- Email: [email protected]
Frequency of Elevated Serum Aluminum Levels in Adult Dialysis Patients
Frequency of Elevated Serum Aluminum Levels in Adult Dialysis Patients Jared A. Jaffe, DO, Carol Liftman, MS, RD, and Joel D. Glickman, MD ● Background: Aluminum toxicity, a complication described in dialysis patients exposed to aluminum-containing phosphate binders or water used in dialysate preparations containing elevated levels of aluminum, can cause significant morbidity. Although physicians monitor patient aluminum levels, the frequency of abnormal aluminum levels has not been reported. Methods: We retrospectively examined 1,410 measurements of serum aluminum in 207 dialysis patients at the Franklin Dialysis Center (FDC; Philadelphia, PA) from January 1, 2000, through April 3, 2003. We also surveyed serum aluminum levels from a nationwide dialysis provider (DaVita) to more than 43,000 patients. Local dialysis facilities were surveyed to determine the annual frequency of aluminum testing in their patients. Results: Of 1,410 serum aluminum measurements performed at FDC, only 30 abnormal levels (2.1%) were found. DaVita measured approximately 117,000 aluminum levels annually during the last 3 years, of which 2.5% were abnormal. Frequencies of abnormal aluminum levels declined significantly each year at FDC and DaVita. Conclusion: The current frequency of abnormal aluminum levels in our dialysis facility and a national dialysis provider is extremely low and has significantly declined. Am J Kidney Dis 46:316-319. © 2005 by the National Kidney Foundation, Inc. INDEX WORDS: Aluminum; dialysis; phosphate binders.
A
LUMINUM TOXICITY is a complication described in dialysis patients exposed to aluminum in dialysate preparations and aluminum-containing phosphate binders. Manifestations of aluminum toxicity include encephalopathy, erythropoietin-resistant anemia, and bone disease.1-4 Aluminum was eliminated successfully from dialysate in the early 1980s by using reverse osmosis and deionization techniques. Thereafter, aluminum-containing phosphate binders became the predominant source of aluminum exposure in dialysis patients.5,6 Recommendations were made to follow up serum aluminum levels as a screen for patients at risk for aluminum toxicity at a time when aluminum exposure was common in dialysis patients. 5 Aluminum-containing phosphate binder–related aluminum toxicity led to the preferential use of non–aluminum-containing phosphate binders.6,7 From the Department of Medicine and Section on Renal Diseases and Hypertension, Pennsylvania Hospital; and Franklin Dialysis Center (DaVita), Philadelphia, PA. Received December 9, 2004; accepted in revised form April 18, 2005. Originally published online as doi:10.1053/j.ajkd.2005.04.020 on June 16, 2005. Address reprint requests to Joel D. Glickman, MD, 230 W Washington Square, Ste 100, Philadelphia, PA 19106. Email: [email protected] © 2005 by the National Kidney Foundation, Inc. 0272-6386/05/4602-0016$30.00/0 doi:10.1053/j.ajkd.2005.04.020 316
The National Kidney Foundation–Kidney Disease Outcomes Quality Initiative, an opinion-based guideline that covers aluminum testing, recommends at least yearly aluminum level measurement to assess aluminum exposure and risk for aluminum toxicity and determinations every 3 months in patients administered aluminum-containing medications. 8 Although many physicians order screening aluminum levels, the usefulness of this test has not been evaluated. We report the frequency of abnormal aluminum levels in 2 cohorts of dialysis patients. METHODS A retrospective cohort study was performed at the DaVita Franklin Dialysis Center (FDC) located in Philadelphia, PA. A computerized database was used to retrieve serum aluminum levels routinely collected in 207 hemodialysis and peritoneal dialysis patients from January 1, 2000, through April 3, 2003. Measurements were reported quarterly during the indicated period. Serum aluminum levels were measured on blood samples by using trace mineral tubes with the technique of graphite furnace atomic absorption spectrophotometry. The method of aluminum testing of serum and dialysate did not change during the study period at FDC. Recent Kidney Disease Outcomes Quality Initiative opinionbased guidelines recommend that baseline aluminum levels be less than 20 g/L (⬍0.742 mol/L), but the basis for this cutoff value is not stated. Additional testing (deferoxamine test) is recommended for levels greater than 60 g/L (⬎2.226 mol/L), but no specific intervention is recommended for levels between 20 and 60 g/L (0.742 and 2.226 mol/L).8 We elected to use an abnormal cutoff value greater than 40 g/L (⬎1.484 mol/L) for FDC dialysis patients, as recommended by previous studies.4 Patients with abnormal serum aluminum levels were identified. Medical records of patients
American Journal of Kidney Diseases, Vol 46, No 2 (August), 2005: pp 316-319
ALUMINUM LEVELS IN ADULT DIALYSIS PATIENTS
with abnormal serum aluminum levels were reviewed to determine aluminum exposure. We reviewed aluminum levels measured in the dialysate water source at FDC biannually from August 1999 through January 2003. Dialysate aluminum values were reported in parts per million (ppm) or milligrams per liter. The reference range for aluminum content in water is defined as less than 0.01 ppm or less than 0.01 mg/L (10 g/L or 0.371 mol/L).5,6 We surveyed data from a nationwide dialysis provider (DaVita). DaVita also has a computerized database to identify elevated aluminum levels in their hemodialysis and peritoneal dialysis patient population. The cutoff value set by the laboratory identified patients with aluminum levels greater than 50 g/L (⬎1.855 mol/L). The same method was used for aluminum measurements and did not change during the study period. Medical records were not available for these patients. Thirty-three local outpatient dialysis units were surveyed randomly by telephone to determine how often they perform routine serum aluminum determinations. The area surveyed was limited to the Philadelphia region. Differences between proportions of groups were assessed by means of chi-square analysis or Fisher exact test, when appropriate. P less than 0.05 is considered statistically significant.
RESULTS
We retrospectively examined 1,410 serum aluminum measurements in 207 patients at FDC from January 1, 2000, through April 3, 2003. Of 1,410 serum aluminum measurements, only 30 abnormal levels (2.1%) were detected in 16 patients. Of the 16 patients with elevated aluminum levels, levels normalized in 15 patients on subsequent determinations. One patient died within 3 months of starting dialysis therapy, and no additional aluminum level was measured. Review of medical records showed that 6 of 16 patients (37.5%) with elevated aluminum levels were not administered aluminum-containing phosphate binders and had a single sporadic abnormal level that subsequently normalized spontaneously without intervention. The other 9 of 16 patients were administered aluminum-containing phosphate binders. However, all aluminum levels normalized after switching to non–aluminum-containing phosphate-binding therapy. The frequency of elevated aluminum levels decreased significantly each year at FDC (Table 1). In 2000, there were 19 abnormal aluminum levels (5.0%) in 377 serum aluminum determinations. In 2002, the frequency of abnormal aluminum levels declined to 2 of 405 serum aluminum determinations (0.5%; P ⬍ 0.0001). Similarly, in 2003, the
317 Table 1. Frequency of Abnormal Serum Aluminum Levels in 207 Adult Dialysis Patients at FDC
Year
No. of Aluminum Levels
Abnormal*
2000 2001 2002 2003
377 412 405 216
19 8 2 1
% Abnormal
5.0† 1.9‡ 0.5§ 0.5§
NOTE. Level of significance is P ⬍ 0.05. *Reference for abnormal aluminum level is greater than 40 g/L (⬎1.484 mol/L). †P ⫽ 0.0187, 2000 versus 2001; P ⬍ 0.0001, 2000 versus 2002; P ⫽ 0.0017, 2000 versus 2003. ‡P ⫽ 0.1074, 2001 versus 2002; P ⫽ 0.1744, 2001 versus 2003. §P ⫽ 1.0000, 2002 versus 2003.
frequency of abnormal aluminum levels remained at 0.5%. Results for DaVita nationwide aluminum determinations are listed in Table 2. The frequency of abnormal aluminum determinations at DaVita showed similar declines each year. In 2000, of 118,296 total serum aluminum determinations, 3.5% were abnormal. By 2002, the frequency of abnormal aluminum levels declined to 0.8% (P ⬍ 0.0001). Aluminum levels were ordered 4 times a year per patient at FDC. At the time of this study, DaVita provided dialysis services to approximately 43,000 patients. Therefore, given 118,000 aluminum determinations per year, on average, there were 2.7 aluminum determinations per patient per year. Aluminum concentrations in the water source used for dialysate were measured as less than 0.005 ppm on 2 occasions in August 1999 and Table 2. Frequency of Abnormal Serum Aluminum Levels in a Nationwide Dialysis Provider (DaVita)
Year
No. of Aluminum Levels
Abnormal*
2000 2001 2002
118,296 115,072 117,065
4,178 3,650 891
% Abnormal
3.5† 3.2‡ 0.8‡
NOTE. Level of significance is P ⬍ 0.05. *Reference for abnormal aluminum level is greater than 50 g/L (⬎1.855 mol/L). †Chi-square ⫽ 23; P ⬍ 0.001, 2000 versus 2001; chisquare ⫽ 2143; P ⬍ 0.0001, 2000 versus 2002. ‡Chi-square ⫽ 1,758; P ⬍ 0.0001, 2001 versus 2002.
318
JAFFE, LIFTMAN, AND GLICKMAN
January 2000, then less than 0.001 ppm on subsequent determinations. Thirty-three local outpatient dialysis centers were surveyed at random by telephone to determine the frequency of routine serum aluminum testing in the Philadelphia area. Results of the telephone survey are listed in Table 3. Twentyfour dialysis centers responded, whereas the remaining 9 centers would not disclose this information. Numbers of aluminum determinations ranged from 0 to 12 per year. A mode of 4 determinations per year was found in 9 of the 24 dialysis centers. DISCUSSION
Serious complications of aluminum toxicity include acute aluminum neurotoxicity, aluminum encephalopathy, and aluminum-related bone disease. Typically, aluminum levels are dramatically elevated in patients with acute aluminum toxicity (400 to 1,000 g/L [14.84 to 37.1 mol/ L]), as well as in those with aluminum encephalopathy (150 to 300 g/L [5.57 to 11.13 mol/ L]).8 Serum aluminum levels are not good indicators of chronic aluminum exposure, but rather reflect recent exposure to aluminum. Although aluminum levels correlate with aluminumrelated bone disease, used as a screening test, no aluminum level completely discriminates between patients with and without aluminumrelated bone disease. Using a cutoff value of 40 g/L (1.484 mol/L) for abnormal aluminum level, the sensitivity of the test was only 65.2% and 76.7% at a time when the prevalence of the disease was relatively high, respectively.4 The frequency of abnormal aluminum levels in dialysis patients has not been evaluated before this study. Historically, aluminum screening was performed in an attempt to identify patients with aluminum toxicity or those at risk for the disease.5 Aluminum has been effectively eliminated from dialysate, and aluminum-containing phosphate binders have been replaced with non–aluminum-containing phosphate-binding therapy.6,7 Despite these changes, routine serum aluminum testing continues in dialysis patients, as shown at FDC, DaVita, and Philadelphia-area dialysis facilities. Results of our study show that the frequency of abnormal aluminum levels in our dialysis facility and a national dialysis provider are extremely low. Moreover, the frequency of abnormal aluminum levels signifi-
Table 3. Telephone Survey of 33 Philadelphia Dialysis Centers to Determine the Prevalence of Routine Serum Aluminum Testing Determinations/Year
No. of Dialysis Centers
0 1 2 4 12 Unknown
3 7 2 9 3 9
NOTE. Unknown refers to dialysis centers that would not respond to the telephone survey.
cantly declined at both FDC and DaVita during a 3-year period as newer non–aluminum-containing phosphate binders appeared on the market. The majority of abnormal aluminum levels at FDC were related to aluminum exposure. Discontinuation of aluminum-containing phosphate binders resulted in normalization of aluminum levels. However, 37.5% of abnormal levels were unrelated to ingestion of aluminum-containing compounds and were sporadically and intermittently abnormal. Thus, these abnormal values likely were erroneous and did not reflect aluminum-related disease. Alternatively, transient isolated aluminum level elevations may have been caused by brief undetected exposure to aluminum that could not be determined by a retrospective review of medical records. Retrospectively, we could not reliably quantitate the use of phosphate binders in FDC patients or document the use of aluminum-containing compounds before the initiation of dialysis therapy. Therefore, we can only speculate that the decreased frequency of abnormal aluminum levels shown in this study is caused by and consistent with the prevailing practice patterns of decreased use of aluminum-containing phosphate binders. It also is conceivable that the decrease in frequency of abnormal aluminum levels may reflect changes in characteristics of prevalent dialysis patients during the study period. Over time, vintage dialysis patients with high aluminum exposure may have died and were replaced with incident dialysis patients with less aluminum exposure. We cannot prove this suggestion. This study does not address the predictive value of abnormal serum aluminum levels or clarify the role of aluminum testing in patients who have aluminum exposure or may have alu-
ALUMINUM LEVELS IN ADULT DIALYSIS PATIENTS
minum-related diseases. Nevertheless, when the frequency of abnormal levels in a population is extremely low, the utility of that test as a screening tool becomes negligible. With an abnormal frequency of 0.5% to 0.8% and at a significant monetary cost, routine serum aluminum determinations are unlikely to be cost-effective. Instead, aluminum testing may be more useful in patients with aluminum exposure or manifestations of aluminum toxicity. Routine testing of dialysate aluminum should continue, and the use of aluminum-containing compounds should be avoided. Although the frequency of abnormal aluminum levels is extremely low, aluminum toxicity can still occur. Nephrologists should remain aware of the potential for aluminum toxicity and be vigilant in preventing aluminum exposure. ACKNOWLEDGMENT The authors thank Debra Hollister, Director Quality Management (DaVita), for providing DaVita data; DaVita for allowing us to use data from their patients; and Pedro C. Fernandez, MD, for his insight and encouragement in preparing this report.
319
REFERENCES 1. Parkinson I, Feest T, Ward M, Ward R, Kerr D: Fracturing dialysis osteodystrophy and dialysis encephalopathy: An epidemiological survey. Lancet 1:406-409, 1979 2. McGonigle R, Parsons V: Aluminum-induced anaemia in haemodialysis patients. Nephron 39:1-9, 1985 3. Alfrey A, LeGendre G, Kaehny W: The dialysis encephalopathy syndrome: Possible aluminum intoxification. N Engl J Med 294:184-188, 1976 4. Kausz A, Antonsen J, Hercz G, et al: Screening plasma aluminum levels in relation to aluminum bone disease among asymptomatic dialysis patients. Am J Kidney Dis 34:688693, 1999 5. Savory J, Berlin A, Courtoux C, Yeoman B, Wills M: Summary report of an international workshop on “The role of biological monitoring in the prevention of aluminum toxicity in man: Aluminum analysis in biological fluids.” Ann Clin Lab Sci 13:444-451, 1989 6. Slatopolsky E: The interaction of parathyroid hormone and aluminum in renal osteodystrophy. Kidney Int 31:842854, 1987 7. Delmez JA, Slatopolsky E: Hyperphosphatemia: Its consequences and treatment in patients with chronic renal failure. Am J Kidney Dis 19:303-317, 1992 8. National Kidney Foundation: K/DOQI Clinical Practice Guidelines for Bone Metabolism and Disease in Chronic Kidney Disease. Am J Kidney Dis 42:S1-S202, 2003 (suppl 3)
A
LUMINUM TOXICITY is a complication described in dialysis patients exposed to aluminum in dialysate preparations and aluminum-containing phosphate binders. Manifestations of aluminum toxicity include encephalopathy, erythropoietin-resistant anemia, and bone disease.1-4 Aluminum was eliminated successfully from dialysate in the early 1980s by using reverse osmosis and deionization techniques. Thereafter, aluminum-containing phosphate binders became the predominant source of aluminum exposure in dialysis patients.5,6 Recommendations were made to follow up serum aluminum levels as a screen for patients at risk for aluminum toxicity at a time when aluminum exposure was common in dialysis patients. 5 Aluminum-containing phosphate binder–related aluminum toxicity led to the preferential use of non–aluminum-containing phosphate binders.6,7 From the Department of Medicine and Section on Renal Diseases and Hypertension, Pennsylvania Hospital; and Franklin Dialysis Center (DaVita), Philadelphia, PA. Received December 9, 2004; accepted in revised form April 18, 2005. Originally published online as doi:10.1053/j.ajkd.2005.04.020 on June 16, 2005. Address reprint requests to Joel D. Glickman, MD, 230 W Washington Square, Ste 100, Philadelphia, PA 19106. Email: [email protected] © 2005 by the National Kidney Foundation, Inc. 0272-6386/05/4602-0016$30.00/0 doi:10.1053/j.ajkd.2005.04.020 316
The National Kidney Foundation–Kidney Disease Outcomes Quality Initiative, an opinion-based guideline that covers aluminum testing, recommends at least yearly aluminum level measurement to assess aluminum exposure and risk for aluminum toxicity and determinations every 3 months in patients administered aluminum-containing medications. 8 Although many physicians order screening aluminum levels, the usefulness of this test has not been evaluated. We report the frequency of abnormal aluminum levels in 2 cohorts of dialysis patients. METHODS A retrospective cohort study was performed at the DaVita Franklin Dialysis Center (FDC) located in Philadelphia, PA. A computerized database was used to retrieve serum aluminum levels routinely collected in 207 hemodialysis and peritoneal dialysis patients from January 1, 2000, through April 3, 2003. Measurements were reported quarterly during the indicated period. Serum aluminum levels were measured on blood samples by using trace mineral tubes with the technique of graphite furnace atomic absorption spectrophotometry. The method of aluminum testing of serum and dialysate did not change during the study period at FDC. Recent Kidney Disease Outcomes Quality Initiative opinionbased guidelines recommend that baseline aluminum levels be less than 20 g/L (⬍0.742 mol/L), but the basis for this cutoff value is not stated. Additional testing (deferoxamine test) is recommended for levels greater than 60 g/L (⬎2.226 mol/L), but no specific intervention is recommended for levels between 20 and 60 g/L (0.742 and 2.226 mol/L).8 We elected to use an abnormal cutoff value greater than 40 g/L (⬎1.484 mol/L) for FDC dialysis patients, as recommended by previous studies.4 Patients with abnormal serum aluminum levels were identified. Medical records of patients
American Journal of Kidney Diseases, Vol 46, No 2 (August), 2005: pp 316-319
ALUMINUM LEVELS IN ADULT DIALYSIS PATIENTS
with abnormal serum aluminum levels were reviewed to determine aluminum exposure. We reviewed aluminum levels measured in the dialysate water source at FDC biannually from August 1999 through January 2003. Dialysate aluminum values were reported in parts per million (ppm) or milligrams per liter. The reference range for aluminum content in water is defined as less than 0.01 ppm or less than 0.01 mg/L (10 g/L or 0.371 mol/L).5,6 We surveyed data from a nationwide dialysis provider (DaVita). DaVita also has a computerized database to identify elevated aluminum levels in their hemodialysis and peritoneal dialysis patient population. The cutoff value set by the laboratory identified patients with aluminum levels greater than 50 g/L (⬎1.855 mol/L). The same method was used for aluminum measurements and did not change during the study period. Medical records were not available for these patients. Thirty-three local outpatient dialysis units were surveyed randomly by telephone to determine how often they perform routine serum aluminum determinations. The area surveyed was limited to the Philadelphia region. Differences between proportions of groups were assessed by means of chi-square analysis or Fisher exact test, when appropriate. P less than 0.05 is considered statistically significant.
RESULTS
We retrospectively examined 1,410 serum aluminum measurements in 207 patients at FDC from January 1, 2000, through April 3, 2003. Of 1,410 serum aluminum measurements, only 30 abnormal levels (2.1%) were detected in 16 patients. Of the 16 patients with elevated aluminum levels, levels normalized in 15 patients on subsequent determinations. One patient died within 3 months of starting dialysis therapy, and no additional aluminum level was measured. Review of medical records showed that 6 of 16 patients (37.5%) with elevated aluminum levels were not administered aluminum-containing phosphate binders and had a single sporadic abnormal level that subsequently normalized spontaneously without intervention. The other 9 of 16 patients were administered aluminum-containing phosphate binders. However, all aluminum levels normalized after switching to non–aluminum-containing phosphate-binding therapy. The frequency of elevated aluminum levels decreased significantly each year at FDC (Table 1). In 2000, there were 19 abnormal aluminum levels (5.0%) in 377 serum aluminum determinations. In 2002, the frequency of abnormal aluminum levels declined to 2 of 405 serum aluminum determinations (0.5%; P ⬍ 0.0001). Similarly, in 2003, the
317 Table 1. Frequency of Abnormal Serum Aluminum Levels in 207 Adult Dialysis Patients at FDC
Year
No. of Aluminum Levels
Abnormal*
2000 2001 2002 2003
377 412 405 216
19 8 2 1
% Abnormal
5.0† 1.9‡ 0.5§ 0.5§
NOTE. Level of significance is P ⬍ 0.05. *Reference for abnormal aluminum level is greater than 40 g/L (⬎1.484 mol/L). †P ⫽ 0.0187, 2000 versus 2001; P ⬍ 0.0001, 2000 versus 2002; P ⫽ 0.0017, 2000 versus 2003. ‡P ⫽ 0.1074, 2001 versus 2002; P ⫽ 0.1744, 2001 versus 2003. §P ⫽ 1.0000, 2002 versus 2003.
frequency of abnormal aluminum levels remained at 0.5%. Results for DaVita nationwide aluminum determinations are listed in Table 2. The frequency of abnormal aluminum determinations at DaVita showed similar declines each year. In 2000, of 118,296 total serum aluminum determinations, 3.5% were abnormal. By 2002, the frequency of abnormal aluminum levels declined to 0.8% (P ⬍ 0.0001). Aluminum levels were ordered 4 times a year per patient at FDC. At the time of this study, DaVita provided dialysis services to approximately 43,000 patients. Therefore, given 118,000 aluminum determinations per year, on average, there were 2.7 aluminum determinations per patient per year. Aluminum concentrations in the water source used for dialysate were measured as less than 0.005 ppm on 2 occasions in August 1999 and Table 2. Frequency of Abnormal Serum Aluminum Levels in a Nationwide Dialysis Provider (DaVita)
Year
No. of Aluminum Levels
Abnormal*
2000 2001 2002
118,296 115,072 117,065
4,178 3,650 891
% Abnormal
3.5† 3.2‡ 0.8‡
NOTE. Level of significance is P ⬍ 0.05. *Reference for abnormal aluminum level is greater than 50 g/L (⬎1.855 mol/L). †Chi-square ⫽ 23; P ⬍ 0.001, 2000 versus 2001; chisquare ⫽ 2143; P ⬍ 0.0001, 2000 versus 2002. ‡Chi-square ⫽ 1,758; P ⬍ 0.0001, 2001 versus 2002.
318
JAFFE, LIFTMAN, AND GLICKMAN
January 2000, then less than 0.001 ppm on subsequent determinations. Thirty-three local outpatient dialysis centers were surveyed at random by telephone to determine the frequency of routine serum aluminum testing in the Philadelphia area. Results of the telephone survey are listed in Table 3. Twentyfour dialysis centers responded, whereas the remaining 9 centers would not disclose this information. Numbers of aluminum determinations ranged from 0 to 12 per year. A mode of 4 determinations per year was found in 9 of the 24 dialysis centers. DISCUSSION
Serious complications of aluminum toxicity include acute aluminum neurotoxicity, aluminum encephalopathy, and aluminum-related bone disease. Typically, aluminum levels are dramatically elevated in patients with acute aluminum toxicity (400 to 1,000 g/L [14.84 to 37.1 mol/ L]), as well as in those with aluminum encephalopathy (150 to 300 g/L [5.57 to 11.13 mol/ L]).8 Serum aluminum levels are not good indicators of chronic aluminum exposure, but rather reflect recent exposure to aluminum. Although aluminum levels correlate with aluminumrelated bone disease, used as a screening test, no aluminum level completely discriminates between patients with and without aluminumrelated bone disease. Using a cutoff value of 40 g/L (1.484 mol/L) for abnormal aluminum level, the sensitivity of the test was only 65.2% and 76.7% at a time when the prevalence of the disease was relatively high, respectively.4 The frequency of abnormal aluminum levels in dialysis patients has not been evaluated before this study. Historically, aluminum screening was performed in an attempt to identify patients with aluminum toxicity or those at risk for the disease.5 Aluminum has been effectively eliminated from dialysate, and aluminum-containing phosphate binders have been replaced with non–aluminum-containing phosphate-binding therapy.6,7 Despite these changes, routine serum aluminum testing continues in dialysis patients, as shown at FDC, DaVita, and Philadelphia-area dialysis facilities. Results of our study show that the frequency of abnormal aluminum levels in our dialysis facility and a national dialysis provider are extremely low. Moreover, the frequency of abnormal aluminum levels signifi-
Table 3. Telephone Survey of 33 Philadelphia Dialysis Centers to Determine the Prevalence of Routine Serum Aluminum Testing Determinations/Year
No. of Dialysis Centers
0 1 2 4 12 Unknown
3 7 2 9 3 9
NOTE. Unknown refers to dialysis centers that would not respond to the telephone survey.
cantly declined at both FDC and DaVita during a 3-year period as newer non–aluminum-containing phosphate binders appeared on the market. The majority of abnormal aluminum levels at FDC were related to aluminum exposure. Discontinuation of aluminum-containing phosphate binders resulted in normalization of aluminum levels. However, 37.5% of abnormal levels were unrelated to ingestion of aluminum-containing compounds and were sporadically and intermittently abnormal. Thus, these abnormal values likely were erroneous and did not reflect aluminum-related disease. Alternatively, transient isolated aluminum level elevations may have been caused by brief undetected exposure to aluminum that could not be determined by a retrospective review of medical records. Retrospectively, we could not reliably quantitate the use of phosphate binders in FDC patients or document the use of aluminum-containing compounds before the initiation of dialysis therapy. Therefore, we can only speculate that the decreased frequency of abnormal aluminum levels shown in this study is caused by and consistent with the prevailing practice patterns of decreased use of aluminum-containing phosphate binders. It also is conceivable that the decrease in frequency of abnormal aluminum levels may reflect changes in characteristics of prevalent dialysis patients during the study period. Over time, vintage dialysis patients with high aluminum exposure may have died and were replaced with incident dialysis patients with less aluminum exposure. We cannot prove this suggestion. This study does not address the predictive value of abnormal serum aluminum levels or clarify the role of aluminum testing in patients who have aluminum exposure or may have alu-
ALUMINUM LEVELS IN ADULT DIALYSIS PATIENTS
minum-related diseases. Nevertheless, when the frequency of abnormal levels in a population is extremely low, the utility of that test as a screening tool becomes negligible. With an abnormal frequency of 0.5% to 0.8% and at a significant monetary cost, routine serum aluminum determinations are unlikely to be cost-effective. Instead, aluminum testing may be more useful in patients with aluminum exposure or manifestations of aluminum toxicity. Routine testing of dialysate aluminum should continue, and the use of aluminum-containing compounds should be avoided. Although the frequency of abnormal aluminum levels is extremely low, aluminum toxicity can still occur. Nephrologists should remain aware of the potential for aluminum toxicity and be vigilant in preventing aluminum exposure. ACKNOWLEDGMENT The authors thank Debra Hollister, Director Quality Management (DaVita), for providing DaVita data; DaVita for allowing us to use data from their patients; and Pedro C. Fernandez, MD, for his insight and encouragement in preparing this report.
319
REFERENCES 1. Parkinson I, Feest T, Ward M, Ward R, Kerr D: Fracturing dialysis osteodystrophy and dialysis encephalopathy: An epidemiological survey. Lancet 1:406-409, 1979 2. McGonigle R, Parsons V: Aluminum-induced anaemia in haemodialysis patients. Nephron 39:1-9, 1985 3. Alfrey A, LeGendre G, Kaehny W: The dialysis encephalopathy syndrome: Possible aluminum intoxification. N Engl J Med 294:184-188, 1976 4. Kausz A, Antonsen J, Hercz G, et al: Screening plasma aluminum levels in relation to aluminum bone disease among asymptomatic dialysis patients. Am J Kidney Dis 34:688693, 1999 5. Savory J, Berlin A, Courtoux C, Yeoman B, Wills M: Summary report of an international workshop on “The role of biological monitoring in the prevention of aluminum toxicity in man: Aluminum analysis in biological fluids.” Ann Clin Lab Sci 13:444-451, 1989 6. Slatopolsky E: The interaction of parathyroid hormone and aluminum in renal osteodystrophy. Kidney Int 31:842854, 1987 7. Delmez JA, Slatopolsky E: Hyperphosphatemia: Its consequences and treatment in patients with chronic renal failure. Am J Kidney Dis 19:303-317, 1992 8. National Kidney Foundation: K/DOQI Clinical Practice Guidelines for Bone Metabolism and Disease in Chronic Kidney Disease. Am J Kidney Dis 42:S1-S202, 2003 (suppl 3)