Brown University Logo

BI 108- Spring 2007

Major Recalls of Organ Replacement Devices



Recall Information

Baxter Dialyzer

Bjork-Shiley Heart Valve

Guidant AAA Stent

Guidant ICD

Saint Gobain Desmarquest Hip Implant

Sulzer Hip Implant


Brown University

BI 108 Website



Baxter Dialyzer Recall

Baxter flag [1]


In 2001, Baxter announced a voluntary recall of its AX, AF, and A series dialyzers due to more than 50 patient mortalities worldwide after using these dialyzers. The models of dialyzers recalled, labeled either Baxter or Althane are: Series A11, A15, A18, A22; Series AF150, AF180, AF220; and Series AX1500, AX2200. The A and AF series dialyzers, composed of cellulose diacetate hollow fibers, were a part of Baxter’s acquisition of Althin Medical AB in March 2000. The hemodialyzers were originally manufactured from melt-spun fibers in Miami Lakes, Florida, and then processed in the Ronneby plant in Sweden. The affected dialyzers were distributed from January, 1998 to October 15, 2001 to dialysis centers in more than 50 countries around the globe. [2]

Affected patients displayed various symptoms, which include chest tightness, shortness of breath, progressive dyspenea, circulatory collapse, heart attack, stroke, and shortness of breath. Autopsies of five patients in Spain indicated they died from multiple organ failure. These patients developed an acute onset of illness either during the dialysis session or in the few hours after. Once ill, the patients did not respond to any resuscitation efforts in any of the cases, and the patients died as a result of respiratory distress.

Baxter Dialyzer

Figure 1: A Baxter Dialyzer[3]

Return to Top


Timeline of Events

  • January, 1998 – Baxter starts manufacturing the A, AF, and AX series dialyzers.
  • December, 2000 – Baxter starts using PF5070 in its manufacturing process.
  • Late August, 2001 – 12 deaths of hemodialysis patients using Baxter dialyzers are reported in Spain.
  • Mid October, 2001 – 23 deaths of hemodialysis patients using Baxter dialyzers are reported in Croatia.
  • Mid October, 2001 – Baxter issues a voluntary recall of the A, AF, and AX series dialyzers.
  • Mid October, 2001 – Other deaths of hemodialysis patients using Baxter dialyzers are reported throughout the world. The mortality count surpasses 50 patients.
  • October 15, 2001 – Baxter’s CEO releases a statement proclaiming it will stop manufacturing the A, AF, and AX series dialyzers.
  • November 21, 2002 – FDA issues an Enforcement Report for the Class I recall notice.
  • November 28, 2001 – Baxter announces it has settled with the families of the patients in Spain.
  • December, 2002 – Baxter receives a subpoena from the Department of Justice. [4]

Return to Top



The function of the Baxter dialyzer was simply to filter the patient’s blood during dialysis. The recalled dialyzers were sterile, single use dialyzers for patients suffering from either acute or chronic renal failure.

Dialysis Process

Figure 2: A diagram of the dialysis process for blood cleansing [5]

Return to Top


Device Failure


Figure 3: Commercialized PF5070 [6]

After multiple investigations, it was determined that the dialyzers failed, because of a manufacturing error in their assembly in a Baxter plant in Ronneby, Sweden. In some of the Althane dialyzers was found a residual liquid that proved to be toxic. The liquids were identified as perfluorocarbon (PF5070) and perfluoroheptane (C7F16), both of which are performance liquids.

Before the deaths, PF5070 was not considered to be a dangerous substance to humans. The essential property of PF5070 that led to the deaths of the dialysis patients is that although it is a liquid at room temperature, it becomes a gas at body temperature. PF5070 is insoluble in plasma, and thus is unable to cross the pulmonary alveolar capillary membrane to dissipate from the patient’s body. [7]

Once the perfluorocarbon seeps into the patient’s blood from the dialyzer, the patient has a deadly problem. Bubbles of O2, CO2, N2, PF5070, and H2O vapor form in the capillary bed and in the bloodstream causing the total volume of gas in the capillary bed to become greater than the bed’s capacity to expand. This causes the retrofilling of the pulmonary artery and the right heart with gas, which interferes with oxygen delivery in the body. Consequently, this leads to the blockage of the lung as well as visceral perfusion, which will ultimately lead to death. When autopsies were done on patients in Croatia, it showed that a foam material was obstructing the pulmonary capillaries resulting from blood emulsion, caused by PF5070. The foam material is indicative of massive gas formation in the right ventricle, the vena cava, and the pulmonary-alveolar capillary bed.


Figure 4: A micrograph of an emulsion [8]

Previous experimenting and testing with PF5070 and similar substances reveals a mix of effects, both good and bad. Although there were no tests previous to this recall with PF5070, there were animal tests that were done with a similar substance, a fluorocarbon, FC80. The results of the tests showed that animals exposed to the fluorocarbon had the same fate as the deceased dialysis patients. Subsequent examination revealed that the cause of death for both animal and dialysis patient came from the same toxic phenomenon.

The FC80 findings may seem like a prior indication that PF5070 may be toxic to humans, but previous studies were also conducted with the introduction of intra-tracheal perfluorocarbon into animals to test the possibility of liquid ventilation. Currently, perfluorcarbons are being considered for use in lung injury treatments as oxygen carrier. These experiments with the perfluorocarbon, unlike the experiments with the fluorocarbon, however, showed perfluorocarbons had no direct adverse toxic effects on the lung. These results indicated that there is no reason to worry about the use of perfluorocarbons in the medical world.

Unfortunately, after the deaths of the dialysis patients, experiments were subsequently done with PF5070, and PF5070 was found to be toxic. In these experiments, rabbits were injected with the same dosage of PF5070 that was found in the dialyzers. Like the dialysis patients, the rabbits also died. The rabbits also had the same foam material in their lung capillaries as the dialysis patients did in their autopsies. This experiment confirmed the toxicity of PF5070 as the reason for the failure of the Baxter dialyzers. [9]

PF5070 is one of many perfluorocarbons that is used throughout the dialysis industry in the manufacturing of dialyzers. More specifically, perfluorocarbons play a prominent role in the process known as “dialysis repair” and has been used throughout the dialysis industry for more than 30 years without a problem. Baxter only used the process of “dialysis repair” in its plant in Sweden. In the Ronneby manufacturing plant, PF5070 was used in the repair of dialyzers that failed the initial air pressure test. If more than five fibers were leaking during the air pressure test, the device was taken out of the normal manufacturing process. The leaking fibers of the defective dialyzers were then exposed to PF5070, repaired, and returned into the normal production lot. What went awry was that once the dialyzers were repaired, PF5070 was supposed to evaporate from the dialyzers, and it did not. [10]

Fiber Dialyzer

Figure 5: Parts of a dialyzer, including the fibers [11]

The process of dialysis repair is simply an added manufacturing step to redeem defective dialyzers. First, the header caps of the dialyzer are taken off and one end of the dialyzer is submerged into a PF5070 bath. The blood compartment then becomes introduced to PF5070 to prime the dialyzer. With the bottom still submerged in PF5070, air is pumped into the dialysate compartment under pressure. This allows the inspector to visualize the air bubbles that show up on the upper end of the dialyzer. The air bubbles approximate the area where the leakages in the dialyzer are, which then allows the areas of leakage to be manually sealed up. Once the leaks in the fibers are sealed, the once defective dialyzers are returned to the production line with the other non-defective dialyzers.

The process of dialysis repair obviously had some problems. If the dialyzers are not completely cleaned of PF5070, the dialyzers become lethal to dialysis patients. The presence of residual PF5070 in the dialyzers can be traced back to problems with the process of dialysis repair itself. In the defective dialyzers with broken fibers, the broken fibers are filled up with PF5070. Since one end of the fiber is sealed, however, it would be difficult to access the PF5070 and any kind of rinsing to rid the fiber of PF5070 would not be very effective. The PF5070 would come out very slowly, because the only way for the liquid to come out would be through the small cross-sectional area of the fiber or any other breaks in the fiber. Any type of rinse to speed up this process would have to be very thorough, which many times it is not. An air rinse would have to be continued for a very long time to remove residual PF5070, and a saline rinse would not be effective since PF5070 is insoluble in water. The presence of PF5070 in Baxter’s final product, the dialyzers, indicates a failure of the rinsing or air blowing process employed to clean out the PF5070 in the dialyzers at the manufacturing plant. [12]

Return to Top




Baxter Dialyzer 2

Figure 6: A Baxter dialyzer series [13]

The total death count as a result of the use of Baxter’s affected dialyzers globally is over 50 hemodialysis patients. Spain, which had the first cases reported, had a total of 12 deaths. The highest mortality occurred in Croatia, which had 23 deaths. Four deaths occurred in the U.S. with two in Austin, Texas, and two in Kearney, Nebraska. Deaths also occurred in Colombia, Italy, Germany, and Taiwan.

Overall, only a small percentage of Baxter dialyzers were affected by the recall. The output of the Ronneby, Sweden plant accounts for less than 30% of the total global production of dialyzers for Baxter.The process of  “dialysis repair” is used in less than 10% of Baxter’s A and AF series dialzyers, because it is only used in those that require extra manufacturing steps to repair them after they fail the quality control test. About 150 million dialyzers are produced annually, with Baxter having about a 22% share of the market. Thus, Baxter roughly makes about 33 million dialyzers a year. Assuming the output from Ronneby is about 25% of Baxter’s total output, the Ronneby plant produces about 8.25 million dialyzers a year, which were all recalled, whether they were toxic or not. Subsequently assuming that about 10% of the dialyzers underwent dialysis repair, there are approximately 825,000 potentially affected dialyzers produced a year. Since PF5070 was used for about 10 months, approximately 687,500 dialyzers were potentially affected. [14]

On November 28, 2001, Baxter announced that it would offer a settlement amount of 55 million pesetas, or approximately $289,000 U.S. dollars to each of the patients’ families in Spain. [15] Baxter estimates that it will cost them about $100 to $150 million after taxes in the fourth quarter to discontinue the manufacture of the dialyzers and to compensate families around the globe. It expects that after taxes, the impact will be at most $50 million. The cost of the recall and all associated expenses will be more than the $130 million Baxter paid to acquire Althin in 2000. [16]

For the year 2001, Baxter has paid greatly for its error. In the fourth quarter of 2001, Baxter fell into the red, and its net income for 2001 decreased by 17% from a year earlier. In addition, Baxter CEO, Harry Kraemar, slashed annual bonuses for top executives by 80%, and slashed his own bonus by 60%. He essentially voluntarily took a 33% pay cut as a result of this fiasco. [17]

The risk of a dialysis patient potentially having a dialyzer with PF5070 residue is very slim. There were only about 687,500 potentially affected dialyzers (AF, A, AX series dialyzers that underwent dialysis repair with PF5070). All medical device companies around the world produce about 150 million dialyzers annually. This implies that about 0.4583% of all dialyzers may be toxic. Each dialysis patient undergoes about 150 dialysis treatments per year. Thus, each patient uses about a 150 dialyzers a year. Each time a patient undergoes a dialysis treatment, they have a 0.4583% chance (or less) of becoming sick and dying. Thus, each patient has about less than a 0.68% chance of mortality. [18]

Return to Top



The first deaths were reported in late August 2001 in Spain. Ten ESRD patients regularly treated by dialysis died suddenly a few hours after a dialysis session. An eleventh patient undergoing acute dialysis also died. The patients were treated in three different dialysis centers, but they all used the same brand dialyzers. As a result, the batch of dialyzers was recalled for reassessment. Baxter subsequently put a worldwide hold only on its A series dialyzers. Baxter and the TUV, the European regulatory agency for medical device approval, each conducted separate tests, but neither test revealed a problem in the safety or performance of the dialyzers. Thus, the dialyzers were released even though there was no conclusion on the cause of the deaths of the patients. The total number of mortalities in Spain is reported as twelve patients.

In mid-October 2001, more deaths associated with Baxter dialyzers were reported in Croatia. Twenty-three ESRD patients died in four different dialysis centers over a span of six days. The usual hemodialysis death rate in Croatia is about three to six patients per week. Croatia’s Health Minister Ana Stavljenic-Rukavina stated that the death rate for hemodialysis returned to normal once the Baxter dialyzers were replaced with other brands. At this news, Baxter initiated a voluntary global recall of its AX, AF, and A series dialyzers. It also brings in independent dialysis experts to help it in its investigation. The total body count reported in Croatia is 23.

Also in October 2001, deaths were reported in the United States, Italy, Germany, Taiwan, and Colombia. Twenty hemodialysis patients died in dialysis centers across the globe. In the U.S., There were four mortalities in the U.S., seven in Taiwan, five in Germany, four in Italy, and two in Colombia. Of the four deaths that occurred in the U.S., two were in Austin, Texas, and two were in Kearney, Nebraska. Baxter responded by sending medical investigators into Croatia, Spain, and Texas. Baxter also sent an independent body to inspect its manufacturing plant in Sweden. The company at this point, however, refused to accept any blame, maintaining that there was no evidence linking its equipment with the cause of the deaths. A representative stated that it was too “premature” to state such ideas, even though all the patients who died across the globe all used the same single use dialyzer from the same manufacturer. [19]

While Baxter was denying any defects of its device, the health authorities of the various countries as well as the TUV conducted investigations. They conducted audits and tests, reviewed the practices of the dialysis centers, and inspected Baxter’s manufacturing facilities. When the FDA and Baxter investigated the situation simultaneously, they discovered an unknown liquid substance in the dialyzers. At first, Baxter stated that after its initial investigation, PF5070 is not toxic, but at the same time, it is not labeled for human use. After the investigation with the FDA, however, Baxter discovers that PF5070 is indeed toxic. Experiments were done with rabbits with the same dosage of PF5070 as found in the dialyzers, and the rabbits also died. The rabbits also had the same foam material in their lung capillaries like the patients in Croatia. The results of this experiment indicated that the perfluorocarbon emulsifies in blood at body temperature, increases partial pressure in the capillary beds of the lungs, and forms bubbles in the pulmonary capillary circulation. This leads to blockage of the lung and visceral perfusion. [20]

The release of the results of this experiment forces Baxter to voluntarily continue to take more actions after the global recall. The company announces that it will discontinue its manufacture of the A, AX, and AF series dialyzers in mid-October. It also sent its finding, that PF5070 is toxic, to different regulatory agencies around the world, such as the FDA and the TUV, to the manufacturers in Sweden, and to other companies that make dialyzers as a precaution.

On November 5, 2001, Baxter officially announced that its investigation points to a processing fluid, PF5070, used in the manufacture of its dialyzers in its Sweden plant, may have contributed to the deaths of the patients. Harry Jansen Kraemer, Jr., Baxter’s chairman and CEO, released a statement:

“We are greatly saddened by the patient deaths and I would like to extend my personal sympathies to family members of those patients. We have a responsibility to make public our findings immediately and take swift action, even though confirmatory studies remain under way. While a small number of our A and AF dialyzers appear to have played a role in some of these tragic events, we believe there remain substantive gaps in information about the facts associated with many of the patient deaths. Therefore, we have decided that in the interest of patient safety, the most prudent course of action is to permanently cease manufacturing these dialyzers.” [21]

On November 28, 2001, Baxter announced that it would offer monetary compensation to the families of the hemodialysis patients who died in Spain.The attorney for the families affected stated, “We feel that the settlement each family reached with Baxter is fair and reasonable. Baxter has treated the families with great respect during this unusually difficult time. The families are comforted in knowing that Baxter continues looking into the science surrounding these tragic events to ensure it will never happen again.” Baxter is giving each of the families 55 million pestas, which as about $289,000 U.S. dollars. [22]

Still unsatisfied, Manuel Mata, who is a lawyer representing eight of the families in Spain, stated in November that he was still considering filing lawsuits against Baxter in the U.S. He explained that the families are still upset because even though the dialyzers were taken away from Madrid in August 18th, they were not removed everywhere else. Patients in Valencia continued to use the dialyzers and as a result, more patients died. These deaths were avoidable, according to the lawyer. In the end, however, no lawsuits ended up filed in Spain.

Around the world, reactions to Baxter ’s settlements varied. In the U.S., lawyer Kenneth Moll, who filed the first lawsuit in the U.S. against Baxter, declared that the settlement amount in Spain was “inadequate” and that he wanted to settle for between $2 to $15 million for each case. In Croatia, no settlement amount has been decided upon. The health minister, however, has stepped down and intends to file her own private lawsuit against Baxter. Most recently, in December of 2002, the U.S. Justice Department issued a subpoena to Baxter for documents for its investigation. Baxter is fully cooperating with the government’s investigation. [22]

In the future, to prevent more deaths, Baxter might consider investigating why fatalities continued to increase after its initial recall. Thirty-one patients had died using its dialyzers already, so why were they still in the market? After the initial recall, the total number of patient deaths increased to more than fifty. Specifically, there should be an investigation into why the dialyzers were still being used after the recall.

To prevent similar errors in the future that occurred with Baxter’s hemodialyzers, strict precautionary steps and ongoing quality assessments should be made to ensure the safety of all products. The production of dialyzers should be continuously controlled for quality according to strict standards as well as be observed carefully for any new hazards that might arise, such as the use of a new chemical in the production line.

The presence of PF5070 in Baxter’s final product, the dialyzers, shows a failure of the rinsing or air blowing process to clean out the PF5070 in the dialyzers at the manufacturing plant. There is either a problem with quality testing or manufacturing practices in the medical device world. Perhaps a review of the manufacturing process is necessary to ensure safety. [23]

This recall provides a lesson for medical device manufacturers, such as Baxter, in the future. Baxter did not appropriately clean off its product before releasing it into the public. Their technique was not adequate nor was their quality control testing, because they were unable to detect the presence of any residual fluid.
In addition, the TUV, which is supposed to regulate medical devices, had tested the incriminated dialyzers after the first deaths and had deemed them safe according to their strict standards. Safety standards, thus, should maybe be re-evaluated to determine if additional testing is needed to protect patients. Perhaps there needs to be continuous quality improvement to ensure that any dangerous problems with hemodialysis devices are found and fixed. This incident proves that the different tests employed by different regulatory bodies, such as the TUV and the FDA, is not sensitive enough to detect these hazards. New criteria should be developed to ensure quality in the manufacture and testing of dialyzers. [24]

Return to Top



Other Baxter recalls

  • Recall of Medisystems blood tubing of Meridian dialysis machines
    • Incidents occurred in Mishawaka, Indiana and Grand Rapids, Michigan
  • June 2004 – Alert of contaminated dialysis equipment
    • In British Columbia, a substance resembling blood leaked into an internal component of the hemodialysis machine. This puts patients at risk for contracting a disease, such as HIV or hepatitis.
  • June 21, 2005 – FDA Class I recall of volumetric infusion pumps
    • 225,000 pumps were recalled because they could automatically shut down, depriving patients of necessary medicine.








[7] Shaldon, Stanley. Letter to the Editor. Artificial Organs. 2002; 26(10): 894-896.


[9] Canaud, Bernard, et al. Pathochemical toxicity of perfluorcarbon-5070, a liquid test performance fluid previously used in dialyzer manufacturing, confirmed in animal experiment. American Society of Nephrology. June 2005; 16(6): 1819-23.



[12] Shaldon, Stanley. Letter to the Editor. Artificial Organs. 2002; 26(10): 894-896.








[20] Daly, Emma and Petersen, Melody. Baxter Finds Possible Link in 53 Deaths. New York Times.
6 November 2001. <>




[24] Canaud, Bernard. Performance liquid test as a cause for sudden deaths of dialysis patients:
perfluorohydrocarbon, a previously unrecognized hazard for dialysis patients.
Nephrology Dialysis Transplantation. April 2002; 17(4): 545-548.


Return to Top


Last Updated: April 30, 2007