Using Latent Class Analysis to Identify Different Clinical Profiles Among Patients With Advanced Heart Failure

Abstract

Context: Although palliative care is guideline-indicated for patients with advanced heart failure (HF), the scarcity of a specialty-trained palliative care workforce demands better identification of patients who are most burdened by the disease.

Objectives: We sought to identify latent subgroups with variations regarding symptom burden, functional status, and multimorbidity in an advanced HF population.

Methods: We performed a latent class analysis (LCA) of baseline data from a trial enrolling advanced HF patients. As LCA input variables, we chose indicators of HF severity, physical and psychological symptom burden, functional status, and the number of comorbidities.

Results: Among 563 patients, two subgroups emerged from LCA, Class A (352 [62.5%]) and Class B (211 [37.5%]). Patients in Class A were less often classified as NYHA class III or IV (88.0% vs. 97.5%, P < 0.001), as compared to Class B patients. Class A patients had fewer symptoms, fewer comorbidities, only 25.9% had impairments in activities of daily living (ADL), and virtually none suffered from clinically significant anxiety (0.4%) or depression (0.9%). In Class B, every patient reported more than three symptoms, almost all patients (92.6%) had some impairment in ADL, and nearly a third had anxiety (30.2%) or depression (28.3%). All-cause mortality after 12 months was higher in Class B, as compared to Class A (18.5% vs. 12.5%, P = 0.047).

Conclusion: Among advanced HF patients, we identified a distinct subgroup characterized by a conjunction of high symptom burden, anxiety, depression, multimorbidity, and functional status impairment, which might profit particularly from palliative care interventions. J Pain Symptom Manage 2022;000:1-9.

Keywords: Advanced heart failure; activities of daily living; anxiety; depression; functional status; latent class analysis; multimorbidity; symptom burden.

Fig. 1.

Characteristics of latent subgroups. Shown are baseline characteristics in percent stratified by latent class for each LCA model construction variable. Asterisks indicate significant difference (P < 0.05) between latent classes. Anxiety and depression are based on the Hospital Anxiety and Depression Scale. ADL, activities of daily living; HF hosp., heart failure hospitalizations within the last 12 months; ICD, implantable cardioverter-defibrillator; LCA, latent class analysis; NYHA, New York Heart Association class. For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.

Fig. 2.

Outcomes of interest during 12-month follow-up. Kaplan-Meier-curves for all-cause mortality (Panel A) and hospitalization (Panel B) during 12-month follow-up. Subgroup differences were compared using the log-rank test. For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.