Related Experiment Video
Updated: Sep 16, 2025
03:57
Generation of Hypoparathyroid Rats via Carbon-Nanoparticle-Assisted Parathyroidectomy
Published on: July 14, 2023
2.0K
Predictive Modeling of Hyperparathyroidism in Patients With Benign Thyroid Nodules: A Cohort Study Using the Vizient
Christopher S Hollenbeak1, Qiang Hao1, Melody Greer2
1Department of Health Policy and Administration, College of Health and Human Development, The Pennsylvania State University, Pennsylvania, USA.
Head & Neck
|July 12, 2025
Summary
Predictive models can identify primary hyperparathyroidism (pHPT) in patients with benign thyroid nodules, aiding earlier diagnosis. Machine learning models showed comparable performance to logistic regression in predicting pHPT.
Area of Science:
- Endocrinology
- Medical Informatics
- Predictive Analytics
Background:
- Primary hyperparathyroidism (pHPT) is a leading cause of hypercalcemia, with a significant percentage of cases remaining undiagnosed.
- Early diagnosis and treatment of pHPT are crucial for managing hypercalcemia and its associated complications.
Purpose of the Study:
- To evaluate the efficacy of predictive modeling using a large clinical database to identify pHPT in patients with benign thyroid nodules.
- To compare the performance of logistic regression with machine learning algorithms for pHPT prediction.
Main Methods:
- Retrospective analysis of the Vizient Clinical Database (CDB) from over 1000 hospitals (2020-2023).
- Development of a predictive model for pHPT using logistic regression and comparison with Gaussian naive Bayes, stochastic gradient descent, and histogram-based gradient boosting classifiers.
- Controlled analyses for demographics, comorbidities, and medication use; outcome measured by ICD-10 codes; model performance assessed by ROC curve analysis.
Main Results:
- The histogram gradient boosting model achieved an area under the ROC curve of 68.7%, slightly outperforming logistic regression (68.1%).
- Classification accuracy was high across models, with logistic regression, gradient descent, and histogram gradient boosting achieving 80.4% and 80.5% correct classifications, respectively.
- Logistic regression at a 5% threshold yielded 38.5% sensitivity and 81.8% specificity for pHPT detection.
Conclusions:
- Predictive modeling is feasible for identifying pHPT in patients with benign thyroid nodules using large clinical datasets.
- The developed predictive models, particularly machine learning algorithms, can potentially be integrated into clinical decision support systems.
- Alerting clinicians to potential undiagnosed pHPT can facilitate timely diagnosis and treatment, improving patient outcomes.
Related Concept Videos
00:59The Parathyroid Glands
2.6K
The two pairs of parathyroid glands embedded within the posterior surface of the thyroid gland are restricted by a dense capsule around them. These glands comprise two distinct cell populations—parathyroid oxyphil and parathyroid principal cells- pivotal in calcium homeostasis.
Oxyphil cells, whose functions remain elusive, emerge during late puberty, adding a layer of complexity to the parathyroid gland's intricacies. In contrast, principal parathyroid cells undertake a vital role by...
Oxyphil cells, whose functions remain elusive, emerge during late puberty, adding a layer of complexity to the parathyroid gland's intricacies. In contrast, principal parathyroid cells undertake a vital role by...
2.6K
01:13Sensitivity, Specificity, and Predicted Value
673
In healthcare diagnostics, laboratory tests play a crucial role in identifying and diagnosing a wide range of medical conditions. However, interpreting test results is not always straightforward. An abnormal test result does not always confirm the presence of a disease, just as a normal result does not guarantee its absence. To assess the reliability of these diagnostic tools, healthcare practitioners rely on two key statistical indicators: sensitivity and specificity.
Sensitivity is the...
Sensitivity is the...
673