Cluster Analysis in Nursing Research: An Introduction, Historical Perspective, and Future Directions

Dunn, Heather; Quinn, Lauretta; Corbridge, Susan J.; Eldeirawi, Kamal; Kapella, Mary C.; Collins, Eileen G.

doi:10.1177/0193945917707705

Cited by 43 publications

(40 citation statements)

References 35 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In 11 of 12 studies, a cluster analysis was used; Norton and colleagues used a latent profile analysis. Cluster analysis and latent profile analysis are both analytical techniques that sort cases into several groups based on a predetermined set of variables such that there is maximal within‐group similarity and maximal between‐group difference (Dunn et al, 2018). Table 2 summarizes software and analysis techniques.…”

Section: Resultsmentioning

confidence: 99%

A systematic review of illness representation clusters in chronic conditions

Rivera

Corte

DeVon

et al. 2020

Research in Nursing & Health

Self Cite

View full text Add to dashboard Cite

A person's beliefs about their chronic condition (illness representations) influence health and treatment outcomes. Recently, researchers have used clustering approaches to identify subgroups with different patterns of beliefs about their illness, with some subgroups having more favorable health outcomes than others. To date, these findings have not been synthesized. The purpose of this systematic review of the literature was to synthesize results of studies that used clustering approaches to analyze illness representation in chronic disease populations, in order to characterize the clusters and their relationship to health outcomes. Using Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines we searched CINAHL, PsycInfo, and PubMed. To be included, studies had to be (a) peer reviewed, (b) in English, (c) performing a cluster analysis (CA), latent class analysis (LCA), or latent profile analysis (LPA), (d) using only illness representation (IR) subscales to form clusters, (e) measuring illness representation with the Illness Perception Questionnaire (IPQ-R), (f) in a chronic condition sample, and (g) measuring health-related outcomes. Twelve studies were included. Across studies, the number of clusters found ranged from two to three. In all studies, an association was found between illness representation group and at least one of their health outcomes. Illness representation clusters associated with favorable outcomes usually included lower disease-related consequences, fewer symptoms, less negative emotion, and a more stable disease pattern. The results of this review indicate that the relationship between the patterns of the illness representation profiles and health outcomes transcend diseases. Additionally, some dimensions of illness representation may be more important drivers of group membership than others. K E Y W O R D S chronic condition, cluster analysis, illness perception, illness representation, systematic review

show abstract

Section: Resultsmentioning

confidence: 99%

A systematic review of illness representation clusters in chronic conditions

Rivera

Corte

DeVon

et al. 2020

Research in Nursing & Health

Self Cite

View full text Add to dashboard Cite

show abstract

“…Cluster analysis. There are various methodological approaches to identifying clinically important subgroups and one method is to identify clusters of characteristics that differentiate people using Cluster analysis 30,31 . Cluster analysis was applied using SPSS's two-step clustering algorithm 31 .…”

Section: Discussionmentioning

confidence: 99%

Fatigue Profiles in Patients with Multiple Sclerosis are Based on Severity of Fatigue and not on Dimensions of Fatigue

Beckerman

Eijssen

Meeteren

et al. 2020

Sci Rep

View full text Add to dashboard Cite

Fatigue related to Multiple Sclerosis (MS) is considered a multidimensional symptom, manifesting in several dimensions such as physical, cognitive, and psychosocial fatigue. This study investigated in 264 patients with severe primary MS-related fatigue (median MS duration 6.8 years, mean age 48.1 years, 75% women) whether subgroups can be distinguished based on these dimensions. Subsequently, we tested whether MS-related fatigue consists of a single common unidimensional factor. Subscale scores on four self-reported fatigue questionnaires, including the Checklist of Individual Strength, the Modified Fatigue Impact Scale, the Fatigue Severity Scale and the SF36 vitality, were used in a cluster analysis to identify patients with similar fatigue characteristics. Next, all 54 items were included in exploratory factor analysis to test unidimensionality. Study results show that in patients with a treatment indication for primary MS-related fatigue, fatigue profiles are based on severity and not on the various dimensions of fatigue. The three profiles found, suggested one underlying fatigue dimension, but this could not be confirmed. Factor analysis of all 54 items resulted in 8 factors, confirming the multidimensional construct of the included fatigue questionnaires. Multiple sclerosis-related fatigue is a disabling symptom that affects most patients during the neuroprogressive course of the disease 1-4. MS-related fatigue is a puzzling interplay between multiple genotypic and phenotypic factors 5,6. Fatigue can be caused either directly by the MS disease process in the central nervous system (i.e. primary fatigue), initiated by inflammatory processes associated with immune activation, demyelination, axonal loss or neuroendocrine disturbance, or indirectly by other problems (i.e. secondary fatigue) such as insomnia, sleep disturbance due to urge incontinence or spasticity, acute infections, thyroid disorders, physical inactivity and deconditioning, or depression. In all cases, a correct differential diagnosis is required for proper clinical decision making. There are many definitions of fatigue (see Supplementary Information 1 for references regarding Fatigue Definitions) and over 250 ways to measure fatigue 7. Examples of frequently used definitions of MS-related fatigue include: the reduction in performance following either prolonged or unusual exertion, together with feelings of sensory, motor, cognitive or subjective fatigue 8 ; a subjective lack of physical and/or mental energy that is perceived by the individual or caregiver to interfere with usual and desired activities 9 ; the perception of decreased mental or physical energy that may restrict routines of daily activities 10 ; the failure to initiate and/or sustain attentional tasks (mental or cognitive fatigue) and physical activities (physical fatigue) 5. Dittner et al. 11 recognized the wide range of fatigue definitions as a so-called Catch-22 situation: "Before a concept can be measured, it must be defined, and before a definition can be agreed, there must ...

show abstract

“…To classify homogeneous groups for the expression of disgust, hierarchical cluster analysis using a between-group linkage method with Euclidean distance was utilized. 18 Values of P < 0.05 were considered statistically significant. All statistical analyses were based on two-tailed probabilities.…”

Section: Discussionmentioning

confidence: 99%

Characteristics of facial muscle activity during voluntary facial expressions: Imaging analysis of facial expressions based on myogenic potential data

Kuramoto

Yoshinaga

Nakao

et al. 2019

Neuropsychopharm Rep

View full text Add to dashboard Cite

Purpose Facial expressions are formed by the coordination of facial muscles and reflect changes in emotion. Nurses observe facial expressions as way of understanding patients. This study conducted basic research using facial myogenic potential topography to visually determine changes in the location and strength of facial muscle activity associated with voluntary facial expression to examine relationships with facial expressions. Methods Participants comprised 18 healthy adults (6 men, 12 women; mean age, 24.3 ± 4.3 years). Facial myogenic potentials were measured from 19 electrodes arranged concentrically on the face, and topographic analysis was conducted. Using potential changes and topograms, the muscle activity associated with nonvoluntary facial expression and voluntary facial expressions of happiness and disgust were classified according to the characteristics of expressions. To classify homogeneous groups among the reaction of disgust, hierarchical cluster analysis was utilized. Results One characteristic of the facial expression of happiness was activity in areas including the greater zygomatic muscle. With the facial expression of disgust, characteristic changes were seen in areas including the corrugator supercilii. Cluster analysis of the expression of disgust showed four homogeneous subgroups. Conclusion With facial myogenic potential topography, facial expressions can be evaluated objectively without being influenced by face shape or countenance. Color changes in topograms showed subtle changes in expressions that could not be supplemented with statistical processing alone, and these were useful in identifying individuality. Topography is thus expected to be utilized to supplement basic knowledge of facial expressions for a better understanding of patients.

show abstract

Cluster Analysis in Nursing Research: An Introduction, Historical Perspective, and Future Directions

Cited by 43 publications

References 35 publications

A systematic review of illness representation clusters in chronic conditions

A systematic review of illness representation clusters in chronic conditions

Fatigue Profiles in Patients with Multiple Sclerosis are Based on Severity of Fatigue and not on Dimensions of Fatigue

Characteristics of facial muscle activity during voluntary facial expressions: Imaging analysis of facial expressions based on myogenic potential data

Contact Info

Product

Resources

About