Citrate metabolism and its complications in non-massive blood transfusions: Association with decompensated metabolic alkalosis+respiratory acidosis and serum electrolyte levels

Transfusion and Apheresis Science xxx (2014) xxx–xxx

Contents lists available at ScienceDirect

Transfusion and Apheresis Science

journal homepage: www.elsevier .com/ locate/ t ransc i

Citrate metabolism and its complications in non-massive bloodtransfusions: Association with decompensated metabolicalkalosis + respiratory acidosis and serum electrolyte levels

http://dx.doi.org/10.1016/j.transci.2014.03.0021473-0502/� 2014 Elsevier Ltd. All rights reserved.

⇑ Corresponding author. Present address: Kafkas University Faculty ofMedicine, Unit of Pediatric Hematology, Kars, Turkey. Tel.: +90 312 33609 09/7448; fax: +90 312 334 03 52.

E-mail addresses: [email protected] (Z. Bıçakçı), [email protected] (L. Olcay).

Please cite this article in press as: Bıçakçı Z, Olcay L. Citrate metabolism and its complications in non-massive blood transfusions: Associatidecompensated metabolic alkalosis + respiratory acidosis and serum electrolyte levels. Transf Apheres Sci (2014), http://dx.doi.org/1j.transci.2014.03.002

Zafer Bıçakçı ⇑, Lale OlcayDr. Abdurrahman Yurtaslan Ankara Oncology Training and Research Hospital, Unit of Pediatric Hematology, Demetevler, Ankara, Turkey

a r t i c l e i n f o a b s t r a c t

Article history:Received 10 November 2013Received in revised form 11 February 2014Accepted 4 March 2014Available online xxxx

Keywords:Citrate metabolismCarbon dioxideMetabolic alkalosisRespiratory acidosisElectrolyte imbalance

Background and objectives: Metabolic alkalosis, which is a non-massive blood transfusioncomplication, is not reported in the literature although metabolic alkalosis dependent oncitrate metabolism is reported to be a massive blood transfusion complication. The aimof this study was to investigate the effect of elevated carbon dioxide production due to cit-rate metabolism and serum electrolyte imbalance in patients who received frequent non-massive blood transfusions.Materials and methods: Fifteen inpatients who were diagnosed with different conditionsand who received frequent blood transfusions (10–30 ml/kg/day) were prospectively eval-uated. Patients who had initial metabolic alkalosis (bicarbonate > 26 mmol/l), who neededat least one intensive blood transfusion in one-to-three days for a period of at least 15 days,and whose total transfusion amount did not fit the massive blood transfusion definition(<80 ml/kg) were included in the study.Results: The estimated mean total citrate administered via blood and blood products wascalculated as 43.2 ± 34.19 mg/kg/day (a total of 647.70 mg/kg in 15 days). Decompensatedmetabolic alkalosis + respiratory acidosis developed as a result of citrate metabolism. Therewas a positive correlation between cumulative amount of citrate and the use of fresh fro-zen plasma, venous blood pH, ionized calcium, serum-blood gas sodium and mortality,whereas there was a negative correlation between cumulative amount of citrate and serumcalcium levels, serum phosphorus levels and amount of urine chloride.Conclusion: In non-massive, but frequent blood transfusions, elevated carbon dioxide pro-duction due to citrate metabolism causes intracellular acidosis. As a result of intracellularacidosis compensation, decompensated metabolic alkalosis + respiratory acidosis and electro-lyte imbalance may develop. This situation may contribute to the increase in mortality. Inconclusion, it should be noted that non-massive, but frequent blood transfusions mayresult in certain complications.

� 2014 Elsevier Ltd. All rights reserved.

1. Introduction

Citrate intoxication is a frequent complication aftermassive blood transfusions and often presents itself as met-abolic alkalosis. The reason this condition occurs is due tothe conversion of citrate, which is used as an anticoagulantin blood bags, to bicarbonate, and this conversion occurs

on with0.1016/

http://dx.doi.org/10.1016/j.transci.2014.03.002

mailto:[email protected]




http://www.sciencedirect.com/science/journal/14730502

http://www.elsevier.com/locate/transci



2 Z. Bıçakçı, L. Olcay / Transfusion and Apheresis Science xxx (2014) xxx–xxx

predominantly in the liver [1]. During orthotopic livertransplantation, metabolic alkalosis associated with mas-sive blood transfusions developed in 40–64% of the caseson approximately the third and fourth days after transplan-tation [2–5]. Also, metabolic alkalosis was detected in 49%and 52% of pediatric patients who underwent open heartsurgery within 24.3 h and 2.7 ± 1.5 days, respectively [6,7].

Many patients, particularly hematology patients withbone marrow deficiency such as leukemia and aplasticanemia receive blood transfusions intensively eventhough it is not defined as massive. Patients who are fol-lowed up with a diagnosis of leukemia or aplastic anemiacan be transfused with 10–30 ml/kg/day erythrocyte andthrombocyte suspension at a rate of approximately 15–20 unit/month in the first one or two months of admis-sion. In cases with disseminated intravascular coagulation(DIC) which developed due to various causes, additionalfresh frozen plasma (FFP) that contains the highestamount of citrate (14 mEq/l) can be administered to pa-tients. The aforementioned amounts do not fit into the‘‘massive blood transfusion’’ definition, but neverthelesscorrespond to a considerably extensive transfusionamount. This is because the amount of blood or bloodproduct transfusion within 24 h does not exceed 80 ml/kg.

While blood transfusion is usually a life-saving proce-dure, it implies various complications as it is a kind of tis-sue implantation. Transfusion associated complicationsare classified as immunological complication and non-immunological complications. Non-immunological com-plications include circulation overload, transfusion-dependent sepsis, hemosiderosis, anticoagulant complica-tions (citrate toxicity and metabolism), gas embolism,cold-induced thrombopathy and viral transmission [8].

According to our literature search, metabolic alkalosisis reported as a well-known complication of massiveblood transfusion, but it is not mentioned as a complica-tion for non-massive blood transfusions. While theamount of estimated citrate these patients received in aday (six to twenty-four hours) due to massive transfusionwere 9164 ± 4870 mg citrate/day [2], the patients in thisstudy were administered the same amount of citrate inapproximately fifteen days using non-massive bloodtransfusion (647.70 mg/kg), and the difference betweenthese types of transfusions (massive and non-massive) isthe duration.

A correlation between the increase in the number ofthrombocyte transfusions and the increase in mortalitywas reported in patients who have thrombocytopeniafor various reasons and who are administered thrombo-cyte suspension in newborn intensive care units. Theunderlying reason for this correlation is not known [9].The association between blood transfusion and age andmortality is not highly reported except for the newbornperiod.

This study aimed to examine the association betweencarbon dioxide production, which is elevated as a resultof citrate metabolism and serum electrolytes in patientswho were followed up with a diagnosis such as leuke-mia/aplastic anemia and who received non-massive bloodtransfusions.

Please cite this article in press as: Bıçakçı Z, Olcay L. Citrate metabolism anddecompensated metabolic alkalosis + respiratory acidosis and serum elecj.transci.2014.03.002

2. Materials and methods

2.1. Study population

Fifteen patients who were monitored in Dr. Abdurrah-man Yurtarslan Ankara Oncology Training and ResearchHospital Pediatric Hematology Clinic with various diagno-ses between March 2008–March 2011, and who received‘‘frequent blood transfusion’’ (10–30 ml/kg/day) in the last15 days of monitoring were prospectively evaluated: themeasurements are indicated below. Approval was obtainedfrom the institutional Ethics Committee.

2.2. Inclusion criteria

The inclusion criteria included the following: Beingscheduled for transfusion; initial serum actual bicarbonatelevel > 26 mmol/l, which is analyzed one day after transfu-sion; no kidney failure; no diuretic and bicarbonate treat-ment; at least one intensive blood transfusion usingerythrocyte, thrombocyte, granulocyte suspensions and/or FFP (approximately 10–30 ml/kg/day) during at leastthe last fifteen days and in less than three days; non-massive blood transfusion (less than 80 ml/kg); absenceof any pathology that would explain observed electrolyteimbalance, such as insufficient uptake, excessive sweating,burns, central nervous system (CNS) diseases, hormonaldisorders, loss from kidneys (including diuretic uptake)and gastrointestinal system.

2.3. Metabolic alkalosis definition

Metabolic alkalosis was defined as (i) a base excess (BE)value higher than �2.5 and/or an actual bicarbonatelevel P 26 mmol/l in successive controls and (ii) presenceof one metabolic alkalosis in more than two successiveblood gas measurements [2,10].

2.4. Analysis of blood gas

Venous blood was used during the blood gas measure-ments, as the patients had thrombocytopenia, and dissem-inated intravascular coagulation (DIC) and arterial blooddrawing possess high risk for bleeding. As known, if themeasurement of partial oxygen pressure is not required,venous blood gas analysis is usually sufficient to evaluatethe acid–base equilibrium. The pH of venous blood is0.04 units lower than arterial pH, whereas venous bloodPCO2 is higher than 5 mmHg. In addition, venous PO2 andbicarbonate levels are also lower. However, venous bicar-bonate measurement is considered as equally informativefor arterial base deficit in intensive care patients [11]. Inthis study, patients who required transfusion in 3 days ora shorter time in the last 15 days, and will receive transfu-sion were chosen and venous blood samples were taken24 h after the new transfusion. Blood and urine electrolyteswere studied with co morbid blood gas. The highest actualbicarbonate level was determined among these blood sam-ples. The amount of blood and blood products given to thepatients in 15 days was indicated prior to the highest

its complications in non-massive blood transfusions: Association withtrolyte levels. Transf Apheres Sci (2014), http://dx.doi.org/10.1016/



Z. Bıçakçı, L. Olcay / Transfusion and Apheresis Science xxx (2014) xxx–xxx 3

actual bicarbonate level determined. The estimated citrateamount within the blood and blood products administeredto the patient in the last 15 day-period was calculated.

2.5. Determining blood amount for transfusion

Patients who had hemoglobin levels lower than 7 g/dlwere administered 10 ml/kg erythrocyte suspension. Ran-dom thrombocyte suspension (1 unit/10 kg) was adminis-tered to some patients who had thrombocyte levelsbetween 10 and 20,000/mm3 (depending on the conditionand diagnosis) and to all patients who had thrombocytelevels lower than 10,000/mm3. The dose of erythrocyte sus-pension for newborns ranged between 5 and 10 ml/kg. Eachunit of apheresis thrombocyte suspension was assumed tobe equivalent to 6–8 units of random thrombocyte suspen-sion. 10 ml/kg of FFP transfusion was performed for pa-tients who had severe sepsis, widespread intravascularcoagulation and who had Hemophilia A. A collected granu-locyte unit was administered in severe and non-recoveringneutropenia (granulocyte number < 100/mm3) [12].

2.6. Calculation of citrate in administered blood products

The approximate blood volume of a normal adult(70 kg) is 80 ml/kg [13]. Massive blood transfusion is de-fined as the administration of a stored blood sample thatis equal to the patient’s whole blood volume (for instance,administration of 3200 ml of blood product to a 40 kgchild, with a dose of 80 ml/kg) in six or twenty-four hours[14]. To put it another way, ‘‘massive blood transfusion’’ isdefined as the administration of ten or more blood units(80 ml/kg or higher in children) to an adult in six ortwenty-four hours [15]. A whole blood bag containsapproximately 450 ml of blood and 63 ml of CPD (citratephosphate dextrose), which in turn contains 1.863 g of cit-rate. Thus, the amount of citrate administered to the bodyof a 70 kg adult via massive blood transfusion is at least298 mg/kg (approximately 70 kg � 80 ml/kg = 5600 ml).

Here, mmol/l was used as the concentration unit incompliance with the international system. Trisodium cit-rate is equivalent to 9.86 mmol/l (9.9 mmol/l is approxi-mately 2.9 g/l) of citrate (C6H5O7), which is equivalent toapproximately 29.6 (2.9 g/l) milliequivalent/l [16].

The amount of citrate administered to each patient withtransfusion and the citrate content of each blood productwere calculated according to the guidelines by Driscoll etal. [17]. In our hospital, citrate–phosphate–dextrose–adenine is used as anticoagulant, and each blood bag con-tains approximately 18 mEq/l of citrate. Each unit of FFP(200–250 ml) contains approximately 14 mEq citrate, ran-dom thrombocyte suspension (50–75 ml) contains 3 mEqof citrate, cryoprecipitate (30–40 cc) contains 1 mEq of cit-rate and erythrocyte suspension (220–240 cc) contains0 mEq of citrate [4]. Each bag contained 18 mEq/l of pro-tective anticoagulant citrate solution, which is approxi-mately equivalent to 1.863 g of citrate. Thus, each unit ofadministered blood and blood products is assumed to con-tain the following amounts of citrate: 1400 mg of citratefor FFP, 300 mg of citrate for random thrombocyte suspen-sion, 100 mg of citrate for cryoprecipitate, approximately


63 g of citrate for erythrocyte suspension, 1.29 g of citratefor apheresis thrombocyte suspension, and 0.8889 mg ofcitrate for granulocyte suspension [17].

2.7. Other examinations

2.7.1. Apheresis thrombocyte and granulocyte preparationApheresis thrombocyte and apheresis granulocyte sus-

pension was prepared using the Fresenius Com.Tec model,continuous current apheresis instrument, Granulo/MNCprotocol and P1Y/C5L thrombophoresis kit.

2.7.2. Fluid, electrolyte treatments and chemotherapyprotocols

Generally, patients were given 1500–3000 ml/m2/dayfluid treatment for varying time periods depending onthe cases. Hypokalsemia was treated with 4� 0.5–1 ml/kg of 10% calcium gluconate, and hypokalemia was treatedintravenously by including 20–60 ml/l of KCl in the dailyfluid. Adjusted calcium was calculated using ‘MeasuredCa2+ + 0.8 � (4-patient’s albumin)’ formula.

TRALL-BFM 2000 chemotherapy protocol was used forALL, and BFM 2004 chemotherapy protocol was used forAML [18,19].

2.7.3. Data recordingFor each patient, certain demographic properties such

as diagnosis, age and gender were recorded; biochemicaland blood gas parameters were analyzed in the hospital’sCentral Laboratory.

3. Statistical analysis

Data were analyzed using SPSS for Windows 11.5 soft-ware. Shapiro Wilk test was used to analyze the fitnessof the continuous parameters to normal distribution. Stu-dent’s t-test was used to analyze the significance of the dif-ferences in parameters that were normally distributedbetween the groups, whereas Mann–Whitney U-test wasused to analyze the significance of the differences inparameters that were not normally distributed betweenthe groups. Spearman’s correlation test was used to ana-lyze the significance of correlation between the continuousparameters. P values < 0.05 were considered as statisticallysignificant.

4. Results

Nine out of 15 patients (60%) were female, and sixpatients (40%) were male. The mean age was 11.0 ± 6.0(1–18) years, and the mean body weight was 36.2 ± 22.1(9–83) kilograms. Six patients had AML, four patients hadALL, one patient had CML, one patient had Hemophilia Aand one patient had Evans syndrome. Seven patients(46.7%) deceased during metabolic alkalosis. In seven pa-tients (46.7%), metabolic alkalosis disappeared as theirneed for transfusion decreased. These patients currentlyfollow up and are treated at our clinic (Table 1). Onepatient (6.6%) (Case 13) deceased six months after recover-ing from metabolic alkalosis.




Table 1Demographic properties of the patients.

Patient code Gender Age (Years) Body weight (kg) Diagnosis Outcome

1 F 15 47 ALL-L1 Alive2 M 17 83 AML-M0 Alive3 F 16 38 AML-M2 Alive4 M 10 35 Acquired aplastic anemia Alive5 M 18 53 ALL-L2 Ph(+) (Relapse) Alive6 M 5.3 15 ALL-L1 (Relapse) Deceased7 M 5.4 17.5 Hemophilia A (Intermediate) Alive8 F 3.2 10 AML-M2 Deceased9 F 2.5 10 Evans syndrome Deceased

10 F 15.8 40 AML-M0 (Refractory) Deceased11 F 10 34 ALL-L2 Ph(+) (Relapse) Alive12 F 16 44 CML (Blastic phase AML-M2)(Refractory) Deceased13 M 17 35 Post-osteosarcoma treatment/Acquired aplastic anemia Deceased14 F 12.5 72 AML-M2 Deceased15 F 1 9 AML-M7 DeceasedMean 9 F (60.0%)

6 M (40.0%)11.0 ± 6.0(min:1–max:18)

36.2 ± 22.1(min:9–max:83)

Alive 8 (53.3%)Deceased 7 (46.7%)

Deceased�: Case 13 was deceased 6 months after exiting metabolic alkalosis.


Total amount of blood and blood products given to thepatients in 15 days prior to the highest actual bicarbonatelevel determined, and the amount of the citrate transfusedthrough these blood products are presented in Table 2. Theestimated citrate content of these transfused products wascalculated according to the guidelines by Driscol et al. [17](Table 2). The mean estimated citrate load was calculatedas 43.2 ± 34.19 (11.3–120.7) mg/kg/day (a total of647.70 mg/kg in 15 days).

The only significant correlation was between theamount fresh frozen plasma and estimated citrate amount,and estimated citrate amount increased with increasingamount of fresh frozen plasma (r = 0.738 and p = 0.002).There was no significant correlation between other blood/blood products and the estimated citrate amount (p > 0.05).

PCO2, pH, base excess, actual bicarbonate, standardbicarbonate and total carbon dioxide levels generally in-creased with increasing estimated citrate amount depend-ing on the transfusion number, frequency and amount(Table 3).

Table 2Blood and blood products administered to the patients in 15 days and estimated

Patient codenumber

Erythrocytesuspension

Random thrombocytesuspension

Aferez thrombsuspension

1 9 22 32 8 14 73 10 11 94 2 11 25 7 17 26 4 2978 2 3 49 4 6

10 6 14 411 4 16 712 10 913 4 6 314 16 16 715 4 6 1Total 90 171 58Mean 6.0 ± 4.09 114 ± 8.14 3.9 ± 3.20


The only significant correlation was between the pH le-vel and estimated citrate amount, and the pH levels in-creased as estimated citrate amount increased (r = 0.546and p = 0.035). There was no significant correlation be-tween other blood gas indicators and estimated citrateamount (p > 0.05). Adjusted calcium levels were calculatedas 8.444 mg/dl (7.9 ± 1.13 + 0.544), which is lower than thenormal range (8.6–10.3). Other laboratory findings of thepatients are presented in Table 4.

There was a significant correlation between Na and esti-mated citrate amount, and Na levels increased withincreasing estimated citrate amount (r = 0.550 andp = 0.034). There was a significant inverse correlation be-tween calcium and estimated citrate amount, and calciumlevels decreased with increasing citrate amount(r = �0.551 and p = 0.041). In addition, there was a signifi-cant inverse correlation between phosphorus and esti-mated citrate amount, and phosphorus levels decreasedwith increasing estimated citrate levels (r = �0.586 andp = 0.045). There was no significant correlation between

citrate amount.

ocyte Fresh frozenplasma

Granulocytesuspension

Estimated citrate load(mg/kg/day)

18.069 21.152 29.91

11.4412.76

13 120.6716 85.33

41.245 60.34

16.2327.61

10 39.7511.28

29 4 54.617 97.3491 4 43.186.1 ± 8.34 0.3 ± 1.03 43.2 ± 34.1(11.3–

120.7)9




Table 3Patients’ blood gas results and estimated citrate amounts.

Patientcodenumber

PCO2

(45 mmHg)PO2

(40 mmHg)pH(7.31–41)

Base excess(+2.5 mmol/l)

AktüelHCO3((22–26 mmol/l)

Standart HCO3

(22–26 mmol/l)Total CO2

(25–29 mmol/l)

Estimated citrateload (mg/kg/day)

1 41.2 34.2 7.419 1.3 26.1 25.5 24.4 18.062 50.1 29.0 7.350 1.4 27.0 25.0 25.9 21.153 36.4 73.3 7.492 4.3 27.3 27.4 26.0 29.914 48.3 20.9 7.371 0.9 27.4 25.6 26.0 11.445 41.9 49.4 7.434 3.3 27.4 26.7 26.2 12.766 39.3 100.8 7.462 4.3 28.3 – 29.5 120.677 52.8 22.5 7.370 2.6 29.8 27.1 28.0 85.338 53.9 37.3 7.541 6.9 30.0 – – 41.249 40.2 41.7 7.496 6.8 30.4 29.9 28.8 60.34

10 47.2 38.5 7.449 7.2 32.0 30.0 29.7 16.2311 49.1 58.9 7.459 9.7 34.1 32.1 31.9 27.6112 50.4 61.2 7.466 12.7 36.7 35.2 38.2 39.7513 54.1 79.6 7.448 13.5 37.7 35.8 39.4 11.2814 51.6 102.2 7.519 19.4 42.5 41.7 44.0 54.6115 54.4 36.4 7.566 27.5 49.8 49.0 51.4 97.34Mean 47.4 ± 6.04 44.8 ± 18.53 7.5 ± 0.06 8.1 ± 7.49 32.4 ± 6.72 31.6 ± 7.19 32.1 ± 8.09 43.2 ± 34.19

(11.3–120.7)

Table 4Various laboratory findings of patients with metabolic alkalosis.

Patientcodenumbers

Na+ mmol/l (131–140)

K+

mmol/l(3.5–5.1)

Cl mmol/l (98–107)

BUN mg/dl (9–21)

Creatininemg/dl(0.6–1.3)

Uric acidmg/dl(2.6–7.2)

Ca++ mg/dl (8.6–10.3)

P mg/dL(3.1–6.2)

Mg++

mg/dl(1.6–2.6)

Totalprotein g/dl (6.4–8.6)

Albumin(g/dl)(3.8–5.6)

1 135 3.5 – 13 0.5 2.2 8.6 3.3 1.4 5.9 4.12 138 4.0 – 14 0.8 2.6 9.2 3.6 2.1 5.9 3.93 138 2.9 – 16 0.4 1.4 8.0 4.2 – 4.9 2.84 140 3.0 – 20 0.9 3.0 9.1 4.5 – 6.8 3.65 137 4.6 – 16 0.5 3.7 8.5 – – 6.1 4.46 150 2.6 93 12 0.4 2.1 6.2 3.3 – 6.3 3.67 140 3.0 – 10 0.6 4.1 – – – – –8 136 1.4 91 6 0.4 1.3 7.5 2.2 – 5.5 3.39 153 3.6 – 11 0.4 2.1 9.0 – – 5.3 3.0

10 135 2.6 94 3 0.4 2.3 7.0 4.5 1.4 5.6 3.411 144 3.9 103 12 0.1 3.2 8.9 2.7 – 4.5 2.612 140 3.5 101 16 0.4 2.6 6.5 4.4 – 4.8 2.713 137 3.3 102 8 0.6 3.5 8.5 4.8 1.3 5.7 3.114 144 2.7 104 21 0.8 5.0 6.7 4.5 – 5.2 2.815 138 1.8 88 22 0.4 5.0 6.2 1.9 1.2 5.3 3.3Mean 140.3 ± 5.3 3.1 ± 0.8 97.0 ± 6.1 13.3 ± 5.4 0.50 ± 0.2 2.9 ± 1.1 7.9 ± 1.1 3.7 ± 0.9 1.5 ± 0.3 5.6 ± 0.6 3.3 ± 0.5


other laboratory findings and estimated citrate levels, andbetween lactate and estimated citrate levels (p > 0.05).

The mean urine (spot) electrolyte levels in patients arepresented in Table 5.

There was a significant inverse correlation betweenonly chloride level and estimated citrate amount, andchloride levels decreased as estimated citrate amountincreased (r = �0.667 and p = 0.500). There was no signifi-cant correlation between the remaining urine electrolytesand estimated citrate amount (p > 0.05).

5. Discussion

5.1. Estimated citrate amount-metabolic acidosis association

Metabolic alkalosis is a major transfusion complicationresulting from the metabolism of citrate that is uptaken via


blood transfusion. Once within the circulation, citrate isoxidized in the tricarboxylic acid (TCA) cycle in the mito-chondria in mainly the liver, and in the skeletal muscleand renal cortex to a less extent, and is rapidly convertedinto carbon dioxide and water, and energy is released inthe form of heat. Citrate is first metabolized to carbondioxide in 1:3 ratio, and then to bicarbonate [20,21]. Ittakes approximately 3–4 days for the kidneys to eliminatethe elevated bicarbonate levels [11]. The reason that thepatients in this study were required to receive blood/bloodproduct transfusion at least once in a period less than3 days is that bicarbonate is eliminated through the kid-neys in approximately 3 or 4 days. The finding that therewas a significant correlation only between the pH leveland the estimated citrate amount implied that citratewas first metabolized into carbon dioxide, and then itwas converted into bicarbonate, and repeated transfusionscaused an increase in the cumulative bicarbonate levels.




Table 5Urine electrolytes in patients with metabolic alkalosis.

Patient codenumbers

Na+ mmol/l (40–220)(>30)

K+ mmol/l (2.5–125)(>15)

Cl mmol/l (15–40)(>20)

Ca++ mg/dl (2–6)(>3.5 mg/kg)

P mg/dl(15–35)

Cr mg/dl(5.2–41)

1 81 11.1 – – – 72 139 65.9 138 8.8 26.4 1983 115 29.3 – 5.6 31.8 304 122 31.5 143 10.7 63.4 60.85 122 19.2 152 0.2 29.7 33.646 96 25.4 94 1.2 3.8 24.37 – – – – – –8 53 28.3 51 0.3 11.4 169 – – – – – –

10 81 6.2 – – – –11 61 12.6 62 10.8 0 15.312 73 23.34 62.5 – – 37.213 101 8.2 99 1.4 13.2 23.514 36 27.6 20 1.4 4.6 140.915 – – – – – –Average 86.4 ± 36.92 23.9 ± 15.96 88.1 ± 49.30 3.3 ± 4.01 20.5 ± 19.92 58.0 ± 61.60


Metabolic alkalosis was reported in 40% [4], 52.1% [2]and 100% [17] of patients who received massive bloodtransfusion due to orthotopic liver transplantation, in 40–64% of adult patients after surgery[22,23], and in 49–52%of pediatric patients who underwent open heart surgery[6,7] on the third [4] and fourth [2] days after the opera-tion. While the amount of estimated citrate these patientsreceived in a day (six to twenty-four hours) due to massivetransfusion were 9164 ± 4870 mg citrate/day [2], and 620–1000 mg/kg/day [4,17], the patients in this study wereadministered the same amount of citrate in approximatelyfifteen days using non-massive blood transfusion(647.70 mg/kg), and the difference between these typesof transfusions (massive and non-massive) is the duration.

Bicarbonate accumulation occurs due to repeated trans-fusions in an approximately three to four day period, with-out full elimination and excretion of the bicarbonateproduced as a result of previous transfusions. In this case,we believe that metabolic alkalosis developed as a resultof citrate administration, which was higher than that ofmassive transfusions (approximately 298 mg/kg/day), in afifteen-day period via more than one transfusion.

5.2. Metabolic alkalosis-age relationship

The mean age of pediatric patients who have metabolicalkalosis as a result of massive blood transfusion wasreported as 22.6 ± 22.2 months[6] and 15 weeks (2 days–95 weeks) [7]. Seventy-two percent of these patients weresmaller than 12 months [6], and the frequency of meta-bolic alkalosis development was inversely correlated withage [7]. While there was no history of diuretic use in thesepatients’ histories, lower age is emphasized as a factorincreasing the frequency of metabolic alkalosis [6,7]. Theresults of the patients implied that the use of diuretics in-creased the mortality.

5.3. Thrombocyte transfusion frequency-mortality associationin newborns

Various studies on newborn patients with severe throm-bocytopenia reported that the incidence of mortality was


proportional to the number of thrombocyte transfusions[9,24–27], but not with low number of thrombocytes[24,25]. The incidence of mortality was 2% in newbornswho did not receive transfusion, 11% in newborns who re-ceived 1–2 thrombocyte transfusions, and 35% in newbornswho received 20 or more thrombocyte transfusions [9].

When we calculated the estimated citrate amountadministered to these patients, 278 patients received 30–60 mg/kg of citrate (1–2 times, 5–10 ml/kg), 167 patientsreceived 100–300 mg/kg of citrate (3–10 times, 15–50 ml/kg), 25 patients received 300–600 mg/kg of citrate(11–20 times, 55–100 ml/kg) and 24 patients received600–900 mg/kg of citrate (over 20 times, 105–150 ml/kg).In other words, more than two to threefold citrate, as com-pared to massive blood transfusion (298 mg/kg/day), wasadministered with these transfusions in the aforemen-tioned study [9]. While this association between the in-creased mortality and increased number of thrombocytetransfusions in the newborn intensive care unit sometimesseems impossible to be quantified or known just like thedisease severity, it is supported by susceptibility tests [9].

The underlying cause of the increased mortality rate dueto increased number of thrombocyte suspension transfu-sions is that carbon dioxide production is elevated as a resultof citrate metabolism, which in turn causes carbon dioxideaccumulation leading to electrolyte imbalance. We believethat mortality rate is higher compared to other age periods,as the ability of children to compensate hyperkapnia andrespiratory acidosis is limited in the period between theneonatal period and two years of age [28]. Despite the insuf-ficient number of patients, we found it striking that the ageof one deceased patient (Case 15) was one.

5.4. Metabolic alkalosis and hypochloremia association

In respiratory acidosis, serum bicarbonate concentra-tions rise via a renal compensation mechanism. It can takea short time, like an hour, for the kidney to give these re-sponses after a considerably slow change in PCO2. Fullcompensation depends on sufficient renal functions. Sincebicarbonate is eliminated through the kidneys, it takes3–4 days for its excretion [11]. Following the compensa-





tion of respiratory acidosis, secondary hypokalemia, post-hypercapnic metabolic alkalosis and chloride deficiencydevelops in the patients [11,29]. In an aforementionedstudy that evaluated 56 pediatric cases who received openheart surgery, chloride levels were 98.5 ± 4.1 mmol/l lowerin patients with metabolic alkalosis. Chloride loss can de-velop depending on various factors (medication, fluid,etc.), in addition to the use of diuretics [6]. Despite the factthat patients received intravenous NaCl- and KCl-contain-ing solutions, serum chloride was lower than normal andurine chloride was higher than normal; we cannot link thisobservation to chloride loss that causes metabolic alkalosisand subsequent decrease in intravascular fluid. We believethat the underlying reasons for hypochloremia are as fol-lows: (1) Due to ongoing metabolic alkalosis, bicarbonateis transported via Cl�/HCO3

� antiporter in exchange forchloride [30]. (2) During the export of H+ and HCO3 intothe lumen via H+ pump (H+-ATPase) in the A-type maincells in the cortical region of the collecting duct, the Cl�/HCO3

� antiporter and chloride channel in the luminal mem-brane, K+ and Cl� are imported into the cell [31,32]. (3)Bicarbonate–chloride exchanger in the bone osteoclastsexports bicarbonate and imports chloride to maintain aneutral pH [29,30]. We believe that hypochloremia devel-oped due to the hyperfunction of Cl�/HCO3

� antiporterpresent in three different organs to prevent intracellularacidosis (due to elevated PCO2), chloride accumulation inerythrocytes and osteoclasts, and the loss of chloride inkidney tubule cells.

5.5. Metabolic alkalosis and hypokalemia association

The intracellular potassium concentration/extracellularpotassium concentration ratio is high. This ratio is primarilymaintained via Na+–K+ ATPase (sodium pump) [29]. In astudy that evaluated 123 cases who received orthotopicliver transplantation, metabolic alkalosis and hypokalemiawere observed in 60 cases [2]. Despite the fact that themajority of patients received intravenous KCl (20–60 mEq/l) and were administered potassium via blood transfusion(potassium levels increase as blood is left to stand) [33],hypopotassemia was observed (mean 3.09 mmol/l(1.4–4.6)), and this situation can be explained by the loss ofpotassium in Na+–K+ ATPase activity and sodium gain as aresult of metabolic alkalosis compensation. The potassiumloss of Na+–K+ ATPase activity and sodium gain might havecontributed to the observation that Na+ concentration wasnormal or on the upper limit of the normal range, despitethe fact that most patients received high amounts(2500–3000 ml/m2) of diluted (1/3–1/4 NaCl + 5% dextrose)solution. The high level of mean urine potassium(24.05 mmol/l > 15 mmol/l) in patients indicate thatpotassium is lost with urine. Moreover, urine sodium levelsshould be low (<25 mmol/l) as a result of Na+–K+ ATPaseactivity, whereas high urine sodium levels (90.00 mmol/l >25 mmol/l) were observed. This situation implied inappro-priate ADH syndrome, nephropathy that causes salt loss,mineralocorticoid deficiency and cerebral salt loss [29].However, hyponatremia is not observed in all of these dis-eases. In addition to the absence of hyponatremia, these dis-eases (diagnoses) were ruled out due to absence of patient


history, physical examination and additional laboratory find-ings which would imply hyponatremia. Despite elevatedfluid volume and Na+–K+ ATPase activity, high sodium excre-tion in urine may be explained with elevated ANP secondaryto hyperkapnia [11].

5.6. Hypernatremia development

In our clinic, isotonic NaCl is not used due to the possi-bility of hypernatremia, acidosis or fluid overload. Gener-ally, solutions containing 1/3–1/4 NaCl + 5% dextrose areused. The use of such hypotonic solutions is not physiolog-ical, and even dangerous, leading commonly to hyponatre-mia. Hyponatremia is a common electrolyte imbalance inpediatric patients. Regarding patients who underwent ser-um sodium analysis in the emergency service, 8.2% of thepatients, and 9% of the hospitalized children had hypona-tremia [34]. The prevalence of hypernatremia, dependingon the data used, is estimated in 0.22–1.4% of childrenadmitted to the hospital [35]. In our country, 11.7% of new-ly diagnosed ALL cases had hyponatremia, and 9.5% of thenewly diagnosed ALL cases had hypernatremia [36]. In astudy by Oopik et al. on healthy athletes, the authors deter-mined that oral administration of 0.5 g/kg of sodium cit-rate resulted in an increase in serum sodium levels [37].Most of the patients who received intravenous solutionduring the hospitalization period in the clinic are expectedto develop dilutional hyponatremia depending on theamount and duration of the solution administered. Despitedilutional hyponatremia expectation, sodium concentra-tion was either in the normal range or at the upper limitof the normal range in all six cases (Cases 4, 7, 8, 10, 11,12 and 13), implying the presence of sodium retention.We believe that elevated levels of sodium due to theadministration of solutions in excess and for a prolongedperiod decreased with dilution and returned to the normallimits. In this study, there was a significant correlation be-tween the increase in sodium levels, which is measuredusing an autoanalyzer, and estimated citrate amount(r = 0.550 and p = 0.034), and sodium levels increased withincreasing estimated citrate amount. As a response to theincrease in PCO2, there is an increase in the Na+/H+ antipor-ter along the luminal membrane and in the coupled trans-port of sodium-bicarbonate along the basal membrane(3HCO3

�/1Na+ co-transporter) in the proximal renal tubulecells, resulting in sodium and bicarbonate reabsorption[29,38,39]. As a result of all these compensations, we be-lieve that hyperbicarbotanemia, isonatremia/hypernatre-mia, hypochloremia, hypomagnesemia and hypokalemiadeveloped in our cases.

5.7. Hypocalcemia and heart relationship

Calcium complexes are filtered through the glomeruli inthe kidneys. They bind to certain anions, including phos-phate, citrate and bicarbonate [40]. Citric acid binds to cal-cium to form a diffusible complex, but this complex ionizespoorly. Thus, most of the calcium bound to citrate becomesbiologically inactive. As known, calcium must be in an ion-ized state to carry out its functions in cardiomyocytes,skeletal muscle and coagulation [1]. Mitochondria and





the sarcoplasmic and endoplasmic reticulum are the intra-cellular compartments with the highest concentration ofcalcium. The T-tubule system is more developed in thecardiomyocytes, whereas sarcoplasmic reticulum is lessdeveloped; therefore, the intracellular calcium resourcesare lower in quantity. Therefore, the myocardiumcontraction depends on the extracellular calcium [41].Mitochondria constitute approximately 40% or more of amyocardium [42]. Cardiomyocytes produce more bicar-bonate from citrate as they contain a higher number ofmitochondria compared to other cells. It is known thathypocalcemia frequently develops as a result of bloodtransfusions. There was a significant inverse correlationbetween calcium levels and estimated citrate amount, withdecreased calcium levels with increasing citrate amount(r = �0.551 and p = 0.041). In addition to the hypokalsemiain the systemic circulation, underdevelopment of the sar-coplasmic reticulum and limited intracellular calciumsources may cause local hypokalsemia arrhythmias thatresult from calcium binding by bicarbonate that is pro-duced by mitochondria. As hypokalsemia is immediatelyand properly treated, hypokalsemia dependent convulsionand arrhythmia were not observed in our patients. On theother hand, Ca levels in the deceased group were signifi-cantly lower compared to the survival group (p < 0.001).When calcium is depleted, isolated cardiomyocytes stopthe heart beat in approximately one minute, whereas skel-etal muscles continue contraction for a long time without acalcium source [41]. We believe that low tolerance ofcardiomyocytes to calcium deficiency is a factor that in-creases the incidence of mortality.

5.8. Estimated citrate and mortality relationship

The low estimated citrate and high actual bicarbonatelevels in some patients contrasting with the high citrateand low actual bicarbonate levels in others may originatefrom differences in diagnosis and treatment. Case 6 wasan ALL relapse case who had leukostasis syndrome and la-ter on developed smallpox pneumonia. Case11 was anotherALL relapse case who developed febrile neutropenia shortlyafter starting. The increase in leukocyte count after antibi-otic treatment was associated with more evident signsand symptoms of pulmonary inflammation. Thus, ausculta-tory crepitant rales, and tachypnea following hyperventila-tion became more obvious. Pulmonary involvementobserved in both cases may have caused ventilation perfu-sion disturbance, and thus hypoxia. Hypoxia stimulates therespiratory center potentially more than hypercapniawhich can cause hyperventilation and the elimination ofcarbon dioxide from the body [15]. The elevation of carbondioxide elimination may have balanced the metabolic alka-losis findings expected to be present in blood gas analysis inboth cases. The children in case 7 and 9 received fresh fro-zen plasma indicated for hemophilia A and blood/bloodproducts indicated for Evans Syndrome, respectively. Nei-ther of the cases received chemotherapy, thus did not havecellular dysfunction including the renal epithelial cells.Therefore, adequate respiration and metabolic compensa-tion may have been the cause of lower blood gas values ofactual bicarbonate. Case 10, received multiple agent che-


motherapy and long term (6–7 months) blood transfusions,because of a resistant AML-M0. Case 13 also received longterm (6–7 months) blood/blood product transfusionsnecessitated after the iatrogenic nephrotoxicity and ac-quired aplastic anemia which resulted from cisplatin givenfor osteosarcoma. Chemotherapy causing cellular dysfunc-tion, including the renal epithelial cells, may have pre-vented respiratory and metabolic compensation, thusleading to a higher blood gas level of actual bicarbonate.In the remaining 9 cases managed with a more homoge-neous diagnostic and therapeutic protocol the blood gasvalues of actual bicarbonate increased in a roughly directcorrelation with the amount of the received citrate levels.In previous studies the correlation of mortality rates ofthe neonates with the frequency of thrombocyte transfu-sions was demonstrated. However, citrate loading wasnot mentioned in the previous studies[9,24–27]. Similarwith the previous findings, we also observed an increasedmortality rate which correlated with the increased transfu-sion frequency. However, the findings may be related to theincreased estimated citrate level. In our study the deceasedpatients had significantly lower serum levels of potassiumand calcium (p = 0.024 and p < 0.001) and significantlyhigher estimated citrate levels and pH values (p = 0.029and p = 0.005) in comparison with the surviving patients.Thus, we think that the metabolic state designated by cit-rate metabolism contributes to the correlation of frequentnon-massive blood transfusions with the increased mortal-ity rates, although the mortality rates are most dominantlyaffected by the primary disease, patient’s age, used drugsand the general health status of the patient. The patientsdeceased due to septicemia and multi organ dysfunction.

In the present study, there was a significant inverse cor-relation between phosphorus and estimated citrate amount,and the phosphorus levels decreased with increasing citratelevels (r = �0.586 and p = 0.045). Because transfusiondependent hypocalcemia may cause secondary hyperpara-thyroidism. Secondary hyperparathyroidism, on the otherhand, is believed to cause hypophospatemia by causingphosphorus loss through the kidneys (phosphoturia).

As far as we know, this is the first study showing thatnon-massive but frequent blood transfusions can alsocause the same complications with classical massive bloodtransfusions which is performed within 24 hours’ time. Inaddition, in this study we stressed that the association be-tween blood transfusion and age and mortality is nothighly reported except for the newborn period.

Our study had a small sample and included patientswith different diagnosis. In addition, our patients wereheterogeneous both as age (1–18 year of age) and severityof disease. The study did not have a normal control groupnaturally, since frequent blood transfusions are restrictedto only sick children or those who have heavy healthproblems.

Our findings suggest that further studies determiningthe definitive relationship between citrate load and in-creased mortality rates are necessary which should in-volve multiple groups which are homogenous both as toseverity of disease and age, in order to enable comparison.So, the role of citrate as an anticoagulant should be ques-tioned further, from many aspects.





In conclusion, there is an increase in carbon dioxide pro-duction as a result of citrate metabolism in non-massive, frequent blood transfusions; elevated carbon diox-ide production causes intracellular acidosis; (decompen-sated) metabolic alkalosis + respiratory acidosis andelectrolyte imbalance such as hypocalcemia, hypokalemia,hypochloremia, iso/hypernatremia develops as a result ofthe compensation of intracellular acidosis; this affectedthe increasing mortality rates in non-massive, frequentblood transfusions (while also related to primary disease,age, medications and overall condition), and therefore,patients who are frequently administered with blood andblood products in the clinics should be monitored regardingthese aspects. Our results suggest that the definition of‘massive transfusion’ in children may be reevaluated toinclude the term ‘time’.

References

[1] Yendt ER. Citrate intoxication. Can Med Assoc J 1957;76:141–4.[2] Raj D, Abreo K, Zibari G. Metabolic alkalosis after orthotopic liver

transplantation. Am J Transplant 2003;3:1566–9.[3] Fortunato Jr FL, Kang Y, Aggarwal S, et al. Acid–base status during

and after orthotopic liver transplantation. Transplant Proc1987;19:59–60.

[4] Contreras G, Garces G, Reich J, et al. Predictors of alkalosis after livertransplantation. Am J Kidney Dis 2002;40:517–24.

[5] Jenkins RL, Benotti PN, Bothe AA, et al. Liver transplantation. SurgClin North Am 1985;65:103–22.

[6] Wong HR, Chundu KR. Metabolic alkalosis in children undergoingcardiac surgery. Crit Care Med 1993;21:884–7.

[7] van Thiel RJ, Koopman SR, Takkenberg JJ, et al. Metabolic alkalosisafter pediatric cardiac surgery. Eur J Cardiothorac Surg2005;28:229–33.

[8] Luban NLC. Wong ECC: hazards of transfusion. In: Arceci RJ, Hann IM,Smith OP, editors. Pediatric hematology. Massachusetts: BlackwellPubishing; 2006. p. 724–44.

[9] Baer VL, Lambert DK, Henry E, et al. Do platelet transfusions in theNICU adversely affect survival? Analysis of 1600 thrombocytopenicneonates in a multihospital healthcare system. J Perinatol2007;27:790–6.

[10] Moffett BS, Moffett TI, Dickerson HA. Acetazolamide therapy forhypochloremic metabolic alkalosis in pediatric patients with heartdisease. Am J Ther 2007;14:331–5.

[11] Yorgin PD, Rhee KH. Gas exchange and acid–base physiology. In:Taussig LM, Landau LI, editors. Pediatric respiratorymedicine. Missouri: Mosby; 1999. p. 212–43.

[12] Sloan SR, Friedman DF, Kao G, et al. Transfusion medicine. In: OrkinSH, Nathan DG, Ginsburg D, et al., editors. Nathan and oski’shematology of infancy and childhood. Philadelphia: SaundersElsevier; 2009. p. 1623–75.

[13] Brugnara C. Reference value in infancy and childhood. In: Orkin SH,Nathan DG, Ginsburg D, et al., editors. Nathan and oski’s hematologyof infancy and childhood. Philadelphia: Saunders Elsevier; 2009. p.1769–97.

[14] Robitaille N, Hume HA. Blood components and fractionated plasmaproducts: preparation, indication and administrations. In: Arceci RJ, HannIM, Smith OP, editors. Pediatric hematology. Massachusetts: BlackwellPublishing; 2006. p. 693–723.

[15] Perkins JG, Cap AP, Weiss BM, et al. Massive transfusion andnonsurgical hemostatic agents. Crit Care Med 2008;36:325–39.

[16] Oral rehydration salts. <http://www.who.int/child_adolescent_health/documents/fch_cah_06_1/en/index.html> [accessed 11.10.10].

[17] Driscoll DF, Bistrian BR, Jenkins RL, et al. Development of metabolicalkalosis after massive transfusion during orthotopic livertransplantation. Crit Care Med 1987;15:905–8.


[18] Yüksel-Soycan L, Aydogan G, Tanyeli A, et al. BFM-TR ALL 2000: firstTurkish multicentric study in the treatment of pediatric acutelymphoblastic leukemia. Ped Blood Cancer 2006;47:426–7.

[19] Möricke A, Reiter A, Zimmermann M, et al. Risk-adjusted therapy ofacute lymphoblastic leukemia can decrease treatment burden andimprove survival: treatment results of 2169 unselected pediatricand adolescent patients enrolled in the trial ALL-BFM 95. Blood2008;111:4477–89.

[20] Kramer L, Bauer E, Joukhadar C, et al. Citrate pharmacokinetics andmetabolism in cirrhotic and noncirrhotic critically ill patients. CritCare Med 2003;31:2450–5.

[21] Bender DA. The Citric acid cycle: the catabolism of acetyl-CoA. In:Murray RK, Bender DA, Botham KM, et al., editors. Harper’s illustratedbiochemistry. China: McGraw-Hill Companies; 2009. p. 143–8.

[22] Okusawa S, Aikawa N, Abe O. Postoperative metabolic alkalosisfollowing general surgery: its incidence and possible etiology. Jpn JSurg 1989;19:312–8.

[23] Hodgkin JE, Soeprono FF, Chan DM. Incidence of metabolic alkalemiain hospitalized patients. Crit Care Med 1980;8:725–8.

[24] Baer VL, Lambert DK, Henry E, et al. Severe thrombocytopenia in theNICU. Pediatrics 2009;124:1095–100.

[25] Kenton AB, Hegemier S, Smith EO, et al. Platelet transfusions ininfants with necrotizing enterocolitis do not lower mortality butmay increase morbidity. J Perinatol 2005;25:173–7.

[26] Christensen RD, Henry E, Wiedmeier SE, et al. Thrombocytopeniaamong extremely low birth weight neonates: data from amultihospital healthcare system. J Perinatol 2006;26:348–53.

[27] Garcia MG, Duenas E, Sola MC, et al. Epidemiologic and outcomestudies of patients who received platelet transfusions in theneonatal intensive care unit. J Perinatol 2001;21:415–20.

[28] Schwartz GJ. Potassium and acid–base. In: Barrat TM, Avner ED,Harmon WE, editors. Pediatric nephrology. Pennsylvania: LippincottWilliams & Wilkins; 1999. p. 155–89.

[29] Greenbaum LA. Acid–base balance. In: Kliegman RM, Behrman RE, JensonHB, editors. Nelson textbook of pediatrics. Philadelphia: Saunders; 2007. p.291–309.

[30] Sterling D, Reithmeier RA, Casey JR. A transport metabolon.Functional interaction of carbonic anhydrase II and chloride/bicarbonate exchangers. J Biol Chem 2001;276:47886–94.

[31] Bastani B, McEnaney S, Yang L, et al. Adaptation of inner medullarycollecting duct vacuolar H-adenosine triphosphatase to chronic acidor alkali loads in the rat. Exp Nephrol 1994;2:171–5.

[32] Bastani B, Gluck S. Adaptational changes in renal vacuolar H(+)-ATPase in the rat remnant kidney. J Am Soc Nephrol 1997;8:868–79.

[33] Olson J, Talekar M, Sachdev M, et al. Potassium changes associatedwith blood transfusion in pediatric patients. Am J Clin Pathol2013;139:800–5.

[34] Hoorn EJ, Geary D, Robb M, et al. Acute hyponatremia related tointravenous fluid administration in hospitalized children: anobservational study. Pediatrics 2004;113:1279–84.

[35] Moritz ML, Ayus JC. The changing pattern of hypernatremia inhospitalized children. Pediatrics 1999;104:435–9.

[36] Olgar S, Yetgin S, Cetin M, et al. Electrolyte abnormalities atdiagnosis of acute lymphocytic leukemia may be a clue for renaldamage in long-term period. J Pediatr Hematol Oncol2005;27:202–6.

[37] Oöpik V, Timpmann S, Hackney AC, et al. Ingestion of sodium citratesuppresses aldosterone level in blood at rest and during exercise.Appl Physiol Nutr Metab 2010;35:278–85.

[38] Ruiz OS, Arruda JA, Talor Z. Na–HCO3 cotransport and Na–Hantiporter in chronic respiratory acidosis and alkalosis. Am JPhysiol 1989;256:414–20.

[39] Preisig PA, Alpern RJ. Basolateral membrane H–OH–HCO3 transportin the proximal tubule. Am J Physiol 1989;256:751–65.

[40] McCarthy PL. Pathophysiology of body fluids. In: Behrman RE,Kliegman RM, Nelson WE, et al., editors. Nelson textbook ofpediatrics. Philadelphia: Saunders; 1992. p. 171–214.

[41] Murray RK. Muscle & the cytoskeleton. In: Murray RK, Bender DA,Botham KM, et al., editors. Harper’s illustratedbiochemistry. China: McGraw-Hill; 2009. p. 45–565.

[42] Junqueira LC, Carneiro Jose. Muscle tissue. In: Junqueira LC, CarneiroJose, editors. Basic histology: text & atlas. China: McGraw-Hill; 2005.p. 183–204.


http://refhub.elsevier.com/S1473-0502(14)00070-6/h0005










































http://www.who.int/child_adolescent_health/documents/fch_cah_06_1/en/index.html

http://www.who.int/child_adolescent_health/documents/fch_cah_06_1/en/index.html
















































































Documents

Citrate metabolism and its complications in non-massive blood transfusions: Association with decompensated metabolic alkalosis+respiratory acidosis and serum electrolyte levels