Comparison Of Predictive Value Of Intelligent Predictive Model And HATCH Scores For Atrial Fibrillation In Patients With Hypertension

OBJECTIVE:Atrial fibrillation?AF?remains one of the most common arrhythmia in clinic.Stroke is the most serious complication of atrial fibrillation.Atrial fibrillation is associated with increased morbidity,mortality,severe disability,rate of recurrence,and length of hospital stay of stroke,which places a heavy burden on society.A lot of clinical risk factors play an important role in the development of AF.Hypertension is one of the independent risk factors for atrial fibrillation.Clinically unrecognized and asymptomatic AF is a potentially leading cause of stroke.Therefore,early identification of atrial fibrillation in high-risk individuals is very significant.However,little is known about the clinical risk assessment system to predict the risk of new-onset atrial fibrillation till now particularly for the Asian population.Studies have confirmed that the HATCH scores and CHADS₂scores are useful for risk assessment and risk stratification especially for new-onset AF.The current study was conducted to evaluate the diagnostic value of HATCH scores and CHADS₂scores for atrial fibrillation in patients with hypertension.Also the study aims to compare the accuracy of HATCH scores,CHADS2 scores and"intelligent prediction model of hypertension with atrial fibrillation"in predicting atrial fibrillation among patients with hypertension.In order to avoid misdiagnosis,early identification,and active intervention and treatment by improving clinicians awareness of the accuracy and importance of atrial fibrillation screening,strengthening the screening of high-risk groups,and adopting the method of detecting24-hour dynamic electrocardiogram or long-term ECG recording with HOLTER implantation to capture whether the patient has paroxysmal atrial fibrillation.Taking active treatment of the primary disease,strengthening the control of the risk factors of atrial fibrillation,reducing the occurrence of complications and improving the long-term prognosis is also important at the same time.Methods:Patients diagnosed with hypertension were admitted to the Heart Center of Zhong Shan Hospital affiliated to Dalian University from January 2015 to September 2018.Patients who with secondary hypertension,rheumatic heart disease,hyperthyroidism,pulmonary circulation and systemic congestion due to heart failure,such as interstitial pulmonary edema and cardiac cirrhosis,were excluded.From the database,review the patient cases?including basic information,history,auxiliary examination,clinical diagnosis?,collect the patient's clinical data?including high blood pressure,age,stroke or TIA,chronic obstructive pulmonary disease,heart failure,diabetes?.HATCH scores and CHADS2 scores were calculated for each selected patient according to the scoring rules.The normal distribution test was performed for the measurement data,which was expressed as mean±standard deviation?x±s?,and the classification variables were described as frequency and percentage[n?%?].T-test was used for continuous variables,and chi-square test was used for comparison of classification data.The receiver operating characteristic?ROC?curve was used to calculate the area under the curve?AUC?of the HATCH score and CHADS2 score,and the specificity and sensitivity were calculated.Z-test was used to compare the c-statistics of each score.RESULTS:A total of 2065 patients were enrolled into the study,including 716patients with hypertension complicated with atrial fibrillation and 1349 patients without hypertension complicated with atrial fibrillation.Age,diabetes,heart failure in patients with atrial fibrillation and unincorporated atrial fibrillation contrast difference are statistically significant?p<0.001?.The area under the receiver operating characteristics curves were 0.609?95%confidence interval[CI]:0.583 0.635,p<0.001?for the HATCH score,0.616?95%CI:0.590 0.641,p<0.001?for the CHADS₂score,there were significant differences of c-statistics among HATCH score and CHADS₂score?p<0.001?.The optimum cut-off points for both HATCH score and CHADS₂score were 1.5,the sensitivity and specificity of HATCH score were50.7%,30%,65.9%and 45.7%for the CHADS₂score.CONCLUSIONS:In our study,the predictive value of HATCH score and CHADS₂score for atrial fibrillation in patients with hypertension is limited,CHADS₂score was found to have a slightly but significantly better predictive performance than HATCH score.Through the comparison of specificity,sensitivity and the area under the receiver operating characteristics curves,the intelligent prediction model have a relatively high performance for atrial fibrillation prediction in patients with hypertension,compare to HATCH score and CHADS₂score,which has more vital clinical significance.