The Impact of Enhanced Critical Care Training and 24/7 (tele-ICU) - - PowerPoint PPT Presentation

The Impact of Enhanced Critical Care Training and 24/7 (tele-ICU) Support on Medicare Spending and Post-Discharge Utilization Patterns

June 27, 2017

Matthew Trombley (Abt Associates), Andrea Hassol (Abt Associates), Jennifer Lloyd (Center for Medicare & Medicaid Services), Timothy Buchman (Emory Critical Care Center), Allison Marier (Abt Associates), Alan White (Abt Associates), Erin Colligan (Center for Medicare & Medicaid Services)

The contents of this presentation are solely the responsibility of the authors and do not necessarily represent the official views of the U.S. Department of Health and Human Services or any of its agencies.

Abt Associates | pg 2

Background

• Hospital intensive care units (ICUs) treat the most

critically ill patients, requiring highly trained staff and sophisticated resources

• Use of intensivist physicians associated with better

quality intensive care

• Only 37% of all ICU patients in US are currently

covered by intensivist physicians

• In many ICUs there are no intensivist physicians

present on-site at night or on weekends

The contents of this presentation are solely the responsibility of the authors and do not necessarily represent the official views of the U.S. Department of Health and Human Services or any of its agencies.

Abt Associates | pg 3

Background

• Integrating physicians’ assistants and nurse practitioners

into the ICU team is demonstrably cost-effective and may maintain or decrease length of stay (LOS) and mortality

• Electronic ICUs (tele-ICUs) allow continuous remote

monitoring of patients, including nights and weekends

• Tele-ICU implementation is correlated with reductions in

ICU and hospital LOS and mortality

• Recent research has identified potentially large savings to

hospitals

• Prior literature did not consider post-discharge patient
• utcomes or potential benefits to payers

The contents of this presentation are solely the responsibility of the authors and do not necessarily represent the official views of the U.S. Department of Health and Human Services or any of its agencies.

Abt Associates | pg 4

Program Background

• Emory University Hospital received Healthcare

Innovation Award from CMS in the summer of 2012

• Award funded joint program targeting care in ICU

setting

• Critical care residency for physicians’ assistants and nurse

practitioners

• Tele-ICU to provide remote support to ICUs

The contents of this presentation are solely the responsibility of the authors and do not necessarily represent the official views of the U.S. Department of Health and Human Services or any of its agencies.

Abt Associates | pg 5

Program Background

• Three participating

hospitals

• Emory University
• Emory University Midtown
• St. Joseph’s Hospital
• Eight participating

ICU/coronary care units

• 2 medical/surgical ICUs
• 2 CCUs
• 4 cardiothoracic surgery

ICUs

Source: Google Maps

The contents of this presentation are solely the responsibility of the authors and do not necessarily represent the official views of the U.S. Department of Health and Human Services or any of its agencies.

Abt Associates | pg 6

Program Background

• Critical care residency training program for non-physician

practitioners placed first graduates January 2013, 15 months before the tele-ICU became operational

The contents of this presentation are solely the responsibility of the authors and do not necessarily represent the official views of the U.S. Department of Health and Human Services or any of its agencies.

• 6-12 month residency
• Practical skills, focusing on ICU

procedures

• Leadership training

Source: Emory University

Abt Associates | pg 7

Program Background

• Tele-ICU coverage started

April-May 2014

• Audio-visual connection;

telemetry data feed; physiologic alert

• Staffed 24/7 by experienced

critical care nurses, monitoring telemetry

• Staffed by intensivist physicians

nights and weekends

Source: Emory University The contents of this presentation are solely the responsibility of the authors and do not necessarily represent the official views of the U.S. Department of Health and Human Services or any of its agencies.

Abt Associates | pg 8

Evaluation Design

• Difference-in-difference evaluation design

– Allows us to estimate program impacts despite non- experimental program design – Before and after program start date, at participating hospitals and matched comparisons – Intervention group: Medicare fee-for-service beneficiaries with relevant ICU/CCU stay at one of participating hospitals – Comparison group: Medicare fee-for-service beneficiaries at

• ne of 9 comparison hospitals located in Atlanta Hospital

Referral Region with at least 250 beds

The contents of this presentation are solely the responsibility of the authors and do not necessarily represent the official views of the U.S. Department of Health and Human Services or any of its agencies.

Abt Associates | pg 9

Intent to Treat Sample Selection

• Emory provided a registry of all patients treated in ICUs

covered by tele-ICU

• Linked registry data to Medicare claims
• ICU revenue codes alone insufficiently sensitive/specific

to define the intervention population for purposes of drawing a matched comparison sample

• Added ICD-9 codes for permutations of primary and secondary

diagnoses present on claims from patients in the registry, to improve accuracy in defining intervention population

• Definition of intervention population applied to patients in

comparison hospitals to define comparable group using same criteria

The contents of this presentation are solely the responsibility of the authors and do not necessarily represent the official views of the U.S. Department of Health and Human Services or any of its agencies.

Abt Associates | pg 10

Intent to Treat Sample Selection

• Example:
• Emory registry included beneficiaries with primary diagnosis

heart failure and secondary diagnosis sepsis; but none with primary sepsis and secondary heart failure

• All inpatient stays with ICU code 0200 or 021X, and primary

diagnosis of heart failure, and secondary diagnosis of sepsis, included in analytic sample

• Inpatient stays with primary diagnosis sepsis and secondary

diagnosis heart failure not included

The contents of this presentation are solely the responsibility of the authors and do not necessarily represent the official views of the U.S. Department of Health and Human Services or any of its agencies.

Abt Associates | pg 11

Intent to Treat Sample Selection

• All ICU stays meeting

these criteria initiated an “episode of care” if:

• The beneficiary did not

die in the hospital

• The beneficiary had not

had an episode of care in the prior 120 days

• Beneficiaries could be in

sample multiple times if episodes of ICU care were separated by at least 120 days

Final Sample of Episodes

Emory Comparison Baseline Period 1/1/10 - 4/24/14 6,129 17,136 Intervention Period 4/24/14 - 6/30/15 3,093 4,002

The contents of this presentation are solely the responsibility of the authors and do not necessarily represent the official views of the U.S. Department of Health and Human Services or any of its agencies.

• Episode of care includes

inpatient stay and subsequent 60 days post-discharge

Abt Associates | pg 12

Cost and Utilization Measures

• Total Part A and B Medicare payments per 60-day

episode (inpatient hospitalization through 60 days post-discharge)

• All-cause inpatient readmissions within 30 or 60 days

after discharge

• ED visits within 30 or 60 days after discharge
• Inpatient length of stay
• Discharge destination
• Home, Home Health, Institutional post-acute care facility,

“other”

The contents of this presentation are solely the responsibility of the authors and do not necessarily represent the official views of the U.S. Department of Health and Human Services or any of its agencies.

Abt Associates | pg 13

Statistical Analyses

• Regression-adjusted difference-in-difference
• Quarter and hospital fixed effects
• Controls for
• patient demographics
• prior health status
• Major Diagnostic Category (MDC)

Outcome Statistical Model Total Medicare Spending OLS Inpatient Readmissions Logit ED Visits Logit Length of Stay Negative Binomial Discharge Destination Multinomial Logit

The contents of this presentation are solely the responsibility of the authors and do not necessarily represent the official views of the U.S. Department of Health and Human Services or any of its agencies.

Abt Associates | pg 14

Descriptive Analysis – Independent Variables

The contents of this presentation are solely the responsibility of the authors and do not necessarily represent the official views of the U.S. Department of Health and Human Services or any of its agencies.

Emory Baseline Period (N=6,129) Comparison Baseline Period (N=17,136) Difference in Baseline Period Differential Change Female (%) 44.4 49.4 5.1***

• 2.3*

Nonwhite (%) 41.0 28.5 12.5***

• 0.5***

Age 70.2 + 14.0 72.2 + 12.8

• 2.0***

0.0 Hierarchical Condition Category Score 2.9 + 3.1 2.7 + 2.7 0.3*** 0.3*** Medicaid (%) 64.1 66.8

• 2.7***

0.0 Non-disabled (%) 63.5 64.8

• 0.1*

0.2 Hospital Transfer (%) 22.2 6.8 15.4***

• 8.7***

Other Health Transfer (%) 3.3 5.4

• 2.1***

5.0*** Charlson Comorbidity Index Score 3.1 + 2.1 2.9 + 1.9 0.2***

• 0.2***

*p<0.10 **p<0.05 ***p<0.01 + indicates standard deviation for continuous measures Differential changes for (%) measures are reported as percentage point changes.

Abt Associates | pg 15

Descriptive Analysis – Independent Variables

The contents of this presentation are solely the responsibility of the authors and do not necessarily represent the official views of the U.S. Department of Health and Human Services or any of its agencies.

Emory Baseline Period (N=6,129) Comparison Baseline Period (N=17,136) Difference in Baseline Period Differential Change Female (%) 44.4 49.4 5.1***

• 2.3*

Nonwhite (%) 41.0 28.5 12.5***

• 0.5***

Age 70.2 + 14.0 72.2 + 12.8

• 2.0***

0.0 Hierarchical Condition Category Score 2.9 + 3.1 2.7 + 2.7 0.3*** 0.3*** Medicaid (%) 64.1 66.8

• 2.7***

0.0 Non-disabled (%) 63.5 64.8

• 0.1*

0.2 Hospital Transfer (%) 22.2 6.8 15.4***

• 8.7***

Other Health Transfer (%) 3.3 5.4

• 2.1***

5.0*** Charlson Comorbidity Index Score 3.1 + 2.1 2.9 + 1.9 0.2***

• 0.2***

*p<0.10 **p<0.05 ***p<0.01 + indicates standard deviation for continuous measures Differential changes for (%) measures are reported as percentage point changes.

Abt Associates | pg 16

Descriptive Analysis – Dependent Variables

The contents of this presentation are solely the responsibility of the authors and do not necessarily represent the official views of the U.S. Department of Health and Human Services or any of its agencies.

Emory Baseline Period (N=6,129) Comparison Baseline Period (N=17,136) Difference in Baseline Period Differential Change

Total 60-day Medicare Spending ($) 12,744 + 18,187 10,969 + 16,234 1175***

• 1882***

Utilization 30-Day Inpatient Readmissions (%) 20.6 19.4 1.2**

• 1.3

60-Day Inpatient Readmissions (%) 30.4 28.6 1.8**

• 2.7**

30-Day Emergency Department Visits (%) 26.2 24.4 1.8***

• 0.5

60-Day Emergency Department Visits (%) 36.6 35.2 1.4**

• 1.3

Length of Stay (days) 11.6 8.6 3.0***

• 0.6***

Discharge Destination Home (%) 46.6 52.2

• 5.6***

2.4** Home Health (%) 21.5 17.1 4.5*** 4.2*** Institutional Post-Acute Care (%) 22.3 21.2 1.1*

• 7.4***

Other (%) 9.5 9.5 0.0 0.9

*p<0.10 **p<0.05 ***p<0.01 + indicates standard deviation for continuous measures Differential changes for (%) measures are reported as percentage point changes.

Abt Associates | pg 17

Regression Results

Measure Average Treatment Effect Standard Error Percent Change Total 60-day Medicare Spending ($)

• 1486.27***

458.65

• 11.66

Utilization 30-Day Inpatient Readmissions

• 0.89

1.14

• 4.29%

60-Day Inpatient Readmissions

• 2.14*

1.28

• 7.05%

30-Day Emergency Department Visits 0.21 1.26 0.60% 60-Day Emergency Department Visits

• 0.54

1.37

• 1.46%

Length of Stay (days)

• 0.08

0.22

• 0.66%

Discharge Destination Discharge Home 0.19 1.31 0.41% Home Health 4.85*** 1.25 22.53% Institutional Post-Acute Care

• 6.90***

0.94

• 30.98%

Other 1.86** 0.87 19.50%

*p<0.10 **p<0.05 ***p<0.01 All inferences are based on Huber-White robust standard errors. All changes in measures of utilization and discharge destination, besides length of stay, expressed as percentage points.

The contents of this presentation are solely the responsibility of the authors and do not necessarily represent the official views of the U.S. Department of Health and Human Services or any of its agencies.

Abt Associates | pg 18

Sensitivity Analyses

• Sensitivity analyses conducted on the 60-day Medicare

spending measure

• Linear time trends in baseline period not significantly different

between intervention and comparison samples

• Pseudo start dates January 1 2012, 2013, and 2014 yielded

insignificant estimated impacts

• The baseline period includes 14 months during which

residency grads were in the ICUs without tele-ICU support. Removing these 14 months from the baseline yielded larger estimated savings for the combined program

• Including decedents in the model did not affect estimates
• Results from nonlinear regression model slightly larger

The contents of this presentation are solely the responsibility of the authors and do not necessarily represent the official views of the U.S. Department of Health and Human Services or any of its agencies.

Abt Associates | pg 19

Limitations

• Sample Medicare FFS beneficiaries only
• Literature suggests that Medicare FFS population

generalizable to Medicare managed care population

• Sample defined by clinical criteria rather than simply the

ICUs in which patients received care (because claims don’t contain sufficient detail about specialty ICUs)

• Intent-to-treat sample excluded all episodes with diagnoses

that never caused admission to a participating ICU

• MDCs are less precise than other clinical controls used in

the literature (e.g., Acute Physiology and Chronic Health Evaluation scores)

• Standard errors will be larger but point estimate still unbiased

The contents of this presentation are solely the responsibility of the authors and do not necessarily represent the official views of the U.S. Department of Health and Human Services or any of its agencies.

Abt Associates | pg 20

Conclusion

• Combined critical care residency and tele-ICU

programs significantly reduced average Medicare spending per 60-day episode of care by $1,486

• Estimated aggregate savings of $4.6 million over 14 months
• Likely driven by decline in rate of 60-day readmissions and

shift from institutional post-acute care to home-based care

• Demonstrates that benefits of ICU interventions

extend beyond discharge

• Demonstrates benefits to payers

The contents of this presentation are solely the responsibility of the authors and do not necessarily represent the official views of the U.S. Department of Health and Human Services or any of its agencies.

Abt Associates | pg 21

Acknowledgements

Funding for this project was provided by the Center for Medicare and Medicaid Services under contract # GS- 10F-0086K Task Order # HHSM-500-2013-00310G “Evaluation of Hospital Settings Interventions Group supported by Health Care Innovation Awards (HCIA)” For more information, please contact Matthew Trombley at matthew_trombley@abtassoc.com

The contents of this presentation are solely the responsibility of the authors and do not necessarily represent the official views of the U.S. Department of Health and Human Services or any of its agencies.