Tactics to Make Telehealth Pay Off

Transcription

1 Tactics to Make Telehealth Pay Off Lin Li, Ph.D. Philips Research North America Kathleen Sullivan, RN, MSN Dignity Health 2013 NAHC Annual Meeting Oct, 2013 Today s discussion Insights gained from market research Home Monitoring Economic Modeling Lin Li, Ph.D. Member Research Staff Philips Research North America Techniques for Making Telehealth Pay Off A Provider s Perspective Kathleen Sullivan, RN, MSN Vice President, Post Acute Care Services Dignity Health 2 1

2 I. Introduction 3 Healthcare Landscape across the Care Continuum Financial Risk Government Agencies Insurance Companies Seniors Seniors Hospitals Primary Care Healthcare Delivery Network Population Health Management Seniors Home Health Agencies Skilled Nursing Facilities Seniors Need Resources, Tools, and Expertise for Economic Modeling to Evaluate Intervention Effectiveness (e.g. Telehealth Technology) Senior Living Community 4 2

3 Research Objective Model and quantitatively evaluate the Intrinsic value Commercial viability of a home monitoring technology Help care provider make better-informed decisions: Exert the right efforts to the right direction Allocating resources to those patients in greatest need Accelerate time to ROI 5 Payers perspective: Clinical impact and cost impact of home monitoring Clinical effects lie in reduction in: Mortality rate Re-admission rate Reduction in length of stay (LOS) Economic effects: Incur cost through HM infrastructure and services save cost through reduction of admission and LOS Payers Whether the cost saving (through avoid admissions) will offset the costs of launching and maintaining the home-based monitoring is a key consideration (and can be modeled) Cost triggering via infrastructure and home-based services Cost saving via reduction of admission and LOS 6 3

4 Market restraints of home monitoring Reimbursement Cost Effectiveness These restraints force the market participants to take a sterner look at the actual performance and economic impact of their systems Major competitive factors are mainly focused on the capability for the system to be cost effective and easily explained as cost efficient. Early analysis of net cost benefits and impact on payers healthcare budget are of great importance in the current economics and healthcare marketplace. 7 II. Home Monitoring Economic Modeling 8 4

5 Generic Healthcare economics methods: comparative effectiveness analysis Determine better strategy t to manage patients No intervention or current state-of-the-art intervention Δ Cost Δ Effect = ICER New intervention Model and compare the two or more alternative management strategies 9 Economic Model on CHF CHF Markov Process Markov Model is the MOST commonly used decision-analytic model to simulate health outcomes and costs in CHF. 80% of heart failure economics journal papers are utilizing Markov Model Determine better strategy to manage patients No intervention New intervention Δ Cost Δ Effect 10 5

6 Model structure derived from a peer-review journal paper Our approach: We first duplicated d the results of Chan model to make sure our model structure is correct We then modify the model inputs and refine the structure to meet our business need Sheena Liu, Rui Xiang, Nan Liu et al Heart failure tele-monitoring : when it becomes cost saving, In preparation. 11 Methodology Overview - CHF Death Higher risk NYHA H=0 H=1 H=2 H=3 H=4+ No hosp Index hosp Rehos 1 Rehos 2 Rehos 3+ Admission History Higher risk All models have to be at certain level of abstraction. 12 6

7 Methodology Overview - Home-based monitoring Test = the home-based solution to test of possible exacerbations. It includes one or more of the following: Activity/symptoms/biomarkers/ monitoring Questionnaire Smart algorithms Treat = the home-based early intervention upon the detection of the exacerbation. It includes one or more of the following: Healthcare professionals to review data Telephone triage Personnel home visit Initiation of home medication package Abstract the Efficacy/Cost of our Technology with 6 Parameters Clinical parameters: Sensitivity of the home exacerbation detection (SEN) Specificity of the home exacerbation detection (SPE) Acute exacerbation reversion ratio (ERR), or, reduction rate of hospitalization LOS reduction ratio (LRR) through early detection and early intervention. Economic parameters: Cost of Test (Cost of Early Detection, or CED) Cost of Treat (Cost of Early Treatment, or CET) 13 Clinical impact and cost impact of home monitoring Post Discharge Ground Truth Exacerbation + Without Exacerbation - Cost incurred TEST Cost incurred TEST Test result TP + FN - FP + TN - Cost incurred Cost saved + TREAT Cost saved No action Cost incurred TREAT No action - Severity of exacerbation has been reduced (reflected in reduced LOS) Exacerbation (re-admission) is successfully reverted FPs will incur unneeded treatment, thus incur unnecessary cost FNs (missed diagnosis) will be omitted from the early treatment to prevent exacerbation, and thus will not save the hospitalization cost (but incur monitoring cost) 14 7

8 Model inputs derived from our 10-year literature meta-analysis Our approach: We found most of meta papers have included trials prior to What we did is we first duplicate Klersy s s paper to make sure our methodology is correct (we verified the results which are consistent with the publication results) We extended the meta analysis from 2008 to 2012 with more trials included We now have an excel sheet with all the raw data for all the literatures: we can continue expand our data pool as needed, and perform all kinds of analysis. Rui Xiang, Lin Li, Sheena Liu, Meta-analysis and meta-regression analysis of home-based management of heart failure patients, Journal of Telehealth and Telecare (Accepted), year literature meta-analysis of HF Telehealth: articles selections Rui Xiang, Lin Li, Sheena Liu, Meta-analysis and meta-regression analysis of home-based management of heart failure patients, Journal of Telehealth and Telecare (Accepted),

9 10-year literature meta-analysis : patients statistics Total papers: 33 randomized control trial (RCT) 26 (78.8%) papers tele monitoring 7 (21.2%) papers case management Year distribution: : 8 (24.2%) : (21.2%) 2%) 2008: 7 (21.2%) 2009: 5 (15.2%) : 6 (18.2%) Trial Country distribution: Germany: 4 (12.1%) Italy: 4 (12.1%) Netherland: 2 (6.1%) UK: 5 (15.2%) USA: 14 (42.4%) Austria Belgium Sweden Canada: 4 (12.1%) Patient characteristics Total trial patients: Average age: 69 Average male percentage: 65.8% 21 out of 33 (63.6%) papers had reported NYHA information: o NYHA I: 1.9% o NYHA II: 40.3% o NYHA III: 46.9% o NYHA IV: 3.8%. Follow up duration 12 months: 17 (51.5%) 24 months or more: 2 (6.1%) 6 months: 11 (33.3%) Less than 6 months: 3 (9.1%) A sample table Rui Xiang, Lin Li, Sheena Liu, Meta-analysis and meta-regression analysis of home-based management of heart failure patients, Journal of Telehealth and Telecare (Accepted), year literature meta-analysis: Results Meta-Analysis Random Effect Model using REML for All Caused Mortality Study RR and 95% CI Weight RR and 95% CI Antonicelli et al., 2008 Balk et al., 2008 Blum et al. (Abstract only), 2007 Capomolla et al., 2004 Cleland et al., 2005 de Lusignan et al., 2001 Giordano et al., 2009 Goldberg et al., 2003 Kielblock et al. (Abstract only), 2007 Mortara et al., 2009 Soran et al., 2008 Villani et al., 2007 Woodend et al., 2008 Zugck et al. (Abstract only), 2008 Jolly et al., 2009 Schwarz et al., 2008 Lynga et al., 2012 Koehler et al., 2011 Weintraub et al., 2012 Wade et al., 2011 Kashem et al., 2008 Dendale et al., 2012 Scherr et al., 2009 Blue et al., 2001 Kasper et al., 2002 Krumholz et al., 2002 Dunagan et al., 2005 RE Model 1.47% 0.62 [ 0.16, 2.36 ] 2.93% 1.26 [ 0.50, 3.14 ] 6.14% 0.81 [ 0.45, 1.44 ] 2.12% 0.70 [ 0.24, 2.11 ] 7.30% 0.71 [ 0.42, 1.18 ] 1.10% 0.67 [ 0.14, 3.17 ] 7.14% 0.66 [ 0.39, 1.10 ] 4.96% 0.44 [ 0.22, 0.85 ] 11.13% 0.54 [ 0.37, 0.77 ] 3.68% 1.37 [ 0.61, 3.04 ] 4.32% 0.63 [ 0.30, 1.29 ] 2.52% 0.74 [ 0.27, 2.00 ] 1.63% 1.19 [ 0.34, 4.22 ] 0.56% 1.55 [ 0.17, ] 2.08% 1.42 [ 0.47, 4.29 ] 1.89% 0.57 [ 0.18, 1.83 ] 2.12% 0.58 [ 0.19, 1.72 ] 11.57% 1.11 [ 0.79, 1.57 ] 0.58% 0.24 [ 0.03, 2.15 ] 2.07% 0.93 [ 0.31, 2.81 ] 0.37% 1.00 [ 0.07, ] 2.22% 0.29 [ 0.10, 0.83 ] 0.27% 0.33 [ 0.01, 8.01 ] 8.31% 0.96 [ 0.61, 1.53 ] 3.15% 0.52 [ 0.22, 1.24 ] 4.18% 0.69 [ 0.33, 1.45 ] 4.21% 1.17 [ 0.56, 2.44 ] % 0.76 [ 0.64, 0.90 ] Tele-health group has 24% mortality reduction as compared to control group. Results are statistically significant Tele-health group has 28% readmission reduction as compared to control group. Results are statistically significant. 18 9

10 Model inputs/outputs Model inputs and outputs INPUTS OUTPUTS # of Months on Tele-health Patients Risk Distributions (admission history + NYHA) Technology Choice Infrastructure and Service Cost USUAL CARE TELE-HEALTH META-ANALYSIS: CUMULATIVE EFFICACY OF TELE-HEALTH # of Admissions Saved Mortality/Life Years Saved Net Cost Saved Customers economic viability 19 II. Prototype Tool 20 10

11 Sensitivity Analysis Key question: What is the maximum monthly recurring service price to make our solution long-term cost-saving for the payers: for different patient risk categories in different product performance scenarios Economic effects considered: Incur cost through HM infrastructure and services save cost through reduction of admission and LOS 21 Customize the Intervention scenario 22 11

12 Cost-effectiveness analysis cross candidate scenarios 23 Identify the target patient group with the best ROI 24 12

13 25 Conclusion Model: Structure: peer-reviewed journal Data: peer-reviewed updated meta-analysis Generalize the model: The model can be extended to other technological solutions It is important to stratify patients based on their clinical conditions and demands, and apply different levels of services efficiently and cost-effectively. Whether the cost saving (through avoid admissions) will offset the costs of launching and maintaining the home-based monitoring is a key consideration (and can be modeled) To achieve sustainability, the focus must be on identifying the financial risk holders return on investment (ROI) and demonstrating economic viability

14 Dignity Health Homecare Services Techniques for Making Telehealth Pay Off A Provider s Perspective Kathleen Sullivan, RN, MSN Vice President, Post Acute Care Services September 2013 Value Proposition Reduced hospitalizations and emergent care Improved clinical outcomes More effective disease management across the continuum Enhanced quality of life for patients Better aftercare for acute patients with heart failure Care transitions support Support for a stretched primary care system Reduced skilled nursing visits; overall cost savings to home health agencies 28 14

15 History of the Program Robert Wood Johnson Foundation Grant to develop web enabled Care Management System for CHF 2002 Telephonic Care Management Program for CHF in place for 8 years Center for Technology and Aging Remote Patient Monitoring Grant Awardee 2011 Remote patient monitoring in place for two years 29 Goals of the Remote Patient Monitoring Program Build a network of distance health service delivery using integrated technology that supports patient and clinician allies Improve patient compliance with medications, diet, weight monitoring and symptom management Contribute to the development of system wide innovations: telemedicine using IPAD/Web Ex, text messaging with Diabetes Care, Asthma Care using GPS inhalers and building e health communities Reduction in avoidable hospitalization 30 15

16 Patient Monitors Selection Criteria 60 years or older Hx non compliance Prior hospitalization(s) within six months More than two co morbidities Stage III IV Heart Failure Multiple ED encounters within six months prior to referral 32 16

17 Dignity Health Central Coast Chronic Care Management Model Heart Failure Hospital Physician Heart Failure Care Management Program Outpatient Services Home Health 33 Center for Technology and Aging Grant Results 51 Patients placed on monitors 39 Patients Monitored for > 6 Months 27 Patients Discharged: 7 Stable 2 No longer wanted 6 Graduated 2 Fractured hip 6 Expired 1 Declined 3 Hospice NYHA Classification: 20% Class II 59% Class III 22% Class IV Average Age: 75 Gender: 37% Male 63% Female 34 17

18 Results TELEHEALTH PATIENT SATISFACTION SURVEY - 8 Weeks 80.0% 70.0% 60.0% 50.0% 40.0% 30.0% 20.0% 10.0% SURVEY QUESTIONS 0.0% DEFINITELY NOT I DON'T THINK SO MAYBE YES, I THINK SO YES, DEFINATELY Telehealth equipment was easy to use 1.9% 0.0% 0.0% 25.9% 72.2% I would recommend Telehealth to others 0.0% 1.9% 1.9% 29.6% 66.7% N = 51 Updated w ith data available as of 07/30/12 35 Results, cont. Reduction in Re Hospitalization (07/01/11 thru 06/30/12) 58% reduction in hospitalization within 30 days as compared to control group (The monitored patients experienced an 9% readmission rate within 30d while the non monitored group experienced a 21 22% readmission rate) 59% reduction in post intervention re admissions at six months as compared to prior 6 months* 58% reduction in cost of care (ACF and E.D.) post intervention at six months as compared to prior 6 months* * Test group 36 18

19 Lessons Learned Planning Vendor selection and contract process Program Infrastructure I.T. Support Clarity with ihpatient selection 37 Lessons Learned, cont. Implementation Timely deployment post hospitalization SNF patients Role of Palliative Care Telephonic vs. RPM Home Health vs. Outpatient Model 38 19

20 Lessons Learned, cont. Next Steps Embed into Population Health Strategy Define future program state Sustainability Complete ROI 39 Thank You 20