Mermel and associates [1] raised important safety issues in their study of the risk of midline catheterization. We have used approximately 500 Landmark midline catheters in our unit during the past 4 years and have seen seven adverse reactions in five adults with cystic fibrosis. In all cases, patients had rapid onset of dyspnea within minutes of insertion and either before or after the line was flushed. We noted four cases of erythematous skin reaction and one case of angioedema. Other reactions seen on one or more occasions were nausea, vomiting, wheeze, cough, tachycardia, and rigor. Cyanosis occurred in two patients who had severe pulmonary disease. Symptoms rapidly resolved immediately after the line was removed. On one occasion, however, symptoms resolved when the line was left in place; the line was then used for 2 weeks. Intravenous hydrocortisone was given to two patients, and adrenaline was given to one.

On two occasions, the reaction occurred on the patient's second exposure to the catheter. In two cases, polyurethane midline catheters were subsequently used without problems; in another two cases, a rechallenge to a Landmark catheter resulted in an identical adverse reaction.

We are aware that the manufacturer's investigation of similar adverse reactions has not identified a cause. Our cases, like those described by others, appear to be acute hypersensitivity reactions. We recently did skin-prick tests using aqueous extracts of Landmark catheter, latex, and a polyurethane catheter. We found no evidence of sensitivity in the three patients challenged.

Pending clarification of the risks, any institution using this device should watch for adverse reactions. We are now using a polyurethane midline catheter in most patients, although we continue to use the Landmark catheter in patients who prefer it and in those in whom the catheters have been well tolerated. We endorse the sentiments of the editorial accompanying the report [2] and agree that the reporting of adverse events associated with medical devices must not be overlooked or considered less important than the surveillance of drugs.