Clustering of patients with end-stage chronic diseases by symptoms: a new approach to identify health needs.

Background: End-stage chronic obstructive pulmonary disease (COPD), chronic heart failure (CHF) and chronic renal failure (CRF) are characterized by a high burden of daily symptoms that, irrespective of the primary organ failure, are widely shared.

Aims: To evaluate whether and to which extent symptom-based clusters of patients with end-stage COPD, CHF and CRF associate with patients' health status, mobility, care dependency and life-sustaining treatment preferences.

Methods: 255 outpatients with a diagnosis of advanced COPD (n = 95), advanced CHF (n = 80) or CRF requiring dialysis (n = 80) were visited in their home environment and underwent a multidimensional assessment: clinical characteristics, symptom burden using Visual Analog Scale (VAS), health status questionnaires, timed "Up and Go" test, Care Dependency Scale and willingness to undergo mechanical ventilation or cardiopulmonary resuscitation. Three clusters were obtained applying K-means cluster analysis on symptoms' severity assessed via VAS. Cluster characteristics were compared using non-parametric tests.

Results: Cluster 1 patients, with the least symptom burden, had a better quality of life, lower care dependency and were more willing to accept life-sustaining treatments than others. Cluster 2, with a high presence and severity of dyspnea, fatigue, cough, muscle weakness and mood problems, and Cluster 3, with the highest occurrence and severity of symptoms, reported similar care dependency and life-sustaining treatment preferences, while Cluster 3 reported the worst physical health status.

Discussion: Symptom-based clusters identify patients with different health needs and might help to develop palliative care programs.

Conclusion: Clustering by symptoms identifies patients with different health status, care dependency and life-sustaining treatment preferences.