Appropriateness of Using Vitamin K for the Correction of INR Elevation Secondary to Hepatic Disease in Critically ill Patients: An Observational Study.

BACKGROUND: Using vitamin K for correction of coagulopathy in critically ill patients is controversial with limited evidence. This study aims to evaluate the efficacy and safety of vitamin K in the correction of international normalized ratio (INR) elevation secondary to liver disease in critically ill patients.
METHOD: A retrospective study of critically ill patients with coagulopathy secondary to liver disease. The primary outcome was to evaluate the association between vitamin K administration and the incidence of new bleeding events in critically ill patients with INR elevation; other outcomes were considered secondary. Patients were categorized into two groups based on vitamin K administration to correct INR elevation. The propensity score was generated based on disease severity scores and the use of pharmacological DVT prophylaxis.
RESULTS: A total of 98 patients were included in the study. Forty-seven patients (48%) received vitamin K during the study period. The odds of the new bleeding event was not statistically different between groups (OR 2.4, 95% CI 0.28-21.67, P = .42). Delta of INR reduction was observed with a median of 0.63 when the first dose is given (P-value: <.0001). However the INR reduction with other subsequent doses of vitamin K was not statistically significant.
CONCLUSION: The administration of vitamin K for INR correction in critically ill patients with coagulopathy secondary to liver disease was not associated with a lower odds of new bleeding events. Further studies are needed to assess the value of vitamin K administration in critically ill patients with liver diseases related coagulopathy.

Introduction

Liver plays a major role in blood homeostasis via regulating the activation or inhibition of coagulation factors. Therefore, any failure in liver functionality to synthesize both antithrombotic and prothrombotic factors may lead to coagulopathy.[1,2] The extent of coagulopathy depends on the functionality degree of liver parenchymal cells that are responsible for producing most fibrinolytic and clotting factors.[3,4]

Patients with end-stage liver disease (ESLD) have an imbalance coagulation profile and are at risk for excessive clotting and bleeding. Recently, it is recognized that hemostasis in patients with chronic liver disease is rebalanced, as the fall in anticoagulant proteins is accompanied by a parallel fall in clotting factor levels. However, compared to patients without liver disease, hemostasis balance is relatively unstable, and the risk for bleeding in cirrhotic patients seems to be particularly elevated during decompensation, infections, or acute-on-chronic liver failure (ACLF).[6-8]

Liver disease complications in critically ill patients such as bleeding and thrombosis are prevalent, challenging, and may not be preventable.[7,9] For example, the incidence of new-onset major bleeding in the intensive care unit (ICU) was found in 17% to 20% of cirrhotic patients. One of the reasons behind the hemostatic imbalance in patients with chronic liver disease is decreased synthesis of vitamin K-dependent and independent clotting factors. Vitamin- K plays a vital role in synthesizing active forms of coagulation and anticoagulation factors, including prothrombin, factor VII, XI, X, protein C, and S. Additionally, vitamin- K has a strong correlation with international normalized ratio (INR) elevation in those patients. However, conventional coagulation tests in those patients do not fully reflect the actual derangement in hemostasis and demonstrated to be ineffective in either predicting or guiding the management.[5,6,10]

Patients with liver disease usually experience hemostatic imbalanced risk for bleeding and thrombosis.[8,13] Due to coagulopathy, these patients may have an elevated INR, and some clinicians still making important clinical decisions based on the INR interpretation as a general indication of a patient's overall bleeding risk due to auto-anticoagulation. Management of coagulopathy by correcting INR elevation is common, especially before major procedures, because some procedures progress to either hemorrhage or thrombosis depending on the patient's status and risk factors. Therefore, the management of coagulopathy secondary to liver diseases is controversial and not fully elucidated. This study aims to evaluate the efficacy and safety of using vitamin K in critically ill patients with INR elevation secondary to liver disease.

Methods

Study Design

A retrospective observational study of adult ICU patients with coagulopathy secondary to liver disease admitted between January first, 2018, and December 31st, 2018. All the patients who met the inclusion criteria during the study period were included. Patients were categorized into two groups: patients who received any dose via any route of administration of phytonadione (vitamin K) compared to patients who did not receive vitamin K for INR correction secondary to liver disease. Patients were followed daily during ICU stay until ICU discharge after improving, or in-hospital death, whichever occurred first.

Participants

Patients were enrolled in the study if they were critically ill aged ≥18 years old with hepatic disease (s) and significant INR elevation (defined as INR≥ 1.5) within 24 h of ICU admission. Exclusion criteria include receiving prolonged antibacterial therapy for ≥ 21 days, prior administration of any oral anticoagulant, receiving prothrombin complex concentrate (PCC) or fresh frozen plasma (FFP) simultaneously with vitamin K. Additionally, patients with comorbidity that cause coagulopathy other than hepatic diseases (i.e., systemic lupus erythematosus, lupus anticoagulant-hypoprothrombinemia syndrome, antiphospholipid antibody syndrome [APAS]), or that increase bleeding/thrombosis risk (i.e., Hemophilia A, Hemophilia B, von Willebrand disease, Protein C or S deficiency, Antithrombin III deficiency, APAS) were excluded based on the chart-review.

Setting

This study was conducted in the adult ICUs at King Abdulaziz Medical City (KAMC), a tertiary-care academic referral hospital in Riyadh, Saudi Arabia. The ICUs admit medical, surgical, trauma, burn patients and operate as closed units with 71 ICU beds capacity with 24/7 onsite coverage by critical care board-certified intensivists and clinical pharmacists.

Data Collection

Demographic and clinical data including age, gender, weight, body mass index (BMI), associated co-morbidities, laboratory baseline including but not limited to coagulation profile (e.g., INR, aPTT, d-dimer), liver function tests (LFTs), complete blood count within 24 h of ICU admission. Additionally, Glasgow Coma Scale (GCS), Vasoactive Inotropic Score (VIS), Acute Physiology and Chronic Health Evaluation (APACHE II) score, Sequential Organ Failure Assessment (SOFA) score, and the use of pharmacological DVT prophylaxis were recorded for eligible patients on the first day. Moreover, mechanical ventilation, endoscopy, receiving proton pump inhibitor (PPI) treatment dose, transfusions during ICU stay (i.e., RBCs/platelets, fresh frozen plasma [FFP] transfusion, cryoprecipitate), radiology finding (i.e., CT scan, ultrasound, magnetic resonance imaging [MRI]), and surgery during ICU stay were reviewed and recorded.

Outcomes

This study aims to investigate the efficacy and safety of using Phytonadione (Vitamin K) to correct INR elevation secondary to liver disease in critically ill patients. The primary outcome was to evaluate the association between vitamin K administration and incidence of bleeding (ie, major and minor bleeding) in critically ill patients with INR elevation. Secondary outcomes were to evaluate the degree of INR correction with vitamin K, the incidence of thrombosis, RBCs/Platelets transfusion requirement, mechanical ventilation duration, ICU length of stay (LOS), and 30-day mortality.

Thrombosis/infraction was defined using the International Classification of Diseases, Tenth Revision, Clinical Modification (ICD10-CM) code (ie, ischemic stroke, pulmonary embolism, deep vein thrombosis) during ICU stay.

Major bleeding was defined as clinically overt bleeding associated with a fall ≥Hgb 20 g/L, transfusion of ≥2U PRBC or whole blood, retroperitoneal or intracranial bleeding, or requiring urgent medical intervention. While minor bleeding was defined as those not fulfilling the criteria of major or clinically significant bleeding.

Acute kidney injury (AKI) was defined using Acute Kidney Injury Network (AKIN) definition.

Data Management and Statistical Analysis

Collected data was entered in Microsoft Excel after being coded. There were two arms considered in this study, patients who received Vitamin K versus non-Vitamin K. As expected in an observational study, differences in baseline characteristics between the two treatment groups may exist. To adjust for these differences, a propensity score for the use of Vitamin K was generated with APACHE II, SOFA scores and the use of pharmacological DVT prophylaxis. Assuming a confidence interval of 90%, based on standard deviation of 0.5, with an acceptable margin of error of 5%. The sample size needed based on calculation is 270 patients for our study to be powered.

We summarized categorical variables as number (percentage) and numerical variables (continuous variables) as mean and standard deviation (SD) or median and interquartile range (IQR) as appropriate. The normality assumptions were assessed for all numerical variables using statistical test (ie Shapiro–Wilk test) and also using graphical representation (ie histograms and Q-Q plots). We compared categorical variables using the chi square or Fisher exact test, normally distributed numerical variables with the t test, and other quantitative variables with the Mann-Whitney U test. Baseline characteristics, baseline severity and outcome variables were compared between the two groups. The INR changes were compared using Wilcoxon signed-rank test in Vitamin K group alone.

We assessed model fit using the Hosmer-Lemeshow goodness-of-fit test. Generalized linear regression and multiple linear regression were also used to find out the relationship between treatments and the different outcomes considered in this study, adjusting for the generated propensity score. The odds ratios (OR) and estimates with the 95% confidence intervals (CI) were reported for the associations. We considered a P value of < .05 statistically significant and used SAS version 9.4 for all statistical analysis.

Results

Patient Characteristics

A total of 1864 patients were screened. Ninety-eight patients were included; 47 (48%) had received vitamin K during the study period. Table 1 depicts baseline characteristics between the two groups. Patients who received vitamin K were observed to be more diabetic, have a lower GCS baseline, higher rate of chronic kidney disease, ischemic heart disease, atrial fibrillation, and heart failure. On the other hand, patients who did not receive Vitamin K had a higher Vasoactive Inotropic Score (VIS) baseline, lower bilirubin, and D-dimer levels within 24 h of ICU admission. When adjusted for propensity score using severity score within 24 h of ICU admission (APACHE II, SOFA score, and the use of pharmacological DVT prophylaxis), all these differences became insignificant.

Table 1.

Baseline characteristics of the Vitamin K and Non-Vitamin K groups.

	Vitamin K (N = 47)	Non Vitamin K (N = 51)	P value
Age (years), mean (SD)	63.5 (11.43)	58.2 (17.00)	.0924^
Body mass index (BMI)(kg/m2), mean (SD)	29.1 (8.06)	27.1 (7.75)	.3110^
Gender, n(%)
Male	26 (55.3)	28 (54.9)	.9669^^
Female	21 (44.7)	23 (45.1)
Weight (kg), mean (SD)	75.8 (23.56)	68.4 (18.22)	.2129^
Hypertension (HTN), n (%)	22 (46.8)	19 (37.3)	<.0001^^
Diabetes Mellitus (DM), n (%)	28 (59.6)	16 (31.4)	<.0001^^
Chronic kidney disease (CKD), n (%)	11 (23.4)	5 (9.8)	<.0001^^
Ischemic heart disease (IHD), n (%)	4 (8.5)	2 (3.9)	<.0001**
Atrial fibrillation (AFib.), n (%)	3 (6.4)	1 (2.0)	<.0001**
Heart failure (HF), n (%)	3 (6.4)	1 (2.0)	<.0001**
Acute coronary syndrome (ACS), n (%)	1 (2.1)	1 (2.0)	<.0001**
Dyslipidemia (DLP), n (%)	5 (10.6)	8 (15.7)	<.0001^^
Hypothyroidism, n (%)	4 (8.5)	2 (3.9)	<.0001**
Previous use of Abx. within 3 months of ICU admission, n (%)	23 (48.9)	10 (19.6)	.0021^^
Previous admission within 3 months, n (%)	26 (56.5)	11 (28.2)	.0087^^
Baseline within 24 h of ICU admission
Sodium, median (Q1, Q2)	135 (130.0, 139.5)	136 (132.3, 138.8)	.9362^
Acute Kidney Injury, n (%)	18 (38.3)	9 (17.6)	.0223^^
Serum creatinine (mg/dL), mean (SD)	251.2 (205.65)	212.1 (220.23)	.3268^
Blood urea nitrogen (BUN), mean (SD)	16.9 (10.65)	13.4 (8.69)	.2225^
Total bilirubin, median (Q1, Q3)	94 (50.1, 301.5)	19.1 (9.9, 49.2)	<.0001^
International Normalized Ratio (INR), mean (SD)	2.2 (0.71)	1.5 (0.62)	<.0001^
Platelets count, mean (SD)	14.5 (21.69)	4.5 (7.75)	.0164^
Activated partial thromboplastin time (aPTT), mean (SD)	51.0 (23.33)	51.5 (22.25)	.9410^
Gamma-glutamyl transferase (GGT), mean (SD)	116.1 (71.35)	217.8 (171.62)	.2932^
Alanine aminotransferase (ALT), mean (SD)	235.7 (435.05)	232.1 (391.20)	.6612^
Aspartate aminotransferase (AST), mean (SD)	475.4 (1005.58)	230.2 (359.88)	.2897^
Albumin, mean (SD)	29.9 (8.13)	27.4 (5.36)	.1294*
White blood cells (WBCs), mean (SD)	14.7 (15.24)	10.9 (5.88)	.3526^
Fraction of inspired oxygen (FiO2) requirement (%), mean (SD)	40.6 (20.37)	37.4 (14.68)	.7372^
Glasgow Coma Scale (GCS), mean (SD)	10.2 (4.62)	13.5 (2.89)	.0018^
Blood Glucose Level (mmol/L), mean (SD)	14.4 (22.02)	15.5 (5.81)	.0025^
Mean arterial pressure (MAP), mean (SD)	53.80 (14.33)	61.19 (14.56)	.0130^
Lactic acid, mean (SD)	8.0 (7.66)	5.9 (6.66)	.1885^
Hematocrit (Hct), mean (SD)	0.3 (0.07)	0.3 (0.07)	.0480*
Fibrinogen, mean (SD)	2.0 (1.10)	2.6 (2.07)	.6077^
D-dimer, mean (SD)	17.2 (13.04)	11.1 (7.08)	.3039^
Vasoactive Inotropic Score (VIS) 24hrs, mean (SD)	58.4 (130.07)	71.3 (218.68)	.0013^
APACHE II, Median (Q1,Q3)	22.00(14.00, 29.00)	17.00(9.00, 21.00)	.0056^
SOFA, Median (Q1,Q3)	11.00(9.00, 15.00)	7.00(5.00, 9.00)	.0001^
Pharmacological DVT prophylaxis, n (%)	6 (12.5)	17 (33.3)	.012

-Denominator of the percentage is the total number of patients

*T -Test / ^ Wilcoxon rank sum test is used to calculate the P-value.

^^Chi-square test is used to calculate the P-value.

**Fisher Exact test is used to calculate the P-value.

Vitamin K was given by IV route of administration to 93.6% of the patients, with a median dose of 10 mg and for a median duration of three days. Neither all bleeding events (OR 2.4, 95% CI 0.28-21.67, P = 0. 42) nor thrombosis (OR 1.2, 95% CI 0.32-4.85, P = .76) were statistically significant between the two groups. Patients who received vitamin K had a lower requirement of RBCs transfusion; however, it was not statistically significant (Table 2).

The rate of venous thromboembolism (VTE) was 17% (8/47) in the vitamin K group as compared to 3.9% (2/51) in the non-vitamin K group. Patients who received vitamin K were 2.4 times more likely to get VTE than patients who did not. However, this association was not statistically significant when adjusted for propensity score (OR 2.4, 95% CI 0.45- 13.12, P = .30). Patients with any reported thrombosis cases were more likely to receive a cumulative dose of ≥ 35.3 mg of vitamin K (Table 2).

Patients who received vitamin K for INR correction have longer ICU LOS (Estimates [STD] 0.67, 95% CI 0.22-1.12, P = .003), and mechanical ventilation duration (Estimates [STD] 0.93, 95% CI 0.23-1.62, P = .01 (Table 2). Using Vitamin K for correction of INR secondary to liver disease was observed at the first dose (P-value: <.0001); other subsequent doses of vitamin K were not statistically significant for correction. The changes were calculated as the difference between pre-measurements to post-measurements at each dose (Table 3). Table 4 shows all bleeding events as well as all thrombosis events after stratification based on the level of severity using the Child-Pugh and MELD scores. Patients with a Child-Pugh score of C's had a higher rate of all thrombosis cases (8.33%) and all bleeding events (9.38%) than Child-Pugh B and A, respectively. The risk of bleeding and thrombosis increased proportionally with increased MELD score severity.

Discussion

Assessing the degree of coagulopathy in hepatic patients using PT and INR may not represent the best practice or evidence-based approach to patient care. Moreover, administering vitamin K is a common practice that has been identified to manage patients with hepatic disease-associated coagulopathy with an effort to augment the synthetic function of clotting factors. However, this approach may not be predictable, and the appropriateness of using vitamin K is not fully elucidated in critically ill patients in terms of safety and efficacy. Critically ill patients have multiple risk factors for bleeding which may not be correlated with an increased INR reading.[6,8] Contributing factors that would increase the risk of bleeding in critically ill patients may include post-operative bleed, thrombocytopenia, critical illness induced hypofibrinogenemia.

The liver is responsible for synthesizing nearly all procoagulants (eg, factor I, II, III, X, Fibrinogen) and anticoagulant proteins (eg, protein C, S, Z, Antithrombin) responsible for maintaining hemostasis. Hemostasis in cirrhotic patients is a dynamic balance. Therefore, patients with liver diseases usually experience an imbalanced hemostatic variable.[8,13] Due to coagulopathy, these patients may have an elevated INR, and some clinicians still making important clinical decisions based on the INR interpretation as a general indication of a patient's overall bleeding risk due to auto-anticoagulation. Both Child-Pugh and MELD scores incorporate INR as part of their validated estimate for liver disease and mortality estimation, subsequently. This may lead clinicians to observe a paradoxical phenomenon for increased risk of bleeding and thrombosis at the same time. It is noteworthy to be reminded that an increased bleeding risk does not rule out the development of a new thrombus in this patient population. Therefore, it is important to note that elevation of INR secondary to liver diseases poses the risk of thrombosis and bleeding as well, suggesting using a different technique to assess this population's hemostasis (eg, Viscoelastic tests).

In the current study, we evaluated the efficacy and the safety of vitamin K administration to correct INR elevation secondary to liver diseases. The APACHE II and SOFA were calculated within 24 h of ICUs admission and adjusted between the two groups. In our study, the first dose of vitamin K administration was statistically significant in correcting INR with a median of 2.64 (1.52) pre-dosing to 2.01 (0.67) post vitamin K. However, our study shows that other subsequent doses of vitamin K were not having a significant reduction in INR. Also, using vitamin K to correct INR elevation did not decrease bleeding odds, RBCs, platelets, or FFP transfusions requirements. In parallel to our finding, Rivosecchi et al. founds that the first dose of vitamin K was associated with INR reduction by 30%; however, this result appeared only in 16% of patients who received vitamin K. Rivosecchi R et al. concluded that the use of vitamin K is not efficient in correcting coagulopathy secondary to liver cirrhosis as about 62.3% of this patients failed to reach these reductions. Moreover, Saja et al. showed a similar conclusion when evaluating vitamin K administration efficiency in series of coagulation parameters in the four stages of liver diseases. Similarly, several studies have provided data that suggested to assess the effectiveness and safety of vitamin K, and it should be avoided to correct an abnormal laboratory result only without significant bleeding in these patients. Other clinical factors may have better utility to assess the risk of bleeding in patients with cirrhosis such as previous history of variceal bleeding events, instead of using INR guided assessment.

In general, cirrhotic patients who are truly vitamin-K deficient are roughly less than 15%. Administration of vitamin K only to correct an abnormal laboratory result should be minimized. Our study shows that patients who receive vitamin K were 2.4 times more likely to get VTE (whether DVT or PE) than the non-vitamin K group and more prone to thrombotic events; however, it was not statistically significant when adjusted using the propensity score. In addition, all bleeding events were not statistically significant between the two groups. Also, our finding shows that patients with any reported case of thrombosis were more likely to receive an increased cumulative dose of ≥ 35.3 mg of vitamin K, emphasizing the need to determine the appropriate indication (s), dosing, and duration when using vitamin K to correct and manage coagulopathy for patients with hepatic disease.

This study is distinctive and has a set of strengths in which the severity of illness between two groups were not significant (i.e., APACHE 2 and SOFA score). In addition, patients who received prolonged antibacterial therapy for ≥ 21 days or at high risk of thrombosis and coagulopathy were excluded. However, some potential limitations have been noted, as the model's effect is based on a retrospective design. In addition, the underpowered sample size in our study led to a lower capacity to detect differences. Besides, heterogeneity of ICU cases (e.g., medical, surgical patients) and secondary liver disease (eg, acute on top of chronic disease) were not determined. Therefore, they are subjected to confounding that may influence our model findings. Further well-designed studies with a larger sample size are warranted to confirm our findings.

Conclusion

Using vitamin K for INR correction in critically ill patients with coagulopathy secondary to liver disease was not associated with a lower incidence of new bleeding events in our study. The routine prescription of vitamin K for coagulopathy secondary to liver disease in critically ill patients warrants further evaluation. Further studies with proper sample size are needed to assess the value of vitamin K administration in critically ill patients with liver diseases related coagulopathy.

Table 2.

Outcomes of the Vitamin K and Non-Vitamin K groups.

	Vitamin K (N = 47)	Non-Vitamin K (N = 51)	P value	Odds Ratio (OR) (95%CI)	P value
Major Bleeding, n (%)*&	3/27 (11.1)	1/51 (2.0)	.12^^	2.4 (0.28, 21.67)	.42
Minor bleeding, n (%)*&	0/27 (0)	0/51 (0.0)	NA	NA	NA
Bleeding (All cases), n (%)*&	3/27 (11.1)	1/51 (2.0)	.12^^	2.4 (0.28, 21.67)	.42
VTE (DVT / PE), n (%)	8 (17.0)	2 (3.9)	.04**	2.4 (0.45, 13.12)	.30
Thrombosis (All cases), n (%)	9 (19.2)	6 (11.8)	.31^^	1.2 (0.32, 4.85)	.76
30-day mortality, n (%)	29 (61.7)	7 (13.7)	<.0001^^	2.7 (0.79, 9.01)	.11
				beta coefficient (Estimates) (95%CI)	P-value
Red blood cells (RBCs) Transfusion (U), mean (SD)	5.3 (4.62)	4.2 (9.41)	.006^	−0.12 (−0.76, .51)	.71
Platelets Transfusion (U), mean (SD)	14.5 (21.69)	4.5 (7.75)	.02^	0.36 (−0.71, 1.43)	.51
ICU LOS, Median (Q1, Q3)	10.0(4.00, 18.00)	4.0 (2.00, 10.00)	.0003^	0.67 (0.22, 1.12)	.003
MV duration, Median (Q1, Q3)	3.0(1.00, 8.00)	0.0(0.00, 2.50)	.002^	0.93 (0.23, 1.62)	.01

*& Denominator of the percentage is non-bleeding at the time of vitamin K administration.

*T -Test / ^ Wilcoxon rank sum test is used to calculate the P-value.

**Fisher Exact /^^Chi-square test is used to calculate the P-value.

$*propensity score adjusted Generalized linear model is used to calculate estimates and P-value.

$ Propensity score adjusted Logistic regression is used to calculate Odds ratio and P-value.

Table 3.

Comparison of INR change between Vitamin K and Non-Vitamin K groups.

	Vitamin K Group (N = 47)		P value
	Pre	Post
INR change at dose# 1, median (IQR)	2.64(1.52)	2.01(0.67)	<.0001
INR change at dose# 2, median (IQR)	2.09(0.63)	1.98(0.59)	.2839
INR change at dose# 3, median (IQR)	2.19(1.39)	2.21(1.45)	.5442
INR change at dose# 4, median (IQR)	2.83(1.80)	2.47(0.96)	.6698

Wilcoxon rank sum test is used to calculate the P-value.

Table 4.

All bleeding and thrombosis cases among different Child Pugh /MELD Score.

	Child Pugh Score
	Child Pugh A (N = 29)	Child Pugh B (N = 26)	Child Pugh C (N = 41)
Bleeding (All case), n (%)	1(1.04%)	3 (3.13%)	9 (9.38%)
Thrombosis (All case), n (%)	2(2.08%)	5(5.21%)	8(8.33%)
	MELD score ≤19 (N = 28)	MELD score 20 to 30 (N = 34)	MELD score >30 (N = 34)
Bleeding (All case), n (%)	1 (1.04%)	4 (4.16%)	8 (8.33%)
Thrombosis (All case), n (%)	4 (4.16%)	5 (5.20)	6 (6.25%)