Early diagnosis of cardiac decompensation is essential to improve therapy of heart failure as one of the most cost-intensive chronic diseases worldwide. This can be achieved by remote monitoring of pulmonary artery pressure. An implantable pressure sensor system is developed in a joint research project (COMPASS), funded by the BMBF. The application of pulse contour analysis, which is well established for stroke volume estimation based on aortic blood pressure, to pulmonary artery pressure (PAP) is challenging due to different signal morphology and limited signal resolution. Especially the incisure in the pressure signal which marks valve closure is often not visible. Therefore, it was investigated whether valve closure can be detected by additional use of heart sounds more reliably. Blood pressure of an anesthetized Rhoen sheep was measured in the pulmonary artery with the implantable pressure monitor COMPASS and, for reference, additionally with commercial pressure catheters (Millar Instruments) in the pulmonary artery, aorta, left and right ventricles. Heart sounds were detected with a 3D accelerometer fixed at the animal's chest. For reference, valve opening and closure were derived from left ventricular pressure. With PAP only, 6.3% of all cardiac cycles were unusable for PCA at low and medium heart rates, as either valve opening or closure were missed or not correctly identified. At very high heart rates the loss rate increased up to 87%. When both PAP and accelerometer data were used, the loss rate was reduced to 0.1%. 0.4% for all heart rates. Combining blood pressure analysis with heart sounds improves the detection of valve opening and closure and thus SV estimation, especially at high heart rates.

We developed a rule-based data filter for the automatic interpretation of data transmitted from implantable cardioverter defibrillators (ICDs). The feasibility and user acceptability of the data filter were tested in a multicentre study. Fifteen European centres analysed 10 cases each. The cases represented ICD follow-up findings, e.g. new tachycardia, battery depletion or sensing defects. The mean follow-up period was 68 days (SD 35). A questionnaire was used to collect information regarding the functionality and general concept of automatic data interpretation. A score of five or above (range 1-9) was classified as acceptable. According to the questionnaires, there was a high degree of satisfaction with the general concept of automatic data interpretation (mean 6.7, SD 1.2) and with user guidance (mean 7.1, SD 0.8). Safety (mean 7.0, SD 1.4) and accuracy (mean 6.7, SD 1.4) of the evaluation of device-related and clinical problems were regarded as high. Support in daily routine was considered to be high (mean 7.3, SD 1.1) as the system was easy to understand (mean 7.5, SD 0.9). The results indicated a high user acceptance with easy system handling.