Comparison of heparin dosing based on actual body weight in non-obese, obese and morbidly obese critically ill patients.

BACKGROUND: Obesity is endemic in the United States and obese patients are at increased risk of thromboembolism but little data are available for dosing unfractionated heparin (UFH). We evaluated the relationship between obesity and UFH efficacy during critical illness by examining UFH infusions in non-obese, obese, and morbidly obese critically ill patients.
MATERIALS AND METHODS: Retrospective review of UFH infusions in non-obese, obese, and morbidly obese critically ill patients. Heparin was initiated without a bolus at 16 units/kg/h or 12 units/kg/h in obese and morbidly obese patients. Demographics, UFH dosage/therapy duration, laboratory values, and bleeding events were reviewed for patients receiving UFH for >24 h. Steady state (SS) was defined as the dosage that resulted in three consecutive activated partial thromboplastin times (aPTT) within target range.
RESULTS: Sixty-two patients were analyzed including 21 non-obese (mean body mass index (BMI) 24.2 ± 2.3); 21 obese (BMI 34.1 ± 3.1); and 20 morbidly obese (mean BMI 55.3 ± 13.7). Patients had otherwise similar demographics. Although 92% had at least one therapeutic aPTT, only 55% of patients reached SS. Six patients developed minor bleeding, but no major hemorrhagic complications. The dosing of heparin based on actual body weight (units/kg/h) and time to first therapeutic aPTT was similar between groups, but dose was statistically higher at steady state in the non-obese (16.3 ± 5.3 non-obese, 11.6 ± 5.5 obese and 11.1 ± 1.2 obese, P = 0.01) with similar times to steady state.
CONCLUSIONS: Dosing of UFH in morbidly obese and obese critically ill patients based on actual body weight and a reduced initial dose was associated with similar time to first therapeutic aPTT and steady state.

INTRODUCTION

The prevalence of obesity, defined as a body mass index (BMI) ≥30 mg/m2, continues to increase in the United States.[1] According to recent data, the prevalence is 33.8% for adults and the rate of morbid obesity (BMI ≥40 kg/m2) is 5.7%.[1] It is estimated that by 2015 more than 40% of the population will be obese.[2] Consequences of endemic obesity include increased risk for cardiovascular disease, diabetes mellitus, cancer, several gastrointestinal disorders and overall mortality.[3] The economic burden associated with obesity is significant, with nearly 10% of the 2002 United States’ health expenditures going toward care of the overweight and obese.

Not surprisingly, with the increasing prevalence of obesity in the United States, more obese patients are being treated in the intensive care unit (ICU).[4] Monitoring of obese critically ill patients is complex. Obtaining good quality radiographs and vascular access can be difficult due to size limitations.[4] The obese critically ill patient has changes in lung and cardiac physiology.[4] Although the forced vital capacity and forced expiratory volume in 1 second increase, functional residual capacity, vital capacity and total lung capacity are maintained.[4] As a result tidal volume is generally based on ideal body weight.[5] Morbid obesity is characterized by increased total blood volume and cardiac output at rest but a depressed ejection fraction and left ventricular contractility at rest and exercise.[4] Cardiac monitoring can be difficult especially when using indexes. For example, a cardiac index in a 5′10˝ 250-kg male patient with a cardiac output of 8 L/min is 2.16 L/min but it is 3.56 L/min in a 5′10˝ 100-kg male also with a cardiac output of 8 L/min. If care is being titrated to cardiac index alone without considering the cardiac output the obese patient may experience adverse effects.

Similarly, dosing of medications in obese patients can be problematic, particularly during critical illness. This is because distribution, metabolism, and protein binding may all be altered by the physiologic changes associated with obesity.[4] Extrapolation of manufacturer dosing recommendations or standard guidelines designed for non-critically ill, non-obese patients can potentially lead to treatment failure or drug toxicity in the obese.[6] Obese and morbidly obese patients have a larger blood volume than normal weight patients due to the additional vasculature required to perfuse the excess adipose tissue.[67] Comparatively, the blood volume of adipose tissue is less than lean tissue.[67] For medications, including unfractionated heparin (UFH), total drug required for therapeutic efficacy is dependent on the volume of distribution and the total blood volume. Medication dosing in obese patients can be further complicated by volume shifts associated with shock and resuscitation during critical illness.

Obesity is a risk factor for venous thromboembolism and there seems to be an incremental risk with increasing BMI.[89] Critically ill obese patients are among those at highest risk for morbidity and mortality of venous thromboembolism.[8] Unfractionated heparin is a commonly used anticoagulant and many studies and guidelines support the dosing based on actual body weight.[610111213141516] Unfortunately, few patients in published studies were obese or morbidly obese, and it is unclear whether actual or ideal body weight should be used to dose these patients.[6] The development and modification of weight-based nomograms for UFH have lagged behind the rapid increase in the number of morbidly obese patients.[17] Two retrospective studies have been recently published comparing dosing of UFH in obese, morbidly obese and non-obese patients.[616] Bauer et al., concluded there were no differences in proportion of first activated partial prothrombin time (aPTT) measurements within goal range in obese patients receiving weight-based heparin dosing but the percentage of critically ill patients was not specified.[16] Riney et al., showed in a study of 273 hospitalized patients, (including 83 critically ill patients) that morbidly obese patients required smaller UFH rates compared to non-obese patients based on the infusion rate per kilogram of actual body weight at the time of the first therapeutic aPTT.[6]

In our surgical ICU since indexes and tidal volumes are based on ideal body weight it was discussed if ideal body weight should be used to dose continuous infusions of medications throughout the health system's ICU. A point of disagreement was dosing of UFH, especially if there would be a difference in time to first therapeutic aPTT. Based on the limited data available to guide UFH dosing in critically ill patients, this study was conducted to compare dosing rates of UFH in the critically ill with and without obesity.

MATERIALS AND METHODS

After approval from the Institutional Review Board, a retrospective review of patients admitted to the surgical or medical ICUs at a tertiary care medical center receiving continuous UFH infusions for >24 h between July and December 2007 was performed using clinical and pharmacy computer systems. Patients aged <18 years, pregnant patients, and prisoners were excluded. Patients were considered non-obese if their BMI was between 20.0-29.9 kg/m2, obese if BMI was between 30.0-39.9 kg/m2 and morbidly obese for BMI >40.0 kg/m2. Our protocol is to treat patients requiring UFH (without a bolus) based on actual body weight at an initial infusion rate of 16 units/kg/h if non-obese or 12 units kg/kg/h if obese or morbidly obese. The initial rates were chosen based on our institution's quality improvement evaluations of heparin. A standardized nomogram is used to achieve a goal aPTT of 57-84 sec (normal range, goal aPTT 1.5-2.5 × normal) or 57-70 sec (low range, 1.5-2 × normal, [Table 1]).

Table 1

Heparin protocol

Patient charts were reviewed for demographics (age, gender, height, weight), indication for UFH, dosing information, laboratory values including aPTT, and bleeding complications. Major bleeding was defined as documented cerebral, gastrointestinal or retroperitoneal bleed. Minor bleeding was defined as ecchymosis, epistaxis, hematoma, hematuria hemoptysis, petechiae or oozing. The primary endpoint was UFH dosage in units/h, unit/kg ideal body weight (IBW)/h and units/kg actual body weight/h at first therapeutic aPTT and time to first therapeutic aPTT. Secondary outcomes included achievement of steady state (defined as a three consecutive aPTTs in therapeutic range [aPTT 57-84 sec for normal range or 57-70 sec for low range]), time to steady state, dosage at steady state aPTT and percentage of aPTTs in the subtherapeutic, therapeutic and supratherapeutic ranges for each group. Continuous data were analyzed using ANOVA for parametric data and Kruskal-Wallis test for non-parametric data. Parametric data are presented as mean ± standard deviation and non-parametric data are presented as median [25%-75% intraquartile range]. Categorical data were analyzed using Fisher's exact test and are presented as frequency distributions. All statistical analyses were performed using SPSS Version 18.0 (SPSS Inc. Chicago, IL) with P < 0.05 considered to be statistically significant.

RESULTS

Sixty-two critically ill patients were included in analysis, 21 non-obese, 21 obese and 20 morbidly obese. Baseline characteristics were similar between the groups [Table 2]. Obesity did not seem to influence achievement of at least one therapeutic aPTT (∼92%) or achievement of steady state (∼60%) [Table 3]. Time to the first therapeutic aPTT as well as time to steady state was also similar between groups.

Table 2

Demographics

Table 3

Heparin dosage

To determine what dosing strategy would be most effective at reaching a therapeutic aPTT, total hourly dosages were evaluated along with actual and ideal weight-based dosages (calculated retrospectively). At the time of the first therapeutic aPTT the mean total doses of UFH were significantly higher in obese patients (878 ± 341 units/h non-obese, 1051 ± 347 units/h obese, vs. 2007 ± 648 units/h morbidly obese, P < 0.001), suggesting an influence of weight. When dosing was corrected for weight using ideal body weight (IBW), there were similar statistically significant differences in these dosages (14.3 ± 4.8 units/kg/h non-obese, 18.0 ± 5.9 units/kg/h obese, vs. 30.1 ± 8.4 units/kg/h morbidly obese, P < 0.001). In contrast, when dosing using actual body weight there were no significant differences (13.5 ± 4.0 units/kg/h non-obese, 11.7 ± 4.5 units/kg/h obese, vs. 12.5 ± 2.9 units/kg/h morbidly obese, P = 0.35).

We next determined the influence of weight on UFH dose required to reach steady state. Similar to the first therapeutic aPTT, the mean total UFH doses at steady state were statistically different between groups (1106 ± 447 units/h non-obese, 1023 ± 401 units/h obese, vs. 1915 ± 594 units/h morbidly obese, P < 0.001). When UFH dosing at steady state was corrected using IBW, the per kg doses remained significantly different (18.1 ± 7.6 units/kg/h non-obese, 17.7 ± 7.0 units/kg/h obese, 29.0 ± 7.0 units/kg/h morbidly obese, P < 0.001). Unlike dosing at first therapeutic aPTT, correction of doses using actual body weight did not resolve the discrepancy in doses for achieving steady state (16.3 ± 5.3 units/kg/h non-obese, 11.6 ± 5.5 units/kg/h obese, vs. 11.1 ± 1.2 units/kg/h morbidly obese, P = 0.01).

Approximately half the aPTT values measured were in the therapeutic range with 52% of cases targeting the lower range of aPTT values [Table 4]. The median proportion of aPTTs in the therapeutic range was 47% (17.5-64.5%) non-obese, 45% (33-66.5%) obese, and 60% (50-64%) for morbidly obese (P = 0.33). The median proportion of aPTT values were subtherapeutic in 33% (12.5-55%) of non-obese, 19% (7.5-33%) of obese, and 17.5% (1.75-39.75%) of morbidly obese (P = 0.17), and the median proportion of aPTTs that were supratherapeutic were 20% (9-31.5%) non-obese, 33% (8.5-45.5%) obese and 20% (2.25-39.5%) morbidly obese (P = 0.43). Approximately 15% of all patients had an aPTT greater than 180 sec including 23.5% non-obese, 9.5% obese and 15% of morbidly obese, P = 0.45. No patients developed a major bleeding event and approximately 10% developed minor bleeding including 4.8% of non-obese, 9.5% of obese and 15% of morbidly obese, P = 0.57. None of the patients that developed minor bleeding had an aPTT value greater than 180 sec.

Table 4

Outcomes

DISCUSSION

In this study, similar outcomes were achieved in critically ill patients when heparin was dosed on actual body weight using our institutional nomogram. The initial heparin dose was 16 units/kg/h in the non-obese and 12 units/kg/h in the obese and morbidly obese. Appropriate dosing of weight-based medications, particularly UFH, continues to be a challenge in the morbidly obese, especially given the potential for adverse outcomes with both under- and overdosing.[17] Unlike previous studies addressing UFH dosing in morbidly obese patients, this study focuses on UFH use in the critically ill.[614161718] With its unpredictable pharmacokinetics, dosing of heparin in the critically ill may be difficult and may explain why approximately 60% of patients had three consecutive aPTT values in goal range regardless of body mass index.[7] In morbidly obese critically ill patients we found they required larger total UFH infusion rates (units/h) and rates based on IBW to achieve first therapeutic aPTT and steady state which is not surprising since dosing of heparin is based on total blood volume which is increased in obesity. Conversely, we found no difference in dosages based upon actual body weight to achieve first therapeutic aPTT. Based on these data, we decided to continue to dose continuous intravenous medication based on actual body weight.

This is the first dedicated study to specifically look at dosing of UFH in critically ill obese patients. Most previous studies comparing the dosing of UFH in the obese to non-obese do not include the morbidly obese or focus only on the initial aPTT values or the time to first therapeutic aPTT value. In the first published study comparing dosing of UFH in 20 obese to 20 non-obese, the authors concluded that there was not a significant difference in the time to first target aPTT, initial or final infusion rates between groups.[14] The mean body weight of the obese group was 95 + 14.4 kg with only 6 patients weighing more than 100 kg, and the results may not be comparable to our study due to these differences in weight. In the largest study published to date of 1,054 patients, Bauer et al., concluded that there was no difference in proportion of the first aPTTs in goal range between groups.[16] Barletta et al., also concluded there was no difference in proportion of patients with therapeutic initial aPTT in a retrospective study of 101 morbidly obese and non-morbidly obese patients treated with UFH.[17] In the study by Bauer et al., all patients received a UFH bolus and statistically more of non-morbidly obese patients received a UFH bolus (95%) compared to the morbidly obese (79%, P = 0.018) in the study by Barletta et al., which may have confounded their results. The use of only the first aPTT as an endpoint in both these studies may be more reflective of use of the bolus dose compared to continuous infusion dose.

The results of our study are similar to the few studies that report data beyond the initial one or two aPTT measurements.[618] Riney et al., prospectively examined 273 patients receiving UFH, including 83 critically ill patients where UFH was dosed based on actual body weight. Approximately a third of patients received a bolus with no difference between groups.[6] The mean rate of UFH required to achieve the first therapeutic aPTT and two consecutive aPTTs in goal range was significantly lower in those with a BMI greater than 40 kg/m2. Dee et al., published the results of their UFH dosing strategy in 55 patients where all patients received an 80 units/kg bolus (maximum bolus 10,000 units) followed by 18 units/kg/h except for those who were more than 50% of their ideal body weight where the initial rate was 15 units/kg/h.[18] There were no differences between study groups with regards to the proportion of patients with at least one therapeutic aPTT (primary endpoint) or time to therapeutic aPTT. The results of both these studies are similar to ours in that patients with significant obesity had improved anticoagulation parameters with the use of a reduced initial dose based on actual body weight.[618]

At our institution, UFH was not bolused and dosing was based on actual body weight. Our results are similar to the studies by Riney et al., and Dee et al., suggesting that UFH should be based on actual body weight with lower initial dosages based on actual body weight. Many of the previously published studies have looked at the dosage at the first aPTT usually drawn 6 h after initiation of UFH. Unfortunately, either all the patients or some of them were administered boluses which confounds the results at the first aPTT. In the study by Riney et al., the mean times to first therapeutic aPTT were similar if patients did or did not receive a bolus. These results are similar to our data where the mean time to first therapeutic aPTT was 16.8 ± 11 h in the morbidly obese, 20.8 ± 11.6 h in the obese, and 17.2 ± 14.1 h in the non-obese.

This study has several limitations. Because of its retrospective nature and small sample size, it may not be sufficiently powered to detect differences between groups. In addition about half of the patients in this study achieved steady state which further decreases the evaluable population and the statistical power for this endpoint. Although weight is the most significant predictor, age, race, gender, renal function, tobacco use, indication and history of diabetes mellitus or thyroid disease may all potentially contribute to variability in UFH dosing,[61019202122] and these variables were not evaluated in this study. Finally, this study was conducted in a single institution with a specific UFH dosing protocol and results may not be applicable to other populations.

CONCLUSION

Obese and morbidly obese critically ill patients required lower dosing of UFH based on actual body weight than non-obese critically ill patients. Dosing of UFH in morbidly obese and obese critically ill patients based on actual body weight and a reduced initial dose was associated with similar time to first therapeutic aPTT and steady state. The initial rate of weight-based UFH should be reduced in obese and morbidly obese critically ill patients.