Transcript SRTR Transplant Program Quality and Surveillance

Consensus Conference on
Transplant Program
Quality and Surveillance
Arlington, VA
February 13-15, 2012
Consensus Conference on Transplant
Program Quality and Surveillance
Co-Chairs
Bertram Kasiske (SRTR) and Maureen McBride (OPTN/UNOS)
Steering Committee
Robert Gaston (AST)
Mitch Henry (ASTS)
Thomas Hamilton/Karen Tritz (CMS)
Danielle Cornell (ACOT)
Kenyon Murphy (Patient/Public)
Funding
HRSA
Dennis Irwin (Optum Health)
Nancy Metzler (OPTN-TAC)
John Roberts (OPTN-MPSC)
Alan Reed (OPTN-PAIS)
Stuart Sweet (OPTN-POC)
Key Questions
•
•
•
•
•
•
•
•
What is the SRTR’s mandate?
Who uses PSRs and why?
Are there unintended consequences?
What can we learn from others?
What statistical methods should we use?
How should we adjust for risk?
What outcomes should we use?
What data should we collect?
SRTR Actions Following the Consensus
Conference
Consensus Conference: 2/13 - 2/15
Reviewed the recommendations with the STAC on 2/23.
Developed a prioritized list of actions.
Forming an STAC Subcommittee to oversee PSR
development activities.
Immediate Priorities
1) Produce an SRTR PSR Manual of Operations detailing the
SRTR process for risk model development and oversight.
2) Explore using a 3-year cycle of model development with
formal input and involvement from the OPTN committees.
3-year Model Building Cycle
Year 1
• Kidney
• Pancreas
Year 3
Year 2
• Liver / Intestine
• OPO Yield
• Heart
• Lung
Immediate Priorities
3) Continue to explore making the reports more
accessible/friendly to the public.
Full graphical PSR was provided to transplant programs in June 2012.
Current Summary Page
Graphical Summary Page
Immediate Priorities
4) Add funnel plots to the new graphical PSRs.
To be included in the January 2013 PSRs.
Funnel plots
Immediate Priorities
5) Add time trend data to the new graphical PSRs.
6) Provide more data on waiting list risk and outcomes.
Targeted improved waiting list metrics for the January 2013 PSR .
7) Move forward with production of CUSUM charts to be
included on the programs’ secure SRTR website. Possibly
start with kidney and/or liver programs.
Explore methods and production options during Q3 & Q4 of 2012.
Longer-Term Priorities
8) Study Bayesian Hierarchical Modeling Strategies.
Explore as part of the first tasks for the STAC-PSR subcommittee.
Methods: Use of hierarchical models with
(Bayesian) suggested performance criteria
Christiansen CL, Morris CN. Ann Intern Med. 1997;127:764.
Longer-Term Priorities
9) Study setting missing data values to the least beneficial
value when risk adjustment models are applied.
10)Study inclusion of short-term living donor outcomes in
the PSRs.
Possible inclusion in January 2013 PSR for liver and kidney donors.
11)Study Life-Years from Listing (LYFL) for inclusion in the
PSRs.
Longer-Term Priorities
12)Study using period analyses to estimate survival rather
than the historical cohorts currently used. This will
possibly enable estimates of long-term patient
outcomes based on the most recent data available.
Incorporated some into the upcoming SRTR/OPTN annual data report.
13)Study measures of center “aggressiveness” for inclusion
in the PSRs.
14)Study inclusion of data regarding what future PSRs will
look like given events already observed. Likely for the
secure site only.
Explore modifications to the Excel-based tools.
Longer-Term Priorities
15)Study Balanced Score Card approaches, including
the current CPM under study by MPSC and
extensions of the CPM to include pre- and posttransplant metrics.
16)Continue to support the MPSC’s exploration of
alternative flagging methodologies, possibly
incorporating results from the Bayesian modeling.
Longer-Term Priorities
17)Work with the OPTN to explore the concept of excluding
patients from that PSR cohorts that are in approved
clinical studies.
Discussed at the March 2012 MPSC meeting; John Roberts and Alan
Reed are working toward assembling a subcommittee of the MPSC to
explore this idea.
Longer-Term Priorities
18)Explore the potential of additional risk adjustment for:
a. More detailed data on donor/organ risk, possibly including
more data from DonorNet®.
b. Socioeconomic status (SES) based on recipient ZIP code
c. Ventricular Assist Devices
d. Coronary heart disease, revascularizations (needs additional
data collection by OPTN)
e. Peripheral vascular disease, revascularizations, amputations
(needs additional data collection by OPTN)
f. More detailed data on diabetes (needs additional data
collection by OPTN).
Longer-Term Priorities
19) Study measures of organ function, e.g. acute rejections,
eGFR, FEV1.
20) Study use of external data sources to estimate rehospitalization or cause-specific post-transplant
hospitalization rates, e.g., Medicare data.
21) Study inclusion of data-completeness indicators in the PSRs.
22) Study the possibility of including quality of life as an outcome
(currently no data available to the SRTR).
Consensus Conference Manuscript