2- 2007 revista de nefrología, diálisis y trasplante volumen 27 - n° 2 - 2007

ansmis-rol Dial

Vascular Access Maintenance:A systematic approach to an unsolved question

idez JL.iemodí-low-up, Eduardo Castiglioni (1), Jorge Bezic (2), Jorge Antonelli (1), Manuel Canalis (1), Marcelo Unzue (1) y

Gaston Alvarez (1).{L, Hsuased se-tepatitis

(1) Hospital Alemán, Dialysis Clinic, Buenos Aires, Argentina; and(2) Nacional University of Entre Ríos, Enqineering School, Paraná - Entre Ríos, Argentina.

}fa sirn-:patitis e

ore anti-genomic88.

~1"f,mc:iOIlanrlientoadecuado de los accesos vascularesaCtcfe;lu:iFpliCic!ntlesen hemodiálisis, constituye un eslabón

lograr la eficacia de los tratamientos sus-~"'~~l1_(l,l",,,,,,Desafortunadamente, no éxistierori.importantes en este campo durante las últimas

.' ª,'~lec:aaas, y la falla del AV sigue siendo considerada:t:;~;J:ru;~llfllHielas principales morbilidades y causas de

por esta realidad, establecimos en 1999 unvigilancia, considerándolo como parte de

r~í~;"~;~~':;,,,,~nmás amplio de administración del ciclo decon un grupo de trabajo interdisciplinario

1,"~~~g"c:,gllll.pIDnenteprincipal.bibliográficos reconocen las ventajas

revista de nefrología, diálisis y trasplante volumen 27 - n° 2 - 2007

INTRODUCTIONMaintenance of VA patency in hemodialysis (HD) pa-tients is nowadays a major expense that consumes asignificant fraction of the budget for healthcare, arous-ing the attention of more than just the patients andheaIthcare staff. Despite extensive clinical and scien-tific efforts, VA-related problems currently accountfor more than 25% of aIl hospitalizations inend-stagerenal disease (ESRD) patients' totaling more than 1bil-lion doIlars per year being spent on access-related careonly in the United States-.The construction of native fistulae (VAF) is currentlythe preferred choice of VA for chronic HD in view oftheir superior patency and low complication tates',However, VA grafts (VAG) are frequently utilized be-cause of comorbid factors such as diabetes and/or age,which often limit the VAF implantation success rate.Accordingly, 20-60% of HD patients in Europe and theUS, respectively, depend on VAG for permanent VN·5.In view of the oñgoing globardia5etesgf6wihglrópü~-lation records", a sharp increase in ESRD incidence canbe expected in patient groups likely to require VAGrather than VAF.The VAG failure has a consistent pattern offibromuscu-lar intimal hyperplasia uncovered, most often at or nearthe venous anastomosis; the progressive venous ob-struction just distal to the graft outflow tract (stenosis)caused by this Iesion reduces blood flow and ultimatelyleads to thrombosis. Salvaging a thrombosed access isusuaIly an emergent procedure that, if not immediatelysuccessful, causes delays in dialysis treatment and mayrequire placement of temporary dialysis catheter withfurther endanger of the patient s life", added costs andthe creation of inconvenience for the patient-".Up to now, aIl tested pharrnacological and surgical in-

o terventions have not resulted in better outcomes for theHD patient s VA matter. On the other hand, periodi-caIly measuring access flow (Qa) plus dynamic andlorstatic venous and arteriªlQrt':s'§'llr~Il1oni!.oringcªn i.clen=tify failing grafts and fistulas before they thrombose,aIlowing elective intervention without interrupting thepatient s dialysis schedule, and avoid other compli-cations'". Several studies suggest that monitoring notonly identifie~Y,'\Jhªt is mOft': likely toJail but,.~heI! .combined with timely intervention, also prolongs theaccess' lifell.\2.I3.\4. The tradeoff is an increase in angi-ography, angioplasty or vascular surgery revision, butthe patient-benefitsfrom reducing hospitalizations andonear elimination of temporary catheters", The net eco-nomic effect is a considerable reduction in the provid-er s costs'v'".

Driven by aIl these ideas, we have established in 1999an aggressive surveillance (specific VA measurements:flow, pressure) and monitoring (clinical observations:thrill, pulse) program, with a main character: the Mul-tidisciplinary Vascular Access Team (MVAT). WithinMVAT, with the nephrologists coordinating activ-ity\8.\9, a group of specialists composed by the vascu-lar surgeon, the interventional radiologist, the nursingstaff29•2\ and other support specialties (biomedical en-gineers, social workers), works very closely to analyzeand decide the course of action regarding VA issues.With our accumulated experience and results as weIl asthe discrepancies observed over these years regardingthe VA surveillance and monitoring (VA SIM) versuspatency rates issue, we decided to face this controver-sial matter on the following pages:1) Analyzing the controversial points from a systematicpoint of view,2) And presenting our VA SIM program s

a. biophysical foundationsb. outcomes and results

Systematic analysisIn figure 1we can see a lifecycle systematic approachto the VA subject, and the relevance of SIM programsunder MVAT management. In every stage there are sorneissues to consider, i.e., how much mapping to do, howmuch time to wait for first-using a VA, on which VAvariables to rely on to monitor its function, what warn-ing levels to use related to those variables, and so on.The overall VA patency depends on the result of each ,stage: a poor planning usually leads to a more frequent trepairing or directly to a higher thrombosis rate, repair- jing interventions made by a vascular surgeon without :.deep experience on the dialysis field.tusually.in patients.freferred by social security agencies) could be less ben-eficial than those made by an specialized surgeon.Hence, within each phase there are sorne critical sub- .7jects that must be properly addressed in order to maxi-j'mize outcomes, measured as the ideal characteristics of ~VA for HD: a minimum blood flow must be reached so 'that an adequate dialysis" can be achieved, an extend-ed(but in a reasonable timeframe consideringpatient related factors) life or patency with minimumcomplications for the patient (no thromboses, no "11Itpf"_ccLtions, no hemodynamicalterations).StatingthdlASIM program as the tool fortion of VA' lifecycle, we will focus on sorneological questions that we identify as controversialthe VA monitoring programs literature, and al so as thereasons of sorne reported failures in this field.

66

1M. The concept of S&M program as VA'lifecycle administration under management of MVATMultidiscipli naryVascular Access

2 - 2007

Vascularmapping

Selecticnof bestlocationforVA

~

revista de nefrología, diálisis y trasplante

n 1999:ments:rations:le Mul-Withinaetiv-vaseu-nursing .•.....-1----- ...-cal en-malyze .-1-.-"_-'"sues. 1r~[;'~0~'"well asgarding

Placement &maturation

ltematie

Preventiveactions

illustrated that these hemody-~4e--arG-tIEHlIFbulentflowpatterns, specially

b~~J:l:IDtilC.rlegion:!9,?O.This turbulence results.in;an4::I/órre~ducedlevels of shear stress during the

with a dramatic impaet on the function. differences between

n.1icprpnf VAinside the same patient.:91!llrnend(!d by the KlDOQI guidelines", a pa-

be referred for a diagnostic fistulogram ifla-,ClLLr:,,,, flow is less than 600 mLlmin or if the

volumen 27 - n° 2 - 2007

Team

MVAT

ManagementresponS¿fliility for

.VascQlar.Aécess

. unreinanceand';~Mo¡'ftQring .

'Y;- - - - - ~'i,.--':'." >~

rºgral11" .'

blood flow is les s than 1000 mlzrnin and has decreasedby more than 25% over a 4-month period (sorne modi-fications have been made in the last published KlDOQIguidelines'" regarding this issue). Both high and lowflows depending on hemodynamic eonditions, wereseen in patients withfistulas an~grafts33;. butts clearthat when the VAG flow falls below a given limit, therisk of graft thtonibosis inereases dramatically=P-" ..Thís.Iimit.could be found within the 600-800 ml/minrange as stated in DOQI guidelines, but other investi-gators also found high thrombosis rates when the flowis :::1300 ml/min".Krivitski" suggests in a recent work that this fixedpointsneed to be.modiñed.to.hígher values, sinee a 50% ste-nosiswere related 10 very different flows depending onthe VA initial conditions considered. Correspondingly,regarding the pressures" (venous and arterial,both statícanddynamic~-anrrintra:ªcc:ess),there are similar prob-lems; boosted by the faet that the fixed values eited bythe literature are outdated (most investigations in the late80's and 90's were done using 16G needles, and cur-rently it s widespread the use of ISG needles).

67

revista de nefrología. diálisis y trasplante volumen 27 - n" 2 - 2007

There are other drawbacks, i.e., needle misplacementand orientation, patient arterial pressure variation;sorne of that were addressed by Besarab' s group defin-ing the concept of intra-access to mean arterial pressureratio", AIl these pressure mea sures (also the methoddeveloped by Besarab), if used isolated (not correlatedto flow measurement), do not detect stenosis in sornetypical places, and in more complicated lesions likemultiple stenosis in series caused by repetitive catheterplacement".We will further analyze this point through a simplemathematical model, comparing two hemodynamicaldifferent VA. As we expected, the use of fixed points,regardless the value, could lead to a late or early inter-vention, always with increased costs (see "Simulationresults"). Nevertheless, even the detractors ofmonitor-ing programs agree that measuring helps to detect thedysfunction prior to VA thromboses, and several stud-ies show a good correlation between sorne measure offlow (value or variation) and/or pressure ami an ana-tomical finding (outlet or inlet stenosis); but they findthat when this lesions are treated by preemptive PTA(percutaneous transluminal angioplasty), the second-ary (assisted) paí~I1cy does not differ from the patientstreated after thrombosis'". This 1S the second point inthe "controversies analysis".

2) Repair effectiveness:The relative lack of effectiveness of the repair proce-dures in achieving a longer VA life was recently estab-lished in sorne clinical trials. This cori.clusion is flawedby several methodological problems:• The repair of the stenotic lesion can be done usingeither a surgical approach or a transluminal procedure.The success of PTA strongly depends on the elastic-ity of the lesion, and the feasibility of surgery dependson the lesion s site. High PTA rates were seen in thoseinvestigations. showing at least a strong preference forthis method over ~.sllE~i~aLllEPE~.ll~.!J:lis prefere11~~could be based on several already stated advantages ofintraluminal methods over the more invasive surgery(morbidity, economics), but in sorne cases the elasticrecoil shown by stenosis treated by PTA should be ad-..

2 - 2007 revista de nefrología, diálisis y trasplante volumen 27 - n° 2 - 2007

differ-year, a

Furthermore, the error for US devices depends mostlyon the operator s skills; and stated in this way, the fig-ure becomes more acceptable for the general medicalcommunity: it s common to hear comments about howIlluch the results of Doppler studies depend on the op-erator ability.What do these "errors" values mean? Let s take a com-

.--'¡"n-n1~fl-taltlon for a Doppler measurement setting, with"á"coloservatlve 15% error:

.~""éé==,,-.-_-, "Real" unknown value: 1000 ml/min---.'.Me~asllrelnelrH1: 1100 ml/min_._c.JVlt;i:1Ii111CJlIICJlll2:900 ml/min

throm-lis pro-~aphicing theagain~ombo-since aieved"ilities).of re-rresult

ionitor-:edVA.

.,._LY.lVu".lr~~~~· 1: 1100 +/- 165 ml/min;:=M~asuremlent 2: 900 +/- 135 ml/min

uccess-ctionalboth in

error theory frame, in spite of the apparent,..""'-''''''"l'''''" 1100 and 900 ml/min as an absolute

must concIude that,because óf measüre---the two values are NOT really different (in

that two statistical hypotheses are not. a given.i'p"). Note that if the limit be-

lUlJ'''LlVUJ'U>', and not-functioning access is 1000then the measurement 1 implies no action, and

IIl'pn1p,nt 2 implies an angiography and/or PTA,

if no other considerations about errors are made. Thesame applies to every other physical magnitude usedfor VA S/M. Therefore, considering the variability ofmeasures due to measurement technique itself, plusthe intra-patient variation because of variable physi-ological adjustment, one must concIude that a betterapproach is to look at tendencies and the integration,than focus on isolated values and variables.

I

4) Monitoring and surveillance programThe composition and interaction within MVAT, the de-cisions made regarding the course of action for eachVA, and the quality and timing of decision are deter-minants, both in our experience and in the others";for a program with good results. In the several papersthataddress the VA monitoring issue, there is no de-scription-of how,' when and using which proceduresthe group in charge works. This is not a minor subject,since the whole success relies on the decisions madeby the group.Given the complexity of the mechanisms contributingto VA graft failure (such as hemodynamic and biophys-ics factors, compliance mismatch, endothelial damage,inflammation, platelet activation, growth factor re-lease, etc), and until the science find different approachfor treatments which could hold a promise for optimíz-

urement 1equip-fy theirarisons

ger be-

Two measurements with realinstruments (Le.with errors).The real value Ii~§ ()n anovérlapplnq-zone whichindicates thatthetwomeasured values, whencorrectly interpreted, are notreally different.

Real unknown value

:0 Tran- Measurement 2

(value ± error)

69

revista de nefrología, diálisis y trasplante

T

volumen 27 - n° 2 - 2007 re- ista

ing AV graft patency rates, the surveillance programsare useful+The thorough knowledge of every VA, andthe decisions made on every step of its care, can betranslated into huge improvements in both the healtheconomics and a more specific measure of care: thepatient s quality of Iife.

Biophysics approach of our SIM program andSimulation ResultsWe will introduce the fundaments and concepts behindour S/M program, and through a simulation, we willcomplete the analysis of the controversial points we havediscussed about on the preceding pages. We have chosento develop a graft model because a better and direct cor-relate between anatomical zones and model componentsalong with failure sites can be shown. With sorne modi-ficatioris, the model can be extended also to AVEThe simulations use the well known fluid flow laws,which we present here rewriting the mathématicálterms so that an immediate correspondence with ana-tomical and physiological entities could be stated:

Q = (Pi-Po)lRh, withRh = 8:rl.LI(7r.R4)

(por an in-depth discussion and applications, see refer-ences= 56. 57. 5S)This way of writing the Hagen-Poiseuille law, allowsan explanation of the flb\V' from two standpoints:a) ¡he driving force, the differential pressure betweentifebeginning (Pi) and the end (Po) of the vascular seg--ment considered.

b) the opposition to this driving force, the hydraulic Finallresistance (Rh) given by the vessel geometrical char- bloodacteristics (longitude L, radius R) and a blood related mentrheological parameter (blood viscosity 11)· tanceThis law strictly deals with the linearcomponents which min/ndissipate energy within the hydraulic circuit. But in the tive fIhuman circulation, other non-linear dissipating enti- All thties are present, Iike the anastomosis and stenosis. One are sitechnique to consider these two effects is to modify the equatparameters in the Poiseuille equation conveniently, not circulto reflect just the linear components, but to include, in a the PIsimplified way, also the non-linear ones (i.e., consider- 4 arteriing sorne of the factors raised to 2nd, 3rd ••• power) conce

¡:¡

Other approach, and the one we have employed, is to start ~ the ](at typical measured pressures profiles'" (which include E fromall the effects within each physical "elernent" ofthe vas- ! the accular access), and td compute from this normal protiles e bloodthe normal hydraulic resistance of each VA' partoAfter ~ As ththis step, the variation of the resistances which reflects i endsthé pathology being simulated is calculated. I one eThe prosthetic graft can be modeled splitting it into i resist3 segments: a resistance for the arterial anastomosis a quen:(Rart), one for the graft itself (Rg), and another one for I effecithe venous anastomosis (Rven). Flowing through the I at R\three resistances connected in a series topology, there l... (folleis the access flow Qa.: wevIn figure 3, we present also a typical pressure protile; dantthrough the graft, beginning at the arterial pressure (as IThe tmean arterial pressure, MAP) and on the other end, the ] unitsvenous central pressure PVen.This is a simplification,but l.... gardiuseful for our use of the model for comparison purposes. accoi

I~~~;lI~

I-i'J

I.\\l

!1j~

Arterialside

Pressures ModelAnatomic sketch

Venousside

Sketch of VAG,pressure profileand intravascularresistances model.

MAP

Rv25

RartRa30Qa

1Rg Re35

Rven

Pven

70

1° 2 - 2007 revista de nefrología, diál isis y trasplante volumen 27 - n° 2 - 2007

Finally, when the patient is connected to a HD filter, theblood circuit with all the components of the HD equip-ment is also depicted (figure 4): RNa and RNv (resis-tance for needles and tubing, typical value 0.28 mmHg.

its which min/ml), R filter (resistance for HD filter), Qb eff (effec-lut in the tive flow at blood pump, typical value 400 ml/min).ing enti- All the parameters normalIy used in VA SIM programsisis. One ·~~l}-l;nU-1.vu uu the figure 4, along with the resultingodify the and initial conditions used in our model: re-ently, not as a percent of blood pump flow (Rec%),lude, in a measured at HD equipment sensors, bothconsider- and venous (PSVen). Regarding thezer) recirculation (in our model we consider only

access related recirculation), only appearspoint of view when the flow trhough

~",,_>.¿" '" less than the flow through the externaleff.

starts to develop the failure which finalIy. nothing is done to correct it) in thromboses,

wíthin theVAraises itshydraulic(usualIy because of stenosis, and as conse-

"""«""'_,,,,, reduction can be found). As the finalour model, weconcentrate the stenosis effect

that it becomes a time-variant resistancea time-squared dependence). The constantfor simulating the effect of a time depen-

';ñ""",.,¡,,,, for Rven is 0.00118 l/s2.in our simulation is parametric with arbitrarythough it can be considered as weeks, re-clinical experience. The same criteria (inwith clinical observations) was used totime-squared dependence for stenosis de-

therefore for Rven).

iydrauliccal char-d related

We discuss the effect of a fixed value over the decisionsmade about VA care comparing the results obtainedwith two simulations. These simulation correspond totwo hemodynamicalIy different situations: a high-flowcondition for "Case A" (e.g. a straight arm graft) and alow-flow condition for "Case B" (e.g. a forearm loop).As inicial conditions, we have chosen a flow of 2000ml/min for case A, and 1400 ml/min for case B. As thethresholds in this hypothetical surveillance program,for the flow parameter we use 1200 ml/min, and fordynamic venous pressure, 200 mmHg.

Simulation resultsThe simulation, using the complete equations set, wasmade using the program Mathematica. The time depen-dant effects ofRven on variables (stenosis percentage, andother relevant model variables) are shown infigure 5.Using the thresholds for flow mentioned ab ove, the fig-ureshows that in Case B an "early warning" is issuedwith a significant but low stenosis (32%), meanwhilefor Case A the warning arrives at a more difficult situ-ation, with a stenosis level over 50%. Recirculation,as stated also by several investigators in thís field, isa very late warning, at least for VAG (and !)f limitedutility in the VAF case). The only real.application ofisolated recirculation measure;~(not as an"ir¡~rme-diate tool for flow calculati~l"l~fthe correcti0n of he-modialysis time to achieve the prescribed K~ ..In the pressures analysis case (figure 6), thé PSVen(pressure at venous HD equipment sensor) is'al&o pooras an indicator, in the sense that it fluctuares ne';." theselected 200 mmHg level when stenosis still continúes

.. "

Model

elarxíel.

EquationsPSArt =MAP -Rart x Qa -RN a x Qb eft

Rec %= Qa/Qb eft

PSVen =Pven +Rven x Qa +RN a x Qb eft

~ ~~ ~

circuparameters, andgoverning equations,along the Rvendependence over time.

71

revista de nefrología, diálisis y trasplante volumen 27 - n° 2 - 2007 rev ista l

MAM The effect on variables Q and %Rec of two hemodynamically different VA. Note in (a) and (b)the imprecision zone for threshold determination due to the measurement error for flow.

I1.1:,

Flow, mllmin2000

1800

1600

1400

1000

;¡

Stenosls 32%

% (Rec & Stenosis)100%

90%

800

600

400

200

2 4 8 8

advancing, or even never reaches this level in case ofmediurn-level stenosis, or in sorne cases of collateralbranching. To overcorne these sensing problerns, sorneauthors suggest a modification over the pressure meth-od60•61, all trying to find the value for Qb that irnprovessorne detection characteristic (sensitivity, specificity),but with no definitive results.

Outcomes of our S/M programSince January-199962, we have established a5rMpro-gram (101 patients with 125 access, VAF: 74 and VAG:51) based on a strong biophysical basis'", aiming at ste-nosis prediction, detection, confirmation and treatmentbefo_r(!_aJ!trgl!l:~2!is_eventoccurs. Thebasis of our pro-gram, as sketched in the model discussion aboye, is thetrend analysis considering the history of the particularVA, and every time a significant reduction of flow'"has.been.detected.tusing Blood Temperature MonitorBTM, Fresenius Medical Care), with a concomitantvariation in pressures. This device employs a thermalbolus in the dialysate side, wich changes in tum theblood ternperature, and this change is measured by thearterial sensor at the module when a recirculation ap-

, -' 1-

% Stenosis

80%

70%

% RecB 50%

40%

30%

20%

10%

Time (arbitrary units)

10 12 14 16 18

pears (an artificially high recirculation is created by in-terchanging the patient needlesj.A closer surveillanceover the implied VA is assurned.When the MVAT has a confirmation of the variablesrnovernent towards an indication of stenosis, an imag-ing technique is carried out (preferred method angiog-raphy), and after the confirmation, the MVAT decidesthe angioplasty (transluminal or by surgery)\falesionis detected, or decides the course of action for a newVA placernent if necessary.WeTíitróduc:ethe results for the January- 1999 to De-cember-2004 period, using Kaplan-Meier survivalanalysis (KMSA). For KMSA an evenf is defined asVA placernent, VA failure, surgery repair, PTA or VA.replacernent, without considering as an event the diag-nostic studies-(DopprefofangiogtaphY)· We have takenas an occurrence for KMSA the VA failure or catheter,and as censored events: facility change, kidney trans-plantation, or death, __Infigure 7, we present the cumulated survival curves,along a surnmary for assisted patency. These valuesVAG are close to those suggested by DOQI.Over the 6 year period to achieve this assisted patency,the 74 VAF have dernanded 39 procedures (0.09 inter-

72

~:cro.oeQ.

ro>.~:::JCJ)

'OQ)-ro:::JEE:::Jo

ventiocedureprogre0.05 tI

nd (b)IV.

In* The effect on variables PVen of two hemodynamically different VA. Note in (a) and (b) theimprecision zone for threshold determination due to the measurement error for flow.

n° 2 - 2007 revista de nefrología, diálisis y trasplante volumen 27 - n'' 2 - 2007

enosis)Xl%

90%

Recirculation90%

70%

90%

50%

40%

30%

20%

10%

0'''14 16

PVenA

t PVen >50% in both cases

6 B

Years

-il.MKaplan-Meier survival analysls.

curve representcensored events. Also the summaryfor assisted patency is presented.

atrisk), and the 51 VAG, 94 pro--year at risk). Other

indicator, the thrombosis rate, waspatient-year at risk.

CONCLUSIONThe understanding and use ofbasic hemodynamic prin-cipIes within the VA context can help in planning anoptimum approach to prevent and treat access failure.

73

revistade nefrología,diálisisy trasplante volumen27 - n°2 - 2007

The basis of this failure, triggering of thrombosis, issecondary to low flows" and the activation of differentendothelial messaging signals at ceIlular level, whichlead to a progressive increase of one or more segmentsofVA' resistance, with aIl the concomitant variable al-terations: lowering of flow, increasing of blood veloci-ties and waIl shear stress in sorne sites and lowering inothers, and pressure alterations.In the last 10 years there have been several papers ad-dressing the different factors that finally lead to con-troversial statements about the real value of a VA S/Mprogram: no better accumulated patency with more in-terventions and therefore expenditures.We have identified and examined several methodologi-cal problems, such as including few patierits in thestudy in order to draw statistically significant conclu-sions, and showing several design problems: whichvariables and values to use, how the inclusion of al-ready thrombosed and salvaged VA affects the real out-come, and which exact aridstriCfproceaures to úsétóconduct the study.One of the main problems, already stated by several in-vestigators as Besarabs.Veselyand Krivitski, is the useof fixed values as a decision point to trigger a corree-tive action. Instead, we have used a more physiologicalapproach, which is the analysis of tendencies and inte-gration of related variables to define the course of ac-tion of every VA, taking also into account its history.Besides the "hard numeric" and good values related toassisted patency or survival curves we have obtainedfrom our work, the main outcome of a well-designedandapplied program (being the team work of MVATthe most important characteristic) is the falling of "VAproblems" far low in the list of worries and fears ofpatients, nurse s and nephrologists. During the consis-tent application ofVA S/M program, several indicatorsshow improvements: a better perceived qüality of lifefor patients, a less global cost and almost the elimi-nation.of. "crisis .situation'Lthat were.seea.añer.ezerx.;thrombosis episode in the past.The VA S/M programs must be considered as a "bridge"to a more radical solution in the futurefor the VA prob-Iem. In fact, it seems that the better approach will be theamelioration.of.tissue. response tojnjuryoLendothelialwall vessel after shuntplacement,follº~iIlgstllcliesªtmolecular and genetic levels to block this response",

BIBLIOGRAPHY1.FeldmanH, Kobrin S,Wasserstein A. Hemodialysis VascularAccess Morbidity. J Am Soc Nephrol 1996; 7:523-535.2. National Institute of Health and National Institute of

Diabetes, Digestive and Kidney Diseases. USRDS data report2002. Atlas of end-stage renal disease in the United States,2002.3. Hemphill H, Allon M. How can the use of arteriovenousfistulas be increased? Sem Dialysis 2003; Vol 16,3:214-16.4. Pisoni RL, Young EW. Dykstra DM, Grenwood RN,Hecking E, Gillespie B, Wolfe RA, Goodkin DA, Held PJ.Vascular access use in Europe and the United States: resultsfrom the DOOPS. Kidney rnt 2002: 61: 305-316.5. Young E, Dyhsfra D, Goodkin D, Mapes D, Wolfe R. HeldP. Hemodialysis vascular access preferences and outcomes inthe Dialysis Outcomes and Practice patterns Study (DOPPS).Kidney rnt 2002; 61:2266-71.6. King H, Aubert RE, Herman WH. Global burden of diabetes(1995-2025): prevalence, numerical estimates and projections.Diabetes Care 1998; 21:1414-1431.7. The Dialysis Outcomes and Practice Patterns Study. TheDOPPS report. Volume V, Issue 1,October 2003.8. Talat Alp Ikizler, Jonathan Himmelfarb. Trials and trade-off in hemodialysis' vascular access monitoring. Nephrol DialTransplant 2006; 21: 3362-63.9. Schwab SJ. Assessing the adequacy of vascular accessand its.relationship to patient outcome. Am J Kidney Dis.1994;24:316-20.10. Clinical Practice Guidelines for Vascular Access. Am JKidney Dis 2006 Jul; 48 (1 Suppl 1):S177-S247.l l.Schwab SJ, Raymond JR, Saeed M, Newman GE, DennisPA, Bollinger RR. Prevention of hemodialysis fistulathrombosis. Early detection of venous stenosis. Kidney Int1989; 36:707-711.12Winds DW, Audrain J, Vanderson r, Jendrisak Md, PicusD, Delmez JA. Optimization of high efficiency hemodialysisby detection and correction of fistula dysfunction. Kidney Int1990; 38: 337-341.13.Depner TA, Krivitski NM. Clinical measurement of bIoodflow in hemodialysis access fistulae and grafts by ultrasounddilution. ASArO J 1995;41: 745-749.14.Besarab A. Access monitoring is worthwhile and valuable.Blood Purif 2006; 24:77-89.15. Sands JJ. Vascular access monitoring improves outcomes.Blood Purif2005; 23, 45-49.16. VeseIy TM. Optimizing funS:9()l!and --.!!:.e.~tmentofhemodialysis grafts and fistulae. Seminars in interventionalradiology, Vol 21, No 2, 2004.17. McCarley P, Wingard RL, Shyr W, Hakim RM, Ikizler_TA.__Vascular. access blcod flow monitoring reduces accessmorbidity and costs. Kidney Int 2001; 60:1164-1172.18. Brouwer D. Directing care for the vascular access: Firststeps. Nephrol News Issues, 2002; 16(6)AO~41..19.: WeIch KA, Pflederer TA, Knudsen J, Hocking MK.Establishing the vascular access coordinator: Breaking groundfor-betteroutcomes.Nephrol News Issues..l998; 12(11):43-46.20. Waterhouse D. Vascular access: A role for a renal nurseclinician. EDTNA ERCA J 2002; 28(2):64-69.21. Leitch R, Ouwendyk- M, Ferguson E, Clement L, Peters -K, Heidenheim AP, Lindsay RM. Nursing issues related topatieht-Selection, vascuIiraccess,_and educaiion in quotidianhemodialysis. Am J Kidney Dis 2003;42(lSuppl):56-60.22. Hakim RM, Breyer J, Ismael N, Schulman G. Effects ofdose of dialysis on morbidity and mortality. Am J Kidney Dis1194;23: 661-669.23. Chaudhury P, Sukhatme VP, Cheung AK Hemodialysis

74

n° 2 - 2007 volumen 27 - n° 2 - 2007

Zijlstra JJ. De Jong GHT. Van Den Dorpel MA, BlankestijnPJ. Short and long-term functional effects of percutaneoustransluminal angioplasty in hemodialysis vascular access. JAm Soc Nephro12002; 13: 715-720.43. Chalmers N. The role of vascular radiology in hemodialysisaccess. Sem Dialysis 2002; 15(4): 259-268 .44. Lilly RZ, Carlton D, Barker J, Saddekni S, Hamrick K,Oser R, Westfall AO, AlIon M. Predictors of arteriovenousgraft patency after radiologic intervention in hemodialysispatients. Am J Kidney Dis 2001; 37 (5):945-953.45. Amin MZ. Vesely TM, Pilgram T. Correlation of intragraftblood fIow with characteristics of stenosis found during d iagnosticfistulography. J Vasc Interv Radio12004; 15:589-593.46. Mason R, Gunst R, Hess I. Statistical design and analysisof experiments. Wiley-Interscience, 2003.47. Li S, Hoskins PR, Anderson T, McDicken WN.Measurement of mean velocity during pulsatile fIow usingtime-averaged maximum frequency of Doppler ultrasoundwaveforms. Ultrasound Med Biol 1993; 19(2):105-13.48. GiIl RW. Measurement of blood fIow by ultrasound:accuracy and sources of error. Ultrasound Med Biol1985;11(4):625-41.49. Heiserman JE. Measurement error of percent diametercarotid stenosis determined by conventional angiography:implications for noninvasive evaluation. Am J Neuroradiol2005;26(8):2102-2107.50. Lui EY, Steinman AH, Cobbold RS, Johnston KW.Human factors as a source of error in peak Doppler velocitymeasurement. J Vasc Surg 2005; 42(5):972-979.51. Schneditz D, Wang E, Levin NW. Validation ofhaemodialysis recirculation and access blood fIow measuredby thermodilution. Nephrol Dial Transplant 1999; 14: 376-383.52. Schneditz D, Kaufman AM, Levin N. Surveillance ofaccess function by the Blood Temperature Monitor.SemDialysis 2003; 16(6); 483-487.53. Beathard GA. Improving dialysis vascular access. DialTransplant 2002; 31(4): 210-219.54. Smits JH, van der U,Hagen EC, Modderkolk-CammmeraatEC, Feith GW, Koomans HA, van den Dorpel MA, BlankestijnPI. Graft surveillance: venous pressure, access fIow, orcombination? Kidney Int 2001; 59: 1551-1558.55. Castiglioni E, Bezic J. Vascular accessforhernodíalysis.In "Ternas de Insuficiencia Renal, diálisis y transplante",(Cusumano A and Hermida O, editors), Buenos Aires, 2000.56. Grevious MA et al. Technical considerations in the

~~,HI~I'-~_'>LU\.!'j.fi.l'um;y.JlllL ..Cl:'-"r __)¿__ UJJ."""UJU':'._ .•..••.- .,,- ." .•_..__..... - .. construction - of . vascular- anastomoses. In "Biornedicaltechnology and devices handbook" (Moore J, Zouridakis G.editors). CRC Press, 2004.57. Belloni F. Teaching the principIes of hemodynamics.Advan Physiol Educ 1999; 277:187-202.58. Lopot F, Nejedly B, Válek M. Vascular access monitoring:methods and-preeeduresr=somethíng tostandardize? BloodPll.rif :20Q5;23:36·::-44.59. Hassell DRM, Van Der Sande FM, Kooman JP, TordoirJP; Leunissen KML. Optimizing dialysis doseby increasingblood flow Rate in patients with.reci.uc.ed vascular-access fIowrateo Am J Kidney Dis2001;-38E5);948-955;60. Sirken GR, Shah C, Raja R. Slow-flow venouspressure fordetection of arteriovenous graft malfunction. Kidney Int 2003;63(5):1894-1898.61. Agarwal R, Davis JL. Monitoring interposition graft venouspressures at higher blood-f1ow rates improves sensitivity in

jata reportted States,

eriovenous:214-16.vood RN,., Held Pltes: resu lts

Vascular Access Dysfunction: A Cellular and MolecularViewpoint. Roy-. J Am Soc Nephrol 2006; 17: 1112-1127.24. Garcia-Cardeña. G, Comander. JI, Blackman BR, Anderson!(R, Gimbrone Jr, MA. Mechanosensitive Endothelial Gene:Exp,reS1;¡OnProfiles. Ann NY Acad Sci 2001; 947:1-6.

Parmar KM, Larman HB, Dai G, Zhang Y, Wang ET,SN, Kratz JR. Lin Z, Jain MK, Gimbrone Jr MA.,

ía-Cardeña G. Integration of flow-dependent endothelial;"\'",nntvn,f'< by Kruppel-like factor 2. J Clin Invest 2006; Vol

fe R. Heldatcomes in(DOPPS).

HS, White S, Glagov S, Choi E, Giddens DP.Anastomotic intimal hyperplasia: mechanical

flow induced. J Vasc Surg 1992; 15: 708-716.RM, Adams MC, Clowes AW. Kinetics of vein

association with tangential stress. J Vasc5: 126-136.X. London G, Boutouyrie P; Mourad ]J, Safar M,

S. Remodeling of the radial artery in response to arne-mcrease in shear stress. Hypertension 1996;27:799-803;

NH. Ulrich M, Winther J, Pedersen EM, How T,H. The anastomosis angle does change the flow fieldsend-to-side anastomoses in vivo. J Vasc Surg 1995;

ofdiabetesirojectíons.

Study. The

and trade-ephrol Dial

llar access.idney Dis.

:m, Dennissis fistulaKidney Int

SE, Bhambhani GH. Doppler color fIow images"''''n",r, graft end-to-side distal anastomotic models.

Med Bio11993; 19: 257-267.MC, Noon GP,ij~ang NH. Flow profiles and walldistribution at a hemodialysis venous anastomosis:study. Biorheology 1987; 24: 723-735.Kidney Foundation. K/DOQI Clinical Practice

for Vascular Access, 2000. Am J Kidney Dís 2001;l):S137-S181.

••_ _.. SU, Pupim LB, Shyr Y, Hakim R, Ikizler TA.serial vascular access flow measurements. Am J

Dis 1999 Sep; 34 (3):471-477.PJ, Boereboom FT, Eikelboom BC, KoomansPI. Graft flow as a predictor of thrombosis in

llJU'i1'V';'.' grafts. Kidney Int 1998; 54: 1726-1730.NR, Ikizler TA, May RE, Himmelfarb J, Schulman

Y, Hakim RM. Change in access blood flow over timevascular access thrombosis. Kidney Int 1998; 54:

Md, Picus:modialysisKidney Int

Id valuable.

s outcomes.

eatment ofterventional RE, Himmelfarb J. Yenicesu M, Knights S, Ikizler

~--=c.c:.::.:--=:..:_G, Hermanz-Schulman M, Shyr y, Hakimmeasures of vascular access thrombosis. A~M, Ikizler

Dialysis Access Consortium (DAC) Aggrenoxof access stenosis tria\. Clinical Trials 2005; 2:

Der Linden J, Smits JHM, Assink JH, Wolterbeek DW,

75

revista de nefrología, diálisis y trasplante volumen 27 - n° :2 - 2007

predicting graft failure. Am J Kidney Dis 1999;34(2):212-7.62. Castiglioni E, Bezic J, Unzué M, Tessey A, Zahon J,Avila H, ViIlegas S, Sassone L. Vascular access monitoring:Outcomes of a multidisciplinary program (Abstraer, inspanish). XII Congreso Argentino de Nefrología, 2000.63. Bezic J, Castiglioni E, Unzué M, Tessey A, Zahon J, AvilaH, Vi llegas S, Sassone L. Vascular access biophysics as itsdisfunction origino XII Congreso Argentino de Nefrología, 2000.64. Bezic J; Rosa Diez G; Diaz C; Crucelegui S; Torres J;Mendilharzu A; Musso, C; Algranati L; Dos Ramos FariasE. Vascular access flow measurement by thermodilutiontechnique: method validation. XII Congreso Argentino deNefrología, 2000.65. Ram SJ, Work J, Caldito GC, Eason JM, Pervez A, PaulsonWD. A randomized controlled trial of blood flow and stenosissurveillance of hemodialysis grafts. Kidney Int 2003; 64: 272-280.66. Besarab A. Preventing Vascular Access Dysfunction: wichpolicy to follow. Blood Purif 2002; 20:26-35.

AppendixHere we show sorne details about.formulae and further.,explanations for the equations and concepts used in themode!.The volume of a.liquid substance that goes through agiven surface pe~Íiendicular to this movement per timeunit is called flow, and when this liquid is blood, thenthe units norrnally used in the biomedical field is ml/minoThis flow is driven by the geometric characteris-tics of the vessel, the pressure diference between thethe input and output, and sorne particular flow charac-teristics.The Hagen-Poiseuille equation describes this relatio-nship when the flow is laminar (ie, the trajectory of aparticle inside the flow its a line). In sorne cases anddepending on several factors,the blood flow goes fromlaminar to turbulent, and this equation loses validity.Anyway, it could be used with sume considerations.The components of this equation are as follows:

Q = (Pi-Poj/Rh, with

Q: blood flow. in ml/min.Pi, Po: inlet pressure, outlet pressure, in mmHg.h: viscosity, in Pa.s or Poise.L: vessel length, in cm.R: vessel radius, in cm.

For using this equation directly, is necesary ro inclu-de ~he factors converting the used units to compatibleones. This equation states several useful informations,as examples:• a short catheter has less resistance than a long one(affecting L in the equation);• a VA located in upper arm, will have more flow thanother located in the forearm, because of the diferencein Pi, the "arterial" or inlet pressure for the shunt.The model uses the equation sector by sector along thevascular access, and also for modelling the physics be-hind the extracorporeal system. The model shown inFigure 4, state also the equations we use for the calcu-lation of sevéral variables. As an example, the calcula-tion of PSart:Qa = 2000 ml/min, MAP =100 mmHg, Qb eff = 400ml/min, Rna = 0.28 mmHg.minlml

Vascular access resistances, initial values (time =0):Rart í = 30 mmHg/Qa=30 mmHg/2000 ml/min= 0.015mmHg.min/mlRg i = 35 mmHg/Qa = 0,0175 mmHg.min/mlRven i = 25 mmHg/Qa = 0.0125 mmHg.min/ml

From this values, the caIeulation for PSart is made:PSart = MAP - Rart x Qa - RN a x Qb eff =PSart=lOO mmHg - 0.015 mmHg.mín/ml x 2000 milmin - 0.28 mmHg.min/ml X 400 ml/minPSart= 100 mmHg - 30 mmHg - 112 mmHg = -42mmHg.

Then, the pressure wich the machine measures at PSart._ls.-:42mmHg, a value that could be found in real situa-tions when the patient has a very good VA (high Qa),and normal values for needles and tubing.

Recibido enforma original: U2 de febreróde2007- ....En su forma corregida: 27 de marzo de 2007Aceptación Final: 21 de mayo de 2007

Hospital Alemán - Servicio de Diálisis. Buenos AiresUniversidad Nacional de Entre Ríos - Escuela de Ingenieria de ParanáEntre Ríos - ArgentinaSerrano 505 piso 3° "A"(CI414DEK) Ciudad de Buenos Aires -ArgentinaTe!. (54 11) 4855-6706E-mail: ecastigli@fiberte!.com.ar

76