Mo110are Big-Data Techniques Useful in Screening for Primary Hyperaldosteronism? Results of a Pilot Study and Their Prospective Application

Abstract

Background and Aims Primary aldosteronism is a rare cause of high blood pressure (BP). Guidelines recommend a screening in several situations, such as hypertension with hypokalemia, sleeping apnea or adrenal mass, by determining the baseline aldosterone/renin ratio. Recent studies point out that hyperaldosteronism may be more frequent than expected in patients who do not have an arbitrary high aldosterone/renin ratio. The application of big data techniques could generate “red flags” that help clinicians to anticipate the diagnosis. We have evaluated the usefulness of big data in the diagnosis of hyperaldosteronism. Method We have used 3 data sources: patient management, laboratory data management and clinical record. With them, Oracle database management system extracts demographic data, blood pressure and analytical values in a single collection. This data are stored in a MongoDB database, where two collections are generated: "patients" (data we are collecting) and "variables" (values obtained from the identifiers of the latest data obtained from the source systems are saved to guide the collection process). Since MongoDB is a schemaless system, it is possible to easily add new variables and their results to the collection "patients", if it is necessary, in order to generate alerts. In order to diagnose hyperaldosteronism, age between 18 and 55 years, BP > 139/89 mmHg, K 3.5 mEq/l and plasma bicarbonate > 26 mEq/L were used. Results From January 2019 to 30th March 2020, 952,520 data were collected and 281,661 outcome variables from 80,410 patients were analyzed. The method generated alerts of possible hyperaldosteronism in 33 patients (18 men, mean age 46.8 + 6.8 years old). The retrospective study of the clinical record of these patients showed that 11 (33%) had sustained hypertension and 54.5% of them needed 3 antihypertensive drugs. 2 patients suffered a severe cardiovascular event (intracranial hemorrhage). Of these 33 patients, aldosterone/renin screening was performed in 2, 2 other patients had an abdominal CT scan, in which no adrenal masses were found. No specific studies were performed to definitively rule out hyperaldosteronism in any patient. In this period of application of big data, the hospital archive coded 3 hyperaldosteronisms: 1 of them was due to Bartter syndrome, another was a misdiagnosis (androgenic syndrome) and the third one was due to medical background of hyperaldosteronism which had been operated years ago. These results were followed by a prospective screening carried on from 30th March 2020 to 1th January 2021. 776,878 supplementary data were extracted, and 232,425 variables from 13,958 patients were analyzed. The system have generated an alert for 12 patients: 2 of them have been diagnosed of primary hyperaldosteronism, another one have an adrenal mass, which is in surgical waiting list, and 3 patients (total: 50%) are waiting medical tests due to a high suspicion of hyperaldosteronism. Conclusion Big data techniques allow us to create “red flags” for the screening of rare diseases and might be an important tool in their diagnosis. Its systematic application guides us to perform specific diagnostic tests in at least 50% of the cases, in order to improve the diagnostic accuracy in these selected patients.

The Metabolism Grey Prediction Model Based on Big Data and Internet of Things Technology

“…If the conditional probability change process of metabolic diffusion under active intrusion of complex network is stable to the direction set potential in a limited time [31,32], the joint distribution probability of metabolic diffusion of active intrusion of complex network is obtained as follows:…”