Supplemental Materials: QC and PT data

Routine quality control (QC), verification, calibration and maintenance of the on-board Beckman Coulter HmX analyzers and off-site Siemens Sysmex CA-500 SERIES analyzer were performed to ensure accuracy and precision ( Dry run replicates (within day precision) were performed at the beginning of each site on split samples from healthy volunteers. Precision of the repeats and differences between the split samples did not exceed the coefficient of variance (CVs) and limits shown in Supplemental Table 2. Reproducibility was also assessed daily by measuring one sample 11 times every 6 hour shift, and did not exceed the limits shown in Supplemental Table 3. Precision studies for fibrinogen resulted in CVs <2.5% across the measuring range (Supplemental Table 4). The external QC program involved proficiency testing using College of American Pathologists (CAP) FH6 surveys performed three times a year, consisting of 5 whole blood aliquots, as well as Quality Management Program – Laboratory Services (QMP-LS) surveys consisting of three CBC and automated leukocyte differential surveys. All samples tested on the mobile clinics met the performance evaluation criteria of both CAP and QMP-LS programs, indicating that the quality of performance was typical for the Beckman Coulter family of analyzers (Supplemental Table 5; Supplemental Figure 1).

Supplemental Table 2. Dry-run replicate precision results and repeating limits for CBC parameters run during cycle 1 and cycle 2*
Analyte/parameter / Method CV (%) / #Allowable difference (%)
WBC / 2.5 / 7.5
RBC / 2.0 / 6.0
Hemoglobin / 1.5 / 4.5
MCV / 2.0 / 6.0
Platelets / 5.0 / 15.0
MPV / 3.0 / 9.0
Neutrophils / ≤4.8 / ≤14.4
Lymphocytes / ≤4.8 / ≤14.4
Monocytes / ≤3.2 / ≤9.6
Eosinophils / ≤1.6 / ≤4.8
Basophils / ≤1.6 / ≤4.8
*At the beginning of each MEC 6 week rotation, dry runs were performed on volunteer blood samples. The samples were split and analyzed immediately on-board.Precision of the repeats and differences between the split samples did not exceed the CVs and limits shown.
#Allowable difference limits on split samples were taken from the manufacturer’s information sheet.
Supplemental Table 3. Daily reproducibility limits for within-day precision
Analyte/parameter / Absolute change / CV (%)
WBC / ± 0.4 x 103 cells/l / 5
RBC / ± 0.20 x 106 cells/l / 2
Hemoglobin / ± 0.3 g/dL / 2
Platelets / ± 20 x 103 cells/l / 7
*Reproducibility was performed on 1 sample analyzed 10-11 times every 6 hour shift. Daily runs did not exceeded either an absolute change or CV as shown.
Supplemental Table 4: Within-day precision for fibrinogen*
Fibrinogen concentration (mg/dL) / CV (%)
89 / 0
95 / 2.3
110 / 2.4
260 / 0
280 / 0
*Siemens Thrombin Reagent assay was used to measure fibrinogen concentrations in normal and pathological controls and patient pools. Eight determinations per day over 5 days (n = 40) were performed using the Sysmex CA-1500 analyzer
Supplemental Table 5: Proficiency testing results for CAP and QMP-LS surveys
aCollege of American Pathologist (CAP) surveys / bQuality Management Program – Laboratory Services (QMP-LS) surveys
Analyte/parameter / Cycle 1 & 2 results
(FH6 A-C survey) / Analyte / cCycle 1 & 2
(HEMA survey)
WBC / No flags / WBC / No flags
RBC / No flags / RBC / No flags
Hemoglobin / No flags / Hemoglobin / No flags
Hematocrit / No flags / Hematocrit / No flags
MCV / No flags / MCV / No flags
RDW / No flags / Thrombocyte / No flags
Platelet count / No flags / Lymphocytes / No flags
Granulocytes / No flags / Monocytes / No flags
Lymphocytes / No flags / Granulocytes / No flags
Monocytes / No flags / Eosinophils / No flags
Eosinophils / No flags / Basophils / No flags
Basophils / No flags
aCAP QC material were run at different sites in 2007-2010. All samples tested were within the standard deviation index (SDI) and received 100% satisfaction scores.
bQMP-LS Comprehensive Hematology with Automated Differential HEMA surveys were run in 2007-2010. All samples tested were within the acceptable percent allowable differences (PAD) scores and ranges given by QMP-LS.

Supplemental Figure 1. Proficiency testing data. Example of a 2010 survey report from CAP (A) and QMP-LS (B)demonstrating satisfactory performance of all analytes/parameters run on-board the Beckman Coulter HmX analyzer.