http://optn.transplant.hrsa.gov/media/1161/ddps_03-2015.pdfHigh-Level Conclusions Based on OPO and CIG Subcommittee AnalysesBased on findings from the OPO and CIG subcommittee analyses, approximately 3
5,000-40,000 potential deceased organ donors may be available per year from inpatient hospital deaths. This estimate represents a 5-fold increase over the current state of transplantation, in which organs from about 8,000 deceased donors are transplanted each year. The estimates of unrealized potential likely include many imminent and eligible deaths already identified by OPOs, so it is important to recognize that not all of the “gap” between actual and potential donors is currently unrecognized as having donation potential. . Because of limitations in the available data sources, possible factors such as organ-specific laboratory measurements and serologies could contraindicate organ transplantation from a fraction of the estimated 35,000-40,000 donors, lessening this potential donor range estimate to some degree. Still, these two largely independent analyses, for which the results closely converged, strongly suggest that
significant donor potential exists that is not currently being realized.Further, the analyses have revealed that under current practice, unrealized organ donation potential exists across all age groups. About half of the estimated donor potential is currently being realized among decedents less than 35 years of age. However, less than 10% of the estimated potential is being achieved among decedents over age 64, and less than 20% for decedents ages 50-64. These results indicate that potential donors over age 50 may offer the most significant opportunity for substantially expanding the number of transplant donors available each year.
The aforementioned analyses focused on donation potential from deaths occurring in hospitals among patients admitted with inpatient status. OPO and CIG subcommittee analyses of outpatient and emergency department death records (Appendix B) found that around 3,000 additional potential donors may be available from this largely untapped source. Due to added uncertainties with emergency department data, however, this estimate cannot be cited with high confidence. Given this increased uncertainty and practical challenges associated with obtaining consent and recovering viable organs for transplantation in an outpatient setting, this estimate of 3,000 potential donors was not added to the overall estimate of 35,000-40,000 potential donors from inpatient deaths. Still, the outpatient and emergency department potential donor analyses highlighted that many deaths in the emergency department are younger individuals involved in accidents or victims of stroke, both causes of death associated with higher likelihood of organ donation viability.
Of course, to achieve an expansion in deceased organ donation,
changes in clinical practice are imperative. Key practice areas that should be revisited in light of these findings include, but are not limited to, the accurate recognition of potential organ donors by referring hospitals and OPOs, as well as transplant centers' willingness to accept less-than-ideal organs. Centers’ willingness to accept organs that would confer a net benefit to their patients, vis-a-vis remaining on the waitlist, is a key practice area with the potential to increase donor supply. Revision of OPO and transplant program performance metrics and/or the way they are used may be needed to spur changes in practice that would result in the increased use of older donors, where a large fraction of unrealized donor potential resides.
Projecting the Future Number of Potential DonorsBased on results from multiple forecasting methods, t
he number of potential donors is not anticipated to change substantially over the coming decade. The most likely outcome is an approximately 5% increase in donors between 2010 and 2020, since similarly shallow rates of change (about 0.5% per year) were predicted by two independent approaches.
Under assumptions most closely aligned with stakeholder recommendations (e.g. length of stay ≤ 14 days) and an empirically derived age ≤ 75 cutoff, approximately 35,000 to 40,000 potential donors were estimated to be available in 2010. Applying a 5% growth rate over a decade to this range estimate results in a forecast of approximately 37,000 to 42,000 potential donors in 2020.
As illustrated by confidence and prediction intervals for both the CIG and OPO subcommittee- based regression based forecasts, inherent statistical uncertainty is present in these forecasts, due to residual year-to-year variation. The uncertainty is greater for forecasts further into the future.
Although the forecasted number of potential donors available in 2020 contains uncertainty, the fact that two independent methods resulted in similar estimates increases confidence that the rate of change in potential donors is likely to be modest over the next 10 years. Furthermore, though the findings from the CIG’s regression-based approach have been deemphasized due to the high degree of statistical uncertainty, that approach forecasted 43,854 potential donors in 2020, a prediction not very far from the range estimate of 37,000-42,000.
Perspectives and Recommendations Drawn from this StudyThe organ transplantation system in the United States is characterized by wide variation in transplant program and OPO characteristics. There is significant variation in size, organizational structure, institutional resources, financial resources, and populations served, which vary by demographics such as urban versus rural, ethnicity, race, and economics. OPOs differ by the number of local transplant programs served, and the number and type of donor hospitals that comprise their referral base. There are significant challenges produced by variations of geography. Transplant programs similarly differ in size and resources and by various patient demographic factors. The organ transplant system is also complicated by layers of regulatory oversight that results from the requirements of multiple organizations including the OPTN, CMS, the Joint Commission on Hospital Accreditation, and third-party payers. This complexity results in significant challenges for OPOs and transplant programs in setting attainable goals and uniform clinical practice patterns. The development of uniform national transplant policy directed to achieving agreed upon system wide goals, while adequately accounting for these wide and largely immutable variations, is a significant and important challenge.
Major conclusions of the work reported here are several.
First, it is clear there is significant deceased donor potential that is currently unused. It is important to acknowledge that some of this potential is already recognized by the transplant community under current practice but may not be realized due to factors such as inability to obtain consent, delayed referral from the donor hospital, and risk aversion among both OPOs and transplant programs. Some of the unrealized potential may also be unrecognized under current practice, with organ donation not pursued due to misperceptions of the suitability for donation and concerns about impact of pursuit on performance metrics.
The majority of the untapped deceased donor potential likely resides in the donor population older than fifty years of age, although there is unrealized potential in all age groups. Potential donors in older age groups present challenges to both OPOs and transplant programs because of co-morbid conditions affecting organ quality and likely lower numbers of organs per donor. Making optimal use of organs from these donors will likely require changes in current policy by the OPTN and CMS, as well as changes in clinical practice. Additionally, an important study finding is that based on current policy and practice, there is only minimal projected growth of the potential donor population through 2020. A preliminary analysis of geospatial patterns in actual and potential donors8 revealed that unrealized donor potential does not appear to be uniformly distributed geographically.
Current performance goals for OPOs arose from the work of the Organ Donation Breakthrough Collaborative. These include aggressive goals for the number of organs transplanted per donor, increasing the percentage of DCD donors, and increasing the donor conversion rate. OPO performance oversight also results from analysis of data by the SRTR resulting in statistically- derived expectations for evaluating both the total number of organs transplanted from each donor, as well as organ–specific transplant counts. OPOs that fall statistically short of expected results are flagged for review by the OPTN Membership and Professional Standards Committee, which is charged by HRSA with program performance oversight. Although the Breakthrough Collaborative initiative resulted in significant increases in organ donor procurement in the early 2000’s, this progress has since stalled and the annual number of organs procured and transplants performed in the United States has stagnated since approximately 2006. It is believed by many stakeholders that the current structure of OPO goals does not reflect the medical characteristics of the current donor population and provides a disincentive to maximizing the number of organs procured. (13)
The performance metrics applied to transplant program evaluation as well as the current third- party payer environment also affect transplant program clinical practice. Though statistical risk adjustment accounts for much of the disparity in demographic and clinical characteristics associated with transplant outcomes, this approach still results in a conservative risk-averse clinical practice, which translates into lower numbers of transplants performed. (21) OPO and transplant center performance are not independent: increased risk aversion among an OPO’s local transplant center(s) may affect the subset of potential donors the OPO chooses to pursue. Policy changes and resulting clinical practice changes go hand in hand to affect both OPO and transplant center performance.
Policy Related Recommendations1. Set specific, attainable, evidence based, performance goals for national transplant policy to emphasize an increased numbers of transplants.
2. Set evidence based policy goals and metrics for OPOs designed to increase the number of transplantable organs procured by removing disincentives to procurement of less than ideal donors. Current measurement of OPOs based on conversion rates and organs per donor may limit pursuit of donors likely to yield lower organ numbers, despite statistical risk adjustment.(13) Much of the estimated unused donor potential resides in older donor populations where donation is less likely to yield multiple organs.
3. Revise transplant center performance metrics to allow increased use of organs from less than ideal donors. This should be in conjunction with revision of OPO metrics encouraging procurement in this population as well. Broader focus on program structure and resources and the use of pre-transplant metrics may result in transplant programs meeting a goal of increased transplant numbers. Performance goals should move beyond the current overemphasis on patient and graft survival among the subset of patients that receive transplants, to also include performance in meeting the needs of the population of end-stage organ failure patients through by far the most effective treatment modality, transplantation. This may result in increased patient life years following diagnosis with organ failure.
4. Explore strategies to link OPO and transplant center performance assessments.
5. Harmonize regulatory policy to the extent possible, accounting for the specific needs of the OPTN, HRSA, and CMS. Additionally, policy should emphasize mutual accountability among OPOs, transplant programs, and donor hospitals.
6. Devise policy to promote early recognition of potential donors and timely referral from donor hospitals.
7. Adjust OPO and transplant center goals to account for geographic and demographic variation in potential donor populations and in recipient populations.
Practice Related Recommendations1. Identify OPO best practices, including further exploration of geographic disparities in donor realization rates, and devise strategies to disseminate these to the OPO community and stakeholders.
2. Identify transplant program that have successfully used marginal organs, achieving good outcomes. Identify their best practices and disseminate these to other transplant centers.
3. Develop predictive algorithms that can assess the outcomes of specific OPO practice changes.
4. Develop predictive algorithms that can be used to predict the outcomes of specific transplant program clinical practice changes.
5. Update educational efforts promoting early recognition and donor referral by donor hospitals to reflect current clinical conditions. This may permit the identification of a wider population of potential donors by donor hospitals.
6. Increase the use of DCD donors and evaluate new technology designed to improve organ quality.
Recommendations Related to Future Research
Further Research Aimed at Refining and Improving Potential Donor Estimation1. Refine filtering assumptions through augmenting or linking national databases.
Since there is no information in either the NCHS or NIS datasets indicating which decedents actually became organ donors, it is not possible to formally validate the filtering assumptions used to estimate donation potential. Such a confirmation would allow an assessment of whether a significant number of donors having one or more exclusionary criteria (or not having any inclusionary criteria) are being recovered and used for transplantation under current practice. If that were the case, such a validation exercise would suggest modifications in the filtering logic that could further refine the donor potential estimate. In other words, if certain types of cases identified through the filtering process as potential donors in reality never actually resulted in donation, refinements to the filter could be made by revisiting key assumptions and revising exclusionary and/or inclusionary factors.
Augmenting either the NCHS or NIS to include information on whether each death resulted in donating organs for transplantation would allow such a validation exercise. Alternatively, linkage of either of these national datasets to OPTN data would also identify decedents that became donors and help this type of analysis.
Additionally, opportunities to incorporate more detailed clinical information such as laboratory measures, serology results, and whether death was determined by cardiac or neurological indication, into national databases would also be valuable for more precisely identifying donor potential.
Finally, more complete collection of data on clinical conditions and use of procedures among emergency room decedents is needed. Although hospital emergency room decedents may represent an untapped source of potential donors, application of the donor potential filter to the NEDS data identified relatively few deaths that could potentially lead to donation. Emergency department data are known to be inaccurate and incomplete, however, limiting confidence in these estimates. Further, most existing practices in organ recovery are set-up for the in-hospital area of clinical practice. Subject matter experts on the OPTN DDPS Stakeholder Committee members expressed a desire to further investigate the clinical practices within emergency departments for opportunities to better clinically integrate this area into the field’s practice.
2. The impact of changes in billing and coding practices should be further studied to better understand their impact on the reporting of diagnosis and procedure codes in national administrative databases.
Coding (reporting of diagnosis and procedure codes) practices are known to change over time, and may be influenced by reimbursement issues and changes to Medicare Diagnostic Related Groups (DRG’s). Since these codes are integral to the methods used in this study, changes in the way they are reported could influence the overall estimates and trends in donor potential. The NEDS data in particular is believed to suffer from underreporting of procedure and secondary diagnoses. Understanding the nature and degree of reporting changes over time in both the NIS and NEDS would allow for better understanding of the limitation of the results and possibly enable further refinements.
3. Pursue a better understanding of the surprising increase in ICD-10 “R99” codes reported for three states (New Jersey, Ohio, West Virginia) in 2009.
Researchers from the NCHS documented an unexpectedly large increase in the number of deaths with ICD-10 code “R99: Other ill-defined and unspecified causes of mortality)” In three states. The increase in R99 codes for these three states resulted in fewer deaths meeting the potential donor filter’s inclusionary criteria. This issue affected both the estimated number of potential donors for 2009, as well as the forecasts of potential donors. So a better understanding of whether the increase in “ill-defined and unspecified” causes of death was a real phenomenon or the artifact of a coding or data processing problem is important for more precisely assessing the reliability of estimates produced using NCHS data.
Research Related to Characterizing Organ Donation Potential1. Further explore gaps in the estimated “donor realization rates” beyond just decedent age.
The OPO Subcommittee’s analysis revealed that more unrealized potential for deceased donation may exist among older (age 50+) potential donors as well as very young donors (<1 year old Preliminary analyses also found large disparities in organ donor realization in different parts of the country, as well as among different ethnic and racial groups. Since these analyses revealed great differences in the current use of potential donors by decedent characteristics, it is likely that analyses of other factors – cause of death, gender, etc. – would reveal disparate utilization gaps as well. Interactions between geography and decedent characteristics may be particularly powerful in identifying areas in which deceased organ donation may have the greatest opportunities to expand.
2. Analyses of whether certain types of hospitals (e.g., large, urban, teaching) have higher proportions of potential donors relative to total inpatient deaths could provide significant insights into opportunities to expand organ donation. Further research is needed to ascertain whether institutional factors that vary between regions and jurisdictions such as diagnostic guidelines, referral procedures and patterns of practice in hospitals and/or among coroners and medical examiners might be affecting the identified pool of medically suitable potential donors.
3. Pursue an increased understanding of combinations of factors that substantially lower clinicians’ perception of organ utility.
Both OPO and CIG Subcommittee surveys asked clinicians about the acceptability for transplant of organs from various types of donors. Though these surveys asked for responses separately depending on whether the decedent had the potential for brain death donation or donation after circulatory death, the surveys did not incorporate combinations of factors that together might result in some clinicians considering the organs unsuitable for transplantation. In fact, several respondents to the OPO Subcommittee’s survey suggested (in free-form, text responses) various combinations of factors that they would consider rule-outs to donation. Both cumulative effects of many factors together and interaction effects may exist. A survey instrument that allows the user to indicate what combinations of factors [for example, elevated age and questionable social history and U.S. Public Health Service (PHS) high risk status] could help identify such combinations, which could then be incorporated into a more refined potential donor filter.
A more interactive approach, where individuals are presented case studies in a simulated setting designed to mimic the environment in which clinicians receive organ offers, might provide even more insightful information. Profiles of donors with varying characteristics would be shown to the clinician, who would have to decide whether to accept the offer for one or more of his hypothetical waitlisted patients. Donor profiles would be varied systematically such as through the use of conjoint analysis, for example, to identify key factors and interaction effects that lead to the decision to accept or decline organ offers.
Finally, regression analysis of OPTN data could be used to complement the above approaches. OPTN data have been analyzed to determine factors and combinations of factors that influence the likelihood of organs to be discarded, as well as the likelihood of organ offers to be declined by individual transplant centers (23-25). Future refinements could focus on identifying combinations of factors associated with high rates of organ offer refusal or discard but not having significantly adverse effects on graft (delayed graft function, long-term survival) and patient (complications, long-term survival) outcomes.
Opportunities for ActionIt is clear from this study and many previous ones that the transplant community is far from achieving the deceased donor potential in the U.S. This can be deduced also from the wide DSA9 and region-
specific variability in various performance measures that quantify donation, recovery, and transplant rates across regions and institutions. (22) Where there is variability, there is opportunity for improvement. A gap between the potential number of donors and the actual number undoubtedly exists in all regions and DSAs. Certainly the gap is larger in some areas than in others.
Opportunities for action lie within direct control of the transplant community and could be pursued as a collaborative effort by a multidisciplinary group of transplant professionals as well as representatives of CMS and HRSA. One such opportunity lies in the refinement of donation-related definitions currently used for quantifying categories of donor potential, particularly what it means to be an “eligible” donor. There is broad agreement, supported by the results of this study, that many more deaths are eligible for organ donation than are included in the current OPTN definition of “eligible death.” The total pool of potential donors includes not only individuals declared brain dead, but also those whose conditions are consistent with brain death as well as donors after cardiac death.
In areas of the country where data suggest the gap between potential and actual could be larger than others, death records and referral data could be mined and further analyzed. Data sources could be regional or national for the purpose of quantifying 1) deaths that, under revised donation definitions, could be considered eligible deaths; 2) eligible deaths not referred; and 3) referrals that did not become donors. Data linked to identified for selected areas of the country could be used to identify the types sorts of donors and reasons for non-donation that are specific to each DSA, for the purpose of regional collaboratives and performance improvement initiatives aimed at bridging donor potential gaps in each area.
Another opportunity lies in cross-organizational discussions that have already begun, for the purpose of examining any unintended consequences of various policies, metrics, and systems on organ recovery and use. In various contexts, organizations including the OPTN/UNOS, HRSA, the Alliance, AOPO, CMS, the SRTR, AST, ASTS, and others are meeting and planning how to effectively pursue, together, our the mutual goal of increasing the number of safe, high quality, timely transplants for patients with end stage organ disease. This work needs to continue quickly and conclude with decisions that can be implemented and make a significant difference in the number of effective transplants.
Topic: Deceased Donor Potential Study Results: Action on Unrealized Potential (Read 3328 times)