Stability of samples collected in citrate-containing tubesStability of WB samples from patients treated with UFH

Anti-factor Xa activities (IU/mL), measured with the Siemens reagent/analyser combination (n = 33), decreased from 0.28 (0.11–0.55) at T1WB to 0.27 (0.12–0.52) at T4WB (p < 0.05) and to 0.27 (0.12–0.53) at T6WB (p < 0.001), but the bias between T1WB and T6WB was limited to -0.02 IU/mL (Fig. 1A,B). For the Stago reagent/analyser combination (n = 31), anti-factor Xa activities decreased from 0.32 (0.12–1.00) at T1WB to 0.28 (0.09–0.90) at T4WB (p < 0.001) and to 0.29 (0.09–0.89) at T6WB (p < 0.001), and the bias between T1WB and T6WB was limited to -0.04 IU/mL (Fig. 1C,D). The agreement between T1WB and T6WB was good for the Siemens (kappa score Ks = 0.73) and excellent for the Stago (Ks = 0.80) reagent/analyser pair.

Fig. 1

Stability of citrated whole blood samples from patients treated with UFH. Anti-factor Xa activity scatter plots (A and C) and Bland–Altman plots (B and D), and aPTT ratio scatter plots (E and G) and Bland–Altman plots (F and H) after analysis with the Siemens and Stago reagent/analyser combination, respectively. Results are plotted as the difference (y axis) in anti-factor Xa activity or aPTT ratio between the analysis after 6 h (T6WB), 4 h (T4WB) and 1 h (T1WB) of storage at room temperature. *P < 0.05, **P < 0.01, ***P < 0.001, NS: Not significant (ANOVA or non-parametric Friedman test)

The aPTT ratio, measured with the Siemens reagent/analyser combination (n = 33), remained stable: 1.55 (0.87–3.13) at T1WB and 1.57 (0.91–2.97) at T6WB (p = 0.07). The bias between T1WB and T6WB was -0.02 (Fig. 1E,F). With the Stago reagent/analyser combination (n = 29), an upward, but not significant, trend was observed for the aPTT ratio: 2.84 (1.19–5.27) at T1WB and 3.33 (1.23–5.52) at T6WB (p = 0.40). The bias between T1WB and T6WB was 0.06 (Fig. 1G,H). The aPTT agreement between T1WB and T6WB was good for both the Siemens (Ks = 0.64) and Stago (Ks = 0.79) reagent/analyser combinations.

Stability of plasma samples from patients treated with UFH

Anti-factor Xa activities measured with the Siemens reagent/analyser combination (n = 33) increased from 0.28 (0.11–0.55) at T1P to 0.29 (0.12–0.60) at T4P (p < 0.05) and to 0.29 IU/mL (0.12–0.62) at T6P (p < 0.05). The bias between T1P and T6P was limited to 0.01 IU/mL (Fig. 2A,B). Using the Stago reagent/analyser combination (n = 31), anti-factor Xa activities decreased from 0.32 (0.12–1.00) at T1P to 0.28 (0.10–0.95) at T6P (p < 0.01), but remained stable between T1P and T4P [0.29 (0.10–0.96); p = 0.11)]. The bias between T1P and T6P was -0.03 IU/mL (Fig. 2C,D). The agreement between T1P and T6P was good for both the Siemens (Ks = 0.76) and Stago (Ks = 0.67) reagent/analyser combinations.

Fig. 2

Stability of citrated plasma samples from patients treated with UFH. Anti-factor Xa activity scatter plots (A and C) and Bland–Altman plots (B and D), and aPTT ratio scatter plots (E and G) and Bland–Altman plots (F and H) obtained using the Siemens and Stago analyser/reagent pairs, respectively. Results are plotted as the difference (y axis) in anti-factor Xa activity or aPTT ratio between the analysis after 6 h (T6P) or 4 h (T4P) and 1 h (T1P) of storage at room temperature. *P < 0.05, **P < 0.01, ***P < 0.001, NS: Not significant (ANOVA or non-parametric Friedman test)

The aPTT ratio, assessed with the Siemens reagent/analyser pair (n = 32), increased from 1.53 (0.87–2.82) at T1P to 1.83 (1.05–3.86) at T4P (p < 0.01). The bias between T1P and T4P was quite high: 0.29 (Fig. 2E,F). Using the Stago reagent/analyser combination, the aPTT ratio (n = 27) remained stable over time: 2.69 (1.19–5.27) at T1P and 3.24 (1.34–5.32) at T6P (p = 0.47). The bias between T1P and T6P was 0.08 (Fig. 2G,H). The aPTT ratio agreement was moderate for the Siemens (Ks = 0.51, between T1P and T4P) and good for the Stago reagent/analyser combination (Ks = 0.70, between T1P and T6P).

Stability of samples from patients treated with LMWH

In sample stored as WB, anti-factor Xa activities measured using the Siemens analyser/reagent pair (n = 16) remained unchanged: 0.23 (0.11–0.59) at T1WB and 0.24 (0.11–0.58) at T6WB (p > 0.99). The bias was 0.00 IU/mL (Fig. 3A,B). Anti-factor Xa activities measured using the Stago analyser/reagent pair (n = 15) also were stable: 0.21 (0.10–0.54) at T1WB and 0.21 (0.11–0.55) at T6WB (p = 0.82). The bias was 0.03 IU/mL (Fig. 3C,D).

Fig. 3

Stability of citrated whole blood samples from patients treated with LMWH. Anti-factor Xa activity scatter plots (A and C) and Bland–Altman plots (B and D) obtained with the Siemens and Stago analyser/reagent pairs, respectively. Results are plotted as the difference (y axis) in anti-factor Xa activity between the analysis after 6 h (T6WB), 4 h (T4WB) and 1 h (T1WB) of storage at room temperature. *P < 0.05, **P < 0.01, ***P < 0.001, NS: Not significant (ANOVA or non-parametric Friedman test)

In samples stored as plasma, anti-factor Xa activities measured using the Siemens analyser/reagent pair (n = 16) remained stable between T1P and T4P [0.23 (0.11–0.59) vs 0.25 (0.11–0.59); p = 0.23] and increased between T1P and T6P [0.23 (0.11–0.59) vs 0.25 (0.12–0.58); p < 0.05]. The bias between T1P and T6P was limited to 0.01 IU/mL (Fig. 4A,B). With the Stago analyser/reagent pair (n = 15), anti-factor Xa activities were stable between T1P and T4P [0.21 (0.10–0.54) vs 0.21 (0.10–0.56); p = 0.43] and increased between T1P and T6P [0.21 (0.10–0.54) vs 0.22 (0.13–0.56); p < 0.05]. The bias between T1P and T6P was limited to 0.01 IU/mL (Fig. 4C,D).

Fig. 4

Stability of citrated plasma samples from patients treated with LMWH. Anti-factor Xa activity scatter plots (A and C) and Bland–Altman plots (B and D) analysed with the Siemens and Stago analyser/reagent pairs, respectively. Results are plotted as the difference (y axis) in anti-factor Xa activity between the analysis after 6 h (T6P), 4 h (T4P) and 1 h (T1P) of storage at room temperature. *P < 0.05, **P < 0.01, ***P < 0.001, NS: Not significant (ANOVA or non-parametric Friedman test)

Stability of blood samples collected in CTAD tubesStability of WB samples from patients treated with UFH

Anti-factor Xa activities (IU/mL), measured with the Siemens reagent/analyser combination (n = 25), remained stable: 0.43 (0.13–0.80) at T1WB and 0.41 (0.11–0.70) at T6WB (p = 0.25). The bias between T1WB and T6WB was -0.01 IU/mL (Fig. 5A,B). With the Stago reagent/analyser combination (n = 25), anti-factor Xa activities also remained unchanged: 0.31 (0.17–1.09) at T1WB and 0.29 (0.15–1.14) at T6WB (p = 0.20). The bias between T1WB and T6WB was -0.01 IU/mL (Fig. 5C,D). The agreement between T1WB and T6WB was good for the Siemens (Ks = 0.69) and excellent for the Stago (Ks = 0.83) reagent/analyser pair.

Fig. 5

Stability of CTAD whole blood samples from patients treated with UFH. Anti-factor Xa activity scatter plots (A and C), Bland–Altman plots (B and D), and aPTT ratio scatter plots (E and G), Bland–Altman plots (F and H) after analysis with the Siemens and Stago reagent/analyser combination, respectively. Results are plotted as the difference (y axis) in anti-factor Xa activity or aPTT ratio between the analysis after 6 h (T6WB), 4 h (T4WB) and 1 h (T1WB) of storage at room temperature. *P < 0.05, **P < 0.01, ***P < 0.001, NS: Not significant (ANOVA or non-parametric Friedman test)

The aPTT ratio, measured with the Siemens reagent/analyser combination (n = 24), remained stable: 1.82 (0.78–5.23) at T1WB and 1.85 (0.81–4.98) at T6WB (p > 0.99). The bias between T1WB and T6WB was -0.05 (Fig. 5E,F). With the Stago reagent/analyser combination (n = 23), the aPTT ratio increased from 2.06 (1.22–3.45) at T1WB to 2.22 (1.29–3.58) at T6WB (p < 0.01). The bias between T1WB and T6WB was 0.13 (Fig. 5G,H). The aPTT agreement between T1WB and T6WB was excellent for the Siemens (Ks = 1) and good for the Stago (Ks = 0.63) reagent/analyser combination.

Stability of plasma samples from patients treated with UFH

Anti-factor Xa activities measured with the Siemens reagent/analyser combination (n = 25) increased from 0.43 (0.13–0.80) at T1P to 0.43 (0.13–0.78) at T4P (p < 0.01) and to 0.44 (0.14–0.80) at T6P (p < 0.001). However, the bias between T1P and T6P was limited to 0.03 IU/mL (Fig. 6A,B). Using the Stago reagent/analyser combination (n = 25), anti-factor Xa activities remained stable: 0.31 (0.17–1.09) at T1P and 0.29 (0.17–1.11) at T6P (p > 0.99). The bias between T1P and T6P was 0.01 IU/mL (Fig. 6C,D). The agreement between T1P and T6P was excellent for the Siemens (Ks = 0.80) and good for the Stago (Ks = 0.75) reagent/analyser combination.

Fig. 6

Stability of CTAD plasma samples from patients treated with UFH. Anti-factor Xa activity scatter plots (A and C) and Bland–Altman plots (B and D), and aPTT ratio scatter plots (E and G) and Bland–Altman plots (F and H) obtained using the Siemens and Stago analyser/reagent pair, respectively. Results are plotted as the difference (y axis) in anti-factor Xa activity or aPTT ratio between the analysis after 6 h (T6P) or 4 h (T4P) and 1 h (T1P) of storage at room temperature. *P < 0.05, **P < 0.01, ***P < 0.001, NS: Not significant (ANOVA or non-parametric Friedman test)

The aPTT ratio, assessed with the Siemens reagent/analyser pair (n = 21), increased from 1.50 (0.78–3.71) at T1P to 2.21 (1.12–5.24) at T4P, but the difference was not statically significant (p = 0.09). The bias between T1P and T4P was quite high: 0.32 (Fig. 6E,F). Using the Stago reagent/analyser combination, the aPTT ratio (n = 23) increased from 2.06 (1.22–3.45) at T1P to 2.23 (1.25–3.90) at T6P (p < 0.05). The bias between T1P and T6P was 0.17 (Fig. 6G,H). The aPTT ratio agreement between time points was moderate for both the Siemens (Ks = 0.57 between T1P and T4P) and the Stago (Ks = 0.53 between T1P and T6P) reagent/analyser combinations.

Stability of samples from patients treated with LMWH

In samples stored as WB, anti-factor Xa activities measured using the Siemens analyser/reagent pair (n = 18) were stable between T1P and T4P [0.24 (0.11–1.22) vs 0.26 (0.11–1.24); p = 0.20] and increased between T1P and T6P [0.24 (0.11–1.22) vs 0.27 (0.11–1.25); p < 0.05]. However, the bias between T1P and T6P was limited to 0.02 IU/mL (Fig. 7A,B). With the Stago analyser/reagent pair (n = 26), anti-factor Xa activities remained stable between T1P and T4P [0.23 (0.15–0.80) vs 0.22 (0.14–0.82); p > 0.99] and between T1P and T6P [0.22 (0.11–0.79); p = 0.18]. The bias between T1P and T6P was -0.01 IU/mL (Fig. 7C,D).

Fig. 7

Stability of CTAD whole blood samples from patients treated by LMWH. Anti-factor Xa activity scatter plots (A and C) and Bland–Altman plots (B and D) obtained with the Siemens and Stago analyser/reagent pairs, respectively. Results are plotted as the difference (y axis) in anti-factor Xa activity between the analysis after 6 h (T6WB), 4 h (T4WB) and 1 h (T1WB) of storage at room temperature. *P < 0.05, **P < 0.01, ***P < 0.001, NS: Not significant (ANOVA or non-parametric Friedman test)

In samples stored as plasma, anti-factor Xa activities measured using the Siemens analyser/reagent pair (n = 18) increased between T1P and T4P [0.24 (0.11–1.22) vs 0.25 (0.11–1.23); p < 0.05] and between T1P and T6P [0.24 (0.11–1.22) vs 0.28 (0.12–1.24); p < 0.001]. The bias between T1P and T6P was 0.03 IU/mL (Fig. 8A,B). With the Stago analyser/reagent pair (n = 26), anti-factor Xa activities were stable between T1P and T4P [0.23 (0.15–0.80) vs 0.24 (0.14–0.85); p > 0.99] and between T1P and T6P [0.22 (0.11–0.81); p = 0.64]. The bias between T1P and T6P was 0.01 IU/mL (Fig. 8C,D).

Fig. 8

Stability of CTAD plasma samples from patients treated with LMWH. Anti-factor Xa activity scatter plots (A and C) and Bland–Altman plots (B and D) analysed with the Siemens and Stago analyser/reagent pairs, respectively. Results are plotted as the difference (y axis) in anti-factor Xa activity between the analysis after 6 h (T6P), 4 h (T4P)and 1 h (T1P) of storage at room temperature. *P < 0.05, **P < 0.01, ***P < 0.001, NS: Not significant (ANOVA or non-parametric Friedman test)