The Timing and Sequence of Multiple Device. related Complications in Patients with
Indwelling Subcutaneous Ports Christopher Young, BA, James R. Gould, MD, Paducah, Kentucky
BACKGROUND: Multiple complications associated with venous access ports are a common occur- rence. In an effort to define patterns of sequential complications in our community, we undertook a prospective analysis of adult cancer patients in whom a subcutaneous port was inserted.
METHODS: One hundred nineteen consecutive adult cancer patients in whom a subcutaneous port was inserted were observed prospectively for the development of complications.
RESULTS: Complications were identified in 70 of the 91 evaluable patients, while sequential com- plications were identified in 35 patients (38%). In aggregate, 121 complications were identified. The ball-valve effect, the most frequently identified problem, was found to occur disproportionately as a primary complication (52 of 70 versus 26 of 51, P <0.02). In contrast, port-related venous thrombosis was identified most frequently as a subsequent complication (11 of 51 versus 4 of 70, P <0.02). The only identified risk factor for the development of port-related complications was the ball-valve effect, found to be associated with the subsequent development of port-related ve- nous thrombosis (9 of 52 versus 2 of 69, P <0.02).
I ndwelling venous access devices have been available for more than 20 years. * In the oncology community, these devices have become a widely acceptable alternative to
repeated venipuncture, which can sclerose peripheral veins resulting in treatment delay, difficulty accessing specimens, and an increased risk of vesicant extravasation with sub-
From the Oncology Associates of West Kentucky, Paducah, Kentucky.
Requests for reprints should be addressed to James R. Gould, Oncology Associates of West Kentucky, 2005 Broadway, Padu- cah, Kentucky 42001.
Manuscript submitted November 14, 1996 and accepted in re- vised form April 16, 1997.
sequent tissue necrosis, z'3 Since the introduction of the per- cutaneous catheter by Broviac et al, 1 venous access devices have provided a means of repeated entry into the venous system for specimen acquisition and for the administration of cytotoxic drugs, antibodies, blood products, fluids, and nutrition. It has been estimated that approximately 500,000 long-term venous access devices are inserted into patients each year. 4 Device-related impediments such as infection, venous thrombosis, ball-valve effect, occlusion, and dis- lodgement interfere with device function, mandate addi- tional care, and may require removal, s-7 Numerous studies have described complications of indwelling venous access products. 5,s-14
In an attempt to assess accurately the patterns of port- related complications in our community, a prospective ran- domized study of the effect of warfarin sodium on compli- cations in implanted ports'in adult patients with malignant disease was begun in 1994. A provisional analysis of our population revealed that 42% of all identified events oc- curred following a primary complication. Subsequently, a completed analysis of the effects of warfarin revealed no reduction in the occurrence of port-related complications in those patients who received warfarin (data on file). We have subsequently analyzed patterns of occurrence in effort to identify causal relationships.
The current report represents a detailed case study of those patients with totally implanted port systems in which a complication developed. Our focus is on the timing and sequence of port-related problems in an adult population with malignant illness.
P A T I E N T S A N D M E T H O D S Between February 1994 and July 1996, 119 consecutive
adult cancer patients who underwent placement of a long- term subcutaneous port were identified for data collection. One hundred four of these patients participated in a pro- spective analysis of the effects of warfarin on port-related complications (data on file). Patients with a dwell time of less than 45 days and those lost to follow-up were excluded from analysis.
Ports were inserted by one of 15 surgeons with insertion site determined by the operating surgeon. Patient charac- teristics considered included age, gender, underlying diag- nosis (tumor and histology), and the presence of nonmalig- nant chronic diseases. Insertion characteristics recorded included indication, facility (hospital, inpatient or outpa- tient), surgeon, site and side of insertion, and problems at the time of implantation. Ports were assessed for longevity, complications, and removal.
DEVICE-RELATED COMPLICATIONS OF SUBCUTANEOUS PORTS/YOUNG AND GOULD]
All patients and their families were instructed in routine subcutaneous port maintenance prior to hospital discharge. Patients undergoing outpatient insertion were provided this information via either office or hospital nursing personnel.
Port care was undertaken by standard procedures. 4 Briefly, each port lumen was irrigated either monthly or after use with 3 mL to 5 mL heparinized saline (100U/mL) used to irrigate the port to ensure clearance of both the port res- ervoir and the catheter lumen, unless blood specimens were obtained through the device. In this case, the port was ir- rigated with 10 mL to 20 mL normal saline before hepar- inization to remove residual blood ceils. Dressings were ap- plied only when the port was accessed. Transparent dressings were used to allow visualization of the needle in- sertion site. Skin care consisted of application of alcohol and/or povidone-iodine scrub. Individual assessment dic- tated the appropriate procedure. Patients were asked to report any problems to the office as they developed. Office personnel and physician queries assisted in the identifica- tion of complications.
A complication was defined as any process or impediment that altered normal port performance or that mandated ad- ditional care. Complications were compiled to a maximum of four per port. Problems that occurred first were defined as "primary" complications, whereas those that occurred after a prior complication were labeled "subsequent" com- plications.
A port that failed to allow blood aspiration by a 10-mL syringe was characterized as being "unable to aspirate." Suc- cessful restoration was taken as de novo evidence of a re- versible port obstruction and was not studied further.
A port whose ability to aspirate was not restored by uro- kinase was characterized as having the "ball-valve" effect. Because of the tendency for the ball-valve effect to occur repetitively in the same port, once identified, it was consid- ered to be present for the life of the port. A port that became occluded to both flushing and aspiration was defined as "dysfunctional."
Port-related venous thrombosis was suspected if a patient developed persistent pain or swelling of the shoulders or arms or signs of a superior vena cava syndrome not related to tumor burden. The presence of thrombosis was confirmed by injection of contrast material through both the port and the ipsilateral upper extremity venous circulation. Cathe- ter-tip-associated clots or fibrin sheaths were defined as clin- ically significant and included for analysis only if accom- panied by clinical signs of vascular obstruction. Complaints of pain and swelling, not associated with port-related ve- nous thrombosis, were noted and recorded as a separate complication.
Treatment for port-related complications was adminis- tered empirically. Port-related sepsis was managed with broad-spectrum antibiotics. Sepsis was defined as the pres- ence of a positive blood culture from one or more catheter ports without another site of infection identified either clin- ically or bacteriologically. An occluded catheter port not cleared by saline instillation was flushed, thereafter, with 5,000 units urokinase after the method of Rubin. is Ports that failed to clear with urokinase were not studied further unless clinical findings suggested venous obstruction. Ra- diographically identified thrombosis involving the accom- modating vessel or superior vena cava was treated with hep-
TABLE I Patient Characteristics
Number (%) Age
27-41 years 9 (10) 41-55 years 26 (29) 55-69 years 27 (30) 69-83 years 29 (32)
Mean 59 years Range 29-82 years STD 13 Gender female 48 (53) Side of insertion
TABLE II Complicating Events in 91 Subcutaneous Ports
Events Count Incidence per Device All 121 1.30 All primary 70 .75 All subsequent 51 .55
Second 35 .38 Third 12 .13 Fourth 4 .04
TABLE III Complication Frequencies in Subcutaneous Ports
Complication Type Number (%) Ball-valve effect 78 (64) Thrombosis 15 (12) Port-related sepsis 8 (7) Other 20 (17) Total reported episodes 121 (100)
arin and warfarin in a manner analogous to that advised for lower extremity deep venous thrombosis, 16'17 a method also described as therapeutic for Groshong-catheter-associated subclavian venous thrombosis, is Warfarin was administered for the life of the port. Removal of the subcutaneous port was undertaken per preference of the attending physician, being instituted only after treatment for thrombosis (or any other complication) failed. The reason for removal of the port or for the removal of the patient from the study was obtained.
418 THE AMERICAN JOURNAL OF SURGERY • VOLUME 174 OCTOBER 1997
DEVICE-RELATED COMPLICATIONS OF SUBCUTANEOUS PORTS/YOUNG AND GOULD
Dwell time was defined as the number of days between the date of insertion to the date of port removal or the date of the patient's last appointment for patients lost to follow- up. Time to onset of the primary complication was defined as the number of days from insertion to the identification of the initial complication. Time to onset of a subsequent complication was defined as the number of days from pri- mary complication to the identification of the subsequent complication.
Data were analyzed using the Microstat'statistical package. .The evaluation included the use of the contingency table chi-square and Fisher's exact test to estimate the signifi- cance of differences between two groups. .9
R E S U L T S One hundred twenty-one subcutaneous ports were placed
in 119 patients between February 1994 and August 1996. Two patients required the insertion of a second port. The study group contained 79 Lifeports (Strato Medical Cor- poration, Beverly, Massachusetts), 7 Triumph ports (Hori- zon Medical Products, Atlanta, Georgia), and 7 Vital Ports (Horizon Medical Products). All ports studied utilize an open-tipped (nonvalved) silicone vascular access catheter.
Ninety-one patients met criteria for analysis. Evaluable patients were observed a total of 25,208 port days with a mean dwell time of 271.05 days. Patient characteristics are noted in Table I.
Overall, 121 complications were identified. Among the 70 ports with a primary impediment, 51 additional compli- cations were subsequently identified (35 second, 12 third, and 4 fourth complicating events). The overall incidence of complications was 1.30 events per port. A summary of the event count is shown in Table II.
Primary and subsequent complications were compared by type and frequency. In aggregate, the most frequently identified impediment was the ball-valve effect, consti- tuting 64% (78 of 121) of all complications identified (Table III). Among primary events, the most frequently occurring was the ball-valve effect, comprising 74% (52 of 70) of all first complications. Port-related venous thrombosis accounted for only 6% (4 of 70) and sepsis accounted for 4% (3 of 70) of initial complications. Among the 51 subsequent events, the most frequent complication was found again to be the ball-valve effect, comprising 51% (26 of 51) of all subsequent impedi- ments. Venous thrombosis followed in frequency, con- stituting 22% (11 of 51) of subsequent complications. Proportionally, the ball valve effect occurred more corn-
monly as a primary event (52/70 vs 26/51, P <0.02), whereas venous thrombosis occurred more commonly as a subsequent event (11/51 vs 4/70, P <0.02). The dis- tribution of primary and subsequent events is shown in Table IV.
The time to occurrence of primary and secondary com- plicating events was analyzed. The mean time to onset for primary complications was 58.7 days compared with a mean time of 56.0 days between a first complication and the iden- tification of a subsequent complication. Among primary complications, the ball-valve effect was first to occur, iden- tified an average 49.6 days from insertion. Venous throm- bosis was the primary complication to occur latest, an av- erage of 77.0 days from insertion. This sequencing of complications is illustrated in the Figure.
Risk factors were sought for both primary and subsequent complications. The ball-valve effect, the most commonly reported complication, was found to be a risk factor for the occurrence of venous thrombosis as a subsequent event (11 of 51 versus 4 of 66, P <0.02). Further analyses of each primary and secondary complication by type yielded no other factor associated with increased risk for a specific sub- sequent complication. Although preinsertion patient char- acteristics were investigated, neither patient age, sex, tumor type, tumor histology, surgeon, nor preexisting hyperten- sion, anemia, or atherosclerotic vascular disease was found to be of significance for increased frequency of a primary or secondary complication.
C O M M E N T S In this prospective analysis, we found that 42% (51 of
121) of all port-related complications were found to occur after, and in additition to, the identification of a primary impediment to port function. We also note that a relatively "minor" primary complication (ball-valve effect) is a risk factor for the later occurrence of a more serious Subsequent complication (port-related venous thrombosis). This se- quencing suggests a causal relationship between the ball- valve effect and venous thrombosis within venous access ports.
Our report is the only prospective analysis of multiple se- quential port-related complications. Our population of 91 patients encompassed a dwell time of greater than 68 years. The most frequently identified complication in our study was the ball-valve effect, occurring in 64% of the patientsm a rate comparable with that of other studies of venous access devices. 2° The lesser incidence of venous thrombosis (12%) and port sepsis (7%) also concur with other studies of port-
THE AMERICAN JOURNAL OF SURGERY ® VOLUME 174 OCTOBER 1997 419
DEVICE-RELATED COMPLICATIONS OF S U B C U T A N E O U S PORTS/YOUNG A N D GOULD]
1.1111
0.75 '4;"
1 ¢,1
E o.~ 8
O D. N 0.~
0.00
I ".
~- • ".
- - - - q
I I I I I
0 30 60 go 120 150
Days after Insertion
Comolk~tlon (No.~
- - N I (121)
- - - aVE (78)
Thmmbo~ (15)
Figure. The onset of the ball-valve effect precedes, and venous thrombosis follows, the average onset of all port-related complications during the life of the port. BVE -- ball-valve effect.
related' complications. 3'8'9'14'21'22 The overall incidence of complications we report may be higher than that reported by other authors. Our complication rate was influenced di- rectly by the careful manner in which we recorded any im- pediment to normal catheter function that required ther- apy, or perturbed normal function. Also, in contrast to prior studies, we recorded sequential complications, whereas we infer from our prior analyses of catheter function that earlier studies were preoccupied with the "major" complications of thrombus and infection.
A similar relationship between the ball-valve effect and venous thrombosis has been described previously in long- term indwelling percutaneous catheters. ~'z3 In the study by Tolar and Gould, 23 the ball-valve effect was also identified most frequently as a primary complication, and venous thrombosis occurred most often as a subsequent event. In addition, the ball-valve effect was an identified risk factor for subsequent venous thrombosis. It is reasonable to assume that the common thrombogenic thread is the intravascular foreign body, the tubing catheter.
A possible limitation of our study is that several of our patients participated in a concurrent investigation of pro- phylactic warfarin on port-related complications. However, because low-dose warfarin was found to have no effect on the incidence of port-related complications, we feel that the inclusion of these patients in this analysis was appropriate. Any bias engendered by the administration of low-dose war- farin would be toward a lesser incidence of complications than in a nonwarfarin population (data on file).
An awareness of complication sequencing may lead to im- proved strategies for prevention and therapy. As we have shown, a complication occurring early in the port life may have an impact on a later event. Theories on causal rela- tionships will ultimately be predicated on patterns of com- plication sequencing. Too, the utility of prophylactic regi- mens (antibiotics or anticoagulants) may be found to influence the incidence of both primary and secondary events. It is most important to realize that the sequencing of catheter-related complications we reported previously parallels the sequencing of port-related complications, z3
Studies of successful prophylactic therapies for port-re- lated complications should be interpreted with caution, and must include data on the timing and incidence of port- related complications.
R E F E R E N C E S 1. Broviac JW, Cole JJ, Scribner BH. Prolonged parenteral nutri- tion in the home. Surg Gynecol Obstet. 1974;139:24-28. 2. Freyetes CO, Reid P, Smith KL. Long-term experience with a totally implanted catheter system in cancer patients. J Surg Oncol. 1990;45:99-102. 3. Koonings PP, Given FT Jr. Long-term experience with a totally implanted catheter system in gynecologic oncologic patients. J Am Co//Surg. 1994;178:164-166. 4. Alexander HR. Vascuhar Access in the Cancer Patient: Devices, Insertion Techniques, Maintenance, and Prevention and Management of Complications. Philadelphia: JB Lippincott; 1994. 5. Ardalan B, FIores MR. A new complication of chemotherapy administered via permanent indwelling central venous catheter. Cancer. 1995;75(8):2165-2168. 6. Tolar B, Gould JR. The prognostic significance of the ball- valve effect in Groshong catheters. Support Care Cancer. 1996;4:34-38. 7. Tschithart JM, Rao MK. Mechanism and management of persis- tent withdrawal occlusion. Am Surg. 1988;54(6):326-328. 8. Puel V, Caudry M, Metayer P, et al. Superior vena cava thrombosis related to catheter malpositiofi in cancer chemo- therapy given through implanted ports. Cancer. 1993;72(7):2248-2252. 9. Mueller BU, Skelton J, Callender DPE, et al. A prospective ran- domized trial comparing the infectious and noninfectious compli- cations of an externalized catheter versus a subcutaneously im- planted device in cancer patients. J Clin Oncol. 1992;10(12):1943- 1948. 10. Denny DF. Placement and management of long-term central venous access catheters and ports. Am] Roentgenol. 1993;161:385- 393. 11. Pegues D, Axelrod P, McClarren C, et al. Comparison of in- fections in Hickman and implanted port catheters in adult solid tumor patients. J Surg Oncol. 1992;49:156-162. 12. Raad II, Bodey GP. Infectious complications of indwelling vas- cular catheters. Clin Infect DLs. 1992;15:197-210.
420 THE AMERICAN JOURNAL OF SURGERY ® VOLUME 174 OCTOBER 1997
DEVICE-RELATED COMPLICATIONS OF SUBCUTANEOUS PORTS/YOUNG AND GOULD
13. Keung YK, Watkins K, Chen SC, et al. Comparative study of infectious complications of different types of chronic central venous access devices. Cancer. 1994;73(11):2832-2837. 14. Soh LT, Ang PT. Implantable subcutaneous infusion ports. Sup- port Care Cancer. 1993;1:108-I10. 15. Rubin RN. Local installation of small doses of streptokinase for treatment of thombotic occlusions of long-term access catheters. J Clin Oncol. 1983;1:572-573. 16. Bandyk DF. Deep vein thrombosis and thrombophlebitis. In: Rakel RE, ed. Conn's Current Therapy. Philadelphia: WB Saunders; 1990:301-305. 17. Sakman EW, Deykin D, Shapiro RM, Rosenberg R. Manage- ment of heparin therapy: controlled prospective trial. NEJM. 1975;292:1046-1050.
18. Gould JR, Carloss HW, Skinner WL. Groshong catheter-asso- ciated subclavian venous thrombosis. Am.l Med. 1993;95:418-423. 19. Fleiss DL. Statistical Methods for Rates and Proportions. New York: Wiley & Sons; 1981. 20. Peters WR, Bush WH Jr. The development of fibrin sheath on indwelling venous cahteters. Surg Gyneco/Obstet. 1973;137:43-47. 21. Lokich JJ, Bothe A Jr, Benotti P, Moore C. Complications and management of implanted venous access catheters. J Clin Oncol. 1985;3(5):710-717. 22. Eastridge BJ, Lefor AT. Complications of indwelling venous access devices in cancer patients../C//n Oncol. 1995; 13( 1):233-238. 23. Tolar B, Gould JR. The timing and sequence of multiple device- related complications in patients with long term indwelling Gro- shong catheters. Cancer. 19.96;78(6):1308-1313.
THE AMERICAN JOURNAL OF SURGERY ® VOLUME 174 OCTOBER 1997 421
