Here's a look at a study of data mining's use in patient care
Here's a look at a study of data mining's use in patient care
Approach predicted patient survival time
Data mining provided an ideal approach for predicting survival time for kidney dialysis patients, according to one study.
"We are able to predict the survival time of patients who have undergone kidney dialysis," says Andrew Kusiak, PhD, professor of industrial engineering at the University of Iowa in Iowa City, IA.
This information is useful to both clinicians and researchers, the latter because they could use the data to determine which patients would be suitable for enrollment in long-term clinical trials, Kusiak notes.
The data collected included historical information and machine-collected information, he says.
Hospitals were motivated to assist in data collection because the data analyses could help them improve patient care, but it was the data mining approach that answered a question that would have been difficult to answer otherwise, Kusiak says.
For instance, patients on dialysis machines could be impacted by multiple factors, and their survival rates may depend on relationships between demographic and clinical parameters, as well as medications, interventions, and dialysis regimen.1
Investigators collected data from known and unknown parameters at four locations of The University of Iowa Hosptials and Clinics. Then dialysis machine data was organized to summarize each of 188 patients' information.1
The patients' age, total time on dialysis, age when dialysis was started, and age at death — when applicable, also were collected. Then investigators created three decision categories of above-median, below-median, and undetermined for data mining purposes.1
The initial data mining used a rough-set algorithm with two types of decision rules, including certain and approximate. A decision-tree algorithm also was used to analyze the dialysis set, and after classifications were made, a decision-making algorithm was developed to predict outcomes of dialysis patients.1
As a result of the data mining study, investigators were able to predict survival for patients who had at least 15-20 dialysis visits. They found an enormous potential for making accurate decisions for individual patients and a classification accuracy of greater than 75 percent.1
Reference:
1. Kusiak A, et al. Predicting Survival Time for Kidney Dialysis Patients: A Data Mining Approach. Computn Biol Med. 2005;35:311-327.
Data mining provided an ideal approach for predicting survival time for kidney dialysis patients, according to one study.Subscribe Now for Access
You have reached your article limit for the month. We hope you found our articles both enjoyable and insightful. For information on new subscriptions, product trials, alternative billing arrangements or group and site discounts please call 800-688-2421. We look forward to having you as a long-term member of the Relias Media community.