A machine learning approach to evaluate the state of hypertension care coverage: From 2016 STEPs survey in Iran

A decision rule algorithm for the detection of patients with hypertension using claims data

Objectives

Claims data covers a large population and can be utilized for various epidemiological and economic purposes. However, the diagnosis of prescriptions is not determined in the claims data of many countries. This study aimed to develop a decision rule algorithm using prescriptions to detect patients with hypertension in claims data.

Methods

In this retrospective study, all Iran Health Insurance Organization (IHIO)-insured patients from 24 provinces between 2012 and 2016 were analyzed. A list of available antihypertensive drugs was generated and a literature review and an exploratory analysis were performed for identifying additional usages. An algorithm with 13 decision rules, using variables including prescribed medications, age, sex, and physician specialty, was developed and validated.

Results

Among all the patients in the IHIO database, a total of 4,590,486 received at least one antihypertensive medication, with a total of 79,975,134 prescriptions issued. The algorithm detected that 76.89% of patients had hypertension. Among 20.43% of all prescriptions the algorithm detected as issued for hypertension, mainly were prescribed by general practitioners (55.78%) and hypertension specialists (30.42%). The validity assessment of the algorithm showed a sensitivity of 100.00%, specificity of 48.91%, positive predictive value of 69.68%, negative predictive value of 100.00%, and accuracy of 76.50%.

Conclusion

The algorithm demonstrated good performance in detecting patients with hypertension using claims data. Considering the large-scale and passively aggregated nature of claims data compared to other surveillance surveys, applying the developed algorithm could assist policymakers, insurers, and researchers in formulating strategies to enhance the quality of personalized care.

Supplementary Information

The online version contains supplementary material available at 10.1007/s40200-024-01519-y.

Patterns of general and abdominal obesity and their association with hypertension control in the iranian hypertensive population: insights from a nationwide study

BACKGROUND

The coexistence of obesity and hypertension (HTN) is a global health concern due to its association with various health abnormalities. This study targeted the association between uncontrolled HTN-defined according to the JNC8 guidelines- and different obesity patterns (general and abdominal) among adult hypertensive individuals.

METHODS

Data for the present investigation were obtained from the 2021 STEPwise Approach to NCD Risk Factor Surveillance (STEPS) national survey in Iran. Participants were classified based on general obesity (BMI) and different abdominal obesity patterns (waist circumference [WC], waist-to-hip ratio [WHR], and waist-to-height ratio [WHtR]). Data were weighted by sex, age, and residence (rural and urban). Multivariate logistic regression models were performed to determine the association between different obesity patterns and uncontrolled HTN, adjusting for confounders including demographic variables, lifestyle factors, and history of metabolic abnormalities.

RESULTS

A total of 8,692 hypertensive adult subjects ≥ 18 years were recruited from all provinces in Iran. The overall mean age of participants was 55.8 ± 0.15, and 55.6% being women. The prevalence of general obesity among controlled and uncontrolled hypertensive patients was 30.3% and 69.8%, respectively. Regarding abdominal obesity, the prevalence among controlled and uncontrolled hypertensive patients was 29.8% and 70.2% based on WC, 28.4% and 71.6% based on WHR, and 28.8% and 71.2% based on WHtR, respectively. Compared to normal weight, underweight (adjusted odds ratio [AOR] = 0.94, 95% CI: 0.57-1.56), overweight (1.37 [1.16-1.61]), and general obesity (1.47 [1.24-1.75]) were associated to uncontrolled HTN compared to normal weight. Abdominal obesity according to WC (1.30 [1.13-1.51]), WHR (1.31 [1.10-1.53]), and WHtR (1.39 [1.11-1.74]) was also associated with uncontrolled HTN.

CONCLUSION

Both general and abdominal obesity are more prevalent and strongly associated with uncontrolled HTN in hypertensive patients. These findings underscore the need for healthcare providers to implement targeted interventions promoting healthy lifestyle changes to mitigate these risk factors and improve HTN management.

Predicting dyslipidemia incidence: unleashing machine learning algorithms on Lifestyle Promotion Project data

BACKGROUND

Dyslipidemia, characterized by variations in plasma lipid profiles, poses a global health threat linked to millions of deaths annually.

OBJECTIVES

This study focuses on predicting dyslipidemia incidence using machine learning methods, addressing the crucial need for early identification and intervention.

METHODS

The dataset, derived from the Lifestyle Promotion Project (LPP) in East Azerbaijan Province, Iran, undergoes a comprehensive preprocessing, merging, and null handling process. Target selection involves five distinct dyslipidemia-related variables. Normalization techniques and three feature selection algorithms are applied to enhance predictive modeling.

RESULT

The study results underscore the potential of different machine learning algorithms, specifically multi-layer perceptron neural network (MLP), in reaching higher performance metrics such as accuracy, F1 score, sensitivity and specificity, among other machine learning methods. Among other algorithms, Random Forest also showed remarkable accuracies and outperformed K-Nearest Neighbors (KNN) in metrics like precision, recall, and F1 score. The study's emphasis on feature selection detected meaningful patterns among five target variables related to dyslipidemia, indicating fundamental shared unities among dyslipidemia-related factors. Features such as waist circumference, serum vitamin D, blood pressure, sex, age, diabetes, and physical activity related to dyslipidemia.

CONCLUSION

These results cooperatively highlight the complex nature of dyslipidemia and its connections with numerous factors, strengthening the importance of applying machine learning methods to understand and predict its incidence precisely.

Mortality and disability-adjusted life years in North Africa and Middle East attributed to kidney dysfunction: a systematic analysis for the Global Burden of Disease Study 2019

BACKGROUND

The study aimed to estimate the attributable burden to kidney dysfunction as a metabolic risk factor in the North Africa and Middle East (NAME) region and its 21 countries in 1990-2019.

METHODS

The data used in this study were obtained from the Global Burden of Diseases (GBD) 2019 study, which provided estimated measures of deaths, disability-adjusted life years (DALYs), and other epidemiological indicators of burden. To provide a better insight into the differences in the level of social, cultural, and economic factors, the Socio-Demographic Index (SDI) was used.

RESULTS

In the NAME region in 2019, the number of deaths attributed to kidney dysfunction was 296 632 (95% uncertainty interval: 249 965-343 962), which was about 2.5 times higher than in the year 1990. Afghanistan, Egypt, and Saudi Arabia had the highest, and Kuwait, Turkey, and Iran (Islamic Republic of) had the lowest age-standardized rate of DALYs attributed to kidney dysfunction in the region in 2019. Kidney dysfunction was accounted as a risk factor for ischemic heart disease, chronic kidney disease, stroke, and peripheral artery disease with 150 471, 111 812, 34 068, and 281 attributable deaths, respectively, in 2019 in the region. In 2019, both low-SDI and high-SDI countries in the region experienced higher burdens associated with kidney dysfunction compared to other countries.

CONCLUSIONS

Kidney dysfunction increases the risk of cardiovascular diseases burden and accounted for more deaths attributable to cardiovascular diseases than chronic kidney disease in the region in 2019. Hence, policymakers in the NAME region should prioritize kidney disease prevention and control, recognizing that neglecting its impact on other diseases is a key limitation in its management.

Shapely additive values can effectively visualize pertinent covariates in machine learning when predicting hypertension

Machine learning methods are widely used within the medical field to enhance prediction. However, little is known about the reliability and efficacy of these models to predict long-term medical outcomes such as blood pressure using lifestyle factors, such as diet. The authors assessed whether machine-learning techniques could accurately predict hypertension risk using nutritional information. A cross-sectional study using data from the National Health and Nutrition Examination Survey (NHANES) between January 2017 and March 2020. XGBoost was used as the machine-learning model of choice in this study due to its increased performance relative to other common methods within medical studies. Model prediction metrics (e.g., AUROC, Balanced Accuracy) were used to measure overall model efficacy, covariate Gain statistics (percentage each covariate contributes to the overall prediction) and SHapely Additive exPlanations (SHAP, method to visualize each covariate) were used to provide explanations to machine-learning output and increase the transparency of this otherwise cryptic method. Of a total of 9650 eligible patients, the mean age was 41.02 (SD = 22.16), 4792 (50%) males, 4858 (50%) female, 3407 (35%) White patients, 2567 (27%) Black patients, 2108 (22%) Hispanic patients, and 981 (10%) Asian patients. From evaluation of model gain statistics, age was found to be the single strongest predictor of hypertension, with a gain of 53.1%. Additionally, demographic factors such as poverty and Black race were also strong predictors of hypertension, with gain of 4.33% and 4.18%, respectively. Nutritional Covariates contributed 37% to the overall prediction: Sodium, Caffeine, Potassium, and Alcohol intake being significantly represented within the model. Machine Learning can be used to predict hypertension.