Systematic review: diagnostic accuracy of clinical decision rules for venous thromboembolism in elderly.

BACKGROUND

Physicians committed to the care of elderly patients, are challenged with the diagnosis of venous thromboembolism (VTE: deep venous thrombosis and pulmonary embolism) due to a higher incidence, co-morbidities masking signs and symptoms and burdening referrals. Clinical decision rules (CDRs) have been developed and implemented for VTE. Yet, until now, no study has evaluated the existing evidence of the diagnostic accuracy of CDRs for VTE in elderly.

PURPOSE

To assess the effect of increasing age on diagnostic accuracy of CDRs for VTE in elderly.

DATA SOURCES

A computerized systematic search was performed in Medline and Embase from 1950 to 2010. After checking reference lists and field experts, all key journals were hand searched.

STUDY SELECTION

After review of 1538 eligible citations, nine articles were included and critically appraised on methodological quality by two reviewers using the QUADAS criteria.

DATA EXTRACTION

Data on age subgroups, type of CDRs, sensitivity, specificity, safety, efficiency and the prevalence of deep venous thrombosis (DVT) and pulmonary embolism (PE) were extracted.

DATA SYNTHESIS

Although sensitivity and safety of the CDRs for VTE in elderly remained high, the specificity and efficiency decreased substantially in older age groups.

LIMITATIONS

A limited number of studies met our inclusion criteria. Possible referral bias due to inclusion of relatively high risk elderly patients.

CONCLUSIONS

This diagnostic review demonstrates an increase of prevalence of PE with age and a strong decrease of specificity and efficiency for CDRs of VTE in older patients. Moreover, due to referral bias the decrease in specificity in the elderly may even be underestimated. Although the safety of CDRs for VTE is high, adapting these rules for elderly is much needed to make them more efficient for aged patients.