelled as a multiplicative factor to the observational effect estimate precision. Different weights will be applied to inflate the variance in a sensitivity analysis. We will also assess whether funding sources influenced the estimates if data are available.
Summary of findings
We will use the principles of the GRADE system (Guyatt 2008) to assess the quality of the body of evidence regarding frequent adverse events and mortality associated with off-label use of ESAs in critically ill patients and will construct a Summary of findings (SoF) table using the GRADE software. The GRADE approach appraises the quality of a body of evidence based on the extent to which one can be confident that an estimate of effect or association reflects the item being assessed. The quality of a body of evidence reflects within-study risk of bias (methodological quality), directness of the evidence, heterogeneity of the data, precision of effect estimates and risk of publication bias.
We would like to thank Rodrigo Cavallazzi (content editor), Nathan Pace (statistical editor) and Julia Bohlius, Ryan Zarychanski and Stacy Voils (peer reviewers) for their help and editorial advice during the preparation of this protocol for the systematic review.
Appendix 1. Detailed search strategy for Ovid SP MEDLINE
1 critical care. kw,tw,ti,ot.
2 life support care*.af.
3 (critical* adj ill*).af.
4 (icu or icus).kw,tw,ti,ot.
5 intensive care*.kw,tw,ti,ot,sh.
6 (ventilat*or Respirat* or Artificial).af.
7 exp Respiration, Artificial/
8 acute care facilit*.af.
9 (thermal injur* or burn* or trauma or resuscitation*).kw,tw,ti,ot,sh.
10 ((APACHE or intensive) adj2 score*).af.
11 exp Heart Arrest/dt, th or exp Acute Coronary Syndrome/dt, th or *Myocardial Infarction/dt, th or exp Brain Injuries/dt, th or exp Ventricular Dysfunction, Left/dt, th or exp Ischemia/dt, th or exp Brain Infarction/dt, th or exp Vasospasm, Intracranial/dt, th or exp Intracranial Aneurysm /dt, th or exp Brain Ischemia/dt, th or exp Subarachnoid Hemorrhage/dt, th or exp Myocardial Ischemia/dt, th or exp Cerebral Hemorrhage/dt, th or exp Stroke/dt, th
12 (erythropo* or erithropo* or epo or rheupo or rhepo or rhuepo or Hempoietin* or hematopoiet* or hemopoiet* or epoetin* or darbopoiet* or darbepo*).af.
13 (r adj HuEPO).af.
14 (Abseamed or Alfaepoetina or Aranesp or Betapoietin or Binocrit or Bioetin or Biopoin or Bioyetin or Ceriton or Culat or Dynepo or EPIAO or Epiao or Epoade or Epobel or Epocim or Epoch or Epocomb or Epocrin or Epoetin Alfa Hexal or Epoetin Sedico or Epofer or Epofit or Epoform or Epogen or Epogin or Epogis S or Epoglobin or Epojet or Epokine or Epomax or Epomax or Eponis-2K or Epopen or Eporatio or Eporise or Eporon or Eposim or Eposino or Eposis or Epostim or Epotin or Epotop or Epotrex-NP or Epotrust or Epovitan or Epox or Epoxitin or Epoyet or Eprex or Eralfon or Erantin or Eritina or Eritrelan or Eritrogen or Eritrogen or Eritromax or Eritropoyetina or Eritropoyetina Alfa or Erlan or Erykine or Erypo or Erypoietin or Erypro or Erypro Safe or Erythrostim or Erytrostim or Espo or Espogen or Exetin-A or GerEPO or Globuren or Heberitro or Hemapo or Hemax or Hemax-Eritron or Hemoprex or Hepta or Hypercrit or Jimaixin or LG Espogen or Marogen or Mircera or Negortire or NeoRecormon or Nesp or Nespo or PDpoetin or Procrit or Pronivel or Recormon or Relipoietin or Renogen or Repotin or Retacrit or Sepo or Shanpoietin or Silapo or Tinax or Vero-Epoetin or Vintor or Wepox or Yepotin or YiBei or Zyrop).kw,tw,nm,ti,ot.
15 (Receptors, Erythropoietin or Erythropoiesis).sh.
16 (HM10760A or 64FS3BFH5W or 15UQ94PT4P or BM06019 or DRG-0062 or KRN 321 or KRN321 or KRN 5702 or KRN5702 or BI71052 or TYB 5220 or TYB5220 or SNB 5001 or SNB5001 or HSDB 7584 or HSDB7584 or LS 64697 or LS64697).af.
17 (113427-24-0 or 11096-26-7 or 209810-58-2 or 122312-54-3 or 261356-80-3 or 604802-70-2 or 130455-76-4 or 148363-16-0 or 154725-65-2 or 879555-13-2).rn.
20 18 and 19
Appendix 2. Detailed search strategy for Ovid SP EMBASE
1 critical care. kw,tw,ti,ot.
2 life support care*.af.
3 (critical* adj ill*).af.
4 (icu or icus).kw,tw,ti,ot.
5 (intensive adj2 care*).kw,tw,ti,ot.
6 intensive care*. sh.
7 (ventilat* or Respirat* or Artificial).af.
8 exp artificial ventilation/
9 acute care facilit*.af.
10 (thermal injur* or burn* or trauma or resuscitation*).kw,tw,ti,ot,sh.
11 ((APACHE or intensive) adj2 score*).af.
12 exp heart arrest/dt, th or exp ST segment elevation myocardial infarction/dt, th or exp acute heart infarction/dt, th or exp brain infarction/dt, th or exp brain vasospasm/dt, th or exp brain artery aneurysm/dt, th or exp brain ischemia/dt, th or exp subarachnoid hemorrhage/dt, th or exp brain injury/dt, th or exp cerebrovascular accident/dt, th or exp stroke patient/
13 (erythropo* or erithropo* or epo or rheupo or rhepo or rhuepo or Hempoietin* or hematopoiet* or hemopoiet* or epoetin* or darbopoiet* or darbepo*).af.
14 (r adj HuEPO).af.
15 (Abseamed or Alfaepoetina or Aranesp or Betapoietin or Binocrit or Bioetin or Biopoin or Bioyetin or Ceriton or Culat or Dynepo or EPIAO or Epiao or Epoade or Epobel or Epocim or Epoch or Epocomb or Epocrin or Epoetin Alfa Hexal or Epoetin Sedico or Epofer or Epofit or Epoform or Epogen or Epogin or Epogis S or Epoglobin or Epojet or Epokine or Epomax or Epomax or Eponis-2K or Epopen or Eporatio or Eporise or Eporon or Eposim or Eposino or Eposis or Epostim or Epotin or Epotop or Epotrex-NP or Epotrust or Epovitan or Epox or Epoxitin or Epoyet or Eprex or Eralfon or Erantin or Eritina or Eritrelan or Eritrogen or Eritrogen or Eritromax or Eritropoyetina or Eritropoyetina Alfa or Erlan or Erykine or Erypo or Erypoietin or Erypro or Erypro Safe or Erythrostim or Erytrostim or Espo or Espogen or Exetin-A or GerEPO or Globuren or Heberitro or Hemapo or Hemax or Hemax-Eritron or Hemoprex or Hepta or Hypercrit or Jimaixin or LG Espogen or Marogen or Mircera or Negortire or NeoRecormon or Nesp or Nespo or PDpoetin or Procrit or Pronivel or Recormon or Relipoietin or Renogen or Repotin or Retacrit or Sepo or Shanpoietin or Silapo or Tinax or Vero-Epoetin or Vintor or Wepox or Yepotin or YiBei or Zyrop).kw,tw,tn,ti,ot.
16 exp erythropoietin/ or exp recombinant erythropoietin/ or exp erythropoietin receptor/ or exp novel erythropoiesis stimulating protein/
17 (HM10760A or 64FS3BFH5W or 15UQ94PT4P or BM06019 or DRG-0062 or KRN 321 or KRN321 or KRN 5702 or KRN5702 or BI71052 or TYB 5220 or TYB5220 or SNB 5001 or SNB5001 or HSDB 7584 or HSDB7584 or LS 64697 or LS64697).af.
18 (113427-24-0 or 11096-26-7 or 209810-58-2 or 122312-54-3 or 261356-80-3 or 604802-70-2 or 130455-76-4 or 148363-16-0 or 154725-65-2 or 879555-13-2).rn.
22 19 and 20
Appendix 3. Data extraction form for included RCT studies
Verification of eligibility
Study design (verify by a ✓)
Screening and enrolment
Intention-to-treat (verify by a ✓)
An intention-to-treat analysis is one in which all participants in a trial are analysed according to the intervention to which they were allocated, whether or not they received it.
Study population (for verification, put a ✓)
*SD = SE × (sample size).
Outcome measures and results
*Mean, SD, N.
If any result was estimated from graphs, etc. or calculated by you using a formula, this should be stated and the formula given.
Appendix 4. Data extraction form for included observational studies
Verification of eligibility
Study design (verify by a ✓)
Study population (for verification, put a ✓)
Verify by a ✓ if characteristic considered to be a confounder (Conf’r?), for each characteristic, tick last column to indicate whether groups were considered different (Diff’t?) by the researchers.
*SD = SE × (sample size).
Outcome measures and results
Effects of exposure Underline the effect estimate measure in the first row for each AE, and write it down in the second row, as well as for confidence interval or standard error.
If any result was estimated from graphs, etc. or calculated by using a formula, this should be stated and the formula given.
Contributions of authors
Conceiving of the review: BM, HH
Co-ordinating the review: HH
Undertaking manual searches: BM
Screening search results: BM
Organizing retrieval of papers: BM
Screening retrieved papers against inclusion criteria: BM, BHH, DR, HH
Appraising quality of papers: BM, BHH, DR, HH
Abstracting data from papers: BM, BHH, DR, HH
Writing to authors of papers for additional information: BM
Providing additional data about papers: BM
Obtaining and screening data on unpublished studies: BM, BHH
Managing data for the review: BM
Entering data into Review Manager (RevMan 5.2): BM, DR
Handling RevMan statistical data: BM, HH
Performing other statistical analyses not using RevMan: SS, CDW
Interpreting data: BM, SS, CDW, HH
Making statistical inferences: SS, CDW, HH
Securing funding for the review: HH
Performing previous work that was the foundation of the present study: BM, BHH, SS, CDW, HH
Serving as guarantor for the review (one author): HH
Taking responsibility for reading and checking the review before submission: BM
Declarations of interest
Bita Mesgarpour: none known.
Benedikt H Heidinger: none known.
Dominik Roth: none known.
Susanne Schmitz: none known.
Cathal D Walsh: none known.
Harald Herkner: none known.
Sources of support
- Medical University of Vienna, Austria.
- Trinity College Dublin, Ireland.
- No sources of support supplied