Main Article Content

Incidence and predictive factors of lower extremity deep‑vein thrombosis in patients with neurological diseases in a sub‑Saharan tertiary hospital


Francis Chukwuebuka Campbell
Enyereibe Chuks Ajare
Kelechi Onyenekeya Ndukuba
Okwuoma Okwunodulu
Idawarifagha Hart
Sunday Samuel Nnama
Terngu Titus Azaatse
Chika Anele Ndubuisi
Samuel Ohaegbulam

Abstract

Background: Deep‑vein thrombosis (DVT) is a major cause of morbidity among patients with neurological disease and is implicated in 50%–90% of pulmonary embolisms (PE). Aim: This study aims to identify the incidence and predictive factors of lower extremity (LE) DVT in high‑risk neurological patients. Materials and Methods: All patients with neurological diseases who were confined to bed for three days or more, or had elevated D‑dimer over a 22 months were included in the study. They all had routine mechanical and chemical prophylaxis. Serial color duplex ultrasonography of the LE was performed for all the patients, and the results were used to stratify them into two groups: DVT positive and DVT negative, and their clinical parameters were recorded and utilised for statistical analysis. Multivariate logistic regression analysis was used to determine the predictive risk factors. Results: A total of 611 patients were admitted, and 107 met the inclusion criteria. The mean age was 57 ± 17 years, with a male‑to‑female ratio of 1.7:1. Forty‑four (41.4%) patients had DVT (overall incidence was 7.2%). Primary intracranial tumours and craniocerebral trauma accounted for 27.2% and 20.5% of aetiology, respectively. The comparative DVT‑positive versus DVT‑negative group D‑dimer levels were 7.9 versus 4.9 mg/L (P = 0.0065), duration of immobility 65.0 versus 35.3 days (P = 0.0001), and length of hospital (LOH) stay 32.5 versus 18.2 days (P = 0.0001). The proximal LE vessels were involved in 64.2%. The PE risk was 0.98%, with three mortalities. The predictive factors on multivariate logistic regression were elevated D‑dimers (odds ratio [OR] 2.02, 95% confidence interval [CI] 1.1–4.45 P = 0.030) and preoperative immobility >seven days ([OR] 2.38, 95% [CL] 1.28–4.39 P = 0.042). Gender, C‑reactive protein, Glasgow Coma Scale, comorbidities, LOH, and chronic steroid use were not predictive. Conclusion: The overall incidence of DVT and mortality from PE were low among patients with neurological diseases. Most thrombi involved proximal LE veins. Elevated D‑dimers and preoperative immobility >seven days were predictive factors for LE DVT.


Journal Identifiers


eISSN: 2667-0526
print ISSN: 1115-2613