ORIGINAL ARTICLE Neonatal red blood cell transfusions: searching for better guidelines Kavita Kasat1, Karen D. Hendricks-Muñoz2, Pradeep V. Mally2 1

Department of Pediatrics, Division of Neonatology, Cohen Children’s Medical Center of New York, New Hyde Park; 2Department of Pediatrics, Division of Neonatology, New York University Langone Medical Center, New York, USA

Background. Packed red blood cell (RBC) transfusions are often administered to patients in the neonatal intensive care unit. The purpose of this study was to determine whether current blood transfusion guidelines are as useful as care givers' perception in identifying patients in need of a packed RBC transfusion. Design and Methods. Health care providers were asked to complete a pre- and posttransfusion survey on neonates receiving a packed RBC transfusion. These patients were divided into three groups based on reasons for transfusion: (i) guidelines; (ii) care-givers' perceptions of need for packed RBC transfusion; or (iii) both. These three groups were further subdivided into two cohorts according to whether they had a clinical improvement or not. Demographic data and clinical variables were compared between the groups. Results. Seventy-eight care-givers were surveyed. Eighteen patients (23%) were transfused based on guidelines, 36 (46%) based on care givers' perception and 24 (31%) based on both. Neonates transfused based on guidelines alone were more likely to have received the transfusion in the first week of life, had a higher pre-transfusion haematocrit, were less symptomatic and had a higher trend to require mechanical ventilation. Neonates transfused based on caregivers' perception were more likely to be on non-invasive ventilatory support and were more symptomatic. Neonates who improved after a transfusion had a lower pre-transfusion haematocrit (p=0.02), were more symptomatic (p=0.01) and were more likely to be on non-invasive ventilatory support (p=0.002) when compared to the group without a clinical improvement. The group without improvement had an increase in oxygen requirement (+2.8±6.4) after the transfusion (p=0.0004). Tachycardia was the most sensitive predictor of a benefit from packed RBC transfusion [OR 6.48: p=0.005]. Discussion. Guidelines on when to transfuse stable growing neonates with packed RBC should be re-evaluated to include more care giver judgement and perhaps be more restrictive for critically ill neonates. Keywords: neonate, blood transfusion, guidelines, anaemia, oxygen delivery.

Introduction Packed red blood cell (RBC) transfusion is one of the most widely utilised medical interventions practised by clinicians, although there is no single clinical parameter to determine the best time to transfuse. The criteria for packed RBC transfusion in

neonates is based on expert opinion and recent evidence-based medicine. In order to optimise oxygenation, the premature or critically ill neonates who develop anaemia secondary to iatrogenic blood loss, iron deficiency, sepsis or anaemia of prematurity are often given many blood transfusions during their

Blood Transfus 2011;9:86-94 DOI 10.2450/2010.0031-10 86

© SIMTI Servizi Srl

Neonatal RBC transfusions

time in hospital1,2. In seriously ill patients, haematocrit levels are maintained at a pre-defined threshold, whereas in stable neonates, transfusions are often given based on symptoms of anaemia3. Attempts have been made to decrease blood sampling in many neonatal intensive care units (NICU) by sending fewer specimens for laboratory tests, using in-line blood gas and chemistry monitors and developing unit-specific transfusion guidelines for premature neonates4. Despite these measures, many hospitalised neonates will receive multiple transfusions during the post-natal period. RBC transfusion guidelines have been implemented in many NICUs in order to decrease the number of transfusions in the critically ill neonate. Neonates managed without recourse to transfusion guidelines are twice as likely to receive a blood transfusion as those managed in compliance with guidelines5. Strict use of guidelines has been found to decrease the exposure of very low birth-weight neonates to donor blood products and guidelines are, therefore, in place at our and other institutions6,7. In addition to guidelines, the decision to transfuse is often dependent on the patient's clinical status (i.e. increasing apnoea and desaturations). However, these symptoms can be due to other factors including sepsis, gastro-oesophageal reflux, and respiratory distress, apnoea of prematurity and necrotising enterocolitis. Many times, several clinical interventions such as packed RBC transfusions, starting antibiotics and holding feeds, are implemented simultaneously, further confusing the validity of each of the interventions. The main objective of this study was to determine whether current NICU guidelines for packed RBC transfusion are better than caregivers' perception of

symptoms in recognising patients who would benefit from a packed RBC transfusion. Our secondary objective was to ascertain which symptoms of anaemia were most indicative of a patient who would benefit from a packed RBC transfusion. Our hypothesis was that neonates transfused based on guidelines alone were less likely to show significant clinical improvement and that caregivers' perceptions are highly predictive of a need for packed RBC transfusion.

Methods This Institutional Review Board-approved pilot study was conducted at New York University (NYU) Medical Center NICU between July 2006 and October 2007. All neonates who received a packed RBC transfusion during this period were eligible for the study. Guidelines for packed RBC transfusion in our NICU are based on the severity of illness, as shown in Table I. These neonatal institutional guidelines (last revised 2006) were established based on NYU neonatology faculty clinical opinion, internal therapeutic discussion, and recent literature at the time of implementation8-10. Our current practice is to administer a packed RBC volume of 15 mL/kg over 4 hours, using blood that is less than 10 days old. House staff, nurses, nurse practitioners, and physicians' assistants who were involved in the clinical decision-making to transfuse the neonate were asked to complete a structured survey on an optional basis (Table II). The final decision to transfuse the patient was made by the nurse practitioner, physician's assistant and/or physician caring for the child. Twentyfour hours after the transfusion was completed, the same caregiver filled out the second part of the survey

Table I - Guidelines for packed red blood cell transfusion in the NICU. 1. Acutely bleeding infant. 2. Infants with severe cardiopulmonary disease [high frequency ventilation (HFV), nitric oxide (NO), fractional inspired oxygen (FiO2) >50%, mean airway pressure (MAP) >8 cm], transfuse for haematocrit 35%], transfuse for haematocrit 8 cm) with a haematocrit 35%) with a haematocrit