Optimal dosing for thromboprophylaxis in medical inpatients

Clinical Question

What is the optimal dosing regimen for thromboprophylaxis in hospitalized medical patients?

Bottom Line

Until a direct comparison study is performed, the best information available suggests that although 3-times-daily dosing of 5000 units unfractionated heparin (UH) is more effective then twice-daily dosing (approximately 1 fewer pulmonary embolism (PE) and 2 fewer deep vein thromboses (DVTs) per 1000 patient days), it is associated with more major bleeds (1 per 2500 patient days). Remember that both regimens are better than doing nothing for high-risk hospitalized medical patients. (LOE = 1a-)


King CS, Holley AB, Jackson JL, Shorr AF, Moores LK. Twice vs three times daily heparin dosing for thromboembolism prophylaxis in the general medical population: a meta-analysis. Chest 2007;131:507-516.  [PMID:17296655]

Study Design

Meta-analysis (randomized controlled trials)


Unknown/not stated


Inpatient (any location)


Guidelines now recommend thromboprophylaxis with low-dose UH or low-molecular-weight heparin (LMWH) in acutely ill medical inpatients, especially those with heart failure, respiratory disease, who are confined to bed, or who have a previous history of DVT or PE. Some studies have used 5000 units UH subcutaneously given 2 times per day, while others have given the same dose 3 times daily; the 2 different dosing frequencies have never been directly compared. The authors of this meta-analysis carefully searched the literature for all studies of UH prophylaxis with adequate verification of DVT or PE and performed in a nonsurgical population. They identified a total of 447 articles, of which 435 were excluded, mostly because they had studied a surgical or postoperative population. The mean ages of the enrolled patients in the remaining 12 studies ranged from 58 years to 75 years. Nine of the studies had between 38 patients and 223 patients; the remaining 3 studies had 482, 726 and 5776 patients, respectively. Most of the patients in these studies were at moderate- to high-risk for DVT or PE. There were a total of 1664 patients receiving 3-times-daily dosing and 6314 receiving twice-daily dosing in the 12 studies. The authors found no significant difference between the groups regarding the rate of DVT (5.4 for twice daily vs 3.0 for 3 times daily per 1000 patient-days; P = .42), but a trend toward fewer PEs in the 3-times-daily dosing group (1.5 vs 0.5 per 1000 patient-days; P = .09). Bleeding complications were more common in the 3 times per day group (0.73 vs 0.33 per 1000 patient-days; P < .001; number needed to treat to harm = 250). The largest study, accounting for more than 90% of patients receiving twice-daily dosing, had no placebo group, did not clearly describe randomization, was not blinded, and used autopsy to confirm the diagnoses of DVT and PE. It therefore reported much lower rates of these outcomes than the other studies. When this study is excluded, the difference between twice daily and 3 times daily groups regarding the number of venous thromboembolic events increased and became statistically significant, but the differences in bleeding rates lost statistical significance.