Auteur, jaar

Studie-ontwerp

Studiepopulatie

Steekproef-grootte

Interventie(s)

Keeler 2017, Keeler 2019

RCT

Patiënten met een colorectale adenocarcinoom en met anemie op moment van rekrutering werden geincludeerd. Patiënten met metastasen, hematologische aandoeningen, onder behandeling waren met chemotherapie of contra-indicaties voor ijzer suppletie werden geëxcludeerd.

Interventie: n=55

Controle: n=61

Intraveneuze ijzer toediening (ferric carboxymaltose) versus. orale toediening van ijzer (ferrous sulphate)

Edwards 2009

RCT

Patiënten met een colorectaal-carcinoom die gepland stonden voor chirurgie werden geïncludeerd. Patiënten die 6 weken voorafgaand aan de rekrutering orale ijzer suppletie en/of bloedtransfusies ontvingen werden geëxcludeerd. Er werden subgroep analyses uitgevoerd voor deelnemers die anemie hadden op moment van rekrutering.

Interventie: n=34

Controle: n=27

Subgroep analyses (deelnemers met anemie op moment van rekrutering):

Interventie: n=9

Controle: n=9

Intraveneuze ijzer toediening (iron sucrose) versus. placebo

Lidder 2007

RCT

Alle patiënten (met een colorectaal carcinoom) die operabel waren.

Interventie: n=24

Controle: n=25

Orale toediening van ijzer (ferrous sulphate) versus. standaard klinische zorg (niet gedefinieerd)

Auteur, jaar

Interventie / Controle

Uitkomstmaat (klinisch relevant verschil)

Resultaat

Keeler 2019

Interventie:

Intraveneuze ijzer toediening (ferric carboxymaltose)

Controle:

Orale toediening van ijzer (ferrous sulphate)

Hemoglobine niveau (toename van 2mg/dl in 2 weken)

Niet gerapporteerd

Complicaties

Niet gerapporteerd

Mortaliteit

Niet gerapporteerd

Morbiditeit

Niet gerapporteerd.

Noodzaak voor bloedtransfusies

Duur van ziekenhuisverblijf

Niet gerapporteerd

Patiënt-tevredenheid / lasten

Kosteneffectiviteit

Niet gerapporteerd

Kwaliteit van leven

SF-36 Scores weren naar een 0-100 schaal getransformeerd.

SF-36, fysiek functioneren, mediane score (IQR) tijdens de rekrutering en in de follow-up op de dag van operatie, en poliklinisch:

I_recrutering: 60 (20-84)

I_operatie: 65 (30-90)

I_poliklinisch: 74 (45-95)

C_recrutering: 65 (31-84)

C_operatie: 65 (40-85)

C_poliklinisch: 70 (32-86)

Geen significante verschillen tussen groepen.

SF-36, lichamelijke pijn, mediane score (IQR) tijdens de rekrutering en tijdens de follow-up op de dag van operatie, en poliklinisch:

I_recrutering: 68 (41-100)

I_operatie: 80 (51-100)

I_poliklinisch: 84 (74-100)

C_recrutering: 72 (51-100)

C_operatie: 72 (46-92)

C_poliklinisch: 74 (52-100)

Geen significante verschillen tussen groepen.

SF-36, rol limitatie (door pijn), mediane score (IQR) tijdens de rekrutering en in de follow-up op de dag van operatie, en poliklinisch:

I_recrutering: 12.5 (0-100)

I_operatie: 50 (0-100)

I_poliklinisch: 100 (0-100)

C_recrutering: 25 (0-100)

C_operatie: 25 (0-100)

C_poliklinisch: 25 (0-100)

Significante verschillen tussen de groepen tijdens de poliklinische follow-up (p=0.01)

SF-36, rol limitatie (door emotie), mediane score (IQR) tijdens de rekrutering en tijdens de follow-up op de dag van operatie, en poliklinisch:

I_recrutering: 61 (47)

I_operatie: 60 (46)

I_poliklinisch: 80 (37)

C_recrutering: 57 (44)

C_operatie: 68 (44)

C_poliklinisch: 74 (43)

Significante verschillen tussen groepen tijdens de poliklinische follow-up (p=0.03)

SF-36, vitaliteit, gemiddelde score (SD) tijdens de rekrutering en tijdens de follow-up op de dag van operatie, en poliklinisch:

I_recrutering: 44 (24)

I_operatie: 59 (22)

I_poliklinisch: 72 (16)

C_recrutering: 47 (24)

C_operatie: 52 (21)

C_poliklinisch: 59 (19)

Significante verschillen tussen groepen op de dag van chirurgie (p=0.048) en in de poliklinische follow-up (p=0.00)

SF-36, sociaal functioneren, mediane score (IQR) tijdens de rekrutering en tijdens de follow-up op de dag van operatie, en poliklinisch:

I_recrutering: 63 (34-88)

I_operatie: 75 (50-100)

I_poliklinisch: 100 (88-100)

C_recrutering: 75 (50-86)

C_operatie: 62.5 (50-100)

C_poliklinisch: 75 (50-100)

Significante verschillen tussen groepen tijdens de poliklinische follow-up (p=0.03)

SF-36, mentale gezondheid, mediane score (IQR) tijdens de rekrutering en tijdens de follow-up op de dag van operatie, en poliklinisch:

I_recrutering: 76 (64-88)

I_operatie: 80 (72-92)

I_poliklinisch: : 92 (88-92)

C_recrutering: 76 (65-87)

C_operatie: 78 (64-89)

C_poliklinisch: 84 (72-92)

Significante verschillen tussen groepen tijdens de poliklinische follow-up (p=0.00)

SF-36, generieke gezondheid, mediane score (IQR) tijdens de rekrutering en tijdens de follow-up op de dag van operatie, en poliklinisch:

I_recrutering: 55 (45-77)

I_operatie: 63 (52-77)

I_poliklinisch: 77 (65-86)

C_recrutering: 57 (46-72)

C_operatie: 62 (50-77)

C_poliklinisch: 62 (50-77)

Significante verschillen tussen groepen tijdens de poliklinische follow-up (p=0.002)

SF-36, samenvatting fysieke component, gemiddelde score (SD) tijdens de rekrutering en tijdens de follow-up op de dag van operatie, en poliklinisch:

I_recrutering: 41 (11)

I_operatie: 43 (11)

I_poliklinisch: 47 (9)

C_recrutering: 42 (11)

C_operatie: 43 (10)

C_poliklinisch: 43 (9)

Geen significante verschillen tussen groepen.

De FACT-An scores werden niet gerapporteerd tijdens rekrutering en op de dag van chirurgie. De FACT-An subschalen ‘fysiek welzijn’ en ‘sociaal / familie welzijn’ werden niet gerapporteerd tijdens de poliklinische follow-up.

FACT-An, emotioneel welzijn, mediane score (IQR) tijdens poliklinische follow-up:

I: 22 (21-24)

C: 21 (20-22)

Significant verschil tussen groepen (p=0.033)

FACT-An, functioneel welzijn, mediane score (IQR) tijdens poliklinische follow-up:

I: 26 (23-28)

C: 22 (15-25)

Significant verschil tussen groepen (p=0.001)

FACT-An, anemie specifiek, mediane score (IQR) tijdens poliklinische follow-up:

I: 71 (66-77)

C: 66 (55-72)

Significant verschil tussen groepen (p=0.002)

FACT-An, trial specifieke index, mediane score (IQR) tijdens poliklinische follow-up:

I: 121 (113-124)

C: 108 (90-123)

Significant verschik tussen groepen (p=0.003)

FACT-An, totale score, mediane score (IQR) tijdens poliklinische follow-up:

I: 168 (160-174)

C: 151 (132-170)

Significant verschil tussen groepen (p=0.003)

Directe vergelijkingen tussen groepen op de EQ5D5I (subschalen) voor kwaliteit van leven werden niet gerapporteerd.

Keeler 2017

Interventie:

Intraveneuze ijzer toediening (ferric carboxymaltose)

Controle:

Orale toediening van ijzer (ferrous sulphate)

Hemoglobine niveau (toename van 2mg/dl in 2 weken)

Hemoglobineniveau op de dag van operatie, gemiddelde g/dl (95%BHI):

I: 11.9 (11.5 tot 12.3)

C: 11.0 (10.6 tot 11.4)

Significant verschil tussen groepen (p=0.002)

Hemoglobineniveau, mediaan van de verschillen (IQR) tussen de rekrutering en de dag van operatie:

I: 1.55 g/dl (0.93 tot 2.58)

C: 0.50 g/dl (-0.13 tot 1.33)

Significant verschil tussen groepen (p<0.001)

Complicaties

Er was geen verschil in de ernst van de complicaties (p=0.995) of de complicatie rate (p=0.305) tussen de groepen. Er was geen verschil in graad van infectiecomplicaties (p=0.083) en de rate van infectiecomplicaties (p=0.091). De frequentie van complicaties werd niet gerapporteerd.

Mortaliteit

Mortaliteit in de follow-up tot ontslag, n (%):

I: 5/51 (9.8%)

C: 4/57 (7%)

Geen significant verschil tussen groepen (p=0.872)

Mortaliteit in de eerste 90 dagen, n (%):

I: 3/51 (5.9%)

C: 2/57 (3.5%)

Geen significant verschil tussen groepen (p=0.906)

Morbiditeit

Niet gerapporteerd.

Noodzaak voor bloedtransfusies

Algehele aantal patiënten die een bloedtransfusie nodig hadden, n (%):

I: 10/55 (18.2%)

C: 14/61 (23%)

Geen significant verschil tussen groepen. (p=0.470)

Aantal patiënten die een bloedtransfusie nodig hadden op de dag van operatie, n (%):

I: 6/55 (10.9%)

C: 6/61 (9.8%)

Geen significant verschil tussen groepen (p=0.894)

Duur van ziekenhuisverblijf

Niet gerapporteerd

Patiënt-tevredenheid / lasten

Noodzaak voor het postoperatief gebruik van supplementen, n (%):

I: 4/51 (7.8%)

C: 30/57 (52.6%)

Significant verschil tussen groepen (p<0.001)

Kosteneffectiviteit

Niet gerapporteerd

Kwaliteit van leven

Niet gerapporteerd

Edwards 2009

Interventie:

Intraveneuze ijzer toediening (iron sucrose)

Controle:

Placebo

Hemoglobine niveau (toename van 2mg/dl in 2 weken)

Hemoglobineniveau, het gemiddelde verschil in g/dl (95%BHI) tussen rekrutering en dag van operatie:

I: -0.19 (-0.74 tot 0.36)

C: -0.50 (-1.00 tot -0.01)

Geen significant verschil tussen groepen (p=0.273).

Hemoglobineniveau het gemiddelde niveau in g/dl (95%BHI) vóór de operatie (95%CI):

I: 13.4 (12.5 tot 14.3)

C: 13.6 (12.6 tot 14.6)

Geen significant verschil tussen groepen (p-waarde niet gerapporteerd)

Hemoglobineniveau, gemiddelde niveau in g/dl (95%BHI) 1 dag postoperatief:

I: 11.2 (10.4 tot 12.0)

C: 11.3 (10.7 tot 12.0)

Geen significant verschil tussen groepen (p-waarde niet gerapporteerd)

Hemoglobineniveau, het gemiddelde niveau in g/dl (95%BHI) bij ontslag:

I: 11.2 (10.4 to t12.1)

C: 11.7 (10.9 tot 12.5)

Geen significant verschil tussen groepen (p-waarde niet gerapporteerd)

Complicaties

Niet gerapporteerd.

Mortaliteit

Niet gerapporteerd

Morbiditeit

Niet gerapporteerd

Noodzaak voor bloedtransfusies

Aantal patiënten die een bloedtransfusie nodig hadden, n (%):

I: 2/34 (5.9%%)

C: 5/26 (19.2%)

Geen significant verschil tussen groepen (p=0.335).

Duur van ziekenhuisverblijf

Mediaan aantal dagen van het ziekenhuisverblijf:

I: 10 dagen

C: 8 dagen

Geen significant verschil tussen groepen (p=0.273)

Patiënt-tevredenheid / lasten

Niet gerapporteerd.

Kosteneffectiviteit

Niet gerapporteerd.

Subgroep analyse van deelnemers met anemie (n=9 in zowel de interventie- als controlegroep)

Interventie:

Intraveneuze ijzer toediening (iron sucrose)

Controle:

Placebo

Hemoglobine niveau (toename van 2mg/dl in 2 weken)

Hemoglobine niveau, het gemiddeld niveau in g/dl (95%BHI) op moment van rekrutering in de studie:

I: 11.7 (10.6-12.7)

C: 11.8 (10.2-13.5)

Geen significant verschil tussen groepen (p-waarde niet gerapporteerd)

Hemoglobineniveau, gemiddeld niveau in g/dl (95%BHI) vóór de operatie:

I: 11.2 (9.7-12.6)

C: 11.9 (9.9-14)

Geen significant verschil tussen groepen (p-waarde niet gerapporteerd)

Hemoglobineniveau, gemiddeld niveau in g/dl (95%BHI) 1 dag postoperatief:

I: 9.6 (8.9-10.4)

C: 10.7 (9.8-11.6)

Geen significant verschil tussen groepen (p-waarde niet gerapporteerd)

Hemoglobineniveau, gemiddelde niveau in g/dl (95%BHI) bij ontslag:

I: 10.2 (8.4-12.1)

C: 11.3 (10.4-12.2)

Geen significant verschil tussen groepen (p-waarde niet gerapporteerd)

Hemoglobineniveau, het gemiddelde verschil in g/dl (95%BHI) tussen rekrutering en dag van operatie:

I: -0.46 (-0.79 tot -0.12)

C: -0.11 (-0.63 tot 0.41)

Geen significant verschil tussen groepen (p=0.223)

Complicaties

Niet gerapporteerd.

Mortaliteit

Niet gerapporteerd.

Morbiditeit

Niet gerapporteerd.

Noodzaak voor bloedtransfusies

Niet gerapporteerd.

Duur van ziekenhuisverblijf

Niet gerapporteerd.

Patiënt-tevredenheid / lasten

Niet gerapporteerd.

Kosteneffectiviteit

Niet gerapporteerd.

Kwaliteit van leven

Niet gerapporteerd

Lidder 2007

Interventie:

Orale toediening van ijzer (ferrous sulphate)

Controle:

Standaard klinische zorg (niet gedefinieerd)

Hemoglobine niveau (toename van 2mg/dl in 2 weken)

Hemoglobineniveau, het gemiddelde niveau in g/dl (SD) bij opname na de operatie:

I: 13.1 (2.0)

C: 11.8 (2.0)

Significant verschil tussen groepen (p=0.04). De controlegroep had op de baseline meet deelnemers met anemie (56%) dan de interventiegroep (25%).

Complicaties

Niet gerapporteerd.

Mortaliteit

Niet gerapporteerd.

Morbiditeit

Niet gerapporteerd.

Noodzaak voor bloedtransfusies

Aantal patiënten die een bloedtransfusie nodig hadden, n (%):

I: 6/23 (26.1%)

C: 13/22 (59.1%)

Significant verschil tussen groepen (p=0.047), hoewel de controlegroep meer patiënten met anemie op de baseline meting had (25% versus. 56%).

Duur van ziekenhuisverblijf

Niet gerapporteerd.

Patiënt-tevredenheid / lasten

Niet gerapporteerd.

Kosteneffectiviteit

Niet gerapporteerd.

Kwaliteit van leven

Niet gerapporteerd

Summary of findings:

Iron suppletion compared to care as usual or no iron suppletion for patients with a colorectal carcinoma in the post-operative phase (Borstlap 2015 (systematic review) + Keeler 2017 & 2019)

Patient or population: patients with a colorectal carcinoma in the post-operative phase

Setting:

Intervention: iron suppletion

Comparison: care as usual or no iron suppletion

Outcomes

Anticipated absolute effects^* (95% CI)

Relative effect
(95% CI)

№ of participants
(studies)

Certainty of the evidence
(GRADE)

Conclussion

Risk with care as usual or no iron suppletion

Risk with iron suppletion

Haemoglobin level (between recruitment and surgery), g/dl

One RCT compared intravenous iron with placebo. There were no differences between groups in the mean change of haemoglobin level between recruitment and surgery (I: -0.19 (95%CI: -0.74 to 0.36), C: -0.50 (95%CI: -1.00 to -0.01), p=0.273). The RCT also performed a subanalysis for 18 participants (n=9 in each group) who had anema at baseline for change in heamoglobin level in the same period (I: -0.46 (95%CI: -0.79 to -0.12), C: -0.11 (95%CI: -0.63 to 0.41), p=0.223). The uncertainy for the subgroup analysis is higher, due to a very small sample size.

61
(1 RCT)

⨁⨁◯◯
LOW ^a

Intravenous iron suppletion might not increase or deacrease the haemoglobinlvel in the period before surgery.

Haemoglobin level (before surgery / on the day of surgery), g/dl

Two RCTs described the mean hemoglobin level before surgery or on the day of surgery. One of these RCTs found a significant difference on the day of surgery when comparing intravenous iron (I) with oral iron (C) (I: 11.9 (95%CI: 11.5-12.3), C: 11.0 (95%CI: 10.6-11.4), p=0.002) for participants who were anaemic at recruitment. The other RCT did not find a significant difference before surgery when comparing intravenous iron (I) to placebo (C) (I: 13.4 (95%CI: 12.5-14.3), C: 13.6 (95%CI: 12.6-14.6), p-value not reported). For 18 anemic participants (n=9 in each group), the latter RCT performed a subanalysis for the haemoglobin level before surgery as well, however found no differences between groups (I: 11.2 (95%CI: 9.7-12.6), C: 11.9 (95%CI: 9.9-14), p-value not reported).

177
(2 RCTs)

⨁⨁◯◯
LOW ^b,c

Intravenous iron suppletion might or might not increase haemoglobin levels before surgery.

Haemoglobin level (shortly after surgery), g/dl

Two RCTs decribed the haemoglobin level shortly after surgery. One RCT reported this as the mean haemoglobin level 1 day post-operatively and did not find a significant difference between groups receiving intravenous iron (I) versus. placebo (C) (I: 11.2 (95%CI: 10.4-12.0), C: 11.3 (95%CI: 10.7-12.0), p-value not reported). Subgroup analyses were also performed for 18 participants (n=9 in both groups) with anaema at recruitment for one day postoperatively while no significant differences were found (I: 9.6 (95%CI: 8.4-10.4), C: 10.7 (95%CI: 9.8-11.6), p-value not reported). The other RCT reported the mean haemoglobin level at admission after surgery and found a significant difference when comparing oral iron (I) with standard care (C, undefined) (I: 13.1 (SD: 2.0), C: 11.8 (SD: 2.0), p=0.04). However there were more anemic participants in the control group (56%) than in the intervention group (25%) at baseline.

110
(2 RCTs)

⨁◯◯◯
VERY LOW ^d,e

Iron suppletion might or might not increase haemoglobin levels shortly after surgery.

Haemoglobin level (at discharge), g/dl

One RCT described the mean haemoglobin level at the moment of discharge. This RCT compared intravenous iron (I) to placebo (C) and found no significant differences (I: 11.2 (95%CI: 10.4-12.0), I: 11.3 (95%CI: 10.7-12.0), p-value not reported). This RCT also performed a subgroup analysis of 18 participants who were anemic at baseline and found no significant difference (I: 10.2 (95%CI: 8.4-12.1), I: 11.3 (95%CI: 10.4-12.2), p-value not reported). The uncertainy for the subgroup analysis is higher, due to a very small sample size.

61
(1 RCT)

⨁⨁◯◯
LOW ^a

Intravenous iron suppletion might not increase or decrease haemoglobn levels at discharge.

Mortality (in follow-up until discharge)

70 per 1.000

98 per 1.000
(28 to 345)

RR 1.40
(0.40 to 4.92)

108
(1 RCT) ^f

⨁◯◯◯
VERY LOW ^g,h

We are unsure of the effect of iron suppletion on mortality until discharge

Mortality (first 90 days)

35 per 1.000

59 per 1.000
(10 to 338)

RR 1.68
(0.29 to 9.64)

108
(1 RCT)

⨁◯◯◯
VERY LOW ^g,h

We are unsure of the effect of iron suppletion on mortality 90 days postoperatively

Morbidity and complications

There was no difference between groups for severity (p.995) or complication rate (p=0.305). There was no difference between groups for infective complication grade (p=0.083) or rate (p=0.091). However, number of events / grades / rates were not reported.

116
(1 RCT)

⨁◯◯◯
VERY LOW ^a,g

We are unsure of the effect of iron suppletion on morbidity and complications

Need for bloodtransfusion

Data was not pooled since all RCTs had different comparisons. One RCT compared intravenous iron (I) to oral iron (C) and found no differences in the overall need for blood transfusions (I: 10/55, C: 14/61, RR=0.79 (95%CI: 0.38-1.64), p=0.47) and for the need for blood transfusion on the day of surgery (I: 6/55, C: 6/61, RR= 1.11 (95%CI: 0.38-3.24), p=0.894). The second RCT compared intravenous iron (I) to placebo (C) and found no significant differences between groups for the overall need for blood transfusions (I: 2/34, C: 5/26, RR= 0.31 (95%CI: 0.06-1.45), p=0.335). The third RCT found a significant difference between groups when comparing oral iron (I) to standard care (C, undefined) (I: 6/23, C: 13/22, RR= 0.44 (95%CI: 0.20-0.95), p=0.047), however the control group had more anemic participants (56%) than the interventrion group (25%) on baseline.

226
(3 RCTs)

⨁◯◯◯
VERY LOW ^h,i

We are unsure of the effect of iron suppletion on the need for bloodtransfusion

Length of hospital stay

One RCT recorded the length of hospital stay for participants that received intravenous iron (intervention) and compared this to participants who received placebo (control). The median length of stay was 10 days for the intervention group and 8 days for the control group (no significant difference, p=0.273).

61
(1 RCT)

⨁◯◯◯
VERY LOW ^j,k

We are unsure of the effect of intravenous iron suppletion on the length of stay.

patient burden (need for postoperative use of supplements)

526 per 1.000

79 per 1.000
(32 to 205)

RR 0.15
(0.06 to 0.39)

108
(1 RCT) ^f

⨁⨁◯◯
LOW ^g,k

Iron suppletion might decrease the need for postoperative supplements.

Cost-effectiveness - not reported

-

-

The effect of iron suppletion on the cost-effectiveness is unknown.

Quality of life

One RCT described the quality of life in a sample receiving either oral iron (care as usual) or intravenous iron. Quality of life was measured at recruitment, on the day of surgery (before surgery), and at an outpatient visit (approximately 3 monts after surgery). The construct was measured with the EQ5D5I, the FACT-An, and the SF-36. No comparative data between groups was reported for the EQ5D5I at any follow-up time point. No comparative data between groups was reported for the FACT-An at the time points 'at recruitment' and 'on the day of surgery'. At the outpatient follow-up time point, no data was reported for the FACT-An subscales 'physical well-being' and 'social/family well-being'). However, significant differences were reported at the outpaient follow-up time point for the FACT-An subscales 'emotional well-being' (oral: 21 (IQR: 20-22), intravenous: 22 (IQR: 21-24), p=0.033), 'functional well-being' (oral:22 (IQR: 15-25), intravenous: 26 (IQR: 23-28), p=0.001), 'anaemia specific scale' (oral: 66 (IQR: 55-72), intravenous: 71 (IQR: 66-77), p=0.002), 'trial index' (oral: 108, intravenous: 121 (IQR: 113-124), p=0.002), and the total score (oral: 151 (IQR: 160-174), intravenous: 168 (IQR: 160-174), p=0.003). None of the SF-36 scores showed significant differences when groups were compared at recruitment. At the day of surgery time point only the SF-36 subscale 'vitality' differed significantly between groups (oral: 52 (SD: 21), intravenous: 59 (SD: 22), p=0.048). Significant differences between groups at the outpatient follow-up were found for the subscales 'Role limitation due to pain' (oral: 25 (IQR: 0-100), intravenous: 100 (IQR: 50-100), p=0.01), 'General health' (oral: 62 (IQR: 50-77), intravenous: 77 (IQR: 65-86), p=0.002), 'Vitality' (oral: 59 (SD: 19), intravenous: 72 (SD: 16), p=0.00), 'Role limitation due to emotion' (oral: 74 (SD: 43), intravenous: 80 (SD: 37), p=0.03), and the subscale 'Mental health' (oral: 84 (IQR: 72-92), intravenous: 92 (IQR: 88-92), p=0.00)

(1 RCT)

⨁◯◯◯
VERY LOW ^l,m

We are unsure whether iron suppletion affects the quality of life of patients.However some aspects might possibly improve when intravenous suppletion is compared to oral suppletion in anemic patients.

*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).

CI: Confidence interval; RR: Risk ratio

GRADE Working Group grades of evidence
High certainty: We are very confident that the true effect lies close to that of the estimate of the effect
Moderate certainty: We are moderately confident in the effect estimate: The true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different
Low certainty: Our confidence in the effect estimate is limited: The true effect may be substantially different from the estimate of the effect
Very low certainty: We have very little confidence in the effect estimate: The true effect is likely to be substantially different from the estimate of effect

Study reference

Study characteristics

Patient characteristics ²

Intervention (I)

Comparison / control (C) ³

Follow-up

Outcome measures and effect size ⁴

Comments

Keeler 2019

Type of study:

RCT

Setting and country:

Multicenter (hospital), UK

Funding and conflicts of interest:

Funded independently by the National Institute for Health Research (UK). Ferinject was donated by Vifor Pharma to all study centers, except for Nothingham University Hospitals NHS Trust. One author’s (MB) research department received grants from Syner-Med and Vifor Pharma. MB received honoraria and travel support from Vifor Pharma and Merck Sharp And Dohme Limited for consulting or lecturing. Another autho’s (AA) research department receiver grants from Syner-Med, Vifor Pharma, and Pharmacosmos A/S. AA received honoraria and travel support from Ethicon Endosurgery, Johnson and Johnsosn Ltd, Olympus, and Vifor Pharma for consulting or lecturing. A third author (ON) received honoraria and travel support from Pharmacosmos A/S for consulting.

Inclusion criteria:

Patients with colorectal adenocarcinoma, anaematic at recruitment (<11 g/dl for women, <12 g/dl for men). Reported in Keeler 2017.

Exclusion criteria:

Metastatic disease, pre-existing haematological disease, renal failure, currently undergoing chemotherapy, pinors, prisoners, pregnant women, patients with contraindications to iron supplementation. Reported in Keeler 2017.

N total at baseline:

Intervention:

Control:

Important prognostic factors²:

age ± SD:

I: 73.8 (8.9)

C: 76.5 (10.9)

Sex:

I: 35M/20F

C: 37M/24F

Weight, mean kg (SD):

I: 79.1 (15.3)

C: 72.7 (17.2)

Mean inclusion Hb g/l^-1 (SD):

I: 96 (13)

C: 99 (11)

ASA status, n:

I: status I-II: 30, status III-IV: 25

C: status I-II: 43, status III-IV: 18

Groups comparable at baseline?

There are more ASA III-IV in the intervention group than in the control group (45.5% versus 29.5%)

Describe intervention (treatment/procedure/test):

Ferric carboxymaltose (Ferinject) was diluted in 250ml saline (0.9%) and intravenously infused over 15 minutes.

Dose was determined by bodyweight and inclusion haemoglobin value:

1000 mg in 1 dose: ≥10 g/dl haemoglobin and <70kg bodyweight

1500 mg in 2 doses: <10 mg/dl haemoglobin and <70kg bodyweight OR ≥10 g/dl haemoglobin and ≥70kg bodyweight

2000mg in 2 doses: <10 g/dl haemoglobin and ≥70kg bodyweight

A maximum of 1000mg per dose per week was given and a maximum of 2000mg during the whole trial. A second dose was, if required, administered at least 7 days after the first dose.

Describe control (treatment/procedure/test):

Oral iron supplements (ferrous sulphate, 200mg) twice daily.

Length of follow-up:

At day of surgery (before surgery)

In an outpatient setting (approximal 3 months after surgery)

Loss-to-follow-up:

Intervention:

13 (24%) from recruitment to outpatient visit

Reasons: death (n=5), moved out (n=2), did not attend outpatient visit (n=4), did not undergo resection (n=2)

Control:

11 (18%) from recruitment to outpatient visit.

Reasons: death (n=4), did not attend outpatient visit (n=3), did not undergo resection (n=4)

Incomplete outcome data:

Intervention:

13 (24%)

Reasons: see loss-to-follow-up

Control:

11 (18%)

Reasons: see loss-to-follow-up

Missing data were imputed

Outcome measures and effect size (include 95%CI and p-value if available):

SF-36 scores were transformed into a 0-100 scale

SF-36, scale physical functioning, median score (IQR) at timepoints recruitment / day of surgery / outpatient follow-up:

I_recruitment: 60 (20-84)

I_surgery: 65 (30-90)

I_outpatient: 74 (45-95)

C_recruitment: 65 (31-84)

C_surgery: 65 (40-85)

C_outpatient: 70 (32-86)

No significant differences between groups at any time point.

SF-36, scale bodily pain, median score (IQR) at timepoints recruitment / day of surgery / outpatient follow-up:

I_recruitment: 68 (41-100)

I_surgery: 80 (51-100)

I_outpatient: 84 (74-100)

C_recruitment: 72 (51-100)

C_surgery: 72 (46-92)

C_outpatient: 74 (52-100)

No significant differences between groups at any time point.

SF-36, scale role limitations (due to pain), median score (IQR) at timepoints recruitment / day of surgery / outpatient follow-up:

I_recruitment: 12.5 (0-100)

I_surgery: 50 (0-100)

I_outpatient: 100 (0-100)

C_recruitment: 25 (0-100)

C_surgery: 25 (0-100)

C_outpatient: 25 (0-100)

Significant difference between groups at time point: patient follow-up (p=0.01)

SF-36, scale rol limitations (due to emotion), score (SD) at timepoints recruitment / day of surgery / outpatient follow-up:

I_recruitment: 61 (47)

I_surgery: 60 (46)

I_outpatient: 80 (37)

C_recruitment: 57 (44)

C_surgery: 68 (44)

C_outpatient: 74 (43)

Significant difference between groups at time point: patient follow-up (p=0.03)

SF-36, vitality, score (SD) at timepoints recruitment / day of surgery / outpatient follow-up:

I_recruitment: 44 (24)

I_surgery: 59 (22)

I_outpatient: 72 (16)

C_recruitment: 47 (24)

C_surgery: 52 (21)

C_outpatient: 59 (19)

Significant difference between groups at time point: day of surgery (p=0.048) and patient follow-up (p=0.00)

SF-36, scale social functioning, median score (IQR) at timepoints recruitment / day of surgery / outpatient follow-up:

I_recruitment: 63 (34-88)

I_surgery: 75 (50-100)

I_outpatient: 100 (88-100)

C_recruitment: 75 (50-86)

C_surgery: 62.5 (50-100)

C_outpatient: 75 (50-100)

Significant difference between groups at time point: patient follow-up (p=0.03)

SF-36, scale mental health, median score (IQR) at timepoints recruitment / day of surgery / outpatient follow-up:

I_recruitment: 76 (64-88)

I_surgery: 80 (72-92)

I_outpatient: 92 (88-92)

C_recruitment: 76 (65-87)

C_surgery: 78 (64-89)

C_outpatient: 84 (72-92)

Significant difference between groups at time point: patient follow-up (p=0.00)

SF-36, scale general health, median score (IQR) at timepoints recruitment / day of surgery / outpatient follow-up:

I_recruitment: 55 (45-77)

I_surgery: 63 (52-77)

I_outpatient: 77 (65-86)

C_recruitment: 57 (46-72)

C_surgery: 62 (50-77)

C_outpatient: 62 (50-77)

Significant difference between groups at time point: patient follow-up (p=0.002)

SF-36, scale physical component summary, mean score (SD) at timepoints recruitment / day of surgery / outpatient follow-up:

I_recruitment: 41 (11)

I_surgery: 43 (11)

I_outpatient: 47 (9)

C_recruitment: 42 (11)

C_surgery: 43 (10)

C_outpatient: 43 (9)

No significant differences between groups at any time point.

FACT-An, scales at outpatient follow-up: physical well-being, social / family well-being, were not reported.

FACT-An comparative scores were not reported for the time points: at recruitment, and at the day of surgery.

FACT-An, scale emotional well-being, median score (IQR) at outpatient follow-up:

I: 22 (21-24)

C: 21 (20-22)

Significant difference between groups (p=0.033)

FACT-An, scale functional well-being, median score (IQR) at outpatient follow-up:

I: 26 (23-28)

C: 22 (15-25)

Significant difference between groups (p=0.001)

FACT-An, scale anemia specific, median score (IQR) at outpatient follow-up:

I: 71 (66-77)

C: 66 (55-72)

Significant difference between groups (p=0.002)

FACT-An, scale trial index specific, median score (IQR) at outpatient follow-up:

I: 121 (113-124)

C: 108 (90-123)

Significant difference between groups (p=0.003)

FACT-An, total score, median score (IQR) at outpatient follow-up:

I: 168 (160-174)

C: 151 (132-170)

Significant difference between groups (p=0.003)

EQ5D5l comparative scores were not reported for time points: at recruitment, at day of surgery, and at outpatient follow-up for any of the scales.

N=4 (oral group) and n=2 (intravenous group) did not receive surgical resection after treatment allocation.

Keeler 2017

Type of study:

RCT

Setting and country:

Multicenter (hospital), UK

Funding and conflicts of interest:

Not reported

Inclusion criteria:

Patients with colorectal adenocarcinoma, anaematic at recruitment (<11 g/dl for women, <12 g/dl for men)

Exclusion criteria:

Metastatic disease, pre-existing haematological disease, renal failure, currently undergoing chemotherapy, pinors, prisoners, pregnant women, patients with contraindications to iron supplementation

N total at baseline:

Intervention: 55

Control: 61

Important prognostic factors²:

Median age (IQR):

I: 73.8 (67.4-78.6)

C: 74.7 (67.9-80.8)

Sex:

I: 63.6% M

C: 60.6% M

Receiving oral iron at recruitment, n (%):

I: 25/55 (45.5%)

C: 30/61 (49.2%)

ASA grade, n (%):

I: ASA I-II: 30 (54.4%), ASA III-IV: 25 (45.5%)

C: ASA I-II: 43 (70.6%), ASA III-IV: 18 (29.5%)

Tumour details:

I: T1-2: 8 (14.5%), T3-4: 45 (85.5%)

C: T1-2: 5 (8.2%), T3-4 (91.8%)

Groups comparable at baseline?

yes

Describe intervention (treatment/procedure/test):

Ferric carboxymaltose was diluted in 250ml saline (0.9%) and intravenously infused over 15 minutes.

Dose was determined by bodyweight and inclusion haemoglobin value:

1000 mg in 1 dose: ≥10 g/dl haemoglobin and <70kg bodyweight

1500 mg in 2 doses: <10 mg/dl haemoglobin and <70kg bodyweight OR ≥10 g/dl haemoglobin and ≥70kg bodyweight

2000mg in 2 doses: <10 g/dl haemoglobin and ≥70kg bodyweight

A maximum of 1000mg per dose per week was given and a maximum of 2000mg during the whole trial. A second dose was, if required, administered at least 7 days after the first dose.

Describe control (treatment/procedure/test):

Oral iron supplements (ferrous sulphate, 200mg) twice daily.

Length of follow-up:

Until the patients’ routine postoperative outclinic visit (approximately 2-3 months after discharge from the hospital)

Loss-to-follow-up:

Intervention: 2/55 (3.6 %)

Reason: did not undergo surgical resection

Control: 4/61 (6.6%)

Reasons: did not undergo surgical resection

Incomplete outcome data:

Intervention: 2/55 (3.6 %)

Reason: did not undergo surgical resection

Control: 4/61 (6.6%)

Reasons: did not undergo surgical resection

Outcome measures and effect size (include 95%CI and p-value if available):

Need for blood transfusion (overall number of patients transfused), n (%):

I: 10/55 (10.9%)

C: 14/61 (23%)

No significant difference between groups (p=0.470)

Need for blood transfusion (number of patients transfused on day of surgery), n (%):

I: 6/55 (10.9%)

C: 6/61 (9.8%)

No significant difference between groups (p=0.894)

Haemoglobin level at surgery, mean g/dl (95%CI):

I: 11.9 (11.5 to 12.3)

C: 11.0 (10.6 to 11.4)

Significant difference between groups (p=0.002)

Haemoglobin level, median difference (IQR) from recruitment to surgery:

I: 1.55 g/dl (0.93 to 2.58)

C: 0.50 g/dl (-0.13 to 1.33)

Significant difference between groups (p<0.001)

Patient burden (need for supplements postoperatively), n (%):

I: 4/51 (7.8%)

C: 30/57 (52.6%)

Significant differences between groups (p<0.001)

Postoperative length of hospital stay, median days (IQR):

I: 6 (5-10)

C: 6 (4-9)

No significant differences between groups (p=0.95)

Mortality (whole follow-up), n (%):

I: 5/51 (9.8%)

C: 4/57 (7%)

No significant differences between groups (p=0.872)

90-day mortality, n (%):

I: 3/51 (5.9%)

C: 2/57 (3.5%)

No significant differences between groups (p=0.906)

Complications:

There was no difference between groups for severity (p.995) or complication rate (p=0.305). There was no difference between groups for infective complication grade (p=0.083) or rate (p=0.091). However, number of events / grades / rates were not reported.

Protocol deviations in control group (n=11) and in the intervention group (n=4).

Edwards 2009

Type of study: RCT

Setting and country: Hospital, UK

Funding and conflicts of interest:

Not reported

Inclusion criteria:

Scheduled to undergo resection for suspected colorectal cancer.

Exclusion criteria:

<18 years, oral iron supplementation and/or blood transfusion 6 weeks prior of recruitment, date of scheduled surgery within 15 days after recruitment.

N total at baseline:

Intervention: 35

Control: 27

Important prognostic factors (based on n=34 and n=26)²:

Median age:

I: 67

C: 70

Sex (M/F):

I: 22/12

C: 17/9

NSAID / antiplatelet / anticoagulant use, n (%):

I: 13/34 (38.2%)

C: 9/26 (34.6%)

Groups comparable at baseline?

Yes (based on the few baseline variables)

Describe intervention (treatment/procedure/test):

Two infusions of 300mg iron sucrose in 250ml saline (0.9%), provided as 15ml ferric hydroxide with sucrose in a 2% solution. Therefore the total dose was 600mg iron sucrose.

A test dose of 5ml was given and when the patient remained stable, the remainder was infused over a period of 1 hour.

The minimum time between infusions was 24 hours and were completed within a minimum of 14 days before surgery.

Describe control (treatment/procedure/test):

Two infusions of 250ml saline (0.9%).

A test dose of 5ml was given and when the patient remained stable, the remainder was infused over a period of 1 hour.

The minimum time between infusions was 24 hours and were completed within a minimum of 14 days before surgery.

Length of follow-up:

Until discharge (not specified, median time not reported)

Loss-to-follow-up:

Intervention: 2/27 (7.4%)

Reasons: unsuitable for surgery (n=1), did not attend second infusion (n=1)

Control: 2/35 (5.7%)

Reasons: unsuitable for surgery (n=1), did not attend second infusion (n=1)

Incomplete outcome data:

Intervention: 2/27 (7.4%)

Reasons: unsuitable for surgery (n=1), did not attend second infusion (n=1)

Control: 2/35 (5.7%)

Reasons: unsuitable for surgery (n=1), did not attend second infusion (n=1)

Outcome measures and effect size (include 95%CI and p-value if available):

Haemoglobin level (between recruitment and surgery), mean difference g/dl (95%CI):

I: -0.19 (-0.74 to 0.36)

C: -0.50 (-1.00 to -0.01)

No significant differences between groups (p=0.273).

Haemoglobin level (before surgery) mean g/dl (95%CI):

I: 13.4 (12.5 to 14.3)

C: 13.6 (12.6 to 14.6)

No significant differences between groups (p-value not reported)

Haemoglobin level (1 day postoperatively), mean g/dl (95%CI):

I: 11.2 (10.4 to 12.0)

C: 11.3 (10.7 to 12.0)

No significant differences between groups (p-value not reported)

Haemoglobin level (at discharge), mean g/dl (95%CI):

I: 11.2 (10.4 to 12.1)

C: 11.7 (10.9 to 12.5)

No significant differences between groups (p-value not reported)

Length of hospital stay, median days:

I: 10 days

C: 8 days

No significant differences between groups (p=0.273)

Need for blood transfusion, n (%):

I: 2/34 (5.9%%)

C: 5/26 (19.2%)

No significant differences between groups (p=0.335).

Lidder 2007

Type of study:

RCT

Setting and country: Hospital, UK

Funding and conflicts of interest:

Not reported

Inclusion criteria:

All patients fit for surgery

Exclusion criteria:

Not described.

N total at baseline:

Intervention: 24

Control:25

Important prognostic factors²:

age:

I: 69 (47-89)

C: 72 (57-80)

Sex (M/F):

I: 14/8

C: 14/9

Anaemic patients, n (%):

I: 6/24 (25%)

C: 14/25 (56%)

Groups comparable at baseline?

No, more anaemic patients in control group at baseline.

Describe intervention (treatment/procedure/test):

Three times per day: oral ferrous sulphate 200mg until surgery.

Describe control (treatment/procedure/test):

Standard clinical management.

Length of follow-up:

Until discharge (length of period not reported)

Loss-to-follow-up:

Intervention: 1/24 (4.2%) Reasons: unsuitable for surgery

Control: 3/25 (12%)

Reasons: unsuitable for resection

Incomplete outcome data:

Unclear, besides the reported loss to follow-up.

Outcome measures and effect size (include 95%CI and p-value if available):

Haemoglobin level (at admission after surgery), mean g/dl (SD):

I: 13.1 (2.0)

C: 11.8 (2.0)

Significant differences between groups (mean difference: 1.3, 95%CI: 0.06-2.51, p=0.04). However, the control group had more anaemic patients from baseline (25% versus 56%).

Need for allogeneic transfusion, n (%):

I: 6/23 (26.1%)

C: 13/22 (59.1%)

Significant difference between groups (p=0.047). However, the control group had more anaemic patients from baseline (25% versus 56%).

SUBGROUP ANALYSES (anaemic participants, n=9 in each group)

Haemoglobin level (at recruitment), mean g/dl (95% CI):

I: 11.7 (10.6-12.7)

C: 11.8 (10.2-13.5)

No significant difference between groups (p-value not reported)

Haemoglobin level (before surgery), mean g/dl (95% CI):

I: 11.2 (9.7-12.6)

C: 11.9 (9.9-14)

No significant difference between groups (p-value not reported)

Haemoglobin level (1 day postoperatively), mean g/dl (95% CI):

I: 9.6 (8.9-10.4)

C: 10.7 (9.8-11.6)

No significant difference between groups (p-value not reported)

Haemoglobin level (at discharge), mean g/dl (95% CI):

I: 10.2 (8.4-12.1)

C: 11.3 (10.4-12.2)

No significant difference between groups (p-value not reported)

Haemoglobin level (from recruitment to preoperative value), mean change g/dl (95% CI):

I: -0.46 (-0.79 to -0.12)

C: -0.11 (-0.63 to 0.41)

No significant difference between groups (p=0.223)

The authors did not define the standard clinical care.

Length of hospital stay was not reported, although this was stated in the methods.

Study reference

(first author, publication year)

Describe method of randomisation¹

Bias due to inadequate concealment of allocation?²

(unlikely/likely/unclear)

Bias due to inadequate blinding of participants to treatment allocation?³

(unlikely/likely/unclear)

Bias due to inadequate blinding of care providers to treatment allocation?³

(unlikely/likely/unclear)

Bias due to inadequate blinding of outcome assessors to treatment allocation?³

(unlikely/likely/unclear)

Bias due to selective outcome reporting on basis of the results?⁴

(unlikely/likely/unclear)

Bias due to loss to follow-up?⁵

(unlikely/likely/unclear)

Bias due to violation of

intention to treat analysis?⁶

(unlikely/likely/unclear)

Keeler 2019

Randomization in 1:1 ratio via a web-based system (variable block randomization) stratified by sex and age.

Unlikely

Reason: the web-based system was designed, set-up, and run by an independent unit of the University of Nothingham.

Likely

Reason: the nature of the intervention and control procedures prevent blinding. Authors describe the trial as open-label (see abstract).

Likely

Reason: the nature of the intervention and control procedures prevent blinding. Authors describe the trial as open-label (see abstract).

Likely

Reason: Authors describe the trial as open-label.

Likely

Reason: Quality of life outcomes were not compared between groups for the EQ5D5I (all sub scales and time points) and the FACT-An (all subscales for time points ‘at recruitment’ and ‘on the day of surgery’. For the time point ‘out patient follow-up’ the subscales physical well-being, and social / family well-being were not reported).

Unlikely (for reported outcomes between recruitment and surgery)

Reason: dropout is small and reason for dropout is similar in both groups.

Unlikely

Reason: authors There is no indication that participants switched between groups. There were protocol deviations in 5/55 (9.1%) patients who received oral iron

Unclear (for reported outcomes between surgery and outpatient visit)

Reason: drop-out was not similar between the groups (20% versus 11.5%). Although the reasons were similar and participants moving out of the area is considered random. It is unclear how this affected the reported outcome.

Keeler 2017

Randomization in 1:1 ratio via a web-based system (variable block randomization) stratified by sex and age.

Unlikely

Reason: the web-based system was designed, set-up, and run by an independent unit of the University of Nothingham.

Likely

Reason: the nature of the intervention and control procedures prevent blinding. Authors describe the trial as open-label.

Likely

Reason: the nature of the intervention and control procedures prevent blinding. Authors describe the trial as open-label.

Likely

Reason: Authors describe the trial as open-label.

Unlikely

Reason: outcomes in the protocol are reported in the manuscript, albeit minimally for some outcomes.

Unlikely

Reason: dropout is small and reason for dropout is similar in both groups.

Unlikely

Reason: authors state that ITT was followed. There is no indication that participants switched between groups. There were protocol deviations, however those participants were analysed.

Edwards 2009

Block randomization based on a computer-generated sequence provided by a support unit. Randomization was stratified by prerecruitment haemoglobin status.

Unlikely

Reason: allocation codes were concealed in opaque envelopes and stored in a locked room in a remote research unit.

Unlikely

Reason: The drug-giving set was concealed by an opaque sheath for the patients

Unlikely

Reason: Although the drug-administering investigator was not blinded, the clinicians involved in the perioperative care were blinded. Therefore it is presumable that the drug-administering investigator was not involved in the care giving, other than administering the drugs.

Unlikely

Reason: clinicians involved in the perioperative care were presumably the outcome assessors and were blinded, so was the chief investigator.

Unclear

Reason: the authors did not define ‘adverse perioperative events’ and may or may not have addressed this outcome in the results, depending of the definition.

Unlikely

Reason: Loss to follow-up is small and for similar reasons in both groups.

Unlikely

Reason: Authors did not state that they followed ITT, although there are no indications that there were protocol deviations (other than n=2 participants did not show up for a second infusion) or that participants switched between groups.

Lidder 2007

Participants were randomized by phone to a distant centre

Unclear

Reason: allocation concealment procedures were not described.

Likely

Reason: oral iron intake causes a dark stool, which is noticeable by patients

Unclear

Reason: It was stated that the clinical team was blinded, although no blinding procedures were described and it was unclear how participants had access to oral iron supplements (and whether this was through the clinical team)

Unclear

Reason: Although the clinical team may have been blinded, it is not stated that the data collector (even though not involved in the clinical care) was blinded for treatment

Likely

Reason: Length of hospital stay was not reported as an outcome, eventhough this was stated in the methods.

Unlikely

Reason: small drop-out and for similar reasons in both groups.

Unlikely

Reason: Authors did not state that they followed ITT, although there are no indications that there were protocol deviations or that participants switched between groups.