The Greenfield filter as a potential hazard to the operating surgeon.

Abstract: The Greenfield filter and similar devices are placed in the vena cava of patients who have contraindications to anticoagulation, but who would otherwise be at risk of pulmonary embolism without such anticoagulation. Injury to a health care worker from one of these devices has been reported in the case of a pathologist performing a necropsy. To the author's knowledge, injury or near-injury to a member of the surgical team during operation has not been reported in the literature. The authors report on a near-miss injury from such a device to the surgical first assistant during the performance of a pancreaticoduodenectomy. Simple guidelines to avoid similar injuries are provided.

Key Words: Greenfield filter, health care worker injury, HIV seropositivity, injury prevention, pancreaticoduodenectomy

**********

Injury to a pathologist from a Greenfield filter has been reported. (1) The injury occurred during a necropsy performed on a patient who harbored such a device. To our knowledge, this has been the only report in the literature of such an incident. In the current report, a surgeon's first assistant almost suffered a similar injury in the course of a pancreaticoduodenectomy on a patient who also harbored such a device.

Case Report

A 50-year-old HIV-seropositive man presented with jaundice, anorexia, and weight loss. His liver function studies and alkaline phosphatase level, computed abdominal tomography (Figure), endoscopic retrograde cholangiopaucreatography (ERCP), and fine-needle aspiration of the mass in the head of the pancreas supported our clinical impression of malignant biliary obstruction. A past cerebrovascular accident and concomitant deep venous thrombosis required Greenfield venacaval filter placement. The patient underwent pancreaticoduodenectomy. During mobilization of the duodenum, the prongs of the Greenfield filter, which had apparently previously penetrated the caval wall, tore both glove layers of the surgical assistant's retracting double-gloved hand, but not his skin; bloody fluid bathed his hand. The operative specimen yielded a 3.5-cm moderately to poorly differentiated pancreatic adenocarcinoma invading the duodenum with angiolymphatic and perineural involvement. Lymph nodes and margins were tumor-free. The patient convalesced uneventfully, and patient and surgeon were well six months later.

Discussion

According to Gordon Brooks, a senior product manager for Boston Scientific, approximately 600,000 Greenfield filters have reached the market, and about the same number have been implanted in patients, since the device's introduction into clinical practice in 1972. These filters are typically placed to prevent massive pulmonary embolism in patients who have a contraindication to anticoagulation. The use of these devices has in all likelihood saved countless lives, and has prevented profound morbidity from pulmonary embolism.

The typical filter looks somewhat like the frame of a half-opened umbrella. The spines of the filter have sharp barbs that help anchor the device to the wall of the vena cava. These barbs sometimes penetrate the wall of the vena cava; thus, they present a potential danger to the unwary operator who must dissect in the retroperitoneal space around the vena cava. Although documented injury to the surgeon from this device is rare, injury to the patient is more commonly reported and has been well-documented. (2,3) As expected, most of the reported injuries have been to retroperitoneal structures in proximity to the implanted device. In contrast, in the case cited above of the pathologist performing the necropsy, the injury was caused when the pathologist tightly closed his hand around the vena cava at the location containing the filter device. The sharp barbs immediately pierced his skin.

[ILLUSTRATION OMITTED]

Like the necropsy case report, our near-injury occurred during dissection around the vena cava. During this step of the operation, the second portion of the duodenum, which encircles the head of the pancreas, is dissected off the retro-peritoneum and retracted medially. The first assistant then typically holds these structures with the nondominant hand in a sweeping motion to allow the operating surgeon to inspect the area between portal vein, vena cava and pancreas. Hence, one could easily envision the injury. A general injury of this sort has been described in the literature as the "dissector's digital injury." Typically, this injury occurs on the opposition surfaces of the thumb, index, and middle finger of the non-dominant hand. (4) Unlike the case of the pathologist at necropsy, in our case the hand did not encircle the vena cava, so that clenching of the fist was not the mode of injury.

The practice of two simple principles should reduce the potential for injury to the operating surgeon in this type of situation. The most basic principle is to avoid using the hand or fingers as exploring, dissecting, holding, or retracting instruments. In the literature, at least one case (5) of HIV transmission from patient to surgeon attests to the potential consequences of violating this principle. Another principle is to double-glove, and to use universal precautions at all times when there is risk of contamination from the patient. The value of double-gloving is well documented in the literature, (6) and it is well illustrated in our case report.

Conclusions

Our case report presents the potential hazard to the operating surgeon of an implanted device commonly encountered in our patients. Although case reports of injury to patients from this device have been published, the potential risk to the operating surgeon has not been well documented. An injury to a pathologist performing a necropsy has been reported, but this situation is clearly different from that of an operating surgeon performing surgery on a live patient. Finally, we point out very simple principles that have been shown to reduce the risk of injury to the operating surgeon.

Accepted January 8, 2004.

References

1. Abraham JL, Greenfield LJ. Hazard to pathologists and anatomists from vena-caval (Greenfield) filters. Lancet 1995;346:1100.

2. Woodward EB, Farber A, Wagner WH, et al. Delayed retroperitoneal arterial hemorrhage after inferior vena cava (IVC) filter insertion: case report and literature review of caval perforations by IVC filters. Ann Vasc Surg 2000;16:193-196.

3. Jackson SAL, Kennedy RJ, Hakam AG, et al. Delayed transcaval renal penetration of a Greenfield filter presenting as symptomatic hydronephrosis. J Urol 2002;167:1778-1779.

4. O'Brian DS. Patterns of occupational hand injury in pathology. The interaction of blades, needles, and the dissector's digits. Arch Pathol Lab Med 1991;115:610-613.

5. Ippolito G. Scalpel injury and HIV infection in a surgeon. Lancet 1996;347:1042.

6. Gerberding JL, Littell C, Tarkington A, et al. Risk of exposure of surgical personnel to patient's blood during surgery at San Francisco General Hospital. New Engl Journ of Med 1990;322:1788-1793.

RELATED ARTICLE: Key Points

* Although the occasion is rare, the Greenfield filter can cause serious injury to the health care worker or surgeon who engages in anatomical dissection in the retroperitoneum.

* Very simple precautions can be taken to minimize the risk of injury to the health care worker from this device.

Ruben Gomez, MD, PHD, Walter Salwen, MD, and Marc D. Basson, MD, PHD

From the Department of Surgery, John D. Dingell VAMC, and the Wayne State University Medical School, Detroit, MI.

Reprint requests to Marc Basson, MD, Department of Surgery, John D. Dingell VAMC, 4646 John R, Detroit, MI 48201. Email: [e-mail]