Primary care nurse's actions using a clinical decision-making algorithm in a remote dielectric sensing-guided heart failure management program

Background Remote Dielectric Sensing (ReDS) is a non-invasive technology that quantifies lung fluid content. It defines an optimal fluid range of 25-35% and detects levels across the spectrum from hypovolaemia to hypervolaemia. A ReDS-guided heart failure (HF) management programme used a clinical decision-making algorithm (Table 1) based on ReDS values to guide follow-up scheduling, diuretic adjustments, and symptom monitoring. Purpose The study aimed at describing a primary care nurse´s actions using an algorithm in the ReDS-guided HF management programme. Methods Data from a feasibility study of 18 patients (mean age, 81±6 years, male 47%, NYHA III/IV 67%) from one primary care centre was analysed. A total of 36 ReDS measurements were performed (baseline and at one month follow-up) by a primary care nurse. Their actions were examined through intervention notes and interview data. Results At baseline, 7 patients had ReDS values >35% (hypervolemic status), and in 6 of these cases, the values remained high at the one-month follow-up, in spite of actions taken (diuretics adjustments (n =1), telephone follow up after 1.5 weeks (n =1), educational support for weight monitoring (n =1) and referral to a HF clinic (n =1). There were also cases where no action was taken (n =4)). One patient showed a reduction in ReDS values from 38% to an acceptable 35% through nurses’ consultation with the physician and a telephone follow up after 1.5 weeks. Three patients with initially optimal ReDS values exhibited ReDS values above the 35% threshold at the one-month follow-up, and no additional action was reported. Nurse´s actions were categorized as either algorithm-recommended or additional actions and are summarised in Table 2. In all, 16 measurements indicated excess fluid status >35% and full adherence to the algorithm's recommended action was reported in one of these cases. In the remaining cases, actions either partially followed the algorithm or were aligned with recommendations for lower ReDS severity. Scheduled telephone follow-ups replaced weekly follow-up including ReDS measurements. In some instances, no action was taken due to patients appearing to be clinically well. Other reasons for taking alternative actions included uncertainty in interpreting ReDS values, partly because the clinical meaning was unclear especially when high values indicating congestion were not confirmed by other clinical findings. This uncertainty contributed to reduced adherence to the decision-support algorithm, especially when it recommended additional visits or increased diuretic doses. A need for further education was also noted to strengthen confidence in interpreting ReDS values. Conclusions The algorithm recommended actions for ReDS values were not always implemented as the nurse relied more on conventional clinical judgement. Further support in understanding ReDS technology and improvement of the algorithm is necessary to enhance the implementation.