Appl Clin Inform 2016; 07(03): 711-730
DOI: 10.4338/ACI-2016-03-RA-0049
Research Article
Schattauer GmbH

Visual Analytics for Pattern Discovery in Home Care

Clinical Relevance for Quality Improvement
Kavita Radhakrishnan
1  School of Nursing, University of Texas – Austin, Austin, TX
,
Karen A. Monsen
2  University of Minnesota School of Nursing, Minneapolis, MN
,
Sung-Heui Bae
3  Ewha Womans University, College of Nursing, Seoul, Republic of Korea
,
Wenhui Zhang
1  School of Nursing, University of Texas – Austin, Austin, TX
› Author Affiliations
Funding The University of Texas Austin New Faculty Start-up Funds for Dr. Radhakrishnan.
Further Information

Correspondence to:

Kavita Radhakrishnan, PhD RN MSEE
Assistant Professor
School of Nursing
University of Texas – Austin
1710 Red River Street
Austin
TX 78701–1499
UNITED STATES
Phone: (512) 471-7936   
Fax: (512) 471 - 3688   

Publication History

received: 6. 06 April 2016

accepted: 23 June 2016

Publication Date:
19 December 2017 (online)

 

Summary

Background

Visualization can reduce the cognitive load of information, allowing users to easily interpret and assess large amounts of data. The purpose of our study was to examine home health data using visual analysis techniques to discover clinically salient associations between patient characteristics with problem-oriented health outcomes of older adult home health patients during the home health service period.

Methods

Knowledge, Behavior and Status ratings at discharge as well as change from admission to discharge that was coded using the Omaha System was collected from a dataset on 988 deidentified patient data from 15 home health agencies. SPSS Visualization Designer v1.0 was used to visually analyze patterns between independent and outcome variables using heat maps and histograms. Visualizations suggesting clinical salience were tested for significance using correlation analysis.

Results

The mean age of the patients was 80 years, with the majority female (66%). Of the 150 visualizations, 69 potentially meaningful patterns were statistically evaluated through bivariate associations, revealing 21 significant associations. Further, 14 associations between episode length and Charlson co-morbidity index mainly with urinary related diagnoses and problems remained significant after adjustment analyses. Through visual analysis, the adverse association of the longer home health episode length and higher Charlson co-morbidity index with behavior or status outcomes for patients with impaired urinary function was revealed.

Conclusions

We have demonstrated the use of visual analysis to discover novel patterns that described high-needs subgroups among the older home health patient population. The effective presentation of these data patterns can allow clinicians to identify areas of patient improvement, and time periods that are most effective for implementing home health interventions to improve patient outcomes.

Citation: Radhakrishnan K, Monsen KA, Bae S-H, Zhang W. Visual analytics for pattern discovery in home care: Clinical relevance for quality improvement.


#

 


#

Conflicts of Interest

The authors do not report any conflict of interest in conducting this study.


Correspondence to:

Kavita Radhakrishnan, PhD RN MSEE
Assistant Professor
School of Nursing
University of Texas – Austin
1710 Red River Street
Austin
TX 78701–1499
UNITED STATES
Phone: (512) 471-7936   
Fax: (512) 471 - 3688