Exploring Clinical Chemistry Data for Disease Groups
Author Information
Author(s): Jan B. Hemel, Hilko van tier Voet, Roll Hendriks, Frans R. Hindriks, Willem van tier Slik
Primary Institution: University Hospital Groningen
Hypothesis
Can multivariate analysis improve the separability of major disease groups based on clinical chemistry data?
Conclusion
The study found that multivariate analysis provides better separation of disease groups compared to univariate analysis, although some overlap remains.
Supporting Evidence
- Multivariate analysis revealed better distinction between disease groups than univariate analysis.
- Three-dimensional plots provided more insight than two-dimensional plots.
- Non-linear mapping retained distances between data points effectively.
Takeaway
This study looked at blood tests from patients with liver, kidney, and heart diseases to see if we could tell them apart better using fancy math techniques.
Methodology
The study used multivariate analysis techniques on clinical chemistry data from patients with liver, kidney, and heart diseases.
Potential Biases
The selection of patients and assays may introduce bias, as not all relevant assays were performed on every patient.
Limitations
Some valuable information was lost due to the deletion of incomplete variables, and the analysis was limited to patients admitted within 10 days.
Participant Demographics
Patients included 46 with liver diseases, 50 with kidney diseases, and 46 with heart diseases.
Statistical Information
P-Value
p<0.0005
Statistical Significance
p<0.0005
Want to read the original?
Access the complete publication on the publisher's website