Cluster analysis of symptom occurrence to identify subgroups of heart failure patients

A pilot study

Melody A. Hertzog, Bunny Pozehl, Kathleen Duncan

Research output: Contribution to journalArticle

29 Citations (Scopus)

Abstract

Background and Research Objective: The aim of this small-scale study was to explore the use of cluster analysis to identify subgroups of heart failure patients whose patterns of symptoms may help guide clinical management. The empirically derived clusters were compared on (1) demographics, (2) clinical characteristics, and (3) subscales of the Kansas City Cardiomyopathy Questionnaire. Subjects and Methods: A demographics questionnaire, the Kansas City Cardiomyopathy Questionnaire, and the investigator-developed Heart Failure Symptom Survey were mailed to a random sample of 300 patients at a Midwestern outpatient heart failure clinic. Results and Conclusions: Of 139 respondents, 33 (24%) were female and 106 (76%) were male. The mean (SD) age was 70.6 (9.7) years, and all were white, except for a single African American female. Most subjects were married (84%) with a median level of high school education, and 5% were New York Heart Association classification I, 38% class II, 52% class III, and 5% class IV. Hierarchical cluster analysis was used to derive a 3-cluster solution based on the presence or absence of 14 symptoms. Cluster 1 patients had significantly lower incidence of symptoms and were more likely to be New York Heart Association class I or class II, with lower body mass index and higher education levels compared with patients in the other clusters. Both clusters 2 and 3 were more symptomatic than cluster 1. Compared with cluster 3, patients in cluster 2 reported more shortness of breath under circumstances other than activity, and the majority reported difficulty sleeping. They also tended to report greater symptom severity and impact on physical activity and enjoyment of life. Additional differences included comorbidities and percentage of subjects on angiotensin-converting enzyme inhibitors or angiotensin receptor blockers. Examination of the clusters suggested clinical implications related to pharmacological management and raised questions concerning potential influences of duration of the heart failure condition, presence of sleep-disordered breathing, and impact of educational level on self-management behavior and symptom patterns.

Original languageEnglish (US)
Pages (from-to)273-283
Number of pages11
JournalJournal of Cardiovascular Nursing
Volume25
Issue number4
DOIs
StatePublished - Jul 1 2010

Fingerprint

Cluster Analysis
Heart Failure
Cardiomyopathies
Demography
Education
Angiotensin Receptor Antagonists
Sleep Apnea Syndromes
Self Care
Angiotensin-Converting Enzyme Inhibitors
African Americans
Dyspnea
Comorbidity
Body Mass Index
Outpatients
Research Personnel
Surveys and Questionnaires
Pharmacology
Exercise
Incidence
Research

Keywords

  • cluster analysis
  • heart failure
  • signs and symptoms

ASJC Scopus subject areas

  • Cardiology and Cardiovascular Medicine
  • Advanced and Specialized Nursing

Cite this

Cluster analysis of symptom occurrence to identify subgroups of heart failure patients : A pilot study. / Hertzog, Melody A.; Pozehl, Bunny; Duncan, Kathleen.

In: Journal of Cardiovascular Nursing, Vol. 25, No. 4, 01.07.2010, p. 273-283.

Research output: Contribution to journalArticle

@article{7fa4b460213f49aea78f31018c9039a5,
title = "Cluster analysis of symptom occurrence to identify subgroups of heart failure patients: A pilot study",
abstract = "Background and Research Objective: The aim of this small-scale study was to explore the use of cluster analysis to identify subgroups of heart failure patients whose patterns of symptoms may help guide clinical management. The empirically derived clusters were compared on (1) demographics, (2) clinical characteristics, and (3) subscales of the Kansas City Cardiomyopathy Questionnaire. Subjects and Methods: A demographics questionnaire, the Kansas City Cardiomyopathy Questionnaire, and the investigator-developed Heart Failure Symptom Survey were mailed to a random sample of 300 patients at a Midwestern outpatient heart failure clinic. Results and Conclusions: Of 139 respondents, 33 (24{\%}) were female and 106 (76{\%}) were male. The mean (SD) age was 70.6 (9.7) years, and all were white, except for a single African American female. Most subjects were married (84{\%}) with a median level of high school education, and 5{\%} were New York Heart Association classification I, 38{\%} class II, 52{\%} class III, and 5{\%} class IV. Hierarchical cluster analysis was used to derive a 3-cluster solution based on the presence or absence of 14 symptoms. Cluster 1 patients had significantly lower incidence of symptoms and were more likely to be New York Heart Association class I or class II, with lower body mass index and higher education levels compared with patients in the other clusters. Both clusters 2 and 3 were more symptomatic than cluster 1. Compared with cluster 3, patients in cluster 2 reported more shortness of breath under circumstances other than activity, and the majority reported difficulty sleeping. They also tended to report greater symptom severity and impact on physical activity and enjoyment of life. Additional differences included comorbidities and percentage of subjects on angiotensin-converting enzyme inhibitors or angiotensin receptor blockers. Examination of the clusters suggested clinical implications related to pharmacological management and raised questions concerning potential influences of duration of the heart failure condition, presence of sleep-disordered breathing, and impact of educational level on self-management behavior and symptom patterns.",
keywords = "cluster analysis, heart failure, signs and symptoms",
author = "Hertzog, {Melody A.} and Bunny Pozehl and Kathleen Duncan",
year = "2010",
month = "7",
day = "1",
doi = "10.1097/JCN.0b013e3181cfbb6c",
language = "English (US)",
volume = "25",
pages = "273--283",
journal = "Journal of Cardiovascular Nursing",
issn = "0889-4655",
publisher = "Lippincott Williams and Wilkins",
number = "4",

}

TY - JOUR

T1 - Cluster analysis of symptom occurrence to identify subgroups of heart failure patients

T2 - A pilot study

AU - Hertzog, Melody A.

AU - Pozehl, Bunny

AU - Duncan, Kathleen

PY - 2010/7/1

Y1 - 2010/7/1

N2 - Background and Research Objective: The aim of this small-scale study was to explore the use of cluster analysis to identify subgroups of heart failure patients whose patterns of symptoms may help guide clinical management. The empirically derived clusters were compared on (1) demographics, (2) clinical characteristics, and (3) subscales of the Kansas City Cardiomyopathy Questionnaire. Subjects and Methods: A demographics questionnaire, the Kansas City Cardiomyopathy Questionnaire, and the investigator-developed Heart Failure Symptom Survey were mailed to a random sample of 300 patients at a Midwestern outpatient heart failure clinic. Results and Conclusions: Of 139 respondents, 33 (24%) were female and 106 (76%) were male. The mean (SD) age was 70.6 (9.7) years, and all were white, except for a single African American female. Most subjects were married (84%) with a median level of high school education, and 5% were New York Heart Association classification I, 38% class II, 52% class III, and 5% class IV. Hierarchical cluster analysis was used to derive a 3-cluster solution based on the presence or absence of 14 symptoms. Cluster 1 patients had significantly lower incidence of symptoms and were more likely to be New York Heart Association class I or class II, with lower body mass index and higher education levels compared with patients in the other clusters. Both clusters 2 and 3 were more symptomatic than cluster 1. Compared with cluster 3, patients in cluster 2 reported more shortness of breath under circumstances other than activity, and the majority reported difficulty sleeping. They also tended to report greater symptom severity and impact on physical activity and enjoyment of life. Additional differences included comorbidities and percentage of subjects on angiotensin-converting enzyme inhibitors or angiotensin receptor blockers. Examination of the clusters suggested clinical implications related to pharmacological management and raised questions concerning potential influences of duration of the heart failure condition, presence of sleep-disordered breathing, and impact of educational level on self-management behavior and symptom patterns.

AB - Background and Research Objective: The aim of this small-scale study was to explore the use of cluster analysis to identify subgroups of heart failure patients whose patterns of symptoms may help guide clinical management. The empirically derived clusters were compared on (1) demographics, (2) clinical characteristics, and (3) subscales of the Kansas City Cardiomyopathy Questionnaire. Subjects and Methods: A demographics questionnaire, the Kansas City Cardiomyopathy Questionnaire, and the investigator-developed Heart Failure Symptom Survey were mailed to a random sample of 300 patients at a Midwestern outpatient heart failure clinic. Results and Conclusions: Of 139 respondents, 33 (24%) were female and 106 (76%) were male. The mean (SD) age was 70.6 (9.7) years, and all were white, except for a single African American female. Most subjects were married (84%) with a median level of high school education, and 5% were New York Heart Association classification I, 38% class II, 52% class III, and 5% class IV. Hierarchical cluster analysis was used to derive a 3-cluster solution based on the presence or absence of 14 symptoms. Cluster 1 patients had significantly lower incidence of symptoms and were more likely to be New York Heart Association class I or class II, with lower body mass index and higher education levels compared with patients in the other clusters. Both clusters 2 and 3 were more symptomatic than cluster 1. Compared with cluster 3, patients in cluster 2 reported more shortness of breath under circumstances other than activity, and the majority reported difficulty sleeping. They also tended to report greater symptom severity and impact on physical activity and enjoyment of life. Additional differences included comorbidities and percentage of subjects on angiotensin-converting enzyme inhibitors or angiotensin receptor blockers. Examination of the clusters suggested clinical implications related to pharmacological management and raised questions concerning potential influences of duration of the heart failure condition, presence of sleep-disordered breathing, and impact of educational level on self-management behavior and symptom patterns.

KW - cluster analysis

KW - heart failure

KW - signs and symptoms

UR - http://www.scopus.com/inward/record.url?scp=77954005790&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=77954005790&partnerID=8YFLogxK

U2 - 10.1097/JCN.0b013e3181cfbb6c

DO - 10.1097/JCN.0b013e3181cfbb6c

M3 - Article

VL - 25

SP - 273

EP - 283

JO - Journal of Cardiovascular Nursing

JF - Journal of Cardiovascular Nursing

SN - 0889-4655

IS - 4

ER -