Injury Data and Resources
Minutes from the June 2004 Meeting of the International Collaborative Effort on Injury Statistics
Vienna, Austria, June 6, 2004
The meeting opened with introductory remarks including progress on the preparation of the 10 Year Review which we hope will be considered and accepted by both Journals- Injury Prevention and Injury Control and Safety Promotion with the thought being that both will publish it because of their different readerships. If not, we will submit to Injury Prevention first. The group was also introduced to the new injury web pages at NCHS that feature both injury data and resources from NCHS as well as the ICE on Injury.
Yvette Holder began the discussion on the strategic plan update. In addition to the Steering Committee whose role is to guide the functioning of the ICE, two “formal” committees are now in place: the Information/outreach sub-committee and the Research sub-committee. The job of the former is to promote ICE by soliciting material based on current work of the Injury ICE for quarterly columns for Injury Prevention; soliciting new members of the ICE and providing assistance, if necessary, to individual researchers as they seek funding from outside sources. The role of the “Research” committee was redefined so that individual projects (i.e., indicators, selecting a main injury, occupational injury) will function independently but that a Research Oversight committee (comprised of project chairs and some members of the Steering committee) will monitor activities of all projects and provide assistance as needed.
Updates to current projects activities
Barell Matrix-conversion to ICD-10 (Lois Fingerhut)
Because of the continued interest in using the Barell Matrix to present injury diagnosis data, there is a need to convert it to ICD-10 and subsequently to the various clinical modifications of ICD-10. The current status of the matrix is “draft under review”. Final work will be done once the US ICD-9/10 Comparability file is analyzed. In addition, consultation with Ted Miller will continue as he works with the Matrix for the US Cost of Injury Study.
Multiple Injury Profiles (Limor Aharonson Daniel)
MIPs are a tool based on the Barell Matrix for presenting injury statistics and for comparing injury patterns in different populations taking into account systematically details the physiological injury profile. The MIP’s theoretically should eliminate all AIS=1 injuries but in the absence of ICDMAP for ICD-9 this was not possible. The MIPs were built based on the AIS 6 body regions --Head, Face, Chest, Abdomen, Extremities, External. Question- should the effort be put into excluding the Abbreviated Injury Scale (AIS) 1 injuries?
Discussion: Why not just look at the first three injuries rather than all? Can ICISS be used rather than AIS?
Selecting a main injury from multiple causes (Margaret Warner)
With continued interest from the WHO-based Mortality Reference Group, work continues on this project with the current emphasis on identifying common pairs of ICD codes that appear in multiple cause-of-death files. Data from additional countries will be solicited. Consideration is being given to examining the underlying cause of injury death in an effort to better understand which injury diagnosis code(s) led to the death (e.g. firearm injury as the underlying cause with open wound of head as one of the diagnoses).
Discussion: When it is not be possible to determine which injury led to the death, the severity of the injuries will be used to select the main injury. Data need to come from more countries.
Both of the two preceding projects are similar in that they are attempting to define “severe” injuries. ICE participants will work together on these.
Occupational Injury (Nancy Stout)
The Occupational Injury group has made use of the AdvICE list serve during the year in their effort to try and broaden the scope of their work to other countries. Unfortunately, no other country (besides US, Australia and New Zealand) was able to provide comparable data. In addition, a question was posed on the use of ICD-10 activity codes and it seems that no country has had success.
Discussion: Great efforts went in to getting activity codes into the ICD. Hospital data may provide a better source for this
Indicators (Colin Cryer)
Work is ongoing to apply the validation criteria to injury mortality and morbidity data from New Zealand. To arrive at valid indicators of serious nonfatal injury, the ICISS (ICD-based Injury Severity Codes) methodology has been applied. ICISS uses data that are ICD-10 based to derive severity scores for injury diagnoses. ICISS scores are survival probabilities calculated for every ICD-10-AM diagnosis. This allows interpretation of trends in the incidence of serious injury rather than having the measure of injury confounded with access to care and utilization issues. Only cases that are most likely to be hospitalized are counted. Based on this work, trends in mortality and serious injury were expected to be similar, but this was not found to be the case. Several possible explanations were offered including the possibilities of data problems; preventive activities selectively preventing deaths but not serious injury; case fatality rate falling – i.e., effect of better treatment (e.g. Roberts BMJ 1996) or getting more seriously injured people to hospital before they die.
Discussion: It was suggested that subgroup variance accounts for much of the problem. Also, many instances of persons with serious injuries are either not hospitalized for whatever reason or die before getting to the hospital. Other possible sources of data are from emergency departments- i.e. data on specific kinds of fractures. Need for valid measures of change over time.
Possible new project: Injury Severity (Lois Fingerhut and Ellen MacKenzie)
Recognizes that AIS continues to be the standard for injury severity, but the link to the ICD is based on ICD-9 CM via ICDMAP (proprietary). The AIS injury descriptions are used to link to the predictive values of the Functional Capacity Index (FCI). Emerging interest was recognized in ICISS for measuring threat to life and its usefulness in large administrative data sets. Results from Australia and New Zealand are encouraging in terms of its comparability across countries and across versions of ICD. ICISS offers an alternative to AIS but it needs validating. U.S. is planning on convening a meeting in September to discuss its use further.
Discussion: How comparable are Survival Risk Ratios (SRR’s) over time? Co-morbidity needs to be analyzed; ICISS is not useful for predicting threat to disablement. There was general agreement that ICE should work towards a “project” looking at data from several countries.
Household Surveys (Margaret Warner)
A paper on the comparability of injury-related questions and survey methodology (e.g., recall period) was submitted to Injury Prevention.
Discussion: The WHO’s I&P program is publishing guidelines for community health surveys. There was some coordination between the two projects and there was discussion about the mutual benefits of that partnership and the need for further collaboration between ICE on Injury and the WHO’s I & P program. It was pointed out that EUROSTAT is developing a database of household surveys. We will try to collaborate.
Poisoning definition (Lois Fingerhut and Clare Griffiths)
Two issues were raised: 1) should injury matrix be reconsidered in how it defines poisoning deaths and 2) should we recommend that WHO disseminate data in the format of external cause of injury matrix. The first issue deals with the use of F10-F19 in the mental health disorders chapter of ICD-10. Data from US, England and Wales as well as from Canada show the use of these F codes for deaths that are similar to X codes.
EU Public Health Program 2003-2008 (Birthe Frimodt Moller and Saakje Mulder)
The EU Public Health Programme (PHP) has only recently been launched. With regard to injury topics, the tasks will be carried out by a ‘Working Party on Accidents and Injuries’ (WPAI). A secretariat (located in Greece) will co-ordinate meetings and dissemination of information etc. along the same lines as developed during the Injury Prevention Programme, which was phased out during 2003. Actions made possible by the first grants awarded by the EU Commission in 2003 are just beginning to get off the ground. Our presentation shows the highlights of the present work programme and initiatives that may further develop the agenda of injury control and safety promotion in Europe.