| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1 Department of Radiology, St. Joseph Mercy — Oakland, 44405 Woodward
Ave., Pontiac, MI 48341.
2 Present address: Department of Radiology, William Beaumont Hospital, 3601 W.
13 Mile Rd., Royal Oak, MI 48073.
3 Present address: Department of Radiology, Wayne State University, DRH 3L-8,
4201 St. Antoine St., Detroit, MI 48201.
4 Present address: Department of Diagnostic Radiology, Yale—New Haven
Hospital, 20 York St., New Haven, CT 06504.
5 Oakland University, School of Health Sciences, Rochester, MI 48309.
Received September 14, 2001;
accepted after revision February 20, 2002.
Presented at the annual meeting of the American Roentgen Ray Society,
Washington, DC, May 2000.
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MATERIALS AND METHODS. The medical records of 148 consecutive patients who underwent filter placement between December 1995 and February 1999 were retrospectively reviewed for cavography, technical success, complications, and operator specialty.
RESULTS. The records of 143 filter placements in 142 patients were complete (one patient underwent two placements). One hundred twenty filter placements were preceded by cavography, and 23 were not. Three misplacements (2.5%) were preceded by cavography and 10 (43%) were not (p<0.0001). One hundred fourteen filter placements were performed by radiologists and 29 by surgeons. Cavography was performed before 98% of placements by radiologists but in only 28% of placements by surgeons (p<0.0001). Filter misplacement occurred in 12 placements (41%) by surgeons and in only one (0.9%) by radiologists (p<0.0001). Major complications occurred in three placements (10%) by surgeons and in none of the placements by radiologists (p<0.01).
CONCLUSION. Vena cava filters were placed with greater technical success and fewer complications when preceded by cavography. Radiologists placed filters with greater technical success and fewer complications than surgeons, which may be the result of radiologists generally having more training and more familiarity with imaging-guided procedures and adhering more to cavography protocol. A credentialing requirement for physician operators and evaluation using quality improvement standards may be advisable.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Technical success rates have recently been reported to be 99% by radiologist operators [4]. However, early reports of misplacement rates in the surgery literature by Otchy and Elliott [5] and Greenfield et al. [6] were high, 14-18%. Filter misplacement in the renal veins, hepatic veins, and right atrium has been reported to have an incidence of 1-10% [1, 7]. Greenfield [1] and Greenfield et al. [6] attributed subsequent improved technical success to the use of a guidewire. The early surgery literature does not indicate that cavography should be done before filter placement [6, 8, 9]. Proper filter placement was determined by verifying its position relative to the third lumbar vertebra on a radiograph obtained after the procedure. The importance of preplacement imaging of the inferior vena cava was soon recognized in the clinical implementation of vena cava filters [2, 4, 8, 10, 11]. Vena cava evaluation must precede filter insertion to identify the level of the renal veins, to determine vena cava size and the presence of thrombus, and to evaluate for congenital anomalies such as vena cava duplication [2, 8]. Inadequate imaging can lead to disastrous results such as deployment of the filter in the aorta rather than in the inferior vena cava [10].
Inadequate imaging can also lead to errors in filter selection [10]. The dimensions of the inferior vena cava are important considerations when choosing a filter. Only the bird's nest filter (Cook, Bloomington, IN) is approved for use in inferior vena cavae more than 30 mm in diameter because of the potential migration of other devices in large vena cavae. Conversely, a vena cava can be too narrow for a filter. The dome of the Simon nitinol filter (Bard, Covington, GA) may not reform properly in a small-diameter vena cava. The Trapeze filter (Cordis, Warren, NJ) is contraindicated in vena cavae smaller than 18 mm. The vena cava may also be too short for a filter. The bird's nest filter is the longest filter, measuring 6-7 cm in length after insertion. Attempts to place this device in a short infrarenal vena cava may lead to partial deployment in an iliac vein.
Technical success requires proper placement of the filter in the vena cava in such a manner as to protect against pulmonary embolism [2]. The optimal location is in the infrarenal inferior vena cava with the apex of the filter just below the level of the lowest renal vein. At this level, a thrombus trapped in the filter will be exposed to renal vein blood flow, which may promote dissolution by the intrinsic lytic system [12]. A filter placed at or above the renal veins can lead to renal vein thrombosis and deterioration of renal function should the filter become occluded [13, 14]. Because of the few published reports of suprarenal vena cava filters, the risk is reasonable but speculative. In addition, a supra-renal vena cava filter may be more difficult to place and more prone to migration than one placed below the lowest renal vein [14, 15].
To our knowledge, no studies have compared technical success and procedural complications of filter placement with and without the use of cavography and among physicians of different specialties. The purpose of this study was to evaluate factors effecting technical success and procedural complications of filter placement, particularly the use of cavography and operator specialty at a single institution.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Patients
The records of 142 patients were complete for assessment. The patients were
64 men and 78 women ranging in age from 25 to 94 years (mean age, 70.1 years).
The indication for filter placement was pulmonary embolism or deep vein
thrombosis with complications of, contraindication to, or failure of
anticoagulation in 95% of patients. Patient demographics, including age, sex,
and indication for placement, were similar for both radiologists and surgeons
(Table 1).
|
Physician Operators
All physician operators were radiologists or surgeons. The hospital is a
private practice institution that has no operator credentialing requirements
for filter placement. Each referring physician decided individually whether to
refer patients to a radiologist or to a surgeon for filter placement. Other
than cavography, no additional procedures were performed concurrently with
filter placement when performed by surgeons.
Six radiologists placed filters. All radiologists were certified by the American Board of Radiology. Five of the six had subspecialty training in interventional radiology, certificates of added qualifications in vascular and interventional radiology, or both. These six radiologists placed between three and 42 filters each. Seven surgeons placed filters. One surgeon was certified by the American Osteopathic Association and certified in the subspecialty of thoracic surgery. The other six surgeons were certified by the American Board of Surgery. Of these six, two had subspecialty certification in vascular surgery and two in thoracic surgery. These seven surgeons placed between one and nine filters each.
Placement Analysis
Radiographs and cavograms were reviewed to evaluate filter position. When
cavography was available, a correctly positioned filter was defined as being
below the lowest of the renal veins in the inferior vena cava. In case of vena
cava thrombosis, filter placement should be above the level of the thrombus.
For each patient in whom cavography was not available, postprocedure
radiographs were reviewed. In four of these patients, postprocedure CT scans
obtained for other reasons were also available and were used to assess filter
position. The radiographic criterion for correct positioning was defined as
filter location between L1-L2 and L5. Because the renal veins typically join
the inferior vena cava at L2 and the vena cava origin is typically at L5,
these limits were thought to be reasonably sensitive and fairly specific in
identifying misplacements. Very few placements evaluated with radiographs
alone were borderline misplacements. Those that were, were not considered
misplacements.
Complications
Procedure-related complications were recorded. Complications that required
additional procedures or prolonged hospitalization or that resulted in death
were considered major [2].
Follow-up was performed until hospital discharge.
Statistical Analysis
Statistical analysis of the data was performed using the chi-square and
Student's t tests.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
|
|
|
Ninety-eight percent of the filter placements performed by a radiologist were preceded by cavography. Two patients (2%) did not have cavography because of renal failure. Only 28% (8/29) of the placements by surgeons were preceded by cavography. This difference is significant (p < 0.0001). Of the two placements by radiologists in patients in whom cavography was not performed, the level of the renal veins was determined, in both patients, by placing a catheter in the renal veins. Of the 21 placements by surgeons in patients in whom cavography was not performed, no other procedure or imaging, such as a fluoroscopic cavography, renal vein catheterization, CT, MR imaging, or sonography, was performed before filter placement. The performance of cavography was at the discretion of the surgeon. Four surgeons did not perform cavography, two performed cavography in some but not all patients, and one performed it for each placement. Surgeons generally used C-arm fluoroscopy for filter placement in the operating room. Greenfield and Simon nitinol filters were the only filter types used. Radiologists used the Greenfield filter in 89% of procedures (102/114) and the Simon nitinol filter in 11%. Surgeons used the Simon nitinol filter in 59% of procedures and the Greenfield filter in 41%. Radiologists preferred the Greenfield filter, and surgeons preferred the Simon nitinol filter (p < 0.001).
Filter misplacement occurred in 43% (10/23) of placements not preceded by cavography but in only 2.5% (3/120) of placements preceded by cavography. This difference is significant (p < 0.0001). Two Greenfield filters were misplaced and 11 Simon nitinol filters were misplaced. Simon nitinol filters were much more likely to be misplaced (p < 0.001). Filter misplacement occurred in 41% (12/29) of the surgeons' placements but in only 0.9% (1/114) of the radiologists' placements. This difference is also significant (p < 0.0001). Of the surgeons' misplacements, cavography was performed in only two of the 12 instances. None of the misplacements was recognized at the time of the procedure or in the operative notes. In the only misplacement by a radiologist, the filter was deployed in the suprarenal vena cava intentionally because of a suspected thrombus in the infrarenal vena cava. However, on retrospective review, no definitive vena cava thrombus was shown on cavography. Of the surgeons' misplacements, nine were in the suprarenal vena cava and three were in iliac veins. Two misplacements were preceded by cavography and 10 were not. Surgeons had a lower misplacement rate with the use of cavography (25%) than with no use of cavography (48%). However, this difference was not statistically significant. As the individual surgeon placed more filters, a statistically significant increase in malpositions occurred (p = 0.014). Surgeons with more placements did not have a statistically lower rate of malpositions.
Major complications occurred in 2% (3/143) of placements. The complications were all in association with filter misplacements into the iliac vein. Two of the three complications were preceded by cavography. The complications occurred in 10% (3/29) of the filter placements performed by surgeons and in none (0/114) of the filter placements performed by radiologists. This difference is significant (p < 0.01).
One patient developed a pulmonary embolism from an existing deep venous thrombosis in the contralateral extremity (Fig. 2A,2B,2C). The filter in the opposite iliac vein did not provide adequate protection and because the misplacement went unrecognized, an additional filter was not placed. No cavography had been performed. The second complication was a large retroperitoneal hematoma requiring blood transfusion. The hematoma likely resulted from perforation of the iliac vein at the time of placement. Although this misplacement was unrecognized at the time of the procedure, abdominal radiography and abdominal CT were subsequently performed to evaluate pelvic pain and hemodynamic instability, and these studies showed the hematoma and the filter misplacement in the right iliac vein. A cavogram had been performed at the time of the procedure. A second filter was subsequently placed in the vena cava (Fig. 3A,3B,3C). A third iliac vein filter misplacement went unrecognized despite cavography, and an additional filter was not placed. Because appropriate treatment would have required placing an additional filter, this was also a major complication [2]. No consequences of this misplacement were shown.
|
|
|
|
|
|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Vena cava filters are safe and effective for the prevention of pulmonary embolism. However, to our knowledge safety and effectiveness have not been previously studied as a function of technique and operator specialty. Technical success rates and complications in this study varied markedly depending on the specialty of the operator and on whether cavography was performed before filter deployment. Radiologists placed filters with a much lower misplacement rate than did surgeons (0.9% vs 43%), used cavography more (98% vs 28%), and had fewer procedural complications (0% vs 10%). Several reasons likely explain these statistically significant differences.
Low radiologist misplacements rates are likely, in part, because radiologists consistently used vena cava imaging (contrast venography), which is generally accepted as a prerequisite for filter placement, to ensure proper placement [2]. Also, perhaps more important, radiologists—in particular, interventional radiologists—are trained in the use of fluoroscopy, venography, and vena cava filter placement. Such training is not uniformly provided in surgical training programs. Operator specialty was closely related to misplacement rate. Although cavography did not influence the surgeons' misplacement rate with statistical significance, a trend was seen toward a lower misplacement rate with the use of cavography. With additional placements by surgeons, the benefit of cavography may have become significant.
Although the Simon nitinol filter was more likely to be misplaced than the Green-field filter (30% vs 1.8%), these misplacements were almost all performed without cavography. It is unlikely that the Simon nitinol filter is the cause of misplacement because the literature has documented a low misplacement rate with the Simon nitinol filter, similar to that of other filters [19].
All major complications arose as a direct result of filter misplacements. Therefore, the complication rate will increase if the misplacement rate is high. The fact that none of the surgeons' misplacements was recognized at the time of the procedure is also worrisome. If a filter does not provide adequate protection because of misplacement, the treatment is placement of an additional filter or correcting the placement of the existing filter. Such treatment cannot be rendered if the misplacement goes unrecognized.
The surgeons' misplacement rate of 41% in this study is greater than that reported in any previous series. In fact, some of the "appropriate" placements in patients who did not have cavography could be ineffective protection against pulmonary embolism if a circumaortic left renal vein or a duplicated inferior vena cava was present. Thus, under the best of circumstances, the misplacement rate of surgeons was 41% and could have been higher. The highest previously reported misplacement rates of 14-18% are from an earlier generation of vena cava filters and placement techniques [5, 6]. In comparison, the misplacement rate of the radiologists in this study was only 0.9%. This is similar to the best results reported in the literature (1-3%) [3, 4].
A weakness of our study is that it is a single-institution study. Results cannot necessarily be generalized to other institutions. However, the high misplacement rate points out the importance of meticulous technique. Such technique requires adequate training of the operator in the use of fluoroscopy, venography, cavography, and filter placement.
Other weaknesses of this study are that it is retrospective rather than prospective and that it is not randomized. Also, for those patients in whom cavography was not performed, placement evaluation with radiographs was limited and does not take potential venous anomalies into account.
The long-term effects of filter misplacements were not determined in this study. The occlusion rate of the Greenfield filter is 4% [3]. Were an occlusion to occur in a filter misplaced in the suprarenal vena cava, renal vein thrombosis could follow with subsequent renal insufficiency. Also, because the iliac veins are smaller, filters placed in these veins may be more prone to occlude over time, resulting in delayed venous stasis.
To our knowledge, no previous studies have compared the technical success of vena cava filter placement and the complications among operators and specialties. In this single-institution study, radiologists are significantly more proficient than surgeons in placing vena cava filters, probably because of their greater familiarity with imaging-guided procedures and their adherence to preplacement cavography requirements. Our study also confirms that the addition of cavography may significantly diminish the complication rate in the placement of vena cava filters. Because of the skills and techniques required for optimal vena cava filter placement, a credentialing requirement for physician operators may be advisable in many circumstances. The placement of vena cava filters should be evaluated by quality improvement standards. Quality improvement guidelines for vena cava filter placement were recently published [20]. Credentialing criteria and outcome standards for vena cava filter placement should be developed.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
