1. 969
Dialysis Is Not Indicated Immediately
After Administration of Nonionic
Contrast Agents in Patients with
End-Stage Renal Disease Treated
by Maintenance Dialysis
. .‘ c
Carol M. Younathan1 OBJECTIVE. This study was undertaken to determine the necessity of immediate dial-
Jun V. Kaude2 ysis after intravascular injection of contrast material in patients with end-stage renal dis-
Michael D. Cook3 ease who are being maintained on hemodialysis. Although many physicians support
this practice, we could find no reports of studies to confirm or refute its necessity.
Geraldine S. Shaw3
SUBJECTS AND METHODS. We studied 10 patients being treated with hemodialy-
John C. Peterson3 sis who were undergoing 11 diagnostic procedures that required intravascular con-
trast material. The patients received 40-225 ml of nonionic contrast material and were
followed up with clinical examination and laboratory analysis to determine any
adverse effects from contrast administration or the need for dialysis or both.
RESULTS. No significant changes in blood pressure, ECG, total serum protein level
or osmolality, extracellular fluid volume, or body weight occurred after injection of
contrast material. None of the patients had clinical features that necessitated emer-
gent dialysis.
CONCLUSION. We conclude that nonionic contrast material can be given safely to
patients with end-stage renal disease who are being maintained on hemodialysis.
Immediate postprocedural dialysis is unwarranted as a routine practice.
AJR 1 994;163:969-971
Contrast media can cause renal dysfunction [1 , 2]. The use of these agents
also is associated with other complications, including intravascular volume expan-
sion, pulmonary edema, depression of myocardial contractility, and arrhythmias
[3-6]. Contrast medium is almost exclusively excreted unchanged in the urine by
glomerulan filtration, and one could postulate that patients with end-stage renal
disease are at risk of adverse hemodynamic on cardiovascular symptoms devel-
oping until these agents are removed. Contrast agents can be removed efficiently
with hemodialysis.
The interval after contrast administration when these risks become clinically
Received March 11 , 1994; accepted after revi-
significant, if ever, is poorly understood. Many radiologists advocate immediate
sion May 23, 1994.
dialysis after a procedure involving contrast material on the grounds that the con-
This work was supported by Mallinckrodt
Medicals, St. Louis, MO. trast material may precipitate a potentially dangerous hemodynamic or cardiovas-
Presented at the annual meeting of the American
culan state.
Roentgen Ray Society, San Francisco, April 1993. We undertook a prospective clinical study to determine if this approach was jus-
1Diagnostic Imaging Center, 6716 NW. 11th P1., tified when nonionic contrast agents were administered to patients with end-stage
Gainesville, FL 32605. Address correspondence to renal disease. We selected a nonionic contrast agent for this investigation to mini-
C. M. Younathan.
mize the incidence of possible side effects in this group of high-risk patients.
2Department of Radiology, University of Florida
College of Medicine, Gainesville, FL.
3Division of Nephrology and Hypertension, De- Subjects and Methods
partment of internal Medicine, University of Florida
College of Medicine, Gainesville, FL. We studied 10 successive patients with end-stage renal disease who had 11 radiographic
0361-803X/94/1634-969 procedures that required intravascular administration of contrast material. All the patients
© American Roentgen Ray Society were maintained on a routine hemodialysis schedule of three times a week. Their etiology

2. 970 YOUNATHAN ET AL. AJR:163, October 1994
for renal failure varied: six developed renal failure from hyperten- monary edema, cardiac dysfunction or arrhythmias, or who developed
sion, two from diabetes mellitus, one from chronic pyelonephnitis, hyperosmolar syndrome would undergo emergent dialysis. Otherwise
and one from acute tubular necrosis. Approval from the hospital’s the patients had dialysis at their next routinely scheduled session.
institutional review board was obtained before commencement of Statistical analysis was done by repeated measures analysis of
the project, and informed consent was obtained from all patients. variance (ANOVA) to analyze the response time profiles of serial sys-
Participation was voluntary. tolic and diastolic blood pressure, serum osmolality, and total serum
Patients’ tolerance of the contrast load was monitored closely. All protein levels. Tests were done to detect (1) the presence of any
subjects were inpatients routinely maintained on hemodialysis change over time in mean response profiles, (2) differences before
because of end-stage renal failure. Baseline blood pressure, ECG, and after administration of contrast material, and (3) the presence of
total serum protein level and osmolality, average weight gain linear, quadratic, or cubic time trends in the mean response profiles.
between two successive sessions of dialysis, and clinical assess- The response profiles also were analyzed in the repeated-measures
ment of extracellular fluid volume (pedal edema, cardiac arrhythmia, ANOVA framework, adjusting for volume of contrast material infused
and pulmonary edema, if any) were recorded. Three patients had as a linear covariant.
CT, five had cardiac catheterizations, two had angiography of the
peripheral vasculature, and one patient underwent a dialysis fistulo-
gram. All received Optiray 320 (Mallinckrodt Medical, St. Louis, MO) Results
as the contrast medium (150 ml IV for CT and a minimum dose of
40-225 ml to obtain a diagnostic angiographic study). None of the patients had postprocedunai side effects that
All patients were followed up by clinical and laboratory evaluation. warranted dialysis before the next routinely scheduled ses-
Physical examinations were performed twice daily (in the morning and sion (Table 1). No ECG changes were observed after injection
evening) to assess for symptomatic cardiac dysfunction, pulmonary of contrast material, except for one patient (case 1 , Table 1)
edema, and pedal edema. Blood pressure was taken at 4-hr intervals, who had premature ventricular contractions. These were
and ECG was done at 4 and 12 hr after administration of contrast asymptomatic, and he underwent dialysis 24 hr after contrast
medium. Laboratory assessment included serum osmolality and total
administration without additional problems. No patient had a
serum protein levels (assayed at 4, 12, 24, and 36 hr after injection of
clinically significant increase in blood pressure. Systolic blood
contrast medium for serum osmolality and 4 and 12 hr after injection for
pressure rose after injection of contrast material in three
total serum protein levels). Total serum protein levels, changes in body
weight, and physical examination were used to objectively assess intra- patients, but in none was this rise more than 25 mm Hg. In all
vascular and extravascular volume (Table 1). Serial EGG, serum osmo- patients, systolic blood pressure returned to baseline by 20 hr
lality, and total serum protein levels were monitored for possible after the contrast material was injected. The mean change in
hyperosmolar syndrome. Patients who became symptomatic from pul- systolic blood pressure was statistically significant only for
TABLE 1 : Blood Pressure, Weight Gain, and Laboratory Data Before and After Administration of Contrast Material in 1 1 Cases
(10 Patients) with End-Stage Renal Disease
Average % Total Serum Protein Serum Osmolality
Level (g!dl) (mOsm!kg) Weight Gain Time Between
Volume of Increase or
Between Contrast
Case Etiology of Contrast Decrease in Average Average Dialysis Administration
No. Renal Failure Study Material Blood Pressure Before Change Before Change Treatments and Dialysis
(ml) (systolic! Contrast After Contrast After (kg) (hr)
diastolic) Contrast Contrast
1 Hypertension Arteniovenous 85 8.L! 7’!’ 7.0 0.3k 302 101 1.1 24
fistulogram
2 Hypertension CT, chest 125 6! 41 4.5 O.2J 277 41’ 1.5 19
3a Hypertension CT, abdomen/ 150 14! 81 5.2 0.3k 305 21. 11.0 47
pelvis
4 Diabetes mellitus Angiogram 225 81.! 9J. 6.7 0.6J. 297 1.1. 5.0 22
5 Acute tubular CT, abdomen 150 3J! 4 5.4 0.21 312 51’ 0.8 24
necrosis
6 Hypertension Cardiac 140 8V191’ 6.9 oil’ 309 31’ 2.0 16
catheterization
7 Hypertension Cardiac 185 1OJJ10.I 7.3 0.9J 314 7.L 3.8 22
catheterization
8 Chronic Angiogram 40 41 J/21 J 6.2 0.6w 298 41’ 0.2 18
pyelonephritis
9 Hypertension Cardiac 150 17V15J 6.2 O.7J 302 21 0.9 21
catheterization
10 Hypertension Cardiac 150 g’ri 8J 6.5 O.7J 294 61’ 1.2 20
catheterization
11 Diabetes Cardiac 120 2! 0 6.7 0.51 304 41 3.1 20
mellitus catheterization
aSame patient as in case no. 2.
1= Increase, .i.= decrease.

3. AJR:163, October 1994 CONTRAST AGENTS AND END-STAGE RENAL DISEASE 971
patients receiving 150 ml or more of contrast material and gency basis and after regular working hours because of the
was not clinically important. Diastolic blood pressure showed fear of the theoretical risks of delay.
an overall subtle but statistically significant increase 20 hr Despite delaying hemodialysis, and presumably increasing
after injection of contrast material, but this disappeared when the risk of adverse effects by prolonging exposure, none of
an adjustment was made for the volume of contrast material our patients had problems that required unscheduled dialysis.
injected. Mean serum osmolality and total serum protein level Total serum protein level was used as an easily measured
varied inconsistently; the former slightly increased in eight parameter that correlates with changes in volume status [8],
patients, and the latter slightly decreased in eight. The serum and serum osmolality was used as a guide to the effect of the
osmolality increased 4 hr after contrast administration in all osmotic load of the contrast material. A baseline value was
but one patient. This statistically significant increase also dis- obtained, and serial measurements were made after the con-
appeared with volume adjustment. Again these changes, trast material was injected. Neither showed significant clinical
although statistically significant, were of no clinical conse- change. Likewise, clinical assessment of volume and hemo-
quence. All patients gained weight (0.2-11 .0 kg), but only one dynamic status revealed no significant complications that
had an increase greaten than his or hen expected weight gain necessitated emergent dialysis or any other intervention in
between dialysis sessions (case 3, Table 1). This patient’s any patient.
dialysis was delayed because of a gastrointestinal hemor- One criterion for inclusion in this study was that the partici-
rhage that occurred after injection of contrast material. Dialy- pants be inpatients. Thus close clinical monitoring could be
sis was not possible, because of hemodynamic instability done, and serial, scheduled ECGs and laboratory data could
(hypotension), until 47 hr after contrast injection. He gained 11 be obtained. However, the results indicate that hospitalization
pounds primarily from blood and colloid transfusions for vol- is not necessary solely to monitor patients after injection of
ume expansion. The average time interval from contrast contrast material. Performing procedures that require injec-
administration to hemodialysis was 23 hr (range, 16-47 hn). tion of contrast material on an outpatient basis would signifi-
Overall, although some trends were statistically significant, cantly reduce health cane costs.
the changes in the response variables oven time were subtle In conclusion, the results of this study suggest that non-
and were of no clinical consequence. ionic contrast agents may be given safely to patients with
end-stage renal disease being maintained on hemodialysis
without the added expense or inconvenience of emergent
Discussion
postprocedural dialysis. To increase the validity and clinical
This prospective study was done to evaluate a common din- application of these data, additional investigation in a larger
ical practice: immediate dialysis of patients with end-stage renal number of cases is warranted.
disease after injection of contrast material. We were unable to
find sufficient published data to confirm or rebut this conviction, ACKNOWLEDGMENT
which further emphasizes the importance of this study. Further-
The authors thank Paul S. Kubilis for the statistical analysis.
more, physicians (radiologists and clinicians alike) should seek
not only to render state-of-the-art care to the patient, but in
REFERENCES
regard to cost-effectiveness, to avoid unnecessary diagnostic
and therapeutic procedures. Hemodialysis is an expensive pro- 1 . Bems JS, Rudnick MR. Radiocontrast media associated nephrotoxicity. In:
Humes, HD, ed. The kidney (National Kidney Foundation suppi.). Bethesda,
cedune, in particular when performed as a nonscheduled emer-
MD; National Institutes of Health, 1992;24:1-6
genttreatment at odd hours ofthe day or night. 2. Taliercio CP, Vliestra RE, Fisher LD, Burnett JC. Risks for renal dysfunc-
In 1990 approximately 200,000 patients in the United tion with cardiac angiography. Ann Intern Med 1986;104:501-504
States had end-stage renal disease, and the prevalence is 3. Hirshfeld JW Jr. Low-osmolality contrast agents: who needs them? N EngI
rising [7]. Forty percent of these patients have coronary J Med 1992;326:482-484
4. Steinberg EP, Moore RD, Powe NR, et aI. Safety and cost effectiveness of
artery disease, and there are many other associated condi- high-osmolality as compared with low-osmolality contrast material in
tions and complications, such as peripheral vascular dis- patients undergoing cardiac angiography. N EnglJ Med 1992:326:425-430
ease, seizures, and access malfunction [6], which increase 5. Barreft BJ, Parfrey PS, Vavasour HM, O’Dea F, Kent G, Stone E. A compar-
the likelihood of the patient undergoing a study that uses ison of nonionic, low-osmolality radiocontrast agents with ionic high-osmola-
lity agents during cardiac catheterization. N EngIJ Med 1 992;326:431-436
contrast material.
6. Held PJ, Port FK, Gaylin DS, et al. Comorbid conditions and correlations
The uncertainty of the delayed effects of contrast material with mortality risk among 3,399 incident hemodialysis patients. in: Agodoa
in this subset of patients has created considerable difficulty LYC, Held PJ, Port FK, eds. US renal data system 1992 annual report.
and inconsistency in obtaining optimal images and planning Bethesda, MD: National institutes of Health, 1993:21-28
7. Agodoa LYC, Held PJ, Port FK. Executive Summary. In: Agodoa LYC,
follow-up care. Studies are sometimes performed without
Held PJ, Port FK, eds. US renal data system 1993 annual data report.
contrast material, resulting in a suboptimal examination, on Bethesda, MD: National Institutes of Health, 1993:7-17
postponed because dialysis cannot be appropriately coordi- 8. Gowenlock AH. varley’s practicalciinicalbiochemistr,’, 6th ed. Boca Raton,
nated. Conversely, dialysis has been performed on an emen- FL: CRC Press, 1988:419-422