Abstract
Background:
Recently developed blood tubes with a barrier to provide plasma are becoming widespread. We compared 43 biochemical, 35 immunochemical and 7 serology analytes in a BD-Vacutainer® Barricor tube for local clinical validation of this lithium-heparin tube with a barrier.
Methods:
Samples from 70 volunteers were collected in different BD-tubes: a clot-activator tube with gel (SST), a lithium-heparin tube with gel (PST), and a lithium-heparin tube with barrier (BAR). Biases from Bland-Altman plots and 95% confidence intervals were compared with the desirable specification from the Ricos database in order to verify whether measurements from different tubes were significantly different.
Results:
For most of the analytes tested, the measurements using SST, PST or BAR tubes were equivalent. Only BIC, GLU, K, LAD, LPA, P, TP, CTX, Ferritin, HGH, vitD3 and ANTIS showed statistically significant, between-tubes, differences which might have clinical implication.
Conclusions:
The study demonstrates that SST, PST and BAR can be used interchangeably for most of the analytes tested, including serology analytes. This allows the use of the same tube for assaying multiple analytes, increasing the laboratory efficiency while decreasing patients discomfort by minimizing blood withdrawal. (www.actabiomedica.it)
Keywords: blood collection tube, mechanical separator, plasma, serum, serology
B&A plots for chemical analytes: the pair value differences were expressed as real values.
B&A plots for chemical analytes: the pair value differences were expressed as percentage.
B&A plots for immunochemical analytes: the pair value differences were expressed as real values.
B&A plots for immunochemical analytes: the pair value differences were expressed as percentage.
B&A plots for serology analytes: the pair value differences were expressed as real values.
B&A plots for serology analytes: the pair value differences were expressed as percentage.
1. Introduction
The preanalytical phase plays a crucial role in laboratory diagnostic and blood collection is probably its most important aspect (1). Heparin plasma and serum are commonly used matrices. The latter is the preferred specimen for the analysis of biochemical parameters (2,3), nevertheless plasma has some important laboratory advantages like a shorter turnaround time (TAT) due to both the absence of the 30-60 minutes time interval needed for the coagulation process (4) and to a shorter centrifugation step, and allows to obtain a larger volume of sample (about 15-20% more) which increases the number of analysis that can be made on one sample (5). According to the World Health Organization, plasma is preferred to serum because it reflects better the patients’ physiological condition (6) by preventing the changes induced by the coagulation process which causes an increase in some analytes (e.g. potassium) and a decrease of others (e.g. total proteins) (7). In addition, the use of anticoagulant prevents the variations induced by the coagulation factors activated when the needle is inserted. The use of plasma also minimizes the formation of fibrin networks found very frequently in serum tubes for several reasons: the sample arrived quickly in the laboratory (e.g. through pneumatic mail systems) and is centrifuged before clot formation, or because the sample was from patients taking oral anticoagulants or heparin which delayed the formation of the clot (8). The presence in serum of soluble fibrin clots causes, on highly automated analytical lines, frequent sampling alarms requiring re-centrifugation or manual re-run of the sample leading to a large increase of the TAT.
Blood collection tubes (BCT) with a gel separator are often the preferred choice because serum (plasma) is physically separated from clotted whole blood (blood cells) (3). However, some drawbacks may still occur like the non-specific adsorption of the molecule to be analyzed or the release of interfering substances (9). A new BCT, the BD-Barricor tube (BAR), containing lithium heparin as anticoagulant and an innovative mechanical separator has been recently developed. According to the manufacturer BAR will improve the quality of laboratory routine analysis in term of TAT and analytes stability. A few studies comparing BAR with standard plasma or serum tubes have been published (10-13) but they still do not cover the wide range of analytes tested in routine analysis. Furthermore, to the best of our knowledge, the BD-Barricor tube has never been tested before for serological analytes. Given this lack of data, 43 biochemical analytes, 35 immunochemical analytes and 7 serology analytes were tested on a fully automated Roche COBAS8000 instrumentation. The study aimed at verify whether plasma (either standard or BAR tubes) can replace serum for high throughput routine analysis without affecting the normal clinical ranges suggested by the manufacturer or selected by the Laboratory.
2. Materials and methods
2.1. Subjects and blood sampling
A total of 70 apparently healthy volunteers, 29 males and 41 females from the San Raffaele Hospital in Milan, Italy were included in the study during the period April-June 2017. Volunteers were aged between 18 and 70 and had no pregnancy status. During blood collection from the volunteers, no exclusion criteria were applied with the exception of difficulties in blood withdrawal like inability to find a suitable vein. Blood samples were collected after overnight fasting (8-10 hours), between the hours of 08:00 and 10:00 am. Smoking and the consumption of tea or coffee were forbidden from midnight until blood collection. Alcohol consumption was not allowed for 3 days prior to blood sampling. Volunteers were seated in an upright position 1 minute before venipuncture and remained seated during the whole procedure. Blood samples were collected, as described elsewhere (14, 15), into three different BCTs from BD (Becton, Dickinson and Company, NJ): a clot-activator gel-containing tube BD-SST II Advance tube, 3.5 mL, 13x75 mm (SST); a lithium heparinized gel-containing tube BD-PST II, 3.0 mL, 13x75 mm (PST); a lithium heparinized tube with a barrier BD-Barricor, 3 ml, 13x75 mm (BAR). PST and BAR were processed immediately after sample collection whereas SST was incubated for at least 30’ to allow appropriate clotting. Samples were separated by centrifugation at 3000xg for 10’ at 4°C. No visible hemolysis was detected in any sample. Concentration measurements were performed within 4 hours after blood collection.
A total of 85 parameters were measured on a Roche COBAS 8000 device (Roche Diagnostic, Basel, Switzerland). Among them 43 were routine biochemical analytes including albumin (ALB), alkaline phosphatase (ALP), alanine aminotransferase (ALT), amylase (AMS), pancreatic amylase (AMSP), anti-streptolysin O (ASO), aspartate aminotransferase (AST), beta-2 microglobulin (B2MICR), bicarbonate (BIC), total bilirubin (BILT), complement C3 (C3), complement C4 (C4), calcium (Ca), cholinesterase (CHE), creatine kinase (CK), chloride (CL), cholesterol (CHO), creatinine (CREA), C-reactive protein (CRP), iron (Fe), gamma-glutamyl transferase (GGT), glucose (GLU), high-density lipoprotein (HDL), homocysteine (HOM), immunoglobulin A (IGA), immunoglobulin G (IGG), immunoglobulin M (IGM), potassium (K), lactate dehydrogenase (LAD), lipase (LIP), lipoprotein A (LPA), low-density lipoprotein (LDL), magnesium (MG), mucoproteins (MUCO), sodium (NA), procalcitonin (PCT), phosphate (P), rheumatoid factor (RF), total protein (TP), transferrin (TRF), triglyceride (TG), urea (UREA), uric acid (UA). The 35 routine immunochemical analytes include: alpha-fetoprotein (AFP), vitamin B12 (B12), beta human chorionic gonadotropin (BHCG), cancer antigen 125 (CA125), cancer antigen 15-3 (CA153), carcinoembryonic antigen (CEA), creatine kinase-muscle/brain (CKMB), cortisol (CORT), peptide-C (CPEP), calcitonin (CA), serum C-telopeptide (CTX), estradiol (E2), ferritin, folate, follicle-stimulating hormone (FSH), free triiodothyronine (FT3), free thyroxine (FT4), cancer antigen 19-9 (GICA), osteocalcin (GLA), growth hormone (HGH), human prolactin (HPRL), insulin (INS), Luteinizing hormone (LH), myoglobin (MIOG), N-terminal-pro BNP (PROBNP), progesterone (PROG), total prostate-specific antigen (PSA), triiodothyronine (T3), thyroxine (T4), testosterone (TESTO), thyroglobulin antibodies (TGAB), anti-thyroid peroxidase (TPO), high-sensitivity cardiac troponin T (hs-CTnT), thyroid stimulating hormone (TSH), and vitamin D3 (vitD3). A total of 7 routine serology analytes were also measured: anti-hepatitis B antibodies (ANTIS), Cytomegalovirus IgG antibodies (CMVG), Cytomegalovirus IgM antibodies (CMVM), Rubella IgG antibodies (RUBEOG), Rubella IgM antibodies (RUBEOM), Toxoplasma gondii IgG antibodies (TOXOG) and Toxoplasma gondii IgM antibodies (TOXOM).
Table 1 shows a brief description of the method used for each analyte.
Table 1.
List of the analytes measured in this study and their corresponding methodology
| Biochemical | Immunochemical | ||
| Analyte | Method | Analyte | Method |
| ALB | Immunoturbidimetric assay | AFP | electrochemiluminescence |
| ALP | Colorimetric assay | B12 | electrochemiluminescence |
| ALT | Spettrophotometric assay | BHCG | electrochemiluminescence |
| AMS | Enzymatic-colorimetric | CA125 | electrochemiluminescence |
| AMSP | Enzymatic-colorimetric | CA153 | electrochemiluminescence |
| ASO | Immunoturbidimetric assay | CEA | electrochemiluminescence |
| AST | Spettrophotometric assay | CKMB | electrochemiluminescence |
| B2MICR | Immunoturbidimetric assay | CORT | electrochemiluminescence |
| BIC | Enzymatic assay | CPEP | electrochemiluminescence |
| BILT | Colorimetric assay | CT | electrochemiluminescence |
| C3 | Immunoturbidimetric assay | CTX | electrochemiluminescence |
| C4 | Immunoturbidimetric assay | E2 | electrochemiluminescence |
| Ca | Colorimetric assay | Ferritin | electrochemiluminescence |
| CHE | Colorimetric assay | Folate | electrochemiluminescence |
| CK | Spettrophotometric assay | FSH | electrochemiluminescence |
| CL | Potentiometric assay | FT3 | electrochemiluminescence |
| CHO | Enzymatic/colorimetric assay | FT4 | electrochemiluminescence |
| CREA | Colorimetric assay | GICA | electrochemiluminescence |
| CRP | Immunoturbidimetric assay | GLA | electrochemiluminescence |
| FE | Colorimetric assay | HGH | electrochemiluminescence |
| GGT | Enzymatic/colorimetric assay | HPRL | electrochemiluminescence |
| GLU | Enzymatic assay | HS-CTnT | electrochemiluminescence |
| HDL | Enzymatic/colorimetric assay | INS | electrochemiluminescence |
| HOM | Enzymatic assay | LH | electrochemiluminescence |
| IGA | Immunoturbidimetric assay | MIOG | electrochemiluminescence |
| IGG | Immunoturbidimetric assay | PROPNB | electrochemiluminescence |
| IGM | Immunoturbidimetric assay | PROG | electrochemiluminescence |
| K | Potentiometric assay | PSA | electrochemiluminescence |
| LAD | Spettrophotometric assay | T3 | electrochemiluminescence |
| LIP | Enzymatic/colorimetric assay | T4 | electrochemiluminescence |
| LPA | Turbidimetric assay | TESTO | electrochemiluminescence |
| LDL | Enzymatic/colorimetric assay | TGAB | electrochemiluminescence |
| MG | Colorimetric assay | TPO | electrochemiluminescence |
| MUCO | Immunoturbidimetric assay | TSH | electrochemiluminescence |
| NA | Potentiometric assay | VitD3 | electrochemiluminescence |
| PCT | Immunoturbidimetric assay | Serology | |
| P | Spettrophotometric assay | ANTIS | electrochemiluminescence |
| RF | Immunoturbidimetric assay | CMVG | electrochemiluminescence |
| TP | Colorimetric assay | CMVM | electrochemiluminescence |
| TRF | Immunoturbidimetric assay | RUBEOG | electrochemiluminescence |
| TG | Enzymatic/colorimetric assay | RUBEOM | electrochemiluminescence |
| UREA | Enzymatic assay | TOXOG | electrochemiluminescence |
| UA | Enzymatic/colorimetric assay | TOXOM | electrochemiluminescence |
Individuals signed an informed consent authorizing the use of their anonymously collected data for retrospective observational studies (article 9.2.j; EU general data protection regulation 2016/679 [GDPR]), according to the San Raffaele Hospital policy (IOG075/2016).
2.2. Statistical analyses
Statistical analyses and graphs were performed with the software Sigmaplot (Systat-Software, Inc. San Jose, CA, USA) and Excel (Microsoft, Redmond, WA, USA). Comparisons between SST, PST and BAR were assessed by the Bland-Altman (BA) plot (17). To avoid disproportionate weights due to analytes having wide concentration ranges the calculated mean bias, and the corresponding 95% confidence interval (95%CI) were expressed as percentage. The latter was compared with the desirable specification (B%) obtained from the Ricos database (18). ASO, MUCO, PCT, UA, BHCG, CT, vitD3 and serological analytes, for which B% was not available, were compared with a 5% arbitrary threshold. The 95CI% was calculated as: bias ± t(0.025; df=n-1)SE, where bias is the calculated % mean bias, SE the standard error of the n differences, with t from the t distribution with n-1 degrees of freedom. The mean %bias was considered statistically significant if its calculated 95%CI did not included the zero; if the 95%CI also exceeded the B%, the mean %bias was considered clinically significant. However, if the 95CI% exceeded the B% but contained also the zero, we cautiously preferred not to make any statement.
3. Results
Collected blood was first tested for hemolysis by measuring the free hemoglobin (fHb), using the hemolysis index (HI). The Roche instrumentation estimates the HI by dichromatic wavelength paired measurement, providing results as absolute numbers, where one unit corresponds to 0.01 g/L. PST and BAR tubes showed fHb of 0.045 and 0.040 g/L, respectively whereas the SST tubes exhibit slightly higher hemolysis (0.070 g/L). However, after a one way ANOVA test, only SST and BAR showed a statistically significant difference.
Table 2-4 show the summary of the BA comparisons between the three BCTs and the corresponding B%.
Table 2.
Biochemical analytes. For each comparison is shown the number of tests (n), the BA Bias calculated as percentage (%bias), the confidence interval (95CI%) and the biological variation expressed as desirable specification for inaccuracy (B%) (18). Analytes with a 95CI% exceeding B% are highlighted in grey
| Analyte | n | SST vs Barricor | SST vs PST | Barricor vs PST | B% | |||
| %bias | 95CI% | %bias | 95CI% | %bias | 95CI% | |||
| ALB | 69 | 0.9 | 0.32, 1.39 | 0.8 | 0.16, 1.37 | -0.2 | -0.75, 0.37 | 1.4 |
| ALP | 69 | 2.0 | 1.39, 2.92 | 2.7 | 2.41, 3.67 | 0.7 | 0.16, 1.63 | 6.7 |
| ALT | 70 | 3.9 | 1.39, 5.96 | -1.4 | -3.80, 0.64 | -5.3 | -7.52, -2.97 | 11.5 |
| AMS | 70 | -0.7 | -1.20, 0.22 | -0.4 | -0.91, 0.00 | 0.3 | -0.24, 0.77 | 7.4 |
| AMSP | 70 | -0.5 | -1.34, 0.25 | 0.0 | -0.74, 0.74 | -0.5 | -0.04, 1.12 | 8.0 |
| ASO | 70 | -2.0 | -4.51, 0.61 | -0.9 | -3.30, 1.48 | 1.1 | -1.48, 3.30 | 5* |
| AST | 62 | 1.2 | -1.61, 4.02 | -1.2 | -4.52, 1.98 | -2.4 | -0.27, 5.14 | 6.5 |
| B2MICR | 66 | 0.6 | 0.07, 1.22 | 0.7 | 0.16, 1.33 | 0.2 | -0.29, 0.62 | 4.1 |
| BIC | 68 | -5.9 | -7.79, -3.77 | -5.9 | -7.91, -4.00 | -0.3 | -1.78, 1.19 | 1.6 |
| BILT | 69 | -0.6 | -1.91, 0.79 | -0.7 | -1.98, 0.85 | -0.2 | -1.64, 1.29 | 8.9 |
| C3 | 70 | -3.1 | -4.04, 0,75 | -0.3 | -1.13, 0.37 | 2.8 | 0.45, 3.58 | 4.1 |
| C4 | 68 | 0.0 | -0.81, 0.74 | 0.8 | 0.06, 1.49 | 0.8 | 0.02, 1.47 | 8.6 |
| Ca | 69 | 0.8 | 0.08, 1.73 | -0.8 | -1.58, -0.25 | -1.8 | -2.51, -1.10 | 0.8 |
| CHE | 70 | 0.5 | -0.08, 1.18 | 0.8 | 0.21, 1.40 | 0.3 | -0.28, 0.81 | 4.8 |
| CK | 69 | 2.2 | 1.19, 3.19 | 1.6 | 0.70, 2.39 | -0.6 | -1.62, 0.34 | 11.5 |
| CL | 70 | 0.3 | -0.14, 0.46 | -0.1 | -0.38, 0.25 | -0.4 | -0.48, -0,08 | 0.5 |
| CHO | 69 | 0.1 | -0.43, 0.72 | 1.0 | 0.37, 1.68 | 0.8 | 0.22, 1.45 | 4.1 |
| CREA | 70 | -1.5 | -2.45, -0.50 | -1.0 | -1.81, -0.03 | 0.5 | -0.49, 1.61 | 4.0 |
| CRP | 70 | -2.1 | -3.87, -0.13 | -0.3 | -1.47, 0.96 | 1.7 | -3.98, 0.48 | 21.8 |
| FE | 69 | 1.3 | 0.50, 2.00 | 1.6 | 0.85, 2.19 | 0.3 | -0.10, 0.75 | 8.8 |
| GGT | 70 | 2.2 | 0.53, 5.13 | 0.9 | -0.87, 2.50 | -1.3 | -0.37, 4.41 | 11.1 |
| GLU | 70 | 3.9 | 2.42, 5.45 | 2.3 | 0.53, 4.20 | -1.6 | -2.72, -0.47 | 2.3 |
| HDL | 70 | -0.1 | -0.74, 0.59 | 0.0 | -0.43, 0.47 | 0.1 | -0.71, 0.47 | 5.6 |
| HOM | 63 | 0.1 | -0.94, 1.17 | -0.9 | -2.07, 0.23 | -1.2 | 0.23, 2.09 | 8.6 |
| IGA | 70 | -0.2 | -1.08, 0.70 | 0.5 | -0.67, 1.65 | 0.6 | -1.59, 0.29 | 9.1 |
| IGG | 69 | 0.5 | 0.06, 0.95 | 0.7 | 0.23, 1.13 | 0.1 | -0.62, 0.32 | 4.3 |
| IGM | 69 | 0.4 | -0.12, 1.02 | 1.6 | 0.67, 2.47 | 1.1 | -2.01, -0.24 | 11.9 |
| K | 70 | 7.6 | 6.47, 8.62 | 7.2 | 6.25, 8.23 | -0.4 | -1.21, 0.46 | 1.8 |
| LAD | 70 | -3.9 | -6.09, -1.40 | -3.9 | -6.24, -1.24 | 0.1 | -2.38, 2.39 | 4.3 |
| LIP | 68 | 0.1 | -0.52, 0.69 | 0.3 | -0.37, 0.84 | 0.1 | -0.67, 0.53 | 11.3 |
| LPA | 54 | -4.7 | -8.24, -1.22 | -4.4 | -7.70, -1.03 | -0.3 | -1.87, 2.48 | 3.7 |
| LDL | 68 | 0.3 | -0.19, 0.88 | 0.6 | 0.06, 1.19 | 0.3 | -0.83, 0.23 | 5.5 |
| MG | 69 | 0.1 | -0.61, 0.82 | 0.3 | -0.31, 0.92 | 0.2 | -0.83, 0.40 | 1.8 |
| MUCO | 70 | 0.2 | -0.87, 0.90 | 0.9 | 0.09, 1.61 | 0.6 | -1.45, 0.23 | 5* |
| NA | 69 | 0.0 | -0.19, 0.19 | 0.2 | 0.01, 0.39 | 0.2 | -0.47, -0.09 | 0.2 |
| PCT | 64 | -4.9 | -19.19, 7.69 | -34.3 | -50.87, -17.69 | -33.5 | -15.51, -51.39 | 5* |
| P | 70 | 7.6 | 6.67, 8.31 | 5.0 | 3.96, 6.05 | -2.6 | -1.42, -3.28 | 3.4 |
| RF | 69 | 0.5 | -0.02, 0.66 | 0.7 | -0.01, 1.27 | 0.3 | -0.15, 0.66 | 6.5 |
| TP | 69 | -4.5 | -5.20, -3.70 | -4.1 | -4.95, -3.25 | -0.4 | -0.98, 0.25 | 1.4 |
| TRF | 68 | 0.1 | -0.70, 0.98 | 1.3 | 0.42, 2.04 | -1.2 | -1.91, -0.41 | 1.3 |
| TG | 70 | 0.8 | -0.01, 1.56 | 3.5 | 2.59, 4.40 | 2.7 | 3.46, 1.90 | 9.6 |
| UREA | 70 | -1.5 | -2.49, -0.57 | -0.5 | -1.41, 0.26 | 1.0 | -1.76, 0.01 | 5.6 |
| UA | 70 | 0.3 | -0.58, 1.10 | 0.0 | -0.69, 0.57 | -0.3 | -0.93, 0.50 | 5* |
Table 4.
Serological analytes. For each comparison was shown: the number of tests (n), the BA Bias calculated as percentage (%bias), the confidence interval (CI95%) and the biological variation expressed as desirable specification for inaccuracy (B%) (18). Analytes with a 95CI% exceeding B% are highlighted in grey.
| Analyte | n | SST vs Barricor | SST vs PST | Barricor vs PST | B% | |||
| %bias | CI95% | %bias | CI95% | %bias | CI95% | |||
| ANTIS | 64 | 10.4 | 6.23, 14.64 | 9.9 | 6.60, 13.46 | -1.2 | -2.82, 0.33 | 5* |
| CMVG | 64 | 2.0 | -0.90, 3.01 | 2.3 | -1.53, 3.04 | 0.4 | -1.95, 1.14 | 5* |
| CMVM | 60 | 1.9 | 1.05, 2.79 | 2.2 | 1.07, 3.30 | 0.2 | -0.52, 1.01 | 5* |
| RUBEOG | 63 | -0.1 | -0.72, 0.61 | -0.4 | -1.11, 0.35 | -0.3 | -0.36, 0.85 | 5* |
| RUBEOM | 61 | 0.0 | -0.72, 0.72 | 0.1 | -0.69, 0.93 | 0.1 | -0.59, 0.54 | 5* |
| TOXOG | 66 | 0.1 | -0.39, 1.78 | 0.1 | -2.39, 1.87 | 0.0 | -0.88, 2.73 | 5* |
| TOXOM | 58 | -0.3 | -0.97, 0.38 | -0.8 | -1.39, 0.01 | -0.6 | -0.06, 1.20 | 5* |
Table 3.
Immunochemical analytes. For each comparison was shown: the number of tests (n), the BA Bias calculated as percentage (%bias), the confidence interval (95CI%) and the biological variation expressed as desirable specification for inaccuracy (B%) (18). Analytes with a 95CI% exceeding B% are highlighted in grey.
| Analyte | n | SST vs Barricor | SST vs PST | Barricor vs PST | B% | |||
| %bias | 95CI% | %bias | 95CI% | %bias | 95CI% | |||
| AFP | 63 | 0.4 | -0.39, 1.20 | 0.2 | -0.89, 1.28 | -0.2 | -1.25, 0.80 | 11.8 |
| B12 | 68 | 0.9 | -0.31, 2.11 | 1.5 | 0.31, 2.66 | 0.8 | -0.23, 1.84 | 17.7 |
| BHCG | 70 | -1.2 | -3.04, 0.61 | -0.7 | -2.27, 0.79 | 0.5 | -0.50, 1.46 | 5* |
| CA125 | 65 | 1.6 | 0.88, 2.37 | 2.4 | 1.68, 3.06 | 0.7 | -0.11, 1.60 | 15.0 |
| CA153 | 65 | 1.7 | -0.06, 3.43 | 2.7 | 0.88, 4.46 | 1.0 | -0.80, 2.78 | 15.8 |
| CEA | 65 | 2.0 | 1.03, 2.93 | 2.5 | 1.42, 3.47 | 0.5 | -0.64, 1.58 | 14.3 |
| CKMB | 64 | 2.3 | -0.09, 4.73 | 1.5 | -2.3, 3.49 | 0.1 | -2.05, 2.26 | 7.8 |
| CORT | 64 | -1.1 | -4.02, 1.74 | 1.2 | -1.89, 4.35 | 2.1 | -0.93, 5.19 | 7.6 |
| CPEP | 64 | -1.3 | -3.52, 0.98 | -0.9 | -2.83, 1.05 | 0.4 | -1.47, 2.22 | 7.1 |
| CT | 65 | -0.4 | -2.09, 1.30 | -3.8 | -4.91, 1.66 | -3.5 | -4.98, -1.50 | 5* |
| CTX | 64 | -8.3 | -9.92, -6.40 | -5.2 | -6.92, -3.46 | 3.1 | 1.77, 4.41 | 8.1 |
| E2 | 68 | -2.4 | -12.2, 7.49 | 8.7 | -1.99, 19.32 | 9.1 | -0.68, 18.6 | 8.3 |
| Ferritin | 70 | -3.9 | -6.80, -1.18 | -0.2 | -2.11, 1.31 | 3.7 | 1.04, 6.17 | 5.2 |
| Folate | 69 | -3.2 | -6.36, 0.04 | -3.7 | -7.19, 0.05 | -0.5 | -3.27, 2.25 | 19.2 |
| FSH | 66 | 1.0 | 0.60, 1.40 | 0.9 | 0.35, 1.35 | -0.1 | -0.58, 0.32 | 12.1 |
| FT3 | 68 | -0.4 | -2.51, 1.78 | -0.6 | -2.74, 1.48 | -0.3 | -1.58, 2.12 | 4.8 |
| FT4 | 68 | 0.5 | -0.10, 1.01 | 0.7 | 0.25, 1.29 | 0.3 | -0.19, 0.83 | 3.3 |
| GICA | 64 | 0.4 | -0.15, 0.91 | 0.6 | -0.02, 1.12 | 0.2 | -0.81, 0.32 | 32.9 |
| GLA | 57 | -1.6 | -3.34, 0.21 | -1.3 | -2.96, 1.29 | 0.3 | -1.94, 1.31 | 7.9 |
| HGH | 64 | 13.4 | 4.38, 21.18 | 12.9 | 4.01, 20.94 | -0.3 | -0.31, 0.96 | 12.2 |
| HPRL | 62 | -1.1 | -1.74, -0.51 | -0.7 | -1.03, 0.01 | 0.3 | -1.02, 0.38 | 10.5 |
| HS-CTnT | 70 | -0.6 | -4.27, 3.01 | -3.1 | -5.17, 0.06 | -2.4 | -0.73, 5.67 | 7.0 |
| INS | 65 | 2.9 | -0.71, 6.53 | 5.4 | 2.65, 8.15 | 2.5 | -5.08, 0.10 | 15.5 |
| LH | 64 | -3.8 | -4.55, 2.97 | -3.5 | -4.23, -2.85 | -0.2 | -0.71, 0.36 | 8.9 |
| MIOG | 64 | -3.3 | -5.44, -1.18 | -2.9 | -4.44, -1.43 | 0.4 | -2.25, 0.80 | 8.2 |
| PROBNP | 66 | 0.2 | -2.01, 2.33 | 0.6 | -1.41, 2.59 | 0.5 | -1.87, 2.93 | 4.7 |
| PROG | 66 | -10.3 | -19.39, 0.25 | -2.0 | -14.83, 10.90 | 6.9 | -16.47, 2.63 | 13.5 |
| PSA | 26 | 0.1 | -2.23, 2.30 | -0.7 | -3.19, 1.83 | -0. | -1.04, 2.48 | 18.7 |
| T3 | 64 | 0.6 | -0.25, 1.39 | 0.4 | -0.50, 1.13 | -0.3 | -0.52, 1.02 | 5.2 |
| T4 | 65 | 1.7 | 1.16, 2.30 | 2.4 | 1.86, 2.92 | 0.7 | -1.21, -0.10 | 3.0 |
| TESTO | 28 | -1.7 | -3.78, 0.34 | 0.6 | -1.91, 3.13 | 2.3 | -5.31, 0.66 | 6.0 |
| TGAB | 63 | -32.8 | -39.40, -26.26 | -28.2 | -34.54, -21,85 | 5.3 | -10.21, 0.23 | 20.6 |
| TPO | 61 | -2.5 | -17.95, 13.03 | -18.5 | -31.25, 0.28 | -16.6 | -0.65, 25.44 | 5.7 |
| TSH | 68 | 2.2 | 1.23, 3.22 | -0.2 | -1.12, 0.64 | -2.5 | 1.54, 3.39 | 9.7 |
| vitD3 | 65 | 1.2 | -0.65, 3.75 | -3.2 | -5.30, -0.94 | -4.4 | -6.49, -2.39 | 5* |
3.1. Biochemical analytes
BIC, K, LAD, LPA, P, and TP showed 95CI% clinically significant only when serum was compared to plasma. Within the same analyte, the 95CI% were similar regardless of the type of plasma tube used. In contrast Ca (which was associated to a rather small B%) and GLU, showed significantly different 95CI% in all of the three comparisons. Na and TRF, also associated to small B%, showed significantly different 95CI% only when PPT was used whereas SST and BAR were equivalent.
PCT showed 95CI% significantly different from the arbitrary adopted B% when PPT was used, however, the 95CI% amplitudes were very large. In contrast, when SST was compared to BAR the 95CI% became almost ten time smaller, but still exceeded the B%.
3.2. Immunochemical analytes
HGH and TGAB showed 95CI% significantly different from B% only when serum was compared to plasma. It must be noted that very large %bias were observed for TGAB on these two comparisons. In contrast, Ferritin showed significant differences only when BAR was used whereas, for vitD3, only when PST was used. CTX showed a 95CI% significantly different from B% only when SST was compared to BAR. E2, TPO and PROG showed 95CI% exceeding B%, and containing the zero, in all of the three comparisons.
3.3. Serological analytes
All of the analytes showed 95CI% smaller than the arbitrary adopted 5% threshold, with the exception of ANTIS which showed a 95CI% significantly different from B% when serum was compared to either PST or BAR.
4. Discussion
Among the biochemical analytes BIC, K, LAD, LPA, P and TP are clearly associated to a matrix effect which was considered clinically significant (table 5). The TP and K differences between plasma and serum were expected and attributed to the coagulation process (19). Calcium and GLU showed both a matrix effect and an influence of the new BAR mechanical separator. In the case of GLU the two effects add up when SST is compared to BAR whereas for Ca they have opposite signs. Furthermore the B% for Ca was so small (0.8%) that, although the %bias were significant, we considered them clinically irrelevant (table 5). The same was true for Na and TRF (B%: 0.2 and 1.3% respectively) which showed matrix effects and influences of the new mechanical separator so small as to be considered clinically irrelevant (Table 5).
Table 5.
Differences observed when comparing SST, PST and BAR tubes. NS: no statistically significant difference; SS-CLI: statistically significant difference and likely clinically relevant (highlighted in grey); SS: statistically significant difference only, clinically irrelevant
| Analyte | SST vs BAR | SST vs BAR | BAR vs PST |
| Biochemical: ALB, ALP, ALT, AMS, AMSP, ASO, AST, B2MICR, BILT, C3, C4, CHE, CK, CL, CHO, CREA, CRP, FE, GGT, HOM, HDL, IGA, IGG, IGM, LIP, LDL, MG, MUCO, RF, TG, UA, UREA | NS | NS | NS |
| BIC | SS-CLI | SS-CLI | NS |
| Ca | SSa | SSa | SSa |
| GLU | SS-CLI | SS-CLI | SS-CLI |
| K | SS-CLI | SS-CLI | NS |
| LAD | SS-CLI | SS-CLI | NS |
| LPA | SS-CLI | SS-CLI | NS |
| NA | NS | SSa | SSa |
| PCT | ?b | ?b | ?b |
| P | SS-CLI | SS-CLI | NS |
| TP | SS-CLI | SS-CLI | NS |
| TRF | NS | SSa | SSa |
| Immunochemical: AFP, B12, BHCG, CA125, CA153, CEA, CKMB, CORT, CPEP, Folate, FSH, FT3, FT4, GICA, GLA, HPRL, HS-CTnT, INS, LH, MIOG, PROPNB, PSA, T3, T4, TESTO, TSH | NS | NS | NS |
| CTX | SS-CLI | NS | NS |
| E2 | ?b | ?b | ?b |
| Ferritin | SS-CLI | NS | SS-CLI |
| HGH | SS-CLI | SS-CLI | NSD |
| PROG | ?b | ?b | ?b |
| TGAB | ?b | ?b | NS |
| TPO | ?b | ?b | ?b |
| vitD3 | NS | SS-CLI | SS-CLI |
| Serology: CMVG, CMVM, RUBEOG, RUBEOM, TOXOG, TOXOM | NS | NS | NS |
| ANTIS | SS-CLI | SS-CLI | NS |
aBecause of the relatively small desirable specification (B%), the %bias, although significantly different, was considered clinically irrelevant.
bWe, cautiously, did not state whether the two measurements were equivalent or not.
PCT showed %biases higher than 30% in the SST vs PST and BAR vs PST comparisons (likely arising from the low concentration data associated with the healthy condition of the individuals tested) whereas a %bias lower than the arbitrary adopted 5% was observed in the SST vs BAR comparison. However, in the latter comparison the 95CI% contained the zero and exceeded the desirable specification interval. Thus, we cautiously did not state whether the two measurements were equivalent or not (Table 5).
Among the immunochemical analytes HGH and TGAB were associated to a matrix effect only. Because of the very large %biases and 95CI% observed for TGAB, we prudently did not draw any conclusion for these measurements. We might speculate that the large %biases were consistent with the Roche recommendation for the exclusive use of serum for TGAB determination (table 5). For CTX the matrix effect (SST vs PST) and the mechanical separator effect were both insignificant. However the two effects adds up resulting in a significant difference between SST and BAR (Table 5). A significant difference was observed for Ferritin as well which showed no matrix effect but a pronounced effect of the mechanical separator (Table 5). In contrast, the matrix effect and the effect of the mechanical separator (both significant) observed for vitD3 were of the opposite sign. As a result SST and BAR can be used interchangeably, for vitD3 measurements, whereas replacing SST with PST might have significant clinical implications (Table 5). For E2, PROG and TPO the plasma vs serum comparisons all gave confidence intervals which exceeded B% and, at the same time, contained the zero. This was likely the consequence of the many results falling in the low concentration range and associated with the healthy condition of the individuals tested (Figure S2A-B). Thus we, cautiously, did not state whether the measurements were equivalent or not.
B% was not available for the serology analytes thus an arbitrary 5% threshold was adopted. Among them only ANTIS showed a significant matrix effect which was considered clinically significant (table 5).
5. Conclusion
We demonstrated that plasma tubes, including the new BAR tube, can be used interchangeably with SST for most of the standard biochemical and immunochemical analytes as well as for serology analytes. This is of particular importance because using the same tube for assaying multiple analytes significantly increases the efficiency and effectiveness of the laboratory while decreasing patient discomfort.
For the few analytes showing clinically significant between-tubes differences (Table 5), a new normal clinical range should be calculated in order to guarantee the patients’ safety. It must be also noted that the results showed in this study refers to a Roche COBAS 8000 device and its related assays. Thus, laboratory equipped with different instrumentations might show different outcomes.
Conflict of interest:
Each author declares that he or she has no commercial associations (e.g. consultancies, stock ownership, equity interest, patent/licensing arrangement etc.) that might pose a conflict of interest in connection with the submitted article
References
- 1.Lippi G, Cornes MP, Grankvist K, Nybo M, Simundic AM. EFLM WG-Preanalytical phase opinion paper: Local validation of blood collection tubes in clinical laboratories. Clin Chem Lab Med. 2016;54(5):755–60. doi: 10.1515/cclm-2015-1274. [DOI] [PubMed] [Google Scholar]
- 2.Plebani M, Banfi G, Bernardini S, et al. Serum or plasma? An old question awaiting for new answers. V Biochimica Clinica. 2019;43:178–86. [Google Scholar]
- 3.Lippi G, Giavarina D. A survey on sample matrix and preanalytical management in clinical laboratories. Biochim Clin. 2017;41(2):142–7. [Google Scholar]
- 4.Yu Z, Kastenmüller G, He Y, et al. Differences between human plasma and serum metabolite profiles. PLoS One. 2011;6(7) doi: 10.1371/journal.pone.0021230. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Giavarina D, Fortunato A, Barzon E, et al. Evaluation of BD Vacutainer® PSTTMII tubes for a wide range of immunoassays. Clinical Chemistry and Laboratory Medicine. 2009;47:237–41. doi: 10.1515/CCLM.2009.034. [DOI] [PubMed] [Google Scholar]
- 6.World Health Organization. Use of anticoagulants in diagnostic laboratory: stability of blood, plasma and serum samples. WHO [Internet] 2002:1–62. Available from: http://apps.who.int/iris/bitstream/10665/65957/1/WHO_DIL_LAB_99.1_REV.2.pdf?ua=1 . [Google Scholar]
- 7.Guder WG, Narayanan S, Wisser H, Zawta B. Samples: From the Patient to the Laboratory: The Impact of Preanalytical Variables on the Quality of Laboratory Results: Third Revised Edition. Samples: From the Patient to the Laboratory: The Impact of Preanalytical Variables on the Quality of Laboratory Results: Third Revised Edition. 2007:1–107. [Google Scholar]
- 8.Kocijancic M, Cargonja J, Delic-Knezevic A. Evaluation of the BD Vacutainer? RST blood collection tube for routine chemistry analytes: Clinical significance of differences and stability study. Biochem Medica. 2014;24(3):368–75. doi: 10.11613/BM.2014.039. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Bowen RAR, Hortin GL, Csako G, Otañez OH, Remaley AT. Impact of blood collection devices on clinical chemistry assays. Clinical Biochemistry. 2010;43:4–25. doi: 10.1016/j.clinbiochem.2009.10.001. [DOI] [PubMed] [Google Scholar]
- 10.Arslan FD, Karakoyun I, Basok BI, et al. The local clinical validation of a new lithium heparin tube with a Barrier: BD vacutainer® barricor LH plasma tube. Biochem Medica. 2017;27(3) doi: 10.11613/BM.2017.030706. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Cadamuro J, Mrazek C, Leichtle AB, et al. Influence of centrifugation conditions on the results of 77 routine clinical chemistry analytes using standard vacuum blood collection tubes and the new bd-barricor tubes. Biochem Medica. 2018;28(1) doi: 10.11613/BM.2018.010704. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Padoan A, Zaninotto M, Piva E, et al. Quality of plasma samples and BD Vacutainer Barricor tubes: Effects of centrifugation. Clin Chim Acta. 2018;483:271–4. doi: 10.1016/j.cca.2018.05.018. [DOI] [PubMed] [Google Scholar]
- 13.Dupuy AM, Badiou S, Daubin D, et al. Comparison of BarricorTM vs. lithium heparin tubes for selected routine biochemical analytes and evaluation of post centrifugation stability Anne. Biochem Med. 2018;28(2):1–7. doi: 10.11613/BM.2018.020902. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Ferrari D, Manca M, Premaschi S, Banfi G, Locatelli M. Toxicological investigation in blood samples from suspected impaired driving cases in the Milan area: Possible loss of evidence due to late blood sampling. Forensic Sci Int. 2018;288:211–7. doi: 10.1016/j.forsciint.2018.04.038. [DOI] [PubMed] [Google Scholar]
- 15.Ferrari D, Manca M, Banfi G, Locatelli M. Alcohol and illicit drugs in drivers involved in road traffic crashes in the Milan area. A comparison with normal traffic reveals the possible inadequacy of current cut-off limits. Forensic Sci Int. 2018;282:127–32. doi: 10.1016/j.forsciint.2017.11.005. [DOI] [PubMed] [Google Scholar]
- 16.Ferrari D, Lombardi G, Strollo M, Pontillo M, Motta A, Locatelli M. Association between solar ultraviolet doses and vitamin D clinical routine data in European mid-latitude population between 2006 and 2018. Photochem Photobiol Sci. 2019;18(11):2696–706. doi: 10.1039/c9pp00372j. [DOI] [PubMed] [Google Scholar]
- 17.Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Int J Nurs Stud [Internet] 2010;47(8):931–6. Available from: http://dx.doi.org/10.1016/j.ijnurstu.2009.10.001. [PubMed] [Google Scholar]
- 18.WESTGARD J. Westgard QC [Internet] 2014 Available from: https://www.westgard.com/ [Google Scholar]
- 19.Er T-K, Tsai L-Y, Jong Y-J, Chen B-H. Selected Analyte Values in Serum Versus Heparinized Plasma Using the SYNCHRON LX PRO Assay Methods/Instrument. Lab Med. 2006;37(12):731–2. [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
B&A plots for chemical analytes: the pair value differences were expressed as real values.
B&A plots for chemical analytes: the pair value differences were expressed as percentage.
B&A plots for immunochemical analytes: the pair value differences were expressed as real values.
B&A plots for immunochemical analytes: the pair value differences were expressed as percentage.
B&A plots for serology analytes: the pair value differences were expressed as real values.
B&A plots for serology analytes: the pair value differences were expressed as percentage.