Skip to main content

Table 4 Results of the second stage of two-stage clustering (described using the prevalence of first stage cluster membership)

From: Could the clinical interpretability of subgroups detected using clustering methods be improved by using a novel two-stage approach?

Domain   Cluster 1 Cluster 2
Subgroup label   Uncertain of cause and fearful, but not heart-related - with local thoracic 5/6 palpation findings. Believes pain to be heart- or musculoskelet-al-related with crushing pain, and local thoracic 2/3 signs and pectoral tenderness.
Cluster size   51 % 49 %
Cluster membership probability Median (interquartile range) 100 % (99 % to 100 %) 100 % (98 % to 100 %)
Demographic Younger 42 %* 62 %
  Older 58 % 38 %
Previous history   0 % 0 %
  Only one cluster 100 % 100 %
Psychological Uncertain of cause of pain and fearful 88 % 2 %
  Believes cause from heart 1 % 44 %
  Believes both from heart and musculoskeletal 0 % 31 %
  Believes cause from musculoskeletal and less fearful 6 % 14 %
  Believes cause from other cause and less fearful 5 % 8 %
Pain Crushing pain 74 % 89 %
  Not crushing pain 26 % 11 %
Activity limitation None or light activity limitation 88 % 76 %
  Moderate activity limitation 8 % 13 %
  Severe activity limitation 4 % 11 %
Diagnostic
classification
Not heart 71 % 1 %
  Typical angina 13 % 61 %
  Maybe angina 16 % 38 %
Palpation No palpation findings 23 % 23 %
  Local 5/6 signs and intercostal tenderness 36 % 1 %
  Local 2/3/4 signs and pectoral tenderness 4 % 32 %
  Pectoral tenderness 3 % 30 %
  Paraspinal and pectoral tenderness 23 % 8 %
  Generalised tenderness 11 % 6 %
  1. *Proportions are those of the people within each cluster on each domain (vertical proportions), as each category within a domain is mutually exclusive