Download Central venous catheters for haemodialysis: looking for optimal

Nephrol Dial Transplant ( 1997) 12: 1689–1691
Nephrology
Dialysis
Transplantation
Brief Report
Central venous catheters for haemodialysis: looking for optimal
blood flow
G. Jean, C. Chazot, T. Vanel, B. Charra, J. C. Terrat, E. Calemard and G. Laurent
Centre de Rein Artificiel de Tassin, Tassin, France
Abstract Central venous catheters are commonly used
for haemodialysis patients and represent, in our centre,
about 15% of the permanent vascular accesses with a
total number of more than 230 central venous catheters
over the last 10 years. Inadequate blood flow may
occur and upsets the nurses, the patients, and the
nephrologist.
The aim of this study was to identify the factors of
the catheter dysfunction. We studied prospectively 25
chronic haemodialysed patients with central venous
catheters, 14 women and 11 men, 65±16 ( 55–89 ) years
of age, treated with haemodialysis for 6.7±7 (1–26)
years. Catheters were tunnelled silicone twin catheters
( Permcath QuintonA n=18, Twincath HemotecA n=
7 ) in right (n=19) and left internal jugular (n=6)
inserted by percutaneous Seldinger techniques. We
studied the localization of the catheter tip (superior
vena cava, right atrium, right ventricular, inferior vena
cava), the central venous pressure before and after
haemodialysis, the blood pressure ( BP) before and
after haemodialysis, the interdialytic weight gain, the
number of symptomatic hypotensions during the 10
last dialyses. The patients were divided into two groups:
group I with usual adequate catheter function (n=18)
and group II with frequent dysfunctions (n=7).
Central venous pressure before dialysis was significantly higher in group I with adequate blood flow and
the catheter’s tip was more frequently found localized
in the right cardiac cavities than in the vena cava.
When central venous pressure before dialysis was over
5 mmHg, no dysfunction occurred. Blood pressure was
not dierent between the two groups. We found no
correlation between central venous pressure and BP,
interdialytic weight gain and symptomatic hypotensions. We could not predict the central venous pressure
from the mean BP but there was a higher frequency
of hypotensions in the hypovolaemic patients.
Optimal haemodynamic conditions will be provided
by a catheter tip in the right cardiac cavities and
a central venous pressure over 5 mmHg which can
be provided with vascular filling or dry weight
revaluation.
Key words: haemodialysis; blood flow; catheter; venous
pressure; right atrium
Introduction
For about 10 years, central venous catheters have been
in wide use as temporary or long-term vascular access
in haemodialysis [1–3 ]. This frequency is explained by
the ageing population and the increasing number of
diabetics patients with poor cardiac and vascular conditions, leading to a limited use of native or synthetic
arteriovenous fistulas. In case of emergency, catheters
are also useful for an immediate dialysis. The main
complications of catheters are local or general infections, venous stenosis or thrombosis [4,5]. Inadequate
blood flow episodes are certainly the most frequent
problem leading to a less ecient haemodialysis delivery. A low blood flow may arise from multiple causes:
formation of an intraluminal or periluminal catheter
clot which can be treated or prevented by general or
local heparinization, daily low-dose warfarin, local
thrombolysis, or J-guidewire insertion [6,7]; sometimes
the low blood flow is related to the catheter tip position
against the atrium or vena cava wall and the catheter
can be moved by the radiologists or has to be changed
[8,9]. Commonly, a low blood flow problem is treated
by the nursing sta initially, and sometimes successfully, by the arterial and venous lumen inversion which
may increase access recirculation, the shift of the
patient posture, or a syringe catheter aspiration.
The aim of this study was to identify some other
anatomical or haemodynamical factors that may
explain some of the dysfunction episodes resistant to
these manoeuvres and not avoided by a general
decoagulation.
Methods
Correspondence and oprint requests to: Guillaume Jean, Centre de
Rein Artificiel de Tassin, 42 avenue du 8 mai 1945, 69160 Tassin,
France.
Central venous catheters represent 15% of the permanent
vascular accesses in our centre with a total of over 230
© 1997 European Renal Association–European Dialysis and Transplant Association
1690
catheters used in the last 10 years. We have studied prospectively 25 haemodialysed patients with a double-lumen tunnelled catheter, 14 women and 11 men, 65.6±16 (mean±SD)
years old, treated with haemodialysis since 6.8±7. (1–26)
years. Nephropathies were: 3 polycystic kidney diseases, 6
diabetes, 4 interstitial nephropathies, 4 glomerular nephropathies, 1 myeloma, 1 amyloidosis, and 6 undetermined.
Haemodialysis sessions were performed three time a week
with acetate (n=18) or bicarbonate (n=7) buer, for 5 h
(n=6 ) or 8 h (n=19) and with 1–1.8 m2 cuprophane membranes. Silicone catheters were tunnelled (18 Permcath
QuintonA double-lumen with a dacron cu, and 7 Twincath
HemotecA double catheters without cu ), inserted by
Seldinger percutaneous technique in the right (n=19) or left
(n=6 ) internal jugular vein. Prescription of these catheters
had always been due to the deficiency or the absence of
arteriovenous native or prosthesis fistulae. Mean survival
rate of catheter was 12 months ( 1–52 months). Catheters
were heparinized locally with 5000 u/ml heparin and obstruction episodes were treated by local thrombolysis.
We have studied the localization of the catheter tip on a
chest X-ray: superior or inferior vena cava, right atrium, or
ventricle. We measured the central venous pressure through
the catheter arterial lumen with a water column before and
after the dialysis, after flushing internal blood, and before
saline restitution. Systolic and diastolic arterial blood pressures before and after the dialysis treatment, were also
recorded. None of the patients was taking antihypertensive
medication. These measurements were recorded twice in two
successive midweek dialysis sessions. We have recorded the
interdialytic weight gain before each session. Averages of
the two sessions were taken in account. We also recorded
the perdialytic symptomatic hypotension episodes of the 10
last haemodialysis sessions, allowing a representative vision
of the haemodynamical stability, dry weight remaining stable.
The patients were divided into two groups: group I (n=
18 ) with a catheter allowing the prescribed blood flow
( 220–300 ml/min), and group II (n=7 ), which included the
catheters with a regular dysfunction not improved by the
local thrombolysis and the use of daily low doses of warfarin
or aspirin. We verified that the two groups were homogeneous
for the central vein used, the type of catheter, age, sex,
buer, daily low dose of warfarin, diabetes, and cardiovascular events. Using the two samples t test, and specially the
Mann–Whitney-Wilcoxon non-parametric test because of the
small number of patients and the non-Gaussian distribution,
we compared the two groups for the studied parameters. We
also made a x2 test between group I and II according to the
catheter tip localization in right cardiac cavities (atrium and
ventricle) or in the vena cava.
Results
The characteristics of the two groups are shown in
Table 1. The only significant dierence is the more
frequent use of warfarin in group II related to the
more frequent episodes of low blood flow. Mean
haematocrit was 32.6%, not dierent between the two
groups. No infection occurred during the study period.
Results are shown in Table 2: systolic and diastolic
arterial BP, before and after dialysis, were not significantly dierent between the two groups but there was
a tendency to a systolic and diastolic higher BP in
group II with frequent catheter dysfunctions. We found
G. Jean et al.
Table 1. Comparisons of the two groups for central vein side ( RIJ,
right internal jugular in %), TC ( Twincath catheter in %), age in
years, diabetes ( DB, %), cardiovascular events (CV, %), sex in % of
male, acetate buer (%), patients treated with preventive warfarin
( WF ) in %
Group
RIJ
TC
Age
DB CV Male sex
Acetate
WF
I
II
P
78
72
NS
27
28
NS
64.5
66
NS
27
14
NS
83
71
NS
16
71
0.005
22 47
14 41
NS NS
no correlation between the level of systolic or diastolic
BP and central venous pressure and some patients in
group II had higher BP with lower venous pressure
than in group II. Hypotension episodes were significantly more frequent in group II (P=0.05), and associated with significantly lower interdialytic weight gain
(P<0.005), with slightly higher systolic BP (NS ) and
lower initial central venous pressures (P=0.009).
Because of haemodynamically unstable bicarbonate
selected patients, hypotension was more frequent is
this little group (n=7, mean hypotension 2.5±0.9 per
10 dialyses) than in the acetate group (n=18 mean
hypotension 0.5±0.2) P<0.05.
All but one bicarbonate-buered patient were on
5-h dialyses and all the acetate-buered patients on
8 h. So long acetate session seems to ensure a better
haemodynamic stability (P<0.05).
In group I, the tip of the catheter was more frequently localized in the right cardiac cavities than in
vena cava ( x2 test P=0.05). The localization may
aect the measured central venous pressure level. The
catheters whose central venous was above 5 mmHg,
were seldom associated with dysfunction.
Discussion
Central venous catheters are increasingly used in
chronic haemodialysis patients. Their facility of assemblage, their possible immediate use, the blood flows
achieved with a low recirculation rate are qualities
counterbalanced by the infectious and thrombosis
risks. Eectively in our experience of more than 200
catheters, access recirculation is low. Measured with
30 s low flow, we found mean 3% recirculation in
normal, and 4.5% in AV inversion. This will make a
very small dierence in urea reduction ratio ( URR
mean 79%) or Kt/V (mean 1.9) in 5- or 8-h dialyses.
Frequent low blood flows have multiple causes
linked to a thrombotic process or a mechanical problem. Among these factors, we have studied the localization of the catheter tip, (vena cava, right atrium, or
ventricle), and the central venous blood pressure measured in the catheter.
Looking for dry weight and normal blood pressure
without antihypertensive medications is among our
daily preoccupations and we sometimes maintain some
slightly hypertensive patients in relative hypovolaemia,
as is seen in group II with low initial central venous
Central venous haemodialysis catheters and optimal blood flow
1691
Table 2. Results of mean±SD for central venous pressure before (CVP1) and after (CVP2) dialysis session (mmHg); systolic ( BPs) and
diastolic ( BPd) blood pressure (mmHg); interdialytic weight gain (IWG ) ( kg), symptomatic hypotension (SH) episodes per dialysis;
catheter tip position in % of intracardiac localization. P=t test non-parametric, except x2 for position.
Group
CVP1
CVP2
BPs1
BPs2
BPd1
BPd2
IWG
SH
Position
I
II
P
7.3±1.7
0.6±1
0.009
3.7±0.9
0.1±0.9
0.1
136±7
150±11
ns
115±6
136±10
ns
65±3
82±6
ns
63±3
67±3
ns
1.5±0.2
0.6±0.1
0.005
0.6±0.3
2.4±0.9
0.05
88
57
0.05
Group I (n=28) with good blood flow and group II (n=7 ) with frequent dysfunctions.
pressure and a tendency to higher BP. Even if central
venous pressure is not a perfect reflection of hydration,
predialysis hypovolaemia is possibly related to a nonvolume-dependent hypertension.
In addition to predialytic BP recording, the evaluation of the volaemia as measured by symptomatic
hypotension may be a factor to take into account in
dry-weight determination. Our results confirm the preliminary thoughts that motivated this study: hypovolaemic patients (some patients begin their dialysis with
a central venous pressure ∏5−mmHg), have significantly more frequent low blood flow incidents than
normovolaemic patients with a positive pressure. A
characteristic of these low blood flows is their delayed
occurrence after a variable ultrafiltration duration frequently associated with a slight blood-pressure drop.
Although that seems very logical, we did not find
reported data studying this factor in the literature.
The other important factor already reported [8 ] is
the position of the catheter tip in the venous system.
Our study highlights the importance of placing the tip
of the catheter in the right cardiac cavities and ideally
in the middle of the right atrium in order to avoid
vena cava wall suction worsened by hypovolaemia.
Placing the catheter tip in the right ventricle was not
a deliberate decision. We found this position in two
cases in catheters in place for more than 3 years with
an excellent blood flow. No complications occurred
but central venous pressure was higher, with a systolic
peak. We do not recommend this localization especially
in cases needing endoluminal manoeuvre.
The type of catheter, Permcath or Twincath, the side
of the jugular vein, age, and sex, are not predictive of
catheter function.
Haemodynamic factors represent only a part of
catheter dysfunction causes with regional and local
thromboses, kinking, or wall suction phenomena.
Hypovolaemia, reflected by low central venous pressure, is one cause of catheter dysfunction in which a
simple intervention is possible.
Conclusions
Optimal haemodynamic conditions of catheter function
are achieved by the placement of the catheter tip in right
cardiac cavities and the maintenance of an initial central
venous pressure above +5 mmHg, that remains in a
physiological range. In haemodialysis patients it is easy,
using a catheter, to measure central venous pressure. A
vascular filling at the dialysis onset or later during
symptomatic hypovolaemia may represent a simple treatment for catheter low flow in hypovolaemic patients.
References
1. Blake PG, Huraib S, Wu G, Uldall PR. The use of a dual lumen
jugular venous catheters as a definitive long term access for
hemodialysis. Int J Artif Organs 1990; 13( 1): 26–31
2. Canaud B, Saunier F, Beraud JJ, Joyeux H, Mion C. La
canulation jugulaire interne avec deux cathéters silastic. Une
nouvelle méthode d’accés vasculaire pour hémodialyse.
Néphrologie 1986; 7: 57–61
3. Tesio F, De-Baz H, Panarello G, Quaia P, Raimondi A, Schinella
D. Double catheterization of the internal jugular vein for
hemodialysis: indications, techniques and clinical results. Artif
Organs 1994; 18( 4):301–304
4. Schwab SJ, Buller GL, McCann RL, Bollinger RR, Stickel DL.
Prospective evaluation of a Dacron cued hemodialysis catheter
for prolonged use. Am J Kidney Dis 1988; 11( 2): 166–169
5. Shusterman NH, Kloss K, Mullen JL. Successful use of doublelumen, silicone rubber catheters for permanent hemodialysis
access. Kidney Int 1989; 35: 887–890
6. Gibson SP, Mosquera D. Five years experience with Quinton
Permcath for vascular access. Nephrol Dial Transplant 1991;6:
269–274
7. Shrivastava D, Lundin Ap, Dosunmu B, Rao TKS, Beyer MM,
Friedman EA. Salvage of clotted jugular vein hemodialysis
catheters. Nephron 1994; 68: 77–79
8. McGonigle DJ, Shrock LG, Hickman RO. Experience using
central venous access for long-term hemodialysis. Am J Surg
1983; 145: 571–573
9. Moss AH, McLaughlin MM, Lempert KD, Holley JL. Use of
silicone catheter with a Dacron cu for dialysis short-term
vascular access. Am J Kidney Dis 1988; 12(6 ): 492–498
Received for publication: 9.1.97
Accepted in revised form: 11. 4.97