Patients' self-evaluation of symptoms, signs and compliance to therapy for heart failure surveillance: A pilot study on identification of worsening heart failure

In patients with chronic heart failure (HF), it is unclear whether self-evaluation of HF-related symptoms and signs may contribute to the surveillance of clinical stability over time or the early identification of worsening HF. Therefore, HF-related symptoms and signs, and compliance to therapy were investigated by a self-administrated questionnaire in patients hospitalized due to worsening HF (cases) as well in those with stable HF (controls). The temporal interval of interest for the questionnaire focused on 2 to 7 days before the hospitalization or the control visit. A method called "classic" extrapolated a surveillance score by 4 questions revealing HF-related symptoms or signs (higher score indicative of more symptoms or sings reported). Alternatively, 3 additional questions enriched the "classic" questionnaire, including one on whether the health status was perceived as stable, improved or worsening. The score from "perceived health status" inquiry was used as "conditional factor" to computed the surveillance score as: conditional factor + [conditional factor X (the sum of the points yielded by the responses to each question on symptoms, sings and compliance to therapy)]. Among 51 patients hospitalized due to worsening HF and 104 with chronic HF (retrospective study), the area under the curve (receiver operating characteristic) discriminating worsening HF was 0.90 by the classic, and 0.96 by the conditional method (both p<0.001), yielding error rates of 2 in 10 by the classic and 1 in 10 by the conditional method. In a pilot longitudinal study in 37 ambulatory HF patients from a different population source, predicting worsening HF by alternative questionnaires within 12 weeks from the visit yielded consistent results. In conclusion, patients' self-monitoring symptoms, signs and compliance by a standardized questionnaire using a conditional method for generating a relative score, may be a promising method for HF stability surveillance.