Prediction models for burden of caregivers applying data mining techniques

Sunmoo Yoon; Maria Patrao; Debbie Schauer; Jose Gutierrez

doi:10.3934/bdia.2017014

Article Contents

2017, Volume 0. Doi: 10.3934/bdia.2017014

Previous Article Next Article

Early Access

Early Access articles are published articles within a journal that have not yet been assigned to a formal issue. This means they do not yet have a volume number, issue number, or page numbers assigned to them, however, they can still be found and cited using their DOI (Digital Object Identifier). Early Access publication benefits the research community by making new scientific discoveries known as quickly as possible.

Readers can access Early Access articles via the “Early Access” tab for the selected journal.

Prediction models for burden of caregivers applying data mining techniques

1.
School of Nursing, Columbia University Medical Center, New York, NY, 10032, USA
2.
Department of Nursing, Columbia University Medical Center, New York, NY, 10032, USA
3.
Department of Neurology, Columbia University Medical Center, New York, NY, 10032, USA

* Corresponding author: Sunmoo Yoon, RN, PhD, Associate Research Scientist, Columbia University, sy2102@columbia.edu
* Corresponding author: Sunmoo Yoon, RN, PhD, Associate Research Scientist, Columbia University, sy2102@columbia.edu

Early access: October 2017

Abstract Full Text(HTML) Figure(2) / Table(2) Related Papers Cited by

Abstract

Introduction

Caregiver stress negatively influences both patients and caregivers. Predictors of caregiver difficulty may provide crucial insights for providers to prioritize those with the highest risk of stress. The purpose of this study was to develop a prediction model of caregiver difficulty by applying data mining techniques to a national behavioral risk factor data set.

Methods

Behavioral data including 397 variables on 2,264 informal caregivers, who provided any care to a friend or family member during the past month, were extracted from a publicly available national dataset in the U.S (N = 451,075) and analyzed. We applied several classification algorithms (J48, RandomForest, MultilayerPerceptron, AdaboostM1), to iteratively generate prediction models for caregiving difficulty with 10-fold cross validation.

Results

44.7% of informal caregivers answered that they faced the greatest difficulties while they took care of patients. Among those who faced the greatest difficulties, the reasons were creating emotional burden (45%). Patient cognitive alteration (e.g. cognitive changes in thinking or remembering during the past year), care hours, and relationship with a caregiver appeared as the main predictors of caregiver stress (classified correctly 63%, difficulty AUC = 65%, no difficulty AUC = 65%).

Conclusions

Data mining methods were useful to discover new behavioral risk knowledge and to visualize predictors of caregiver stress from a multidimensional behavioral dataset.This study suggests that health professionals target dementia family caregivers who are anticipated to experience patients' neuro-cognitive changes, and inform the caregivers about importance of limiting care hours, burn out and delegation of caregiving tasks.

Keywords:

Mathematics Subject Classification: Primary:97R50;Secondary:97R71.

Citation:

FullText(HTML)

Figure 1. Iterativesteps of the data mining process to build a prediction model from a large dataset

Download: Full-size image PowerPoint slide

Figure 2. Burden of caregivers

Download: Full-size image PowerPoint slide

Table 1. Characteristics of Caregivers (n=2,264)

Patient age (mean, SD)	69.87	20.53
Caregiver age (mean, SD)	56.14	15.46
Race/Ethnicity
White	2,049	90.50%
Black	61	2.69%
Hispanic	69	3.05%
Others	56	2.47%
Patient Gender
Male	795	35.11%
Female	1,455	64.27%
Employment
Employed for wages	1,035	45.72%
Self-employed	220	9.72%
Unemployed	423	18.69%
Retired	582	25.71%
Income
< ＄35,000	577	25.49%
< ＄50,000	299	13.21%
< ＄75,000	344	15.19%
≥＄75,000	734	32.42%
Relationship
(Grand) Parents	915	40.41%
Spouse	371	16.39%
Child, sibling, relatives	504	22.26%
Friends	451	19.92%
Patient status
Cognitive changes	1,156	51.06%
No cognitive changes	1,038	45.85%
Not sure	29	1.28%

| Show Table

DownLoad: CSV

Table 2. Characteristics of Caregivers -Cont'd (n=2,264)

Caregiving duration
≤ 1 year	769	33.97%
≤ 5 years	907	40.06%
> 5 years	497	21.95%
Caregiving frequency
≤ 10 hours/week	1,344	59.36%
≤ 30 hours/week	380	16.78%
≤ 100 hours/week	201	8.88%
> 100 hours/week	92	4.06%
Most needs
Cleaning, managing ＄, prepare meals	614	27.12%
Transportation outside of the home	503	22.22%
Something else	317	14.00%
Self care -eating, dressing, bathing	302	13.34%
Relieving anxiety or depression	184	8.13%
Caregiving difficulties
No difficulty	1,013	54.0%
Difficulty	1,178	44.7%
Not sure/ Don't know	28	1.25%
Refused	24	1.07%
Greatest difficulties having difficulties
Creates emotional burden	528	44.82%
Not enough time for yourself	165	14.01%
Other difficulty	113	9.59%
Creates financial burden	95	8.06%
Affects family relationships	85	7.22%
No enough time for your family	84	7.13%
Interferes with your work	71	6.03%
Aggravates health problems	37	3.14%

| Show Table

DownLoad: CSV

Related Papers

Cited by

References

[1]	R. D. Adelman, L. L. Tmanova, D. Delgado, S. Dion and M. S. Lachs, Caregiver burden: A clinical review, Jama, 311 (2014), 1052-1060. doi: 10.1001/jama.2014.304.
[2]	A. Barfar and B. Padmanabhan, Predicting presidential election outcomes from what people watch, Big Data, 5 (2017), 32-41. doi: 10.1089/big.2017.0013.
[3]	C. M. Bishop, Neural Networks for Pattern Recognition, Oxford university press, 1995.
[4]	L. Breiman, Random forests, Machine Learning, 45 (2001), 5-32.
[5]	C.-Y. Chiao, H.-S. Wu and C.-Y. Hsiao, Caregiver burden for informal caregivers of patients with dementia: A systematic review, International Nursing Review, 62 (2015), 340-350. doi: 10.1111/inr.12194.
[6]	G. DePalma, H. Xu, K. E. Covinsky, B. A. Craig, E. Stallard, J. Thomas Ⅲ and L. P. Sands, Hospital readmission among older adults who return home with unmet need for ADL disability, The Gerontologist, 53 (2013), 454-461. doi: 10.1093/geront/gns103.
[7]	C. for Disease Control and Prevention, Behavioral risk factor surveillance system survey data, atlanta, georgia. u. s.
[8]	J. E. Gaugler, D. L. Roth, W. E. Haley and M. S. Mittelman, Can counseling and support reduce burden and depressive symptoms in caregivers of people with Alzheimer's disease during the transition to institutionalization? results from the new york university caregiver intervention study, Journal of the American Geriatrics Society, 56 (2008), 421-428. doi: 10.1111/j.1532-5415.2007.01593.x.
[9]	P. E. Greenberg and H. G. Birnbaum, The economic burden of depression in the us: Societal and patient perspectives, Expert Opinion on Pharmacotherapy, 6 (2005), 369-376. doi: 10.1517/14656566.6.3.369.
[10]	S. Gupta, G. Hawker, A. Laporte, R. Croxford and P. Coyte, The economic burden of disabling hip and knee osteoarthritis (oa) from the perspective of individuals living with this condition, Rheumatology, 44 (2005), 1531-1537. doi: 10.1093/rheumatology/kei049.
[11]	M. Hall, E. Frank, G. Holmes, B. Pfahringer, P. Reutemann and I. H. Witten, The weka data mining software: An update, ACM SIGKDD Explorations Newsletter, 11 (2009), 10-18. doi: 10.1145/1656274.1656278.
[12]	Y. LeCun, Y. Bengio and G. Hinton, Deep learning, Nature, 521 (2015), 436-444. doi: 10.1038/nature14539.
[13]	S. J. Lupien, B. S. McEwen, M. R. Gunnar and C. Heim, Effects of stress throughout the lifespan on the brain, behaviour and cognition, Nature Reviews Neuroscience, 10 (2009), 434-445. doi: 10.1038/nrn2639.
[14]	P. C. J. Navas, Y. C. G. Parra and J. I. R. Molano, Big data tools: Haddop, mongodb and weka, in International Conference on Data Mining and Big Data, Springer, 2016,449-456.
[15]	U. D. of Health Huma Service., 2011 poverty guideline, Federal Register, 76 (2010), 3637-3638.
[16]	B. D. Ripley, Pattern Recognition and Neural Networks, Cambridge university press, 2007.
[17]	J. W. Rowe, T. Fulmer and L. Fried, Preparing for better health and health care for an aging population, Jama, 316 (2016), 1643-1644. doi: 10.1001/jama.2016.12335.
[18]	J. Schmidhuber, Deep learning in neural networks: An overview, Neural Networks, 61 (2015), 85-117. doi: 10.1016/j.neunet.2014.09.003.
[19]	B. C. Spillman and S. K. Long, Does high caregiver stress predict nursing home entry?, INQUIRY: The Journal of Health Care Organization, Provision, and Financing, 46 (2009), 140-161. doi: 10.5034/inquiryjrnl_46.02.140.
[20]	C. H. Van Houtven, S. D. Ramsey, M. C. Hornbrook, A. A. Atienza and M. van Ryn, Economic burden for informal caregivers of lung and colorectal cancer patients, The Oncologist, 15 (2010), 883-893. doi: 10.1634/theoncologist.2010-0005.
[21]	I. H. Witten, E. Frank, M. A. Hall and C. J. Pal, Data Mining: Practical Machine Learning Tools and Techniques, Morgan Kaufmann, 2016.
[22]	X. Wu, V. Kumar, J. Ross Quinlan, J. Ghosh, Q. Yang, H. Motoda, G. J. McLachlan, A. Ng, B. Liu and P. S. Yu, et al., Top 10 algorithms in data mining, Knowledge and Information Systems, 14 (2008), 1-37. doi: 10.1007/s10115-007-0114-2.
[23]	E. Yan and T. Kwok, Abuse of older Chinese with dementia by family caregivers: An inquiry into the role of caregiver burden, International Journal of Geriatric Psychiatry, 26 (2011), 527-535. doi: 10.1002/gps.2561.
[24]	Q. Yang and X. Wu, 10 challenging problems in data mining research, International Journal of Information Technology & Decision Making, 5 (2006), 597-604. doi: 10.1142/S0219622006002258.