TRAGIC EVENTS SURROUNDING CHRONIC HEMODIALYSIS
PATIENTS IN CROATIA
Vladimir Gašparović1, Rajko Ostojić1
1
Department of Medicine Rebro, Zagreb University Hospital Centre, Zagreb, Croatia
key words: chronic hemodialysis, sudden deaths, perfluorocarbon
SUMMARY
Until recently mortality rates of chronic hemodialysis patients in Croatia was 10,26% and did
not greatly differ from those in other countries. The tragic events that occurred in October
2001 resulted in 23 unexpected deaths of patients on chronic hemodialysis in a single week.
Five patients died during hemodialysis and an additional 18 patients died within several hours
after hemodialysis had been completed. The clinical presentation was characterized by
dyspnea and hypotension, followed by cardiac arrest. Standard resuscitation measures were
unsuccessful. These deaths pointed to an acute incident which had to be promptly identified
in order to prevent further loss of life. To illustrate how dramatic these incidents were, let us
just say that five patients died in a small regional center in a single day.
Once the initial epidemiologic data was obtained, simulation of hemodialysis on a model was
performed. In all 23 patients who died during or shortly after hemodialysis P15 or P18
dialyzers were used. The mentioned dialyzers (P15 and P18) were manufactured by Baxter
and distributed by Pliva. Effluent samples were gathered to be study by gas chromatography.
Dialysate, AV lines, water and staff were different and no correlation was found between
these factors and the deaths. Fourteen patients underwent an autopsy. The autopsy findings
consistant with end stage renal disease were found as were unusual foaming of blood and
dilatation of heart chambers. According to the data provided by the manufacturer,
perfluorocarbon which is used in the fiber stabilization process in the mentioned dialyzers
may be responsible for the noted clinical picture since similar observations were made in an
experimental model.
INTRODUCTION
The most commonly reported causes of in patients of chronic hemodialysis are cardiovascular
incidents independent of extracorporeal circulation. Reactions to hemodialysis are most
commonly allergic in origin, but according to available data they occur mainly at its
beginning and do not result in a lethal outcome (1). Sudden death on hemodialysis is not
frequent, and the literature reports mainly cardiovascular events as causes of death in chronic
renal failure patients on hemodialysis (2,3). A number of reactions to hemodialysis have
been noted, but as a rule without a lethal outcome. These are usually allergic reactions to a
foreign substance (4,5). End-stage renal failure in Croatia is mainly managed with chronic
intermittent dialysis in centers for dialysis, most often found in hospitals. Home hemodialysis
is still underdeveloped (only six patients). According to the year 2000 annual report 2270
patients with chronic renal failure were dialyzed in Croatia. Of these, 143 were on peritoneal
dialysis, and all others were on hemodialysis. The reported mortality rate was 10.26% (7).
Mortality rates of patients on chronic hemodialysis in our country do not differ from those in
the world. Information about four sudden deaths that occurred in one day, October 12, 2001,
in the Center for Hemodialysis in the Požega General Hospital, pointed to the gravity and
tragedy of events to follow. Once a connection with dialysis centers throughout Croatia was
established, we were unfortunately faced with 24 fatal events, most of which were
unexpected. Immediately upon arrival of the first epidemiological data we started to
investigate the causes and found one common factor - dialyzer manufactured by Baxter/Pliva.
PATIENTS AND METHODS
A questionnaire made up of general and specific information was distributed among
hemodialysis centers in which deaths had been reported. The clinical presentation, time at the
which the incident was reported, previous medical history, type of hemodialysis equipment,
types of membranes used, as well as solution types, specific AV lines and needles composed
the data that was thoroughly analyzed. Autopsy findings were also recorded. One of the
patients dialyzed on the hemofane membrane committed suicide and was therefore excluded
from the investigated group.
The remaining 23 patients were dialyzed in six different centers throughout Croatia (Zagreb,
Požega, Karlovac, Rijeka, Dubrovnik, Pula) on cellulose diacetate membrane P15 (15
patients) and P18 (eight patients). One patient in this group who died five hours after the
beginning of dialysis had metastatic uterine cancer with ascites, but she was still in a
satisfactory condition and ambulatory. Upon completion of dialysis the patient went home
and died. Despite her malignant disease, her death was not immediately expected. The
previous medical history of another patient was significant for an acute myocardial infarction
eight days prior to death. Dialysis was performed with the P18dialyzer, and as in the former
patient the death was unexpected. Death occurred 37 hours after the dialysis, and it is not
certain that these two cases are associated with the dialyzer (Table 1). In all other patients on
chronic hemodialysis, i.e. 21 patients, death was unexpected and was found to be associated
with the dialyzer because all other materials were different (water, dialysis solutions, AV
lines, needles, devices). All preserved incriminated dialyzers P15 had the same control
number 2001 F 07 P (11 pieces), and dialyzers P18, on which the events occurred, probably
the same control number 2001 B 17 R, manufactured by Baxter, distributed by Pliva.
Gas chromatography determined certain differences between the effluents taken from the
incriminated dialyzers P15 and P18 when compared to the same type of dialyzers but with
different control number.
The quality of water used for dialysis purposes was investigated by conductivity and the
values ranged from 0 to 10.1 microsiemens.
Examination of the concentrate matched the declaration on the solution label. In patients who
underwent an autopsy, in addition to findings commonly seen in chronic end-stage renal
failure, foaming of blood in heart chambers and dilation of heart chambers was noted. At an
autopsy in Požega, opening of the heart under water resulted in copious release of air bubbles
from the heart.
Effluent samples from the incriminated dialyzers P15, control number 2001 F07 P, were
marked by different findings of gas chromatography, i. e. the presence of a substance (peak)
not found in the control cellulose diacetate dialyzers of the same manufacturer but of a
different control number. Collected data on used devices, type of bicarbonate capsules,
hemodialysis solutions, AV lines, needles and methods of device disinfection were different.
According to the data provided by the manufacturer, the perflourocarbon solution used for
fiber stabilization for the mentioned dialyzers is a volatile solution which under certain
circumstances can assume a gaseous form.
The results are presented in table 1.
DISCUSSION
Mortality of patients on hemodialysis in the world is 9-10%, and in Croatia it does not
substantially differ from this number According to the national annual report the mortality
rate in chronic hemodialysis patients in Croatia is 10.26% (6).
The unexpected deaths among hemodialysis patients in Croatia that occurred within a
week led to a medical crisis. These events alarmed all hemodialysis patients, their
families as well as the medical personnel and the media throughout Croatia. They led
to the resignation of the Croatian Minister of Health and will certainly have legal
consequences. Once the problem was identified appropriate measures were
undertaken, yet the death toll of these tragic events remains a horrific reminder(7,8).
Problems with cellulose acetate membrane were consequent to long storage of a dialyzer, and
were characterized by eye redness and tearing, loss of hearing, tinnitus and pain in the bones
(1-5).
On the day of the first noted death the hemodialysis procedure started in the usual
fashion. The events that followed occurred rapidly. Patients complained of chest pain
and dyspnea. Five deaths occurred during hemodialysis. An additional 16 patients died
within 10 hours of hemodialysis completion. Only two patients died more than 10
hours after hemodialysis.
The clinical picture was dominated by suffocation, chest pain, sweating, and in 2 cases with
generalized convulsions. Clinical manifestations of dyspnea and restlessness lasted briefly
from the beginning of the event to cardiac arrest, most often less than 30 minutes. All
incidents resulted in a lethal outcome despite resuscitation attempts in hospital settings. Some
patients went home without complaints, and according to the information provided by family
members the presentation corresponded to in-hospital deaths. Similar incidents happened in
two Spanish towns, Madrid and Valencia, but data concerning these incidents is not quite
clear (9). According to an oral communication by the representative of the Food and Drugs
Administration (FDA), there were two deaths during hemodialysis in the U.S.A., and two
patients were hospitalized with severe respiratory problems, so that in at least one case
artificial ventilation was needed. No satisfactory explanations of these events were offered
(10). The patients in Croatia were dialyzed on monitors from different companies with
different solutions produced by different manufacturers. Different AV lines, needles, as well
as vascular accesses, were used. Information concerning incidents related to certain Baxter
and Althin monitors secondary to blood acidification was not applicable to the Croatian
setting because we use a bicarbonate powder for alkalinization.
Collection of epidemiological data was impeded by the brevity of time in which the events
happened, as well as by insufficient communication between centers for dialysis. All
elements including the sudden occurrence of the incident, very similar clinical manifestations,
unsuccessful resuscitations, autopsy findings, different consumables in the dialysis procedure
pointed to the only common factor in these tragic events: cellulose diacetate membrane
manufactured by Baxter. There have been no reports in the literature in which the mechanism
of such events was disclosed. This includes the reports from Spain.
Sudden deaths on hemodialysis or in the first hours after dialysis are related with cellulose
diacetate dialyzers manufactured by Baxter, distributed by Pliva, named P15 and P18.
The clinical presentation pointed to a possible gas embolization of both the arterial and the
venous circulation. The macroscopic autopsy findings from Pula and Požega Centers pointed
to gas embolization of the great arteries and foaming of blood within both the arterial and
venous pools. These reports were crucial for our investigation to focus on possible causes of
gas embolism. According to the most recent communication by the manufacturer, it is
possible that the liquid for preparation and stabilization of fibers, perfluorocarbon PF 5070, is
responsible for these events. This liquid is used in Sweden for stabilization of fibers for
dialyzers Althin, which Baxter has in the meantime taken over (11). The offered information
pointed that a 1 μl of perflourocarbon may produce up to 1 ml of gas, which is a 1000 fold
volume enlargement. It would seem that flushing the filter with a 1000 ml of normal saline
enabled small quantities of perflourocarbon to remain within the filter. These would turn to
their gaseous form once they entered the blood stream and produce gas embolisms. These
results suggest that a serious manufacturing error has been made and that the manufacturing
process need to be radically altered.
After the withdrawal of P15 and P18 dialyzers no new dialysis deaths have been recorded in
our country. We believe that the identification of the incriminated substance is of paramount
importance for the prevention of such future tragedies.
LITERATURE:
1. Averbukh Z, Modaai D, Sandbak J, Berman S, Cohn m, Galperin E, Cohen N,
Dishi V, Weissgarten J. Red exe syndrome: clinical and experimental experience in
a new aspect of diffuse eosinophilic infiltration? Artif Organs 2001; 25:437-40.
2. Takeda K, Harada A, Okuda S, Fujimi S, Oh Y, Hattori F, Motomura K, Hirakata
H, Fujishima M. Sudden death in chronic dialysis patients. Nephrol Dial
Transplant 1997; 12(5):952-5.
3. Cohle SD, Graham MA. Sudden death in hemodialysis patients. J Forensic Sci
1985; 30(1):158-66.
4. Noh H, Lee YK, Kan SW, Choi KH, Ha DS, Lee HY. Acute eosinophilic pneumonia
associated with amitriptyline in a hemodialysis patient. Yonsei Med J 2001;
42:357-9.
5. Moore MA, Kaplan DS, Picciolo GL, Wallis RR, Kowolik MJ. Effect of cellulose
acetate materials on the oxidative burst of human neutrophilis. J Biomed Mater
Res 2001; 55:257-65.
6. Kaštelan A. Annual report of Croatian center for dialysis and tissue
transplantation in Croatia in 2000, Zagreb, Urology Clinic, 2001.
7. Marušić Ana, Croatian Health Minister resignts after dialysis deaths. Lancet
2001;338:1431,
8. V. Gašparović, R. Ostojić, Ira Gjenero-Margan, P. Kes. Sudden deaths in chronic
renal failure patients on haemodialysis in Croatia. Cro Med J, 2001; 42:606-610
9. Bosch X: Baxter withdraws dialyser after 12 patients die; Fatal incidents on
hemodialysis in Spain, BMJ. 2001; 323:529-529a.
10. CDC Conference Bridge Confirmation, FDA, Conference call. Oct 17, 2001.
11. Baxter; Corporate news: Following extensive investigation, Baxter identifies
probable cause of recent hemodialysis patient deaths, November 5, 2001