APR DRG Data Discovery - midasplus.com · APR DRG Data Discovery Vanessa Dorr RN MSN Midas+ DataVision Clinical Consultant Henry Johnson MD, MPH Midas+ Vice President and Medical

APR DRG Data Discovery

Vanessa Dorr RN MSN Midas+ DataVision Clinical Consultant

Henry Johnson MD, MPH Midas+ Vice President and Medical Director

Agenda

• Review:

– 3M APR DRG Methodology

– DataVision: Web and Server Applications

• Data Discovery:

– Resource use and Length of Stay:

• Wren Memorial, Cardiac Services

• Michael Memorial Orthopedic Services

– Mortality

• Vitale Memorial Hospital, Percutaneous Coronary Intervention (PCI)

• Summary and Questions

21st Annual Midas+ User Symposium • May 20–23, 2012 • Tucson, Arizona 2

Severity Adjustment Systems

• All Patient Diagnostic Related Group (AP DRG) – 1987: New York passed prospective payment for all

non-Medicare patients

• All-Payer Severity-adjusted DRG (APS DRG) – Different methodology, also developed from the DRG

model

• All Patient Refined DRG (APR DRG) – Researched and developed by New York Department

of Health (NYDH) in partnership with 3M™ Health Information Systems (HIS)

• Others: – STS, APACHE, Disease Staging, MedisGroups

21st Annual Midas+ User Symposium • May 20–23, 2012 • Tucson, Arizona

3M™ APR DRG

• Expanded the scope of DRGs to add:

– Severity of Illness (SOI)

• The extent of physiologic decompensation or organ system loss of function

– Risk of Mortality (ROM)

• The likelihood of dying


Development Process: Clinically Coherent with Sufficient Case Volume

• Start with All Patient DRGs (AP DRGs)

• Consolidate – Combine the “with CC” and “without CC” DRGs

– Each APR DRG has a four-point “CC” scale built in

• Pediatric additions

• Add APR DRGs for mortality – Eliminate any unique DRGs for deaths only


Subclass Assignment

• Assign each secondary diagnosis a SOI and ROM level (subclass) from 1 to 4

• In an iterative fashion, assign a subclass number based on dominant secondary diagnoses

• Some modifications for age

– Hypertension in children is considered major as a secondary

• Some modifications for specific APR DRGs

– Chronic Renal Failure as a secondary is Moderate, except if the APR DRG is diabetes, in which case it is Major


Testing

• After the clinical model for SOI and ROM is developed for each base APR DRG, it is evaluated with historical data to be sure the subclasses predict resource use (SOI) and mortality rate (ROM).

• Version 20.0 was evaluated with a national database of 8.5 million discharges, all payers, from 1,000 hospitals from 10 states, and 47 children’s hospitals in the U.S.


Beginning with APR DRG v.20

• For each APR DRG subclass, a Relative Weight (RW) was calculated to reflect the resources required for a patient in that APR DRG subclass.

• The subclass RW allows calculation of an overall Case Mix Index (CMI) for the hospital or by physician or physician group, which is severity-adjusted.

• Allows for more refined reimbursement


APR DRG v20: Relative Weights

Base Rate RW Payment

4,113.00$ 0.5804 2,387.19$

4,113.00$ 0.793 3,261.61$

4,113.00$ 1.2736 5,238.32$

4,113.00$ 3.0052 12,360.39$


Cautions

• APR DRG SOI and ROM subclasses are ranked categories unique to each APR DRG and cannot be averaged across APR DRGs (they are not really numbers).

– For example, congestive heart failure patients in severity subclass 4 are not comparable to patients with a fractured leg in severity subclass 4.

• In version 20.0 or later, you can average the RW.

• Beware choosing a population by MS-DRG or ICD-9 diagnosis and running reports or doing analysis for the whole population by APR DRG SOI or ROM.


DataVision Architecture


Midas+ Care Management

Midas+ Solutions: Tucson AZ

Standardized Measures:

DataVision

Navigator

National Comparisons:

DataVision

Web Application

Hospital Server

DataVision APR DRG Reports


APR DRG Reports

• APR-DRG Lives/Days Saved Report

– Compares the facility's mortality and average length of stay (ALOS) data observed/expected (O/E) ratio statistics with the Midas+ comparative database (CDB) mean

– Provides site percentile rank in comparison with the CDB

– Predicts the number of lives and days that might be saved if outcomes improved to a different percentile

• Hospital APR-DRG Service Line Profile

– Lists all 3M™ APR DRG–defined service lines for the facility

– Provides mortality, ALOS, and Potentially Preventable Readmission (PPR) O/E ratio or delta statistics based on the CDB mean


APR DRG Reports

• Hospital APR-DRG Ranking Profile

– Lists all APR DRGs and provides mortality, ALOS, and Potentially Preventable Readmission (PPR) O/E ratio or delta statistics based on the CDB mean

• APR-DRG Comparative Trend Analysis

– A two-part report that displays the length of stay O/E ratio and the mortality O/E ratio for the selected APR DRG, comparing the facility’s performance with percentiles from the CDB


APR DRG Reports

• APR-DRG Subclass Detail

– A three-part report that examines facility

• LOS for the APR DRG, stratified by four severity-of-illness (SOI) levels

• Mortality rates for the APR DRG, stratified by four ROM levels

• Potentially Preventable Readmissions, stratified by four SOI levels

– This report includes percentile comparisons with the comparative database

• Single APR-DRG Comparison Profile

– Lists all providers for a specific APR DRG and provides mortality and ALOS statistics for each provider

– Can be viewed by provider name or code

– Parameter allows report to be viewed by attending or principal procedure provider


APR DRG Reports

• APR-DRG Specialty Comparison Report – Compares a single provider to her peers for LOS and mortality, within

the hospital and with the comparative database

– Based on mapping of specialty groups specialties with admitting or attending activity

• Provider APR-DRG Detail Report – Lists all APR DRGs for a selected provider and provides mortality and

LOS statistics for each APR DRG as well as a specialty comparison

• All APR-DRG Provider Summary – Observed and expected mortality and LOS data for each of the facility’s

providers

– This report is useful for identifying high-volume providers


• What is in the ToolPack?

– 256 fields . . .


What Is New on Your Server?

Description # fields

Patient Demographics 14

Encounter description, including LOS 15

Admitting/Attending Physician 8

Discharging Physician 3

Payer, charge, and cost data 6

DRG (CMS) related 10

Principal Diagnosis 3

Principal Procedure 5

Count of Diagnoses, Count of Not Present on Admission 2

Secondary Diagnoses (total 25 possible, including principal) 72

Secondary Procedures (total 20 possible, including principal) 95

Consultants (total 5) 10

Primary Care physician 2

APR DRG-related 10

Birthdate 1


New fields in Excel Toolpack Export


APR DRG 174

APR DRG DESC Percutaneous cardiovascular

procedures w AMI

APR DRG Severity 2

APR DRG Mortality Risk 2

APR DRG MDC 5

CDB APR DRG All Inpatient Expected 2.9

CDB APR DRG All Inpatient Expected

Mortality

0.41%

CDB APR DRG Acute Care Expected 2.9

CDB APR DRG Acute Care Expected

Mortality

0.41%

20 21st Annual Midas+ User Symposium • May 20–23, 2012 • Tucson, Arizona

Steps to Data Discovery

Resource Use and Length of Stay:

Wren Memorial, Cardiac Services


Question: What determines LOS?

• Discharge planning

• Events that prolong stay

Why is looking at long LOS important?

Basic Steps:

• Start with DataVision web application:

– SmartReport

– APR-DRG Reports

• Next, move to server for patient data

– Report Toolpack













Total of 1,214.3 days, or 3.3 beds/day if we can go to the mean.

Can we do better?




Wren March 2011 to February 2012

APR DRG

CodeDescription

#

Cases

Obs

ALOS

Expec

ted

LOS

Ratio

Perce

ntile

Rank

20th Percentile

Pred ALOSDays saved # hospitals

194 Heart failure 1313 5 4.69 1.07 78th 3.81 1553.51 676

201Cardiac arrhythmia & conduction

disorders615 3.62 3.24 1.12 82nd 2.67 580.16 671

175Percutaneous cardiovascular

procedures w/o AMI384 3.53 2.91 1.21 82nd 2.47 407.00 457

190 Acute myocardial infarction 186 5.23 4.19 1.25 93rd 3.15 386.92 653

174Percutaneous cardiovascular

procedures w AMI312 4.34 3.87 1.12 79th 3.37 303.54 434

If we can perform better than 80% of the hospitals in the comparative database, for these five APR-DRGs in the cardiac service line, we will save 3,231 bed days, or almost 9 extra beds/day – more room for transferring ICU patients and decreasing waits in the Emergency Department.




To the Server:





DataVision

Navigator


DataVision

Web Application

Hospital Server







Wren

APR DRG 194

Heart Failure

C:/Documents and Settings/HCJ/My Documents/_Conferences 2012/Symposium 2012/APR DRG DV Data Discovery/DATA_2/A Wren/Wren APR 194 Heart Failure Jan2011 to Dec2011 2012 04 27.xlsx


Wren

APR DRG 174

PCI w AMI

C:/Documents and Settings/HCJ/My Documents/_Conferences 2012/Symposium 2012/APR DRG DV Data Discovery/DATA_2/A Wren/Wren APR 194 Heart Failure Jan2011 to Dec2011 2012 04 27.xlsx

../DATA_2/A Wren/Wren APR 174 PCI w AMI Tlpk Jan2011 Dec2011 2012 04 24.xlsx


Wren

APR DRG 175

PCI w/o AMI

C:/Documents and Settings/HCJ/My Documents/_Conferences 2012/Symposium 2012/APR DRG DV Data Discovery/DATA_2/A Wren/Wren APR 175 PCI wo AMI Tlph Jan2011 Dec2011 2012 04 28 1415.xlsx



Wren

APR DRG 190

AMI

C:/Documents and Settings/HCJ/My Documents/_Conferences 2012/Symposium 2012/APR DRG DV Data Discovery/DATA_2/A Wren/Wren APR 190 AMI Jan2011 Dec2011 2012 04 28 1450.xlsx



Wren

APR DRG 201

Card Arrhy

Condn Disords

C:/Documents and Settings/HCJ/My Documents/_Conferences 2012/Symposium 2012/APR DRG DV Data Discovery/DATA_2/A Wren/Wren APR 201 Card Arrthym and Conductn Disords Jan2011 Dec2011 20112 04 28 1500.xlsx




• Web application: reviewed

SmartReport

• Hospital APR-DRG Ranking Profile

• Set parameters to Days Delta

• Take contents to Excel to view and

rank

• Hospital APR-DRG Service Line

Profile

• Drill down to a service line

• Days Saved Report – bring out to

Excel

• Estimate total Days Saved at 20th

percentile

• Check by Attending and Principal

Procedure Provider

• To server: build APR Indicators

• Run Toolpacks

Resource Use and Length of Stay:

Michael Memorial,

Orthopedics










To the Server:





DataVision

Navigator


DataVision

Web Application

Hospital Server




Michael Mem

APR DRG 301

Total Hip Repl

Christopher

APR DRG 301

Total Hip Repl

C:/Documents and Settings/HCJ/My Documents/_Conferences 2012/Symposium 2012/APR DRG DV Data Discovery/DATA_2/D Michael/Michael APR 301 Total Hip Repl Jan2011 Dec2011 2012 04 29 1330.xlsx

C:/Documents and Settings/HCJ/My Documents/_Conferences 2012/Symposium 2012/APR DRG DV Data Discovery/DATA_2/B Christopher/Christopher Mem APR 301 Tot Hips Jan2011 Dec2011 2012 05 09.xlsx



• Web application: review

SmartReport

• APR-DRG Service Line Profile

• Set parameters to Days Delta

• Review APR-DRG 313 . . . .

• Choose APR-DRG that is easily

“actionable”

• Look at AP-DRG 301, Total Hips

• To server: build APR Indicators

• Run Toolpack

Mortality: Vitale Memorial Percutaneous Coronary Intervention

• Consider:

– Does the hospital have a hospice unit?

• Recall that on the DataVision web application

– Acute Care is All Patients, excluding

» Obstetrics

» Rehabilitation

» Behavioral Health

» Hospice

– Does the hospital use code V66.7 Palliative Care?























To the Server:





DataVision

Navigator


DataVision

Web Application

Hospital Server


Vitale

APR DRG 174

PCI w AMI

C:/Documents and Settings/HCJ/My Documents/_Conferences 2012/Symposium 2012/APR DRG DV Data Discovery/DATA_2/C Vitale/APR DRG 174 Vitale Jan2011 Dec2011 2012 04 26.xlsx



• Review SmartReport, find high

mortality rate

• Run Comparative Trend Analysis

• APR-DRG Ranking Profile

• Facility/Mortality Data

• Hospice present

• Review SmartReport

• Facility/Mortality Data: No hospice

• Lives/Days Saved Report: APR 174

• Remove V66.7 coded patients

• View CTA and subclass detail

• Profile by Doctor

• To server: build Indicator, run

ToolPack

Questions:
