Early detection and simplified management of obstructed hickman and broviac catheters

Early Detection and Simplified Management of Obstructed Hickman and Broviac Catheters By Dennis C. Stokes, Bhaskar N. Rao, Joseph Mirro, Jr, Paul W, Mackert, Brad Austin, Marci Colten, and Michael L. Hancock

Memphis, Tennessee 9 T h r o m b o t i c occlusion of Hickman and Broviac central venous catheters is a serious obstacle to their long-term use. Because resistance to flow (R) through a catheter of lumen radius, r, is proportional t o 1 / r 4, w e hypothesized t h a t measurement of R would provide an objective and sensitive monitor for partial occlusions. Our measurements showed that median R at a flow of 17 m L / m i n w a s 0.7 c m H 2 0 / m L / m i n in normally functioning Hickman catheters, and 4.1 c m H z O / m L / m i n in Broviac catheters. In obstructed catheters, which by subjective standards resisted flushing or blood withdrawal, median R was 3.0 c m H z O / m L / m i n for Hickman and 5.6 c m H z O / m L / m i n for Broviac catheters, representing significant increases. In a series of obstructed lines in which urokinase was administered, R decreased from 7.7 t o 4.5 in Hickman catheters and from 5.6 to 4.2 in obstructed Broviac catheters. The elevated resistance in Hickman catheters after urokinase suggested that residual catheter obstruction was present even though catheter function returned t o normal. Elevated R was seen with abnormal venograms in seven of 13 patients. Four patients had normal R values and abnormal venograms, and two patients had elevated R values w i t h normal venograms. Measurement of resistance in Hickman and Broviac catheters provides a simple technique that can supplement or replace venography in the serial assessment and treatment of partial obstruction. 9 1989 by (;rune & Stratton, Inc.

'. catheter occlusions. Resistance to fluid flow (R) through the catheter is approximated by Poiseuille's law for laminar flow in a rigid tubeT:

INDEX WORDS: Thrombosis; venous indwelling catheters: Hickman catheters; Broviac catheters; urokinase.

We studied pediatric ontology patients followed at St Jude Children's Research Hospital. Most of the children were enrolled in a prospective study of central venous catheters and their complications. The study was approved by the institution's clinical trials review committee, and informed consent for participation in the study was obtained from patients or, when appropriate, their parents. Catheters from a single manufacturer (Evermed, Inc, Kirkland, WA)--cither Hickman or adult Broviac type--were used in all patients. Malfunctioning catheters were defined as those with subjectively judged difficulty in flushing by a syringe or persistent inability to withdraw blood from the catheter, or both.

CCLUSION OF Hickman and Broviac central venous catheters is a serious problem in their long-term use. ~5 Difficulty with fluid infusion or withdrawal of blood from the catheter limits its usefulness and ultimately may require their premature removal. Evaluation usually requires radiographic studies of the catheter and venous system. 4'5 Because catheter obstruction may occur several times over the course of extended use, multiple radiographic studies may be needed. In most clinical situations, catheter dysfunction is recognized only when extreme difficulty with fluid infusion or withdrawal of blood through a syringe is noted. Because dangerous pressures can be generated within an obstructed catheter when using a syringe, there is a risk of rupture of the catheter. 6 Earlier determination of occlusion may prevent serious complications and early use of thrombolytic therapy may avoid early catheter failure. Changes in resistance to flow in the catheter provide a sensitive and objective method for measuring partial

O

Journal of Pediatric Surgery, Vo124, No 3 (March), 1989: pp 257-262

R = (length) • (fluid viscosity) lr • (lumen radius) 4 This equation shows that small changes in the catheter's radius will cause large changes in R when viscosity and length remain constant. For example, decreasing the radius by one half will increase R by a factor of 16. Measurements of R are thus sensitive to the early development of thrombosis and to slight decreases in catheter lumen size, whereas clinically 9 detected R usually represents extremely narrowed catheters (or total obstruction). We measured and evaluated R values in normal and malfunctioning Hickman and Broviac catheters and compared them with radiographic study results. In a series of obstructed catheters, R changes after the instillation of urokinase were also measured. MATERIALS A N D METHODS

Patients and Catheters

From the Cardiopulmonary Division, the Departments of Surgery and Hematology-Oncology, and the Biostatistics Section, St Jude Children's Research Hospital; and the Department of Pediatrics, University of Tennessee, Memphis. Supported in part by the American Lebanese Syrian Associated Charities (ALSAC). Dr Mirro is the recipient of a Clinical Oncology Career Development Award from the American Cancer Society. Date accepted: April 30, 1988. Address reprint requests to Dennis C. Stokes, MD, St Jude Children's Research Hospital, 332 N Lauderdale, Memphis, TN 38101. 9 1989 by Grune & Stratton, Inc. 0022-3468/89/2403-0007503.00/0 257

25B

STOKES ET AL

Measurement of Resistance The setup for measurement of catheter resistance is shown in Fig 1. Pressure at the distal end of the catheter was estimated by measuring the pressure ( P , in mmHg) in the catheter with no infusion. The diaphragm of a fluid-filled pressure transducer (Bently Model 800; American Pharmaceal, Irving, CA) was placed at the level of the right atrium for pressure measurements. For the serial measurements of R, this level was marked on the chest wall to ensure the same reference point for the measurements. After determination of P~, normal saline was then infused with a volumetric pump (Model 560; IVAC Corp, San Diego) that delivered at a constant rate of 16.6 mL/min. Proximal mean pressure at the catheter hub (P2, in mmHg) during ten to 15 seconds of infusion was calculated using a digital monitor (Sirecust 404; Siemens, West Germany). The data were recorded on a strip chart recorder (Western Graphtec, Irving, CA). The infusion was immediately interrupted if the peak pressure recording was > 150 rnmHg in order to prevent possible catheter rupture. For this reason, the maximum recorded R was approximately 10 emH20/mL/min. All R values were derived by use of the following equation:

f ~

,

~

-

\

R in cmH20/mL/min = [(P2 - Pl)mmHg x 1.36 cmH20/mmHg ] 16.6 mL/min

Urokinase in Obstructed Catheters Resistance in obstructed catheters was measured before and after instillation of urokinase (5,000 IU/mL) into the catheter.6 The volume of urokinase was determined by the type of catheter--1.8 to 2.0 mL for Hickman catheters and 0.8 mL for Broviac catheters. After urokinase had been in the catheter for 60 minutes, the catheter was aspirated and flushed with heparinized saline.

/

Radiographic Studies In malfunctioning catheters, the position of the catheter tip was first determined on a chest radiograph. Catheters with their tips placed in the superior vena cava (SVC) or junction of the SVC and innominate were considered in satisfactory position. Additional radiographic contrast studies were performed by hand injection of contrast agent (Conray 60; Mallinckrodt, Ine, St Louis) through the catheter during fluoroscopy.

Statistical Methods Analysis of R was based on ranked measures. To evaluate the potential correlation of repeated measures within an individual, variability within and among subjects was estimated. The results indicated that the observations were essentially noncorrelated, and the different groups were compared using the Kruskal WaUis test. Comparison of the paired preurokinase and posturokinase changes were based on the Wilcoxon signed rank test.8

RESULTS Patients and Catheters The median patient age in the study was 10.2 years (range, 3.5 to 18.2 years) for those with Hickman catheters and 2.9 years (range, 0.1 to 13.0 years) for those with Broviac catheters. Thirty-four percent of the Hickman catheters and 47% of the Broviac catheters were used in patients with acute leukemias; the remainder of the catheters were in patients with lym-

Fig 1. Setup for catheter resistance measurements. # volume infusion pump (at 17 mL/min); T, pressure transducer: P1, pressure at distal end of catheter at beginning of infusion; P2, pressure at catheter hub during infusion.

phomas or solid tumors. Sixty-nine percent of Hickman catheters had been placed surgically and 31% by the percutaneous method. All of the Broviac catheters were placed surgically. Normal Resistance Measurements R measurements were first made in a Hickman and an adult Broviac catheter before placement. Each catheter was 65 cm long. R in the Hickman catheter was 0.7 (mean of five determinations, _+.04, SD) cmH20/mL/min and 4.6 (_+.12, SD) in the Broviac catheter. This 6.2-fold difference in R between the two types of catheters was close to the value expected (6.6-fold difference) from the differences in their lumen diameters (1.6 v 1.0 mm) using Poiseuille's law. Twenty-nine measurements were made in 14 subjects with nonoccluded Hickman catheters, and ten measurements were made in nine patients with nonoccluded Broviac catheters. The median R was 0.7

OBSTRUCTED HICKMAN A N D BROVIAC CATHETERS

259

c m H : O / m L / m i n in the Hickman catheters and 4.1 c m H 2 0 / m L / m i n in the Broviac catheters. In the 29 baseline measures of R (one to four observations per subject), between-subject variability accounted for approximately 6% of the total variance.

BROVIAC

HICKMAN

>1-

9 -

In 63 measurements (22 Hickman, 41 Broviac), R was measured because of difficulty with the catheter. In the malfunctioning catheters, the median R was 3.0 c m H 2 0 / m L / m i n for Hickman catheters and 5.6 for Broviac catheters, both significantly greater than the values for normally functioning catheters (P < .001). At the time of measurement, malfunctioning catheters had been in place longer than their control catheters (Hickman median, 217 v 12.5 days, P = .002; Broviac median, 380 v 72.5 days, P = .08).

Obstructed

o - Normol function

Obstructed Catheter Studies

._c E E

E 13C

Urokinase Infusion Paired R measurements where urokinase was administered were available in five patients with malfunctioning Hickman catheters and in 13 with Broviac catheters. Urokinase was administered on more than one occasion in two patients with Hickman catheters and four with Broviac catheters. One patient had studies done on two separate Broviac catheters. R values before and after urokinase in these patients are shown in Fig 2 and Table 1, along with the clinical result of urokinase infusion. R in the Hickman catheters fell from a preurokinase median value of 7.7 c m H 2 0 / m L / m i n (range, 3.2 to >10) to a posturokinase value of 4.5 (range, 2.0 to 5.9) (P < .05). In the Broviac catheters, R fell from a median of 5.6 c m H 2 0 / m L / m i n (range, 4.1 to >10) to 4.2 (range, 2.5 to 6.8) (P < .05). In ten of 17 Broviac catheters (59%), R was elevated at the time of obstruction, but after urokinase, only one catheter had an elevated R; that catheter then had a free flow of blood. In 12 of 17 obstructions (71%) in Broviac catheters, urokinase administration successfully established free flow of blood and median R measurements returned to baseline (P = .96 for the comparison). All Hickman catheters had an elevated R, and all but one remained elevated after urokinase, even though free blood flow was reestablished in four of six (67%). Median R remained elevated compared with the baseline value (P < .01) in the Hickman catheters.

Correlation With Radiologic Studies Results of the radiologic studies are shown in Table 1 and are compared with R values obtained at the same time in these 18 patients. Several different types of abnormal venograms were seen, but the majority showed irregular filling defects around the catheter

if

BEFORE

AFTER

BEFORE AFTER

Urokinose Infusion

Fig 2. Change in catheter resistance before and after urokinase infusion. Shaded areas are the 95%-confidence limits for resistance in normally functioning Broviac and Hickman catheters. Closed circles indicate difficulty flushing or no blood withdrawal; open circles indicate normal function.

tip, consistent with thrombosis or thrombi adherent to the catheter. Three contrast studies were normal before urokinase despite elevated R values, but in two of these, urokinase successfully reestablished free flow in the catheter. Four patients had normal R values despite having clinical difficulty and abnormal venograms. The other seven patients with abnormal venograms also had elevated R values, and urokinase was successful in all seven. DISCUSSION

Catheter occlusion is a common problem with prolonged use of tunneled catheters of the Hickman or Broviac type, occurring at a rate of 4% to 10% in most reported series. Hurtubise et al 9 found that up to 25% of these catheters had to be removed prematurely because of thrombotic occlusion. In a recent series of 204 consecutive Hickman and Broviac catheter placements at our institution, catheter occlusion was the

260

STOKES ET AL

Table 1. Resistance Measurements and Urokinase Administration Catheter Type/ Patient No.

Resistance Pre-U

Post-U

% Change

Normal Function

10.0

4.6

-- 5 4

Yes

Venogram

Hickman

1

Abnormal Normal (post)

3.2

3.8

ND

0.7

-- 7 8

Yes (next day)

No

Abnormal ND

3.5

ND

--

No

Abnormal

5.3

4.4

- 17

No

Normal (post)

> 10

5.9

--41

No

Normal

>10

5.8

--42

Yes

ND

5

4.0

2.0

-- 5 0

Yes

ND

Broviac 1

5.2

4.1

- 21

Yes

Abnormal

2

5.6

4.6

- 18

Yes

Abnormal

4.9

4.3

-- 12

Yes

Abnormal

3

6.1

4.2

-- 2 4

Yes

Abnormal

4

6.7

4.1

-- 3 9

Yes

Abnormal

5.2

5.2

0

No

ND

6.9

6.8

-- 1

Yes

Abnormal

8.0

3.4

Yes

ND

5

--58

6

5.1

4.4

-- 14

No

Abnormal

*

7.8

3.0

-- 61

Yes

Abnormal

7

4.1

4.8

+ 17

No

Abnormal

8

> 10

3.4

-- 6 6

Yes

Normal

9

5.4

3.7

--32

Yes

Normal

10

5.6

4,8

- 14

Yes

ND

11

6.3

4.6

-- 14

Yes

ND

12

6.5

2.5

-61

No

ND

13

4.4

3.9

-- 11

No

ND

NOTE. Resistance in units of c m H 2 0 / m i n / m L is given for catheters before and after urokinase. Venograms were done before urokinase unless otherwise noted. Abbreviations: U, urokinase; ND, not done; Abnormal, consistent with thrombosis. *Second catheter.

second most common complication (after infection) that developed, and in 14% of cases, there was either difficulty in blood withdrawal or fluid infusion before the planned removal of the catheter. Four percent of these catheters had to be removed and replaced) ~Even though fluids may still be infused through a partially obstructed catheter, the inability to withdraw blood ("withdrawal occlusion") H is a serious limitation in the pediatric oncology patient because alternative peripheral sites for drawing blood are limited. In addition to the added patient discomfort and risks of a second catheter placement caused by obstruction, there may be a link between fibrin deposition on the catheter tip and subsequent development of catheterassociated infection and major thrombotic complica-

tions such as pulmonary emboli and thrombosis of the

SVC.12,13 There are several mechanisms for obstruction in Hiekman and Broviac catheters and that lead to increased R. 4'5'14'15 Obstruction may be due to intraluminal fibrin or other precipitates, to adherence of the catheter to the vessel wall by mural thrombi, or to buildup of extraluminal "sleeve" clots around the catheter tip (Fig 3). These clots act as a "ball-valve" and can collapse into the catheter opening during infusion, resulting in difficulty withdrawing blood. During infusion, the catheter may empty through a circuitous route; a break in the catheter may be suspected because the dye empties outside of the catheter above the level of the clot (Fig 3D). The

OBSTRUCTED HICKMAN AND BROVIAC CATHETERS

Fig 3. Mechanisms of obstruction in Hickman and Broviac catheters. (A) Intraluminal obstruction by thrombosis or deposited inorganic matter. (B) Partial obstruction of the lumen by a catheter thrombosis adherent to the venous wall. (C) Loose strands of attached thrombi capable of producing ball-valve obstruction. (D) Complete thrombosis of the catheter tip, with a false channel outside the catheter emptying proximal to the catheter end.

261

/

/

geometry and fluid mechanics of these types of obstruction are undoubtedly more complicated than the assumptions that we made to calculate R, but the Poiseuille relationship provides a useful approximation. Radiographic studies of occluded catheters are generally performed to evaluate obstruction. Unfortunately, occlusions may occur several times during the useful life of the catheter, necessitating several radiographic studies. We evaluate R values as a supplement to radiographic studies, particularly when the catheter tip is known to be in satisfactory position. Radiographic studies alone cannot be used as the standard for detecting thrombosis because some types, such as fibrin irregularities within the catheter lumen, may be missed by venography. Two patients apparently had this type of obstruction, with clearly elevated catheter R values and normal venograms. Four patients had normal R values despite having withdrawal occlusion and abnormal venograms. In those cases, the thrombosis did not interfere with the flushing of the catheter but produced a ball-valve obstruction upon attempted blood withdrawal. We used a simple setup with commonly available clinical equipment for measuring R. Despite the fact that the catheters varied in length, we were able to develop a useful range of normal values for R and to demonstrate significantly elevated R values in malfunctioning catheters, even in the majority of catheters for which there was not subjective difficulty with infusion. This study included more Broviac than Hickman catheters, but in our consecutive series of catheters, episodes of obstruction occurred with equal frequency in both types of catheters. ~~ Only a small amount of fibrin deposition or adherence to the venous wall is required to obstruct the small lumen of the Broviac catheters. Such changes are generally easy to reverse with urokinase, which was demonstrated by the return

B to baseline R and return of normal catheter function. In the Hickman catheters, obstruction may be more commonly associated with intraluminal fibrin deposition or other types of obstruction. The persistent elevations of R after urokinase (despite large percentage decreases in R) are consistent with residual obstruction in Hickman catheters by one of the mechanisms shown in Fig 3. The residual obstruction may also serve as a nidus for future problems; four of five Hickman catheters and five of 13 Broviac catheters had more than one episode of obstruction (Table 1). We recommend this sequence for evaluating partially obstructed catheters: (1) Obtain a chest radiograph to confirm good catheter tip position and rule out spontaneous tip migration. 16 (2) Measure R and compare it with normal values for Hickman or Broviac catheters, or to a normal baseline measurement for the patient. (3) If R is elevated, administer urokinase as described previously, and measure R again. (4) If R is at the baseline level and the catheter is functioning normally, no further evaluation is done. (5) If R remains elevated or catheter malfunction persists, obtain a venogram. This approach provides a rational clinical method to evaluate catheter function that may reduce the number of radiographic studies necessary in partially obstructed catheters. Early detection of fibrin deposition at the end of the catheter by a simple study such as R may prove useful in preventing catheter-related infections, occult pulmonary emboli, or development of permanent occlusion. A larger prospective study of R measurements will be necessary to evaluate the predictive merit of increased R values for complications of prolonged catheter use. ACKNOWLEDGMENT

The authors thank Dr Diane Faircloughfor reviewingthe manuscript, Ann Morris for editing it, and Chris Winston for secretarial assistance.

STOKES ET AL

262

REFERENCES

1. Hickman RO, Buckner CD, Clift RA, et al: A modified right atrial catheter for access to the venous system in m~rrow transplant recipients. Surg Gynecol Obstet 148:871-875, 1979 2. Broviac JW, Cole J J, Scribner BH: A silicone rubber atrial catheter for prolonged parenteral alimentation. Surg Gynecol Obstet 136:602-606, 1973 3. Lazarus HM, Lowder JN, Herzig RH: Occlusion and infection in Broviac catheters during intensive cancer therapy. Cancer 52:2342-2348, 1983 4. Brismar B, Hardstedt C, Jacobson S: Diagnosis of thrombosis by catheter phlebography after prolonged central venous catheterization. Ann Surg 194:779-783, 1981 5. Hoshal VL, Ause RG, Hoskins PA: Fibrin sleeve formation on indwelling subclavian central venous catheters. Arch Surg 102:353358, 1971 6. Winthrop AL, Wesson DE: Urokinase in the treatment of occluded central venous catheters in children. J Pedlatr Surg 19:536538, 1984 7. Green JF: Mechanical Concepts in Cardiovascular and Pulmonary Physiology. Philadelphia, Lea & Febiger, 1977, p 11 8. Conover W J: Practical Nonparametric Statistics. New York, Wiley, 1980, pp 229-237, 280-288 9. Hurtebise MR, Bottino JC, Lawson M, et al: Restoring

patency of occluded central venous catheters. Arch Surg 115:212213, 1980 10. Mirro J Jr, Rao BN, Stokes DC, et al: A prospective study of Hickman/Broviac catheters and implantable ports in pediatric ontology patients. J Clin Oncol (in press) 11. Lokich J J, Bothe A, Benotti P, et al: Complications and management of implanted venous access catheters. J Clin Oncol 3:710-717, 1985 12. Brismar B, Hardstedt C, Malmborg A-S: Bacteriology and phlebography in catheterization for parenteral nutrition. Acta Chir Stand 146:115-119, 1980 13. Tchekmedyian NS, Newman K, Moody MR, et al: Case report: Special studies of the Hickman catheter of a patient with recurrent bacteremia and candidemia. Am J Med Sci 291:419-423, 1986 14. Fleming CR, Barham SS, Ellefson RD, et al: Analytical assessment of Broviac catheter occlusion. J Parenter Enter Nutr 9:314-316, 1985 15. Ross AHM, Griftith CDM, Anderson JR, et al: Thromboembolic complications with silicone elastomer subclavian catheters. J Parenter Enter Nutr 6:61-63, 1982 16. Krasnow SH, Rhodes G, Boyer M, et al: Hickman catheter tip displacement. South Med J 78:1327-1329, 1985