Use of Simulation to Assess the Impact of a Remote Monitoring System.

OBJECTIVES: Remote monitoring (RM) is useful in chronic diseases, however evidence is scarce in home dialysis. SharesourceTM, a two way connectivity platform with RM capabilities, should enable earlier intervention and healthcare savings. This abstract describes a simulation study designed to estimate the type and importance of the savings. METHODS: This study used an approach similar (although in a virtual environment) to that for diagnostic tests , i.e., measuring the impact of additional information on treatment pathways. Automated peritoneal dialysis (APD) patient profiles (e.g., therapy non-compliance, fluid overload, etc.) were developed by a group of in-house nephrologists and nurses experienced in managing APD patients. Each patient profile was developed in 2 scenarios: “with RM” information (treatment data, blood pressure, weight) and “without RM”. Nephrologists (US=4, Germany=3, Italy=2), experts in managing APD patients, validated the “without RM” resources. The “with RM” resources were evaluated by 15 (US=7, Germany=4, Italy=4) APD teams (1 nephrologist & 1 nurse). Resources were compared with t-tests. RESULTS: Patient profile development required several rounds of individual reviews and group meetings over a 6 month-period to ensure clarity of clinical event sequence and clinician management response. During the process, 8 profiles were abandoned, leaving 12 profiles for the simulation. Recruitment of experts and APD teams was moderately easy, but the biggest hurdle was the contract signature in some countries (e.g., Italy). Participants submitted responses, usually within two weeks. Overall, results indicate that RM could avoid in the USA and Germany 49-75 healthcare resources (4.1-6.2 per patient profile), including clinic calls/visits (48-72), hospitalizations (2), emergency room visits (3.5-5) and transfer from APD to in-center hemodialysis (1-2.25). Numbers were lower in Italy. These cases were assessed by the APD experts as being representative of 15-37% of the APD patient population. CONCLUSIONS: This simulation was a rapid and affordable way of estimating the type and importance of the potential healthcare resource consumption reduction to be expected by RM in APD patients while waiting for real-world evidence. USE OF SIMULATION TO ASSESS THE IMPACT OF A REMOTE MONITORING SYSTEM Laplante S1, McLeod K2, Danek J1, Kudelka T1, Sloand J1, Gellens M1 1-Baxter Healthcare Corporation, Deerfield, IL; 2-Xcenda, Palm Harbor, FL Poster No. PRM240 Background Remote monitoring (RM) involves a device or software in the home setting. RM collects patient’s clinical data, such as blood pressure, weight, and temperature and sends it to a healthcare professional that monitors and analyzes the data received, giving opportunities for preventative therapeutic intervention. RM has been shown to decrease hospitalization and death in patients with chronic diseases such as heart failure. A Cochrane review and meta-analysis published in 2011 estimated that RM in the form of telemonitoring of patient’s weight and blood pressure could reduce disease-related hospitalization by 21% and death by 9%.[1] However, evidence is scarce in home dialysis. SharesourceTM, a two-way connectivity platform with RM capabilities, should enable earlier intervention that could lead to a reduction in usage of more expensive healthcare resources such as hospitalizations or emergency room visits. Objectives This simulation study was designed to estimate the type and importance of the resources that could be saved. Conclusions This simulation was a rapid and affordable way of estimating the type and importance of the potential healthcare resource consumption reduction to be expected by RM in APD patients while waiting for realworld evidence. References 1. Inglis SC et al. Structure telephone support or telemonitoring programmes for patients with chronic heart failure (Review). Cochrane Database of Systematic Reviews 2010, Issue 8. Art. No.: CD007228. DOI: 10.1002/14651858.CD007228.pub2. 2. Lacourcière Y et al. Impact of Levovist ultrasonographic contrast agent on the diagnosis and management of hypertensive patients with suspected renal artery stenosis: a Canadian multicenter pilot study. Can Assoc Radiol J. 2002 Oct;53(4):219-27. Results • Patient profile development required several rounds of individual reviews and group meetings over a 6 month-period to ensure clarity of clinical event sequence and clinician management response. During the process, 8 profiles were abandoned, leaving 12 profiles for the simulation. • Recruitment of experts and APD teams was moderately easy, but the biggest hurdle was the contract signature in some countries (e.g., Italy). • Participants submitted responses, usually within two weeks. • Overall, in Germany and the USA, RM could avoid consumption of 49-75 healthcare resources (4.1-6.2 per patient profile), including: • 48-72 clinic calls/visits • 2 hospitalizations • 3.5-5 emergency room visits • 1-2.25 transfer from APD to in-center hemodialysis • In Italy, the numbers were much lower but both the number of KOLs and participant teams were smaller. • These cases were perceived as representative of 17-37% of the APD patient population. • An example of resources consumption with and without RM is displayed for Germany in Figure 2. Figure 3 gives the healthcare resources avoided for all 3 countries. Methods The methodological approach was that of measuring the impact of the additional information provided by the RM system on patient treatment pathways. This approach is traditionally used for diagnostic test.[2] Step 1 APD patient profiles: Twelve (12) APD patient profiles with therapy non-compliance, fluid overload, low drain/unidentified alarms, or missing/forged data were developed by a group of nephrologists and PD nurses at Baxter Healthcare Corporation in two versions: a) Without RM: description of patient demographic, disease history and current events with dialysis clinic actions in a usual care environment (i.e., without information from the RM software) b) With RM: description of patient demographic, disease history and current initial event with information (weight, blood pressure, treatment data) from the RM software. An example of a patient profile without and with RM is shown in Figure 1. Figure 2: Healthcare resource estimation with and without RM (Germany)