External validation of the In-hospital Mortality for PulmonAry embolism using Claims daTa (IMPACT) multivariable...

Area of Science:

• Health Services Research
• Clinical Informatics
• Epidemiology

Background:

• Pulmonary embolism (PE) poses a significant mortality risk.
• Accurate prediction of in-hospital mortality is crucial for effective patient management and resource allocation.
• Existing clinical prediction rules require validation using large-scale administrative datasets.

Purpose of the Study:

• To validate the In-hospital Mortality for PulmonAry embolism using Claims daTa (IMPACT) multivariable prediction rule.
• To assess the rule's performance in predicting in-hospital mortality for pulmonary embolism (PE) patients using admission claims data.
• To evaluate the IMPACT rule's utility for payers and hospitals in determining optimal hospitalization duration.

Main Methods:

• Retrospective analysis of Humana admission claims data from January 2007 to March 2014.
• Inclusion of adult patients admitted for their first PE, identified using ICD-9 codes (415.1x).
• Application of the IMPACT rule (age + 11 comorbidities) to estimate mortality probability and classify risk (≤1.5% in-hospital mortality as low-risk).

Main Results:

• The study included 23,858 PE patients, with an in-hospital mortality rate of 3.3%.
• The IMPACT rule demonstrated high sensitivity (97.6%) and negative predictive value (99.2%) for in-hospital mortality.
• The rule showed a c-statistic of 0.70, indicating good discriminatory ability, and was also evaluated for patients under 65.

Conclusions:

• The IMPACT prediction rule is validated for use with claims data, particularly Medicare claims.
• Further external validation and comparison with existing rules are recommended.
• IMPACT shows potential as a valuable tool for healthcare payers and hospitals to assess PE patient management and hospitalization length.