Stable warfarin dose prediction in sub-Saharan African patients: A machine-learning approach and external validation of a clinical dose-initiation algorithm.
Ontology highlight
ABSTRACT:
SUBMITTER: Asiimwe IG
PROVIDER: S-EPMC8752108 | biostudies-literature |
REPOSITORIES: biostudies-literature
Similar Datasets
Project description:BackgroundWarfarin dosing in cardiac surgery patients is complicated by a heightened sensitivity to the drug, predisposing patients to adverse events. Predictive algorithms are therefore needed to guide warfarin dosing in cardiac surgery patients.ObjectiveThis study aimed to develop and validate an algorithm for predicting the warfarin dose needed to attain a therapeutic international normalized ratio (INR) at the time of discharge in cardiac surgery patients.MethodsWe abstracted variables influencing warfarin dosage from the records of 1031 encounters initiating warfarin between April 1, 2011, and November 29, 2019, at St Michael's Hospital in Toronto, Ontario, Canada. We compared the performance of penalized linear regression, k-nearest neighbors, random forest regression, gradient boosting, multivariate adaptive regression splines, and an ensemble model combining the predictions of the 5 regression models. We developed and validated separate models for predicting the warfarin dose required for achieving a discharge INR of 2.0-3.0 in patients undergoing all forms of cardiac surgery except mechanical mitral valve replacement and a discharge INR of 2.5-3.5 in patients receiving a mechanical mitral valve replacement. For the former, we selected 80% of encounters (n=780) who had initiated warfarin during their hospital admission and had achieved a target INR of 2.0-3.0 at the time of discharge as the training cohort. Following 10-fold cross-validation, model accuracy was evaluated in a test cohort comprised solely of cardiac surgery patients. For patients requiring a target INR of 2.5-3.5 (n=165), we used leave-p-out cross-validation (p=3 observations) to estimate model performance. For each approach, we determined the mean absolute error (MAE) and the proportion of predictions within 20% of the true warfarin dose. We retrospectively evaluated the best-performing algorithm in clinical practice by comparing the proportion of cardiovascular surgery patients discharged with a therapeutic INR before (April 2011 and July 2019) and following (September 2021 and May 2, 2022) its implementation in routine care.ResultsRandom forest regression was the best-performing model for patients with a target INR of 2.0-3.0, an MAE of 1.13 mg, and 39.5% of predictions of falling within 20% of the actual therapeutic discharge dose. For patients with a target INR of 2.5-3.5, the ensemble model performed best, with an MAE of 1.11 mg and 43.6% of predictions being within 20% of the actual therapeutic discharge dose. The proportion of cardiovascular surgery patients discharged with a therapeutic INR before and following implementation of these algorithms in clinical practice was 47.5% (305/641) and 61.1% (11/18), respectively.ConclusionsMachine learning algorithms based on routinely available clinical data can help guide initial warfarin dosing in cardiac surgery patients and optimize the postsurgical anticoagulation of these patients.
| S-EPMC10733832 | biostudies-literature
Project description:Despite some previous examples of successful application to the field of pharmacogenomics, the utility of machine learning (ML) techniques for warfarin dose predictions in Caribbean Hispanic patients has yet to be fully evaluated. This study compares seven ML methods to predict warfarin dosing in Caribbean Hispanics. This is a secondary analysis of genetic and non-genetic clinical data from 190 cardiovascular Hispanic patients. Seven ML algorithms were applied to the data. Data was divided into 80 and 20% to be used as training and test sets. ML algorithms were trained with the training set to obtain the models. Model performance was determined by computing the corresponding mean absolute error (MAE) and % patients whose predicted optimal dose were within ±20% of the actual stabilization dose, and then compared between groups of patients with "normal" (i.e., > 21 but <49 mg/week), low (i.e., ≤21 mg/week, "sensitive"), and high (i.e., ≥49 mg/week, "resistant") dose requirements. Random forest regression (RFR) significantly outperform all other methods, with a MAE of 4.73 mg/week and 80.56% of cases within ±20% of the actual stabilization dose. Among those with "normal" dose requirements, RFR performance is also better than the rest of models (MAE = 2.91 mg/week). In the "sensitive" group, support vector regression (SVR) shows superiority over the others with lower MAE of 4.79 mg/week. Finally, multivariate adaptive splines (MARS) shows the best performance in the resistant group (MAE = 7.22 mg/week) and 66.7% of predictions within ±20%. Models generated by using RFR, MARS, and SVR algorithms showed significantly better predictions of weekly warfarin dosing in the studied cohorts than other algorithms. Better performance of the ML models for patients with "normal," "sensitive," and "resistant" to warfarin were obtained when compared to other populations and previous statistical models.
| S-EPMC6987072 | biostudies-literature
Project description:Populations used to create warfarin dose prediction algorithms largely lacked participants reporting Hispanic or Latino ethnicity. While previous research suggests nonlinear modeling improves warfarin dose prediction, this research has mainly focused on populations with primarily European ancestry. We compare the accuracy of stable warfarin dose prediction using linear and nonlinear machine learning models in a large cohort enriched for US Latinos and Latin Americans (ULLA). Each model was tested using the same variables as published by the International Warfarin Pharmacogenetics Consortium (IWPC) and using an expanded set of variables including ethnicity and warfarin indication. We utilized a multiple linear regression model and three nonlinear regression models: Bayesian Additive Regression Trees, Multivariate Adaptive Regression Splines, and Support Vector Regression. We compared each model's ability to predict stable warfarin dose within 20% of actual stable dose, confirming trained models in a 30% testing dataset with 100 rounds of resampling. In all patients (n = 7,030), inclusion of additional predictor variables led to a small but significant improvement in prediction of dose relative to the IWPC algorithm (47.8 versus 46.7% in IWPC, p = 1.43 × 10-15). Nonlinear models using IWPC variables did not significantly improve prediction of dose over the linear IWPC algorithm. In ULLA patients alone (n = 1,734), IWPC performed similarly to all other linear and nonlinear pharmacogenetic algorithms. Our results reinforce the validity of IWPC in a large, ethnically diverse population and suggest that additional variables that capture warfarin dose variability may improve warfarin dose prediction algorithms.
| S-EPMC8585774 | biostudies-literature
Project description:BackgroundSurgical complications have tremendous consequences and costs. Complication detection is important for quality improvement, but traditional manual chart review is burdensome. Automated mechanisms are needed to make this more efficient. To understand the generalizability of a machine learning algorithm between sites, automated surgical site infection (SSI) detection algorithms developed at one center were tested at another distinct center.Study designNSQIP patients had electronic health record (EHR) data extracted at one center (University of Minnesota Medical Center, Site A) over a 4-year period for model development and internal validation, and at a second center (University of California San Francisco, Site B) over a subsequent 2-year period for external validation. Models for automated NSQIP SSI detection of superficial, organ space, and total SSI within 30 days postoperatively were validated using area under the curve (AUC) scores and corresponding 95% confidence intervals.ResultsFor the 8,883 patients (Site A) and 1,473 patients (Site B), AUC scores were not statistically different for any outcome including superficial (external 0.804, internal [0.784, 0.874] AUC); organ/space (external 0.905, internal [0.867, 0.941] AUC); and total (external 0.855, internal [0.854, 0.908] AUC) SSI. False negative rates decreased with increasing case review volume and would be amenable to a strategy in which cases with low predicted probabilities of SSI could be excluded from chart review.ConclusionsOur findings demonstrated that SSI detection machine learning algorithms developed at 1 site were generalizable to another institution. SSI detection models are practically applicable to accelerate and focus chart review.
| S-EPMC8679130 | biostudies-literature
Project description:VKORC1 and CYP2C9 genotypes explain less variability in warfarin dose requirements in African Americans compared with Europeans. Variants in BCKDK and GATA-4 gene regions, purported to regulate VKORC1 and CYP2C9 expression, have been shown to play an important role in warfarin dose requirements in Europeans and Asians, respectively. We sought to determine whether rs56314408 near BCKDK or GATA-4 rs2645400 influence warfarin dose requirements in 200 African Americans. Unlike the strong linkage disequilibrium (LD) between rs56314408 and VKORC1 rs9923231 in Europeans, they were not in LD in African Americans. No associations were found on univariate analysis. On multivariable analysis, rs56314408 was associated (P = 0.027) with dose in a regression model excluding VKORC1 rs9923231, and GATA-4 rs2645400 was associated (P = 0.032) with dose in a model excluding CYP2C (CYP2C9*2, *3, *5, *6, *8, and *11, CYP2C rs12777823) variants. Neither variant contributed to dose in the model that included both VKORC1 rs9923231 and CYP2C variants. Our results do not support contributions of the studied variants to warfarin dose requirements in African Americans. However, they illustrate the value of studies in African descent populations, who have low LD in their genome, in teasing out genetic variation underlying drug response associations. They also emphasize the importance of confirming associations in persons of African ancestry.
| S-EPMC7993290 | biostudies-literature
Project description:The main aim of this study was to screen various genetic and nongenetic factors that are known to alter warfarin response and to generate a model to predict stable warfarin maintenance dose for Indian patients. The study comprised of 300 warfarin-treated patients. Followed by extensive literature review, 10 single-nucleotide polymorphisms, that is, VKORC1-1639 G>A (rs9923231), CYP2C9*2 (rs1799853), CYP2C9*3 (rs1057910), FVII R353Q (rs6046), GGCX 12970 C>G (rs11676382), CALU c.*4A>G (rs1043550), EPHX1 c.337T>C (rs1051740), GGCX: c.214+597G>A (rs12714145), GGCX: 8016G>A (rs699664), and CYP4F2 V433M (rs2108622), and 5 nongenetic factors, that is, age, gender, smoking, alcoholism, and diet, were selected to find their association with warfarin response. The univariate analysis was carried out for 15 variables (10 genetic and 5 nongenetic). Five variables, that is, VKORC1-1639 G>A, CYP2C9*2, CYP2C9*3, age, and diet, were found to be significantly associated with warfarin response in univariate analysis. These 5 variables were entered in stepwise and multiple regression analysis to generate a prediction model for stable warfarin maintenance dose. The generated model scored R2 of .67, which indicates that this model can explain 67% of warfarin dose variability. The generated model will help in prescribing more accurate warfarin maintenance dosing in Indian patients and will also help in minimizing warfarin-induced adverse drug reactions and a better quality of life in these patients.
| S-EPMC6714673 | biostudies-literature
Project description:With complex traits and diseases having potential genetic contributions of thousands of genetic factors, and with current genotyping arrays consisting of millions of single nucleotide polymorphisms (SNPs), powerful high-dimensional statistical techniques are needed to comprehensively model the genetic variance. Machine learning techniques have many advantages including lack of parametric assumptions, and high power and flexibility.We have applied three machine learning approaches: Random Forest Regression (RFR), Boosted Regression Tree (BRT) and Support Vector Regression (SVR) to the prediction of warfarin maintenance dose in a cohort of African Americans. We have developed a multi-step approach that selects SNPs, builds prediction models with different subsets of selected SNPs along with known associated genetic and environmental variables and tests the discovered models in a cross-validation framework. Preliminary results indicate that our modeling approach gives much higher accuracy than previous models for warfarin dose prediction. A model size of 200 SNPs (in addition to the known genetic and environmental variables) gives the best accuracy. The R(2) between the predicted and actual square root of warfarin dose in this model was on average 66.4% for RFR, 57.8% for SVR and 56.9% for BRT. Thus RFR had the best accuracy, but all three techniques achieved better performance than the current published R(2) of 43% in a sample of mixed ethnicity, and 27% in an African American sample. In summary, machine learning approaches for high-dimensional pharmacogenetic prediction, and for prediction of clinical continuous traits of interest, hold great promise and warrant further research.
| S-EPMC3087957 | biostudies-literature
Project description:The ideal dose of the oral anticoagulant warfarin varies widely among patients, mainly due to genetic factors. Genetic variations that impact warfarin pharmacokinetics and the vitamin K cycle are plausible candidates for being associated with warfarin dose requirements. Therefore, the aim of this study was to assess whether polymorphisms in the ABCB1 and CYP4F2 genes were associated with stable warfarin dose requirements in Brazilian patients. This retrospective study included samples from 309 individuals. Genotyping of ABCB1 c.3435C>T and CYP4F2 c.1297G>A were performed by polymerase chain reaction followed by melting curve analysis (HRM-PCR) and TaqMan® genotyping assay, respectively. Stable doses were adjusted in a linear multiple regression model for age, gender, body mass index, self-reported race, use of amiodarone, CYP2C9 (*2 and *3), VKORC1 c.1639G>A, and ABCB1 c.3435C>T or CYP4F2 c.1297G>A. By performing a univariate analysis of variance, we found that the warfarin patients who carry ABCB1 c.3435T variant alleles (CT and TT genotypes) need fewer warfarin stable doses in comparison with the individuals that are CC wild-type: 2.5 (p = 0.003) and 4.3 (p < 0.001) mg/week less, respectively, for the overall group of patients on stable anticoagulation therapeutics (n = 309); and 5.5 (p = 0.006) and 10.2 (p < 0.001) mg/week less, respectively, for the self-declared non-white stable subgroup (n = 76). No statistically significant differences in dose requirements were observed according to CYP4F2 genotypes. In conclusion, our results suggest ABCB1 c.3435C>T variant may influence warfarin dose requirements in Brazilian patients, when associated with other genotypic, demographic and clinical factors.
| S-EPMC5975540 | biostudies-literature
Project description:The objective of this study was to determine whether warfarin dosing algorithms developed for Caucasians and African Americans on the basis of clinical, environmental, and genetic factors will perform better than an empirical starting dose of 5 mg/day. From April 2002 through December 2005, 259 subjects (Caucasians and African Americans) who started using warfarin were prospectively followed until they reached maintenance dose. The Caucasian algorithm included 11 variables (R(2) = 0.43). This model (which predicted 51% of the doses to within 1 mg of the observed dose) performed better than 5 mg/day (which predicted 29% of the doses to within 5 +/- 1 mg). The African-American algorithm included 10 variables (R(2) = 0.28). This model predicted 37% of the doses to within 1 mg of the observed dose, representing a small improvement compared with 5 mg/day (which predicted 34% of the doses to within 1 mg of 5 mg/day). These results were similar to the results we obtained from testing other published algorithms. The dosing algorithms explained <45% of the observed variability in Caucasians, and the algorithms performed only marginally better for African Americans when compared with giving 5 mg empirically.
| S-EPMC2538606 | biostudies-literature
Project description:Variants in the CYP2C9 (i.e. *2 and *3) and VKORC1 (i.e. 1173C/T or -1639G/A) genes have been shown to influence warfarin dose requirements. However, these factors seem to explain less of the dose variability in African Americans who have a lower prevalence of the CYP2C9*2 and *3 and VKORC1 1173T alleles.In African Americans, the VKORC1 rs17886199 variant was statistically significantly associated with log-transformed warfarin maintenance dose, independent of the influence of VKORC1 1173C>T and CYP2C9*2 and *3. However, replication of our finding is needed to confirm the association of rs1786199 SNP in African Americans, since Limdi et al.[3] did not examine the effect of this SNP because the prevalence of the rs1786199 A-allele was too low.To raise hypotheses with regards to whether genetic variants in the VKORC1, CYP2C9, EPHX1, GGCX and ALB genes might influence warfarin dose in African Americans and Caucasians, independent of the effects of the VKORC1 1173C>T and CYP2C9*2 and *3 variants.From a prospective cohort study, we obtained additional DNA on 36 Caucasian and 22 African American warfarin users who reached maintenance dose and genotyped them for tagSNPs (r2<0.8) in VKORC1, EPHX1, GGCX and ALB genes, and one exonic CYP2C9 SNP. Linear regression models were fitted to estimate the relationship (P value) between log-transformed maintenance dose and each SNP and the amount of the warfarin dose variability accounted for by each SNP (partial R2).In African Americans, the VKORC1 rs17886199 A-allele was associated with a lower dose (GG=46.3 mg and GA=25.6 mg; P=0.002), independent of the VKORC1 1173C>T and CYP2C9*2 and *3 variants. Even after applying Bonferroni correction, the P value would still be considered statistically significant. The VKORC1 rs17886199 variant was not found in Caucasians. In Caucasians, the EPHX1 rs1051741 T-allele was associated with a lower dose (CC=41.3 mg and CT=30.0 mg; P=0.04). The latter was no longer statistically significant after applying Bonferroni correction.Our pilot study suggests that the VKORC1 rs17886199 variant could influence warfarin maintenance dose among African Americans, even after accounting for the influence of the VKORC1 1173C>T variant. Future studies with a larger sample size will be needed to confirm our findings.
| S-EPMC2949912 | biostudies-literature