Introduction
Materials and methods
Study objectives
Study endpoints
Search strategy
Study eligibility
Data extraction
Data synthesis and statistical analysis
Results
Description of studies
Reference | Study design | Patients, number | Hemodynamic instability (number) | Timing of BNP sampling | Thrombolysis, number | Age, years | Male, percentage | Follow-up | CHF, number |
---|---|---|---|---|---|---|---|---|---|
Kucher, et al. [1] (2003) | Prosp | 73 | Yes | Admission (<4 hours) | 10 | 61 ± 18 | 59 | In hosp | NA |
Ten Wolde, et al. [2] (2003) | Prosp | 110 | Excl | Admission | NA | 58 ± 18 | NK | 3 months | NA |
Pieralli, et al. [3] (2006) | Prosp | 61 | Excl | Admission (<1 hour) | 6 | 75 ± 14 | 26 | In hosp | Excl |
Krüger, et al. [4] (2004) | Prosp | 42 | Yes (10) | Admission | 22 | 57 ± 17 | 64 | In hosp | 8 Excl |
Tulevski, et al. [5] (2006) | Prosp | 28 | Excl | Admission (<1 hour) | NA | 53 ± 18 | 43 | 90 days | Excl |
Logeart, et al. [6] (2007) | Prosp | 67 | Excl | Admission | 6 | 64 ± 16 | 60 | In hosp | Excl |
Ray, et al. [7] (2006) | Prosp | 51 | NA | Admission | 0 | 79 ± 10 | NA | In hosp | NA |
Pruszczyk, et al. [8] (2003) | Prosp | 79 | Yes (9) | Admission | 8 | 63 ± 17 | 37 | In hosp | NA |
Kucher, et al. [9] (2003) | Prosp | 73 | Yes (14) | Admission (<4 hours) | 10 | 61 ± 18 | 59 | In hosp | 6 |
Kostrubiec, et al. [10] (2005) | Prosp | 100 | Excl | Admission | 5 | 62 ± 18 | 35 | 40 days | 17 |
Binder, et al. [11] (2005) | Prosp | 124 | Yes (9) | Admissionand at 4, 8, and 24 hours | 12 | 60 ± 18 | 40 | In hosp | NA |
Maziere, et al. [12] (2007) | Prosp | 60 | Excl | Admission | NA | 73 ± 14 | 40 | In hosp | 20 |
Brain natriuretic peptides assays
Reference | BNP/NT-proBNP | Assay | Manufacturer | Kind of assay | Cutoff, pg/mL | Elevated BNP, percentage |
---|---|---|---|---|---|---|
Kucher, et al. [1] | BNP | Fluorescence immunoassay | Biosite (San Diego, USA) | Quantitative | 90 | 43.8 |
Ten Wolde, et al. [2] | BNP | Immunoradiometric assay | Shionoria (Osaka, Japan) | Quantitative | 21.7 | 33 |
Pieralli, et al. [3] | BNP | Fluorescence immunoassay | Biosite | Quantitative | 527 | 67 |
Krüger, et al. [4] | BNP | Immunofluorometric assay | Biosite | Quantitative | 90 | 40 |
Tulevski, et al. [5] | BNP | Immunoradiometric assay | Shionoria | Quantitative | 10 | 50 |
Logeart, et al. [6] | BNP | Fluorescence immunoassay | Biosite | Quantitative | 100 | 70 |
Ray, et al. [7] | BNP | Fluorescence immunoassay | BioMérieux (Marcy l'Etoile, France) | Quantitative | 200 | 43 |
Pruszczyk, et al. [8] | NT-proBNP | ECLIA | Roche (Basel, Switzerland) | Quantitative | NA | 83.5 |
Kucher, et al. [9] | NT-proBNP | ECLIA | Roche | Quantitative | 500 | 57 |
Kostrubiec, et al. [10] | NT-proBNP | ECLIA | Roche | Quantitative | 600 | 39 |
Binder, et al. [11] | NT-proBNP | ECLIA | Roche | Quantitative | 1,000 | 54 |
Maziere, et al. [12] | NT-proBNP | ECLIA | Roche | Quantitative | 1,000 | 43 |
Outcome measures
Death
Endpoints | OR (95% CI) | Sensitivity (95% CI) | Specificity (95% CI) | LR+ (95% CI) | LR- (95% CI) | PPV (95% CI) | NPV (95% CI) |
---|---|---|---|---|---|---|---|
Short-term death (12 studies, 868 patients) | 6.57 (3.11–13.91) | 0.93 (0.85–0.98) | 0.48 (0.44–0.51) | 1.64 (1.39–1.94) | 0.34 (0.19–0.61) | 0.14 (0.11–0.18) | 0.99 (0.97–1.00) |
Death resulting from PE (10 studies, 684 patients) | 6.10 (2.58–14.25) | 0.92 (0.81–0.98) | 0.52 (0.48–0.56) | 1.76 (1.33–2.34) | 0.37 (0.19–0.71) | 0.13 (0.10–0.17) | 0.99 (0.97–1.00) |
Serious adverse events (9 studies, 688 patients) | 7.47 (4.2–13.15) | 0.89 (0.83–0.93) | 0.48 (0.44–0.52) | 1.70 (1.44–2.01) | 0.28 (0.17–0.48) | 0.33 (0.29–0.38) | 0.94 (0.90–0.96) |
Cause-specific death resulting from pulmonary embolism
Serious adverse events
Discussion
Limitations
Conclusion
Key messages
-
Elevated brain natriuretic peptide (BNP) levels can help to identify patients with acute pulmonary embolism at high risk of short-term death and adverse outcome events.
-
Although elevated BNP levels have a high sensitivity to detect patients at risk of death, the specificity is low.
-
The positive predictive value of elevated BNP levels alone remains low and its high negative predictive value is more useful to identify individuals with a likely favorable outcome.