1 Measuring Patient Safety and Disparities using the Medicare Patient Safety Monitoring System (MPSMS) 9 th Annual Maryland Patient Safety Center Conference, April 5, 2013 Noel Eldridge, MS Agency for Healthcare Research and Quality Center for Quality Improvement and Patient Safety [email protected]301-427-1127 1 MPSMS Background Key Aspects of the Medicare Patient Safety Monitoring System (MPSMS) – Based on chart review by abstractors at CDAC 21 specific measures – Patient charts come from RHQDAPU and IQR Abstraction with MPSMS tool comes after other reviews – Sample was: all-diagnosis (18+) Medicare for 2002-2007 All payer (18+) 4-principal diagnosis groups (AMI, HF, SCIP, Pneumonia) for 2009-2011, Potentially “global” (18+) starting in 2012 2
9
Embed
Disparities in Patient Safety - Presentation from 2013 Maryland Patient Safety Center Conference
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
1
Measuring Patient Safety and Disparities using the Medicare Patient Safety Monitoring
System (MPSMS)
9th Annual Maryland Patient Safety Center Conference, April 5, 2013
Noel Eldridge, MSAgency for Healthcare Research and Quality
– AEs associated with femoral artery puncture for angiography
– AEs after Hip surgeryCAUTI
– CVCBSI
– VAP
– C. difficile
– MRSA
– VRE
– Post-op pneumonia
– AEs after Knee surgery
– Post-op cardiac events
Others Pressure Ulcers
Falls
3
Role of MPSMS data in HHS’s Partnership for Patients
MPSMS supplies over 90% of the measured Hospital-Acquired Conditions (HACs) that we are using to track q ( ) gnational changes from 2010 thru 2013
ADEs, CAUTIs, CLABSI, Falls, Pressure Ulcers, VAP, & VTE (7 of 9 focus areas) will be measured at the national level with MPSMS
SSIs are based on CDC’s NHSN data
Obstetric AEs are based on AHRQ PSIs
Additional MPSMS and PSIs are used to represent “all-other” HACs in the estimate
Baseline is 145 HACs per 1,000 discharges in 2010
– based on MPSMS, PSI, NHSN and HCUP data4
3
PFP-Measured National HACs Baseline Pie Chart (2010)
Percent of Total Measured HACs –PFP 2010 Baseline (4.745M)
Adverse Drug Events (57% Hypoglycemic Events & 42% Anticoagulant Drug Events)Pressure Ulcers
34.2%
5 5%1.7%
0.8%
0.4%0.3%
18.8%
Catheter-Associated Urinary Tract Infections
Falls
Surgical Site Infections
Obstetric Adverse Events
27.8%
8.4%
5.5%2.0% Ventilator-Associated Pneumonia
Central Line-Associated Bloodstream Infections
Venous Thromboembolism
All Other HACs -- based on 14 other specific measures (from C diff Infection to Contrast Nephropathy) 5
National Measurement –Hospital-Acquired Conditions (HACs)
We have good measures for the 9 HACs but they are not perfect; forHACs, but they are not perfect; for example…– VTE measure is post-op only
– ADE measures miss many ADEs – e.g., those due to narcotics or allergies, etc.
– CAUTIs or CLABSIs caused by inpatient fprocesses but manifesting post-discharge cannot
be counted
– SSI measure covers only 17 selected procedures
– OB measures count injuries to mother only
6
4
Disparities in 2002-2007 data(background)
Publication by Metersky, et al (including Moy) in Medical Care (May 2011) covered 2004 2007Medical Care (May 2011) covered 2004-2007 data
Design and Subjects: Abstraction of 102,623 Medicare charts of non-Hispanic white and black patients to assess frequency of patient safety events in 4 domains: – adverse drug events due to anticoagulants and
hypoglycemic agents,
– selected nosocomial infections,
– selected procedure-related adverse events,
– general (pressure ulcers and falls) 7
Disparities in 2002-2007 data(summary findings)
Blacks had a higher risk than whites of suffering one of the nosocomial infections (1 34; 95%one of the nosocomial infections (1.34; 95% confidence interval, 1.17-1.55; P < 0.001) and one of the adverse drug events (1.29; 95% confidence interval, 1.19-1.40; P < 0.001).
Patients in hospitals with the highest percentages of black patients were at increased risk of experiencing one of the nosocomialinfections (1.9% vs. 1.5%; P < 0.001) and adverse drug events (8.7% vs. 7.8%; P < 0.01).
8
5
A quick look at potential disparities in 2010 -2011 data
Preliminary analyses of all measures– Years and types of events not aggregated
Measured grouped as follows: ADEs, HAIs, Surgery & Procedures, General
Two sets of slides for each grouping of measures:measures:– Exposure Rates
– Adverse Event Rates among those exposed
9
ADE “Exposure” Rate(Preliminary Data)
This is a measure of the percentage of patients that are subject to the AE, i.e., mostly a measure of those that “get” the drug.
15%
20%
25%
30%
35%
40%
45%
50% Exposure rate (%) WHITE PATIENTS 2010 (12-month, n=27528)
Exposure rate (%) WHITE PATIENTS 2011 n=27699
Exposure rate (%) BLACK PATIENTS 2010 (12-month, n=4088)
Pressure Ulcer and Falls Adverse Event Rate (Preliminary Data)
In MPSMS 100% of patients are considered “exposed “ for Pressure Ulcers and Falls. Pressure Ulcers are of all stages (I to IV), and Falls i l d ll f ll ( t j t f ll ith i j )include all falls (not just falls with injury).
(Preliminary Data)
3.0%
4.0%
5.0%
6.0%
Observed adverse event rate (%) WHITE PATIENTS 2010 (12-month, n=27528)Observed adverse event rate (%) WHITE PATIENTS 2011 n=27699Observed adverse event rate (%) BLACK PATIENTS 2010 (12-month, n=4088)Observed adverse event rate (%) BLACK PATIENTS 2011 n=4086
Observed adverse event rate (%) OTHER PATIENTS 2011 n=2182
16
9
Discussion Items
MPSMS disparities studies are constrained by the small sample size forconstrained by the small sample size for racial/ethnic subpopulations– And many measures only pertain to even
smaller subgroups, e.g., CVC patients, knee surgery patients, warfarin patients, etc.
– Combining multiple years of data for severalCombining multiple years of data for several related measures (e.g., all ADEs) reduces the ability to see trends, but may provide more insights re disparities (as was done for 2004-2007) from 2010 to date