Der Anaesthesist

Leitthema

Anaesthesist 2019 · 68 (Suppl 1):S1–S14https://doi.org/10.1007/s00101-017-0290-8Published online: 10 April 2017© Springer Medizin Verlag GmbH 2017

M. Rehm · N. Hulde · T. Kammerer · A. S. Meidert · K. Hofmann-KieferDepartment of Anaesthesiology, LMU Munich, Munich, Germany

State of the art in fluid andvolume therapyA user-friendly staged concept. Englishversion

Prologue

Water is a friendly element for anyonewho is familiar with it and knows how tohandle it.1Johann Wolfgang von Goethe

Adequate infusion therapy representsa challenge for modern surgical, emer-gency, and intensive care medicine. Vol-ume replacement solutions have beenused for this purpose for decades. Astheir name indicates, these solutionsserve as a volume replacement, which istheir original and also essential indica-tion. As with all medications, volumereplacement products have dose-depen-dent side effects too and should thereforeonly be applied in the presence of thecorresponding indication. In recentyears, several multicenter studies havestimulated an intense and sometimesemotional debate on the benefits andrisks of hitherto existing pathophysiol-ogy-based treatment concepts.

Butwhatistheobjectiveofanadequatevolume therapy?

In healthy individuals, blood volumeremains constant within very tight limits

1 Quote is translatedfromGerman.

Extended and updated version of the articleby Hulde N and Rehm M (2016) Grundlagender Infusionstherapie. In: Refresher Course –AktuellesWissen fürAnästhesisten,No.42. AktivDruck&Verlag,Ebelsbach,pp.73–81

English version of: Rehm M, Hulde N, Kam-merer T, Meidert SA, Hofmann-Kiefer K (2017)Stand der Technik in der Flüssigkeits- undVolumentherapie. Ein anwenderfreundlichesStufenkonzept. Anaesthesist doi:10.1007/s00101-017-0272-x

during the course of a day. The objec-tive must be to maintain the patient’snormovolemia (. Fig. 1) as far as possi-ble. Normovolemia is the prerequisite forensuring that, for example, cardiac out-put and oxygen supply are always suf-ficient and can increase in response tostress at any time. Traditionally, a liberalfluid regime was propagated: “In doubt,give volume!” However, various publica-tions indicate that “restrictive” volumemanagement has advantages over “lib-eral” volume management. This effecthas been impressively demonstrated formajor abdominal surgery: the incidenceof potentially life-threatening complica-tions, such as pulmonary edema, anas-tomotic insufficiency, impaired woundhealing, and coagulation problems couldbe reduced by fluid restriction, and post-operative intestinal motility significantlyincreased [1–6]. However, in this con-text it is important to note that the terms“restrictive” and “liberal” are not stan-dardized, and their use is unfortunatelynot uniform. Therefore, when speakingof volume therapy, “restrictive” is beingincreasingly replaced by the term “needs-based”.

Theprimarygoalofperioperativefluidtherapy is the extracellular space, whichis divided into two functional compart-ments by the intact vascular barrier [2].The intention behind adequate volumetherapy is maintaining normovolemia,extracellularhomeostasisbetween the in-tra- and extravascular spaces (. Fig. 1).In the perioperative period it seems rea-sonable todistinguishbetweenfluid losses(extravascular space)without colloid-os-motically activemolecules (such as urine

and perspiratio) and volume losses (in-travascular space), i. e., fluid with on-cotically active macromolecules (bloodor plasma losses from the intravascu-lar space). These different entities aretreated by the differentiated applicationof infusion solutions, crystalloids, andcolloids [8]. The way this can be suc-cessfully achieved is described below bya fundamental presentation of therapeu-tic options and their significance in termsof the risk–benefit ratio and by consider-ation of a practical case that could be en-countered in daily practice. The conceptof loss-adapted fluid and volume therapyis basedon thepatient’s preoperative nor-movolemia. This is normally present inpatients without internal medicine com-plaints, despite fasting [9]. As men-tioned above, under normal conditions,the blood volume of a healthy individ-ual remainsphysiologically very constantover the course of the entire day; it variesby only a few deciliter. Drinking is by nomeans theonlyregulatory factor: theper-manentfilteringandreabsorptionoffluidoutof and fromthe interstitial space playsamajor role. Adetaileddescriptionof theperioperative changes in hemodynamicsinduced by volume deficit or expansionwould exceed the scope of this article;thus, this is intentionally not presented.However, in an illustrative case exampleof a situation that can arise any day in theoperating roomthe effect ofhypovolemiaon hemodynamics and organ perfusionis exemplarily portrayed.

Der Anaesthesist · Suppl 1 · 2019 S1

Leitthema

Hypovolemia:– O2 mismatch– Hypotension– Organ dysfunc�on– Lac�c acidosis

Hypervolemia:– ANP release– Glycocalyx damage– Edema forma�on– Coagulopathy

Outc

ome

Periopera�ve fluid balance

Correla�on betweenoutcome and fluid balance

Hospital stayComplica�onsMortality

Fig. 18 Correlationbetweenfluidbalanceandperioperativeoutcome.ANPatrial natriuretic peptide.(Modified fromGlassfordNJ et al. [7])

The foundation: efficiency,effectiveness, and evidence

Careful evaluationof efficiency, effective-ness, and evidence is essential for suc-cessful application of a particular med-ical treatment. To enable the reader tomake his/her own best possible evalua-tion of the current data on fluid therapy,including understanding a viable infu-sion strategy, the genuine meanings ofthe terms efficiency, effectiveness, andevidence are explained in the following:Efficiency is the effort, in relation to com-pleteness and accuracy, that is requiredto attain a particular objective. In con-trast, effectiveness describes the degree towhich the objective is attained (efficacyand quality). Toenable these twoterms tobe understood more easily, the followingpractical examplemaybeconsidered: Af-ter sport and sweating correspondingly,a fluid deficit of 500ml can be efficientlyand effectively treated by drinking water.Leaving aside cost considerations, drink-ing 500ml of prosecco or even whiskeywould be just as efficient in this instance.In terms of treatment of the fluid deficitalone, all three “therapeutic” regimenscan be considered equally efficient andeffective; however, drinking prosecco orwhisky is associatedwith additional “sideeffects.” The effectiveness of a treatmentmust, thus, always be seen broader thanasanachievementof the therapeutic goal.Efficiency is important, but if the wrongthings are efficiently accomplished, thenthis constitutes a waste of resources and

a risk for the patient. Strategy (effective-ness) thus comes before efficiency in fluidtherapy, too. Completing the circle, thequestion of the safety of a therapy andthe corresponding evidence demonstrat-ing that the benefits outweigh the risksalways arises, too. Evidence – and thisis very important – means “proof.” Evi-dence in medicine refers to the empiricalverification of the benefit of a diagnos-tic or therapeutic action. This forms thebasis of evidence-based medicine.

Evidence inmedicinemustnotbecon-fused with the German word “Evidenz”whichmeans obviousness/no proof is re-quired and therefore just describes thecognitive comprehensibility of a possiblerelationship. This is illustrated by the fol-lowing simple example: Two potentiallynephrotoxic infusion solutions are testedby comparing them with one another intwo study arms. No differences are foundin renal function or renal replacementtherapy, so there is nothing to obviouslyimplyakidney-damagingeffectofthetwosubstances. However, this of course doesnotprove the safety (nodamaging side ef-fects in terms of renal function) of eitherof the solutions. In order to produce “evi-dence,” a valid control group is necessary,as are appropriate function tests, obser-vationperiods, and patient/control num-bers (“power”). This of course representsa significant challenge, and not all ques-tions can be answered with the help ofevidence-based medicine particularly inview of limited resources. The CochraneCollaboration has set itself the task of

describing evidence-basedmedicine andworks to generate, update, and dissemi-nate systematic review articles. The anal-yses arebasedonthe “CochraneDatabaseof Systematic Reviews,” which is renewedquarterly as part of the Cochrane Li-brary. The Cochrane Collaboration de-fined the types of Cochrane evidencelisted in . Table 1. Systematic reviewsbased on data from randomized con-trolled trials (RCTs) of sufficient size andpower represent the highest level of ev-idence. The opinions and recommen-dations of reputable authorities and ex-pert panels represent the lowest level ofevidence according to this list, and canneither replace nor refute the results andfindings of large RCTs [10]. Recommen-dations based on low-level evidence fre-quently find their way into guidelines,which may also be related to the ex-perts’ conflicts of interests. Althoughthese must be disclosed, the declarationdoes not balance out or even negate theconflicts of interests:

Standards for the evaluation and man-agement of conflicts of interests in guide-lines are lacking and should be urgentlydeveloped2 [11].

Therefore, it can not be excluded thatguidelines are, to some extent, influ-enced by the evaluations or even wishesof pharmaceutical companies.

There are two fundamental principlesfor the use of evidence-based medicine:1. “Absence of evidence is not evi-

dence of absence!” This means: Ifthere is no evidence for a therapy,this still is not a license to pro-ceed at will. Consider the followingcurrently relevant example of livertransplant and alcohol consumption,which also has to do with fluids:In light of the fact that there is noevidence demonstrating the benefitof abstinence from alcohol prior totransplantation, its importance iscurrently the subject of debate. “Therequired 6-month period of abstinencefrom alcohol excludes patients froma potentially life-saving treatment. In

2 Quote is translatedfromGerman.

S2 Der Anaesthesist · Suppl 1 · 2019

light of the far-reaching consequencesof this regulation, special demandsare to be placed on the reliability ofthe underlying evidence: The datashould demonstrate clearly that thesurvival times of relapsed patients aresignificantly reduced by the renewedconsumption of alcohol” [12]. Theconcept of evidence-based medicineis misunderstood here: It is wellknown that alcohol (even in smallquantities) can have hepatotoxic ef-fects. Clinical experience has shownthat in a nonnegligible number of pa-tients a rapid improvement in organfunction due to alcohol abstinencebecause of their potential listing fora potential liver transplant rendersthe transplant no longer necessary.Furthermore, it cannot be doubtedthat alcohol will continue to have anhepatotoxic effect after the transplan-tation – a transplanted liver is not“alcohol resistant.” Purely becauseno “evidence” is available, i. e., thereare no sufficiently large controlledstudies clearly proving these correla-tions in this special population, theuniversally valid pathophysiology or“common knowledge” must not beignored. This is also true for infusiontherapy, e. g., gelatin: There is a lackof large controlled trials with thecorresponding long-term follow-up to prove that gelatin is not safe(see also below “Evidence in volumetherapy: Gelatin”). However, the lackof these data of course does not provethe safety of this infusion solution.

2. “Presence of evidence is evidentlypresent!” For example, if it has beenshown in controlled studies, i. e.,blinded prospective randomizedtrials, that the adverse effects ofa treatment outweigh the benefits ina particular patient collective, thisresult cannot be “topped” or refutedby studies of a lower-level evidencetype, such as register or observationalstudies. Available evidence cannotbe negated in this manner, evenif certain representatives of profit-oriented pharmaceutical companieswould understandably wish to see itotherwise.

Abstract · Zusammenfassung

Anaesthesist 2019 · 68 (Suppl 1):S1–S14 https://doi.org/10.1007/s00101-017-0290-8© Springer Medizin Verlag GmbH 2017

M. Rehm · N. Hulde · T. Kammerer · A. S. Meidert · K. Hofmann-Kiefer

State of the art in fluid and volume therapy. A user-friendly stagedconcept. English version

AbstractAdequate intraoperative infusion therapy isessential for the perioperative outcome ofa patient. Both hypo- and hypervolemia canlead to an increased rate of perioperativecomplications and to a worse outcome.Perioperative infusion therapy shouldtherefore be needs-based. The primaryobjective is the maintenance of preoperativenormovolemia using a rational infusionstrategy. Perioperative fluid losses shouldbe differentiated from volume losses dueto surgical bleeding or protein lossesinto the interstitial space. Fluid loss viaurine excretion or insensible perspiration(0.5–1.0 ml/kg/h) should be replaced withbalanced, isooncotic, crystalloid infusionsolutions in a ratio of 1:1. Volume therapystage 1: intraoperative volume losses up toa blood loss corresponding to 20% of thepatient’s total blood volume are compensated

for by balanced crystalloids in a ratio of4–5:1. Stage 2: blood losses exceedingthis level are to be treated with isooncoticcolloids (preferably balanced) in a 1:1 ratio.In this regard taking into considerationthe contraindications, e. g., sepsis, burns,critical illness (usually patients in theintensive care unit), impaired renal functionor renal replacement therapy, intracranialhemorrhage, or severe coagulopathy, artificialcolloids such as hydroxyethyl starch (HES)can be used perioperatively for volumereplacement. Stage 3: if an allogeneic bloodtransfusion is indicated, blood and bloodproducts are applied in a differentiatedmanner.

KeywordsInfusions · Colloid · Hydroxyethyl starch ·Crystalloid · Perioperative volume balance

Stand der Wissenschaft in der Flüssigkeits- und Volumentherapie.Ein anwenderfreundliches Stufenkonzept

ZusammenfassungEine adäquate intraoperative Infusionsthe-rapie ist wesentlich für das perioperativeOutcome eines Patienten. Sowohl Hypo-als auch Hypervolämie können zu einererhöhten Rate perioperativer Komplikationenführen und somit zu einem schlechterenBehandlungsergebnis. Daher sollte dieperioperative Flüssigkeitstherapie bedarfs-gerecht und zielorientiert durchgeführtwerden. Das Hauptziel ist die präoperativeNormovolämie durch eine rationaleInfusionstherapie aufrechtzuerhalten.Perioperative Flüssigkeitsverluste solltendabei von Volumenverlusten durch chirur-gische Blutungen oder Proteinverlusten insInterstitiumdifferenziert werden. Flüssigkeits-verluste via Urinexkretion oder Perspiratioinsensibilis (0,5–1,0 ml/kg/h) sollten 1:1 mitbalancierten, isoonkotischen, kristalloidenInfusionslösungen ausgeglichen werden.Volumentherapie, Stufe 1: IntraoperativeVolumenverluste bis zu einem Blutverlust von

20% des Gesamtblutvolumens werden mitbalancierten Kristalloiden im Verhältnis 4–5:1ausgeglichen. Stufe 2: Darüber hinausge-hende Blutverluste sind im Verhältnis 1:1mit isoonkotischen Kolloiden (bevorzugtbalanciert) zu behandeln. In dieser Hinsichtund unter Beachtung der Kontraindikationenwie Sepsis, Verbrennungen, kritischeErkrankung (i. d. R. Patienten auf Intensivsta-tion), eingeschränkte Nierenfunktion oderNierenersatztherapie, intrakranielle Blutungoder schwere Gerinnungsstörungen könnenperioperativ auch künstliche Kolloide, wie z. B.HES, zum Volumenersatz verwendet werden.Stufe 3: Wenn die Indikation zur Gabe vonFremdblut besteht, erfolgt ein differenzierterEinsatz von Blut und Blutprodukten.

SchlüsselwörterInfusionen · Kolloid · Hydroxyethylstärke ·Kristalloid · Perioperative Volumenbilanz

Der Anaesthesist · Suppl 1 · 2019 S3

Leitthema

Table 1 Types of evidence according to Cochrane

Level Type of evidence

Ia At least one systematic review based on methodologically high-quality randomizedcontrolled trials (RCTs)

Ib At least one sufficiently large, methodologically high-quality RCT

IIa At least one high-quality non randomized study

IIb At least one high-quality study of another type of quasi-experimental studies

III More than one methodologically high-quality non experimental study

IV Opinions and recommendations of respected authorities (from clinical experience);expert commissions; descriptive studies

Table 2 Directlymeasured volume effects of volume replacement solutions

Infusion solution Volume effect in% Reference

Human albumin 5% 85 Rehm et al. [16]

Human albumin 20% 185 Jacob et al. [21]

HES 6% 200/0.5 90 Rehm et al. [16]

HES 6% 130/0.4 98 Jacob et al. [18]

Ringer’s lactate 18 Jacob et al. [21]

Since in clinical routine there are some-times situations in which clinicians occa-sionally do not know what to do, volumetherapy is sometimes not performed ac-cording to the findings of the large stud-ies and metaanalyses, but rather empir-ically, based on “gut feeling.” However,which sustained findings concerning in-fusion therapy can be provided by phys-iology and evidence-based medicine?

Physiology and effectiveness ininfusion therapy

Fluid therapy comprises administrationof crystalloids and colloids. Crystalloidsare solutions of water and electrolytes;colloids also contain substances withwater-binding properties, e. g., albumin,hydroxyethyl starch (HES), or gelatin.The type and composition of the solu-tion to be infused should be adapted tothe fluid compartment that is the targetof therapy. This is the reason for thedevelopment of, e. g., isotonic infusionsolutions. As already mentioned, it mustbe discriminated between fluid losseswith and without loss of oncoticallyactive macromolecules. From this re-sults the differentiation of volume lossesout of the intravascular and intersti-tial spaces (loss of oncotically activemacromolecules, e. g., albumin) fromfluid losses such as urine, insensibleperspiration (quantitative 0.5ml/kg/h,

with maximal eventration 1ml/kg/h),or preoperative dehydration (no loss ofoncotically active macromolecules, butpure fluid deficit). With each episodeof blood loss, patients lose a sometimesconsiderable amount of protein – de-pending on the amount of blood loss.Therefore, if blood loss is compensatedfor to a large extent using pure crystallinesolutions, the colloid osmotic pressurein the blood vessel system will dropsignificantly. Moreover, during majortumor surgeries, considerable quanti-ties of macromolecules – protein andHES – are lost into the interstitial space,the so-called protein [13] or HES shift[14–16]. These are losses of protein orHES that can not be explained by surgi-cal bleeding or urine. For example, inovarian cancer patients, in addition tothe protein loss resulting from surgicalbleeding (1700ml, 67 g protein loss),a mean intraoperative protein shift of49 g occurred [13]. In patients withradical hysterectomy, it was shown that30% of the applied HES molecules leftthe vascular bed (mean loss via surgicalbleeding: 18 g, mean excretion of HES inurine: 18 g, mean HES shift: 30 g) [16].Both HES or protein loss via surgicalbleeding and the shift of protein or HESinto the interstitial space reduces theintravascular colloid osmotic pressure,thus, exacerbating edema formation dueto extravasation of fluid.

In order to develop a viable strategyto replace all of these fluid and colloidosmotic pressure losses in a demand-ori-ented manner, it is important to knowhow effective the various available infu-sion solutions are.

The “volume effect” of infusion solu-tions refers to the proportion of the infu-sion solution that remains in the vascularsystem after intravenous application. Ifthis is 100%, the entire infusion remainsin the intravascular space; if the valueis 20%, this indicates a high degree ofextravasation (80%), with the danger ofinterstitial edema. If the volume effect is180%, this means that the infusion solu-tiondrawsfluid fromthe interstitial spaceback into the vascular system due to thehigh intrinsic colloid osmotic pressure.Many data are available on this effect andevidenceof this effect is listed in.Table2.The most reliable method for investigat-ing the actual volume effect has provedto be direct measurement of blood vol-ume before and after the application ofan infusion solution (. Table 2; [8, 14,16–20]). The crucial pharmacodynamicproperty of the colloids is their relativelylarge volume effect [8]. These are repre-sented in clinical routine by human albu-min(HA),HES, andgelatinpreparations.The application of colloids for the treat-ment of volume losses can thus be con-sidered very effective. As can be seen in. Table 2, the volume effects of isoon-cotic colloids applied to replace bloodlosses (acute normovolemic hemodilu-tion, ANH) are about five times thoseof crystalloids. This means that in or-der to approximately maintain normo-volemia, 4–5 l of crystalloid have to beinfused to replace a blood loss of 1 l.Therefore, the application of crystalloidsalone is not only not very effective, it alsoleads to a pronounced positive fluid bal-ance (interstitial edema), particularly incase of large blood losses. This, in turn,has an unfavorable impact on treatmentoutcome. In contrast, a blood loss of1 liter can be adequately treated usingonly 1 l of isooncotic HES or albumin.The volume effect of colloids is, however,crucially dependent on the integrity ofthe vascular barrier.

S4 Der Anaesthesist · Suppl 1 · 2019

Erythrocyte

Glycocalyx

Thrombocyte

HES

Albumin

vWF AntithrombinICAM/VCAM

HES shiftProtein shift

a

b

Fig. 29 a Struc-ture of the intact en-dothelial glycoca-lyx. The integrationand concentrationof plasma proteinsgenerates a highcolloid – osmoticpressure in the gly-cocalyx, which re-tains fluid in thevascular system.b Pathophysiologicsituation follow-ing degradationof the endothelialglycocalyx. Releaseof the plasma pro-teoglycans leads toadhesion of bloodcells and efflux ofmacromoleculesfrom the vascularsystem. vWF vonWillebrand Factor,ICAM intercellularadhesionmolecule,VCAM vascular ad-hesionmolecule,HEShydroxyethylstarch. (Modifiedfrom [29])

The vascular barrier

Thevascular barrier separating the bloodvessels from the interstitial space is a dou-ble barrier. It is composed of endothe-lial cells and the endothelial glycoca-lyx, a layer of proteoglycans and gly-cosaminoglycans at the luminal side ofthe endothelial cells (. Fig. 2a). Due toits charge properties, this layer binds al-bumin and concentrates it, such that theglycocalyx develops its own colloid os-motic competence. This layer is respon-sible for the maintenance of the Starlingcolloid osmotic pressure gradient. Di-rectly beneath the glycocalyx, the col-loid – osmotic pressure is very low. Inthe interstitial space itself, it once againcorresponds approximately to the colloid– osmotic pressure of the intravascularspace [22, 23]. Provided the endothelialglycocalyx is intact, the volume effects ofcolloidsarevery large, asdescribedabove.However, if the glycocalyx has been de-stroyed (. Fig. 2b), as can occur, e. g., insepsis, then the volume effects of col-loids are greatly reduced. It is likely that

the effectiveness of isooncotic colloids inthis situation is no longer five times, butrather only about twice that of crystal-loid solutions. An inappropriate infusiontherapy can also be the cause of a leakyvascular barrier. Volume loading is asso-ciated with small volume effects, but alsowith adverse events [24]. This owes tothe fact that volume loading (infusion inthe absence of corresponding fluid loss)causes expansion of the atria and thus re-lease of atrial natriuretic peptide (ANP)[25–27]. This peptide alters the endothe-lial glycocalyx, thus, increasing extrava-sation of macromolecules (colloids) andfluid. The release of glycocalyx compo-nents into the circulation can also impaircoagulation due to the heparinoid effectof the released heparan sulfate. This canincrease the postoperative risk of throm-bosis as the freely circulating glycocalyxcomponents are excreted via the kidneys.Afterwards, plasmatic coagulation is nolonger impaired by these components;however, continuously uncovered by theglycocalyx endothelial cells promote ad-hesionof thrombocytes. Oncedestroyed,

it likely takes 7–10days for the glycocalyxto be completely rebuilt. Volume loadingis thus not recommended under normo-volemic conditions. In the absence of anintact vascular barrier (e. g., in sepsis),extravasation, particularly extravasationof synthetic colloids, should be kept ata minimum (to avoid undesired deposi-tion of colloids in tissue) [28]. Based onpathophysiological considerations, albu-min should be preferred to the artificialcolloids HES or gelatin in this situation.However, this raises the following ques-tion: Is the application of infusion solu-tions, particularly colloids, safe?

Evidence in fluid therapy

Crystalloids

Perioperatively and in the intensivecare units, the use of balanced iso-tonic crystalline solutions is increasing.This is because physiologic saline so-lution causes hyperchloremic acidosisin a dose-dependent manner [30]. Incomparison to balanced solutions, thisreduces splanchnic bed and renal per-fusion, reduces urine excretion, andincreases the incidence of nausea andvomiting [31]. There are also indicationsthat balanced solutions are associatedwith advantages in terms of patient sur-vival [32]. But in contrast, a recentlypublished study found no survival ben-efit of balanced solution compared tosaline solution [33]; however, the appliedamount of infusion was so small in thisstudy (<2 l applied slowly during thecourse of several days) that neither thechloride concentration nor the pH inthe saline group could increase aboveor drop below the respective normalvalues. At ISICEM 2016, Kellum pre-sented results from a very large cohortof patients demonstrating that largeramounts of saline solution were associ-ated with increased mortality comparedto balanced solutions. These results werepublished as the current article was inpress [34]. Contrary to widespread be-lief, the potassium content of balancedsolutions does not cause hyperkalemiain renal insufficiency patients [35, 36].In fact, it has been shown for kidneytransplantation that even in the anuric

Der Anaesthesist · Suppl 1 · 2019 S5

Leitthema

phase a potassium-containing (up to5mmol/l) balanced solution is superiorto 0.9% NaCl [35]. This is because thehyperchloremic acidosis caused by salinesolution promotes potassium drift fromthe intracellular space into the intravas-cular space, which favors hyperkalemia.Even infused in large amounts, a bal-anced infusion solutionwith a potassiumcontent of 4–5mmol/l cannot cause hy-perkalemia, since it is not hyperkalemic.Only if the patient has a pronouncedpure water diuresis could balanced solu-tions theoretically lead to hyperkalemia(concentration effect due towater loss viathe kidneys). In summary, it thereforeappears rational to use balanced infusionsolutions in intensive care medicine andin the OR, and to apply isotonic salinesolutions only in special situations, e. g.,cholera (substained losses of chlorideand sodium).

Albumin

Albuminisa585-aminoacidproteinwitha molecular weight of 69 kDa. Albu-min-containinginfusionsolutionsareex-tracted from human blood (and thus re-quire a batch manufacturing record), areconsidered as virus safe, and can be usedto treat hypoalbuminemia. Although al-bumin level is a frequentlymeasuredclin-ical parameter, no generally valid min-imum limit can at present be definedbelow which replacement should takeplace. It has been repeatedly demon-strated that pronounced hypoalbumine-mia in critically ill patients is associatedwith poor outcome [37]. Whether or nothypoalbuminemia was indeed the causeof the poor outcome or rather symp-tomatic of the severity of each particu-lar disease (e. g., nutritional status, de-gree of inflammation) is still controver-sially discussed. However, in cases of se-vere hypoalbuminemia (<2.5 g/dl), thereare many indications that albumin ad-ministration seems to be beneficial interms of mortality, organ function, theamount of fluid to be applied (fluid bal-ance), and tolerance of enteral nutrition[37–41]. The endothelial glycocalyx re-quires a low but nevertheless basal albu-min concentration for maintenance of itsbarrier function. For a long time, albu-

min was considered the virtually idealvolume substitute, due to its colloid os-motic properties and long intravascularretention time. The volume effect of20% albumin solution is 185%, whichmeans that this infusion solution can re-cruit fluid from the interstitial space forthe intravascular space. The volume ef-fect is higher than the entire volume ofthe infused solution. Albumin 5% hasa volume effect of 85%, meaning that al-most all of the infused fluid remains inthe vascular system. Albumin was thusused for decades as the reference sub-stance during development of artificialcolloidal plasma replacement products.Since the development of cheaper artifi-cial colloids, the application of albuminas a volume replacement product has de-creasedconsiderably. AnolderCochraneanalysis demonstrating increased mor-tality among critically ill patients receiv-ing albumin contributed significantly tothis trend [42]. Volumereplacementwithalbumin was subsequently investigatedin several large-scale prospective trials.The Australian SAFE trial investigatedthe safety of infusion therapy with hu-man albumin in intensive care patients[43, 44] and was unable to show any se-vere adverse effects compared to 0.9%saline infusion; nor, however, was theoutcome of the albumin-treated patientsbetter [43, 44]. The ALBIOS trial evalu-ated a patient collectivewith severe sepsisor septic shock. The experimental groupreceived albumin replacement (not forvolume therapy, but rather to increase theserum albumin concentration to≥3 g/dl)and the control group received only crys-talloid solution [45]. There was no dif-ference in terms of mortality betweenthe two groups [45]. In the similarlyconceived EARSS trial, a reduction inabsolute and relative mortality of 2.2%and 8.4%, respectively, was observed inthe human albumin group compared tothe crystalloid group after 28 days (p >0.05) [46]. Wiedermann’s meta-analy-sis [47] of pooled data from the SAFE,EARSS, and ALBIOS trials reported thatalbumin application in patients with se-vere sepsis or septic shock was associ-ated with significantly reduced mortality(p = 0.046). An almost simultaneouslypublishedmeta-analysis did not arrive at

this result (p = 0.08); however, this studyalso included patients from two othervery small studies, which were not de-signed and standardized for addressingthe endpoint “outcome” (unfortunately,this is a frequent shortcoming of meta-analyses). Additionally, this studydidnotuse the correct patient numbers from theoriginal ALBIOS study publication [48].In contrast to hyperoncotic HES solu-tion, no negative impact of albumin onrenal function could be shown [49].

Whether the positive effects of infu-sions are realized or are outweighed bytheir adverse effects, however, always de-pends on the applied dose and a correctlyestablished indication. It was shown thatmortality in children with a severe in-fection significantly increased followingapplicationofa largefluidbolus (albuminor isotonic saline solution; 20–40ml/kg).However, in comparison over 90% of thechildren had no clinical signs of hypo-volemia [50]. Application of high dosesof infusionsolutions intheabsenceofvol-ume requirements must be viewed crit-ically, irrespective of the type of solu-tion. However, use of human albumin asvolume replacement for stabilization ofhemodynamics is nowadays widely ac-cepted.

In summary, the base of high-qualitydata (level of evidence Ia) confirms theeffectiveness, efficiency, and safety of hu-man albumin. These data are not onlyavailable from sufficiently large RCTs,but also from significant meta-analysesof these RCTs (. Table 1). Comparedto synthetic colloids, human albumin isa very expensive treatment option [51].For this reason, albumin should not beusedwhere other therapeutic alternatives(e. g., crystalline solutions or syntheticcolloids) are possible. Twopost hoc anal-yses demonstrated reduced mortality insevere sepsis (SAFE) and septic shock(ALBIOS) achieved by albumin appli-cation. It would be wrong to concludethat this demonstrates that albumin is aneffective treatment for sepsis. Adminis-tration of antibiotics as early as possibleand a causal treatment (e. g., curing ofinfection) is undoubtedly far more effec-tive. However, many available data in-dicate that hypoalbuminemia (<2.5 g/dl)in sepsis patients should be treated early

S6 Der Anaesthesist · Suppl 1 · 2019

by appropriate administration of albu-min [43, 47, 52, 53]. Therapeutic targetshould be an albumin concentration ofover 2.5 g/dl, which is most effectivelyachieved with a hyperoncotic albuminsolution (albumin 20%).

Hydroxyethyl starch (HES)

HES solutions are synthesized fromplant-derivedstarch(waxycornorpotatostarch). A proportion of the cyclic glu-cose rings in the starch is substitutedwith hydroxyethyl groups. For example,the indication 130,000/0.4 means thatthe starch, with a molecular weight of130,000 Da, carries hydroxyethyl groupson 40% of the glucose rings. The volumeeffect of HES 130,000 is almost 100%and can thus be described as ideal interms of volume replacement. Althoughvarious smaller studies in different pa-tient cohorts found no negative effectsdue to the application of HES as partof a combined HES and electrolyte so-lution therapy, studies in intensive caremedicine have revealed a different pic-ture. The VISEP trial in patients withsevere sepsis found a negative effect onrenal function, which was frequentlyobserved only at later time points (ob-servation time was 90 days) [54]. Inthis study, an older HES preparation wasused in a hyperoncotic form (10%) andadministered to a high cumulative dose.The blinded randomized prospective6S trial [55] compared treatment with6% HES 130,000/0.42 with Ringer’s ac-etate application in sepsis patients. Thisstudy also revealed an increased rateof renal replacement therapy and, fur-thermore, increased mortality followingHES therapy. Another blinded random-ized prospective study, the CHEST trial,investigated a large mixed cohort of pa-tients requiring postoperative intensivecare, but of whom the majority were notseptic patients [56]. This study showedan increased rate of renal replacementtherapy following HES treatment (p =0.04) and no difference in mortalityafter 90 days, although circulatory sta-bilization was achieved more quicklyusing HES volume replacement. Alsothe CRYSTMAS trial [57] investigateda sepsis patient cohort and demonstrated

that HES 130,000/0.4-based volume re-placement therapy resulted inmore rapidhemodynamic stabilization compared to0.9% saline solution. This trial found nosignificant difference between the groupsin terms of renal function and mortality,which may be due to the comparativelysmall number of patients (n = 174).

Meta-analyses containing these inten-sivecaremedicineRCTs(levelofevidenceIa) demonstrate that in critically ill pa-tients, particularlyinthosesufferingfromsepsis or septic shock, application ofHESis associated with an increased incidenceof renal failure and a higher rate of re-nal replacement therapies [58–60]. Sub-sequently, on the basis of these results,in October 2013, the PharmacovigilanceRisk Assessment Committee (PRAC) ofthe European Medicines Agency issuedthe recommendation that HES shouldno longer be used in critically ill patientsas well as those suffering from sepsis orburns. Additionally, it was also recom-mended not to continue HES treatmentfor more than 24 h and to monitor renalfunction for 90 days [61, 62].

However, thecontroversialdebatesur-rounding the safety of HES was not si-lenced by these measures. On the con-trary, the available studies were criticizedfor their lackofcomparability (patientco-hort, patients’ status at randomization,use of volume products, maximum HESdose). Meybohm et al. [63] reevaluatedthe prospective randomized studies anddefined a six-point catalog of criteria thatshould be fulfilled by future studies in or-der to be able to be used for comparisonof different infusion regimes:4 short time interval between shock

event and randomization,4 limited use of initial fluid therapy,4 consistent use of a fixed infusion

algorithm until hemodynamic stabi-lization,

4 reproducible criteria for hypo-volemia,

4 maximum doses of HES,4 exclusion of patients with preexisting

impaired renal function or renalreplacement therapy.

The studies mentioned above only par-tially fulfill these criteria [63]. So far,controlled trials fulfillingall criteriamen-

tioned above, with high case numbersand long observation period of at least90 days, are unfortunately not available.Whether HES 130,000/0.4 has a morefavorable safety profile in sepsis patientsand is thus associatedwith a lower rate ofrenal function-related adverse events inlong-term follow-up compared to olderstarch preparations is currently unclear[58]. Therefore, no HES preparationsat all should be used in this special pa-tient collective (see also section“Practicalconclusion – the synopsis”). By way ofprecaution, the German Society of Anes-thesiology and Intensive Care Medicine(DGAI) recommends that the applica-tion of modern low molecular HES so-lutions in intensive care patients shouldbe evaluated critically on an individualbasis and that their use should be re-stricted to patients with acute life-threat-ening blood and volume losses not man-ageable by other means. This means, forexample, that a patient having undergoneaortic stent grafting, who is postopera-tively only in the intensive care unit formonitoring purposes and suddenly suf-fers massive hypovolemic shock due toaorticrupture, canofcoursereceive largeramounts of HES as an already initiallylifesaving treatment, even though the sit-uation takes place in the intensive careunit.

Do these considerations also applyto the use of HES preparations asa volume replacement solution inthe OR?After the regulatory authorities initiallysuspended the entire approval for HESpreparations, it was also no longer per-mitted to use these during surgery fora while. In two industry-supportedmeta-analyses, the safety of cornstarch130,000/0.4 was investigated in a largenumber of surgical patients. The au-thors found no significant difference tothe control group and attested to thesafety of this cornstarch. However, themajority of the included studies hadinsufficiently long observation periodsand no real control group. A furtherstudy without industry funding cameto the same conclusion with the sameshortcomings of the included studies[64]. To prove the safety of HES in

Der Anaesthesist · Suppl 1 · 2019 S7

Leitthema

the OR – analogous to intensive caremedicine – correspondingly conceived,randomized, controlled, and blindedtrials are also required. Results fromstudies of this type are not yet available[65].

Where evidence is insufficient, patho-physiologic considerations will help:The potential nephrotoxic propertiesof HES 130,000/0.4 have almost exclu-sively been demonstrated in seriouslyill intensive care patients to date. Itseems highly likely (even though thismay be speculative) that patients withnormal renal function are considerablyless sensitive to the potential nephro-toxic effects of HES 130,000/0.4 thanintensive care patients who are alreadyseriously ill. If these patients are nottreated with colloids in the event ofloss of larger amounts of blood, eithervery large amounts of crystalloids arerequired – with all the adverse effectsof the resulting positive fluid balance –or the unavoidably arising hypovolemiawith hypotension and subsequent va-sopressor application will very likelyendanger renal function. However, thisdilemma can be solved by consideringthe effectiveness: If a nephrotoxic ef-fect of HES 130,000/0.4 is present atall in otherwise healthy patients in theOR, then the data from intensive caremedicine indicate that HES-associatedperioperative renal failure is likely to bea rare event, since even in this patientcollective very high case numbers arerequired in order to detect this effect. Incontrast, prerenal acute kidney failureis much more likely to be caused bysevere hypovolemia. This “effect-based”risk analysis clearly favours the use ofHES 130,000/0.4 for correctly indicatedtreatment of larger blood losses in theOR. If one wishes to eliminate everyrisk, even a theoretical risk, then the ex-pensive albumin preparations discussedabove must also be used in the OR.

A further potential adverse effect ofHES 130,000 is its – albeit minor – nega-tive impact on blood coagulation, whichgoes beyond the effect of simple dilutionof coagulation factors [51]. This and thepotential impairment of renal functionprovide the reason for dose limitation(30ml/kg). A recently published meta-

analysis showed thatHES application canresult in a larger intraoperative blood losscompared to crystalloid therapy alone[66]. However, it must be noted thatlarger blood losses are often not appro-priately treated by the sole applicationof crystalloid. The drop in blood pres-sure caused by hypovolemia can itselfdecrease blood losses. Slightly reducedblood losses should not be an argumentfor neglecting the objective of demand-oriented maintenance of normovolemia,since the predominant application of va-sopressors for the compensation of a vol-ume deficit can seriously impair perfu-sion of organs such as the kidneys.

The concept propagated in the past,namely that patients be prophylactically“volume loaded” prior to surgery, e. g., byHES application, should also be viewedcritically, due to the possible impair-ment of the vascular barrier function[8]. This also applies to Caesarean sec-tion patients, since the safety of artificialcolloids for mother and child has notyet been sufficiently proven (by RCTswith high case numbers and long-termfollow-up). Pregnant women generallyhave a high normal blood volume priorto parturition per se, which does nothave to be and should not be expandedby artificial colloids. In summary, con-sidering the contraindications, the useof HES preparations for compensationof larger blood losses in the OR canbe endorsed. As long as sufficientlylarge controlled studies with long-termfollow-up are lacking for the perioper-ative phase, it is impossible to excludecompletely a residual risk (e. g., for renalfunction).

Gelatin

Alongside HES, gelatin is another artifi-cial colloid employed in the clinical use.Gelatin is a very heterogeneous polypep-tide mixture made from the collagenousconnective tissue of cattle specially bredfor this purpose. Since a part of the ap-plied amount is eliminated via the kid-neys already during its infusion, a vol-ume effect of 50–100% can be expected,depending on the preparation [20, 51].However, this volume effect is onlymain-tained for a maximum of 2 h. The ef-

fects on blood coagulation are even lesspronounced thanwithHESpreparations.However, intolerance reactions of all de-grees of severity have been described forgelatin preparations. It is certain thatgelatin preparations elicit the highest rateof anaphylactic reactions of all artificialand natural colloids [61]. Due to thebovine raw material, contamination ofgelatin with the agent responsible forcausing the new variant of Creutzfeldt-Jakob disease cannot be completely ruledout. Although a chemically very aggres-sive manufacturing procedure can mas-sively reduce the prions, it cannot com-pletely eliminate them. In the era preced-ing the first cases of bovine spongiformencephalopathy (BSE) in Germany, thetheoretic risk of BSE transmission viagelatin was calculated to be <1:1000000[61]. In the USA, gelatin solutions areno longer approved; in Germany an obli-gation for batch documentation is dis-cussed. Animal experiments have shownthat gelatin can cause histopathologicchanges in thekidneys, suchasvacuoliza-tion of tubule cells (“osmotic nephrosis-like lesions”)[61]. Incritically illpatients,a cumulativegelatindose>33 ml/kgbodyweight, is associated – as for HES – witha higher incidence of renal failure [61].

In contrast to the large amount of dataavailable for HES preparations or albu-min, there are currentlyno large random-ized studies demonstrating the safety andbenefit of treatment with gelatin prepa-rations [67, 68]. According to a 2012Cochrane analysis, data from RCTs areavailable for HES from 9147 patients andfor albumin from 9920; gelatin, however,has only been subject to controlled in-vestigation in this context in 506 patients[69]. “Absence of evidence” of coursedoes not mean that the preparations canbe applied in intensive care medicine orin the OR without hesitation. The factthat there is currently no sufficient evi-dence for the safety of gelatin solutions –neither in intensive care units nor in OR– does not support their use, but must beassessed as a considerable disadvantageof gelatin compared to albumin andHES[61, 67].

S8 Der Anaesthesist · Suppl 1 · 2019

The three-stage concept ofvolume therapy

As mentioned in the introduction, fluidtherapy should be target-directed and loss-adapted, and thus performed in a needs-based manner. The objective of treat-ment must be normovolemia. There iscurrently no evidence that volume re-placement with colloids reduces patientmortality compared to volume replace-ment with crystalloids [69]. However,if the volume effects of crystalloids andcolloids are considered, colloids are fivetimes more effective than crystalloids,provided they are used to treat the appro-priate indication of hypovolemia. Thisis also confirmed by the results of theCRISTAL trial, which could show a re-duction in mortality in patients with hy-povolemic shock achieved by volume re-placement with colloids [70]. The pe-rioperative infusion therapy comprisesdifferent phases, for which different rec-ommendations have to be given:4 Stage one: basal crystalloid therapy

Fluid losses (urine and insensibleperspiration) are replaced in a 1:1ratio with a balanced crystalloidinfusion solution. Additionally,intravascular blood loss must bereplaced in a 4–5:1 ratio. Thisis where the limitation of basalcrystalloid therapy is encountered,since this means at the same time theinfusion of 80% “interstitial edema”.As already mentioned, establishmentof a highly positive fluid balanceshould not be considered harmless[1, 3–5, 8, 24, 71]. A limit musttherefore be defined, above whichbasal crystalloid therapy is replacedby a treatment of higher effectiveness,in order to avoid the complicationsand adverse events associated withfluid overload. As a rule of thumb,blood loss corresponding to 20%of the initial blood volume canbe stated as the limit. Since mostpatients in the OR only experienceblood losses within this order ofmagnitude, according to this concept,the majority of patients can be treatedwithout colloid.

4 Stage two: blood losses exceeding 20%of the initial blood volume

Blood loss should be replaced by aneffective colloid therapy to preventthe fluid balance becoming too posi-tive. Fluid losses (urine and insensibleperspiration) are still replaced in a 1:1ratio with crystalloids. The largestvolume effect has been describedfor 20% human albumin (. Table 2).However, this is also by far the mostexpensive colloid. Considering thecontraindications, artificial isoon-cotic colloids such as HES may alsobe used in stage two (1:1 ratio inrelation to the blood losses exceedingthat of stage 1).

4 Stage three: for very large blood lossesThe use of infusion solutions (stages 1and 2) is limited by the increasingdevelopment of anemia and impair-ment of blood coagulation resultingfrom the dilution. In addition to thedilution of coagulation factors by themaintenance of normovolemia withdecreasing blood volume, gelatin andHES 130,000 also have direct negativeeffects on blood coagulation. Anemiaand the loss of coagulation factorsnecessitate a differentiated and de-mand-oriented transfusion of bloodproducts (erythrocyte concentrates,EC; fresh frozen plasma, FFP; throm-bocyte concentrates; coagulationfactors).

Application of this staged concept en-sures a loss-adapted and therefore needs-based volume therapy. However, thethree stages should not be followed toodogmatically! For example, should oneof the large veins suddenly tear – leadingto loss of a large amount of blood withinjust a few minutes – stage 2 should andmust be implemented immediately, sincethe volume loss must be treated withinfusion solutions of high effectivenesswithout delay. In this scenario, sufficientamounts of colloids should be appliedimmediately (e. g., 1–2 l of HES) to pre-vent intraoperative hypovolemia. Bloodproductsmay also be administered early.

The example in the following sectionhighlights the importance of correctlydefining the fluid and volume require-ments; it illuminates the perils of intraop-erative infusion therapy on one hand, but

also provides important simple hints andtricks on the other.

A practical example – volumetherapy explained using simplemath

Consider a 52-year-old ovarian cancerpatient undergoing exploratory laparo-tomy. The procedure is conducted un-der general anesthesia with a previouslyplaced epidural catheter. The patienthas no cardiovascular complaints andstill does sports; she weighs 67 kg witha height of 168 cm. During surgery theanesthetist endeavors to maintain nor-movolemia and the mean arterial bloodpressure is kept at 60mmHg at least.Theevents, balance status, and laboratoryparameters (hemoglobin concentration,Hb and hematocrit, Hct) at various timepoints during surgery are listed below:4 Induction:

jgeneral anesthesia and administra-tion of epidural anesthesia

jHb 12.3 g/dl, Hct 39.3%jnorepinephrine 0.3mg/h

4 1 h after the beginning of surgery:jHb 9.2 g/dl, Hct 29.4%jnorepinephrine 0.5mg/hjcumulative 1500ml of crystalloidinfused

4 2 h after the beginning of surgery:jestimated blood loss in towels andaspirators 1500ml

jurine excretion 500mljHb 8.2 g/dl, Hct 26.2%jnorepinephrine 1.5mg/hjcumulatively 5000ml of crystalloidinfused

4 3 h after the beginning of surgery:jsurgeon complains of impairedblood coagulation and sustainedsurgical bleeding

jadministration of three fresh frozenplasma (FFP) (750ml)

jHb 7.0 g/dl, Hct 22.4%jnorepinephrine 1.8mg/hjinfusion of another 1000ml ofcrystalloid solution

jadministration of four red cellconcentrates (cumulative total of1200ml)

4 4 h after the beginning of surgery:jHb 7.7 g/dl, Hct 24.6%jnorepinephrine 1.2mg/h

Der Anaesthesist · Suppl 1 · 2019 S9

Leitthema

Table 3 Overview of fluid balance at the end of surgery

Output Input

Urine 700ml Crystalloid 6000ml

Blood 3000ml Erythrocyte concen-trate

1200ml

Perspiration 335ml Fresh frozen plasma 1500ml

Sum 4035ml Sum 8700ml

Fluid balance +4665ml

jadministration of another threeFFP (cumulative total of 1500ml)

jthereafter Hb 7.5 g/dl and Hct24.0%

jestimated total blood loss 3000mljtotal urine excretion 700ml

4 End of surgery 5 h after the beginningof surgery:jHb 7.9 g/dl, Hct 25.3%jnorepinephrine 0.6mg/h

The patient is extubated and moved tothe intensivecareunitwithnoradrenaline0.7mg/h. Due to increasing retentionparameters, renal replacement therapy isinitiated. This has to be continued for3 days due to acute renal failure. Renalfunction normalizes hereafter, and after11 days in the intensive care unit thepatient is transferred to a general ward(. Table 3).

Why did circulatory / cardiovascularinstability and renal failure develop in thispatient?

The mathematical solution of thiscase

Preoperative situationBased on the patient’s height and weight,a body surface area (BSA) of 1.76m2 canbe calculated [72].

BSA =

√

height [cm] x weight [kg]3600

It isassumedthat thepatienthadanor-mal preoperative bloodvolume, since pa-tients with a healthy cardiopulmonarystatus generally have an even high nor-mal morning blood volume despite sev-eral hours of fasting [9]. Using the mea-sured normal blood volumes per ml2 ofBSA (listed in . Table 4; normal female

BV: 2245ml/m2), the blood volume (BV)can be calculated .

Blood volume = body surface area×

normal values of blood volume perm2

BVNormovolemia

= 1.76m2×2245ml/m2

= 3951ml

Practical tip: BSA can be calculated us-ing tools available online. For a simplifiedcalculation of blood volume, BV=60ml/kgcan be used, although this is less precisethan calculations using BSA.

FromtheHctvalue, theredcellvolume(RCV) can be calculated.

red cell volume =blood volume×hematocrit

RCVpreop = 3951ml×39.3% = 1553ml

Practical tip: if only Hb values areavailable, Hct can be calculated using: Hb(g/dl) × 3.2 = Hct. This requires a normalmean corpuscular hemoglobin concentra-tion (MCHC), which, however, is the casein the vast majority of patients.

Provided the patient remains indeednormovolemic, using the blood volume,it is possible to calculate at every particu-lar timepoint the corresponding erythro-cyte volume and thus the change in ery-throcyte volume (ΔEV) as well as bloodloss using the mean Hct value.

For example, the situation 2 h after thebeginning of surgery:

Intraoperative red cell volume= BVNormovolemia×HctintraopRCV = 3951ml×26.2 % = 1035ml

Change in red cell volume= preoperative red cell volume–intraoperative red cell volume

Table 4 Composition of total blood vol-ume perm2of body surface area [9, 73]

Compartment Gender ml/m2

Erythrocyte volume Female 850

Male 1100

Plasma volume Female 1395

Male 1578

Blood volume Female 2245

Male 2678

ΔRCV = RCVpreop–RCVintraop

= 1553ml–1035ml = 518ml

Mean Hct (Hctm) is calculated fromthe sumofallHctmeasurementsmadeupuntil the point in time 2 h after incisiondivided by the number of measurements.Inthisexample,Hcthadbeendeterminedn = 3 times by the point in time 2 h afterincision

Hctm =(39.3 + 29.4 + 26.2)

3= 31.6 %

Via Hctm, the blood loss can be calcu-lated from the erythrocyte loss (blood =erythrocytes and plasma):

Calculated blood loss

=

change in red cell volumemean hematocrit

resulting in

Blood lossintraop~ 2h =518ml31.6%

= 1639ml

Evaluation of volume therapy atthe point in time 2 h after incisionThe estimated and the calculated bloodloss values were almost identical at thispoint in time. With a cumulative admin-istration of 5000ml crystalloid solution,the fluid balance was +2727ml (calcula-tion: 5000ml of crystalloid – 134ml ofperspiration – 500ml of urine – 1639mlof blood loss). However, despite this,the administered volume therapywas notsufficient.

The following balance 2 h after thestart of surgery considers the volumeeffect of the volume replacement ther-apy selected in this example and demon-strates the intravascular volume balance.In this case, crystalloids replacing fluid lossdue to urine or perspiration are accredited

S10 Der Anaesthesist · Suppl 1 · 2019

Table 5 Overview of the volume balance for the intravascular space 2h after incision

Output Input

Urine and perspi-ration

637ml Crystalloid for fluid replace-ment

637ml

Blood loss 1629ml Amount of crystalloid ac-creditable to compensationof blood loss

873mlCalculation:(crystalloid at measurement –637ml) × 20%= (5000ml –637ml) × 20%

Sum 2266ml Sum 1510ml

Volume balance –756 ml

Table 6 Overview of the volume balance for the intravascular space at the end of surgery

Output Input

Urine and perspi-ration

1035ml Crystalloid for fluid replace-ment

1035ml

Blood loss 4956ml Amount of crystalloid ac-creditable to compensationof blood loss

993mlCalculation:(crystalloid atmeasurement time –1035ml) × 20%= (6000ml –1035ml) × 20%

– – Erythrocyte concentrate 1200ml

– – Fresh frozen plasma 1500ml

Sum 5991ml Sum 4728ml

Volume balance –1263 ml

in full, but those replacing blood losses areonly accredited to 20%, since only 20% re-main intravascularly (the remainder leadsto interstitial edema, . Table 5).

Therefore: At this point in time thepatient had a pronounced intravascu-lar hypovolemia (756ml), which ex-plains the cardiovascular instability andthe requirement for 1.5mg/h of nore-pinephrine. According to the three-stage concept (see section “The three-stage concept of volume therapy”), thecritical value of blood loss correspondingto 20% of the preoperative blood volumeshould have been calculated first. Ina patient with an approximate bloodvolume of 4000ml, this corresponds toa blood loss of 800ml. In stage 1, 4–5-times this amount of blood loss, plusthe losses due to urine excretion andperspiration, should have been replacedwith crystalloid. This corresponds toapproximately 4600ml (5 × 800ml +637ml). Thereafter, stage 2 should havebeen observed and further blood lossesshould have been replaced with colloids.In this example, 800ml of HES wouldhave been indicated.

Situation at the end of surgery

The calculation of blood loss proceedsanalogously to the calculation above forthe timepoint2 hafterthestartofsurgery.

First the loss of erythrocyte volume iscalculated:

BVpostop =3951ml(assumingnormovolemia)

RCVpostop = 3951ml×25.3% = 1000ml

ΔRCV = RCVpreop–RCVpostop

= 1553ml–1000ml = 553ml

Tobeable tofinally calculate the lossofred cells, in addition to the change inred cell volume, the administered redcell concentrates now have to be addedto the equation. One unit of a red cellconcentrate, having a volume of 300 mland a Hct of 65%, corresponds to about200ml of red cells.

Loss of red cellstotal= ΔRCV + (4×200ml)= 553ml + 800ml = 1353ml

The mean Hct value now enables theblood loss to be calculated from the lossof red cells:

Blood loss =Total loss of red cellsmean hematocrit

Hktm (seven measurements in total)

=

⎡

⎢

⎢

⎢

⎢

⎢

⎢

⎣

39.3% + 29.4% + 26.2%

+ 22.4% + 24.0%+ 24.6% + 25.3%

⎤

⎥

⎥

⎥

⎥

⎥

⎥

⎦

7= 27.3%

To calculate the final total blood loss,the total loss of red cells is divided bythe Hctm value calculated from mea-surements made throughout the entiresurgery. This was

Calculated total blood loss

=

total loss of red cellsHctm

Calculated total blood loss

=

1353ml27.3%

= 4956ml

Overall evaluation: Target missed!Pronounced hypovolemia despitea highly positive fluid balanceAt+4665ml, themereperioperative fluidadministration appears very generous;however, the total blood loss was un-derestimated by 1900ml. The followingoverview again considers the volume ef-fect of the selected volume replacementtherapy and thus represents the intravas-cular volume balance (. Table 6).

The patient had a severe hypovolemiapostoperatively (over 1200ml volumelacking)! Furthermore, in addition tobleeding during these surgical proce-dures, large intraoperative protein lossesalso occur, which have not been includedin the calculations presented here.

In summary, in this situation witha blood loss of almost 5 l, the adminis-tered volume therapy was not demand-oriented. Less crystalloid but consid-erably more “volume” (colloid) wouldhave been loss-adapted and the infusionof 2000ml of HES appropriate. Thepresented example also demonstratessomething fundamental: In patients

Der Anaesthesist · Suppl 1 · 2019 S11

Leitthema

NormovolemiaActual mean Hct: 20%

Hematocrit

Bloodvolum

e

Hctm : 32 % with two Hct measurements

Duration of surgery

Hctm : 22 % with seven Hct measurements

40 %

30 %

20 %

Fig. 39 Changein hematocrit withbleeding and simul-taneousmainte-nance of normo-volemiawith in-fusions (dilution).(Hct hematocrit,Hctmmean hema-tocrit)

without internal medicine complaints,if during major surgery a gradual pro-nounced increase in catecholamine re-quirement is observed, this is a verystrong indication that blood loss is beingtreated inadequately! The following is,however, speculative: The pronouncedhypovolemia with correspondingly highcatecholamine requirement had a moresevere impact on renal function than“defense” of normovolemia using HESwould have had in this renally healthy,noncritically ill patient.

Important comments for bedsideblood loss calculationThe calculation presented above is validfor each and every point in time dur-ing surgery. Particularly in apparentlyunclear situations (larger blood losses,transfusionof blood) this calculationwillbring clarity. Asdemonstrated in. Fig. 3,the calculation of blood loss always as-sumes a normovolemic patient. The cor-responding volume replacement thus re-sults in dilution; otherwise Hct does notsink logarithmically upon blood loss, aswould be the case if normovolemia wasmaintained. It is important to realizethat the blood volume can only increaseslightly due to infusions [24]. ANP isreleased immediately, the vascular bar-rier becomes permeable, and the bloodvolume quickly returns to normal high[25–27, 74]. In the case of larger bloodlosses, a moderate – sometimes consid-erable, as demonstrated by the presentedexample – hypovolemia is much morefrequent, caused by an apparently gener-ous but in fact inadequate volume ther-

apy. If blood loss is calculated, e. g., ata point in time at which the blood vol-ume is 500ml less than at the start ofsurgery, then 500ml has to be added tothe calculated blood loss. If, due to hy-povolemia and hemoconcentration, theHct increases, the result of blood losscalculation will be a negative value andgives an approximation of the extent ofhypovolemia.

As shown in . Fig. 3, in the incidenceof bleeding and normovolemic dilutionwith infusion solutions, Hct decreaseslogarithmically [75]. Owing to the sim-pler mathematics, the calculation ofblood loss presented above is not loga-rithmic, but rather linear. If, for example,only two Hct measurements are madeduring a bleeding (example 1), thenHctm will be considerably too high andthe calculated blood loss therefore toolow. It is thus always preferable to inputseveral Hct values into the calculation.In the discussed example, seven Hctmeasurements were performed (. Fig. 3,example 2). Although this is a suffi-ciently high number, the Hctm will stillbe somewhat too high and the calculatedblood loss will therefore be an underes-timation of the actual blood loss. Thismeans that potential hypovolemia andthe linearity of the calculation can resultin the calculated blood loss being lowerthan the actual blood loss. However,as the example demonstrates, even ablood loss calculated still too low pro-vides a far better overview than a pureestimation of blood loss, which is un-fortunately a very unreliable parameter.Factors contributing to this unreliability

include the fact that the losses in tow-els are very difficult to estimate, bloodcan flow into unseen regions which thesurgeon does not aspirate, and even theblood content of aspirators is difficultto estimate due to rinsing solutions,particularly in case of large blood losses.Unfortunately, considerable underesti-mation of intraoperative blood losses isthus a frequent event [76]. Therefore,the bedside blood loss calculation pre-sented herein is extremely useful andcan even be implemented in anesthesiadocumentation programs.

Practical conclusion – the synopsisAn adequate intraoperative infusiontherapy:4 should be needs-based, with the aim

of normovolemia,4 should be loss-adapted and distin-

guish between fluid and volumelosses,

4 should balance out fluid losses(urine and insensible perspiration0.5–1.0 ml/kg/h) with a balancedcrystalloid infusion solution ina 1:1 ratio,

4 should first compensate volumelosses (intraoperative blood loss)corresponding to up to 20% of thetotal blood volume with balancedcrystalloids in a 4–5:1 ratio, and treatblood losses exceeding this level withisooncotic colloids in a 1:1 ratio.

Considering the contraindications4 sepsis,4 burns,4 critically ill patient (usually in the

intensive care unit),4 impaired renal function or renal

replacement therapy,4 intracranial or cerebral bleeding,4 severe coagulation disorder

HES can also be used for volume replace-ment. In patients with a contraindica-tion toHES, albumin solutions should bepreferred to gelatin solutions, since thereare data available supporting the safetyof albumin infusions.

The three-stage concept for loss-adaptedand needs-based volume therapy:4 Stage 1: basal crystalloid therapy.

Fluid losses (urine and insensible

S12 Der Anaesthesist · Suppl 1 · 2019

perspiration) are replaced by a bal-anced crystalloid infusion solution ina 1:1 ratio. Intravascular blood lossmust be additionally compensated ina 4–5:1 ratio.

4 Stage 2: blood losses exceeding 20% ofthe initial blood volume. Applicationof colloids (preferably balanced) ina 1:1 ratio for replacement of bloodlosses exceeding those of stage 1.Fluid losses (urine and insensibleperspiration) continue to be replacedin a 1:1 ratio by a balanced crystalloidinfusion solution.

4 Stage 3: very large blood losses. Differ-entiated and needs-based transfusionof blood products, i. e., red cell con-centrates and FFPs (adapted to thepresent coagulation), as well as co-agulation factors and thrombocyteconcentrates as required.

Bedside blood loss calculation based onnormal blood volume and intraoperativeHb or Hct measurements provides theanesthetist with a very good orientation,which renders the objective –maintenanceof normovolemia in the patient – easierto achieve. Disregarding coincidences,a fluid and volume therapy can only beneeds based if the patients’ requirementsare known.

Corresponding address

Prof. Dr. M. RehmDepartment of Anaesthesiology, LMUMunichMarchioninistr 15, 81377Munich, Germanymarkus.rehm@med.uni-muenchen.de

Compliance with ethicalguidelines

Conflict of interest M. Rehm is currently directingtwo clinical studies, one ofwhich is supportedbyCSLBehring and the other ofwhich is supportedbyFresenius Kabi. He has receivedno travel expensesor honoraria frompharmaceutical companies duringthe past 2 years. N. Hulde, T. Kammerer, A.S.Meidert,andK. Hofmann-Kiefer declare that theyhave nocompeting interests.

This article does not contain any studieswith humanparticipants or animals performedby anyof the au-thors.

References

1. Brandstrup B, Tonnesen H, Beier-Holgersen R etal (2003) Effects of intravenous fluid restrictionon postoperative complications: Comparison oftwo perioperative fluid regimens: A randomizedassessor-blinded multicenter trial. Ann Surg238:641–648

2. Jacob M, Chappell D (2009) Rehm M The ‘thirdspace’ – fact or fiction? Best Pract Res ClinAnaesthesiol23:145–157

3. Holte K, Kehlet H (2006) Fluid therapy andsurgical outcomes in elective surgery: A need forreassessment in fast-track surgery. J AmColl Surg202:971–989

4. Lobo DN, Bostock KA, Neal KR, Perkins AC, Row-lands BJ, Allison SP (2002) Effect of salt andwaterbalance on recovery of gastrointestinal functionafter elective colonic resection: A randomisedcontrolledtrial. Lancet359:1812–1818

5. NisanevichV, Felsenstein I, AlmogyG,WeissmanC,Einav S,Matot I (2005) Effectof intraoperativefluidmanagement on outcome after intraabdominalsurgery. Anesthesiology103:25–32

6. Parquin F, Marchal M, Mehiri S, Herve P, Lescot B(1996) Post-pneumonectomy pulmonary edema:analysis and risk factors. Eur J Cardiothorac Surg10:929–932 (discussion33)

7. Glassford NJ, Myles P, Bellomo R (2012) TheAustralian approach to peri-operative fluidbalance. CurrOpinAnaesthesiol25:102–110

8. Chappell D, Jacob M, Hofmann-Kiefer K, ConzenP, Rehm M (2008) A rational approach toperioperative fluidmanagement. Anesthesiology109:723–740

9. Jacob M, Chappell D, Conzen P, Finsterer U, RehmM (2008) Blood volume is normal after pre-operative overnight fasting. Acta AnaesthesiolScand52:522–529

10. Fachgesellschaften_AAWM (2014) IntravasleVolumentherapie beim Erwachsenen (Register-nummer001–020)

11. Langer T, Conrad S, Fishman L et al (2012)Interessenkonflikte bei Autoren medizinischerLeitlinien. Eine Analyse der Leitlinien deutscherFachgesellschaften 2009–2011. Dtsch Arztebl109:836–842

12. Marckmann G (2015) Alkoholabstinenz vorLebertransplantation: Contra. DtschArztebl 112:A279

13. Rehm M, Haller M, Brechtelsbauer H, Akbulut C,Finsterer U (1998) Extra protein loss not causedbysurgical bleeding in patients with ovarian cancer.ActaAnaesthesiolScand42:39–46

14. Rehm M, Haller M, Orth V et al (2001) Changesin blood volume and hematocrit during acutepreoperative volume loadingwith 5% albumin or6%hetastarch solutions in patients before radicalhysterectomy. Anesthesiology95:849–856

15. Jacob M, Chappell D, Hofmann-Kiefer K, ConzenP, Peter K, Rehm M (2007) Determinants ofinsensible fluid loss. Perspiration, proteinshift and endothelial glycocalyx. Anaesthesist56(747–58):60–64

16. RehmM,OrthVH, KreimeierUetal (2001)Changesin blood volume during acute normovolemichemodilution with 5% albumin or 6% hydro-xyethylstarch and intraoperative retransfusion.Anaesthesist50:569–579

17. RehmM,OrthV,KreimeierUetal (2000)Changes inintravascular volume during acute normovolemichemodilution and intraoperative retransfusion inpatients with radical hysterectomy. Anesthesiol-ogy92:657–664

18. Jacob M, Rehm M, Orth V et al (2003) Exactmeasurement of the volume effect of 6%hydoxyethyl starch 130/0.4 (Voluven) duringacute preoperative normovolemic hemodilution.Anaesthesist52:896–904

19. Rehm MHM, Brechtelsbauer H, Akbulut C,Finsterer U (1998) Changes in plasma volumein immediate pre- and postoperative periods inpatientsmajorgynecologic surgery. InfusionstherTransfusionsmed25:222–228

20. Rehm M, Orth VH, Weninger E et al (2001)Acute “normovolemic” hemodilution with 3.5%polygel (Haemaccel) for patients in theWertheim-Meigs-operation. Blood loss of 87% bloodvolume without perioperative blood transfusion.Anaesthesist50:580–584

21. JacobM,ChappellD,Hofmann-Kiefer Ketal (2012)The intravascular volume effect of Ringer’s lactateis below20%: a prospective study in humans. CritCare16:R86

22. Jacob M, Bruegger D, Rehm M et al (2007) Theendothelial glycocalyx affords compatibility ofStarling’s principle and high cardiac interstitialalbumin levels. CardiovascRes73:575–586

23. Hu X, Adamson RH, Liu B, Curry FE, Weinbaum S(2000) Starling forces that oppose filtration aftertissue oncotic pressure is increased. Am J PhysiolHeartCircPhysiol279:H1724–36

24. Jacob M, Chappell D, Rehm M (2007) Clinicalupdate: Perioperative fluidmanagement. Lancet369:1984–1986

25. Bruegger D, Jacob M, Rehm M et al (2005)Atrial natriuretic peptide induces shedding ofendothelial glycocalyx in coronary vascular bedofguinea pig hearts. Am J Physiol Heart Circ Physiol289:H1993–H1999

26. Bruegger D, Schwartz L, Chappell D et al (2011)Release of atrial natriuretic peptide precedesshedding of the endothelial glycocalyx equallyin patients undergoing on- and off-pumpcoronary artery bypass surgery. Basic Res Cardiol106:1111–1121

27. Chappell D, Bruegger D, Potzel J et al (2014)Hypervolemia increases release of atrial natri-uretic peptide and shedding of the endothelialglycocalyx. CritCare18:538

28. WiedermannCJ (2014) JoannidisMAccumulationof hydroxyethyl starch in human and animaltissues: A systematic review. Intensive Care Med40:160–170

29. Nieuwdorp M, Meuwese MC, Vink H, HoekstraJB, Kastelein JJ, Stroes ES (2005) The endothelialglycocalyx: Apotentialbarrierbetweenhealthandvasculardisease. CurrOpinLipidol16:507–511

30. Scheingraber S, Rehm M, Sehmisch C, FinstererU (1999) Rapid saline infusion produces hyper-chloremicacidosis inpatientsundergoinggyneco-logicsurgery. Anesthesiology90:1265–1270

31. OrbegozoCortesD,RayoBonorA,VincentJL(2014)Isotonic crystalloid solutions: a structured reviewof the literature. Br JAnaesth112:968–981

32. Shaw AD, Bagshaw SM, Goldstein SL et al (2012)Major complications, mortality, and resourceutilization after open abdominal surgery: 0.9%saline compared to Plasma-Lyte. Ann Surg255:821–829

33. Young P, Bailey M, Beasley R et al (2015) Effect ofa buffered crystalloid solution vs saline on acutekidney injury amongpatients in the intensive careunit: The SPLIT randomized clinical trial. JAMA314:1701–1710

34. Senn A et al (2017) Chloride content of fluids usedfor large-volume resuscitation is associated withreducedsurvival. CritCareMed45:e146–e153

Der Anaesthesist · Suppl 1 · 2019 S13

Leitthema

35. O’MalleyCM, FrumentoRJ,HardyMAetal (2005)Arandomized, double-blind comparison of lactatedRinger’s solution and 0.9% NaCl during renaltransplantation. AnesthAnalg100:1518–1524

36. Potura E, Lindner G, Biesenbach P et al (2015) Anacetate-bufferedbalancedcrystalloidversus0.9%saline in patients with end-stage renal diseaseundergoing cadaveric renal transplantation:a prospective randomized controlled trial. AnesthAnalg120:123–129

37. Vincent JL, Dubois MJ, Navickis RJ, Wilkes MM(2003) Hypoalbuminemia in acute illness: is therea rationale for intervention? A meta-analysis ofcohort studies and controlled trials. Ann Surg237:319–334

38. Dubois MJ, Orellana-Jimenez C, Melot C et al(2006) Albumin administration improves organfunction in critically ill hypoalbuminemicpatients:Aprospective, randomized, controlled, pilot study.CritCareMed34:2536–2540

39. HaynesGR,Navickis RJ,WilkesMM(2003)Albuminadministration – what is the evidence of clinicalbenefit? A systematic review of randomizedcontrolledtrials. Eur JAnaesthesiol20:771–793

40. Martin GS, Moss M, Wheeler AP, Mealer M, MorrisJA, Bernard GR (2005) A randomized, controlledtrial of furosemide with or without albumin inhypoproteinemic patients with acute lung injury.CritCareMed33:1681–1687

41. Vincent JL (2009) Relevance of albumin inmoderncritical care medicine. Best Pract Res ClinAnaesthesiol23:183–191

42. Cochrane Injuries Group Albumin Reviewers(1998)Humanalbumin administration in criticallyill patients: Systematic review of randomisedcontrolledtrials. BMJ317:235–240

43. Investigators SS, Finfer S, Bellomo R et al (2006)Effectofbaseline serumalbuminconcentrationonoutcomeof resuscitationwith albumin or saline inpatients in intensive care units: analysis of datafrom the saline versus albumin fluid evaluation(SAFE)study. BMJ333:1044

44. Finfer S, Bellomo R, Boyce N et al (2004) Acomparison of albumin and saline for fluidresuscitation in the intensive care unit. N Engl JMed350:2247–2256

45. Caironi P, TognoniG,MassonSetal (2014)Albuminreplacementinpatientswithseveresepsisorsepticshock. NEngl JMed370:1412–1421

46. Charpentier JJ-PM (2011) Efficacy and toleranceof hyperoncotic albumin administration in septicshock patients: The EARSS study. Intensive CareMed37:115

47. Wiedermann CJ, Joannidis M (2014) Albuminreplacement in severe sepsis or septic shock.NEngl JMed371:83

48. Patel A, Laffan MA, Waheed U, Brett SJ (2014)Randomised trials of human albumin for adultswith sepsis: Systematic review andmeta-analysiswith trial sequential analysis of all-causemortality.BMJ349:g4561

49. Wiedermann CJ, Dunzendorfer S, Gaioni LU,ZaracaF, JoannidisM(2010)Hyperoncotic colloidsand acute kidney injury: A meta-analysis ofrandomizedtrials. CritCare14:R191

50. MaitlandK,KiguliS,OpokaROetal (2011)Mortalityafter fluid bolus in African children with severeinfection. NEngl JMed364:2483–2495

51. Rehm M, Paptistella M, Dieterich HJ (2012)Volumenersatzlösungen, 3 edn. Springer, Berlin,Heidelberg

52. Vincent JL, De Backer D, Wiedermann CJ (2016)Fluid management in sepsis: The potential

beneficial effects of albumin. J Crit Care35:161–167

53. Delaney AP, Dan A, McCaffrey J, Finfer S (2011)The role of albumin as a resuscitation fluid forpatients with sepsis: A systematic review andmeta-analysis. CritCareMed39:386–391

54. Brunkhorst FM, Engel C, Bloos F et al (2008) Inten-sive insulin therapy and pentastarch resuscitationinseveresepsis. NEngl JMed358:125–139

55. Perner A, Haase N, Guttormsen AB et al (2012)Hydroxyethyl starch 130/0.42 versus Ringer’sacetate inseveresepsis. NEngl JMed367:124–134

56. Myburgh JA, Finfer S, Bellomo R et al (2012)Hydroxyethyl starchorsaline forfluidresuscitationin intensivecare. NEngl JMed367:1901–1911

57. Guidet B, Martinet O, Boulain T et al (2012)Assessment of hemodynamic efficacy and safetyof 6% hydroxyethylstarch 130/0.4 vs. 0.9% NaClfluid replacement in patients with severe sepsis:theCRYSTMASstudy. CritCare16:R94

58. Dart AB, Mutter TC, Ruth CA, Taback SP (2010)Hydroxyethyl starch (HES) versus other fluidtherapies: Effects on kidney function. CochraneDatabaseSystRev. doi:10.1002/14651858

59. Wiedermann CJ (2008) Systematic review of ran-domized clinical trials on the use of hydroxyethylstarch forfluidmanagement in sepsis. BMCEmergMed8:1

60. Zarychanski R, Turgeon AF, Fergusson DA et al(2009) Renal outcomes and mortality followinghydroxyethyl starch resuscitation of critically illpatients: systematic review and meta-analysisof randomized trials. Open Med 3:e196–209(ATTENTION: The analysis and conclusions of thisarticle are being revised by the authors. Thisis due to the journal Anesthesia and Analgesia’sretraction of a paper by Dr. Joachim Boldt, anauthor in seven of the studies analyzed in thisreview. As such, the editors of Open Medicinerecommend interpreting this reviewwithextremecaution until Zarychanski et al. publish a newanalysis and interpretation in OpenMedicine. Formore information, see Anesthesia andAnalgesia’spress release)

61. Rehm M (2013) Limited applications for hy-droxyethyl starch: Background and alternativeconcepts. Anaesthesist62:644–655

62. EuropeanMedicines Agency. Assessement reportfor solutions for infusion containing hydroxyethylstarch. EMA/667553/2013

63. Meybohm P, Van Aken H, De Gasperi A et al(2013) Re-evaluating currently available data andsuggestions for planning randomised controlledstudies regarding the use of hydroxyethyl starchin critically ill patients – a multidisciplinarystatement. CritCare17:R166

64. Gillies MA, Habicher M, Jhanji S et al (2014)Incidenceofpostoperativedeathandacutekidneyinjury associatedwith i. v. 6%hydroxyethyl starchuse: systematic review and meta-analysis. Br JAnaesth112:25–34

65. Kammerer T, Klug F, Schwarz M et al (2015)Comparison of 6% hydroxyethyl starch and 5%albumin for volume replacement therapy inpatients undergoing cystectomy (CHART): Studyprotocol for a randomized controlled trial. Trials16:384

66. RasmussenKC,SecherNH,PedersenT(2016)Effectof perioperative crystalloid or colloid fluid therapyon hemorrhage, coagulation competence, andoutcome: A systematic reviewandstratifiedmeta-analysis.Medicine (Baltimore)95:e4498

67. Moeller C, Fleischmann C, Thomas-Rueddel D etal (2016) How safe is gelatin? A systematic review

and meta-analysis of gelatin-containing plasmaexpanders vs crystalloids and albumin. J Crit Care35:75–83

68. Thomas-Rueddel DO, Vlasakov V, Reinhart K et al(2012) Safety of gelatin for volume resuscitation –a systematic review andmeta-analysis. IntensiveCareMed38:1134–1142

69. Perel P, Roberts I (2012)Colloids versus crystalloidsfor fluid resuscitation in critically ill patients.Cochrane Database Syst Rev. doi:10.1002/14651858

70. Annane D, Siami S, Jaber S et al (2013) Effects offluid resuscitation with colloids vs crystalloids onmortality in critically ill patients presenting withhypovolemic shock: TheCRISTAL randomized trial.JAMA310:1809–1817

71. HolteK,SharrockNE,KehletH(2002)Pathophysiol-ogy and clinical implications of perioperative fluidexcess. Br JAnaesth89:622–632

72. MostellerRD(1987)Simplifiedcalculationofbody-surfacearea. NEngl JMed317:1098

73. Pearson TC, Guthrie DL, Simpson J et al (1995)Interpretation of measured red cell mass andplasma volume in adults: Expert Panel onRadionuclides of the International Council forStandardization in Haematology. Br J Haematol89:748–756

74. Jacob M, Saller T, Chappell D, RehmM, Welsch U,Becker BF (2013) Physiological levels of A-, B- andC-type natriuretic peptide shed the endothelialglycocalyx and enhance vascular permeability.BasicResCardiol108:347

75. Jacob M, Bruegger D, Conzen P, Becker BF,Finsterer U, Rehm M (2005) Development andvalidation of a mathematical algorithm forquantifyingpreoperative blood volumebymeansof the decrease in hematocrit resulting fromacute normovolemic hemodilution. Transfusion45:562–571

76. Orth VH, RehmM, Thiel M et al (1998) First clinicalimplications of perioperative red cell volumemeasurement with a nonradioactive marker(sodiumfluorescein). AnesthAnalg87:1234–1238

S14 Der Anaesthesist · Suppl 1 · 2019