Performance of Early Prognostic Assessment Independently Predicts the Outcomes in Patients with Acute Pulmonary Embolism

Despite advances in hospital management in recent years, pulmonary embolism (PE) is the most common cause of vascular death after myocardial infarction and stroke. Guidelines recommend risk stratification of patients with acute symptomatic PE.[1] However, it remains to be determined whether using a prognostication model per se—irrespective of the model results—is predictive of outcomes. Using our prospective single-centre 2003 to 2016 PE registry, we aimed to assess whether clinicians routinely prognosticate patients with acute PE, and whether prognostic assessment was associated with clinical outcomes.

We defined prognostic assessment (yes/no) as documented calculation of a validated clinical prognostic score (i.e. PE severity index [PESI],[2] simplified PESI[3] and Hestia criteria[4]), and/or assessment of right ventricular dysfunction (e.g. echocardiography, spiral computed tomography or brain natriuretic peptide testing), and/or assessment of myocardial injury (e.g. cardiac troponin T or I testing) by the primary practitioners. The primary outcome of the study was all-cause mortality during the first month after the diagnosis of acute PE. To estimate the outcomes of time to death, Kaplan–Meier's probabilities were computed, and differences between the groups were assessed with the log-rank test. We used logistic regression to assess if the variable of prognostic assessment had an independent association with 30-day all-cause mortality, after adjusting for other variables that had a p-value of < 0.05 in the univariate analyses.

The study cohort consisted of 2,096 patients (52% women). Prognostic assessment was performed in 1,754 patients (84%). As shown in [Table 1], patients who did not receive prognostic assessment had a higher prevalence of cancer, anaemia and concomitant deep vein thrombosis compared with those who received prognostic assessment. Patients having prognostic assessment were more frequently haemodynamically compromised, with lower systolic blood pressure, higher heart rate and proportionally more with hypoxaemia and syncope.

Table 1 Baseline characteristics and clinical events in PE patients with and without prognostic assessment

	All patients N = 2,096	Prognostic assessment performed N = 1,754	Prognostic assessment not performed N = 342	p-Value
Clinical characteristics
Age, y (mean ± SD)	68.7 ± 16.6	69.0 ± 16.5	67.0 ± 17.4	0.04
Age > 65 y	1,394 (67%)	1,183 (67%)	211 (62%)	0.04
Male sex	1,004 (48%)	838 (48%)	166 (48%)	0.81
Body mass index, kg m−2 (mean ± SD)	27.2 ± 6.4	27.4 ± 6.6	26.2 ± 4.7	<0.01
Risk factors for VTE
History of VTE	249 (12%)	218 (12%)	31 (9.1%)	0.08
Cancer[a]	432 (21%)	293 (17%)	139 (41%)	< 0.001
Recent surgery[b]	188 (9.0%)	159 (9.1%)	29 (8.5%)	0.84
Immobilization[c]	412 (20%)	342 (19%)	70 (20%)	0.71
Comorbid diseases
Recent major bleeding[b]	70 (3.3%)	57 (3.2%)	13 (3.8%)	0.62
COPD	169 (8.1%)	142 (8.1%)	27 (7.9%)	1.0
Congestive heart failure	116 (5.5%)	103 (5.9%)	13 (3.8%)	0.15
Concomitant deep vein thrombosis (n = 1,630)	916 (56%)	734/1,343 (55%)	182/287 (63%)	< 0.01
Clinical symptoms and signs at presentation
Syncope	300 (14%)	275 (16%)	25 (7.3%)	< 0.001
Heart rate ≥ 110/min	426 (20%)	377 (21%)	49 (14%)	< 0.01
Arterial oxyhaemoglobin saturation < 90%	563 (27%)	491 (28%)	72 (21%)	< 0.01
SBP < 90 mm Hg	75 (3.6%)	70 (4.0%)	5 (1.5%)	< 0.01
Low-risk sPESI[d]	638 (30%)	468 (27%)	170 (50%)	< 0.001
Laboratory findings
Haemoglobin, g/dL (mean ± SD)	13.2 ± 2.0	13.4 ± 2.0	12.6 ± 2.0	< 0.001
Creatinine > 2 mg/dL	78 (3.7%)	67 (3.8%)	11 (3.2%)	0.75
Prognostic assessment
sPESI	811 (39%)	811 (39%)	0 (0%)	–
Hestia criteria	0 (0%)	0 (0%)	0 (0%)	–
Echocardiography and/or CT	1,147 (55%)	1,147 (55%)	0 (0%)	–
BNP	805 (38%)	805 (38%)	–	–
cTnI	1,564 (75%)	1,564 (75%)	–	–
Treatment
Thrombolysis	92 (4.4%)	91 (5.2%)	1 (0.3%)	< 0.001
Inferior vena cava filter	49 (2.3%)	42 (2.4%)	7 (2.0%)	0.70
30-d event rates
All-cause mortality	140 (6.7%)	103 (5.9%)	37 (10.8%)	0.01
PE-related mortality	62 (2.9%)	50 (2.8%)	12 (3.5%)	0.49
Cancer-related mortality	31 (1.5%)	22 (1.3%)	9 (2.6%)	0.08
Recurrent VTE	38 (1.8%)	29 (1.6%)	9 (2.6%)	0.26
Major bleeding	79 (3.8%)	67 (3.8%)	12 (3.5%)	0.88

Abbreviations: BNP, brain natriuretic peptide; COPD, chronic obstructive pulmonary disease; CT, computed tomography; cTnI, cardiac troponin I; PE, pulmonary embolism; SBP, systolic blood pressure; SD, standard deviation; sPESI, simplified pulmonary embolism severity index; VTE, venous thromboembolism.

^a Active or under treatment in the past year.

^b In the previous month.

^c Immobilized patients are defined in this analysis as nonsurgical patients who had been immobilized (i.e. total bed rest with bathroom privileges) for ≥4 days in the month prior to PE diagnosis.

^d Calculated retrospectively.

Overall, 140 out of 2,096 patients died (6.7%) during the first month of follow-up. One hundred and three deaths (5.9%) occurred in the group of patients with prognostic assessment, whereas 37 deaths (10.8%) occurred in the group of patients who did not undergo prognostic assessment (p < 0.01). Patients with acute PE who did not undergo prognostic assessment had a significantly higher cumulative mortality than the patients with acute PE who underwent prognostic assessment (p < 0.01, log-rank test). In univariate analyses, those without prognostic assessment at the time of acute PE diagnosis were significantly more likely to die during follow-up (odds ratio [OR], 1.95; 95% confidence interval [CI], 1.31–2.89). After adjustment, lack of prognostic assessment was independently significantly associated with all-cause mortality (adjusted OR, 1.70; 95% CI, 1.11–2.62; p = 0.01). Among patients without cancer, using the same multivariable model, lack of prognostication was associated with significantly increased 30-day all-cause mortality (adjusted hazard ratio, 1.96; 95% CI, 1.31–2.89; p < 0.01).

This study showed that lack of prognostic assessment for patients with acute symptomatic PE was not infrequent, and was correlated with patient outcomes in adjusted analyses. The risk of death among patients who did not have prognostic assessment performed was approximately two times higher than in patients who received prognostic assessment. The European Society of Cardiology guidelines recommend risk stratification of patients with PE, as it has important implications both for the diagnostic and therapeutic management.[1] We can hypothesize different mechanisms for such association between prognostic assessment and outcomes. First, early performance of prognostic assessment might have enabled the providers to better determine the level of care (e.g. intensive care vs. step-down or regular floor) and associated ancillary therapies. Second, clinicians might have used the findings from prognostic assessment models to guide PE-specific therapies for their patients. Third, use of a prognostication model might be a marker of quality of care, with physicians performing prognostication also providing better care related to PE and/or other processes of care, with a favorable impact on patient outcomes.

In conclusion, our findings suggest that use of a prognostication model for patients with acute PE is associated with short-term mortality, irrespective of the prognostication model's results. Our data in the current format should be interpreted as hypothesis generating. Further, we may have underestimated the performance of prognostic assessment, as it may have happened for some patients but not documented. Further research is required to elucidate the nuances of the association between risk assessment and outcome of patients with PE.

• References

• 1 Konstantinides SV, Torbicki A, Agnelli G. , et al; Task Force for the Diagnosis and Management of Acute Pulmonary Embolism of the European Society of Cardiology (ESC). 2014 ESC guidelines on the diagnosis and management of acute pulmonary embolism. Eur Heart J 2014; 35 (43) 3033-3069 , 3069a–3069k
• 2 Aujesky D, Obrosky DS, Stone RA. , et al. Derivation and validation of a prognostic model for pulmonary embolism. Am J Respir Crit Care Med 2005; 172 (08) 1041-1046
• 3 Jiménez D, Aujesky D, Moores L. , et al; RIETE Investigators. Simplification of the pulmonary embolism severity index for prognostication in patients with acute symptomatic pulmonary embolism. Arch Intern Med 2010; 170 (15) 1383-1389
• 4 Zondag W, Mos IC, Creemers-Schild D. , et al; Hestia Study Investigators. Outpatient treatment in patients with acute pulmonary embolism: the Hestia Study. J Thromb Haemost 2011; 9 (08) 1500-1507