Association between routinely reported symptoms and 3 month hospital admission and mortality risk: a landmark analysis investigation in 86,000 individuals with heart failure in the UK

Objectives In patients with heart failure (HF), investigate the association between symptoms recorded in routine primary care consultations and short-term (3-month) hospitalisation and mortality. Design Landmark analysis using Royston-Parmar flexible parametric survival models fitted at three different timepoints (landmarks): baseline (diagnosis date), and 6 and 12 month post-diagnosis. Setting Primary care database analysis using Clinical Practice Research Datalink (CPRD), linked to hospitalisations and mortality (1998 to 2020). Participants Adults (>40 years) with a first code for HF diagnosis in primary or secondary care records. Exposures Shortness of breath (SOB), ankle swelling, oedema, fatigue, chest pain, depression and anxiety measured in a 3-month time-window prior to HF diagnosis and in survivors at 6 and 12-months. Main outcome measures 3-month first all-cause hospitalisation and mortality, separately. Secondary outcomes were hospitalisations for HF and non-cardiovascular disease. Results There were 86,882 patients with a diagnosis of HF, of whom 62,742 and 54,555 survived 6 and 12 months, respectively. There were differential symptom associations among different landmarks and outcomes. The highest risk associations for symptoms recorded just prior to diagnosis (baseline) were for depression for all-cause hospitalisation (adjusted hazard ratio: 1.26; 95% CI 1.15, 1.39), for non-cardiovascular hospitalisation (1.15; 1.10, 1.21) and death (1.22; 1.09, 1.36) and SOB for HF hospitalisation (1.18; 1.12, 1.26). Patients with HF exhibiting symptoms just prior to 6 and 12 month landmarks had increased risk of hospitalisation and mortality: at 6-months, the highest risk associations were with symptoms of depression for all-cause hospitalisation (1.46; 1.25, 1.70) and non-CVD hospitalisation (1.46; 1.23, 1.74), SOB for HF-specific hospitalisation (1.72; 1.50, 1.98); ankle swelling for mortality (1.49; 1.14, 1.94). At the 12-month landmark, classic HF symptoms had the highest risk associations: ankle swelling for all-cause hospitalisation (1.47; 1.21, 1.79) and non-CVD hospitalisation (1.55; 1.24, 1.93); SOB for HF hospitalisation (1.99; 1.68, 2.35); and fatigue for mortality (1.57; 1.22, 2.03). Conclusions Symptoms reported by patients with HF in routine primary care consultations were associated with short-term risk of hospitalisations and death. These risk associations were more prominent in patients with HF at 6 and 12-months post-diagnosis than at the point of diagnosis.