Phaedrus

Fables (c. 8 A.D.)

The prospective study of midline catheters by Mermel and colleagues in this issue [1] reaffirms this ancient truism.

Reliable vascular access is one of the most basic and yet essential therapeutic modalities of modern-day health care. More than 175 million intravascular devices of various types are sold in the United States each year [2]. Most are 1.24-inch (3.2-cm) peripheral venous catheters made of Teflon or polyurethane, but the usefulness of these catheters for periods longer than 3 days is limited by a high incidence of infusion phlebitis (in the range of 40%) [3]. Central venous catheters reduce the risk for phlebitis but are 20 to 300 times more expensive [4] and are associated with as much as 20-fold higher rates of catheter-related bloodstream infection [2]. Three-inch to 8-inch (7.6-cm to 20.3-cm) catheters (termed midline catheters), inserted in the antecubital fossa or upper arm and terminating in the proximal cephalic or basilic vein or the distal subclavian vein, have been promulgated as an attractive and cost-efficient option for intermediate periods of venous access. The period of access averages 2 weeks in published reports but can be as long as 4 weeks [4-8]. Inserted in a larger peripheral vein, midline catheters appear to cause less phlebitis than peripheral venous catheters and are less expensive than central venous devices of all types [4].

The midline catheters available are made of silicone, polyurethane, or a novel composite polymer in which polyurethane is cross-linked chemically with polyethylene oxide, resulting in an elastomeric hydrogel (Aquavene, Menlo Care, Menlo Park, California). The elastomeric hydrogel catheter has the attractive feature of softening and expanding in length and diameter after insertion. The elastomeric hydrogel midline catheter (Landmark, Menlo Care) has been promoted by the manufacturer as extremely safe and almost uniquely resistant to catheter-related bloodstream infection [9, 10]. It was approved in 1989 by the Food and Drug Administration (FDA) through Section 510(k) of the Food, Drug, and Cosmetic Act [11] on the basis of a manufacturer's claim that the device is "substantially equivalent to a pre-1976 predecessor device." As such, preapproval clinical trials of the new catheter were not required or done.

Approximately 750 000 elastomeric hydrogel midline catheters have been sold. Most of the reports that have been published about the catheters since their introduction have been anecdotal or have involved small numbers of catheters [4-8]; no studies have prospectively sought to identify infection with every catheter studied by routinely culturing catheters at removal.

Mermel and colleagues [1] comprehensively examined the safety of the hydrogel midline catheter and its potential to cause infection. In a prospective study of these catheters in situ for an average of 9 days, they found a relatively low incidence of catheter-related bloodstream infection (0.8 per 1000 catheter-days). More importantly, however, they encountered two unanticipated, severe systemic reactions that occurred shortly after the insertion of a hydrogel catheter; after study completion, a third case was encountered. During the study, no similar reactions were seen with 58 580 peripheral venous catheters made of Teflon, which were inserted by the same research nurses who inserted the hydrogel midline catheters.

After finding a handful of published reports of similar reactions associated with use of the midline catheter [12, 13], Mermel and colleagues reported their experience to the FDA and the Centers for Disease Control and Prevention (CDC). However, only through the Freedom of Information Act were they able to get information from the FDA on adverse reactions associated with the catheter that had been reported by other hospitals. They ultimately obtained data on 53 cases [14]. Most affected patients had had dyspnea, urticaria, or other skin changes and abdominal or back pain, frequently with hypotension. Some of the women affected during pregnancy had spontaneous abortions. At least three patients had temporally related cardiac arrest.

In a meeting with FDA officials in December 1994, the hydrogel catheter manufacturer presented data on 141 reports of systemic reactions among users of the first 566 381 midline catheters sold, an incidence of 1 per 4000 catheters (Walker J, Hardy D. Personal communication). Beginning in mid-1994, the manufacturer recommended in its package insert that the hydrogel midline catheter be inserted very gently and flushed more slowly. In the meeting with the FDA, the manufacturer reported that the rate of reactions had fallen to 1 in 13 000 catheters.

Investigations by the manufacturer have not identified the cause of these reactions (Walker J, Hardy D. Personal communication). Most occurred within 5 minutes of catheter insertion (43% after the first flush) and abated within 15 minutes, usually after removal of the catheter. However, in many cases, the catheter was not removed, and yet the patients' symptoms abated spontaneously. Several affected patients later received a second hydrogel catheter without having another reaction. No common skin antiseptics, flush solutions, or intravenous medications were used with the catheters in these reports, and latex gloves were not worn consistently during insertion. Some institutions have reported multiple cases; others, which have used the catheter extensively, have reported no cases. There has been no clustering by production lot. History of allergies does not appear to be a risk factor, and scratch testing with the elastomeric hydrogel in patients who have had reactions has not shown evidence of hypersensitivity. It has not been possible in vitro to show histamine release from peripheral blood leukocytes exposed to the catheter material [15]. Efforts to identify the release of endotoxin or a chemical toxin from the catheter have failed. Given the clinical features of these reactions, they probably represent an unpredictable (type B) [16] nonallergic, anaphylactoid reaction triggered by the elastomeric hydrogel.

Although the biological mechanism of these reactions has eluded detection—which has led the manufacturer to question the relation between the reactions reported and the elastomeric hydrogel—similar numbers of similar systemic reactions have not been reported in association with the use of hundreds of millions of venous catheters made of Teflon, polyurethane, or silicone during this period. However, peripherally inserted central venous catheters made of the same elastomeric hydrogel by the same manufacturer (Centermark, Menlo Care) have been associated with a similar rate of reactions (1 in 4000), whereas small peripheral catheters made of the hydrogel by the same manufacturer (Streamline, Menlo Care) have had a far lower rate (1 in 190 000) (Walker J, Hardy D. Personal communication).

The manufacturer's calculated rate of reactions associated with hydrogel midline catheters is based on reports from users, and it is probably an underestimation, because users may have identified their reactions as a vasovagal reaction, as part of the hyperventilation syndrome, as cryptogenic sepsis, or as an allergic reaction to a drug [17]. (During a large, nationwide outbreak of thousands of cases of gram-negative bacteremia traced to the contaminated intravenous products of one U.S. manufacturer, many hospitals and most clinicians never associated a marked increase in cryptogenic bacteremias with exposure to the manufacturer's intravenous products e.) More-over, even if a reaction is suspected to be causally related to the insertion of a catheter, it may not be reported to the manufacturer, the FDA, or the CDC. The data of Mermel and colleagues [1] suggest that the true incidence of these reactions is higher than the manufacturer's estimates, possibly as high as 1 per 125 catheters.

The study by Mermel and colleagues has important lessons for health care workers and hospitals, and it also raises questions for the federal agency charged with regulating drugs and medical devices.

1. No medical device should ever be considered totally risk-free with regard to infection, especially if the device is implantable. The risk for catheter-related bloodstream infection associated with the midline catheter appears to be low [1, 4-8], but it is not zero, and it is higher than that found with small peripheral venous catheters [2, 3]. Mermel and colleagues' data indicate that the hydrogel midline catheter provides reliable venous access for as long as 2 weeks, but approximately 20% of catheters have to be removed prematurely because of thrombosis or phlebitis.

2. Many new medical devices can reasonably be considered "substantially equivalent" and can be given FDA 510(k) clearance on the basis of a design and choice of a material that has been used safely in millions of patients. However, preapproval clinical trials to establish efficacy and safety should be mandatory for implantable devices, such as intravenous catheters, when the material is new [19].

3. Midline catheters made of the elastomeric hydrogel appear to cause serious anaphylactoid-like reactions that were associated with cardiac arrest in approximately 5% of reported cases. The incidence of these reactions appears to be low but is probably higher than the estimates by the manufacturer. It seems desirable that the elastomeric hydrogel midline catheter be studied in a large prospective trial, ideally a randomized trial, and compared with a catheter made of another material. This would help to determine the true incidence of systemic reactions, to discover whether these reactions are material-specific, and to delineate the cause of these reactions.

4. Although the FDA has one of the most rigorous preapproval processes in the world, even large clinical trials cannot be expected to uncover every safety problem, especially those that occur with a frequency of 1 in 1000 or less [20]. Manufacturers are required to report all adverse events of which they are aware to the FDA [21]. Surveillance of adverse events after devices and drugs are marketed provides the most important means to detect rare adverse reactions to drugs or devices, especially unpredictable type B reactions [22]. For this surveillance to be maximally effective, however, health care workers must report all significant adverse reactions (reactions that require hospitalization or medical or surgical intervention or that result in death) associated with the use of a drug or device. It is not necessary to prove causality—a suspected association is grounds for a report [21]. Reporting is probably most easily facilitated institutionally: Hospitals are mandated by the Joint Commission on Accreditation of Healthcare Organizations to maintain surveillance of adverse reactions to drugs or devices and to promptly report significant reactions to the FDA [23].

5. The FDA's responsibility for protecting the public [11] involves 1) ensuring that new drugs and devices are safe and effective before granting approval for sale and 2) alerting the public and health care workers to possible hazards, in the case of medical devices, that have been reported. No FDA alert about the elastomeric hydrogel catheters has been issued. The FDA may have concluded that the reactions reported were probably not caused by the hydrogel midline catheter; if they believed that the reactions could have been caused by the catheters, they may have thought that the reactions were too rare to justify an alert. The frequency of the reported reactions appears to have been low, but the true frequency is almost certainly higher, and the reactions have clearly been serious. A very recent article by an employee of the FDA's Center for Devices and Radiologic Health [24] provides information on reactions reported to the FDA that were associated with the hydrogel midline catheters (information similar to that uncovered by Mermel and colleagues [1]), but many users of midline catheters have not seen this article. It is unclear why the FDA did not issue an alert to health care workers and hospitals as early as 1992, encouraging them to report serious reactions associated with all types of vascular catheters. It is not too late.