A Machine Learning Recommender System to Tailor Preference Assessments to Enhance Person-Centered Care among Nursing Home Residents

Gerald C. Gannod, Katherine M. Abbott, Kimberly Sue Van Haitsma, Nathan Martindale, Alexandra Heppner

Research output: Contribution to journalArticle

1 Citation (Scopus)

Abstract

Nursing homes (NHs) using the Preferences for Everyday Living Inventory (PELI-NH) to assess important preferences and provide person-centered care find the number of items (72) to be a barrier to using the assessment. Using a sample of n = 255 NH resident responses to the PELI-NH, we used the 16 preference items from the MDS 3.0 Section F to develop a machine learning recommender system to identify additional PELI-NH items that may be important to specific residents. Much like the Netflix recommender system, our system is based on the concept of collaborative filtering whereby insights and predictions (e.g., filters) are created using the interests and preferences of many users. The algorithm identifies multiple sets of 'you might also like' patterns called association rules, based upon responses to the 16 MDS preferences that recommends an additional set of preferences with a high likelihood of being important to a specific resident. In the evaluation of the combined apriori and logistic regression approach, we obtained a high recall performance (i.e., the ratio of correctly predicted preferences compared with all predicted preferences and nonpreferences) and high precision (i.e., the ratio of correctly predicted rules with respect to the rules predicted to be true) of 80.2% and 79.2%, respectively. The recommender system successfully provides guidance on how to best tailor the preference items asked of residents and can support preference capture in busy clinical environments, contributing to the feasibility of delivering person-centered care.

Original languageEnglish (US)
Pages (from-to)167-176
Number of pages10
JournalGerontologist
Volume59
Issue number1
DOIs
StatePublished - Jan 9 2019

Fingerprint

Nursing Homes
Logistic Models
Machine Learning
Equipment and Supplies

All Science Journal Classification (ASJC) codes

  • Gerontology
  • Geriatrics and Gerontology

Cite this

Gannod, Gerald C. ; Abbott, Katherine M. ; Van Haitsma, Kimberly Sue ; Martindale, Nathan ; Heppner, Alexandra. / A Machine Learning Recommender System to Tailor Preference Assessments to Enhance Person-Centered Care among Nursing Home Residents. In: Gerontologist. 2019 ; Vol. 59, No. 1. pp. 167-176.
@article{20af97999b7c427ab2fb6f44602dfd40,
title = "A Machine Learning Recommender System to Tailor Preference Assessments to Enhance Person-Centered Care among Nursing Home Residents",
abstract = "Nursing homes (NHs) using the Preferences for Everyday Living Inventory (PELI-NH) to assess important preferences and provide person-centered care find the number of items (72) to be a barrier to using the assessment. Using a sample of n = 255 NH resident responses to the PELI-NH, we used the 16 preference items from the MDS 3.0 Section F to develop a machine learning recommender system to identify additional PELI-NH items that may be important to specific residents. Much like the Netflix recommender system, our system is based on the concept of collaborative filtering whereby insights and predictions (e.g., filters) are created using the interests and preferences of many users. The algorithm identifies multiple sets of 'you might also like' patterns called association rules, based upon responses to the 16 MDS preferences that recommends an additional set of preferences with a high likelihood of being important to a specific resident. In the evaluation of the combined apriori and logistic regression approach, we obtained a high recall performance (i.e., the ratio of correctly predicted preferences compared with all predicted preferences and nonpreferences) and high precision (i.e., the ratio of correctly predicted rules with respect to the rules predicted to be true) of 80.2{\%} and 79.2{\%}, respectively. The recommender system successfully provides guidance on how to best tailor the preference items asked of residents and can support preference capture in busy clinical environments, contributing to the feasibility of delivering person-centered care.",
author = "Gannod, {Gerald C.} and Abbott, {Katherine M.} and {Van Haitsma}, {Kimberly Sue} and Nathan Martindale and Alexandra Heppner",
year = "2019",
month = "1",
day = "9",
doi = "10.1093/geront/gny056",
language = "English (US)",
volume = "59",
pages = "167--176",
journal = "The Gerontologist",
issn = "0016-9013",
publisher = "Oxford University Press",
number = "1",

}

A Machine Learning Recommender System to Tailor Preference Assessments to Enhance Person-Centered Care among Nursing Home Residents. / Gannod, Gerald C.; Abbott, Katherine M.; Van Haitsma, Kimberly Sue; Martindale, Nathan; Heppner, Alexandra.

In: Gerontologist, Vol. 59, No. 1, 09.01.2019, p. 167-176.

Research output: Contribution to journalArticle

TY - JOUR

T1 - A Machine Learning Recommender System to Tailor Preference Assessments to Enhance Person-Centered Care among Nursing Home Residents

AU - Gannod, Gerald C.

AU - Abbott, Katherine M.

AU - Van Haitsma, Kimberly Sue

AU - Martindale, Nathan

AU - Heppner, Alexandra

PY - 2019/1/9

Y1 - 2019/1/9

N2 - Nursing homes (NHs) using the Preferences for Everyday Living Inventory (PELI-NH) to assess important preferences and provide person-centered care find the number of items (72) to be a barrier to using the assessment. Using a sample of n = 255 NH resident responses to the PELI-NH, we used the 16 preference items from the MDS 3.0 Section F to develop a machine learning recommender system to identify additional PELI-NH items that may be important to specific residents. Much like the Netflix recommender system, our system is based on the concept of collaborative filtering whereby insights and predictions (e.g., filters) are created using the interests and preferences of many users. The algorithm identifies multiple sets of 'you might also like' patterns called association rules, based upon responses to the 16 MDS preferences that recommends an additional set of preferences with a high likelihood of being important to a specific resident. In the evaluation of the combined apriori and logistic regression approach, we obtained a high recall performance (i.e., the ratio of correctly predicted preferences compared with all predicted preferences and nonpreferences) and high precision (i.e., the ratio of correctly predicted rules with respect to the rules predicted to be true) of 80.2% and 79.2%, respectively. The recommender system successfully provides guidance on how to best tailor the preference items asked of residents and can support preference capture in busy clinical environments, contributing to the feasibility of delivering person-centered care.

AB - Nursing homes (NHs) using the Preferences for Everyday Living Inventory (PELI-NH) to assess important preferences and provide person-centered care find the number of items (72) to be a barrier to using the assessment. Using a sample of n = 255 NH resident responses to the PELI-NH, we used the 16 preference items from the MDS 3.0 Section F to develop a machine learning recommender system to identify additional PELI-NH items that may be important to specific residents. Much like the Netflix recommender system, our system is based on the concept of collaborative filtering whereby insights and predictions (e.g., filters) are created using the interests and preferences of many users. The algorithm identifies multiple sets of 'you might also like' patterns called association rules, based upon responses to the 16 MDS preferences that recommends an additional set of preferences with a high likelihood of being important to a specific resident. In the evaluation of the combined apriori and logistic regression approach, we obtained a high recall performance (i.e., the ratio of correctly predicted preferences compared with all predicted preferences and nonpreferences) and high precision (i.e., the ratio of correctly predicted rules with respect to the rules predicted to be true) of 80.2% and 79.2%, respectively. The recommender system successfully provides guidance on how to best tailor the preference items asked of residents and can support preference capture in busy clinical environments, contributing to the feasibility of delivering person-centered care.

UR - http://www.scopus.com/inward/record.url?scp=85059796260&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85059796260&partnerID=8YFLogxK

U2 - 10.1093/geront/gny056

DO - 10.1093/geront/gny056

M3 - Article

C2 - 29790930

AN - SCOPUS:85059796260

VL - 59

SP - 167

EP - 176

JO - The Gerontologist

JF - The Gerontologist

SN - 0016-9013

IS - 1

ER -