Confirmatory chest radiographs after central line placement: are they warranted?

Objectives: This study was designed to determine the ability of physicians to predict complications associated with the placement of central venous access devices and to decide whether a confirmatory chest radiograph is warranted after placement.

Methods: Patients receiving central venous access on an inpatient and outpatient gynecologic oncology service were studied. Data were collected regarding patient demographics, patient history, procedural details of the placement, and the type of catheter used. The physician then predicted which patients had a reasonable potential for placement complications. All of the patients then underwent radiography, which was then compared with the original prediction.

Results: Ninety-eight patients who had central venous access devices placed were included in the study. Eighty of the 81 central lines thought by the practitioner to have been placed without incident caused no significant complications; one individual in this group had a minor pneumothorax. Two of 17 patients predicted to have complications were noted to have a pneumothorax that required hospitalization. No patients in the low-risk group were hospitalized for a placement complication, whereas two hospitalizations occurred in the high-risk group.

Conclusion: Confirmatory chest radiographs may potentially be omitted in certain cases after line placement when experienced clinicians use good technique, good clinical judgment, and discrimination.

**********

The technique of percutaneous central venous catheterization was described first by Aubaniac (1) in 1952. Aubaniac described his experience with the use of subclavian fluids in military casualties. The use of central venous access expanded when an article about the use of pressure monitoring was published in 1962. (2) Further use of central venous accesses became more common with the advent of total parenteral nutrition in 1968. According to the Food and Drug Administration, approximately 5 to 6 million catheters are placed in patients in the United States on an annual basis. (3)

The placement of these catheters has been associated with various complications including pneumothorax, hemothorax, hydrothorax, chylothorax, arrhythmias, cardiac tamponade, hematoma formation, air embolism, infection, injury to the great vessels, thrombosis, injury to the brachial plexus, injury to the phrenic nerve, and death. (4-9) The rate of complications from placement of these lines ranges from 0.4 to 20%. (10-12) Significant malpositioning of the catheter occurs in approximately 2% of the patients when either a subclavian or internal jugular vein approach is used. (13) Several studies have correlated these stated complications as being inversely proportional in their occurrence rate with the level of experience of the person placing the line. (14-18)

In these study institutions, as in many in the United States today, the standard of care has been to obtain a postprocedural chest radiograph to evaluate for any placement complications. Chest radiographs after line placement may cause delay in the usage of the line and have an added cost both for the radiograph and for the interpretation fee. This study was undertaken to determine whether experienced physicians could reasonably predict patients who might have complications from central line placement.

Patients and Methods

Between October 1997 and July 1999, 98 patients from a gynecologic oncology practice in South Carolina received central access. These were 98 consecutive patients requiring central lines for different medical therapies. Information collected on the patients included height, weight, cancer stage and type of disease, presence of supraclavicular adenopathy, presence of pulmonary metastases, and presence of a pulmonary effusion. A history of prior line placement, previous pneumothorax, or any previous chest surgery was also recorded.

Three separate access devices were used: a Hohn Percutaneous Catheter (Bard Access Systems, Salt Lake City, UT), the Arrow Triple Lumen Catheter (Arrow International, Reading, PA), and the Life Port (Horizon Medical Products, Manchester, GA). The Hohn catheters and triple-lumen catheters were placed in the patient at the bedside with the patient awake and with lidocaine for anesthesia. The Life Ports were placed in patients in the operating suites under conscious sedation with local anesthesia. The location of the catheter entry site was always under the right or left clavicle for a subclavian access. Two board-certified gynecologic oncologists who had extensive experience with line placements placed all lines. The number of attempts at placement, sites of entry, and the type of catheter used were recorded in all patients (if a patient had more than one placement of a catheter during the study period, he or she could be recorded twice for the two separate new placements). The operator was then asked to decide and document whether or not a postprocedural radiograph of the chest was needed on the basis of the perceived difficulty in the placement of the central venous access. A portable upright chest radiograph was obtained after the placement of all central lines and subsequently reviewed by a radiologist to confirm the presence or absence of any complications and to check for malpositioning of the line. The radiologists reviewing the films were board-certified and were blinded as to concern for complications by the gynecologic oncologists. Malpositioning of the central line was considered to be any placement in the right ventricle, the jugular veins, or a doubling back of the catheter tip into the arm. Categorical variables were compared with Fisher's exact test and continuous variables were compared with Student's [tau] test. All reported P values are two-sided. Statistical significance was defined at the 0.05 level.

Results

During the study period from October 1997 to July 1999, attempts were made to place 98 central venous access devices. Hohn catheters and triple lumens were successfully placed in 56 patients (57%) and Life Ports were successfully placed in 41 patients (42%). Placement of the central venous catheter was on the left in 55 patients (56%) and on the right in 42 patients (43%). The mean number of attempts to place catheters was identical for each side (mean, 2.2 attempts). There were three patients (3%) who had attempts performed on both subclavian veins at the same sitting, with successful placement accomplished, and one patient (1%) in whom the attempt to place a line failed. This patient had a pneumothorax.

Postprocedural radiographs were obtained for all 98 patients to diagnose potential complications associated with central venous access. The staff radiologists at the institution reviewed all radiographs where the line was placed. Of the 98 patients, 81 were predicted to have no complications by the physician placing the line. The other 17 individuals were considered to be patients in whom venous access placement was difficult and therefore were thought to need a postprocedural radiograph. Factors that might have complicated placement, the frequency with which they occurred in patients, and other demographic information about the study population are listed in Table 1. The high-risk group was more likely to have had a prior chest surgery (18 versus 1%, P = 0.016) and to have had more attempts on both the left (3.2 versus 1.9, P = 0.103) and the right sides (3.8 versus 1.8, P = 0.021). There were no statistically significant differences between the groups for any of the other factors.

Three pneumothoraces were identified in the study population. One of these complications was identified in the group of patients deemed not to have required a radiograph. This patient had a small pneumothorax of <10% and did not require chest tube placement for treatment. The other two patients who had complications were in the group thought to be at risk for complications; both required hospitalization and the placement of a chest tube. There were two malpositioned catheters in the right ventricle, one in each study group. These were subsequently repositioned. None of the lines were noted in the neck vasculature or in the arm. No arrhythmias were noted in conjunction with these malpositionings.

No other potential complications that have been associated with central venous line placement were found. Therefore, two patients with complications for the entire study required hospitalization, and both of these patients were thought to have been at an increased risk for potential complication because of the difficulty of attempted placement. The proportion of hospitalizations in the high-risk group (2 [12%] of 17) was significantly greater (P = 0.03) than the proportion of hospitalizations in the low-risk group (0 of 81) (Table 2).

We reviewed the demographic information--weight, height, type of catheter, status or location of disease, side of line placement, and number of attempts--for patients who had a pneumothorax and compared the information with that of patients who did not have a pneumothorax. We found no significant risk factors that could be assigned to patients who had pneumothoraxes that might explain why this subgroup had such a complication.

Discussion

The physicians in this study correctly predicted complications that required hospitalization in both instances. Total complications for the study group as a whole included a 3% rate for pneumothorax and approximately 2% for malpositioning of the line. The complications that required hospitalization occurred in 2% of the study population, and again both patients had been predicted to be at an increased risk for complications at the time of placement of the central line.

Several articles in the literature have concluded that a routine chest radiograph after the replacement of a central line over a guidewire is not necessary when good clinical judgment and discrimination are used. (8) (14) Farrell et al (19) found that routine chest radiographs rarely contributed to the diagnosis of any procedural complications and were of little value after internal jugular access in hemodialysis patients. Gray et al (20) evaluated the necessity for postprocedural chest radiographs after catheterization of the central veins or the insertion of pulmonary artery catheters. Their ability to predict the absence of complications after insertion of central venous catheters through the subclavian or internal jugular veins was excellent at 151 of 152. Gray et al (20) also showed, however, that in line placement by house staff, the unexpected complication rate increased to 12%.

Several series in recent years have shown that ultrasound guidance may limit complications from central line placement. (21) (22) Slama et al (21) revealed in 37 patients cannulated with ultrasound guidance that no complications occurred. Likewise, Fry et al (22) showed that in 52 consecutive patients with ultrasound-guided line placement, only one patient developed a pneumothorax.

In this study, the complication rate for pneumothorax was 3%. In those thought not to be at risk for complications, the pneumothorax rate was 1.2% (1 of 81). In this one case, however, the complication was minor and did not require hospitalization or the placement of a chest tube. There were two cases of minor malpositioning of catheters in which the lines were placed in the right ventricle, and these were subsequently repositioned; none of these complications were of significance because none of them required hospitalization or resulted in a serious adverse event. In the group thought to be at higher risk for a complication, the complication rate was 12% (n = 2). The two patients here that suffered from a pneumothorax required both hospitalization and a chest tube. When comparing the two arms of the study (those who were thought to be at low risk versus those at high risk for complications), the difference in complications requiring hospitalization was statistically significant (0 versus 12%, P = 0.03).

The authors raise the question of whether a chest radiograph should be routinely obtained after line placement in all patients or whether some radiographs are unnecessary. Some clinicians might argue that "unnecessary" radiographs, even in large volume, might be justifiable if the even rare occurrence of a life-threatening complication could be detected. It might equally be argued that most life-threatening complications, such as a tension pneumothorax, might be recognizable before the chest radiograph has been read. Delays from the time of the postprocedural film being taken, processed, and read may result in valuable time being lost and certainly can add to the charges of the patient's care. The advent of managed care has compelled all medical participants to be more aware of cost containment within health care facilities.

This study had some limitations in that the lines were placed in a controlled environment. There were no rapid placements of central lines in trauma or emergent situations in the operating suite. Also, the study was limited in that both physicians placing lines were very experienced, and the results may not apply to placement of lines by inexperienced practitioners.

Conclusions

We conclude that routine chest radiographs after central venous access may not be necessary when experienced clinicians use good technique, clinical judgment, and reasonable discrimination in low-risk, nonemergent patients when placing central venous access devices. We think the current climate of cost containment mandates a continuing review of common procedures and practices to determine the best possible patient outcome for each health care dollar spent.

Key Points

* The placement of central venous catheters has been associated with various complications, including pneumothorax, hemothorax, hydrothorax, chylothorax, arrhythmias, cardiac tamponade, hematoma formation, air embolism, infection, injury to the great vessels, thrombosis, injury to the brachial plexus, injury to the phrenic nerve, and death.

* Routine chest radiographs after central venous access may not be necessary when experienced clinicians use good technique, good clinical judgment, and reasonable discrimination when placing central venous access devices in low-risk, nonemergent patients.

* The current climate of cost containment mandates a continuing review of common procedures and practices to determine the best possible patient outcome for each health care dollar spent.

Table 1. Demographic and clinical characteristics of women receiving
central lines for cancer therapy, Greenville Hospital System, October
1997-July 1999, by whether physician believed central line was high or
low risk (a)

                            High-risk      Low-risk
                            placement      placement
Patient characteristics      (n = 17)       (n = 81)

Mean height, cm (SD)        160.9 (6.6)     163.3 (7.4)
Mean weight, kg (SD)         71.8 (23.0)     75.0 (21.3)

Catheter type
  Hohn/Triple-lumen (%)         9 (53)         48 (59)
  Portacath/Life Port (%)       8 (47)         33 (41)

Stage of disease
  I/II (%)                      2 (12)         18 (22)
  III/IV (%)                   15 (88)         63 (78)
Prior central lines (%)         4 (24)         30 (37)
Prior pneumothorax (%)          1 (6)           0 (-)
Prior chest surgery (%)         3 (18)          1 (1) (b)
Pleural effusion (%)            0 (-)           3 (4)
Adenopathy (%)                  0 (-)           4 (5)
Pulmonary metastases (%)        2 (12)          5 (6)

Catheter placement
  Right subclavian (%)          7 (41)         35 (43)
  Left subclavian (%)           9 (53)         46 (57)
  Failed (%)                    1 (6)

Mean No of attempts (SD)
  Left side (n = 10/36)       3.2 (2.7)       1.9 (1.9) (c)
  Right side (n = 12/52)      3.8 (2.5)       1.8 (1.4) (d)

(a) All differences P > 0.10 except as noted.
(b) P = 0.016.
(c) P = 0.103.
(d) p = 0.021.

Table 3. Occurrence of pneumothorax by whether the physician placing the
catheter thought a chest radiograph might be necessary (a)

                                      No pneumothorax
                                        (n = 95) or
                                        pneumothorax
                    Pneumothorax        that did not
Physician thought   requiring              require
chest radiograph    hospitalization    hospitalization
required              (n = 2)            (n = 1) (b)

Yes (high risk)           2                  15
No (low risk)             0                  81

(a) P = 0.029 (Fisher's exact test, two-sided).
(b) The pneumthorax patient who did not require hospitalization was in
the low-risk group.

From the Departments of Gynecologic Oncology, Obstetrics and Gynecology, Vascular Surgery, and Research, Greenville Hospital System, Greenville, SC.

Reprint requests to Larry E. Puls, MD, Department of Gynecologic Oncology, Greenville Hospital System, 900 W. Faris Road, CTC 3rd Floor, Greenville, SC 29605. Email: [e-mail]

Accepted August 21, 2002.

Copyright [c] 2003 by The Southern Medical Association

0038-4348/03/9611-1138

References

1. Aubaniac RL. Intravenous subclavicular injection [in French]. Presse Med 1952;60:1456-1458.

2. Wilson IN, Grow JB, Demong CV, et al. Central venous pressure in optimal blood volume maintenance. Arch Surg 1962;85:563-578.

3. Dudrick SJ, Wilmore DW, Vars HM, et al. Long-term total parenteral nutrition with growth, development, and positive nitrogen balance. Surgery 1968;64:134-142.

4. Collier PE, Blocker SH, Graff DM, et al. Cardiac tamponade from central venous catheters. Am J Surg 1998;176:212-214.

5. Moskal TL, Ray CE Jr. Left mediastinal central line malposition: A case report. Angiology 1999;50:349-353.

6. Murray BH, Cohle SD, Davison P. Pericardial tamponade and death from Hickman catheter perforation. Am Surg 1996;62:994-997.

7. Lo WK, Chong JL. Neck haematoma and airway obstruction in a preeclamptic patient: A complication of internal jugular vein cannulation. Anaesth Intensive Care 1997;25:423-425.

8. Lowell JA, Bothe A Jr. Venous access: Preoperative, operative, and postoperative dilemmas. Surg Clin North Am 1991;71:1231-1246.

9. Ryan JA Jr, Abel RM, Abbott WM, et al. Catheter complications in total parenteral nutrition: A prospective study of 200 consecutive patients. N Engl J Med 1974;290:757-761.

10. Amshel CE, Palesty JA, Dudrick SJ. Are chest X-rays mandatory following central venous recatheterization over a wire? Am Surg 1998;64:499-502.

11. Scott WL. Central venous catheters: An overview of Food and Drug Administration activities. Surg Oncol Clin N Am 1995;4:377-393.

12. Lefrant JY, Cuvillon P, Benezet JF, et al. Pulsed Doppler ultrasonography guidance for catheterization of the subclavian vein: A randomized study. Anesthesiology 1998;88:1195-1201.

13. Kaufman JL, Rodriguez JL, McFadden JA, et al. Clinical experience with the multiple lumen central venous catheter. JPEN J Parenter Enteral Nutr 1986;10:487-489.

14. Palesty JA, Amshel CE, Dudrick SJ. Routine chest radiographs following central venous recatheterization over a wire are not justified. Am J Surg 1998;176:618-621.

15. Herbst CA Jr. Indications, management, and complications of percutaneous subclavian catheters: An audit. Arch Surg 1978;113:1421-1425.

16. Bernard RW, Stahl WM. Subclavian vein catheterizations: A prospective study--Part 1: Non-infectious complications. Ann Surg 1971;173:184-190.

17. Riblet JL, Shillinglaw W, Goldberg AJ, et al. Utility of the routine chest X-ray after "over-wire" venous catheter changes. Am Surg 1996;62:1064-1065.

18. Cullinane DC, Parkus DE, Reddy VS, et al. The futility of chest roentgenograms following routine central venous line changes. Am J Surg 1998;176:283-285.

19. Farrell J, Walshe J, Gellens M, et al. Complications associated with insertion of jugular venous catheters for hemodialysis: The value of postprocedural radiograph. Am J Kidney Dis 1997;30:690-692.

20. Gray P, Sullivan G, Ostryzniuk P, et al. Value of postprocedural chest radiographs in the adult intensive care unit. Crit Care Med 1992;20:1513-1518.

21. Slama M, Novara A, Safavian A, et al. Improvement of internal jugular vein cannulation using an ultrasound-guided technique. Intensive Care Med 1997;23:916-919.

22. Fry WR, Clagett GC, O'Rourke PT. Ultrasound-guided central venous access. Arch Surg 1999;134:738-741.

Larry E. Puls, MD, Carrie Ann Twedt, MD, James E. Hunter, MD, Eugene M. Langan, MD, and Martin Crane, PHD