This article is mentioned in:

Introduction

Valvular heart disease (VHD) represents a prevalent cardiovascular condition, impacting >100 million individuals globally (1). Primarily stemming from rheumatic fever and degenerative valves, VHD manifests as an anomaly in the heart valve structure, disrupting cardiac hemodynamics and diminishing the well-being of affected individuals, thereby posing a substantial risk to their lives (2). Despite a global reduction in the incidence of rheumatic heart disease, it remains a significant health concern in low- to middle-income nations (3). Patients afflicted by this condition often experience a compromised quality of life and face elevated risks of heart failure, stroke and arrhythmia. Surgery to implant artificial heart valves has significantly improved the quality of life for patients and plays a crucial role in treating advanced cardiac valve diseases. With ~30 million heart valves inserted worldwide each year, the replacement of heart valves with prosthetic valves, whether mechanical or bioprosthetic, is an effective solution for severe VHD (4). Individuals who undergo heart valve surgery require long-term anticoagulation therapy to prevent embolic events. Despite the development of new oral anticoagulants that address certain limitations of warfarin therapy, warfarin remains the standard of care for patients with prosthetic heart valves (5).

Warfarin, an established oral vitamin K antagonist, is one of the oldest and most commonly prescribed coagulation inhibitors used in anticoagulant therapy (6). Recognized for its efficacy, safety and cost-effectiveness, it is on the World Health Organization's list of essential medicines (7). Warfarin is widely used to prevent various thromboembolic events and conditions, including prosthetic heart valve replacement, atrial fibrillation, deep venous thrombosis and pulmonary embolism (8). However, monitoring warfarin in a clinical setting poses challenges due to its narrow therapeutic window and heightened sensitivity to drug interactions (9). A notable limitation in warfarin-treated patients is the considerable inter-individual variability in the required dose to reach the target anticoagulant effect (10). To gauge the anticoagulant effects of warfarin, clinicians rely on the index of anticoagulant activity, represented by the international normalized ratio (INR), to attain the desired target range.

The duration taken to attain the effective INR threshold is an essential assessment of anticoagulant performance, leading to the development of numerous warfarin dose prediction models aimed at reaching this stability promptly (11,12). In clinical settings, the patient's response to warfarin exhibits significant variability, with factors such as the initial warfarin dose, interacting drugs, age, liver disease and dietary vitamin K intake all known to influence the prothrombin response (13). Swiftly reaching the therapeutic INR range for warfarin holds potential benefits, including minimizing the length of heparin bridging and reducing treatment costs (14). Conversely, prolonging the time to reach the therapeutic INR range may extend hospital length of stay (LOS) and elevate the risk of cardiovascular complications, particularly in the starting treatment phase. Therefore, it is crucial to investigate the factors related to rapidly reaching the effective INR threshold to expedite the attainment of the therapeutic INR range in patients with VHD (4,7). With this objective, the present study aimed to assess multiple factors that could impact the prompt achievement of the effective INR threshold in warfarin treatment and evaluate its safety in clinical applications.

Patients and methods

Study design and subjects

The retrospective cohort study involved 201 inpatients undergoing heart valve replacement or repair and receiving warfarin prescriptions. Among them, 167 inpatients were first-time users. The enrolment spanned from July 2022 to June 2023 at the First Affiliated Hospital of Zhengzhou University (Zhengzhou, China). Clinical data were collected from an electronic medical records database, focusing on patients who met the following inclusion criteria: i) Age ≥18 years; ii) initiation of oral warfarin therapy with no prior history of warfarin use within 1 month before admission; iii) maintenance of an INR range between 1.5 and 2.5 prior to discharge. Exclusion criteria comprised the following: i) Patients who switched anticoagulant medications and ceased warfarin before discharge; ii) individuals with an INR exceeding 1.3 before commencing warfarin; iii) patients who did not survive until discharge. A cardiac examination of the participants was conducted using color Doppler ultrasonography. The present study was reviewed and approved by the Ethics Committee of the First Affiliated Hospital of Zhengzhou University (Zhengzhou, China).

Data collection and definition

Patient data were gathered, encompassing basic demographics, laboratory results and clinical information obtained during the hospital stay. Basic demographics included gender, age, weight and height. Laboratory results encompassed baseline levels of the creatinine clearance rate, bilirubin, alanine aminotransferase and aspartate aminotransferase, at the start of warfarin therapy, along with the fluctuations in the INR observed throughout hospitalization. Clinical information encompassed the indications for warfarin use, comorbidities (including coronary heart disease, congestive heart failure, diabetes mellitus, stroke and hypertension), alcohol consumption and smoking history, concurrent medications, initial warfarin dose, LOS at the hospital post-medication and adverse effects. Indications for warfarin treatment and comorbidities were derived from the diagnoses recorded at admission and discharge. The initial warfarin dose was categorized as either low or high, with 3 mg/day serving as the cutoff point. Post-medication hospital LOS was calculated as the duration from the commencement of warfarin administration to the day of discharge. The body mass index (BMI) was determined by dividing weight in kilograms by the square of height in meters and was classified into two categories: i) low-BMI (<24 kg/m²); and ii) high-BMI (≥24 kg/m²). Liver dysfunction was characterized by exceeding three times the upper limit of normal in aspartate aminotransferase or alanine aminotransferase levels (or surpassing twice the upper limit of normal for bilirubin levels). A creatinine clearance rate <30 ml/min was used as the criterion for determining renal insufficiency.

The time to reach the therapeutic target INR was identified as the day when the INR level of 1.8 was attained following the start of warfarin treatment during the hospital stay. The patients were categorized into slowly reaching the target INR range group (>7 days) and rapidly reaching the target INR range group (≤7 days) based on the mean time to reach the treatment target INR (1.5-2.5) (7). Adverse outcomes included laboratory-related complications and bleeding episodes (major or non-major). Major bleeding was characterized as a fatal or symptomatic bleeding episode necessitating the transfusion of at least 2 units of red blood cells, a drop in hemoglobin by ≥2 g/dl, or bleeding in critical areas (intraocular, intracranial, pericardial, retroperitoneal, intraspinal, intra-articular or intracompartmental syndrome). Non-major bleeding was characterized as a clinically relevant bleeding that was not classified as major. A laboratory adverse event was defined as an instance in which inpatients had an INR level ≥4.

Statistical analysis

Continuous variables are presented as the mean ± standard deviation or median [interquartile range (IQR)], while categorical variables are expressed as n (%). Statistical significance was assessed using chi-square tests for categorical variables and Mann-Whitney U-tests for continuous variables that were not normally distributed. Factors associated with a low initial dose of warfarin were assessed using both univariate analysis and multivariable logistic regression model analysis. Variables with a significance level of ≤0.20 in the univariate analysis were retained in the final multivariable model. For each variable in the model, the odds ratio (OR) and corresponding 95% confidence interval (CI) were provided. Data analyses were performed using SPSS software (version 21.0; IBM Corp.). P<0.05 was considered to indicate a statistically significant difference.

Results

Patient characteristics

As depicted in Fig. 1, the selection process, guided by the inclusion and exclusion criteria, involved the recording of preoperative baseline clinical data for 201 patients diagnosed with VHD undergoing valve replacement or repair, with primary indications for warfarin treatment episodes. Out of these, 12 patients did not attain the INR values within the target range (1.5-2.5) before discharge. In addition, 8 patients were excluded due to incomplete pre- or postoperative data. Those who changed anticoagulants before discharge (5 cases) and individuals who did not survive until discharge (1 case of severe pneumonia) were also excluded. Among the patients receiving warfarin, a total of 175 patients (87.1%) met the eligibility criteria, achieved the target INR range and were included in the reported results.

Figure 1 Study population and flow chart. INR, international normalized ratio.

In the present study, the main indications for warfarin were VHD necessitating heart valve surgery, encompassing both heart valve replacement and heart valve repair. Among the patients, 36.6, 6.3 and 21.1% underwent mitral valve, tricuspid valve and aortic valve replacement or repair, respectively (data not shown). In addition, 29.1 and 6.9% of the patients underwent double and triple valve procedures (data not shown). All 175 subjects were divided into two groups based on the time taken to reach the therapeutic INR target, using a cutoff of 7 days. Among those receiving warfarin, the mean time to reach the therapeutic target postoperatively was 9.8 days, covering a period of 3-28 days (data not shown). As indicated in Table I, the median duration of hospitalization after medication was notably reduced in the group rapidly reaching the INR target [9 days (IQR, 8-11 days)] compared to the group reaching it more slowly [13 days (IQR, 10-17 days)] (P=0.001).

Table I Patient characteristics for different reaching the target international normalized ratio range groups.

Table I

Patient characteristics for different reaching the target international normalized ratio range groups.

Variable	Totals (n=175)	Slowly reaching the target INR range group (n=78)	Rapidly reaching the target INR range group (n=97)	P-value
Male sex	90 (51.4)	42 (53.8)	48 (49.4)	0.648
Age >65 years	71 (40.6)	25 (32.1)	46 (47.4)	0.045
Body mass index <24 kg/m2	89 (50.9)	32 (41.0)	57 (58.8)	0.020
Smoking history	52 (29.7)	30 (38.5)	22 (22.7)	0.023
Drinking history	59 (33.7)	22 (28.2)	37 (38.1)	0.167
Indications: Thrombosis prophylaxis Comorbidities	34 (19.4)	14 (17.9)	20 (20.6)	0.657
Hypertension	69 (39.4)	28 (35.9)	41 (42.3)	0.391
Diabetes mellitus	21 (12.0)	11 (14.1)	10 (10.3)	0.443
Stroke	7 (4.0)	3 (3.8)	4 (4.1)	0.187
Congestive heart failure	56 (32.0)	25 (32.1)	31 (32.0)	0.990
Coronary heart disease	45 (25.7)	22 (28.2)	23 (23.7)	0.290
Liver dysfunction	21 (12.0)	8 (10.3)	13 (13.4)	0.524
Renal insufficiency	19 (10.9)	10 (12.8)	9 (9.3)	0.454
Concomitant statins or antimicrobials	88 (50.3)	37 (47.4)	51 (52.6)	0.499
Initial dose of warfarin ≥3 mg/day	137 (78.3)	54 (59.2)	83 (85.6)	0.009
Post-medication hospital length of stay, days	12 (9-14)	13 (10-17)	9 (8-11)	0.001

[i] Values are expressed as n (%) or the median (interquartile range). Slowly reaching the target INR range group: Achieving therapeutic target time >7 days; Rapidly reaching the target INR range group: Achieving therapeutic target time ≤7 days. INR, international normalized ratio.

Factors influencing the rapidly reaching INR threshold for warfarin management following heart valve surgery

As indicated in Table I, the study participants had a mean age of 52.7±12.3 years, with 85 (48.6%) being women. Univariate analyses revealed no significant differences in gender between the group slowly reaching the INR target and the group rapidly reaching the INR target (P>0.1). Similarly, there were no statistically significant differences in comorbid conditions (coronary heart disease, congestive heart failure and hypertension), hepatic dysfunction, or concurrent use of statins or antimicrobials between the two groups (P>0.05). Having a BMI <24 kg/m2 (P=0.020), a smoking history (P=0.023) and an initial warfarin dose ≥3 mg/day (P=0.009) were identified as separate factors influencing the rapid achievement of the target INR level for warfarin management.

Factors associated with a reduced initial dose of warfarin

The warfarin doses administered to 175 patients post-operation were examined. Among the patients, 85.1% (149 cases) received an initial warfarin dose of 3 mg, while 14.9% (26 cases) were prescribed an initial dose of <3 mg. As indicated in Table II, multivariable analyses showed no significant differences in features such as gender and age between the group with a low initial dose and the group with a high initial dose. Additional factors with significance values of P≤0.05 were incorporated into the final multivariable framework. A total of six variables were independently related to a reduced initial dose of warfarin, including drinking history (OR, 0.100; 95% CI, 0.016-0.645; P=0.016), thrombosis prophylaxis (OR, 6.885; 95% CI, 1.944-24.386; P=0.003), liver dysfunction (OR, 10.13; 95% CI, 2.533-40.51; P=0.001), renal insufficiency (OR, 4.714; 95% CI, 1.349-16.467; P=0.015), BMI (OR, 3.644; 95% CI, 1.084-12.253; P=0.037) and antimicrobials (OR, 4.141; 95% CI, 1.152-14.885; P=0.029).

Table II Factors associated with the low initial dose of warfarin in the multivariate analysis.

Table II

Factors associated with the low initial dose of warfarin in the multivariate analysis.

Variable	OR (95% CI)	P-value
Female sex	1.399 (0.345-5.683)	0.638
Age, >65 years	2.067 (0.548-7.788)	0.284
Drinking history	0.100 (0.016-0.645)	0.016
Smoking history	0.735 (0.132-4.097)	0.726
Diabetes	1.358 (0.265-6.961)	0.714
Indications: Thrombosis prophylaxis	6.885 (1.944-24.386)	0.003
Liver dysfunction	10.13 (2.533-40.51)	0.001
Renal insufficiency	4.714 (1.349-16.467)	0.015
BMI <24 kg/m2	3.644 (1.084-12.253)	0.037
Antimicrobials	4.141 (1.152-14.885)	0.029

[i] OR, odds ratio; BMI, body mass index.

Warfarin management reduces the risk of adverse events

Bleeding events were assessed as a secondary outcome to compare differences between the groups. Out of the 175 participants, 23 (13.1%) experienced confirmed non-major bleeding events. The locations of non-major bleeding included wounds (15 cases), ecchymosis (2 cases), the gastrointestinal tract (2 cases), occult blood in the stool (2 cases) and microscopic hematuria (2 cases). No major bleeding events were documented across the entire study cohort. As depicted in Fig. 2, those slowly and rapidly reaching the target INR range showed no significant differences in bleeding incidents (12.8 vs. 13.4%, P>0.05). During hospitalization for warfarin management, the occurrence of elevated INR levels (≥4) was observed in 8 patients (4.6%), with a higher occurrence in 7 patients rapidly reaching the target INR range compared with those reaching it slowly (7.7 vs. 2.1%, P=0.045).

Figure 2 Incidence of bleeding events and INR ≥4 events. The occurrence of INR ≥4 events was higher in the rapidly reaching the target INR range group than in the slowly reaching the target INR range group (7.7 vs. 2.1%, P=0.045). INR, international normalized ratio.

Discussion

The current study provided a comprehensive analysis of potential confounding factors influencing the rapid achievement of the therapeutic INR target in patients undergoing heart valve surgery and receiving conventional warfarin management. Notably, the present findings indicated that individuals aged >65 years, with a BMI <24 kg/m2, no smoking history and a starting warfarin dosage >3 mg/day tended to reach the therapeutic INR target rapidly. Furthermore, reaching the INR target quickly was associated with a significant reduction in the post-medication hospital LOS. This outcome aligns with findings from previous research (11-14). The current results suggest that individuals with a higher BMI may necessitate an increased initial warfarin dose vs. the conventional treatment to reach the therapeutic INR more promptly. It is noteworthy that the BMI has not been factored into determining starting warfarin dosing decisions in current recommendations (12).

Furthermore, the present study underscored that being elderly (>65 years of age) stands out as a separate prognostic factor of rapidly reaching the INR target during warfarin management. While a limited number of studies demonstrated that age can considerably affect the warfarin maintenance dose needed to achieve a stable therapeutic INR, there is a scarcity of research exploring the impact of age in determining the warfarin starting dose requirements for expeditious attainment of the therapeutic INR (14-18). The present research contributes data indicating that patients aged ≤65 years old may benefit from commencing warfarin with a higher initial dose, facilitating a more rapid achievement of therapeutic levels and consequently expediting the onset of anticoagulant effects.

Similar to findings in other studies (19-21), the current research highlights that a high warfarin starting dose emerges as a separate prognostic factor for rapidly reaching the INR threshold during warfarin management in patients undergoing heart valve surgery. Consequently, the present study delved into an analysis of situations in which physicians opted for a low initial warfarin dose. The present findings revealed that patients prescribed low initial warfarin doses were more likely to have a drinking history, liver dysfunction, renal insufficiency, concomitant use of antimicrobials and a criterion for thromboprophylaxis in comparison with patients administered higher starting doses of warfarin. Existing guidelines recommend smaller starting doses for aged patients, those experiencing liver dysfunction, heart failure or an increased bleeding risk (22). However, the present results suggested that the warfarin dosing strategies in clinical practices may not completely align with these guidelines.

While statistical significance was not reached in the present cohort, possibly due to its modest size, there is suggestive evidence that severe renal insufficiency could be a factor associated with the rapid achievement of the INR target in warfarin therapy. Other factors explored in previous studies, such as diabetes mellitus and no smoking history, have been specifically linked to the time required to reach the therapeutic INR (18,23). Importantly, the present study suggested that factors including gender, comorbid conditions, justifications for warfarin treatment and coexisting treatments demonstrated no significant impact on the rapid achievement of optimal INR levels for warfarin treatment.

The primary complications associated with warfarin treatment encompass bleeding and overdosing on anticoagulants (24,25). Consequently, the current study aimed to assess the safety of promptly reaching the INR following the commencement of warfarin medication, gauged by the frequency of bleeding events and INR values ≥4. The present findings revealed that the rapid achievement of the INR target with warfarin was not associated with an elevated risk of bleeding events. However, it was associated with an increased occurrence of INR values ≥4, suggesting that patients who swiftly attained the desired INR range were more prone to experiencing an elevated risk of an excessively high INR. Given the strong link between an excessively high INR and associated bleeding risk during warfarin administration (7,18), it is advisable to vigilantly track INR measurements in patients reaching the therapeutic INR rapidly.

The present study possesses certain limitations. Firstly, it is a retrospective study conducted at a single medical center with a relatively small sample size, which restricted the ability to examine all potential influencing factors thoroughly. Furthermore, while the influence of genetic factors on the time required to reach the therapeutic INR range during warfarin therapy is well-documented, the present authors could not include any genetic data in the present analysis due to limitations in genetic testing availability at the hospital. It is worth noting that genetic testing is not routinely performed in several healthcare settings. In clinical practice, physicians often adjust the starting warfarin dose according to individual patient factors. In addition, a limitation of the present study is the lack of further evaluation, including potential additional in vitro and in vivo validation, which would provide a more comprehensive understanding of the findings. Therefore, despite these limitations, these analyses continue to have clinical significance.

The findings of the present study suggested a need for prospective research involving varied warfarin starting doses across various patient populations. Such studies would be instrumental in identifying the most effective warfarin management strategies for reaching the INR threshold rapidly.

The current study assessed the factors influencing the rapid achievement of the target INR threshold for inpatients undergoing heart valve surgery and receiving conventional warfarin management. The results indicated that patients aged >65 years, those with a BMI <24 kg/m2 and those prescribed a starting warfarin dose ≥3 mg/day had a higher likelihood of rapidly attaining the therapeutic INR. Conversely, patients with a BMI ≥24 kg/m2 may require a higher warfarin starting dose. To enhance safety, it is recommended to implement more stringent INR tracking and make effective adjustments to the warfarin dose, particularly for patients reaching an INR ≥1.8 within a week of initiating warfarin therapy.

Acknowledgements

Not applicable.

Funding

Funding: No funding was received.

Availability of data and materials

The data generated in the present study may be requested from the corresponding author.

Authors' contributions

HH and AY: Conception and design of the study, methodology, manuscript writing-editing. HH, AY, XH and YX: Design of the methodology and contribution to data acquisition. XH and LY: Statistical analyses. HH and AY drafted and revised the manuscript. HH and AY checked and confirm the authenticity of all the data. All authors have read and approved the final manuscript.

Ethics approval and consent to participate

This study was reviewed and approved by the Ethics Committee of the First Affiliated Hospital of Zhengzhou University (approval no. 2024-KY-0597).

Patient consent for publication

Not applicable.

Competing interests

The authors declare that they have no competing interests.

References