| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Letters to the Editor |
1 Department of Community Medicine, School of Medicine, West Virginia University, Morgantown, WV
2 Lead Poisoning Prevention Branch, Division of Emergency and, Environmental Health Services, National Center for Environmental Health, Centers for Disease Control and Prevention, Atlanta, GA
3 Inorganic Toxicology and, Radionuclide Laboratories, Division of Laboratory Services, National Center for Environmental Health, Centers for Disease Control and Prevention, Atlanta, GA
aAddress correspondence to this author at: Lead Poisoning Prevention Branch, Centers for Disease Control and Prevention, 4770 Buford Highway, NE (MS-F40), Atlanta, GA 30341. Fax 770-488-3635; e-mail mjb5{at}cdc.gov.
To the Editor:
In May, 2005, a partial recall of LeadCare® portable blood lead testing kits was initiated after proficiency testing revealed a negative error (mean, –25%; 95% CI, –15% to –35%) in blood lead concentrations (often called blood lead level in the US) analyzed with this device (1)(2). This bias exceeded previously recommended 95% error thresholds established with graphite furnace atomic absorption spectroscopy (3). The recall included 8 lots of defective sensors containing enough material for an estimated 500 000 patient tests distributed between September 2003 and May 2005.
The LeadCare device uses an anodic stripping voltammetry method to measure lead. Capillary sampling by fingerstick is the most common method of specimen collection. Assessments of the portable units by testing of venous samples from occupationally exposed workers have demonstrated that measurements obtained with these units show an insignificant positive error that decreases as blood lead concentrations increase (4).
A blood lead concentration 
10 µg/dL (100 µg/L) has been established as the concentration at which children should receive clinical intervention (5).
We contacted 15 US laboratories identified as users of the analyzer by ESA Inc., the manufacturer of the testing unit. A total of 26 883 results were provided by 8 (53%) of 15 testing sites. Capillary samples were used to obtain all results reported in this study.
Data were recoded and classified according to age group (0–6, 7–15, or 16 years), test date (January 1–August 31, 2003, and September 1, 2003–June 30, 2005), recall status, and blood lead concentration category. Time periods for this study were determined by the market availability of defective sensors. Time period comparisons were limited to facilities with data from both time periods and with sufficient sample size (n = 15 024). Of 573 total test results, 12 included the lot number of the sensor. One facility provided both test and retest results (n = 95).
The median age of persons tested was 2 years (range 0–94 years), with 91% of all tests performed on children <6 years old, 7% on children 7–15 years old, and 3% on children and adults 16 years old.
Compared with mean blood lead concentrations collected before September 2003, the mean result of tests performed from September 2003 to May 2005 was 29% lower (P <0.001) for children 
6 years old and 22% lower for children 7–15 years old (P <0.001). For children 6 years old, blood lead concentrations 10 µg/dL (100 µg/L) were found in 2.3% tested before September 2003 and in 1.7% tested after August 2003 (P = 0.01).
Children tested with the defective sensors were 4.5 times more likely to have blood lead concentrations 10 µg/dL (100 µg/L). When a defective sensor was used, the measured percentage of blood lead concentrations 10 µg/dL (100 µg/L) decreased in both the 0–6 year (P <0.001) and 7–15 year (P = 0.01) age categories. Table 1
shows an outcome matrix comparing differences in the initial and retest results. An increase of 49.7% was observed in the prevalence of retest results that were above clinically important threshold values and thus sufficient to place the patient in a higher clinical category.
|
These analyses show a negative error among blood lead measurements obtained with defective sensors. These errors led to substantial misclassification, with nearly a 50% rate of underestimation of the prevalence of concentrations above clinically relevant thresholds, and may have resulted in misclassified patients receiving delayed or no treatment. The findings reported here have several limitations. Our analyses are based on the results from a limited number of testing sites. It is possible that the smaller time period before introduction of the defective sensors may influence results. In addition, our estimate of misclassification among those with retest results is based on a small sample of individuals from a single facility and solely on the percentage of persons whose retest result placed them in a higher clinically significant category. Nevertheless, this estimate is consistent with expected proportions of children with blood lead concentrations higher than selected thresholds, given a log normal distribution.
On the basis of these findings, we recommend retesting of individuals whose blood lead concentrations were 6 µg/dL (60 µg/L) when measured with the LeadCare device between September 2003 and May 2005. We also recommend that LeadCare not be used for confirmatory testing of children with increased blood lead concentrations, because validation studies have evaluated the instrument only in the range of 1–42 µg lead/dL (10–420 µg/L) (4). This study underscores the value of proficiency testing as well as the importance of laboratory procedures that may identify unexpected results over time.
Editor’s Note: Results are expressed in this Letter in conventional units first because these units are the ones familiar to many of the intended readers of this report.
Acknowledgments
Grant/funding support: None declared.
Financial disclosures: None declared.
References
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
