lable at ScienceDirect

e-SPEN Journal 9 (2014) e223ee227

Contents lists avai

e-SPEN Journal

journal homepage: ht tp: / /www.elsevier .com/locate/c lnu

Original article

Barium in intravenous solutions for administration to neonates:Origins and levels of contamination

Denise Bohrer a, *, Vanessa Morschbacher a, Carine Viana b, Paulo Cícero do Nascimento a,Leandro Machado de Carvalho a

a Department of Chemistry, Federal University of Santa Maria (UFSM), Santa Maria, Brazilb Department of Pharmaceutical Sciences, Federal University of Santa Maria (UFSM), Santa Maria, Brazil

a r t i c l e i n f o

Article history:Received 22 January 2014Accepted 25 July 2014

Keywords:Preterm neonatesParenteral nutritionBariumContaminationContainers

* Corresponding author. Tel./fax: þ55 55 3220 887E-mail address: ndenise@quimica.ufsm.br (D. Boh

http://dx.doi.org/10.1016/j.clnme.2014.07.0052212-8263/© 2014 European Society for Clinical Nutr

s u m m a r y

Background & aims: Barium is a glass constituent and a component of plastic additives, and may migratefrom containers and devices into solutions. As barium is a toxic element, all steps involved in thepreparation of intravenous solutions for premature neonates in an intensive care unit were evaluated todetermine to what degree, if any, they contribute to Ba load. Commercial solutions for parenteralnutrition (PN) and medications and the apparatus used for administration were analyzed for their Bacontent. Bags after compounding, medications after their preparation, infusion sets, and syringes werealso evaluated.Materials and methods: Ba concentration was determined by atomic absorption spectrometry.Results: Bags, burettes, syringes, rubber caps, and glass containers yielded Ba in levels ranging from 0.02to 4.38 mg/g. The highest levels among the solutions were found in multivitamins, magnesium sulfate,and calcium gluconate at 262 mg/L, 193 mg/L, and 166 mg/L, respectively. Most medications did not havemeasurable Ba contamination. However, after dilution in syringes, all of them became contaminated, andthe highest level was reached in dexamethasone samples at 1333 mg/L Ba. Compounded PN bags (n ¼ 15)had a mean of 54.6 mg/L Ba. The content of the same bags after percolating the burette had a mean levelof 94.2 mg/L Ba.Conclusion: Barium is leached from container materials into solutions parenterally administered topreterm babies. The handling processes of compounding and delivering nutrition solutions and medi-cines increased the Ba intake by almost 3 fold in relation to its levels in the starting products.

© 2014 European Society for Clinical Nutrition and Metabolism. Published by Elsevier Ltd. All rightsreserved.

1. Introduction

Barium in the form of oxide is a glass constituent, and it is alsopresent in plastics and rubbers in the form of soaps as stabilizers[2]. These materials are widely used in the pharmaceutical industryfor storing solutions and drugs. Nevertheless, official concentrationlimits have not been established for Ba in packaging materials,except for polyvinyl chloride (PVC). The British Pharmacopoeialimits the amount of Ba in PVC bags to 5 ppm [3]. A species presentin the container/closure system can go into the product via aleaching process that depends on several factors, of which theformulation composition plays a key role [4,5].

0.rer).

ition and Metabolism. Published b

Barium is not an element essential to the body and is toxic in theform of soluble salts (the LD50 of BaCl2 for humans, orally ingested,is 11.4 mg/kg). Reports of individuals exposed to high levels ofbarium suggest that the cardiovascular, nervous, and gastrointes-tinal systems are targets of barium toxicity. The likely cause of mostof these effects is barium-induced hypokalemia [6]. Gastrointes-tinal disturbances are usually the first symptoms of acute bariumexposure [7e11]. Hypokalemia, hypertension, and abnormalities inheart rhythm frequently occur shortly afterward [12]. About 90% ofbarium is absorbed in the bones [13], where it replaces calcium andis deposited in the form of phosphate or carbonate [14]. Theremainder can be distributed in soft tissues such as the brain andthe cardiovascular system.

Preparation of parenteral nutrition (PN) and intravenousmedication involves multiple product manipulations. Althoughinfusion solutions (NaCl, glucose, and Ringer's solution) are usually

y Elsevier Ltd. All rights reserved.

D. Bohrer et al. / e-SPEN Journal 9 (2014) e223ee227e224

administered as they are furnished (ready-to-use bags), PN solu-tions are individually prepared from commercial products in hos-pital pharmacies according to the necessity of each patient [1].Parenteral nutrition for neonates demands a particular strategy forthe product and the drugs to be administered to the patients. Fig. 1shows a simplified scheme of this process. The hospital pharmacycompounds the bags using commercial products and sends them tothe neonatal intensive care unit (NICU). Drugs for administration toneonates are usually diluted in syringes and delivered to the patientvia a burette, a graduated cylinder device intended to allow controlover the volume of the infusion administered. These syringes areusually stored in the NICU for up to 7 days. During the course of thisprocess, the final product administered to the patient remains incontact with different packaging materials for relatively long pe-riods of time, which may favor an increase in contamination.

PN products contaminated by metals can lead to metal distri-bution in the body and metal deposition in organs, posing severalrisks to the patients. Previous studies have shown significant con-taminations of PN solutions with metals such as barium, germa-nium, and manganese [15], arsenic [16], aluminum [17], chromium[18], and lead [19]. In this context, this study aimed to investigatethe presence of Ba in packaging materials where pharmaceuticalformulations for parenteral use are stored and to determine thelevel of this metal in medicines and PN solutions administered topreterm neonates.

2. Methods

2.1. Chemicals

All solutions were prepared using analytical grade reagents andultra-pure water obtained in a Milli-Q water purification system(Millipore, Bedford, MA, USA). Reference solutions were preparedby suitable dilutions of the stock solutions containing 1000 mg/L ofBa (NIST, USA) in water. Arginine (Merck, Darmstadt, Germany),ornithine (SigmaeAldrich, St. Louis, USA), glutamic acid (Merck),aspartic acid (Aldrich), EDTA (Merck), calcium gluconate (Sigma-eAldrich) and citric acid (Merck) solutions were prepared by suit-able dissolution of their salts.

2.2. Contamination control

To avoid contamination, only plastic (polyethylene) laboratoryware (pipette tips and volumetric flasks) was used. It wasimmersed for at least 48 h in 10% HNO3 in an ethanol (v/v) mixtureand washed withMilli-Q purified water shortly before use. To avoidcontamination from the air, all steps in the sample and reagentpreparations were performed in a Class 100 clean bench.

Fig. 1. Scheme of the handling process of formulations and medication at the hospitalpharmacy and neonatal intensive care unit (NICU) to the patient administration.*Points at which samples were collected for analysis.

2.3. Apparatus

Barium was measured by atomic absorption spectrometry(AAS). For container materials, the measurements were carried outby flame AAS, and for all other samples by graphite furnace AAS (GFAAS). The flame AAS was a novAA 300 atomic absorption spec-trometer from Analytik Jena AG (Jena, Germany). The flame usedwas C2H2/N2O, the gas flowwas 180 L/h, and the burner height was5 mm. For these measurements, KCl (2% (m/v)) was used as anionization suppressant. The graphite furnace atomic absorptionspectrometer was a ZEEnit600 from Analytik Jena AG (Jena, Ger-many) with transverse Zeeman-effect background correction sys-tem and equipped with anMPE 60z auto-sampler. Calcium chloride(1% (m/v), 5 mL) was used as a chemical modifier.

2.4. Samples

Two groups of samples were collected at the University Hospitalof Santa Maria. The first group, collected at the hospital pharmacy,included the commercial formulations for PN and infusion solu-tions used to prepare the patients' bags and the medicinesadministered to the patients. The second group, collected at theneonatal intensive care unit (NICU), included compounded bagswith their respective administration sets (burette and lines) andthe syringes containing the medications delivered to the patients.The commercial formulations and the medications were frombatches identical to those used for compounding the bags andpreparing the syringes. All samples from the first group had nopreviousmanipulation. The second group consisted of the solutionsleft in the bags, burettes, and syringes after their administration tothe patients. The asterisks in Fig. 1 represent process-handlingpoints where samples were collected.

The commercial samples consisted of infusion solutions ofsterile water for injection (SWFI), 0.9% NaCl, and 5% and 10% glucoseand solutions for PN, including 10% amino acids, 20% lipids, 50%glucose, 10% calcium gluconate, 20% NaCl, 10% KCl, 10% and 50%magnesium sulfate, trace elements, and multivitamins. The medi-cines were aminophylline, dexamethasone, dobutamine, dopa-mine, gentamicin, furosemide, midazolam, morphine, andranitidine. Three samples from the same lots were analyzed. So-lutions from 15 compounded bags (used for different patients)were analyzed after collecting the fluid from both the bag itself andthe burette.

Barium in these samples was measured by GF AAS. For themeasurement, 50 mL HNO3 was added to 1 mL of each sample,which was then diluted to 10 mL with water.

2.5. Container evaluation

An experiment was conducted to determine the concentrationof Ba in each container/closure system and set used to store anddeliver the formulations. The analyzed materials were new emptybags (Baxter, S~ao Paulo, Brazil), clear and amber 10-mL ampoules(Schott, S~ao Paulo, Brazil), polyethylene bottles (Fresenius, S~aoPaulo, Brazil), burette sets (pediatric burette sets, Baxter), rubberstoppers from commercial amino acid solutions and lipid emul-sions bottles, and rubber plungers removed from the syringes. Forthe determination of Ba, plastic and rubber samples were cut intosmall pieces, calcined and dissolved following the protocol estab-lished by the British Pharmacopoeia [3]. Glass samples werebroken, crushed and prepared as described previously [20]. Briefly,the glass containers were crushed into fragments approximately1 mm in size and mixed well. One hundred milligrams of the glassfragments were placed in a PTFE vessel with 5 mL of 48% (m/m)hydrofluoric acid and 5 mL of water, and heated in a domestic

Table 2Barium measured in constituents of parenteral nutrition solutions.

Component Brand Ba (mg/L) þ SD*

Water for injection Baxter 3.0 ± 1.0Amino acids 10% Fresenius-Kabi 105.2 ± 20.3KCl 10% Halex Istar n.dNaCl 20% Halex Istar n.dLipids 20% Fresenius-Kabi 51.9 ± 1.7Glucose 50% Baxter 27.2 ± 10.8Ca gluconate 10% Hypofarma 166.4 ± 22.9Mg sulfate 10% Ariston 193.3 ± 23.7Mg sulfate 50% Hypofarma 129.4 ± 25.0Polyvitamins Farmalab 31.8 ± 19.3Polyvitamins Crist�alia 261.9 ± 13.0Oligoelements Darrow 51.3 ± 4.3

n.d: not detected, *n ¼ 3.

Table 3Barium predicted in the bags originating from the commercial products, measuredin the bags themselves and measured in the bags fluids after passage through theburette set.

Sample Predicted from the measurement Measured Measured in

D. Bohrer et al. / e-SPEN Journal 9 (2014) e223ee227 e225

microwave oven at minimal power (174 W) for 10 min. This pro-cedure was repeated twice. After total dissolution of the sample,the volume was completed to 100 mL with water, and Ba wasmeasured by F AAS.

2.6. Contact test

A cation exchange resin (strongly acidic Hþ-form resin, DOWEX(SigmaeAldrich)) was washed with a 1% (v/v) aqueous nitric acidsolution and water until a neutral pH was attained. Thereafter, theresin was converted into a Ba2þ-form resin by packing it (50 g) in acolumn and percolating 500 mL of a 1 mol/L barium chloride so-lution at a flow rate of 2 mL/min through the column. The resinwasthen washed with water until no precipitate was observed whendrops of sodium sulfate solution were added to the washes. Theresin was removed from the column and dried in an oven at 60 �Cfor 4 h. Solutions of ornithine, arginine, calcium gluconate, citricacid, glutamic acid, aspartic acid, and EDTA were prepared at aconcentration of 0.2 mol/L with the pH adjusted to 7.0 with NaOH(0.1 mol/L) or HCl (0.1 mol/L). These solutions were divided intotwo aliquots of 200 mL; one was placed in contact with the resin(0.1 g), and the other was kept in a previously cleaned plastic flaskwithout resin. A blank test was performed by storing the resin incontact with 200 mL pure water. Aliquots of these solutions werecollected at time intervals of 1, 3, 24, 48, 72, and 96 h, and after 1, 2,3, and 8 months.

3. Results and discussion

Table 1 shows the Ba content of the containers and devicesanalyzed. Although all materials had Ba at some level, the highestamounts were found in the glass samples and in the rubberplungers of the syringes. The PVC bags analyzed showed a mean of0.15 mg/g, a level much higher than that proposed by the BritishPharmacopoeia [4], which limits the concentration of Ba in PVCbags to 5 ppm (0.005mg/g). Barium in glass containers ranged from4.3 to 4.9 mg/g. There are no limits set for Bain glass containers. Infact, barium oxide is a glass constituent, and its amount in Type Iglass (glass used for PNs) may range from 0.1 to 2.0% (1e20 mg/g)[21].

Table 2 shows the Ba found in the commercial products for PN.Except for the potassium chloride and sodium chloride solutions,all products that were tested yielded Ba at some level. Thecontamination ranged from 3.0 to 261.9 mg/L, and the most highlycontaminated products were multivitamins, magnesium sulfateand calcium gluconate.

Table 3 shows the Ba concentration measured in each bag and inits respective burette after delivery to the patients, i.e., 24 h aftercompounding. Table 3 also presents the amount of Ba each bag

Table 1Barium measured in rubber stoppers, containers, and delivery sets.

Container part Material Color Ba(mg/g ± SD*)

Bottle Glass Clear 4.28 ± 0.80Ampoule Glass Clear 4.05 ± 0.20Ampoule Glass Amber 4.94 ± 0.80Stopper Rubber Red 0.04 ± 0.03Stopper Rubber Gray 0.02 ± 0.01Bag PVC Clear 0.15 ± 0.02Bottle PE Clear 0.04 ± 0.02Syringe seal Rubber Black 4.38 ± 0.40Infusion set PVC Clear 0.11 ± 0.02

PE ¼ polyethylene; PVC ¼ polyvinyl chloride; *n ¼ 3.

would contain if only the Ba in the commercial products werepresent. This calculation was performed by taking into consider-ation the volume of each component used to compound the bagand the Ba concentrations in these products (Table 2). The datapresented in Table 3 revealed that there is an increase in the Balevels in the solutions collected from the burette when comparedwith the bag, and in the bag when compared to the sum of thecontributions of the individual products. From the results, one canconclude that compounding and dispensing increased the Ba con-tent of the solutions administered. Containers may be responsiblefor this increase because practically all materials used to manu-facture containers for pharmaceuticals contain Ba either as a con-stituent or as an impurity. Nevertheless, the higher concentration ofBa in the solutions that remained in the burette may have origi-nated from the use of this device for the administration of medi-cation from the syringe.

Table 4 presents the Ba content in medicines before and aftertheir dilution. Most products contained no Ba contamination (anamount below the detection limit of the method). The exceptionswere dexamethasone and gentamicin, whose Ba concentrationsreached 87 and 584 mg/L, respectively. All drugs had Ba at somelevel after being diluted and stored in the syringes, showing thatthe contact with the syringe materials during the storage periodpromoted Ba release. Even in the dexamethasone and gentamicinsolutions, Ba increased after storage. Calcium gluconate, magne-sium sulfate, sodium chloride and potassium chloride were

of individual components in the bag the burette

1 39.1 40.2 62.72 5.4 19.3 23.13 40.0 51.6 55.74 30.9 53.6 78.25 52.4 70.8 74.56 39.4 32.4 48.17 43.9 145.8 151.48 37.2 72.3 130.59 49.3 26.8 336.210 26.3 54.1 85.511 56.3 33.8 38.812 5.4 83.8 129.213 40.7 53.3 58.114 42.0 40.1 47.715 44.0 41.6 67.3Mean 36.8 54.6 92.5

Values given as mg/L.

Table 4Barium measured in medicines delivered from syringes. Barium measured in commercial forms (ampoules) and after dilution and storage (syringes) for administration.

Medicine Concentrationactive ingredient

Ba (ampoules)(mg/L ± SD)

Diluent Conc. after dilution Theoretical Baconcentrationafter dilution(mg/L ± SD)

Measured Ba afterdilution (Syringe)(mg/L ± SD)

No of syringes

Aminophylline 24 mg/mL n.d. A 10 mg/10 mL n.d. 44.3 ± 8.1 6Dobutamine 12.5 mg/mL n.d. B 250 mg/20 mL n.d. 140.4 ± 12.7 3Dopamine 5 mg/mL n.d. B 50 mg/10 mL n.d. 116.1 ± 13.3 6Dexamethasone 2 mg/mL 87.0 ± 0.5 A 20 mg/20 mL 21.5 ± 0.1 1333.0 ± 31.1 6Furosemide 10 mg/mL n.d A 2 mL/20 mL n.d 298.4 ± 2.0 3Gentamicin 40 mg/mL 584.7 ± 10.9 A 40 mg/20 mL 29.2 ± 1.1 158.8 ± 13.4 6Midazolam 5 mg/mL n.d e e n.d. 83.1 ± 4.6 3Morphine 10 mg/mL n.d A 10 mg/10 mL n.d 2.6 ± 0.2 6Ranitidine 25 mg/mL n.d A 50 mg/10 mL n.d 78.1 ± 11.8 3Ca gluconate 10% 166.4 ± 22.9 e e 166.4 ± 22.9 3769.0 ± 33.7 3Mg sulfate 50% 193.3 ± 23.7 e e 193.3 ± 23.7 4659.0 ± 20.0 3KCl 10% n.d e e n.d 568.7 ± 12.2 3NaCl 20% n.d e e n.d 128.4 ± 12.1 3NaCl 0.9% n.d e e n.d 111.3 ± 12.1 3

SD ¼ standard deviation; n.d ¼ not detected (below the limit of detection of the method).A ¼ 0.9% NaCl; B ¼ 5% glucose in 0.9% NaCl.

Fig. 2. Barium removed from resin by contact with solutions 0.02 mM of ornithine,arginine, Ca gluconate, citric acid, glutamic acid, aspartic acid and EDTA, in the timeinterval 1, 3, 24, 48, 72, 96 h, 1, 2, 3 and 8 months.

D. Bohrer et al. / e-SPEN Journal 9 (2014) e223ee227e226

included in Table 4 because they were also stored in syringes forsmall-dose administration. As seen with the drugs, the solutionsbecame contaminated with Ba after contact with the syringecomponents.

To confirm that Ba canmigrate from containers and delivery setsinto the formulations and that solution constituents may influencethis process, a test using a cation exchange resin as a Ba source wasperformed. An earlier study [22] compared glass and cation ex-change resin and demonstrated that resin and glass behave simi-larly as a cation source. The study was carried out with an Al3þ-containing resin and proved that Al can be leached from both glassand resin via the action of solution constituents. It was also shownthat the higher the affinity of the substance for Al, the larger theamount leached.

In this study, the contact test was carried out with a Ba2þ-con-taining resin and was performed with some components of the PNsolutions, namely polar amino acids (acidic and basic) and calciumgluconate, and with the complexing agents citric acid and EDTA,which have a high affinity for Ba. The results showed that the so-lutions that did not remain in contact with the resin did not have Bain any aliquot collected at any time. Fig. 2 shows the amount of Baremoved from the resin during the experiment. Calcium gluconateextracted much more Ba than the other substances, reachingalmost 50 mg/L after 8 months. Although EDTA is a strong com-plexing agent, it reacted like the other species, most likely due tothe pH levels. Because the pH was adjusted to 7, and the formationof EDTA complexes is governed by the pH, the action of EDTA couldhave been repressed by the relatively low pH. Nevertheless, fromthese results, one can infer that the constituents of PN, namely drugcomponents and even pure water, canwithdraw Ba from packagingmaterials.

It should be noted that gentamicin and dexamethasone aresulfate and phosphate salts, respectively, as are magnesium sulfateand calcium gluconate. The anions present in these salts have ahigh affinity for Ba, forming stable compounds. As they had highconcentrations of Ba, even before contact with the syringes and thedelivery sets, the anions contained in these salts were likely able tobind Ba tightly after removing it from any source with which theycame into contact.

4. Conclusion

The presence of barium in containers and devices made of glassand in plastic materials that are used to store and deliver PN

solutions was confirmed. Glass bottles and syringe plungers are theitems containing the highest amounts of Ba. As these materialsremain in direct contact with the formulation solutions, it allowstheir constituents to interact with the Ba in the containers' surfaces,transferring it into the solution. This hypothesis was confirmed bychecking the ability of some constituents of these solutions towithdraw Ba from a cation exchange resin in the form of Ba.Because Ba is a toxic element, this is an undesirable interaction,increasing the risk of metal uptake. Although practically all prod-ucts used to compound PN solutions contained Ba as a contami-nant, contact with the delivery sets increased the contaminationlevel by more than three times. Most medication that did not haveany Ba contamination became contaminated after contact with thesyringe plungers. As this study was conducted with productsadministered to premature neonates, the findings are of greatconcern because in addition to the direct form of administration,premature neonates are more vulnerable to toxicants.

Conflict of interest

The authors have no conflict of interest.

D. Bohrer et al. / e-SPEN Journal 9 (2014) e223ee227 e227

Acknowledgments

The authors are grateful to the nursery staff in the NICU whofacilitated sample collections and to CNPq (Conselho Nacional deDesenvolvimento Científico e Tecnol�ogico, Brazil) for the financialsupport (Process 477258/2010-7).

References

[1] Oliveira SR, Bohrer D, Garcia SC, Nascimento PC, Noremberg S. Aluminumcontent in intravenous solutions for administration to neonates: role ofproduct preparation and administration methods. J Parent Ent Nutr 2010;34:322e8.

[2] Avis KE, Lachman L, Lierbman HA. Pharmaceutical dosage forms: parenteralmedications. New York: Marcel Dekker Incorporation; 1986.

[3] British Pharmacopoeia (BP), Her Majesty's Stationary Office, London, v. I, 2009.[4] Jenke D. Compatibility of pharmaceutical solutions and contact materials:

safety assessments of extractables and leachables for pharmaceutical prod-ucts. Hoboken: Wiley-Interscience; 2009.

[5] Bohrer D, do Nascimento PC, Binotto R, Becker E. Influence of the glass packingon the contamination of pharmaceutical products by aluminium. Part III:interaction container-chemicals during the heating for sterilisation. J TraceElem Med Biol 2003;17(2):107e15.

[6] Ananda S, Shaohua Z, Liang L. Fatal barium chloride poisoning four casesreport and literature review. Am J Forensic Med Pathol 2013;34:115e8.

[7] Payen C, Dellinger A, Pulce C, Cirimele V, Carbonnel V, Kintz P, et al. Intoxi-cation by large amounts of barium nitrate overcome by early massive Ksupplementation and oral administration of magnesium sulphate. Hum ExpToxicol 2011;30:34e7.

[8] Harris CR, Richards M. Acute barium carbonate toxicity. Clin Toxicol 2007;45:614.

[9] Koch M, Appoloni O, Haufroid V, Vincent JL, Lheureux P. Acute bariumintoxication and hemodiafiltration. J Toxicol Clin Toxicol 2003;41:363e7.

[10] Downs JCU, Milling D, Nichols CA. Suicidal ingestion of barium-sulfide-containing shaving powder. Am J Forensic Med Pathol 1995;16:56e61.

[11] Deng JF, Jan IS, Cheng HS. The essential role of a poison center in handling anoutbreak of barium carbonates poisoning. Vet Hum Toxicol 1991;33:173e5.

[12] Phelan DM, Hagley SR, Guerin MD. Is hypokalemia the cause of paralysis inbarium poisoning? Br Med J 1984;289:882.

[13] P�elisser-Alicot AL, L�eonetti G, Champsaur P, Allain P, Mauras Y, Botta A. Fatalpoisoning due to intravasation after oral administration of barium sulfate forcontrast radiography. Forensic Sci Int 1999;106:109e13.

[14] Motta VT. Bioquímica clínica: princípios e interpretaç~oes, metabolismo min-eral e �osseo. 3rd ed. Porto Alegre: M�edica Missau; 2000.

[15] Buchman AL, Neely M, Grossie Jr VB, Truong L, Lykissa E, Ahn C. Organ heavy-metal accumulation during parenteral nutrition is associated with pathologicabnormalities in rats. Nutrition 2001;17:600e6.

[16] Bohrer D, do Nascimento PC, Becker E, de Carvalho LM,, Dessuy M. Arsenicspecies in solutions for parenteral nutrition. J Parent Ent Nutr 2005;29:1e7.

[17] Bohrer D, do Nascimento PC, Binotto R, Becker E, Pomblum S. Contribution ofthe raw material to the aluminum contamination in parenterals. J Parent EntNutr 2002;26:382e8.

[18] Hak EB, Storm MC, Helms RA. Chromium and zinc contamination of parenteralnutrient solution components commonly used in infants and children. Am JHealth-System Pharm 1998;55(2):150e4.

[19] Nascimento PC, Marques MS, Carvalho LM, Schirmer S, Hilgemann M,Bohrer D, et al. Simultaneous determination of cadmium, copper, lead andzinc in amino acid parenteral nutrition solutions by anodic stripping vol-tammetry and sample digestion by UV irradiation. Anal Lett 2006;39:777e90.

[20] Bohrer D, do Nascimento PC, Binotto R, Pomblum SC. Influence of the glasspacking on the contamination of pharmaceutical products by aluminium. PartI: salts, glucose, heparin and albumin. J Trace Elem Med Biol 2001;15:95e101.

[21] Schott. Technical glasses. Physical and technical properties; 2010. Availablefrom: http://www.schott.com/tubing/english/download/schott-tubing_brochure_technical-glasses_english.pdf.

[22] Bohrer D, do Nascimento PC, Martins P, Binotto R. Availability of aluminumfrom glass and an Al form ion exchanger in the presence of complexing agentsand amino acids. Anal Chim Acta 2002;459(2):267e76.