Be Careful What You Wish For - Managing Devices and Data In Your Patient's Home

1. Be Careful What You Wish For - Managing Devices and Data In Your Patient's Home Lee R. Goldberg, MD, MPH Associate Professor of Medicine Heart Failure/Transplant Program University of Pennsylvania September 9, 2008

2. In the “Perfect World” • Chronic diseases would be managed by “daily” monitoring that would allow both clinician intervention and patient self management leading to improved “quality” and “outcomes” • The “savings” could be used for other purposes within the Health Care System like prevention

3. Prior Studies • Hypothesis: We hypothesize that patients with disease X who are treated with home monitoring technology Y will have an improvement in outcome Z. • Little focus on the mechanism of changes in outcomes – what specifically is driving the outcomes (good or bad)

4. Assumptions • Monitoring can impact outcomes and self management • The impact is positive (does not increase cost or cause harm) • Clinicians want or need to know the data • Clinicians can identify when and how to respond from a potentially large volume of data • The data are “actionable” • The data are “reliable” • “Systems” are in place that can quickly and easily incorporate all the data into the patient record

5. The Reality • Many studies have shown improvements in a variety of outcomes from utilization to quality of life to improved survival • These improvements have been difficult to duplicate outside the confines of a single center or research project - “Implementation of Innovations” • Some studies have shown increased costs/utilization (?improved access) or no impact at all • The individual centers involved combined with risk (and access to care) of the population studied seems to drive the outcome • Managing the data and incorporating it into clinical practice is a significant challenge

6. What Could be Going On? Outside of the technology…. • Improved access to care in general • Improved adherence to Guideline Based Care • Improved self-management • Identification of other barriers to care – financial, psycho-social, comorbid illness • Novelty of the technology • Device acts as a “reminder” • Regular human contact…. Need to collect data about these factors during a study to get at “mechanism”

7. What is “Improved Outcome”? • Perspective – who is interested? • Patient • Provider • Payer • Health Care Institutions • Society • Cost (only reduction in costs or effectiveness? Total vs. Hospital?) • “Quality of Life” • Improved adherence to “Evidence Based Medicine” • Safety - improved or not worsened? • System performance – Does the technology perform as designed or intended? • Improved survival Competing Interests

8. Barriers to Successful Implementation of Telemedicine Interventions • Reimbursement for supervision of telemedicine and disease management systems • Trained clinicians to manage the data and the disease • Mechanisms to consistently and reliably review patient data and alerts • Development of appropriate algorithms to respond to patient data in a manner that improves patient outcomes • Medical-legal liability for data collected • Professional licensure across state lines • Lack of evidence for types and frequency of patient data collected and impact • Clinicians’ fear of being replaced by technology • Physician acceptance

9. Lingering Questions • Type of technology - Intensity • Is simple better? – scale versus implantable monitor • Is there too much data? – can we “hurt” people by responding too quickly? • Dose of technology • Daily monitoring necessary? • Duration of intervention • How long to continue? • Withdrawal effect or do patients “learn”? • How should we manage the data? • Where is the magic? • Technology? • People?

10. Our Study - Assessing Quality of Telehealth:Home Heart Failure Care Comparing Patient-Driven Technology ModelsR01 HS015459 • A study comparing 3 different care models of outpatient heart failure care • Usual care • Electronic monitoring (scale, BP cuff, questions, +/- glucometer) with nurse case management • Electronic monitoring with self-management – interactive voice response system

11. Our Primary Hypotheses • Both electronic disease management strategies will be superior to usual care in reducing hospitalizations • The patient self-management electronic disease management arm will not be inferior to nurse case management disease management arm • “Testing data flow and human contact”

12. Our Secondary Hypotheses • “Quality of life” will be improved for the patients in the electronic disease management arms as compared to usual care • “Quality of life” will not be different between the two electronic disease management arms • Adherence to heart failure guideline care will be improved in the electronic disease management arms • Self Management will reduce the cost of HF care more than Case Management by eliminating the cost of nursing case management.

13. Our Secondary Hypotheses • Assessment of Self Management patients’ vital signs and symptoms by the expert clinical decision support system, coupled with tailored self-care algorithms, will improve patients’ self efficacy in the management of their disease more so than in patients in the Case Management group. • Self Management and Case Management patients will have greater satisfaction with care than Standard care patients. • Physician’s satisfaction will be higher with Self Management and Case management approaches to patient management than Standard care.

14. Measured Outcomes • Hospitalizations for HF, cardiovascular and all causes. • Hospital length of stay (LOS) • ER visits for HF, cardiovascular and all causes. • Survival, mortality and fatal and nonfatal myocardial infarctions • Self-efficacy in management of heart failure as well as HRQoL and its dimensions assessed by the Kansas City Cardiomyopathy Questionnaire • Acute care visits to physicians. • Satisfaction with care

15. Technology • Shipped to patient’s home • Connected to phone line • Equipment identical for the two technology arms

16. Implementation:Designing the Intervention • Designing the “IVR” for the electronic only disease management arm • Sensitivity versus specificity • Consensus on the “clinical” content • Review by experienced heart failure clinicians • Patient focused • Easy to use • Easy to understand • Short and to the point • Safe • Many concerns and delays during the design phase

17. Implementation:Safety Pilot of the IVR • Given challenges with the IVR safety pilot using simulated patients was performed • Members of IRB • Family members of study staff • AHRQ staff • Multiple technical and clinical issues identified and corrected • Delayed enrollment but improved safety and understanding of a new patient management system

18. Implementation:Vendor Issues • Technology “up-time” • Many technical issues with IVR • Many technical issues with servers, phone lines, etc. • Troubleshooting with subjects and providers • Support for installation • Support for problems • Equipment issues • Defective • Batteries • Availability of vendor on off hours

19. General Vendor Considerations • Privacy – HIPAA issues • Service guarantee • System monitoring – continuous? • Approved equipment (FDA/FCC) • Support hours • Interface issues • Fax • Web • E-mail • Pager (text messaging) • Integration • ?EMR interface

20. Home IT Implementation Issues Farberow B, Hatton V, Leenknecht C, Goldberg LR, Hornung CA, Reyes B. Caveat Emptor: The Need for Evidence, Regulation and Certification of Home Telehealth Systems for the Management of Chronic Conditions, AJMQ AJMQ 23(3): 208-14, May-June 2008.

21. Home IT Implementation Issues Farberow B, Hatton V, Leenknecht C, Goldberg LR, Hornung CA, Reyes B. Caveat Emptor: The Need for Evidence, Regulation and Certification of Home Telehealth Systems for the Management of Chronic Conditions, AJMQ 23(3): 208-14, May-June 2008..

22. Implementation:Overcoming Provider Resistance • Providers (practices) concerns • Too much time to review data/alerts • Coverage during day and on nights/weekends/holidays – “critical labs” • Medical-legal concerns about responsibility for data – where and how to document • Educate to respond (not just file) • Educate to respond appropriately • Comfort with adjusting medications over the phone • Use of extra visits/ER when appropriate only • “Learning curve” observed with most clinicians

23. Implementation:Subjects • Phone line (land line) • Not cellular only • Not Voice over internet (VOIP) • In the home? (or access daily nearby?) • Ability to install equipment • Ability to hear and see well enough to use the equipment • Ability to stand on the scale

24. Status • 156 subjects randomized • Last patient out May 31, 2008 • Database lock – July 21, 2008 • Data analysis underway

25. Challenges • Several “technology” related challenges • Server down • Communication down • IVR “errors” • Provider issues • “too many alerts” in IVR arm • Alert “fatigue” • Educate around adjusting parameters to make alerts meaningful

26. Subjects • Seem to prefer the nurse case management arm (?is this our bias or just more contact with these subjects??) • Interacting with a “person” • Nurses identify other issues that may increase cost but improve either quality of care or patient satisfaction • Battery replacement • Accuracy of scale questioned • Technical due to carpeting and scale placement?

27. Early Results • Many anecdotes from call center, providers and subjects • Identified serious medication errors • Intervened to avoid ER or hospitalization • Identified several “educational opportunities” • “Missed” data transmission is an important parameter to be followed • Nurse Case Managers seem to promote patient self-care and encourage patient-clinician communication • IHS group – many more hospitalizations and ER visits in all arms (clearly a higher risk group) • But…the IVR group appears to have at least as good outcomes – could this be cost-effective “self-management” – it does not appear to be “inferior”

28. Conclusions • Several challenges to home monitoring • Provider • Vendor • Subject • Data management • Payers • Studies need to be performed to understand what drives changes in outcomes as opposed to focusing on a specific technology or program • Studies need to be performed on “best practice” for data management with standardized HIPAA compliant interfaces with alerts • Desperate need for vendor regulation, standardization and/or certification so that we know what we are testing (and what the subjects are getting) • Despite skepticism from clinicians, the IVR system appears to be “non-inferior” to the nurse case management system • Cost implications • Mechanism implications