Review

(Included trials)

Objectives

Date assessed as up to date / Date of last search

Interventions

Comparison Interventions

Study Population

Number of participants / type of SCD / age of participants

Number of included trials / study design / location of studies / duration of follow‐up

Review outcomes for which data are reported

Review outcomes for which data are not reported

Estcourt 2017a

Blood transfusion for preventing primary and secondary stroke in people with sickle cell disease

(SIT (DeBaun 2014); STOP (Adams 1998); STOP2 (Adams 2005); SWITCH (Ware 2012); TWITCH (Ware 2016))

To determine whether chronic blood transfusion in people with SCD:

1. reduce occurrence of stroke (primary prevention);

2. reduce recurrence of stroke (secondary prevention;

3. reduce mortality;

4. reduce other complications of SCD including pain crises, ACS and splenic sequestration;

5. are associated with unacceptable adverse events or costs.

04 April 2016 / 04 April 2016

Chronic blood transfusion regimens to maintain HbS level < 30% (+/‐ iron chelation)

Standard care

N = 405 participants

N = 130 children with abnormal TCD velocities and previous history of TIA or stroke (STOP)

N= 79 children with normalised TCD velocities (STOP 2)

N = 196 children with SCIs and normal TCD velocities (SIT)

All 3 trials enrolled children with HbSS or Hb Sβº

% with either 1 not clear majority HbSS

Mean (SD) age (years):

STOP: 8.3 (3.3);

STOP 2: 8.6 (1.2);

SIT: 8 (3).

2 multicentre trials

conducted in USA and Canada (STOP; STOP 2)

Mean (SD) length of follow‐up: transfusion arm: 21.0 (5.7) months; standard care: 18.3 (7.0) months (STOP).

Median time from randomisation to an end‐point event was 3.2 months (range, 2.1 to 10.1), and the mean (SD) was 4.5 (2.6) months (STOP 2).

Both studies were stopped early due to safety concerns.

1 multicentre trial conducted in USA, Canada, France and UK (SIT)

Children were followed for a median of 3 years.

Primary outcomes

· Incidence of clinical stroke

· Mortality

· Transfusion‐related complications

Secondary outcomes

· Incidence of TIA

· Haemoglobin level and haemoglobin S percentage

· Measure of neurological impairment

· Incidence of other sickle cell complications

· QoL, inpatient stay, immobility and disability

Measures of organ damage

Chronic blood transfusion regimens to maintain HbS level < 30% (+/‐ iron chelation)

Hydroxyurea with phlebotomy

N = 254 participants

N = 133 history of stroke, chronic transfusion treatment and iron overload (SWiTCH)

N = 121 children with abnormal TCD velocities but no severe vasculopathy on MRI/MRA, who have received at least 1 year of RBC transfusions (TWITCH)

HbSS:

SWITCH: transfusion: 100% hydroxyurea : 99%

TWITCH: transfusion: 97% hydroxyurea: 100%

Mean (SD) age (years): SWITCH: transfusion: 13.3 (3.8); hydroxyurea: 13.0 (4.0); TWITCH: transfusion: 9.5 (0.68); hydroxyurea: 9.7 (3.2).

2 multicentre trials with 254 participants

Conducted in USA and Canada

Total duration of study treatment was: 30 months after randomisation, with a final study visit scheduled 6 months after discontinuation of study treatments (SWiTCH); 24 months after randomisation with a 6‐month visit after completing exit studies (TWITCH).

Both studies were stopped early, 1 for futility (SWITCH) and 1 for non‐inferiority (TWITCH).

Primary outcomes

· Mortality

· Incidence of clinical stroke

· Transfusion‐related complications

Secondary outcomes

· Incidence of TIA or silent infarction

· Haemoglobin level and haemoglobin S percentage

Measure of neurological impairment

Incidence of other SCD complications

QoL

Inpatient stay

Immobility and disability

Measures of organ damage

Estcourt 2017b

Interventions for preventing silent cerebral infarcts in people with sickle cell disease

(SIT (DeBaun 2014); STOP (Adams 1998); STOP2 (Adams 2005); SWITCH (Ware 2012); TWITCH (Ware 2016))

To assess the effectiveness of red blood transfusions and hydroxyurea alone or in combination and HSCT to reduce or prevent SCI in people with SCD

19 September 2016

Chronic blood transfusion regimens to maintain HbS level < 30% (+/‐ iron chelation)

Standard care

N = 405 participants

N = 130 children with abnormal TCD velocities and previous history of TIA or stroke (STOP)

N= 79 children with normalised TCD velocities (STOP 2)

N = 196 children with SCI and normal TCD velocities (SIT)

All 3 trials enrolled children with HbSS or Hb Sβº

% with either 1 not clear majority HbSS

Mean (SD) age (years): STOP: 8.3 (3.3); STOP 2: 8.6 (1.2); SIT: 8 (3).

2 multicentre trials

conducted in USA and Canada

(STOP, STOP 2)

Mean (SD) length of follow‐up: transfusion arm: 21.0 (5.7) months; Standard care: 18.3 (7.0) months (STOP)

Median time from randomisation to an end‐point event was 3.2 months (range, 2.1 to 10.1), and the mean (SD) was 4.5 (2.6) months (STOP 2).

Both studies terminated early.

1 multicentre trial conducted in USA, Canada, France and UK (SIT)

Children were followed for a median of 3 years.

Primary outcomes

· Proportion of participants developing new or progressive SCI lesions on MRI

· All‐cause mortality

· SAEs associated with different therapies or SCD

Secondary outcomes

· Clinical stroke (according to short‐, medium‐, and long‐term outcomes)

· Cognitive function as assessed by validated scales (such as Wechsler scales) from baseline and at various time intervals as reported in studies (at least 6 months)

· QoL as assessed by validated scales (at least 6 months)

· Any AEs associated with different therapies

QoL and

Cognitive function were not reported in STOP and STOP 2

Chronic blood transfusion regimens to maintain HbS level < 30% (+/‐ iron chelation)

Hydroxyurea with phlebotomy

N = 254 participants

N = 133 history of stroke, chronic transfusion treatment and iron overload (SWiTCH)

N = 121 children with abnormal TCD velocities but no severe vasculopathy on MRI/MRA, who have received at least 1 year of RBC transfusions (TWITCH)

HbSS:

SWITCH: transfusion: 100% hydroxyurea : 99%

TWITCH: 97% transfusion: 97% hydroxyurea: 100%

Mean (SD) age (years): SWITCH: transfusion: 13.3 (3.8); hydroxyurea: 13.0 (4.0); TWITCH: transfusion: 9.5 (2.68); hydroxyurea: 9.7 (3.2).

2 multicentre trials with 254 participants

Conducted in USA and Canada

Total duration of study treatment was:

30 months after randomisation, with a final study visit scheduled 6 months after discontinuation of study treatments (SWiTCH);

24 months after randomisation with a 6 month visit after completing exit studies (TWITCH).

Both studies terminated early.

Primary outcomes

· Proportion of participants developing new or progressive SCI lesions on MRI

· All‐cause mortality

· SAEs associated with different therapies or SCD

Secondary outcomes

· Clinical stroke (according to short‐, medium‐, and long‐term outcomes)

· Any AEs associated with different therapies

Cognitive function as assessed by validated scales (such as Wechsler scales) from baseline and at various time intervals as reported in studies (at least 6 months)

QoL as assessed by validated scales (at least 6 months)

SWiTCH did not report any AEs

Estcourt 2016bPreoperative blood transfusions for sickle cell disease

(Al‐Jaouni 2006; Howard 2013; Vichinsky 1995)

To determine whether there is evidence that preoperative blood transfusion in elective or emergency surgery:

a. reduces mortality;

b. reduces complications directly related to the surgical procedure, such as local infection and bleeding;

c. reduces serious perioperative complications including pain, ACS and the postoperative frequency and severity of infections;

d. is associated with severe AEs (as reported in the included studies).

2. To compare the effectiveness of different transfusion regimens (aggressive or conservative) if preoperative transfusions

are indicated

9 July 2015 / 17 March 2015

Aggressive, designed to decrease the haemoglobin S level to less than 30%

Conservative, designed to increase the haemoglobin level to 100 g/L

N = 551 participants

HbSS: 551 (100%)

75% aged < 20 years; 25% aged 20 years or older.

Review reports on cholecystectomy (N = 230) and tonsillectomy/ adenoidectomy (N = 107) subgroups only due to unit of analysis issues with re‐randomisation in the trial.

1 parallel RCT in USA

30‐day follow‐up

Primary outcomes

· Perioperative mortality (all‐cause)

· Serious complications related to:

§ sickle cell disease

§ surgery

§ transfusion

Secondary outcomes

· Other transfusion‐related complications

· Haemoglobin level and haemoglobin S percentage

· Number of units or volume (mL) of RBCs infused and, where known for exchange transfusions

· Length of stay

· QoL

· Measures of organ damage

Simple or exchange transfusion

No preoperative transfusion

N = 436 participants (2 re‐randomised)

HbSS = 337 (77%)

HbSβº = 76 (17%)

HbSβ⁺ = 14 (4%)

HbSC = 9 (2%)

247 children (57%) (aged up to 16 or 18 years)

189 adults (43%) (only 6 (1%) aged 40 years or older)

2 parallel RCTs:

1 multinational (22 centres in Canada, Ireland, Netherlands and the UK); 1 in Saudi Arabia.

1 study reports a 30‐day follow‐up the other does not report study duration.

Multinational study stopped early before full enrolment.

Primary outcomes

· Perioperative mortality (all‐cause)

· Serious complications related to:

§ sickle cell disease

§ surgery

§ transfusion

Secondary outcomes

· Other transfusion‐related complications

· Length of stay (reported in one trial only)

· QoL

(reported in 1 trial only)

· Haemoglobin level and haemoglobin S percentage (reported in 1 trial only)

· Number of units or volume (mL) of RBCs infused and, where known for exchange transfusions (reported in 1 trial only)

· Measures of organ damage

Okusanya 2016Prophylactic versus selective blood transfusion for sickle cell disease in pregnancy

(Koshy 1988)

To assess the benefits and harms of a policy of prophylactic versus

selective blood transfusion in pregnant women with SCD

30 May 2016 /30 May 2016

Prophylactic transfusion:

to optimise Hb concentration to a specified level

Selective (emergency) transfusion:

when indicated by specific complication or a critically low level of Hb concentration

N = 72

pregnant women with sickle cell anaemia (genotype HbSS) before 28 weeks of gestation

1 multicentre study conducted in 6 secondary and university hospitals in the USA.

Final evaluation conducted 6 weeks postpartum.

Primary outcomes

· Maternal death

· Severe maternal morbidity (e.g. organ failure, pulmonary embolism, fat embolism, stroke, intensive care unit admission; or as defined by trial authors)

· Perinatal death

Secondary outcomes

Mother

· Sickle cell crisis (due to vaso‐occlusion, sequestration or

· haemolysis)

· Total units of blood transfused

· Blood transfusion reaction

Mother

· Iron overload

· postpartum haemorrhage

· Cumulative duration of hospital stay

Infant ·

· Admission to neonatal intensive care

· Haemolytic disease of the newborn

Dastgiri 2016

Blood transfusions for treating acute chest syndrome in people with Sickle cell disease

(PROACTIVE (Styles 2012))

To assess the effectiveness of blood transfusions, simple and exchange, for treating ACS by comparing improvement in symptoms and clinical outcomes against standard care.

25 July 2016/ 25 April 2016

· Confirmed diagnosis of SCD: SS; SC; Sβº; Sβ⁺

· Male or female

· All ages

· Any setting

· RBC transfusions (simple or exchange)

· Standard care

1 study met the inclusion criteria but no study results reported as out of 237 enrolled only 10 were randomised – the rest observational.

In the study publication they report 0/4 in transfusion arm no ACS and 2/6 in standard care arm with ACS.

Dastigiri 2016 Blood transfusions for treating ACS in people with SCD

To assess the effectiveness of blood transfusions, simple and exchange, for treating ACS by comparing improvement in symptoms and clinical outcomes against standard care.

Review

(Included trials)

Objectives

Date assessed as up to date / Date of last search

Study Population

Number of participants/ type of sickle cell/ age of participants

Interventions

Comparison Interventions

Number of included trials/study design/location of studies/duration of follow‐up

Roy 2017

Interventions to treat chronic kidney disease in people with Sickle cell disease

(BABYHUG (Wang 2011); Foucan 1998)

To assess the effectiveness of any intervention in preventing or reducing kidney complications or CKD in people with SCD

(including RBC transfusions, hydroxyurea and ACEi (either alone or in combination with each other)

/07 October 2026

People with all types of SCD of all ages and either gender

• RBCl transfusion

• Hydroxyurea

• ACEi

• In combination with each other

• Compared to each other

• Placebo

• Standard care

Two studies met the inclusion criteria: one study compared hydroxyurea to placebo and the other study compared angiotensin converting enzyme inhibitors to placebo

Chinegwundoh 2004Treatments for priapism in boys and men with sickle cell disease

(Serjeant 1985)

To assess the benefits and risks of the different treatments for both

stuttering and fulminant priapism in SCD

20 August 2010 / 22 July 2010

• SCD: SS; SC; Sβº; Sβ⁺

• Males

• Priapism, fulminant or stuttering

• All ages

• Any race

• Any ethnic origin

• Any setting

• Any treatment for priapism in SCD

• Placebo;

• No treatment;

• Any other intervention

One study comparing Stilboestrol 5 mg daily versus placebo.

Martí‐Carvajal 2014Interventions for treating leg ulcers in people with sickle cell disease

(Baum 1987; La Grenade 1993; McMahon 2010; Serjeant 1977; Serjeant 1997; Wethers 1994)

To determine whether any clinical interventions (used either alone or in combination) are effective and safe when treating leg ulcers in people with SCD

21 October 2014 / 18 September 2014

• Any SCD

• Diagnosed with a leg ulcer

• Treated in community, hospital or both

Systemic interventions

• Vascular drugs

• Antioxidant agents

• Recombinant agents growth factors

• Pharmacologic stimulation of HbF synthesis

• Oral zinc sulphate

Topical interventions

• Antibiotics and antiseptics

• Growth factors and related

• Steroids

• Dressing;

• Debriding agents;

• Ccompression; miscellaneous (such as topical opioids)

Non‐pharmaceutical interventions

• Reconstructive surgery

• Cell therapy

• Laser therapy miscellaneous

• conventional care another treatment

• regimen for leg ulcers in people with SCD

• Conventional care

• Another treatment regimen for leg ulcers in people with SCD

Six studies met the review’s inclusion criteria:

Interventions included:

· Topical antibiotics

· Solcoseryl® and Duo Derm

· Arginine butyrate

· Propionyl‐L‐carnitine

· RGD peptide matrix

Martí‐Carvajal 2016Treatment for avascular necrosis of bone in people with sickle cell disease

(NOTSCA (Neumayr 2006))

To determine if the following treatments have any impact on avascular necrosis of bone in people with SCD in both the

short‐ and the long‐term (efficacy, safety, and adverse events): any surgical versus non‐surgical intervention, including combinations of surgical and non‐surgical treatment. Also, address the following:

1. the role of decompression;

2. the relative effectiveness of surgical approaches (for studies comparing 2 surgical approaches);

3. types of prosthesis (glued or not, with or without bone grafts etc.)

3 July 2014 / 17 March 2014

• Confirmed diagnosis of SCD (trial included HbSS; SC; Sβ⁰)

• Any SCD ‐related avascular necrosis

• All ages

• Male or female

Surgical treatment:

• Joint reconstruction (femoral head replacement, cup arthroplasty, articular surface replacement, total hip replacement);

• Nucleus decompression;

• Bone graft;

• Vascularized bone grafts;

• Osteotomy

Non‐surgical treatment:

• Observation;

• Analgesic drugs;

• Electrical stimulation;

• Physiotherapy;

• Resting of the joint;

• RBC transfusion;

• Stem cell transplant;

• Treatment to prevent sickling;

• Bisphosphonates

• 1 surgical approach compared to another, or to a non‐surgical approach,

• 2 non‐surgical approaches compared

One study: NOTSCA 2006

Randomised open‐label

Hip core decompression and physical therapy (N = 17) versus physical therapy alone (N = 21)

Oniyangi 2015Phytomedicines (medicines derived from plants) for sickle cell disease

(Akinsulie 2005; Wambebe 2001)

To assess the benefits and risks of phytomedicines in people with SCD of all types, of any age, in any setting.

20 January 2015 / 26 March 2015

• People of all ages

• SCD of any genotype in any geographic setting such as: homozygous (HbSS) and compound heterozygotes

• Including SC disease (HbSC) and β‐ thalassaemia (Sβ0/

• Sβ⁺)

• Proven by electrophoresis with family studies or DNA tests as appropriate

• Administration (by any mode: topical; oral; or parenteral) of phytomedicines (defined as a remedy derived directly from plants or plant material and not synthesised).

• Placebo

• Conventional treatment, including blood transfusion and hydroxyurea

Two studies met the inclusion criteria: one study compared Niprisan versus placebo or conventional treatment (blood transfusions not reported)

The second study compared Ciklavit versus placebo or conventional treatment (blood transfusions not reported).

Review

Objectives

Date assessed as up to date / Date of last search

Study Population

Number of participants/ type of sickle cell/ age of participants

Interventions

Comparison Interventions

Estcourt 2016a

Regular long‐term red blood cell transfusions for managing chronic chest complication in sickle cell disease

Whether long‐term blood transfusions show differences in the

following: incidence of chronic chest complications; ’severity’ or progression of established chronic chest complications; mortality associated with chronic chest complications; unacceptable adverse events

16 May 2016 / 25 April 2016

Any SCD:

• Confirmed diagnosis of SCD: SS; SC; Sβº; Sβ⁺

• Male or female

• All ages

• Positive or negative for a history of chronic or acute chest complications

• Regular RBC transfusions (simple top‐up or exchange)

• Alternative treatment;

• No treatment

Martí‐Carvajal 2009

Intervention for treating painful sickle cell crisis during pregnancy

To assess the effectiveness and safety of different regimens of

packed red cell transfusion, oxygen therapy, fluid replacement

therapy, analgesic drugs, and steroids for the treatment of painful sickle cell crisis during pregnancy.

20 December 2009 / December 2007

• Pregnant women with SCD (all types)

• Any age

• Excluded sickle cell trait

• RBC transfusion

• Oxygen therapy

• Fluid replacement therapy

• Analgesia (nonsteroidal

• aAnti‐inflammatory agents or opiates agents)

• Steroids (prednisone, dexamethasone, or methylprednisolone)

• Each other

• Placebo

• Standard care

Martí‐Carvajal 2015a

Intervention for treating intrahepatic cholestasis in people with sickle cell disease

To assess the benefits and harms of the interventions for treating

intrahepatic cholestasis in people with SCD

13 March 2015 /10 October 2014

• SCD with intrahepatic cholestasis

• Any age

• Male or female

• Hospital or community setting

• Simple RBC transfusion

• Exchange transfusion

• Liver transplant

• Bile acid binding resins (cholestyramine colestipol)

• Rifampin

• Opiate antagonists (naltrexone, nalmefene)

• Sertraline

• Dexamethasone

• Guar gum

• Activated charcoal

• Conventional care

• Each other

Oringanje 2016

Hematopoietic stem cell transplantation for people with sickle cell disease

To determine whether stem cell transplantation can improve survival and prevent symptoms and complications associated with SCD. To examine the risks of stem cell transplantation against the potential long‐term gain for people with SCD

31 March 2016 / 11 December 2015

• Children and adults with SCD of all phenotypes, either gender and in all settings

• Methods of stem cell transplantation

• Methods of stem cell transplantation compared with each other

• Ssupportive care (e.g. periodic transfusion, use of hydroxyurea,

• Antibiotics, pain relievers, supplemental oxygen)

Owusu‐Ofori 2015

Splenectomy versus conservative management for acute sequestration crises in people with sickle cell disease

To assess whether splenectomy (total or partial), to prevent acute splenic sequestration crises in people with SCD, improved survival and decreased morbidity in people with SCD, as compared with regular blood transfusions.

15 July 2015 / 10 June 2015

Any SCD:

• Confirmed diagnosis of SCD: SS; SC; Sβº; Sβ⁺

• Experienced at least one acute splenic sequestration crisis

• Full or partial splenectomy

• Conservative management

• No treatment

• Rregimen of regular RBC transfusions (e.g. 4‐weekly)

AMSTAR Criteria^{1, 2}

Estcourt 2017a

Blood transfusion for preventing primary and secondary stroke in people with sickle cell disease

Estcourt 2017b

Interventions for preventing silent cerebral infarcts in people with sickle cell disease

Estcourt 2016b

Preoperative blood transfusions for sickle cell disease

Okusanya 2016

Prophylactic versus selective blood transfusion for sickle cell disease in pregnancy

Dastgiri 2016

Blood transfusions for treating acute chest syndrome in people with sickle cell disease

¹ We applied a modified AMSTAR based on the univariable questions developed for the overview by Pollock 2014. See Appendix 2 AMSTAR Checklist

² Modified AMSTAR answers are reported as: Y = yes; N = no; U = unsure; NA = not applicable

1. Was an ’a priori’ design provided

Y

Protocol available ‐ objectives, participants, interventions and comparisons and primary and secondary outcomes clearly defined.

Y

Protocol available ‐ objectives, participants, interventions and comparisons and primary and secondary outcomes clearly defined.

Y

Protocol available ‐ objectives, participants, interventions and comparisons and primary and secondary outcomes clearly defined.

Y

Protocol available ‐ objectives, participants, interventions and comparisons, and primary and secondary outcomes clearly defined.

Y

Protocol first published: Issue 2, 2009.

2. Was there duplicate study selection and data extraction?

Y

Two independent review screened all electronically‐derived citations and abstracts of papers identified in the search for relevance. We excluded trials that were clearly irrelevant at this stage based on a review of the abstract.

Two independent review authors formally assessed the full texts of all potentially‐relevant trials for eligibility against the criteria outlined above. We resolved all disagreements by discussion without the need for a third review author.

Two review authors conducted the data extraction according to the guidelines proposed by Cochrane. We resolved disagreements between the review authors by consensus.

Y

Two independent review authors screened all electronically‐derived citations and abstracts of papers identified by the review search strategy for relevance.

Two independent review authors (LE, PF) formally assessed the full texts of all potentially relevant trials for eligibility against the criteria outlined above. We requested additional information from study authors as necessary.

The two review authors discussed the results of study selection and resolved our discrepancies.

Two review authors conducted the data extraction according to the guidelines proposed by the Cochrane Collaboration. Disagreements were resolved by consensus.

Y

Two independent review authors screened all electronically‐derived citations and abstracts of papers identified by the review search strategy for relevance.

Two independent review authors (PF, LE) formally assessed the full texts of all potentially‐relevant trials for eligibility against the criteria outlined above.

We resolved all disagreements by discussion without the need for a third review author.

Two review authors conducted the data extraction according to the guidelines proposed by the Cochrane Collaboration. Disagreements were resolved by consensus.

Y

Two review authors independently assessed for inclusion all the potential studies identified as a result of the search strategy.

For eligible studies, both review authors independently extracted the data using the agreed form. We resolved discrepancies through discussion.

Y

Two authors independently assessed the abstracts of trials identified from the searches. The two review authors independently assessed the full text papers independently and resolved any disagreement on their eligibility for this review through discussion and consensus; or if necessary through a third party.

3. Was a comprehensive literature search performed?

Y

Searched Haemoglobinopathies Trials Register. We also searched the following databases for RCTs and SRs on April 4 2016:

The Cochrane Library (CENTRAL & DARE) – issue 4, 2016; issue 2, 2015 respectively MEDLINE (OvidSP, 1946 to April 4 2016) EMBASE (OvidSP, last 6 months to April 4 2016) CINAHL (EBSCOHost, to April 4 2016) PubMed (e publications only to April 4 2016) Transfusion Evidence Library (1950 to April 4 2016) LILACS (1982 to April 4 2016) IndMed (1986 to April 4 2016) KoreaMed (1997 to April 4 2016) Web of Science (Conference Proceedings Citation Index‐ Science (CPCI‐S) ‐1990 to April 4 2016) We also searched the following trial databases for ongoing trials on 4 April 2016: ClinicalTrials.gov; WHO International Clinical Trials Registry Platform (ICTRP).

Y

Searched haemoglobinopathies trials register and the following databases for RCTs on 19 September 2016.

The Cochrane Library (CENTRAL, DARE, HTA, NHSEED) MEDLINE (OvidSP, Epub Ahead of Print, In‐Process & Other Non‐Indexed Citations, Ovid MEDLINE Daily and

Ovid MEDLINE, 1946 to 19 September 2016) PubMed (Epub Ahead of Print, In‐Process & Other Non‐Indexed Citations, Embase (OvidSP) CINAHL (EBSCOHost, 1937 to 19 September 2016)Transfusion Evidence Library (1950 to 19 September 2016) LILACS (1982 to 19 September 2016) Web of Science (Conference Proceedings Citation Index‐ Science (CPCI‐S) ‐ 1990 to 19 September 2016) We also searched the following trial databases for ongoing trials on 19 Sepetember 2016.

ClinicalTrials.gov

WHO International Clinical Trials Registry Platform (ICTRP).

Y

Searched haemoglobinopathies trials register and supplemented searches with a search of current MEDLINE, Embase, clinical trial registries, included mesh terms, key words and dates. Search strategies in appendix.

Y

The Cochrane Pregnancy and Childbirth Group’s Trials Register which includes at least 2 major databases – and handsearches was searched 30 May 2016.

Provided Mesh terms but no key words. Trial registries were not searched.

Y

We conducted searches in the Cochrane Cystic Fibrosis and Genetic

Disorders Group’s Haemoglobinopathies Trials Register using the terms: sickle cell AND acute chest syndrome. We also searched the (www.ClinicalTrials.gov) database to identify any additional relevant clinical studies and registered clinical trials of the same topic. Regarding the relatively high prevalence of hereditary sickle cell disease (SS) and others associated disorders in south provinces of Iran (Zandian 2012), we tried to find any additional RCT studies registered in Iranian Registry of

Clinical Trials (www.irct.ir). Other databases and trial sites not search such as EMBASE and ICSTR.

4. Was the status of publication

(i.e. grey literature) used as an Inclusion criterion?

Y

no limits on language or publication status.

Y

Searches for unpublished work included in search of the Register. No language restrictions.

Y

Searches for unpublished work included in search of the Register. No language restrictions.

Y

Searches for unpublished work included in search of the Register. No language restrictions.

Y

We contacted the trial authors of the PROACTIVE trial for further data and information, but to date have received no.

5. Was a list of studies (included

and excluded) provided?

Y

Includes study flow diagram.

Y

Includes a flow diagram.

Y

Includes a flow diagram.

Y

Includes study flow diagram.

N

No study flow diagram provided.

6. Were the characteristics of the included studies provided?

Y

Detailed descriptions of participants interventions and outcomes of included studies provided

Y

Detailed descriptions of participants interventions and outcomes of included studies provided.

Y

Detailed descriptions of participants interventions and outcomes of included studies provided.

Y

Detailed descriptions of participants interventions and outcomes of included studies provided.

Y

Detailed descriptions of participants interventions and outcomes provided.

7. Was the scientific quality of the Included studies assessed and documented?

Y

Assessed by two authors independently with Cochrane Risk of Bias tool, and Summary of findings table included and documented.

Y

Assessed by two authors independently with Cochrane Risk of Bias tool and Summary of findings table included and documented.

Y

Assessed by two authors independently with Cochrane Risk of Bias tool and Summary of findings table included and documented.

Y

Assessed by two authors independently with Cochrane Risk of Bias tool, and Summary of findings table included and documented.

Not applicable.

8. Was the scientific quality of the Included studies used appropriately in formulating conclusions?

Y

The quality of the evidence was very low to moderate across different outcomes according to GRADE methodology. This was due to the trials being at a high risk of bias due to lack of blinding, indirectness and imprecise outcome estimates. Due to lack of evidence this review cannot comment on management for adults with HbSS disease or children and adults with HbSβ⁰, HbSC or HbSβ⁺ disease.

Y

The quality of the evidence was very low to moderate across different outcomes according to GRADE methodology. This was due to: trials being at high risk of bias because they were unblinded; indirectness (the available evidence was for children with HbSS); and imprecise outcome estimates. There is low‐quality evidence that long‐term RBC transfusions may reduce the incidence of SCI in children with abnormal TCD velocities but have little or no effect on children with normal TCD velocities. In children who are at higher risk of stroke and have not had previous long‐term transfusions, there is moderate‐quality evidence that long‐term RBC transfusions.

Y

Overall, the quality of the evidence was rated as low to very low across different outcomes according to GRADE methodology. This was due to two of the studies being at high

or unclear risk of bias, and many of the outcome estimates being imprecise.

Due to lack of evidence this review cannot comment on management for people with HbSC or HbSβ⁺ disease or patients with high baseline haemoglobin concentrations, or management of patients undergoing low risk surgery.

Y

Evidence from one small trial of very low quality suggests that prophylactic blood transfusion to pregnant women with sickle cell

anaemia (HbSS) confers no clear clinical benefits when compared with selective transfusion. Currently, there is no evidence from randomised or quasi‐randomised trials to provide reliable advice on the optimal blood transfusion policy for women with other variants of sickle cell disease (i.e. HbSC and HbS_Thal). The available data and quality of evidence on this subject are insufficient to advocate for a change in existing clinical practice and policy.

Not applicable.

9. Were the methods used to combine the findings of studies

appropriate?

Y

Risk ratio reported and Peto odds ratio reported for outcomes with low event rates all confidence intervals included. Heterogeneity defined, There was no statistical heterogeneity, but if statistical heterogeneity was found to be above 75%, we would identify a reason for clinical heterogeneity and not perform a meta‐analysis but comment instead on the results as a narrative.

Y

Risk ratio reported and Peto odds ratio reported for outcomes with low event rates all confidence intervals included. Heterogeneity defined, There was no statistical heterogeneity, but if statistical heterogeneity was found to be above 75%, we would identify a reason for clinical heterogeneity and not perform a meta‐analysis but comment instead on the results as a narrative.

Y

Risk ratio reported and Peto odds ratio reported for outcomes with low event rates all confidence intervals included. Heterogeneity defined, identified and reported.

Y

For dichotomous data, we presented results as summary risk ratio with 95% confidence intervals. No continuous data were analysed in this review. In future updates, if appropriate, we will use the mean difference if outcomes are measured in the same way between trials.We will use the standardised mean difference to combine trials that measure the same outcome, but use different methods.

Not applicable.

10. Was the likelihood of publication bias assessed?

Y

Mentioned that too few studies to assess publication bias.

Y

Mentioned that too few studies to assess publication bias.

Y

Mentioned that too few studies to assess publication bias.

Y

Mentioned that too few studies to assess publication bias.

Not applicable.

11. Was the conflict of interest

stated?

Y

Lise Estcourt is partly funded by an NIHR Cochrane Programme Grant. Patricia Fortin is funded by an NIHR Cochrane Programme Grant. Sally Hopwell is partly funded by an NIHR Programme Grant. Marialena Trivella is partly funded by an NIHR Programme Grant. Winfred Wang was a PI on several of the included trials.

Y

Lise Estcourt is partly funded by an NIHR Cochrane Programme Grant.

Carolyn Doree: none to declare. Patricia Fortin is funded by an NIHR Cochrane Programme Grant.

Sally Hopwell is partly funded by an NIHR Programme Grant.

Marialena Trivella is partly funded by an NIHR Programme Grant.

Y

Lise Estcourt is partly funded by an NIHR Cochrane Programme Grant.

Carolyn Doree: none to declare. Patricia Fortin is funded by an NIHR Cochrane Programme Grant.

Sally Hopwell is partly funded by an NIHR Programme Grant.

Marialena Trivella is partly funded by an NIHR Programme Grant.

Y

None known.

Y

There are no financial conflicts of interest and the review authors declare that they do not have any associations with any parties who may have vested interests in the results of this review.

AMSTAR

Criteria^{1, 2, 3}

Chinegwundoh 2004

Treatments for priapism in boys and men with sickle cell disease

Martí‐Carvajal 2014

Interventions for treating leg ulcers in people with sickle cell disease

Martí‐Carvajal 2016

Treatment for avascular necrosis of bone in people with sickle cell disease

Roy 2017

Interventions for chronic kidney disease in people with sickle cell disease

Oniyangi 2015

Phytomedicines (medicines derived from plants) for sickle cell disease

¹ We applied a modified AMSTAR based on the univariable questions developed for the overview by Pollock 2014. See Appendix 2 modified AMSTAR questions

² Modified AMSTAR answers are reported as: Y = yes; N = no; U = unsure; NA = not applicable

³ Items 5 ‐ 10 of AMSTAR were not assessed in the quality appraisal as the reviews did not include any studies with a red cell transfusion comparison

1. Was an ’a priori’ design provided

Y

Protocol available – participants, interventions and comparisons and outcomes clearly defined.

Y

Protocol available – participants, interventions and comparisons and outcomes clearly defined.

Y

Protocol available – participants, interventions and comparisons and outcomes clearly defined.

Y

Protocol available – participants, interventions and comparisons and outcomes clearly defined.

Y

Protocol available – participants, interventions and comparisons and outcomes clearly defined.

2. Was there duplicate study selection and data extraction?

Y

Each author independently identified potentially relevant trials. Each author independently extracted data. No disagreement occurred, so no other expert in the field was invited to make an independent assessment based on the selection criteria.

Y

Two authors independently selected studies for inclusion, and extracted data. A third author was always included when two authors disagreed, all disagreements resolved by group discussion.

Y

The authors screened the search results for potentially relevant trials and independently assessed them. Each author used this form to extract data from each relevant study. We independently extracted information from the papers.

Y

Two independent review authors screened all citations and abstracts of papers identified by the search strategy for relevance.

Two review authors independently and formally assessed the full texts of all

potentially relevant trials for eligibility. Two review authors conducted the data extraction.

Y

Two authors (OO, DC) independently selected the trials for inclusion in the review using pre‐defined inclusion criteria. two authors (NC, OO) had independently

extracted the data and resolved differences by referring to the original trial. The process was repeated for this update by both authors.

3. Was a comprehensive literature search performed?

Y

Studies identified in the Group’s Haemoglobinopathies Trials Register. Searched Embase (1974 to 2003) (Appendix 1) and the Internet (December 2003).

Y

We identified relevant trials from the Group’s Haemoglobinopathies Trials Register. Also searched– LILACS and several other databases and websites.

U

Trials were identified from the Group’s Haemoglobinopathies Trials Register.

No Embase or trial registry search.

Y

We identified trials from the Cochrane Cystic Fibrosis and Genetic Disorders Group's Haemoglobinopathies Trials Register. In addition to this we searched the following databases for RCTs on 05 April 2016. CENTRAL, the Cochrane Library (current issue) MEDLINE (Epub Ahead of Print, In‐Process & Other Non‐Indexed Citations, Ovid MEDLINE Daily and Ovid

MEDLINE, Embase CINAHL PubMed;Transfusion Evidence Library LILACS; IndMed; KoreaMed;

PakMediNet; Web of Science;

ClinicalTrials.gov; WHO International Clinical Trials Registry Platform (ICTRP).

Y

Relevant trials were identified from the Group’s Haemoglobinopathies Trials Register. Relevant trials were also identified from the International Standard Randomised

Controlled Trial Number Registry (ISCTN). An additional search of the Allied and Complimentary Medicine

(AMED) bibliographic database.

4. Was the status of publication (i.e. grey literature) used as an Inclusion criterion?

Y

Attempts were also made to identify any unpublished trials through contact with experts in the field and personal communication

with known authors.

Y

We searched for trials, irrespective of publication status (trials may be unpublished or published as an article, an abstract, or a letter), language or country. No limit was applied with respect to the period of follow‐up.

U

Did not mention if any language restrictions.

Stated that bibliographic references of all retrieved literature were reviewed for additional reports of trials.

Y

We did not limit searches by language, year

of publication or publication type.

U

The bibliographic references of all retrieved trials and reviews were assessed for additional reports of trials; lead authors were to be contacted if necessary.

No mention of limitation by language.

11. Was the conflict of interest stated?

Y

None known.

Y

In 2004 Arturo Martí‐Carvajal was employed by Eli Lilly to run a four‐hour workshop on ’How to critically appraise clinical trials. In 2007 Arturo Martí‐Carvajal was employed by Merck to run a four‐hour workshop on ’How to critically appraise clinical trials. Jennifer Knight‐Madden and María José Martínez‐Zapata: none known.

N

None known.

Y

Noemi Roy: none known.

Patricia Fortin: funded by the NIHR Cochrane Programme Grant ‐ Safe and Appropriate Use of Blood Components.

Katherine Bull: none known.

Carolyn Doree: none known.

Sally Hopewell: partly funded by the NIHR Cochrane Programme Grant ‐ Safe and Appropriate Use of Blood Components. Marialena Trivella: partly funded by the NIHR Cochrane Programme Grant ‐ Safe and Appropriate Use of Blood Components.

Lise Estcourt: partly funded by the NIHR Cochrane Programme Grant ‐ Safe and Appropriate Use.

Y

None known.

AMSTAR

Criteria^{1, 2, 3}

Estcourt 2016a

Regular long‐term red blood cell transfusions for managing chronic chest complications in sickle cell disease

Martí‐Carvajal 2009

Interventions for treating painful sickle cell crisis during pregnancy

Martí‐Carvajal 2015a

Interventions for treating intrahepatic cholestasis in people with sickle cell disease

Oringanje 2016

Hematopoietic stem cell transplantation for people with sickle cell disease

Owusu‐Ofori 2015

Splenectomy versus conservative management for acute sequestration crises in people with sickle cell disease

¹ We applied a modified AMSTAR based on the univariable questions developed for the overview by Pollock 2014. See appendix xx modified AMSTAR questions

² Modified AMSTAR answers are reported as: Y = yes; N = no; U = unsure; NA = not applicable

³ Items 5 ‐ 10 of AMSTAR were not assessed in the quality appraisal as the reviews did not include any studies with a red cell transfusion comparison

1. Was an ’a priori’ design provided

Y

Protocol available – participants, interventions and comparisons and outcomes clearly defined.

Y

Protocol available – participants, interventions and comparisons and outcomes clearly defined.

Y

Protocol available – participants, interventions and comparisons and outcomes clearly defined.

Y

Protocol available – participants, interventions and comparisons and outcomes clearly defined.

Y

Protocol available – participants, interventions and comparisons and outcomes clearly defined.

2. Was there duplicate study selection and data extraction?

Y

Two review authors independently screened all the remaining references for relevance against the full eligibility criteria. We retrieved full‐text articles for all references for which a decision on eligibility could not be made from title and abstract alone. The two review authors discussed the results of study selection and resolved any disagreements on trial inclusions by consensus.

Y

We planned to screen the results of our search strategy for potentially relevant trials, and independently assess them. We will independently extract the data.

Y

Two authors will screen the search results for potentially relevant trials and independently assess them. One author will extract data from the included trials using a spreadsheet data extraction form and one author will check the data entered.

Y

Two authors independently screened the ten trials found by the initial search of all the databases and reference lists to identify papers with potential relevance to the review. We obtained the full text of selected articles. Two authors independently selected trials for inclusion by applying the inclusion

criteria to all potential trials.

Y

We did not apply the process described below, as we were not able to identify any trials eligible for inclusion. However, if we include any trials in future updates of this review, we will apply the following methods. The two authors will independently identify potentially relevant

trials from the results of the searches. Each author will independently extract data on trial information.

3. Was a comprehensive literature search performed?

Y

Potentially relevant trials were identified from the Group’s haemoglobinopathies Trials Register We also searched for RCTs in the following databases:

CENTRAL & DARE; MEDLINE); Embase; CINAHL; Pubmed; Transfusion Evidence Library; LILACS(BIREME/PAHO/WHO); KoreaMed; IndMed; PakMediNet; Web of Science; ClinicalTrials.gov; The WHO International Clinical Trials Registry (ICTRP); the ISRCTN Register EU Clinical Trials Register and the Hong Kong Clinical Trials Register.

Y

Searched the Cochrane Pregnancy and Childbirth Group’s Trials Register. In addition, we searched using the terms using the terms sickle cell AND pregnancy: Cystic Fibrosis and Genetic Disorders Group’s Trials Register (October 2007); LILACS database (1982 to Dec 5 2007); ClinicalTrials.gov www.clinicaltrials.gov

Y

Relevant studies from Disorders Group’s Haemoglobinopathies Trials Register using the terms:

Also searched LILACS; Epistemonikos; WHO International Clinical Trials Registry.

U

We searched the Cochrane Cystic Fibrosis and Genetic Disorders Group Group’s Haemoglobinopathies Trials Register. We searched the Meetings of the American society for Blood and

Marrow Transplantation; Center for the International Blood and Marrow Transplant Research; and European Group for Blood and Marrow

Transplantation .

No additional databases or clinical trials sites searched.

Y

Relevant trials were identified from Relevant trials were identified from the Group’s Haemoglobinopathies Trials Register. In addition to the above, further subject specific electronic searched of MEDLINE and Embase.

4. Was the status of publication (i.e. grey literature) used as an Inclusion criterion?

Y

We did not limit searches by language or publication

status.

U

Did not mention if any language restrictions

Stated that bibliographic references of all retrieved literature were reviewed for additional reports of trials.

Y

We did not adopt any language or publication restrictions.

U

All randomised controlled and quasi‐randomised studies.

Not stated if any language or publication status limitations.

Y

A comprehensive search strategy was formulated in an attempt to identify all relevant trials regardless of language or publication status (published, unpublished, in press, and in progress).

11. Was the conflict of interest stated?

Y

Lise Estcourt: partly funded by the NIHR Cochrane Programme Grant ‐ Safe and Appropriate Use of Blood Components.

Patricia Fortin: funded by the NIHR Cochrane Programme Grant ‐ Safe and Appropriate Use of Blood Components. Sally Hopewell: partly funded by the NIHR Cochrane Programme Grant ‐ Safe and Appropriate Use of Blood Components. Marialena Trivella: partly funded by the NIHR Cochrane Programme Grant ‐ Safe and Appropriate Use of Blood Components. Ian Hambleton: none declared.

Gavin Cho: none declared.

N

None known.

Y

In 2004 Arturo Martí‐Carvajal was employed by Eli Lilly to run a four‐hour workshop on ’How to critically appraise clinical trials

In 2007 Arturo Martí‐Carvajal was employed by Merck to run a four‐hour workshop on ’How to critically appraise clinical trials

Daniel Simancas‐Racines: none known.

Y

Chioma Oringanje: none known.

Eneida Nemecek: none known.

Oluseyi Oniyangi: none known.

Y

Dr Shirley Owusu‐Ofori: none known.

Tracey Remmington: Managing Editor of the Cochrane Cystic Fibrosis and Genetic Disorders Group.

Review

Intervention

Population

N Participants

(N RCTs)

Relative effect estimate (95% confidence interval)

GRADE quality rating

Interpretation

Estcourt 2017a

Primary and secondary stroke

Long‐term RBC transfusions versus standard care

Children with no previous long‐term RBC transfusions

(Follow‐up 18 to 36 months)

326

(2 RCTs)

No deaths

⊕⊝⊝⊝

Very low^{1, 2,3}

We do not know whether long‐term RBC transfusions decrease mortality

Children and adolescents with previous long‐term RBC transfusions

(Follow‐up 18 months)

79

(1 RCT)

Peto OR 8.0 (0.16 to 404.12)

⊕⊝⊝⊝

Very low^{1, 2,3}

We do not know whether long‐term RBC transfusions decrease mortality

RBC transfusion versus disease‐modifying agents

Primary prevention

Children with abnormal TCD velocities but no severe vasculopathy on MRI/MRA, who have received at least one year of RBC transfusions

(6 months follow‐up)

121

(1 RCT)

No deaths

⊕⊝⊝⊝

Very low^{1, 2,3}

We do not know whether hydroxyurea and phlebotomy decreases mortality

Secondary prevention

Children and adolescents with previous stroke, at least 18 months of RBC transfusions, and evidence of iron overload

(6 months follow‐up)

133

(1 RCT)

Peto OR 0.98 (0.06 to 15.92)

⊕⊝⊝⊝

Very low^{1, 2,3}

We do not know whether hydroxyurea and phlebotomy decreases mortality

Estcourt 2017b

Silent cerebral infarcts

Long‐term RBC transfusions versus standard care

Children or adolescents with abnormal TCD velocities 18 to 21 months follow‐up

130

(1 RCT)

No deaths

⊕⊝⊝⊝

Very low^{1, 2,3}

We do not know whether long‐term RBC transfusions decrease mortality

Children or adolescents with normalised TCD velocities 18 months of follow‐up

79

(1 RCT)

Peto OR 8.00 (0.16 to 404.12)

⊕⊝⊝⊝

Very low^{1, 2,3}

We do not know whether long‐term RBC transfusions decrease mortality

Children with normal TCD velocities and silent stroke

(median 3 years follow‐up)

196

(1 RCT)

No deaths

⊕⊝⊝⊝

Very low^{1, 2,3}

We do not know whether long‐term RBC transfusions decrease mortality

RBC transfusion versus disease‐modifying agents

Primary prevention

Children with abnormal TCD velocities but no severe vasculopathy on MRI/MRA, who have received at least one year of red cell transfusions

(6 months follow‐up)

121

(1 RCT)

No deaths

⊕⊝⊝⊝

Very low^{1, 2,3}

We do not know whether hydroxyurea and phlebotomy decreases mortality

Secondary prevention

Children and adolescents with previous stroke, at least 18 months of RBC transfusions, and evidence of iron overload

(6 months follow‐up)

133

(1 RCT)

Peto OR 1.02 (0.06 to 16.41)

⊕⊝⊝⊝

Very low^{1, 2,3}

We do not know whether hydroxyurea with phlebotomy decreases mortality

Estcourt 2016b

Preoperative blood transfusions for sickle cell disease

Restrictive (to increase the total haemoglobin) versus liberal (to decrease the haemoglobin S level below a specified percentage) RBC transfusion strategy

People with SCD undergoing cholecystectomy

(Follow‐up: 30 days)

230

(1 RCT)^a cholecystectomy

No deaths

⊕⊝⊝⊝

Very low^{1, 2,3}

We do not know whether aggressive RBC transfusions decrease mortality

Short‐term RBC transfusion versus standard care

People with SCD undergoing low to medium risk surgery or any surgery other than cardiac surgery

(Follow‐up 30 days in one trial; not stated in 1 trial)

434

(2 RCTs)

No deaths

⊕⊝⊝⊝

Very low^{1, 2,3}

We do not know whether preoperative RBC transfusions decrease mortality

Okusanya 2016

Prophylactic versus selective blood transfusion for sickle cell disease in pregnancy

Long‐term RBC transfusions versus RBC transfusions to treat complications

Pregnant women with SCD

(Follow‐up for 21.5 to 24.5 weeks of prenatal care)

72

(1 RCT)

No deaths

⊕⊝⊝⊝

Very low^{1, 2,3}

We do not know whether prophylactic RBC transfusions decrease maternal mortality

Perinatal mortality

72 (76 births)

(1 RCT)

RR 2.85 (0.61 to 13.22)

⊕⊝⊝⊝

Very low^{1, 2,3}

We do not know whether prophylactic RBC transfusions decrease perinatal mortality

¹ High Risk of bias in one or more domains using the Cochrane Risk of Bias tool

² Indirectness as specific population or procedure reported may not be generalisable to review question

³ Imprecision as effect has wide confidence intervals and/or study has small number of people or low event rates

GRADE summary of findings for SCD‐related serious adverse events^a

Systematic Review

Interventions

Population

(Follow‐up)

N Participants (N RCTs)

Reported outcomes and time points

Relative effect estimate (95% confidence interval)

GRADE quality rating

Interpretation

Estcourt 2017a

Primary and secondary stroke

Long‐term RBC transfusions versus standard care

Children with no previous long‐term RBC transfusions

(Follow‐up 18 to 36 months)

326
(2 RCTs)

Clinical stroke

RR 0.12
(0.03 to 0.49)

⊕⊕⊕⊝

Moderate²

Long‐term RBC transfusions probably reduce the risk of stroke

326
(2 RCTs)

ACS

RR 0.24
(0.12 to 0.48)

⊕⊕⊝⊝

Low^1,2

Long‐term RBC transfusions may reduce the incidence of ACS

326
(2 RCTs)

Painful crises

RR 0.62 (0.46 to 0.84)

⊕⊕⊝⊝

Low^1,2

Long‐term RBC transfusions may reduce the incidence of painful crisis

Children and adolescents with previous long‐term RBC transfusions

(Follow‐up 18 months)

79

(1 RCT)

Clinical stroke

RR 0.22 (0.01 to 4.35)

⊕⊝⊝⊝

Very low^1,2,3

We do not know whether continuing long‐term red cell transfusions reduces the incidence of clinical stroke

79

(1 RCT)

ACS

_

⊕⊝⊝⊝

Very low^1,2,3

No comparative numbers reported

79

(1 RCT)

Pain crisis

_

⊕⊝⊝⊝

Very low^1,2,3

No comparative numbers reported

RBC transfusion versus disease‐modifying agents

Primary prevention

Children with abnormal TCD velocities but no severe vasculopathy on MRI/MRA, who have received at least one year of red cell transfusions

(6 months follow‐up)

121

(1 RCT)

Clinical stroke

No stroke occurred in either arm

⊕⊝⊝⊝

Very low^1,2,3

We do not know whether switching to hydroxyurea and phlebotomy has any effect on the risk of stroke

121

(1 RCT)

ACS

RR 2.03

(0.39 to 10.69)

⊕⊝⊝⊝

Very low^1,2,3

We do not know whether switching to hydroxyurea and phlebotomy has any effect on ACS

121

(1 RCT)

Pain crisis

RR 5.08

(0.61 to 42.23)

⊕⊝⊝⊝

Very low^1,2,3

We do not know whether switching to hydroxyurea and phlebotomy has any effect on pain crisis

Secondary prevention

Children and adolescents with previous stroke, at least 18 months of RBC transfusions, and evidence of iron overload

(6 months follow‐up)

133

(1 RCT)

Clinical stroke

RR 14.78
(0.86 to 253.66)

⊕⊝⊝⊝

Very low^2.3

We do not know whether switching to hydroxyurea and phlebotomy has any effect on the risk of stroke

133

(1 RCT)

Global SCD SAEs

RR 3.10 (1.42 to 6.75)

⊕⊕⊝⊝

Low ^1,2

Switching to hydroxyurea and phlebotomy may increase the risk of SCD related serious adverse events

133

(1 RCT)

ACS

RR 0.33 (0.04 to 3.08)

⊕⊝⊝⊝

Very low^1,2,3

Very uncertain whether switching to hydroxyurea and phlebotomy has any effect on the risk of ACS

133

(1 RCT)

Pain crisis

RR 3.15 (1.23 to 8.11)

⊕⊕⊝⊝

Low ^1,2

Switching to hydroxyurea and phlebotomy may increase the risk of painful crisis

Estcourt 2017b

Silent cerebral infarcts

Long‐term RBC transfusions versus standard care

Children and adolescents with abnormal TCD velocities

(18 – 21 months follow‐up)

124
(1 RCT)

New or progressive SCI lesions

RR 0.11 (0.02 to 0.86)

⊕⊕⊝⊝

Low^{1, 2}

Long‐term RBC transfusions may reduce the incidence of SCIs

130

(1 RCT)

ACS

RR 0.30

(0.11 to 0.87)

⊕⊕⊝⊝

Low ^{2 3}

Long‐term RBC transfusions may reduce the incidence of ACS

130

(1 RCT)

Pain crisis

RR 0.90

(0.44 to 1.86)

⊕⊝⊝⊝

Very low^1,2,3

Long‐term RBC transfusions may be no different than standard care in reducing the incidence of pain crisis

Children and adolescents with normalised TCD velocities

(18 months follow‐up)

77

(1 RCT)

New or progressive SCI lesions

RR 0.29

(0.09 to 0.97)

⊕⊕⊝⊝

Low ^{1, 2}

Continuing RBC transfusions may reduce the incidence of SCIs

79

(1 RCT)

ACS

‐

⊕⊝⊝⊝

Very low^1,2,3

No comparative numbers reported

79

(1 RCT)

Pain crisis

‐

⊕⊝⊝⊝

Very low^1,2,3

No comparative numbers reported

Children with normal TCD velocities and silent stroke

(median 3 years follow‐up)

196

(1 RCT)

New or progressive SCI lesions

RR 0.70

(0.23 to

2.13)

⊕⊕⊝⊝

Low ^2,3

Long‐term RBC transfusions may be no different than standard care in reducing the incidence of SCIs

196

(1 RCT)

Acute chest syndrome

RR 0.20

(0.08 to 0.51)

⊕⊕⊝⊝

Low ^2,3

Long‐term RBC transfusions may reduce the incidence of ACS

196

(1 RCT)

Pain crisis

RR 0.56

(0.40 to 0.78)

⊕⊕⊝⊝

Low ^2,3

Long‐term RBC transfusions may reduce the incidence of pain crisis

RBC transfusion versus disease‐modifying agents

Primary prevention

Children with abnormal TCD velocities but no severe vasculopathy on MR/MRA, who have received at least one year of red cell transfusions

(6 months follow‐up)

121

(1 RCT)

New or progressive SCI lesions

No SCIs occurred in either study arm

⊕⊝⊝⊝

Very low^1,2,3

We do not know whether switching to hydroxyurea and phlebotomy has any effect on SCIs

121

(1 RCT)

Global SCD SAEs

RR 1.52 (0.58 to 4.02)

⊕⊝⊝⊝

Very low^1,2,3

We do not know if switching to hydroxyurea and phlebotomy has any effect on total SAEs

121

(1 RCT)

ACS

RR 2.03

(0.39 to 10.69)

⊕⊝⊝⊝

Very low^1,2,3

We do not know whether switching to hydroxyurea and phlebotomy has any effect on ACS

121

(1 RCT)

Pain crisis

RR 5.08

(0.61 to 42.23)

⊕⊝⊝⊝

Very low^1,2,3

We do not know whether switching to hydroxyurea and phlebotomy has any effect on pain crisis

Secondary prevention

Children and adolescents with previous stroke, at least 18 months of RBC transfusions, and evidence of iron overload

(6 months follow‐up)

133

(1 RCT)

New or progressive SCI lesions

Peto OR 7.28

(0.14 to

366.91)

⊕⊝⊝⊝

Very low^2,3

We do not know whether switching to hydroxyurea and phlebotomy has any effect on SCIs

133

(1 RCT)

ACS

RR 0.33

(0.04 to

3.08)

⊕⊝⊝⊝

Very low^1,2,3

We do not know whether switching to hydroxyurea and phlebotomy has any effect on ACS

133

(1 RCT)

Pain crisis

RR 3.15

(1.23 to

8.11)

⊕⊕⊝⊝

Low ^{1, 2}

Switching to hydroxyurea and phlebotomy may increase the risk of pain crisis

Estcourt 2016b

Preoperative blood transfusions

Restrictive (to increase the total haemoglobin) versus liberal (to decrease the haemoglobin S level below a specified percentage) RBC transfusion strategy

People undergoing cholecystectomy

(Follow‐up: 30 days)

(results for the tonsillectomy / adenoidectomy groups are similar)

230 (1)

Neurological complications

No events

No effect estimate

⊕⊝⊝⊝

Very low^1,2,3

We do not know whether aggressive RBC transfusions reduce the risk of acute stroke

230 (1)

ACS

RR 0.84

(0.38 to 1.84)

⊕⊝⊝⊝

Very low^1,2,3

We do not know whether aggressive RBC transfusions reduce the risk of ACS

230 (1)

Painful crises

RR 0.30

(0.09 to 1.04)

⊕⊝⊝⊝

Very low^1,2,3

We do not know whether aggressive RBC transfusions reduce the risk of painful crisis

230 (1)

Renal complications

No events

No effect estimate

⊕⊝⊝⊝

Very low^1,2,3

We do not know whether aggressive RBC transfusions reduce the risk of renal complications

230 (1)

Serious infection

RR 1.75 (0.59 to 5.18)

⊕⊝⊝⊝

Very low^1,2,3

We do not know whether aggressive RBC transfusions reduce the risk of serious infection

Short‐term RBC transfusion versus standard care

People undergoing low‐to medium‐risk surgery or any surgery other than cardiac surgery

(Follow‐up 30 days in 1 trial; not stated in 1 trial)

434 (2)

Neurological complications

Peto OR 7.22 (1.24 to 41.94)

⊕⊝⊝⊝

Very low^1,3

We do not know whether preoperative RBC transfusions increase the risk of neurological complications (no events in no preoperative transfusion group)

65 (1)

ACS

RR 0.11 (0.01 to 0.80)

⊕⊕⊝⊝

Low^1,3

Preoperative RBC transfusions may reduce the risk of ACS the population of African haplotypes

369 (1)^b

RR 4.81 (0.23 to 99.61)

⊕⊝⊝⊝

Very low^1,3

We do not know whether preoperative RBC transfusions increase the risk of ACS (no events in control group) in the population of Arabic haplotypes

434 (2)

Painful crises

Peto OR 1.91 (0.61 to 6.04)

⊕⊝⊝⊝

Very low^1,3

We do not know whether preoperative RBC transfusions reduce the risk of painful crisis

65 (1)

Renal complications

No events

No effect estimate

⊕⊕⊝⊝

Low^1,3

We do not know whether preoperative RBC transfusions reduce the risk of renal complications (no events)

434 (2)

Serious infection

Peto OR 1.29 (0.29 to 5.71)

⊕⊝⊝⊝

Very low^1,3

We do not know whether preoperative RBC transfusions reduce the risk of serious infection

Okusanya 2016

Prophylactic vs selective blood transfusion

Long‐term RBC transfusions versus RBC transfusions to treat complications

Pregnant women with SCD

(Follow‐up for 21.5‐24.5 weeks of prenatal care)

72 (1)

ACS

RR 0.67

(0.12 to 3.75)

⊕⊝⊝⊝

Very low^1,3

We do not know whether prophylactic RBC transfusions reduce the risk of ACS

72 (1)

Painful crisis

RR 0.28 (0.12 to 0.67)

⊕⊕⊝⊝

Low¹

Prophylactic RBC transfusions may reduce the risk of painful crisis

72 (1)

Renal complications

RR 1.00 (0.07 to 15.38)

⊕⊝⊝⊝

Very low^1,2,3

We do not know whether prophylactic RBC transfusions reduce the risk of renal complications

72 (1)

Serious infection

RR 1.00 (0.07 to 15.38

⊕⊝⊝⊝

Very low^1,2,3

We do not know whether prophylactic RBC transfusions reduce the risk of serious infection

^a Sickle cell serious adverse events include: neurological, ophthalmological, respiratory, orthopaedic, vascular, hepatic or renal complications, vaso‐occlusive pain crisis, priapism, infections

^b Rated down by 2 for very serious risk of bias

¹ High Risk of bias in one or more domains using the Cochrane Risk of Bias tool

² Indirectness as specific population or procedure reported may not be generalisable to review question

³ Imprecision as effect has wide confidence intervals and/or study has small number of people or low event rates

Table: Transfusion related adverse events (serious and non‐serious)^a

Systematic Review

Intervention

Population

(Follow‐up)

N Participants (N RCTs)

Reported outcomes and

Relative effect estimate

(95% confidence interval)

GRADE quality rating

Interpretation

Estcourt 2017a

Blood transfusion for preventing primary and secondary stroke in people with sickle cell disease

Long‐term RBC transfusions versus standard care

Children with no previous long‐term RBC transfusions

(Follow‐up 18 to 36 months)

121

(1 RCT)

Alloimmunisation

RR 3.16 (0.18 to 57.17)

⊕⊝⊝⊝

Very low^1,2,3

We do not know if long‐term RBC transfusions increase the risk of alloimmunisation

121

(1 RCT)

Transfusion reactions

RR 5.17 (0.71 to 37.52)

⊕⊝⊝⊝

Very low^{1, 2,3}

We do not know if long‐term RBC transfusions increase the risk of transfusion reactions

121

(1 RCT)

Iron overload (serum ferritin)

Increased in children receiving long‐term RBC transfusions, incidence rate ratio 14.42 (5.41 to 875.17

⊕⊕⊝⊝

Low^2,3

Long‐term RBC transfusions may increase the risk of iron overload

Children and adolescents with previous long‐term RBC transfusions

(Follow‐up 18

79

(1 RCT)

Alloimmunisation

1 participant who was in the continuing transfusion arm developed an alloimmunisation

⊕⊝⊝⊝

Very low^{1, 2,3}

No comparative numbers reported

79

(1 RCT)

Transfusion reactions

7 participants in the continuing transfusion arm had 9 reactions to transfusions. 1 required hospitalisation

⊕⊝⊝⊝

Very low^{1, 2,3}

No comparative numbers reported

RBC Transfusion versus disease‐modifying agents

Primary prevention

Children with abnormal TCD velocities but no severe vasculopathy on MRI/MRA, who have received at least one year of red cell transfusions

(6 months follow‐up)

121

(1 RCT)

Liver iron concentrations

MD ‐1.80 (‐5.16 to 1.56)

⊕⊕⊝⊝

Low ^1,2

Switching to hydroxyurea and phlebotomy may have little or no effect on liver iron concentrations

Secondary prevention

Children and adolescents with previous stroke, at least 18 months of RBC transfusions, and evidence of iron overload

(6 months follow‐up)

133

(1 RCT)

Liver iron concentrations

Hydroxyurea: 17.3 mg Fe/g dry weight iron, (IQR 10.0 to 30.6); transfusion: 17.3 mg Fe/g dry weight iron, (IQR 8.8 to 30)

⊕⊕⊝⊝

Low ^1,2

Switching to hydroxyurea and phlebotomy may have little or no effect on liver iron concentrations

Estcourt 2017b

Interventions for preventing silent cerebral infarcts in people with sickle cell disease

Long‐term RBC transfusions versus standard care

Children or adolescents with abnormal TCD velocities

(18 to 21 months follow‐up)

130

(1 RCT)

Alloimmunisation

10 participants in the transfusion arm developed an alloimmunisation

⊕⊝⊝⊝

Very low^{1, 2,3}

No comparative numbers reported

130

(1 RCT)

Transfusion reactions

12 participants in the transfusion arm had 16 mild reactions to blood products or transfusion

procedures

⊕⊝⊝⊝

Very low^{1, 2,3}

No comparative numbers reported

130

(1 RCT)

Iron overload

Iron overload developed faster than anticipated in children receiving transfusion, with mean (SD)

serum ferritin rising to and 2509 μg/L (974 μg/L) at 24 months

⊕⊝⊝⊝

Very low^{1, 2,3}

No comparative numbers reported

Children and adolescents with normalised TCD velocities

(18 months follow‐up)

79

(1 RCT)

Alloimmunisation

1 participant who was in the continuing transfusion arm developed an alloimmunisation

⊕⊝⊝⊝

Very low^{1, 2,3}

No comparative numbers reported

79

(1 RCT)

Transfusion reactions

7 participants in the continuing transfusion arm had 9 reactions to transfusions. 1 required hospitalisation

⊕⊝⊝⊝

Very low^{1, 2,3}

No comparative numbers reported

Children with normal TCD velocities and silent stroke

(median 3 years follow‐up)

121

(1 RCT)

Alloimmunisation

RR 3.16 (0.18 to 57.17)

⊕⊝⊝⊝

Very Low^1,2,3

We do not know if long‐term RBC transfusions increase the risk of alloimmunisation

121

(1 RCT)

Transfusion reactions

RR 5.17 (0.71 to 37.52)

⊕⊝⊝⊝

Very low^{1, 2,3}

We do not know if long‐term RBC transfusions increase the risk of transfusion reactions

121

(1 RCT)

Iron overload (serum ferritin)

Increased in children receiving long‐term RBC transfusions, incidence rate ratio 14.42 (5.41 to 875.17)

⊕⊕⊝⊝

Low^2.3

Long‐term RBC transfusions may increase the risk of iron overload

RBC transfusion versus disease‐modifying agents

Primary prevention

Children with abnormal TCD velocities but no severe vasculopathy on MRI/MRA, who have received at least one year of red cell transfusions

(6 months follow‐up)

121

(1 RCT)

Liver iron concentrations

MD ‐1.80 (‐5.16 to 1.56)

⊕⊕⊝⊝

Low ^1,2

Switching to hydroxyurea and phlebotomy may have little or no effect on liver iron concentrations

Secondary prevention

Children and adolescents with previous stroke, at least 18 months of RBC transfusions, and evidence of iron overload

(6 months follow‐up)

133

(1 RCT)

Liver iron concentrations

Hydroxyurea: 17.3 mg Fe/g dry weight iron, IQR 10.0 to 30.6; transfusion: 17.3 mg Fe/g dry

weight iron, IQR 8.8 to 30

⊕⊕⊝⊝

Low ^1,2

Switching to hydroxyurea and phlebotomy may have little or no effect on liver iron concentrations

Estcourt 2016b

Preoperative blood transfusions for sickle cell disease

Restrictive (to increase the total haemoglobin) versus liberal (to decrease the haemoglobin S level below a specified percentage) RBC transfusion strategy

People with SCD undergoing cholecystectomy

(Follow‐up: 30 days)

230

(1 RCT)

Alloimmunisation

RR 3.05 (1.14 to 8.20)

⊕⊝⊝⊝

Very low^1,2,3

We do not know whether aggressive RBC transfusions increase the risk of alloimmunisation

230

(1 RCT)

Transfusion reactions

RR 2.18 (0.77 to 6.18)

⊕⊝⊝⊝

Very low^1,2,3

We do not know whether aggressive RBC transfusions increase the risk of transfusion reactions

230

(1 RCT)

Any transfusion complication

RR 1.85 (0.89 to 3.88)

⊕⊝⊝⊝

Very low^{1, 2,3}

We do not know whether aggressive RBC transfusions increase the risk of transfusion complications

Short‐term RBC transfusion versus standard care

People with SCD undergoing low to medium risk surgery or any surgery other than cardiac surgery

(Follow‐up for 30 days in 1 trial; not stated in 1 trial)

434 (2)^b

Serious transfusion complications

No events in 1 trial, circulatory overload reported in the preoperative transfusion group in 1 trial

⊕⊝⊝⊝

Very low^1,3

We do not know whether preoperative RBC transfusions increase the risk of serious transfusion complications

65 (1)^b

Alloimmunisation

Peto OR 7.17 (0.14 to 361.44)

⊕⊝⊝⊝

Very low^1,3

We do not know whether preoperative RBC transfusions increase the risk of alloimmunisation

434 (2)

Transfusion reactions

No events in 1 trial; 2 reactions reported in the preoperative transfusion group in 1 trial

⊕⊝⊝⊝

Very low^1,3

We do not know whether preoperative RBC transfusions increase the risk of transfusion reactions

Okusanya 2016

Prophylactic versus selective blood transfusion for sickle cell disease in pregnancy

Long‐term RBC transfusions versus RBC transfusions to treat complications

Pregnant women with SCD

(Follow‐up for 21.5‐24.5 weeks of prenatal care)

72 (1)

Transfusion reactions

RR 2.00 (0.54 to 7.39)

⊕⊝⊝⊝

Very low^1,3

We do not know whether prophylactic RBC transfusions increase the risk of transfusion reactions

^a Serious and non‐serious transfusion related adverse events can include: acute and delayed transfusion reactions, transfusion‐related acute lung injury, transfusion‐associated circulatory overload, transfusion‐associated dyspnoea, alloimmunisation, iron overload, problems of venous access)

^b Rated down by 1 for serious risk of bias and rated down by 2 for very serious risk of bias

¹ High Risk of bias in one or more domains using the Cochrane Risk of Bias tool

² Indirectness as specific population or procedure reported may not be generalisable to review question

³ Imprecision as effect has wide CIs and/or study has small number of people or low event rates