Restrictive versus liberal fluid resuscitation strategy, influence on blood loss and hemostatic parameters in mild obstetric hemorrhage: An open-label randomized controlled trial. (REFILL study).

BACKGROUND: Evidence for optimal hemostatic resuscitation in postpartum hemorrhage (PPH) is lacking. Liberal fluid administration may result in acidosis, hypothermia and coagulopathy.
OBJECTIVE: We hypothesize that in early PPH a restrictive fluid administration results in less progression to moderate PPH. STUDY
DESIGN: In four Dutch hospitals we recruited women of 18 years and over, and more than 24 weeks pregnant. Exclusion criteria were: anticoagulant therapy, known coagulation disorders, pre-eclampsia, antenatal diagnosis of abnormally adhesive placenta, and a contraindication for liberal fluid therapy. We blindly randomized participants at 500 mL and ongoing blood loss in the third stage of labor between restrictive fluid administration (clear fluids 0.75-1.0 times the volume of blood lost) and liberal fluid administration (clear fluids 1.5-2.0 times the volume of blood lost). The primary outcome was progression to more than 1000 mL blood loss. Analyses were according to the intention-to-treat principle.
RESULTS: From August 2014 till September 2019, 5190 women were informed of whom 1622 agreed to participate. A total of 252 women were randomized of which 130 were assigned to the restrictive group and 122 to the liberal group. In the restrictive management group 51 of the 130 patients (39.2%) progressed to more than 1000 mL blood loss versus 61 of the 119 patients (51.3%) in the liberal management group (difference, -12.0% [95%-CI -24.3% to 0.3%], p = 0.057). There was no difference in the need for blood transfusion, coagulation parameters, or in adverse events between the groups.
CONCLUSIONS: Although a restrictive fluid resuscitation in women with mild PPH could not been proven to be superior, it does not increase the need for blood transfusion, alter coagulation parameters, or cause a rise in adverse events. It can be considered as an alternative treatment option to liberal fluid resuscitation. TRIAL REGISTRATION: NTR3789.

RCT Entities:   Population Interventions Outcomes

Introduction

Postpartum hemorrhage (PPH) is one of the most common reasons for peripartum intensive care unit (ICU) admittance and the main cause of maternal death worldwide. In high resource countries an increase in incidence of PPH is observed [1-4]. In the Netherlands the incidence of PPH rose from 4.1% in 2000 to 6.4% in 2013 [5].

Evidence for optimal hemostatic resuscitation in PPH is currently lacking [6]. The Managing Obstetric Emergencies and Trauma course (MOET) and the Royal College of Obstetricians and Gynecologists (RCOG) instructions advise generous volume resuscitation to restore blood volume and oxygen carrying capacity: about twice the lost volume and up to 3.5 L of fast fluid infusion in patients with more than 1000 mL blood loss or clinical shock [7, 8]. Dutch guidelines recommend to commence volume resuscitation at profuse blood loss, not disclosing a minimum amount of blood loss. This guideline states that there is currently no evidence to administer more fluids than lost [9].

Resuscitation with crystalloids and colloids have their own (dis)advantages and risks. Transfusing large amounts of crystalloids before commencing with blood products may result in acidosis, hypothermia and coagulopathy; the lethal triad [10]. Additionally, the hydroxyethyl starch solutions may impair clot function when used excessively [11].

Restrictive or permissive hypotension has been advocated as an alternative for liberal fluid resuscitation in other areas than obstetric care, although the amount of randomized controlled trials is limited. A restrictive fluid administration policy in other fields has shown to decrease further blood loss and the amount of blood transfused [12-17]. Physiological hemodynamic and hemostatic changes in pregnancy make that these result cannot be readily adopted in the postpartum hemorrhage care. During pregnancy plasma volume and red blood cell count increases 40% and 30% respectively. Cardiac output is increased and the blood pressure decreased in the second trimester and increased again at term [18, 19]. There is a hypercoagulable state during pregnancy which is most pronounced in the third trimester [20, 21]. Two recent retrospective obstetric studies, showed that larger quantities of crystalloid volume administrated in the care of women with severe PPH was associated with a more severe deterioration of coagulation parameters. Fluid resuscitation with more than 4 L of crystalloid infusion was associated with more subsequent bleeding and adverse maternal outcomes (intensive care admittance, embolization, hysterectomy) [22, 23]. To date there are no publications on randomized trials on optimal fluid resuscitation in women with PPH.

The aim of this randomized controlled trial was to determine if a restrictive fluid administration policy in early and mild PPH (500 mL blood loss) leads to a decrease in progression to more than 1000ml blood loss compared to care as usual. We included both women with a caesarean and vaginal delivery. Although the risk of PPH is higher in women with a caesarean delivery, the question on optimal fluid administration is valid for both groups [24, 25]. We hypothesized a decrease in progression and therefore a decrease in adverse outcomes.

Materials and methods

Study design

REFILL was a randomized controlled multicenter trial performed from August 2014 until September 2019 in four hospitals in the Netherlands. This study was approved by the Medical Ethics Committee Maastricht University Hospital (NL4294206813). This trial is registered in the Netherlands Trial Register NTR3789 or NL3623 (date of registration, 11 January 2013). The study protocol was published in 2018 [12]. The study protocol is available online at: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5838856/

Participants

Participants were recruited in four Dutch hospitals, two university hospitals (Maastricht University Medical Center, Radboud University Medical Center) and 2 regional teaching hospitals (Zuyderland Medical Center and Jeroen Bosch Hospital). Maastricht University Medical Center (MUMC) was coordinating center. All four centers have in-house midwives and medical residents (junior and senior), and a supervising gynecologist on call.

All women attending the outpatient clinic or the labor ward, and not in active labor, were considered for eligibility. These women were informed about the study if they met the inclusion criteria. The inclusion criteria were: age 18 years and over, understanding of the Dutch language, pregnant and labor starting after 24+0 weeks, and mentally competent. Both vaginal and caesarean deliveries were included. Exclusion criteria were: prophylactic or therapeutic anticoagulant therapy (carbasalate calcium within the previous 10 days or low molecular weight heparins within previous 48 h), known congenital coagulation disorders, pre-eclampsia, antenatal diagnosis of placenta accreta spectrum (due to likelihood of reaching primary outcome regardless of management), and contraindication for liberal fluid therapy (e.g. cardiac causes, systemic causes (Marfan), renal causes, pulmonary failure). We obtained written informed consent from all participants.

Randomization and masking

Women who gave oral and written consent for the study were randomized if they reached 500 mL and ongoing blood loss postpartum. Enrolment was performed by the treating team of caregivers at that time through sealed opaque envelopes.

Treatment allocation was blinded by the use of sealed opaque envelopes including a trial number. Randomization was stratified per center, in blocks of four in an allocation of 1:1. Sequence was generated online (https://www.randomizer.org/) and the sealed opaque envelopes were created by an independent research nurse or medical student not involved in the randomization of the patient. Local investigators were blinded to block size and allocation. Randomization envelopes were distributed per center by Maastricht University Medical Center.

Procedures

The randomization envelopes were quickly and readily available on the labor ward or operating theatre. In case of 500 mL and ongoing blood loss at the third stage of labor an envelope was opened by the treating physician. The patient was either randomized to the restrictive fluid administration (intervention) group or to the liberal fluid administration (control) group. In the intervention group patients received fluids at 0.75–1.0 times the volume of blood loss. In the control group patients received 1.5–2.0 times the volume of blood loss. Blood loss was measured by weighing the absorption towels after childbirth. The first towel was disposed directly after childbirth and not measured as this also includes amniotic fluids. Blood loss during caesarean section was measured through suction and weighing operative gauzes after childbirth. The first 2000 mL of volume replacement consisted of NaCl 0.9% or Ringers lactate, or a combination of both on room temperature.

At 500 mL and ongoing blood loss allocation took place. Directly after randomization, moment T1 was initiated. If intravenous access was not yet present, intravenous access was established and blood samples for T1 were drawn. At T1 hemoglobin concentration, hematocrit, platelet count, activated partial thromboplastin time, prothrombin time, and fibrinogen concentrations were measured.

Hemodynamic parameters such as blood pressure and oxygen saturation were observed according to local protocol. Additional safety measures were taken in case of systolic blood pressure <90 mmHg, diastolic blood pressure <50 mmHg, a decrease of more than 20 mmHg in blood pressure, or a maternal heart rate of 125 beats per minute or more. In this case 500 mL additional volume was administered in 15 minutes in either group.

The second evaluation, T2, was 45–60 minutes after T1. At T2 a second set of blood samples as stated above were drawn. At T3, 12–18 hours post-partum a third set of blood sample was drawn for hemoglobin and hematocrit analysis only, if patients were still admitted to hospital.

If the patient reached 1500 mL blood loss the study protocol was exited and the patient was treated according to local massive obstetric hemorrhage protocol which also includes the blood transfusion policy. Blood samples as stated above were still drawn and patient data was analyzed on an intention-to-treat basis.

Third stage of labor was actively managed in all participants according to national protocol with the administration of 5 IE of oxytocin directly after childbirth and 10 IE oxytocin infused in 4 hours thereafter. Patients were kept warm. The underlying cause of the PPH was treated according to local and national guidelines (Dutch Society of Gynecology and Obstetrics, NVOG) [9].

All study parameters were collected from the patient chart and study files. The data were collected and stored anonymously in Maastricht University Medical Center in a restricted access file. A trial number was assigned to each patient at time of randomization.

Outcomes

The primary outcome was the frequency of progression to major PPH (defined as blood loss > 1000 mL).

Secondary outcomes were the differences in hemoglobin concentration (mmol L-1) 12–18 hours postpartum (including hemoglobin <5mmol L-1), differences in transfusion requirements (number of units of packed red blood cells, fresh frozen plasma, thrombocytes, and fibrinogen concentrate needed), differences in coagulopathies (platelets <50.10^9 L-1, fibrinogen concentration < 1g L-1 and activated partial thromboplastin time (APTT) and prothrombin time (PT) > 1.5x mean control).

Severe adverse outcomes (SAE), defined as intensive care admittance, the need of four or more units of packed cells, embolization, and hysterectomy, were registered and analyzed by a data safety monitoring board (DSMB).

Statistical analysis

Sample size was calculated with the assumption that, with standard care, around 30% of the women with 500 mL blood loss would progress to more than 1000 mL blood loss [12]. We calculated that, in order to be able to detect a 50% relative reduction, with a power of 0.80 and an alpha of 0.05, 118 patients would be needed in each study arm. To be able to compensate for incomplete data we aimed for 250 inclusions.

Comparative analysis was performed with either a Student’s t test in case of continuous data or the chi-square test in case of dichotomous outcomes. Multivariable linear regression analysis was employed to check whether results were sensitive to controlling for baseline characteristics, including center of inclusion. Analyses were done according to the intention-to-treat principle. Missing data were scarce and not imputed. Analyses were performed using IBM SPSS 24.0 and SAS version 9.4.

A data safety monitoring board was established. The DSMB was notified at each SAE, after the first 2x 25 inclusions, and every 2x 50 inclusions thereafter for which they performed an interim analysis on the primary outcome and SAEs. Throughout the study there was no need to stop the trial prematurely.

The corresponding author had full access to all the data in the study and had final responsibility for the decision to submit for publication.

Results

Between August 2014 and September 2019 5190 patients were assessed for eligibility of which 1622 patients gave informed consent to participate if they reached 500ml blood loss postpartum. A total of 252 patients were randomized, 130 were assigned a restrictive fluid administration strategy, and 122 a liberal fluid administration strategy (Fig 1). Maastricht University Medical Center recruited 74 participants, Radboud University Medical Center 37 participants, Zuyderland Medical Center 134 participants, and Jeroen Bosch Hospital recruited 7 patients.

Fig 1

Flow diagram.

Table 1 shows baseline characteristics, which were similar for the two groups. For all patients risk factors for PPH were evaluated (see S1 File for the risk factors collected). In two women no risk factors for PPH were present, the mean number of risk factors was 3 in both groups. In the liberal resuscitation strategy arm three patients discontinued treatment as they were diagnosed with pre-eclampsia. In the restrictive strategy arm no patient discontinued treatment. 130 patients in the restrictive fluid administration strategy, and 119 patients in the liberal fluid administration strategy were analyzed as intention-to-treat. The total mean crystalloid fluid administration after randomization in the restrictive arm was 1078 mL (SD1029 mL, median 800 (0–5200 mL)) and 1534 mL (SD 957 mL, median 1350 mL (0–4100 mL)) in the liberal arm (p = 0.000). All patients received crystalloids. Additional colloids were administered in 10/130 (7.7%, 48 mL SD 220) women in the restrictive arm versus 8/119 (6.7%, 47 mL SD 171) in the liberal arm (p = 0.957).

Table 1

Baseline characteristics.

	Restrictive (n = 130)	Liberal (n = 119)
Age (years)	31.9 (±3.5; 22–41)	31.6 (±4.5; 19–45)
BMI kg m-2	25.6 (±5.5; 17–45)	26·2 (±5.5; 17–47)
Gestational age (weeks)	39.2 (±1·8; 30.4–42.3)	39.3 (±1.6; 31.5–41·6)
Gravidity	2 (1–7)	2 (1–7)
Parity	0 (0–4)	1 (0–3)
Gestational age (days)	276 (±12.7; 214–297)	276 (±10.9; 222–293)
Risk factors HPP (amount)	3 (0–10)	3 (1–8)
History of:
Manual removal of placenta	10 (7.7)	4 (3.3)
Postpartum hemorrhage	19 (14.6)	15 (12.3)
Blood transfusion with postpartum hemorrhage	8 (6.2)	6 (4.9)
Onset of labor
Spontaneously	31 (23.8)	31 (25.4)
Induction	83 (63.8)	77 (63.6)
Caesarean, planned	16 (12.3)	13 (10.7)
Pain relief
Opioids	24 (18.5)	17 (13.9)
Epidural	57 (43.8)	63 (51.6)
No pain relief	50 (38.5)	28 (23.5)
Outcome
Delivery
Spontaneously	89 (68.5)	84 (68.9)
Ventouse	13 (10)	10 (8.2)
Caesarean	28 (21.6)	27 (22.1)
Caesarean, unplanned	12 (9.2)	19 (12.6)
Augmentation	80 (61.5)	84 (68.9)
Episiotomy	46 (35.4)	51 (41.8)
Vaginal rupture	52 (40)	48 (39.3)
Weight at birth (gram)	3497 (±631; 1470–5130)	3552 (±569; 1922–4740)
Macrosomia (>4kg)	28 (21.5)	26 (21.3)
T1 laboratory parameters
Hb (g l-1)	112.8 (±16.1; 61.2–143.4)	112.8 (±12.9; 67.7–138.6)
Ht (%)	0.33 (±0.05; 0.18–0.43)	0.33 (±0.04; 0.21–0.42)
Thrombocytes (·10^9 l-1)	202 (±61; 75–379)	198 (±48; 80–327)
APTT (sec)	26.9 (±3.6; 21–48)	26.6 (±3.7; 20.0–41.8)
Fibrinogen (g l-1)	4.2 (±0.9; 2–6.5)	4.2 (±0.7; 2.2–5.9)
PT(sec)	10.9 (±1.7; 9.1–16)	10.8 (±1.8; 9–20)

Data are n (%), mean (SD; range)

Gravity, parity and risk factors are presented as median (range)

BMI: body mass index, T1: time 1 at 500ml and start randomization, Hb: hemoglobin, Ht: hematocrit, APTT: activated partial thrombin time, PT: partial thrombin time

Table 2 shows both primary and secondary outcomes. In the restrictive policy 51 of the 130 patients (39.2%) progressed to more than 1000 mL blood loss versus 61 of the 119 patients (51.3%) in the liberal resuscitation policy arm (difference, -12.0% [95%-CI -24.3% to 0.3%], p = 0.057). Total blood loss did not differ significantly. Mean blood loss in the restrictive arm was 1182 mL (SD 761 mL) and in the liberal arm 1242mL (SD 621 mL), (p = 0.5).

Table 2

Primary and secondary outcomes.

	Restrictive policy (n = 130)	Liberal policy (n = 119)	p
Progression to more than 1000 mL blood loss	51 (39.2)	61 (51.3)	0.057
Total blood loss (mL)	1182 (761)	1242 (621)	0.5
Hemoglobin g l-1
T2	105.5 (15.3)	104.1 (15.6)	0.652
T3	92.7 (13.7)	99.9 (83.8)	0.849
Hemoglobin < 80.6g l-1 (n)
T2	6 (4.6)	7 (5.9)	0.404
T3	18 (13.8)	18 (15.1)	0.430
Transfusion (n)
Packed cells	14 (10.8)	11 (9.2)	0.689
Fresh Frozen Plasma	3 (2.3)	5 (4.2)	0.397
Thrombocytes	1 (0.8)	0 (0)	0.338
Fibrinogen	2 (1.5)	3 (2.5)	0.581
Coagulopathy T2
Platelets <50.10^9 l-1	0	0	n/a
APTT > 1.5 times reference range (n (mean seconds))	3 (42.2)	4 (43.0)	0.858
PT > 1.5 times reference range (n (mean seconds))	3 (14.6)	4 (14.5)	0.880
Fibrinogen < 1 gram	0	0	n/a
Adverse events (n)
ICU admittance	1	1	0.157
≥ 4 packed cells	1	2	0.223
Embolization	1	0	0.338
Hysterectomy	0	0	n/a

Data presented are n (%), mean (SD) unless otherwise stated

APTT: activated partial thrombin time PT: partial thrombin time ICU: intensive care unit T1: time 1 (at 500ml blood loss and start of randomization), T2: time 2 (45-60min after T1), T3: time 3 (12–18 hours after T1)

There was no difference in hemoglobin (Hb) levels at T2, and T3 in the restrictive arm and liberal arm respectively. Hb levels < 80.6g L-1 at T2 and T3 are comparable in the restrictive arm and liberal arm respectively.

The number of patients in need for blood products in both groups are comparable. Packed cells were primarily given to those exceeding the 1500 mL blood loss (n = 22/25), fresh frozen plasma and fibrinogen concentrate only in those exceeding 2000 mL of blood loss, and the thrombocytes were given in a case of 6000 mL blood loss. No significant difference in coagulopathy was observed between both policies: thrombocytes <50.10^9 L-1, APTT and PT more than 1.5 times the reference range, and fibrinogen less than 1 g. The use of intrauterine balloon tamponade (n = 3/130 in the restrictive policy versus n = 2/119 in the liberal policy, p = 0.73) and the use of B lynch stitch (n = 2/130 in the restrictive policy versus n = 1/119 in the liberal policy, p = 0.61) are comparable. There was no use of arterial ligation in either group. Adverse events defined as ICU admittance, administration of more than 4 packed cells, embolization therapy, and hysterectomy were not different between both groups.

Causes identified for PPH are presented in Table 3. Main cause of PPH in both groups is uterine atony; 52.3% in the restrictive arm and 63.9% in the liberal arm.

Table 3

Causes identified for PPH.

	Restrictive (n = 130)	Liberal (n = 119)
Uterine atony	68 (52.3)	76 (63.9)
Episiotomy	32 (24.6)	34 (28.6)
Retained placenta	27 (20.8)	26 (21.8)
Incomplete placenta	8 (6.2)	5 (4.2)
Cervical/vaginal trauma	17 (13.1)	11 (9.2)
Uterine rupture	3 (2.3)	0
Inversio uteri	0	0
Coagulopathy	0	0

Data are n (%)

Adjustment, by means of multiple regression, for the small differences in baseline characteristics (augmentation, episiotomy, analgesics) or controlling center of inclusion did not result in any meaningful changes in the effect estimates or in more precision.

Discussion

Principal findings

A restrictive fluid resuscitation in women with mild postpartum hemorrhage could not been proven to be superior (p = 0.057), even though the confidence interval around the effect estimation ranged from decreased progression risk by almost a quarter to near equality in the risk of progression (difference, -12.0% [95%-CI -24.3% to 0.3%]). A restrictive fluid resuscitation management in women with a moderate postpartum hemorrhage does not alter the need for blood transfusion, alter coagulation parameters, or cause a rise in adverse events. A more restrictive fluid resuscitation fluid management strategy could be a safe management choice in early and mild PPH.

Outside the obstetric field there is still no widely implemented consensus on fluid management in peri-operative and trauma care. As outlined in our trial protocol there is little and contradictive evidence for either liberal or restrictive fluid resuscitation regimens [12]. In addition to this outline Myles et al reports an increased risk for acute kidney injury in high risk patients during major abdominal surgery receiving a restrictive fluid management [26]. There was no difference in disability free survival in both groups. The randomized controlled trial of Myles et al. supports the dangers of hypoperfusion. However their study population is a high risk population undergoing major abdominal surgery which is not comparable to a relatively healthy obstetric population. Kwan et al reports, in their systematic review of six randomized controlled trials, no evidence for or against early or larger intravenous fluid administration strategies in uncontrolled hemorrhage in trauma patients [27]. No quantitative assessment could be provided due to diverse patient populations. They stress the necessity for further randomized controlled trials. We showed, in a systematic review, that a restrictive policy in elective surgery was favorable in comparison to a liberal fluid management policy for total complication rate, infection, and transfusion rate [13].

The lack of consensus on perioperative fluid management is reflected in the guidelines of the American Society of Anesthesiologists (ASA) and the European Society of Anaesthesiology (ESA). ASA has not updated their perioperative fluid management since 2008 [28]. In their perioperative blood management guidelines there is no mention of crystalloid or colloid use [29]. Their latest editorial note still marks the disagreement on the matter [30]. However they do agree that the optimal regimen is to replace the losses. The ESA state the lack of evidence to advice upon a perioperative fluid management. They do however advocate to avoid hypoperfusion, and advise a timely and aggressive stabilization of the cardiac preload taking a goal-directed approach [31]. However in their trauma guideline they recommend the use of a restrictive fluid management to achieve target blood pressure [32]. The ESA was also collaborator in the obstetric guideline of Network for the Advancement of Patient Blood Management, Hemostasis and Thrombosis (NATA) published in 2019. This multidisciplinary consensus statement recommend a restrictive fluid crystalloid administration of 1–2 mL crystalloids for every 1 mL blood lost [33]. This is a more liberal trend to how restrictive fluid management is generally advocated. Restrictive fluid resuscitation is based to replace the fluids lost with avoiding fluid overload [34].

The use of colloids can impair clot function by disturbing fibrin polymerization and by faster clot disintegration. Colloids may therefore increase blood loss [35-38]. The use of starches may increase the need for blood transfusion, increase the likelihood of acute kidney injury, and overall more side effects such as pruritis and rash [39-41].

Systematic reviews or randomized controlled trials evaluating fluid management protocols in the obstetric population are lacking. Our trend of a favorable outcome with a restrictive fluid management policy in the obstetric field is supported by the publications of Henriquez and Gillisen [22, 23]. In a retrospective cohort study Gillisen showed deterioration in coagulation parameters correlated with the amounts of crystalloid fluids infused. Levels of hemoglobin, hematocrit, platelet count, fibrinogen, and APTT were all negatively associated with the amount of crystalloid fluids infused. Henriquez performed a retrospective cohort study on women with a severe postpartum hemorrhage, finding that administration of more than 4 L of crystalloid fluids was independently associated with more maternal adverse outcomes (a composite of mortality and severe maternal morbidity defined as hysterectomy, embolization, or ICU admittance). Mean blood loss in both studies were 3.0 and 2.9 L respectively and exceeded our mean blood loss of 1.2 L.

Clinical implications

Our study is prospective and data were gained in a randomized controlled setting. The results of this randomized controlled trial are applicable to a wide obstetric population as most women with PPH do not exceed the 1500ml of blood loss. These data can be used to design and perform new randomized controlled trials in this fragile population and acute setting.

Strengths and limitations

To our knowledge this is the first randomized controlled trial on fluid resuscitation strategies within the obstetric field. We reached our calculated sample size in both resuscitation arms and had no loss to follow-up for the primary outcome. Baseline characteristics were well balanced and adjustment for small differences in the baseline gave similar results.

The randomization envelopes were available at a central point at the labor ward in each location. In case of a cesarian section, an envelope was taken to the operating theatre in case the patient would reach 500 mL blood loss. The operating theatres are not at the labor ward but in a different section of the hospital at all participating locations, making it impractical to pick up an envelope from the labor ward once the patient reached 500 mL blood loss. Unfortunately some of these unopened envelopes were disposed of during clean up instead of returned to their original central point, causing a slight imbalance of 130:122 in treatment assignments. This numerical imbalance does not influence validity. Even though we reached our aimed pre-calculated sample size, and the point estimate of the difference in risk of progression to more than 1000 mL blood loss pointed to a clinically relevant effect (absolute difference, 12.1%; number needed to treat, 9), the difference did not reach statistical significance (p = 0.057). With inclusion of a larger number of participants the power of study would have been higher. Arguably, in retrospect, for our sample size calculation we chose a minimally detectable relative risk that was too conservative (RR 0.5) with the consequence that smaller, but relevant, differences such as the one found would stay statistically non-significant. Pooling of our results (or data) with any similar future studies could yield more precision and enable smaller differences to be more easily detectable.

As this study was the first randomized controlled trial with a solely obstetric patient population, strict precautionary safety measures were in place. One of the main safety measure was abdication of the enrolled resuscitation arm when reaching 1500 mL blood loss. Therefore our results can only be applied to women with a PPH less than 1500 mL blood loss, which is the majority of all women experiencing PPH. Of all women in labor, 1.4–3.9% progress to more than 1500 mL blood loss [3, 42]. Another safety measure was the choice to commence resuscitation at 500 mL of blood loss. Signs of clinical shock can present as early as 750 mL of blood loss, defined as class II hemorrhagic shock [32]. However these limitations do reduce the ability to compare the more serious adverse outcomes such as transfusion need, embolization, IC admittance and hysterectomy between these groups.

Our progression to more than 1000 mL blood loss was 39.2% in the restrictive management group and 51.3% in the liberal management group which is higher than the general percentage of women with a PPH of more than 1000 mL: 3–5.5% [33, 42, 43]. We contribute this to a selected population, we only selected women who had more than 500ml and ongoing blood loss whereas the general percentage applies to the whole population of women giving birth. This is reflected in the percentage in the population who gave informed consent: 6.3% (102/1622) had more than 1000 mL blood loss. This is slightly higher than the reported incidences. Although important, optimal management and the safety of restrictive management or even permissive hypotension in massive PPH, concerning only a small minority of women, is not the scope of the current study. We chose to study the more common and therefore more relevant effect on the effects of fluid resuscitation in mild PPH. Adequately managing mild PPH can improve care for a large group of women and may prevent progression.

Conclusions

Although a restrictive fluid resuscitation in women with mild PPH could not been proven to be superior it does not increase the need for blood transfusion, alter coagulation parameters, or cause an increase in adverse events and therefore can be considered as an alternative treatment option. This study does not allow comments on safety on restrictive management in cases with massive PPH. More randomized controlled trials on fluid resuscitation should be conducted in patients with PPH to establish a evidence based recommendations on this matter.

CONSORT 2010 checklist of information to include when reporting a randomised trial*.

(DOC)

Click here for additional data file.

Risk factors postpartum hemorrhage.

BMI: body mass index PPH: postpartum hemorrhage.

(DOCX)

Click here for additional data file.

Dataset.

(SAV)

Click here for additional data file.

16 Mar 2021

PONE-D-21-04160

Restrictive versus liberal fluid resuscitation strategy, influence on blood loss and hemostatic parameters in mild obstetric hemorrhage: an open-label randomized controlled trial. (REFILL study)

PLOS ONE

Dear Dr. Schol,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

This is the first RCT regarding restrictive versus liberal fluid resuscitation strategy in the management of PPH. The article is clear and very interesting, even if results were found to be negative. Some methodological comments (see reviewer comments) decrease the interest and some modifications should be done in the manuscript, especially in methods. The flow diagram should be checked because there are some errors in the number of patient (see reviewer comments). Some details should also be added in discussion about limitations of the study. To conclude, the authors reported an interesting, but negative RCT, and a lot of work should be done before resubmission.

Please submit your revised manuscript by Apr 23 2021 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols

We look forward to receiving your revised manuscript.

Kind regards,

Guillaume Ducarme, MD, MSc, PhD

Academic Editor

PLOS ONE

Journal Requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2. Please amend your list of authors on the manuscript to ensure that each author is linked to an affiliation. Authors’ affiliations should reflect the institution where the work was done (if authors moved subsequently, you can also list the new affiliation stating “current affiliation:….” as necessary).

3. Your ethics statement should only appear in the Methods section of your manuscript. If your ethics statement is written in any section besides the Methods, please move it to the Methods section and delete it from any other section. Please ensure that your ethics statement is included in your manuscript, as the ethics statement entered into the online submission form will not be published alongside your manuscript.

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

Reviewer #2: Yes

Reviewer #3: Yes

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: No

Reviewer #3: Yes

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: Yes

Reviewer #3: Yes

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

Reviewer #3: Yes

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: Pim Schol et al. presented here a very interesting randomized clinical trial regarding restrictive versus liberal fluid resuscitation strategy in the management of PPH.

This is the first RCT in obstetrics field ++

In a general way, this article is clear, and very interesting, even if results were found to be negative.

Please find some comments that should be taken into account:

Specific comments:

Methods section:

L142: Were there some women treated with 1 to 1.5 times the volume of blood loss?

L.148: please explain more precisely T1, T2 etc… it is unclear.

Please state if women were administered heating blanket in the management of PPH.

L161: how many women exited from the study?

L172: I think that “moderate” should be changed to “severe” ?

L177: what was the policy regarding transfusion of blood product?

L182: Is there a policy regarding the use of intrauterine tamponade?

Results section:

L219: May the authors provide administered volume of colloids?

Table 1: how can the authors explain the rate of episiotomy?

I do not understand the percentage of episiotomy 51 (41·8 %) ? Explain.

Please explain the rate of macrosomia that appears high…?

Table 1: same remark as above: explain the overall of 23 pregnancies…

May the authors indicate BMI, parity in the table?

Discussion section:

L268-270: please moderate this sentence since the benefits do not appear in this study.

L290-295: in which situation? PPH? Surgery with high-risk of bleeding?

L298-300: it is strange since their “restrictive policy” corresponds to twice the volume of blood loss whereas in your study, the restrictive policy corresponded to a 0.75-1 times the volume of blood loss. Therefore, it is difficult to compare. The authors should discuss the definition of “restrictive policy”

L305: are there data regarding colloids use on coagulation parameters (vs cristaloids)?

L307: please provide example.

Please provide few sentences about the use of intrauterine ballon that is not documented in the manuscript.

Reviewer #2: This is an interesting, but ultimately negative clinical trial. I believe the extreme sample size assumption that the investigators wished to see a 50% reduction was overly conservative, and resulted in an underpowered study. There was a trending difference between the groups in the primary outcome, but the sample size was not large enough to get a significant p-value. This should be noted in the "weaknesses" section as one possible reason for the negative primary outcome.

I have numerous comments on how the manuscript and analysis could be improved:

1. The abstract Methods section should be less about inclusion and exclusion and more about the study design.

2. There are no statistical methods listed in your "Statistical Methods" section. You only discuss sample size and what the DMC was monitoring. Without specifying how you are analyzing, sample size computations and monitoring are irrelevant.

3. Since you did not say how you analyzed the study, it appears that you did not conduct a stratified test, which every statistician would require following a stratified (by center) analysis. Please correct this, or specify that you did this already.

4. The DMC does not "abandon the trial prematurely", they determine if there is sufficient evidence to stop a trial early for efficacy, futility, and/or safety.

5. Throughout the paper you refer to the groups with the phrase "there is no difference". As you know, determination of a statistical difference requires more than just observation of numbers. It is more appropriate to discuss that the data appear to be comparable with respect to the given metric.

6. The discussion section should (as noted above) indicate whether the assumptions in the sample size and power computations were realized in the study (e.g., effect size, LTF assumptions, etc.)

7. I have no idea how you would up with a 130:122 imbalance in treatment assignments given that you were using permuted blocks with a fixed size of 4. The maximum imbalance possible is 2 due to an unfilled block, or where there non-compensating incomplete blocks in multiple centers.

8. You do not discuss the clinical centers, where they were, what their staffing was, how many patients they each recruited, and how study procedures were standardized.

9. Fixed block sizes are frowned upon by statisticians--I presume the investigators were blinded to the block size to mitigate selection bias, and that should be stated if so.

Reviewer #3: Dear Author,

I reported my comments on the manuscript titled “Restrictive versus liberal fluid resuscitation strategy, influence on blood loss and hemostatic parameters in mild obstetric hemorrhage: an open-label randomized controlled trial. (REFILL study)”.

I think that the topic is interesting. References list appears up to date and appropriate. The manuscript is clear and easy to understand also for non-specialized reader.

Without doubts, the study have many limits, some of which punctually explained in the manuscript. In my opinion a great limit is the selection of population, there are many confounding factor, for example the choose to recruit women both after vaginal delivery and cesarean section or the recruitment of women with risk factors for PPH. Probably the best way is recruit women without risk factors after a spontaneous vaginal delivery. Anyway, this limit is partially overcame by the fact that the sampling method is the same in both groups and by the decision to analyze data according to the intention-to-treat principle.

Before the acceptance I suggest some minor revisions.

You (at line 143-146) explain that blood loss was measured by weighing the absorption towel after childbirth, but the first towel was disposed directly after childbirth and not measured as this also includes amniotic fluid. I suggest to explain how do you have solved the problem of amniotic fluid interference in case of cesarean section, in which the blood loss was measured through suction and weighing operative gauze.

In the methods, at line 164, I suggest to explain the actions performed in the active management of third stage of labor, for example the type and dosage and route of administration of uterotonic agent.

I suggest to check the flow diagram in the figure 1, because in my opinion could be some errors in the number of patient. Authors say that they have assessed for eligibility 5190 women. After that 1622 women signed informed consent and 3568 women were excluded. In the reasons of exclusion we have 1370 women that not met inclusion criteria and 1946 that declined to participate. In this way we have 252 lacking women (1370+1946=3316, 252 women less of 3568 women excluded). I think that this 252 women are not the women randomized, because the 252 women randomized were in the group of women that signed the consent. So I suggest to check the data, because if the 252 women randomized are added to the women that signed the consent we have not 1622 women but 1874 women. In this way the authors have to check also the data at line 339-340, because the rate 102/1874 is 5,4% in line with the litterature data.

In the discussion at line 270, remove "low risk PPH", because the women in the study are not at low risk PPH. I think that you refer about the early and mild PPH (500 mL blood loss).

After these minor revisions, I suggest to accept this manuscript because in my personal opinion the paper is interesting for at least 3 reasons: the originality (first study on this topic), the study design (prospective, blinded, randomized controlled multicenter clinical trial) and for the results (which certainly deserve to be then confirmed on a larger scale).

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

Reviewer #3: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

21 Apr 2021

PLOS ONE

Senior Editorial Operations Manager R. Yterdal

1265 Battery Street

Floor 2

San Francisco, CA 94111

United States of America

Maastricht, 18th April 2021

Dear dr. Ducarme,

Thank you and the reviewers for the investment in our manuscript. We reviewed the comments and revised our manuscript “Restrictive versus liberal fluid resuscitation strategy, influence on blood loss and hemostatic parameters in mild obstetric hemorrhage: an open-label randomized controlled trial. (REFILL study)” accordingly.

As requested we supply a marked-up copy of our manuscript as well as an unmarked version. Below you will find the points raised and our response to it per point.

We hope you will accept our submission for “PLOS ONE”.

Kind regards,

Pim Schol, M.D.

Please submit your revised manuscript by Apr 23 2021 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

• A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

• A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

• An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols

We look forward to receiving your revised manuscript.

Kind regards,

Guillaume Ducarme, MD, MSc, PhD

Academic Editor

PLOS ONE

Journal Requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

Checked

2. Please amend your list of authors on the manuscript to ensure that each author is linked to an affiliation. Authors’ affiliations should reflect the institution where the work was done (if authors moved subsequently, you can also list the new affiliation stating “current affiliation:….” as necessary).

Adjusted

3. Your ethics statement should only appear in the Methods section of your manuscript. If your ethics statement is written in any section besides the Methods, please move it to the Methods section and delete it from any other section. Please ensure that your ethics statement is included in your manuscript, as the ethics statement entered into the online submission form will not be published alongside your manuscript.

Adjusted

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

Reviewer #2: Yes

Reviewer #3: Yes

________________________________________

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: No

Reviewer #3: Yes

________________________________________

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: Yes

Reviewer #3: Yes

________________________________________

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

Reviewer #3: Yes

________________________________________

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: Pim Schol et al. presented here a very interesting randomized clinical trial regarding restrictive versus liberal fluid resuscitation strategy in the management of PPH.

This is the first RCT in obstetrics field ++

In a general way, this article is clear, and very interesting, even if results were found to be negative.

Please find some comments that should be taken into account:

Specific comments:

Methods section:

L142: Were there some women treated with 1 to 1.5 times the volume of blood loss?

Women were randomized between 0.75-1.0 times and 1.5-2.0 times, there was no separate group of 1-1.5 times. We chose these groups so that the treatment would differ enough to establish difference. Women were analyzed according to intention to treat. However we did analyze ‘as treated’ which yielded no different results.

L.148: please explain more precisely T1, T2 etc… it is unclear.

Adjusted.

Please state if women were administered heating blanket in the management of PPH.

It is part of national protocol to keep women warm at >1000 mL blood loss: https://www.nvog.nl/wp-content/uploads/2018/02/bijlage-checklist-bij-uitgangsvragen-rl-HPP-2015lhee.pdf .This can be also be achieved with prewarmed blankets as well, not only heating blankets. We referenced to this national protocol at line 171 and added the statements “Patients were kept warm”.

L161: how many women exited from the study?

None exited while randomized. In the analysis 3 women with pre-eclampsia during labor were omitted from analysis as this was an exclusion criteria. It is stated in the “Results” section, line 223

L172: I think that “moderate” should be changed to “severe” ?

Depending on the defining source. RCOG defines “minor” and “major”, with major subdivided into “moderate” (1000-2000mL) and “severe” (>2000mL). WHO defines “severe” as >1000mL. We adjusted it to “major”, this will suit both definitions.

L177: what was the policy regarding transfusion of blood product?

According to national guideline (https://www.nvog.nl/wp-content/uploads/2018/02/bijlage-checklist-bij-uitgangsvragen-rl-HPP-2015lhee.pdf), ordering blood products at >1000mL and ongoing blood loss. Administering at 2000mL or when 2L of crystalloids are given. If patients reached 1500 mL of blood loss the randomized arm was exited and patients were treated according to local massive hemorrhage protocol, as stated in line 163. Two centers have thromboelastometry available which makes a patient directed blood transfusion possible beyond the standardized protocol.

Clarified at line 164

L182: Is there a policy regarding the use of intrauterine tamponade?

Mentioned in national guideline (https://www.nvog.nl/wp-content/uploads/2018/02/Hemorrhagia-postpartum-HPP-3.0-14-11-2013.pdf) . To be used at physicians discretion.

Results section:

L219: May the authors provide administered volume of colloids?

Added, as the mean was very low with a larger standard deviation we initially chose to only mention the amount of patients that received the colloids. We now added the mean and standard deviation to the text.

Table 1: how can the authors explain the rate of episiotomy?

I do not understand the percentage of episiotomy 51 (41·8 %) ? Explain.

Please explain the rate of macrosomia that appears high…?

This study was conducted in the hospital which results in a selected population. In the Netherlands the ultimately low risk pregnancies are guided by the midwives at home or in a birth center without medical supervision. Women were informed at our outpatient clinic and gave informed consent before start of the induction or spontaneous labor with a medical indication (macrosomia, intra-uterine growth retardation, cesarian section indication etc). Resulting in the higher need for intervention. Also an episiotomy and macrosomia are a risk factor for postpartum hemorrhage. Resulting in women with an episiotomy and/or macrosomia to be more easily randomized.

Table 1: same remark as above: explain the overall of 23 pregnancies…

Overall 23 pregnancies? We do not understand this remark. The average number of pregnancies is 2.

May the authors indicate BMI, parity in the table?

BMI and parity are already stated in the table at the top section.

Discussion section:

L268-270: please moderate this sentence since the benefits do not appear in this study.

Moderated.

L290-295: in which situation? PPH? Surgery with high-risk of bleeding?

Perioperative fluid management outside the obstetric field. Also indicated at start of the section. Clarified at these lines.

L298-300: it is strange since their “restrictive policy” corresponds to twice the volume of blood loss whereas in your study, the restrictive policy corresponded to a 0.75-1 times the volume of blood loss. Therefore, it is difficult to compare. The authors should discuss the definition of “restrictive policy”

The overall basis of restrictive fluid policy is to replace the volume lost, to avoid fluid overload and hypoperfusion. It is indeed strange for the NATA to therefore recommend a more liberal approach to the restrictive policy. This underlines the point made, that there is no real consensus on fluid resuscitation policy. We added a reference and definition to lines 319-320

L305: are there data regarding colloids use on coagulation parameters (vs cristaloids)?

There are more than enough data on colloids disturbing the coagulation parameters, that is the reason crystalloids are preferred over colloids. We feel that such a detailed outline is beyond the scope of this article. We did however insert a short comment on the effect of colloids on coagulation. Hydroxyethylstarches (HES) are also prone to give more (acute) kidney injury and adverse events.

L307: please provide example.

This is stated at this line prior to revisions: “In a retrospective cohort study Gillisen showed deterioration in coagulation parameters correlated with the amounts of crystalloid fluids infused. Levels of hemoglobin, hematocrit, platelet count, fibrinogen, and APTT were all negatively associated with the amount of crystalloid fluids infused”

We feel that elaborating with exact numbers on a research that is not our own, does not contribute to the readability of our discussion. We feel that we referenced correctly and when exact numbers are wanted, one should read the study itself to interpret the numbers correctly as this study provides a lot of data.

Please provide few sentences about the use of intrauterine ballon that is not documented in the manuscript.

The use of intrauterine balloon tamponade is not one of the primary or secondary outcome measures. We added a sentence about this intervention on request. We also added the use of B lynch stich and arterial ligation to complete the interventions used.

Reviewer #2: This is an interesting, but ultimately negative clinical trial. I believe the extreme sample size assumption that the investigators wished to see a 50% reduction was overly conservative, and resulted in an underpowered study. There was a trending difference between the groups in the primary outcome, but the sample size was not large enough to get a significant p-value. This should be noted in the "weaknesses" section as one possible reason for the negative primary outcome.

Added:

“Even though we reached our aimed pre-calculated sample size, and the point estimate of the difference in risk of progression to more than 1000 mL blood loss pointed to a clinically relevant effect (absolute difference, 12.1%; number needed to treat, 9), the difference did not reach statistical significance (p=0.057). With inclusion of a larger number of participants the power of study would have been higher. Arguably, in retrospect, for our sample size calculation we chose a minimally detectable relative risk that was too conservative (RR 0.5) with the consequence that smaller, but relevant, differences such as the one found would stay statistically non-significant. Pooling of our results (or data) with any similar future studies could yield more precision and enable smaller differences to be more easily detectable.”

I have numerous comments on how the manuscript and analysis could be improved:

1. The abstract Methods section should be less about inclusion and exclusion and more about the study design.

We feel that a clear description of a performed study is crucial for further research. The inclusion and exclusion criteria are provided in lines 115 to 122. These are mentioned and not explained in this paragraph or thereafter in the methods section. The rest of the methods section is focused on the execution of the trial to allow for reproduction by others.

2. There are no statistical methods listed in your "Statistical Methods" section. You only discuss sample size and what the DMC was monitoring. Without specifying how you are analyzing, sample size computations and monitoring are irrelevant.

We already mentioned the method for comparative analysis in our original manuscript (Lines 194-197)

“Comparative analysis was performed with either a Student’s t test in case of continuous data or the chi-square test for dichotomous outcomes. Analyses were done according to the intention-to-treat principle. Missing data were not imputed. All analyses were performed using IBM SPSS 24.0 software.”

We added information on the additional sensitivity analysis that we carried out by means of multiple linear regression: ”Multivariable linear regression analysis was employed to check whether results were sensitive to controlling for baseline characteristics, including center of inclusion.”

3. Since you did not say how you analyzed the study, it appears that you did not conduct a stratified test, which every statistician would require following a stratified (by center) analysis. Please correct this, or specify that you did this already.

We now mentioned in the methods section (statistical analysis): ”Multivariable linear regression analysis was employed to check whether results were sensitive to controlling for baseline characteristics, including center of inclusion.”

Results (in terms of betas and precision) were robust to such adjustment. Therefore the last line of the Results section says:

“Adjustment, by means of multiple regression, for the small differences in baseline characteristics (augmentation, episiotomy, analgesics) or controlling center of inclusion did not result in any meaningful changes in the effect estimates or in more precision.”

4. The DMC does not "abandon the trial prematurely", they determine if there is sufficient evidence to stop a trial early for efficacy, futility, and/or safety.

Adjusted

5. Throughout the paper you refer to the groups with the phrase "there is no difference". As you know, determination of a statistical difference requires more than just observation of numbers. It is more appropriate to discuss that the data appear to be comparable with respect to the given metric.

Adjusted throughout

6. The discussion section should (as noted above) indicate whether the assumptions in the sample size and power computations were realized in the study (e.g., effect size, LTF assumptions, etc.)

In results: “Correction for the small differences in baseline characteristics (augmentation, episiotomy, analgesics) did not result in any meaningful changes in the effect estimates or more precision.”

In discussion: “We reached our calculated sample size in both resuscitation arms and had little loss to follow-up. Baseline characteristics were well balanced and adjustment for small differences in the baseline gave similar results.”

Loss to follow up can only be applied for T3 results as women may already have left the hospital and this may also be viewed as missing data. None of the women were lost to follow up for the primary outcome.

7. I have no idea how you would up with a 130:122 imbalance in treatment assignments given that you were using permuted blocks with a fixed size of 4. The maximum imbalance possible is 2 due to an unfilled block, or where there non-compensating incomplete blocks in multiple centers.

The envelopes were available at a central point at the labor ward in all locations. However these were taken to the operating theatre in case of a cesarian section. The operating theaters are not at the labor ward but in a different section of the hospital (at all participating locations) making it impossible and time consuming to get randomization envelope in case there was 500 mL and ongoing blood loss. The missing envelopes causing the imbalance as mentioned by this reviewer, were not opened but were disposed of during clean up instead of returned to their original central point.

8. You do not discuss the clinical centers, where they were, what their staffing was, how many patients they each recruited, and how study procedures were standardized.

Added, see lines 108-112 and 215-218

9. Fixed block sizes are frowned upon by statisticians--I presume the investigators were blinded to the block size to mitigate selection bias, and that should be stated if so.

Stated already in original manuscript:

“Sequence was generated online (https://www.randomizer.org/) and the sealed opaque envelopes were created by an independent research nurse or medical student not involved in the randomization of the patient.”

Clarified with “Local investigators were blinded to block size and allocation.”

Reviewer #3: Dear Author,

I reported my comments on the manuscript titled “Restrictive versus liberal fluid resuscitation strategy, influence on blood loss and hemostatic parameters in mild obstetric hemorrhage: an open-label randomized controlled trial. (REFILL study)”.

I think that the topic is interesting. References list appears up to date and appropriate. The manuscript is clear and easy to understand also for non-specialized reader.

Without doubts, the study have many limits, some of which punctually explained in the manuscript. In my opinion a great limit is the selection of population, there are many confounding factor, for example the choose to recruit women both after vaginal delivery and cesarean section or the recruitment of women with risk factors for PPH. Probably the best way is recruit women without risk factors after a spontaneous vaginal delivery. Anyway, this limit is partially overcame by the fact that the sampling method is the same in both groups and by the decision to analyze data according to the intention-to-treat principle. Before the acceptance I suggest some minor revisions.

You (at line 143-146) explain that blood loss was measured by weighing the absorption towel after childbirth, but the first towel was disposed directly after childbirth and not measured as this also includes amniotic fluid. I suggest to explain how do you have solved the problem of amniotic fluid interference in case of cesarean section, in which the blood loss was measured through suction and weighing operative gauze.

In the operating theatre the nurse notes the amount of amniotic fluid suctioned during child birth. After child birth blood loss is counted from there onwards. Adjusted at line 147

In the methods, at line 164, I suggest to explain the actions performed in the active management of third stage of labor, for example the type and dosage and route of administration of uterotonic agent.

See line 167-171 for adjustment

I suggest to check the flow diagram in the figure 1, because in my opinion could be some errors in the number of patient. Authors say that they have assessed for eligibility 5190 women. After that 1622 women signed informed consent and 3568 women were excluded. In the reasons of exclusion we have 1370 women that not met inclusion criteria and 1946 that declined to participate. In this way we have 252 lacking women (1370+1946=3316, 252 women less of 3568 women excluded). I think that this 252 women are not the women randomized, because the 252 women randomized were in the group of women that signed the consent. So I suggest to check the data, because if the 252 women randomized are added to the women that signed the consent we have not 1622 women but 1874 women. In this way the authors have to check also the data at line 339-340, because the rate 102/1874 is 5,4% in line with the litterature data.

Thank you for noticing, We reran the numbers and corrected those were needed. The 1622 is the correct number therefore 339-340 didn’t need adjustment.

In the discussion at line 270, remove "low risk PPH", because the women in the study are not at low risk PPH. I think that you refer about the early and mild PPH (500 mL blood loss).

Adjusted

After these minor revisions, I suggest to accept this manuscript because in my personal opinion the paper is interesting for at least 3 reasons: the originality (first study on this topic), the study design (prospective, blinded, randomized controlled multicenter clinical trial) and for the results (which certainly deserve to be then confirmed on a larger scale).

________________________________________

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

Reviewer #3: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

Submitted filename: Respons to reviewers.docx

Click here for additional data file.

12 May 2021

PONE-D-21-04160R1

Restrictive versus liberal fluid resuscitation strategy, influence on blood loss and hemostatic parameters in mild obstetric hemorrhage: an open-label randomized controlled trial. (REFILL study)

PLOS ONE

Dear Dr. Schol,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

==============================

ACADEMIC EDITOR: The paper has undergone extensive revision according to the reviewers' comments. Some minor modifications should be done (see comments Reviewer 2) about the randomization procedure, some sentences should be added in the weaknesses of the study.

==============================

Please submit your revised manuscript by Jun 26 2021 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols. Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols.

We look forward to receiving your revised manuscript.

Kind regards,

Guillaume Ducarme, MD, MSc, PhD

Academic Editor

PLOS ONE

Journal Requirements:

Please review your reference list to ensure that it is complete and correct. If you have cited papers that have been retracted, please include the rationale for doing so in the manuscript text, or remove these references and replace them with relevant current references. Any changes to the reference list should be mentioned in the rebuttal letter that accompanies your revised manuscript. If you need to cite a retracted article, indicate the article’s retracted status in the References list and also include a citation and full reference for the retraction notice.

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #1: All comments have been addressed

Reviewer #2: (No Response)

Reviewer #3: All comments have been addressed

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

Reviewer #2: (No Response)

Reviewer #3: Yes

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: (No Response)

Reviewer #3: Yes

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: (No Response)

Reviewer #3: Yes

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: (No Response)

Reviewer #3: Yes

**********

6. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: (No Response)

Reviewer #2: Since the randomization procedure was compromised due to unopened envelopes, the explanation given in response to my comments should be added to the "Weaknesses of the study" section in the Conclusions. I would also add when stating the unbalanced sample size that "reasons for the imbalance are discussed in the conclusions", or discuss the issue earlier when you talk about the envelope procedure.

Reviewer #3: Congratulations for the manuscript! Interesting Topic, well conducted and written and understandable and of general interest paper.

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

Reviewer #3: Yes: Alessandro Svelato

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

31 May 2021

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #1: All comments have been addressed

Reviewer #2: (No Response)

Reviewer #3: All comments have been addressed

________________________________________

2. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

Reviewer #2: (No Response)

Reviewer #3: Yes

________________________________________

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: (No Response)

Reviewer #3: Yes

________________________________________

4. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: (No Response)

Reviewer #3: Yes

________________________________________

5. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: (No Response)

Reviewer #3: Yes

________________________________________

6. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: (No Response)

Reviewer #2: Since the randomization procedure was compromised due to unopened envelopes, the explanation given in response to my comments should be added to the "Weaknesses of the study" section in the Conclusions. I would also add when stating the unbalanced sample size that "reasons for the imbalance are discussed in the conclusions", or discuss the issue earlier when you talk about the envelope procedure.

We added this to our strength and limitations section of our manuscript, and 347-354.

Reviewer #3: Congratulations for the manuscript! Interesting Topic, well conducted and written and understandable and of general interest paper.

________________________________________

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

Reviewer #3: Yes: Alessandro Svelato

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

Submitted filename: Response to reviewers.docx

Click here for additional data file.

14 Jun 2021

Restrictive versus liberal fluid resuscitation strategy, influence on blood loss and hemostatic parameters in mild obstetric hemorrhage: an open-label randomized controlled trial. (REFILL study)

PONE-D-21-04160R2

Dear Dr. Schol,

We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/, click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

Guillaume Ducarme, MD, MSc, PhD

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #3: All comments have been addressed

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #3: (No Response)

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #3: (No Response)

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #3: (No Response)

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #3: (No Response)

**********

6. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #3: (No Response)

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #3: Yes: Alessandro Svelato

18 Jun 2021

PONE-D-21-04160R2

Restrictive versus liberal fluid resuscitation strategy, influence on blood loss and hemostatic parameters in mild obstetric hemorrhage: an open-label randomized controlled trial. (REFILL study)

Dear Dr. Schol:

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

If we can help with anything else, please email us at plosone@plos.org.

Thank you for submitting your work to PLOS ONE and supporting open access.

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Dr. Guillaume Ducarme

Academic Editor

PLOS ONE