TECHNOLOGIC progress has led to the development of simple and reliable epidural patient-controlled analgesia (PCA) pumps, resulting in their widespread use with the consequent appearance of new challenges. We present a case of PCA pump failure inside a hyperbaric chamber.

Case Report 

A 52-yr-old obese woman with renal failure, who had been receiving hemodialysis for 8 yr and had refused a kidney transplant, developed necrotic lesions on the wall of the lower abdomen as a result of vascular calcification secondary to hyperparathyroidism. Anaerobic infection of the affected area necessitated antibiotics, daily debridement, and hyperbaric oxygen therapy (HBO) in a multiplace chamber (TEDSA, Cartagena, Spain). Pain relief was provided by epidural PCA with a CADD-PCA 5200 PXC pump (Pharmacia Deltec Inc., St Paul, MN). However, during the first HBO treatment period, the warning alarm on the PCA pump sounded. Consequently, the anesthesiologist who initiated the epidural PCA treatment also accompanied the patient during the second HBO session, during which the alarm again sounded, and we were able to confirm the malfunction of the pump.

To verify the functioning of this particular PCA pump in a hyperbaric environment, we conducted a series of 20 pressurizations with the same pump in the same HBO conditions, i.e. , within a multiplace chamber reaching 2.5 absolute atmospheres (ATA), a pressure equivalent to a depth of 49.5 ft of sea water. The pump behaved the same each time. When 2 ATA was reached, the alarm rang, the liquid crystal display screen started blinking, and the pump stopped. On reducing the pressure to 1.9 ATA, the pump resumed working without needing to be reprogrammed. At 2 ATA, it was impossible to reprogram the pump.

To determine if this problem was related just to the pump in question, we submitted it plus another two later models of the pump (CADD-PCA 5800) from the same manufacturer to HBO. They were each programmed identically to release 0.5 cm3/min of physiologic saline solution. The results show that although there were no marked differences in flow (the readout on the pumps does not show decimals, only cubic centimeters released) there was a difference in the pressure at which the pumps stopped working and then resumed again (table 1).

Table 1. Experimental Test: Behavior of the Three Pumps at Different Pressures 

Table 1. Experimental Test: Behavior of the Three Pumps at Different Pressures 
Table 1. Experimental Test: Behavior of the Three Pumps at Different Pressures 


Hyperbaric oxygen therapy provides oxygen at pressures > 1 ATA. 1Its therapeutic and iatrogenic effects are a result of the high concentration of oxygen and the increase in barometric pressure. HBO is used with surgery and antibiotic therapy to treat necrotizing lesions. 1,2However, reasons for alterations in the normal functioning of PCA systems within a hyperbaric environment include the following 3:(1) explosion (the pump motor contains brushes that produce sparks);(2) battery leakage caused by the high pressure;(3) collapse of the protective face plate covering the controls caused by high pressure;(4) entry of air into the reservoir or tubes (this is especially dangerous in intravenous PCA).

In this case, and after the experiment, we believe that the PCA pump in question stopped working because of collapse of the protective face plate on reaching 2 ATA, because when the pressure was reduced, the pump resumed working normally without needing to be reprogrammed. This situation was similar in all three pumps, although at different pressures. The manufacturer does not recommend use of these pumps in oxygen-rich hyperbaric chambers because of the risk of explosion (personal communication, December 1991). Although this risk is indeed real in single-bed chambers in which the pressure is increased with pure oxygen, in multiplace chambers, this pressure increase is conducted with air, and the maximum concentration of environmental oxygen is 0.25, 4so that the risk of explosion is reduced (table 2). However, because of the elevated barometric pressure within the chamber, even the ambient atmosphere could have a partial pressure of oxygen of approximately 475 mmHg—similar to 62% fraction of inspired oxygen at 1 ATA. Similar pumps have been used at pressures of 6 ATA with no noticeable problems. 3 

Table 2. Hyperbaric Oxygen Chambers: Characteristics 

Table 2. Hyperbaric Oxygen Chambers: Characteristics 
Table 2. Hyperbaric Oxygen Chambers: Characteristics 

Despite the great amount of literature concerning PCA, we have found very few references to this technique in a hyperbaric environment, two of which report alterations in pump performance. 5,6As a result of our experience and bearing in mind the very few references available, we believe that PCA pumps should be used with caution, if at all, in hyperbaric chambers because of possible changes in their functioning.

The authors thank Ian Johnstone for his help with the English-language version of the manuscript.


Burnett JS, Camporesi EM: Hyperbaric oxygen in trauma, Textbook of Trauma Anesthesia and Critical Care. Edited by Grande CM. St. Louis, MO, Mosby-Year Book, 1993, pp 1085–93
Tibbles PM, Edelsberg JS: Hyperbaric-oxygen therapy. N Engl J Med 1996; 334:1642–8
Aguiluz L, Vrevich N, Hill RK Jr: Ambulatory patient-controlled analgesia devices in the hyperbaric environment. J Hyperb Med 1990; 5:247–57
Manubens E, Estañol N, Robert M: La medicina hiperbárica y la anestesiología. Rev Esp Anestesiol Reanim 1996; 43:321–6
Radcliffe JJ, Spencer I: Performance of the Baxter disposable patient-controlled analgesia infusor under hyperbaric conditions. Anaesthesia 1994; 49:796–7
Akman MN, Loubser PG, Fife CE, Donovan WH: Hyperbaric oxygen therapy: Implications for spinal cord injury patients with intrathecal baclofen infusion pumps. Case report. Paraplegia 1994; 32:281–4