Fully digital data processing during cardiovascular implantable electronic device follow-up in a high-volume tertiary center

Between November 2013 and November 2015, a total of 18,449 patients presented to the electrophysiology outpatient clinic at the department of cardiology, Medical University Hospital Heidelberg (Germany). A sample cohort of 391 patients scheduled for routine follow-up of cardiac pacemakers, ICDs, or CRT devices were prospectively enrolled. Study patients underwent CIED interrogation, followed by either manual or digital transfer (MediConnect^®; Fleischhacker, Schwerte, Germany) of device data into the electronic patient management system (i.s.h.med^®, SAP, Walldorf, Germany) installed on Windows (Microsoft, Redmond, WA, USA)—based personal computer workstations. The decision between manual and digital data processing was left to the physician’s discretion. Baseline characteristics and procedural data were analyzed. Total duration of the follow-up visit was measured from each study subject entering the examination room to the receipt of written medical reports and updated device identification card by the respective patient.

Device interrogation was performed through telemetry (Fig. 1). Specific programmers were used to retrieve programmed parameters and data stored on pacemakers, ICDs and CRT devices from the following manufacturers: Biotronik (Berlin, Germany), Boston Scientific (Marlborough, MA, USA), LivaNova (London, UK), Medtronic (Minneapolis, MN, USA), St. Jude Medical (St. Paul, MN, USA), and Vitatron (Maastricht, the Netherlands). Operating parameters of the devices were modified if required. Interrogation data were then transferred manually into a Microsoft Word-based file of the i.s.h.med^® patient management system, allowing for electronic documentation and generation of a medical report. Furthermore, the data were manually entered into a paper-based device identification card. Additional tasks were performed at the physician’s discretion (e.g., obtaining medical history or blood pressure, physical examination, auscultation).

Following device interrogation via appropriate programming devices, MediConnect^® software was employed for digital data transfer and for generation of medical documents. During the course of the study, five versions of the software were used owing to software updates: 2.6.0.130 (10/2013–3/2014), 2.6.0.138 (3/2014–8/2014), 2.7.0.10 (8/2014–10/2014), 2.7.0.20 (10/2014–7/2015), and 2.7.0.47 (7/2015–11/2015). Patients underwent CIED follow-up with the current software version at the time of follow-up. The update from 2.6.0.130 to 2.6.0.138 included minor modifications. More substantial changes were carried out during the following update to version 2.6.0.10, involving optimization of user interface and electrocardiogram viewer. Subsequent updates 2.7.0.20 and 2.7.0.47 were performed to improve stability and performance. The list of programming devices and CIEDs recognized by the software was updated with each step. Interrogation data were stored on a permanently installed USB mass storage stick connected to the programming device. The follow-up data were then imported automatically via a switch box from different programmers employed in parallel into MediConnect^® software installed on a personal computer, allowing for standardized reading and storage of data retrieved from different devices (Fig. 1). Medical reports and CIED identification documents were automatically generated by the software package, printed, and stored electronically both in MediConnect^® software and through import into the electronic patient management system (i.s.h.med^®). Thus, the integrated approach allowed for facilitated handling of interrogation data from various manufacturers. Telemedical applications [8] were not included in this work.

Statistical analyses were performed using Prism 6.0 (GraphPad, La Jolla, CA, USA) and Origin 8.0 (OriginLab Corporation, Northampton, MA, USA) software, as well as “R” software [9] (version 3.3.1) with the packages “car” [10] (version 2.1-2) and “xtable” [11] (version 1.8-2). Continuous variables are expressed as mean ± SEM. For between-group comparisons, the unpaired Student’s t test (two-tailed test) was used, assuming that the data were approximately normally distributed. Categorical variables are described as count and percentage and were analyzed using the Chi-square test or Fisher’s exact test (two-tailed tests) where appropriate. Furthermore, a linear regression model for the duration of follow-up visits was applied where study group (FDDP vs. conventional follow-up), age, gender, concomitant cardiac disease, device manufacturer, device type, and physician experience in clinical electrophysiology were included as predictors. Effect estimates, 95% confidence intervals (CI), and p values are provided in the results section. A p value < 0.05 was considered statistically significant. Since this was an exploratory analysis, no adjustment for multiple testing was applied.

A total of 391 patients presenting for ambulatory follow-up of cardiac pacemakers (n = 173; 44%), ICDs (n = 154; 39%), or CRT devices (n = 64; 16%) were included (Tables 1 and 2). The mean age was 70.3 ± 0.7 years, and 249 study subjects were male (64%). Manual transfer of device interrogation data was employed in 126 cases (32%), whereas fully digital data processing (FDDP) using MediConnect^® software was used in 265 patients (68%). The majority of patients exhibited concomitant cardiac disease (manual data transfer, 80%; FDDP, 74%), with ischemic heart disease being the most abundant (Table 1). Sinus rhythm was present in 74% (manual data transmission) and 72% of cases (FDDP). Twenty-three percent of all study subjects was presented in atrial fibrillation. Common CIED indications comprised primary and secondary preventions of sudden cardiac death, higher degree atrioventricular block, and sinus node disease (Table 1). Baseline parameters were not significantly different between groups. Pacemakers and ICDs were equally distributed between the groups (Table 2). Finally, significant intergroup differences between device manufacturers were noted (Table 2).

A total of 18 physicians follow-up were involved in the study. Of these, the majority (n = 11) applied both conventional follow-up and FDDP. Five physicians used FDDP with MediConnect^® software only, and two physicians performed CIED follow-up conventionally in all patients. In the conventional group, 85.7% (108/126) of device interrogations were conducted by investigators routinely using both methods, compared to 86.8% (230/265) in the FDDP cohort (p = 0.77). Physician’s experience in clinical electrophysiology differed between treatment groups. Physicians who performed conventional CIED follow-up displayed 3.2 ± 0.2 years of previous electrophysiological experience, whereas in the FDDP cohort, prior experience was lower (1.8 ± 0.1 years; p < 0.001).

The accuracy of data transfer increased with subsequent software updates (Table 3). Errors included general failure to import interrogation data, incomplete import of single values, insufficient transfer of interrogation values into documents created by the software, and import of incorrect values, respectively. Of note, no errors were observed with the final software update during the course of the study. MediConnect^® software crashes occurred during four of 265 follow-up investigations (1.5%). In these cases, a reboot of the system and a re-login were necessary, which required approximately 2–3 min. Two additional system reboots (0.75%) were actively initiated at the physicians’ discretion when document printing failed. Written patient reports and device identification cards produced through MediConnect^® software were free from errors in 98% of cases (Table 3). Incompletely transferred device interrogation data were manually completed prior to generation of medical documents, if required.

In the present sample cohort, the mean duration of follow-up visits using FDDP and automatic document production yielded 18.7 ± 0.4 min (n = 265) and was not significantly different from patients subjected to manual transfer of interrogation data from the programmer into electronic documentation system and into paper-based device identification cards (18.2 ± 0.8 min; n = 126; p = 0.49) (Fig. 2). While software updates improved the quality of data transfer, mean follow-up duration did not change when comparing early software versions (16.4 ± 1.0 min; n = 38) with the most recent version investigated in the present study (18.4 ± 2.2 min; n = 17; p = 0.37) (Table 4). Multivariate analysis revealed an influence of physician’s experience and device type on device follow-up duration among potential determinants (i.e., age, gender, concomitant cardiac disease, device type and manufacturer, physician’s experience in clinical electrophysiology, and conventional vs. FDDP follow-up) (Table 5). The duration of ambulatory follow-up was shorter when performed by more experienced physicians.

In addition to device interrogation, the completion of reports and CIED identification cards, ambulatory CIED follow-up visits included further physician tasks contributing to the duration and complexity of device follow-up. Obtaining patient’s history, physical examination, and consultation with other physicians were commonly performed without significant differences between study groups (Table 6). Furthermore, interrogation of cardiac pacemakers, ICDs, or CRT devices revealed technical issues such as low battery voltage and extended investigation of ICD leads under advisory in a minority of patients of both groups (Table 7). Finally, additional clinical aspects were addressed during device follow-up (Table 8). Ventricular tachycardia or fibrillation episodes detected by defibrillator devices represented the most frequent events and were encountered in 16% (manual data transfer group) and 23% of patients (FDDP; p = 0.088). In these cases, further evaluation was necessary and options for antiarrhythmic treatment optimization were discussed and initiated when appropriate.

We were able to demonstrate successful implementation of FDDP using MediConnect^® software into the clinical CIED follow-up workflow at a university hospital. Feasibility of FDDP in routine application is indicated by a 82% rate of complete and correct data transfer with the first three software versions and a 100% rate of correct data transfer with the final software version which has been used at the end of the study. Software crashes were rare (1.5%). Furthermore, medical documents were produced by the software with very few errors, and provided the added benefit of electronic documentation of both patient reports and device identification cards, respectively, eliminating the use of hardcopy archiving. In addition, removing the need for manual completion of device identification cards resulted in increased technical quality of medical documentation. A low rate of physician complaints after an initial MediConnect^® software installation and adaptation phase indicated satisfaction with FDDP performance during the course of the study.

Optimized data transfer between the programming device and the electronic documentation system could reduce overall duration of follow-up visits. Mean follow-up duration was 19 min in the present study. Contrary to expectations, there was no significant difference between the follow-up duration using FDDP and manual data transmission. Overall times of outpatient visits were determined by additional tasks that were carried out by the physician in addition to device interrogation in the setting of a specialized Electrophysiology Outpatient Clinic, associated with referral of particularly demanding cases. These actions were not affected by the method of handling device-related data. Thus, the relative contribution of efficient data transfer to follow-up times may have been too low to show significant effects. Increased physician’s experience contributed to shorter follow-up duration among patients with manual data handling as well. Furthermore, there was a tendency towards more CRT devices in the FDDP group than in the manual data handling group (18% vs. 13%). As CRT follow-up can be more time-consuming and requires the investigation of more technical parameters compared to pacemakers or ICDs, this difference may potentially have outweighed the benefits of FDDP. Finally, the duration of MediConnect^® software loading processes was noted during the course of the study as possible factor contributing to a lack of FDDP effects on total follow-up time. Future software improvements are expected to allow for exploitation of further benefits provided by FDDP use.

This study has inherent limitations due to the non-randomized patient assignment to treatment groups. A low recruiting rate resulted in a relatively low number of patients included in the study. It is important to highlight that the subpopulation analyzed here represents only a small “snapshot” of the “real life” situation at the Electrophysiology Outpatient Clinic. In addition, the study was initiated early after new implementation of fully digital data transfer into the clinical workflow. Therefore, benefits of established, long-term FDDP use, expected to result in reduced follow-up duration as device records will be readily available from the database, could not be evaluated. Telemedical applications that may provide significant further improvements in CIED patient follow-up using MediConnect^®-based digital data transfer were not included in this study. We recognize a potential bias as stratification of the results according to device manufacturers or to the respective physicians performing the follow-up investigations was not statistically feasible in the present sample. However, the open study design and resulting uneven distribution of device manufacturers between manual data handling and FDDP groups at the physicians’ discretion, respectively, may offer further insights into FDDP use. It could be speculated that some manufacturers’ devices were more likely used by the established, manual process, while others were more frequently subjected to be interrogated using the new FDDP workflow algorithm. The hypothesis that FDDP may indeed be better suited for some systems, while other manufacturers’ devices may be associated with potential problems or increased work load, could impact daily clinical routine and needs to be further addressed and evaluated in controlled studies in the future. Potential added benefits of digital data transfer using MediConnect^® software were not assessed: Electronic storage of device interrogation and clinical data in the software package facilitates the handling of safety information, device recalls or scientific investigations among CIED patients. While the present study confirms the feasibility of newly implemented digital CIED interrogation data transfer in a high-volume center, a prospective, randomized multicenter study is required to further elucidate characteristics of its long-term use in clinical application.