Peripherally inserted central venous catheter-associated thrombosis: retrospective analysis of clinical risk factors in adult patients.

Background: Peripherally inserted central catheters (PICC) are common venous access devices. Clinical conditions and therapies that increase the risk of PICC-associated thrombosis have not been studied.

Methods: We performed a retrospective case-control analysis of all adult patients who underwent placement of a PICC at our hospital over a three-year period (n = 1296). Clinical variables examined were indication for PICC placement, active cancer treatment, history of DVT, diabetes mellitus, and use of prophylactic anticoagulation.

Results: The overall incidence of PICC-associated DVT was 2% (n = 27). Active cancer therapy was significantly associated with PICC-associated DVT (OR 3.5, 95% CI 1.3-9.8). The use of prophylactic anticoagulation did not reduce this risk.

Conclusions: Patients who suffered a PICC-associated DVT were more likely to be undergoing treatment for cancer. This risk was not lowered by the use of prophylactic anticoagulation. These results suggest a need for prospective studies on effective anticoagulation for patients at high risk for PICC-associated DVT.

Key Words: thrombosis, peripherally inserted central catheter, PICC, DVT

**********

The use of peripherally inserted central catheters (PICCs) for venous access has increased substantially over the past decade. (1) PICCs can be used for the infusion of chemotherapy, parenteral nutrition and antibiotics, and provide convenient venous access for both in-hospital use and at-home therapy. In many institutions, these devices have replaced central venous catheters as the first choice for long-term IV therapy. (2,3) As healthcare costs rise and the movement toward outpatient care continues, the need for long-term venous access and PICC use can be expected to steadily increase.

One complication of long-term venous access is venous thrombosis. While extensive literature exists regarding the incidence, risk factors, and prevention of venous thrombosis associated with central venous catheters (CVCs), (4,5) limited data are available on this subject with respect to PICCs. PICC-specific risk factors for the development of deep venous thrombosis (DVT) have been identified, (2,3,6) but specific medical conditions and therapies that are risk factors for PICC-associated venous thrombosis have not been well defined. In addition, little is known regarding prophylactic regimens that may mitigate the risk of PICC-associated venous thrombosis.

In an effort to identify risk factors associated with PICC-related DVT in an adult population, we performed a retrospective case-control analysis of our institution's patients who experienced a PICC-related DVT. We also reviewed the use of anticoagulation in conjunction with PICC use in a case-control manner as a preliminary step in the evaluation of prevention of PICC-associated venous thrombosis.

Methods

Design

We included in our retrospective study all patients aged 18 years or older who had a PICC inserted at our hospital between January 1, 2000 and June 30, 2004. To limit variation in catheter type, size, and placement technique, the data set was limited to catheters placed only by the vascular nurse PICC team. The PICC team followed these patients until the PICC line was discontinued or until a complication occurred. Follow-up care of eligible patients in this study was facilitated by the closed nature of our military healthcare system, in which patients receive the majority of their medical care at one institution. All of the catheters were placed at the bedside by one of two registered nurses who comprised the PICC team during the study time period. PICCs were placed via a modified Seldinger technique using portable ultrasound guidance. The basilic vein was the vessel of choice during PICC placement. Per PICC team protocol, catheter tip location in the lower one-third of the superior vena cava was confirmed by chest x-ray at the time of PICC placement. The study was approved by the institutional review board of the 59th Medical Wing, Wilford Hall Medical Center.

Identification of Case and Control Patients

A PICC-associated DVT was defined as a symptomatic event prompting a duplex ultrasonography study which confirmed venous thrombosis. Although ultrasound interpretation was not standardized, cases were included only if the original radiology interpretation noted noncompressibility of the vein and presence of echogenic thrombus associated with the PICC line or at the site of a PICC line that had been discontinued no more than 10 days before the radiologic study. In addition, we reviewed all of the upper extremity duplex ultrasonography reports at our institution during the study time period to identify any PICC-associated DVTs not recorded in the PICC team files. Case patients were individually matched to two age- and gender-matched controls selected randomly by computer from the PICC team records.

Data Collection

We reviewed the medical records of the cases and controls for the following data: indication for PICC as denoted by the physician requesting PICC placement (antibiotics, chemotherapy, parenteral nutrition, or other venous access), diameter of PICC, diabetes mellitus, history of prior venous thromboembolism, active cancer therapy, and use of prophylactic anticoagulation. In patients who suffered a PICC-associated DVT, we also noted the duration of time between PICC placement and development of thrombosis. Diabetes mellitus was defined as any clinical laboratory-drawn glucose value of greater than 200 mg/dL or a history of outpatient prescriptions for oral diabetes mellitus medications or for insulin. A history of prior venous thromboembolism (DVT or pulmonary embolus) was obtained from review of the patient's chart. Active cancer therapy was defined as the administration of chemotherapy or radiotherapy within 30 days before PICC insertion. Prophylactic anticoagulation was defined as unmonitored vitamin K antagonist use (1 mg daily of warfarin) for the duration of PICC use. The effect of prophylactic unfractionated or low-molecular weight heparin was not investigated, as these agents were generally limited to the inpatient arena and their use did not extend through the duration of the PICC dwell time.

Statistical Analysis

Demographic characteristics and potential risk factors for case patients and controls were compared for continuous variables with the Student t test and for dichotomous variables with either the Pearson chi-square (for large sample sizes) or Fisher exact (for small sample sizes with cells containing less than five statistics or observations) tests. The odds ratio (OR) was used to estimate relative risk. Logistic regression analyses were attempted for variables that were deemed risk factors by univariate analyses. All statistical analyses were performed using the statistical software package SPSS for Windows version 12.0 (Chicago, IL).

Results

A total of 896 patients underwent placement of 1296 PICCs. The study population included 27 case patients and 54 age- and gender-matched controls, for an overall incidence of symptomatic PICC-associated DVT of 2%. Table 1 summarizes the characteristics of the case patients and controls. There were no differences between the two groups with respect to catheter size, type of infusate, presence of diabetes, or history of prior thrombosis. PICC-associated DVTs occurred more commonly in patients undergoing active cancer treatment. However, the type of malignancy (hematologic versus solid tumor) was not different between the two groups (data not shown). The most common indications for PICC placement in both cases and controls were antibiotics, chemotherapy, and parenteral nutrition. The median time from PICC placement to diagnosis of DVT was 15 days (range 2-155). None of the patients who suffered a PICC-associated DVT had a recurrent DVT during the study time period.

Odds ratios (ORs) for the development of venous thrombosis for each of the postulated clinical risk factors, along with their 95% confidence intervals (CI), were calculated (Table 2). The development of venous thrombosis was significantly associated with active cancer treatment at the time of PICC placement (OR 3.5, 95% CI = 1.3-9.8) and with the use of prophylactic (1 mg daily) warfarin (OR 5.9, 95% CI = 1.4-25.3). These factors were highly correlated ([r.sup.2] = 0.53, P < 0.001), making a logistic regression analysis using both of these variables statistically invalid.

Discussion

PICC lines are increasingly common venous access devices used in hospitalized adult patients. A significant complication of long-term venous access is associated thrombosis, which was symptomatic in approximately 2% of patients in our study. This is consistent with prior studies of symptomatic PICC-associated DVT. (1,6,7) While PICC-specific qualities that increase DVT risk have been identified, (2,3,6) this study helps to define patient-specific risk factors. Specifically, to our knowledge, our study is the first to show that patients receiving current treatment for cancer are at increased risk for PICC-associated DVT and that this risk is not reduced by the use of commonly used prophylactic anticoagulation.

In our study, the risk for developing a PICC-associated DVT appeared to be influenced by one specific factor; namely, active cancer treatment. Statistically, the use of prophylactic dose warfarin was associated with thrombosis. However, clinically, this association is likely attributable to the fact that in our hospital during the time of this study, patients receiving active cancer treatment who had indwelling PICCs were frequently given low doses of warfarin for DVT prophylaxis. Multivariate analysis was not performed in our study due to the interdependence of these two variables. One prior investigation found a diagnosis of cancer to be a risk factor for PICC-associated DVT in univariate analysis, but the study included PICCs of multiple diameters and, in multivariate analysis, malignancy was not found to be a risk factor for thrombosis. (6) In comparison, cancer has been shown to be a risk factor for nonperipherally inserted central venous catheter-associated venous thrombosis. (4)

Although patients undergoing active cancer treatment experienced a higher incidence of thromboses, the infusion of chemotherapy versus other solutions through the PICC, as indicated by the ordering physician, did not appear to increase this risk. Therefore, although these patients appear to be at increased risk for PICC-associated DVT, this elevation in risk may be more attributable to an inherent hypercoagulable state, rather than to the toxic nature of the chemotherapy solution, as has been previously postulated. (6)

Patients with a history of DVT or pulmonary embolism have been shown to be at increased risk for thrombosis recurrence. (8,9) In addition, a retrospective analysis showed that a history of lower extremity DVT was independently associated with an increased risk of upper extremity PICC-associated DVT. (10) In our study, there was a trend toward an increased risk in this population. These patients may indeed be at increased risk for recurrent thrombosis, but the small number of events in our study precludes robust statistical analysis.

It has been postulated that diabetes mellitus is an independent risk factor for venous thrombosis, via impaired endothelial function and an inherent prothrombotic state. (11) In our study, approximately one-half of the patients carried a diagnosis of diabetes mellitus. It does not appear from our data that diabetes mellitus is associated with PICC-associated DVT, although additional qualifying information such as level of glycemic control, duration of disease, and other markers of endothelial dysfunction were not evaluated.

Most cases of upper extremity DVT are thought to be asymptomatic. (2,10,12,13) Reports of patients with PICC-associated thromboses suggest that only a small minority of these patients experience symptoms related to the thrombosis. (2,13) While the majority of PICC-associated DVTs are not clinically evident, the sequelae of both symptomatic and silent thromboses remain significant. Findings suggestive of a pulmonary embolus (PE) on perfusion scanning were present in 36% of patients with upper extremity DVT in one study (10) and evidence of the postthrombotic syndrome (comprised of upper extremity pain, functional limitation, or skin changes on physical examination) was present in 15%. (10) As the complications of PICC-associated DVT can be significant, prophylaxis seems a reasonable option to attempt to prevent PICC-associated DVT in those patients at highest risk. Although several studies in patients with malignancy and nonperipherally inserted central venous catheters suggest that prophylactic anticoagulation with low-dose warfarin may reduce the incidence of DVT, (14,15) a more recent prospective trial of low-dose (1 mg) unmonitored warfarin therapy performed by Heaton and colleagues showed no reduction in the risk of thrombosis in cancer patients with central venous catheters. (16) In addition, published clinical practice guidelines recommend against the use of prophylactic anticoagulation in patients with central venous catheters. (17)

Much of the available literature for preventing catheter-related thrombosis is specific for nonperipherally inserted central venous catheters and we are not aware of any existing data regarding methods or efficacy of pharmacologic prevention of upper extremity DVT in patients with PICCs. To our knowledge, our study results are the first to show that low-dose warfarin appears to confer no benefit for the prevention of PICC-related thrombosis in this population. The lack of benefit of one prophylactic anticoagulation regimen (1 mg a day of unmonitored warfarin) in this patient cohort may be related to the nature of the peripheral catheter. It has been observed that peripheral ports confer an increased risk for thrombosis when compared with central ports. (18,19) The proposed mechanism behind this observation is that a larger percentage of the luminal diameter of the vein is filled, (20) and that a greater length of the venous system is traversed (18) by the peripherally placed catheter.

Previously identified PICC-specific risk factors for the development of upper extremity DVT include increasing diameter of PICC, (6) the use of polyvinyl chloride rather than silicone-coated catheters, (12) and longer catheter dwell times. (3) Although noted to be a risk factor for PICC-associated DVT in other studies, (6) catheter diameter effect on PICC-related DVT is beyond the scope of our study as the vast majority (over 85%) of catheters placed in our study were double-lumen, 5-French catheters. The only other catheter size used was a 4-French catheter. The standardized catheter size used at our institution makes it unlikely that this variable would be found to be an independent risk factor for the development of thrombosis in our patient cohort. We deliberately chose to study this homogeneous population to eliminate the contribution of PICC-specific risk factors and to evaluate the effect of specific medical conditions and anticoagulation regimens on the incidence of PICC-associated DVT.

There are limitations to our study. The retrospective single-center design has inherent limitations and may have resulted in bias and possibly erroneous conclusions, and the results should only be considered as hypothesis generating. The true incidence of DVT in this population is probably higher than that reported for several reasons. Prior investigation has established that the diagnosis of DVT via ultrasound is less sensitive in the upper extremity relative to that of the lower extremity. (4,21) In addition, venous thrombosis was only recorded if the patient presented with symptoms and underwent upper extremity ultrasonography in association with an indwelling PICC.

Conclusion

Our results help define patient characteristics that may predispose to PICC-associated DVT. Specifically, patients receiving therapy for cancer appear to be at increased risk of PICC-associated DVT. Our investigation suggests that one commonly used prophylactic anticoagulation regimen in these higher risk patients may not be effective. As PICC use is likely to increase, prospective studies are needed both to define the actual incidence of PICC-associated DVT in high-risk populations and to investigate means of preventing PICC-associated DVT.

Acknowledgments

Dr. Anneke Bush, Clinical Research Support Squadron, Wilford Hall Medical Center

References

1. Smith JR, Friedell ML, Cheatham ML, et al. Peripherally inserted central catheters revisited. Am J Surg 1998;176:208-211.

2. Allen AW, Megargell JL, Brown DB, et al. Venous thrombosis associated with the placement of peripherally inserted central catheters. J Vase Interv Radiol 2000;11:1309-1314.

3. Gonsalves CF, Eschelman DJ, Sullivan KL, et al. Incidence of central vein stenosis and occlusion following upper extremity PICC and port placement. Cardiovasc Intervent Radiol 2003;26:123-127.

4. Verso M, Agnelli G. Venous thromboembolism associated with long-term use of central venous catheters in cancer patients. J Clin Oncol 2003;21:3665-3675.

5. Joffe HV, Kucher N, Tapson VF, et al. Upper-extremity deep vein thrombosis: a prospective registry of 592 patients. Circulation 2004;110:1605-1611.

6. Grove JR, Pevec WC. Venous thrombosis related to peripherally inserted central catheters. J Vase Interv Radiol 2000;11:837-840.

7. Ng PK, Ault MJ, Ellrodt AG, et al. Peripherally inserted central catheters in general medicine. Mayo Clin Proc 1997;72:225-233.

8. Prandoni P, Lensing AW, Cogo A, et al. The long-term clinical course of acute deep venous thrombosis. Ann Intern Med 1996;125:1-7.

9. Heit JA, Mohr DN, Silverstein MD, et al. Predictors of recurrence after deep vein thrombosis and pulmonary embolism: a population-based cohort study. Arch Intern Med 2000;160:761-768.

10. Prandoni P, Polistena P, Bernardi E, et al. Upper-extremity deep vein thrombosis: risk factors, diagnosis, and complications. Arch Intern Med 1997;157:57-62.

11. Fonseca V, Desouza C, Asnani S, et al. Nontraditional risk factors for cardiovascular disease in diabetes. Endocr Rev 2004;25:153-175.

12. Shah MK, Burke DT, Shah SH. Upper-extremity deep vein thrombosis. South Med J 2003;96:669-672.

13. Luciani A, Clement O, Halimi P, et al. Catheter-related upper extremity deep venous thrombosis in cancer patients: a prospective study based on Doppler US. Radiology 2001;220:655-660.

14. Bern MM, Lokich JJ, Wallach SR, et al. Very low doses of warfarin can prevent thrombosis in central venous catheters: a randomized prospective trial. Ann Intern Med 1990;112:423-428.

15. Boraks P, Seale J, Price J, et al. Prevention of central venous catheter associated thrombosis using minidose warfarin in patients with haematological malignancies. Br J Haematol 1998;101:483-486.

16. Heaton DC, Han DY, Inder A. Minidose (1 mg) warfarin as prophylaxis for central vein catheter thrombosis. Intern Med J 2002;32:84-88.

17. Geerts WH, Pineo GF, Heit JA, et al. Prevention of venous thromboembolism: the Seventh ACCP Conference on Antithrombotic and Thrombolytic Therapy. Chest 2004;126(3 Suppl):338S-400S.

18. Kuriakose P, Colon-Otero G, Paz-Fumagalli R. Risk of deep venous thrombosis associated with chest versus arm central venous subcutaneous port catheters: a 5-year single-institution retrospective study. J Vase Interv Radiol 2002;13:179-184.

19. Shetty PC, Mody MK, Kastan DJ, et al. Outcome of 350 implanted chest ports placed by interventional radiologists. J Vase Interv Radiol 1997;8:991-995.

20. Bodner LJ, Nosher JL, Patel KM, et al. Peripheral venous access ports: outcomes analysis in 109 patients. Cardiovasc Intervent Radiol 2000;23:187-193.

21. Baarslag HJ, van Beek EJ, Koopman MM, et al. Prospective study of color duplex ultrasonography compared with contrast venography in patients suspected of having deep venous thrombosis of the upper extremities. Ann Intern Med 2002;136:865-872.

As if you could kill time without injuring eternity.
--Henry David Thoreau

Melissa M. King, MD, Mark S. Rasnake, MD, Rechell G. Rodriguez, MD, FACP, Nancy J. Riley, RN, and Jason A. Stamm, MD

From the Department of Internal Medicine, Wilford Hall Medical Center, Lackland Air Force Base, San Antonio, TX.

The views contained herein are those of the authors and do not reflect the views or opinions of the United States Air Force or of the Department of Defense.

Reprint requests to Jason A Stamm, MD, 2200 Bergquist Drive, Suite 1, 759th MDOS/MMIM, Wilford Hall Medical Center, Lackland Air Force Base, San Antonio, TX 78236. Email: [e-mail]

Study results presented at the Society of Air Force Physicians annual meeting in Dayton OH on 8 March 2005, and at the Southern Medical Association annual meeting on 10 November 2005.

Accepted June 27, 2006.

RELATED ARTICLE: Key Points

* The overall incidence of symptomatic PICC-associated DVT was 2%.

* In our review, patients receiving active cancer treatment were at increased risk of PICC-associated venous thrombosis.

* We also found that a commonly used prophylactic anticoagulation regimen does not appear to mitigate this risk.

Table 1. Baseline characteristics of case and control patients

                               Case patients  Control patients
Variable                       (n=27)         (n=54)            P value

Age, years
  Median (range)               56 (21-85)     56 (21-85)        0.993
Male                           15 (56)        30 (56)           1.0
Catheter size 5-French         21 (84)        43 (90)           0.367
Indication for PICC placement
  Antibiotics                  17 (63)        36 (67)           0.497
  Chemotherapy                  7 (26)         8 (15)
  Parenteral nutrition          2 (7)          4 (7)
  Other                         1 (4)          6 (11)
Time (days) from catheter      15             N/A               N/A
  placement to thrombosis
  (median)

Values are absolute numbers with percentages in parentheses, unless
otherwise stated.
N/A, not applicable.

Table 2. Clinical variables and the risk of PICC-associated venous
thrombosis

                         Case      Control
                         patients  patients           Odds ratio
Variable                 (n=27)    (n=54)    P value  (95% CI)

Active cancer treatment  12 (44)   10 (19)   0.013    3.52 (1.27-9.80)
Prophylactic warfarin     7 (26)    3 (6)    0.014    5.95 (1.4-25.3)
  (1 mg/d)
History of thrombosis     3 (11)    1 (2)    0.105    6.63 (0.66-67.0)
Indication for PICC
  placement
  Chemotherapy           17 (63)   36 (67)   0.497    2.01 (0.64-6.31)
  Other                  10 (27)   18 (33)
Diabetes mellitus        13 (48)   27 (50)   0.875    0.93 (0.37-2.34)

OR, odds ratio; CI, confidence interval.