Prevalence of Deep Vein Thrombosis and Associated Factors in Adult Medical Patients Admitted to the University Teaching Hospital, Lusaka, Zambia

  • C K Mwandama Department of Internal Medicine, University Teaching Hospital, Lusaka, Zambia
  • B Andrews School of Medicine, University of Zambia, Vanderbilt Institute for Global Health (Nashville, TN, USA), and Christ Community Health Services (Memphis, TN, USA)
  • S Lakhi Department of Internal Medicine, University Teaching Hospital, Lusaka, Zambia



Deep vein thrombosis (DVT) and pulmonary embolism (PE) collectively referred to as venous thromboembolism (VTE) are associated with significant morbidity and mortality worldwide.

DVT is common in hospitalized patients with acute medical illness. Routine use of thromboprophylaxis has been shown to reduce prevalence of DVT in hospitalized patients with acute medical illness. Thromboprophylaxis is not routinely given to hospitalized medical patients in most hospitals in Zambia.



The objectives of this study were: to determine the prevalence and anatomical distribution of DVT in medical patients admitted to the UTH for at least 7 days; establish demographic and clinical characteristics of medical patients with DVT and finally determine the accuracy of the Well’s score for DVT.


This was a descriptive, cross sectional analytical study. The sample size comprised 296 medical patients admitted for at least 7 days. A questionnaire was used to obtain demographic characteristics and relevant clinical history. A focused detailed physical examination was conducted to screen for DVT of the lower limbs and Well’s score for DVT computed. Laboratory tests including HIV test and full blood count were done. Biochemical and genetic studies to evaluate for inherited and acquired thrombophilias could not be done due limited financial resources. Compression ultrasound scans (USS) were done on lower limbs of recruited patients to determine the presence of DVT.                                                                                   Variables of interest were compared by chi-square, Kruskal-wallis and t-tests. Multivariate and univariate logistic regression analysis were used to assess for associations between DVT and independent variables of interest.


The Prevalence of DVT of the lower limbs was 11.1 % (33/296). Prevalence of proximal lower limb DVT was 9.1%. Eighty two percent (27/33) of all patients with DVT had proximal lower limb DVT. Asymptomatic lower limb DVT was noted in 85 % (28/33) of all patients with DVT.

The Mean age of patients with DVT was 42.12 years (SD 12.71).The mean duration of hospital admission was 11.91 (SD 7.77) days. Seventy six percent (25/33) of all patients with DVT had an infectious disease as a primary diagnosis. Tuberculosis was the most common infectious disease among patients with DVT accounting for 60.6 % (20/33) of all infections. Up to 70 % (23/33) of all patients with DVT were HIV positive.

The specificity for the Wells score specificity for DVT was 73.4% while the sensitivity was 100%.  The accuracy was 76.3%.The positive and negative predictive values were 32% and 100% respectively.


Proximal Lower limb DVT is common among HIV positive medical   patients admitted for at least 7 days at the UTH. In our study, lower extremity proximal DVT was more common in patients with Tuberculosis and a low BMI. Up to 85% of lower limb DVT was asymptomatic. Without a high index of suspicion, lower limb DVT is likely to be missed. The pretest Wells score correlated well with the USS findings and could be used as a rule out test for those with suspected DVT.A follow up study to evaluate for genetic and biochemical factors that predispose to DVT need to be undertaken in the near future. There is need to advocate for thromboprophylaxis in medical patients with acute illness and prolonged hospital stay.


Download data is not yet available.