VALUE IN HEALTH 14 (2011) S89 –S92

available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/jval

Short-Term Therapy with Enoxaparin or Unfractionated Heparin for

Venous Thromboembolism in Hospitalized Patients: Utilization Study
and Cost-Minimization Analysis
Catia Argenta, MSc,1 Maria Angélica Pires Ferreira, MSc,2 Guilherme Becker Sander, PhD,2
Leila Beltrami Moreira, PhD1,2,3,*
1
Graduate Program in Medical Sciences, Universidade Federal do Rio Grande do Sul, Porto Alegre, Brazil; 2Pharmacy and Therapeutics Committee,
Hospital de Clínicas de Porto Alegre, Porto Alegre, Brazil; 3Pharmacology Department, Universidade Federal do Rio Grande do Sul; National Institute of
Science and Technology for Health Technology Assessment–CNPq, Porto Alegre, Rio Grande do Sul, Brazil

A B S T R A C T

Objectives: To evaluate the direct costs of venous thromboembolism observed between groups in the number of bleeding events (10.0% and
(VTE) treatment with unfractionated heparin (UFH) and low-molecular 9.4%; P ⫽ 1.00); blood transfusion (2.0% and 2.6%; P ⫽ 1.00); death (8.0%
weight heparin, from the institutional perspective. Methods: This is a and 3.4%; P ⫽ 0.24); and recurrent VTE, bleeding, or death (20.0% and
real-world cohort study that included inpatients treated with UFH or 14.5%; P ⫽ 0.38). Daily mean cost per patient was US$12.63 ⫾ $4.01 for
enoxaparin for deep venous thromboembolism or pulmonary embo- UFH and US$9.87 ⫾ $2.44 for enoxaparin (P ⬍ 0.001). The total costs
lism in a tertiary public hospital. To estimate medical costs we com- considering the mean time of use were US$88.39 and US$69.11.
puted the acquisition costs of drugs, supplies for administration, labo-
Conclusion: The treatment of VTE with enoxaparin provided cost sav-
ratory tests, and hospitalization cost according to the patient ward.
ings in a large teaching hospital located in southern Brazil.
Results: One hundred sixty-seven patients aged 18 to 92 years were
Keywords: heparin, deep venous thrombosis, utilization study, cost
studied (50 treated with UFH and 117 with enoxaparin). The median of
analysis.
days in use of heparin was the same in both groups. Activated partial
thromboplastin time was monitored in 98% of patients using UFH and Copyright © 2011, International Society for Pharmacoeconomics and
56.4% using enoxaparin. Nonstatistically significant differences were Outcomes Research (ISPOR). Published by Elsevier Inc.

which could result in different antithrombotic effects, but there is

Introduction no evidence that any LMWH preparation is better or worse than
Deep venous thrombosis (DVT) in the lower extremities is the another in terms of efficacy or safety outcomes [11].
most frequent manifestation of venous thromboembolism (VTE), Enoxaparin is among the most widely studied treatments for
with an incidence of 0.48 to 1.6 cases/1000 persons-year among VTE [12,13]. Despite LMWH having a greater acquisition cost,
community residents [1–3]. It is a common condition affecting previous pharmacoeconomic analyses have shown that LMWH
mainly inpatients and rates increase with age. The most life- is more cost-effective than UFH [14 –16]. It has been calculated
threatening manifestation is pulmonary embolism (PE), affecting that outpatient treatment with LMWH may save $1641 per pa-
0.23 to 0.69 cases/1000 person-years [4]. The treatment of VTE rec- tient in comparison to UFH hospital treatment [17]. This eco-
ommended by the American College of Chest Physicians [5] in- nomic benefit of outpatient treatment of VTE seems to be pres-
volves short-term low-molecular-weight-heparin (LMWH) or un- ent in different health systems of developed countries,
fractionated heparin (UFH) therapy plus long-term oral warfarin although the same can not be extrapolated to developing na-
therapy. Anticoagulant therapy with UFH followed by warfarin tions. Economic evaluations in developing countries are desir-
prevents thrombus extension, reduces the risk of recurrent able to estimate VTE hospital treatment cost with UFH and
thrombosis, and prevents death in patients with VTE [6,7]. Subcu- LMWH, because people with low income do not have access to
taneous LMWH is as effective and safe as conventional UFH ther- outpatient treatment with LMWH. We conducted a cohort study
apy, but does not require laboratory monitoring and is less likely to to evaluate the direct costs of short-term heparin anticoagula-
cause bleeding, immune thrombocytopenia, and osteoporosis [8 – tion treatment for VTE in a large teaching hospital located in
10]. LMWH preparartions differ considerably in composition, southern Brazil, from the institutional perspective.

Conflicts of interest: The authors have indicated that they have no conflicts of interest with regard to the content of this
article.
* Address correspondence to: Leila Beltrami Moreira, Farmacologia Clínica sala 947, Hospital de Clínicas de Porto Alegre, Ramiro Barcelos,
2350, 90.035-903, Porto Alegre, RS, Brazil.
E-mail : lbmoreira@hcpa.ufrgs.br.
1098-3015/$36.00 – see front matter Copyright © 2011, International Society for Pharmacoeconomics and Outcomes Research (ISPOR).
Published by Elsevier Inc.
doi:10.1016/j.jval.2011.05.017
S90 VALUE IN HEALTH 14 (2011) S89 –S92

tients that received both heparins (Table 1 in Supplemental Ma-

Methods terials found at: doi:10.1016/j.jval.2011.05.017). Patients were 18
This was a cohort study that prospectively included patients to 92 years old; UFH was first prescribed for 50 (29.9%) patients
hospitalized from March 2005 through January 2007 in a univer- and LMWH for 117 (70.1%) patients. DVT was treated in 107
sity-affiliated, general tertiary teaching hospital with 749 beds cases, PE in 43 cases, and both conditions in 17 cases. Diabetes,
in southern Brazil. These patients had been treated with intra- surgery, infectious disease, and renal failure were significantly
venous UFH or subcutaneous LWMH for suspected or confirmed more frequent in the group that began anticoagulation with
DVT or PE. The study was approved by the institutional review UFH. Previous PE was more frequent in LMWH group (P ⫽ 0.015).
board. Only general surgery had crude association with combined end-
Patients receiving UFH or LWMH were identified through the point (death, bleeding, or recurrent VTE). Among nine critically
institution’s computerized prescription system and had their ill patients, six were treated with UFH (12%) and three with
clinical records revised to be included in the study. All patients LMWH (2,6%) (chi-square P ⫽ 0.036). The DVT and PE diagnoses
receiving an anticoagulation dose of heparin to treat VTE were were confirmed in 93.0% and in 51.7% of patients, respectively.
potentially eligible. The LMWH included on the hospital formu- The median of days in use of heparin was the same in both
lary was enoxaparin, based in the lowest price acquisition pol- groups and warfarin was initiated at the first day in 28.1% of 32
icy of the institution. Patients identified with DVT or PE were patients on the UFH group and in 24.2% of 89 patients on LMWH
prospectively followed-up until the end of the heparin antico- group treated with oral anticoagulant. aPTT was monitored in
agulation period. The only exclusion criterion was age younger 98% of patients receiving UFH and in 56.4% of patients receiving
than 18 years. Data about diagnosis confirmation, anticoagula- LWMH. Half of the UFH group had aPTT measured less than
tion regimen, duration of treatment, laboratory monitoring once a day (Table 2 in Supplemental Materials found at: doi:
with activated partial thromboplastin time (aPTT), bleeding, 10.1016/j.jval.2011.05.017). Nonstatistically significant differ-
thrombocytopenia, blood transfusion, and protamine prescrip- ences were observed in the number of bleeding events (10.0% in
tion were recorded for all included patients. Adverse events UFH and 9.4% in LMWH; chi-square P ⫽ 1.00), blood transfusion
potentially related to DVT or its treatment were defined as com- (2.0% and 2.6%; chi-square P ⫽ 1.00), and death (8.0% and 3.4%;
bined endpoint and included in-hospital death, bleeding, or re- chi-square P ⫽ 0.242). Only one patient in the UFH group evolved
current VTE. to PE. Combined endpoint consisting of recurrent VTE, bleeding,
Costs were assessed directly from the hospital’s records on or in-hospital death occurred in 20.0% and 14.5% of patients in
prices paid for each one of the elements involved in patients’ as- the UFH and LMWH groups, respectively (chi-square P ⫽ 0.379).
sistance, considering the prices during their period of hospitaliza- In a logistic regression model, adjusted for propensity score to
tion. This information was provided directly by the hospital, so receive UFH and for the hospital ward (clinical or obstetric unit,
retrospective costs did not have to be estimated based on any surgery unit, or intensive care unit), LMWH group members
other indirect information. Total medication costs were calculated showed no significant risk reduction of combined outcome
for each patient, considering the prospectively collected data on (odds ratio 0.87; 95% confidence interval 0.32–2.41; P ⫽ 0.79).
medicines, supplies, and laboratory tests. To estimate direct med- We calculated the drug cost per day of treatment taking into
ical costs we computed the acquisition costs of drugs and supplies account the actual price for the institution (Table 3 in Supplemen-
associated with UFH treatment, laboratory tests, and hospitaliza- tal Materials found at: doi:10.1016/j.jval.2011.05.017). We excluded
tion cost according to the patient ward. The cost of use of auto- patients who were treated initially with one form of heparin and
matic pump for UFH administration was not computed because then changed to other on the follow-up (12 individuals). Daily
they were not hospital property and its charge was covered by the mean cost per patient was US$12.63 ⫾ $4.01 for UFH and US$9.87 ⫾
cost of pump-specific infusion equipment required. The costs $2.44 for LWMH (t test P ⬍ 0,001). The total cost of short-term
were converted into US dollars considering the mean exchange heparin treatment considering the mean length of use was
rate from April to May 2007. US$88.39 and US$69.11, respectively, representing a cost saving of
Data were analyzed with PASW Statistics version 18.0 (2009, US$19.28 per heparin treatment. The medication itself is the main
SPSS Inc., Chicago, IL) and a level of significance of 0.05 was set. component of the LMWH group costs (92.88%), whereas it repre-
Chi-square statistics were used in the comparison of categorical sents only 5.25% of costs in the UFH group (Fig. 1 in Supplemental
variables and Student t or Mann-Whitney tests were applied to Materials found at: doi:10.1016/j.jval.2011.05.017). Sensitivity anal-
compare continuous variables. Although costs and time of treat- ysis was performed changing the drug cost according to the actual
ment had small skewness deviation, t test and Mann-Whitney price of acquisition in 2010 converted to US dollars (exchange rate
results were similar and t test value was provided to compare in June 26, 2010). The daily mean cost became US$81.48 ⫾ $55.52
means. Logistic regression modeling was applied to analyze the for UFH and US$14.23 ⫾ $8.42 for LWMH (t test P ⬍ 0,001), and the
association of heparin form and composite clinical outcomes, to total unadjusted cost of short-term heparin anticoagulation was
take into account potential confounders identified in crude anal- respectively US$757.31 ⫾ $359.64 and US$570.94 ⫾ $201.76 (t test
yses. Sensitivity analyses were performed during June 2010, con- P ⫽ 0.002), representing a cost saving of US$186.37 per heparin
sidering the actual cost of drugs to account for acquisition prices treatment.
variation. Propensity score for LWMH prescription was computed
to adjust for indication bias.
Discussion
This was a pharmacoeconomic analysis of heparin use to treat
VTE conducted in a developing country, through cost analysis
Results
from a public health perspective. This kind of analysis is justified
From the 200 patients included, 33 were excluded because they based on the therapeutic equivalence between initial treatment of
had been treated for arterial thrombosis. Twelve patients re- DVT and PE with UFH and LMWH [5]. One characteristic that dis-
ceived both heparins and were classified in the group of UFH tinguishes our study is the evaluation of a real cohort rather then
(n ⫽ 7) or LMWH (n ⫽ 5) according to the first drug prescribed. a hypothetical one, concomitantly including groups of patients
The change from the first prescribed drug to the other was more receiving different treatments. This reflects the usual care of these
frequent in the UFH group (14.0% vs. 4.3% P ⫽ 0.044). The char- patients and makes it possible to estimate the actual costs of treat-
acteristics of the groups are similar when excluding the 12 pa- ment in a defined setting.
 VALUE IN HEALTH 14 (2011) S89 –S92 S91

Patients included in the study were younger then those of hy- greater as a result. Because of evidence of a drug contaminant
pothetical cohorts [17,18] and similar to those of other pharmaco- associated with greater incidence of adverse reactions UFH had its
economic analyses that collected data from individual patients commercialization stopped [25]. When commercialization re-
[14,19]. Sex is not a risk factor for VTE and is in accordance with the started, there was a considerable price increase for UFH, inflating
nonsignificant predominance of women in the sample [20]. The costs still further.
18.6% prevalence of confirmed PE in the studied sample is similar
to the 19% in a clinical trial of VTE [12]. In the same trial the major
or minor bleeding rates were lower then those recorded in our Conclusions
cohort but they were not different between the treatment groups
either. Treatment of VTE with enoxaparin provided costs savings in a
There was no difference on clinical outcomes between the two large teaching hospital located in southern Brazil, from the insti-
groups. This is in accordance with others studies [12,13], but clin- tutional perspective. Despite differences in the settings where
ical trials are not all consistent about equivalence between enoxa- studies of economic evaluation were conducted, the findings
parin and UFH. A meta-analysis [21] demonstrated a statistically agree in regard to the economic advantages of use of LMWH. Im-
significant reduction in clinical outcomes with LMWH when com- plementation of a critical pathway for anticoagulation is desirable
bining all trials, but benefit persisted only for reduction of the to promote rational use of heparins and to save costs.
thrombus size when the 11 studies with adequate concealment of Source of financial support: This study was funded by Fundo de
allocation before randomization were considered. Comparison Incentivo à Pesquisa/Hospital de Clínicas de Porto Alegre.
with literature is difficult because the conditions affecting en-
rolled patients in clinical trials are quite different and the small
sample size is a limitation that precludes a definitive conclusion of Supplemental Materials
our study.
We did not expect that LMWH prescription was greater than
Supplemental material accompanying this article can be found in
UFH prescription because the institutional policy during the
the online version as a hyperlink at doi:10.1016/j.jval.2011.05.017,
study period reinforced the use of UFH for VTE treatment based
or if hard copy of article, at www.valueinhealthjournal.com/issues
on equal efficacy and lower acquisition costs. The preferred use
(select volume, issue, and article).
of UFH in renal failure is recommended in view of the plasmatic
accumulation due to delayed elimination and risk of excessive
anticoagulant effect [22]. In surgical patients, the preference REFERENCES
could be related to the potential for reversion of anticoagulant
effect with protamine and the shorter half-life then LMWH in
case of bleeding. On the other hand, the preference for using [1] Anderson FA Jr, Wheeler HB, Goldberg RJ, et al. A population-based
LMWH in face of previous PE is not evidence-based. The cross- perspective of the incidence and case-fatality rates of deep vein
over from one drug to another was not expected either, but we thrombosis and pulmonary embolism. The Worcester DVT Study.
Arch Intern Med 1991;151:933– 8.
did not investigate the reasons for it. The greater proportion of [2] Nordström M, Lindblad B, Bergqvist D, et al. A prospective study of the
infected, critically ill, and diabetic patients in the UFH group incidence of deep-vein thrombosis within a defined urban population.
suggests they were at higher risk and may be responsible for the J Intern Med 1992;232:155– 60.
statistically nonsignificant increase in death rate in this group. [3] Naess IA, Christiansen SC, Romundstad P, et al. Incidence and
mortality of venous thrombosis: a population-based study. Thromb
The propensity score was constructed to adjust for these risk Haemost 2007;5:692–9.
factors and was included in the logistic regression model that [4] Silvestein MD, Heit JA, Mohr DN, et al. Trends in the incidence of deep
showed no independent association of heparin group with the vein thrombosis and pulmonary embolism: a 25-year population-
based study. Arch Intern Med 1998;158:585–93.
composite endpoint. But the nonsignificant relative risk reduc-
[5] Hirsh J, Raschke R. Heparin and low-molecular-weight heparin: the
tion of 17% favoring LMWH may be explained by the small sam- seventh ACCP conference on antithrombotic and thrombolytic
ple size and lack of power. therapy. Chest 2004;126:188s–203s.
The total costs of initial treatment for VTE were less expensive [6] Brandjes DBM, Heijboer H, Buller HR, et al. Acenocoumarol and
heparin compared with acenocoumarol alone in the initial treatment
with enoxaparin than with UFH, providing an economic advantage
of proximal-vein thrombosis. N Engl J Med 1992;327:1485–59.
of 21.0%. It provided cost savings regardless of the abuse in order- [7] McRae S, Eikelboom JM. Latest medical treatment strategies for
ing monitoring laboratory tests when LMWH was prescribed. If we venous thromboembolism. Expert opinion on pharmacotherapy
had computed the costs of automatic infusion pump use this eco- 2007;8:1221–33.
[8] Majerus PW, Tollefsen DM. Bood coagulation and anticoagulant,
nomic advantage would be even greater. The economic advan-
thrombolitic, and antiplatelet drugs. In: Brunton LL, Lazo JS, Parker KL,
tages are in agreement with those found in studies that compared (ed.). The Pharmacological Basis of Therapeutics. (11th ed.). New York:
the costs of UFH and LMWH for VTE treatment [14,15,16,18,23,24] Brunton LL, Lazo JS, Parker KL, 2005.
despite the differences in cost components like home care [14,23], [9] Snow V, Qassem A, Barry P, et al. Management of venous
thromboembolism: a clinical practice guideline from the american
outpatient treatment [14], and the health care providers’ perspec- college of physicians and the american academy of famy physicians.
tive [18]. It must be emphasized that the institution is a tertiary Ann Inten Med 2007;146:204 –10.
university-affiliated hospital that acquires great quantities of the [10] Quinlan DJ, McQuillan A, Eikelboom JW. Low-molecular-weight
drug. Therefore, it is a favored buyer that can purchase LMWH at heparin compared with intravenous unfractioned heparin for
treatment of pulmonary embolism: a meta-analysis of randomized,
prices significantly lower than the average wholesale price. The controlled trials. Ann Intern Med 2004;140:175– 83.
costs saving associated to LMWH may be explained mainly by the [11] Kearson C, Ginsberg JS, Julian JA, et al. Comparison of fixed-dose
supplies associated with UFH treatment. Besides of the same du- weight-ajusted unfractionated heparin and low-molecular-weight
heparin for acute treatment of venous thromboembolism. JAMA 2006;
ration of UFH and LMWH use, computing the costs of hospitaliza-
296:23–30.
tion must implicate in greater economic differences because UFH [12] Merli G, Spiro TE, Olsson CG, et al. Subcutaneous enoxaparin once or
seems to be mainly indicated for more severely ill patients. It must twice daily compared with intravenous unfractionated heparin for
be noted that total cost, including hospitalization cost was not treatment of venous thromboembolic disease. Ann Intern Med 2001;
134:191–202.
adjusted for the clinical disadvantages of patients with UFH. The
[13] Mismetti P, Quenet S, Levine M, et al. Enoxaparin in the treatment of
sensitivity analysis was performed computing actual drugs’ prices deep vein thrombosis with or without pulmonary embolism. Chest
in June 2010. The economic LMWH advantage became even 2005;128:2203–10.
S92 VALUE IN HEALTH 14 (2011) S89 –S92

[14] Boucher M, Rodger M, Johnson JA, et al. Shifting from inpatient to [20] Hureta CJS, Wallander MA, Garcia RLA. Risk factors an short-term
outpatient treatment of deep vein thrombosis in a tertiary care center. mortality of venous thromboembolism diagnosed in the primary care
A cost-minimization analysis. Pharmacoterapy 2003; 23:301–9. setting in the United Kingdom. Arch Intern Med 2007;167:935– 43.
[15] Spyropoulos AC, Hurley JS, Ciesla GN, et al. Management of acute [21] Dongen CJJ, Belt AGM, Prins MH, et al. Fixed dose subcutaneous low
proximal deep vein thrombosis: phamacoeconomic evaluation of molecular weight heparins versus adjusted dose unfractionated
outpatient treatment with enoxaparin vs inpatient treatment with heparin for venous thromboembolism. Cochrane Database Syst Rev
unfractioned heparin. Chest 2002;122:108 –14. 2004;18:CD001100.
[16] Gould MK, Dembitzer AD, Sanders GD, et al. Low-molecular-weight [22] Lim W, Dental F, Eikelboom JW, et al. meta-analysis: low-molecular-
weight heparin and bleeding in patients with severe renal
heparins compared with unfractioned heparin for treatment of acute
insufficiency. Ann Intern Med 2006;144:673– 84.
deep venous thrombosis: a cost-effectiveness analysis. Ann Intern
[23] Spyropoulos AC, Frost FJ, Hurley JS, et al. Costs and clinical outcomes
Med 1999;130:789 –99.
associated uit low-molecular-weight heparin vs unfractioned heparin
[17] Holzheimer RG. Low-molecular-weight heparin (LMWH) in the for perioperative bridging in patient receiving long-term oral
treatment of thrombosis. Eur J Med Res 2004;9:225–39. anticoagulat therapy. Chest 2004;125:1642–50.
[18] Aujesky D, Smith K, Cornuz J, et al. Cost-effectiveness of low- [24] Knight KK, Wong J, Hauch O, et al. Economic and utilization outcomes
molecular-weight heparin for treatment of pulmonary embolism. associated with choice of treatment for venous thromboembolism in
Chest 2005;128:1601–10. hospitalized patients. Value Health 2005;8:191–200.
[19] Huse DM, Cummins G, Taylor DC, et al. Outpatient treatment of [25] Kishimoto TK, Viswanathan K, Ganguly T, et al. Contaminated
venous thromboembolism with low-molecular-weight heparin: an heparin associated with adverse clinical events and activation of the
economic evaluation. Am J Manag Care 2002;(89 Suppl.):S10 – 6. contact system. N Engl J Med 2008;358:2457– 67.