Haloperidol Does Not Activate Thrombogenic Factors in Older, Nonpsychotic Hospitalized Patients

    loading  Checking for direct PDF access through Ovid


To the Editors
Worldwide, haloperidol is the antipsychotic drug used most often to treat delirium in (older) patients admitted to hospital.1 In 2005, the Food and Drug Administration issued a warning about the increased mortality rate in elderly users of antipsychotics.2 Although the pathophysiological mechanism is unknown, the increased death rate might be due to vascular problems. For example, users of antipsychotic drugs are at increased risk of cerebrovascular accidents3 and venous thromboembolism.4,5 Potential mechanisms for the association between antipsychotics and cerebrovascular events include thromboembolic effects, altered platelet function, cardiovascular effects, and the atherosclerotic effects of metabolic dysregulation.3 Markers of thrombogenesis are activated in untreated patients with an acute psychosis,6 which suggests that the psychosis itself may be responsible for the increased risk of morbidity and mortality. However, there is no clear explanation for the increased mortality seen in nonpsychotic elderly patients with dementia.
The main aim of this study was to investigate whether factors of thrombogenesis are activated in older, nonpsychotic hospitalized patients treated with haloperidol.
Patients were a subset of patients included in a randomized, stratified, double-blind, placebo-controlled trial (“Haloperidol Prophylaxis in Older Emergency Department Patients,” HARPOON study).7 This subset consisted of all patients recruited at the Jeroen Bosch Hospital, a teaching hospital in the Netherlands, between June 2014 and March 2015.
In this randomized controlled trial, nonpsychotic elderly patients received halo-peridol (1 mg) or placebo twice daily. Patients were included if they presented to the emergency department, were 70 years or older, had an increased risk of developing delirium, but did not have delirium at presentation, and were admitted to departments of internal medicine or surgery. The primary outcome was the effect of haloperidol versus placebo on coagulation, as evaluated by measuring fibrinogen and D-dimer as markers of thrombogenesis, P-selectin as marker of platelet activation, and von Willebrand factor and osteoprotegerin as markers of endothelial activation. Venous blood was collected on day 1 (baseline) and day 6. All samples were analyzed at the same time.
The effects of haloperidol versus placebo and time (day 1 vs day 6) on coagulation were analyzed using repeated-measures analysis of variance (ANOVA), because this takes potential nonsignificant baseline differences into account and can distinguish between “within groups” (day 1 vs day 6) and “between groups” (haloperidol vs placebo) effects. Covariate adjustment was not considered in the analysis because of the stratified randomization method used in this study. The Statistical Package for the Social Sciences (IBM SPSS 22, Armonk, NY) was used.
This study was approved by the institution’s ethics committee, and all participants provided written informed consent.
In the Jeroen Bosch Hospital, the Netherlands, 52 patients were randomized. Of these, 3 withdrew consent, 3 stopped per protocol treatment, and 12 had no blood sample drawn on day 6. The data of 16 haloperidol patients and 18 placebo patients were analyzed. None of these patients developed delirium. Potentially relevant comedication was equally divided over the 2 groups. In the haloperidol group, 5 patients (31.2%) used aspirin/dipyridamole/clopidogrel versus 8 patients (44.4%) in the placebo group. No patients used nonsteroidal anti-inflammatory drugs. No patients in the haloperidol group used a selective serotonin reuptake inhibitor versus 2 patients (11.1%) in the placebo group. In the haloperidol group, 0 patients used low-molecular-weight heparins, and 4 patients (25%) used a vitamin K antagonist or non–vitamin K oral anticoagulant versus 1 patient (5.6%) with low-molecular-weight heparin and 6 patients (33.3%) with vitamin K antagonists or non–vitamin K oral anticoagulant in the placebo group.
Table 1 shows the main results. There were no significant changes in levels of thrombogenic factors between the haloperidol and the placebo group. Thrombogenic factors were not influenced by haloperidol.
    loading  Loading Related Articles