GBD Expert Group Consultation
Meeting Minutes
When and Where: Jan 22-23 2009, Institute for Health Metrics and Evaluation, Seattle, WA. Followed by a side meeting on Jan 24th.
Attended by: GBD core team members; James and Kavi representing the injury expert group; and leaders of several other expert groups (Water Sanitation and Hygiene, Haemoglobinopathies, Genitourinary, Maternal causes, Vision loss, and Occupation risks). These notes were written by James and Kavi. Thus, "we" in the notes below is shorthand for "James and Kavi".
Purpose of meeting: Given the diversity of approaches being used by the 43 expert groups in the GBD project, this meeting sought to provide early peer-feedback on the methods being used for epidemiological estimation. The GBD core team and leaders of 10-12 expert groups participated in each of four meetings in January and February 2009. In addition to presentations by the expert group (described below), the GBD core team also presented results of their cause of death and mortality work. For more details, see meeting agenda posted on the injury expert group website.
Key issues pertaining to the injury expert group:
- Proposed external cause list: Prior to this meeting, we (James and Kavi) were concerned that we had received no response from the core team regarding our submitted external cause list (flowing from Discussion Topic 12). Thus, we gave emphasis to this topic in our presentations with the aim of conveying our rationale (e.g. the advantages of the matrix approach for reporting in terms of intent and mechanism). Our assessment is that the core team appreciates the merits of our proposal. The main remaining issue with implementation is the availability of suitable data sources from enough regions.
- Upcoming deadline for reporting preliminary incidence estimates: The expert group is required to report preliminary estimates of incidence of injuries by region, age, sex, external cause and nature of injuries in May 2009. This implies that the expert group needs to push ahead with gathering data sources for use and developing the process for deriving population incidence estimates. But please note that we expect that May 2009 is only a preliminary deadline, and that it will be possible to incorporate data sets received well after May 2009. (We don't know the final; we will inform you when we do)
- Other issues discussed: Data hunt process; health state descriptions.
In general, the core team was impressed by the effort and the thoughtful deliberations that have gone into the injury expert group work this far.
Side Meeting - Saturday Jan 24th
When and Where: Jan 24 2009, Institute for Health Metrics and Evaluation, Seattle, WA.
Session 1: Meeting with Harborview Injury Prevention and Research Center (HIPRC) group
Key participants: Fred Rivara, David Zonies, Beth Ebel, James Harrison and Kavi Bhalla
Summary: Discussion covered a wide ranging of issues related with the GBD project. In particular, the HIPRC group offered to help connect with several groups that could potentially help access data sources for the project. It was decided that the HIPRC group would work closely with James to develop the approach for estimating injury morbidity. HIPRC would implement the approach for estimating the morbidity from burn injuries and the approach developed would be generalized to other injury topics.
Session 2: Meeting with cause-of-death subteam of the GBD project
Key participants: Rafael Lozano, Mohsen Naghavi, Theo Vos, James Harrison and Kavi Bhalla. Fred Rivara and David Zonies stayed back for part of this meeting.
Summary: The analytical work being undertaken by the expert group overlaps and compliments much of the work being done by the cause-of-death subteam of the core group. Thus the primary purpose of this meeting was to harmonize our current work and plan future collaborative work.
Mohsen described the current default approach for handling poorly specified ICD codes in mortality data. We left the meeting with the understanding that the core team will send the expert group a spreadsheet that provides the details of the redistribution process so that the expert group can suggest modifications.
The external cause list was also discussed again in this meeting. Unfortunately, some distance still remains between the advice of the injury expert group and the views of the cause-of-death subteam of the project. We agreed that the expert group will continue to collaborate with the cause-of-death subteam of GBD to look at the pragmatic issues surrounding data sufficiency, amount of effort needed for generating the detailed tables, and some analytical issues related with modeling sparse data in the project.
Meeting Presentations
Introduction - Data Sources - Unspecifieds
External Cause List - Health State Descriptions