Cumulative fraction of states passing each policy over the study period.
eTable 1. Description of Types of State Policies for Organ Donation
eTable 2. Complete Parameter Estimates for Relationship between Policy Passage and Organ Donation Rates and Absolute Number of Transplants
eTable 3. Sensitivity Analysis-Policy Impacts with Narrow Time Window (1998-2010)
eTable 4. Sensitivity Analyses - Subgroup Analyses of Policy Impact by Living versus Deceased Donors Using Narrow Window (1998-2010)
eTable 5. Sensitivity Analysis-Policy Impacts with State-Specific Flexible Time Trends
eTable 6. Subgroup Analyses of Policy Impact by Living vs. Deceased Donors with State-Specific Flexible Time Trends
eTable 7. Sensitivity Analysis-Relationship between Policy Passage and Organ Donation Rates with Five-Year Time-lagged Effects
eTable 8. Sensitivity Analysis-Relationship between Policy Passage and Organ Donation Rates using Fixed Effects Binomial Regression Specification
eTable 9. Sensitivity Analysis-Relationship between Number of Policies Passed and Organ Donation Rates
eTable 10. Sensitivity Analysis-Policy Impact of Donor Rates Stratified by Organ Type
eTable 11. Sensitivity Analysis-Policy Impact of Donor Rates Limited to Variables Most Germane to Each Outcome
eFigure. Trends in Organ Donors and Transplants at National Level from 1988 to 2010
Chatterjee P, Venkataramani AS, Vijayan A, Wellen JR, Martin EG. The Effect of State Policies on Organ Donation and Transplantation in the United States. JAMA Intern Med. 2015;175(8):1323–1329. doi:10.1001/jamainternmed.2015.2194
Copyright 2015 American Medical Association. All Rights Reserved. Applicable FARS/DFARS Restrictions Apply to Government Use.
Shortages in transplantable solid organs remain a critical public health challenge in the United States. During the past 2 decades, all states have implemented policies to increase organ supply, although their effectiveness is unknown.
To determine the effects on organ donation and transplantation rates of state policies to provide incentives for volunteer donation.
Design, Setting, and Participants
Using a quasi-experimental design and difference-in-differences regression analyses, we estimated the effect of policies in all 50 states and the District of Columbia on organ donors per capita and the number of transplantations from January 1, 1988, to December 31, 2010. Analyses were also stratified by type of donor (living vs deceased). Data were derived from the United Network for Organ Sharing. All data collection occurred between July 7 and September 27, 2013.
Policies of interest were the presence of first-person consent laws, donor registries, dedicated revenue streams for donor recruitment activities, population education programs, paid leave for donation, and tax incentives. Information on states’ passage of various policies was obtained from primary legislative and legal sources.
Main Outcomes and Measures
The number of organ donors and transplantations per state, per year, during the study period.
From 1988 to 2010, the number of states passing at least 1 donation-related policy increased from 7 (14%) to 50 (100%). First-person consent laws, donor registries, public education, paid leave, and tax incentives had no robust, significant association with either donation rates or number of transplants. The establishment of revenue policies, in which individuals contribute to a protected state fund for donation promotion activities, was associated with a 5.3% increase in the absolute number of transplants (95% CI, 0.57%-10.1%; P = .03). These associations were driven by a 4.9% increase in organ donations (95% CI, 0.97%-8.7%; P = .01) and an 8.0% increase in transplants (95% CI, 3.1%-12.9%; P = .001) from deceased donors as opposed to changes among living donors or transplants from living donors.
Conclusions and Relevance
Nearly all state-level policies to encourage organ donation have had no observable effect on the rate of organ donation and transplantation in the United States. The one exception was the establishment of revenue policies to promote organ donation, which may have led to small increases in organ donations and transplantations from deceased donors. New policy designs are needed to increase donation rates and curtail the widening gap between organ supply and demand.
The shortage in transplantable solid organs is a critical public health challenge in the United States. With nearly 124 000 patients requiring an organ transplant and 79 000 patients remaining on active waitlists for organs nationwide,1 more than 6000 patients are expected to die this year owing to organ shortages.2 Despite clinical and systems-based improvements in organ use,3- 8 current trends in population aging and the prevalence of chronic diseases with end-stage organ dysfunction are expected to worsen organ shortages.9- 11 This persistent gap between supply and demand has drawn the attention of physicians and policymakers, yet barriers to donation and effective organ transplantation remain.
Since the late 1980s, states have enacted numerous policies to increase organ supply. For instance, first-person consent laws that legally register donor intent prior to donation may have increased living donor rates in certain states.12 However, other policies, such as paid leave from employment for activities associated with donation or tax benefits for donors equal to the amount associated with an act of donation, have not increased organ supply.3,13- 15 It remains unclear whether these policies have influenced national trends in organ donation or transplantation. Furthermore, there have been no national data with which to evaluate the diversity or relative effectiveness of these policies. Such information is critical for policymakers seeking ways to increase organ donation and transplantation with limited resources to invest in these efforts.
In this study, we examined the effect of a variety of state policies on organ donation and transplantations in the United States for 2 decades. We studied 3 primary questions. First, what donation and transplantation policies did states adopt, and when? Second, which policies were related to increased organ donors per capita and transplantations per state, and were some policies more effective than others? Finally, did policies have a differential effect by donor status; that is, did some policies (such as paid leave and tax benefits) have a more pronounced effect among living donors while other policies (such as first-person consent and donor registries) have a larger effect among deceased donors?
We examined primary legal sources to create indicator variables for the passage of 6 major types of state policies designed to increase organ donation: (1) dedicated revenue pools comprising individual voluntary contributions and protected state funds for donor recruitment activities, such as community outreach campaigns, worksite campaigns, and hospital-based interventions to encourage donation and improve clinical practices for organ procurement among deceased donors; (2) population education programs on organ donation through classes in public schools or driver’s education programs; (3) leaves of absence for donors working in public and private sectors; (4) first-person consent laws registering individual consent for donation without family consent at the time of donation; (5) donor registries to document consent for donation; and (6) tax benefits for donors equal to the costs associated with the act of donation (eTable 1 in the Supplement). These policies were selected because they are recognized, defined, and publicly reported by Organdonor.gov, the US Department of Health & Human Services’ official site for organ and tissue donation. For each state, the presence of these policies and their passage dates were coded using qualitative data from the United Network for Organ Sharing and Organ Procurement and Transplantation Network databases,16 state-specific legislative codes, and WestLaw Next (a comprehensive legal database of federal and state laws). These sources were reviewed on a state-by-state basis and coded for the presence of each policy, passage dates, and other pertinent details using a standard data extraction template. Where discrepancies between sources existed, we relied on the primary legal documents and resolved coding questions through consensus. All data collection occurred between July 7, 2013, and September 27, 2013. This project was considered exempt from institutional review board approval under federal guidelines owing to the use of publicly available, deidentified aggregate data.
We included measures of economic environment, health systems, and risk of requiring transplantation in all statistical models. State income per capita and state population were acquired from the Bureau of Economic Analysis.17 The percentage of uninsured Americans and the percentage of Medicare patients per state were obtained from the Area Resource File.18 From the United States Renal Data System, we collected the number of patients with end-stage renal disease per state, per year during the study period.19
The primary outcomes were number of solid-organ donors (inclusive of living and deceased donors) per state population and number of solid-organ transplantations per state, per year from January 1, 1988, to December 31, 2010. This information was obtained from publicly available data from the United Network for Organ Sharing and Organ Procurement and Transplantation Network databases.16 Outcome data were limited to 41 states and the District of Columbia reporting the number of donors, as well as 47 states and the District of Columbia reporting the number of transplants.
We created cumulative adoption plots of the percentage of states that passed each policy during the study period. Summary statistics were calculated for the outcome variables of organ donors per capita and total number of transplants; control variables were also summarized.
We assessed the effect of state policies on the primary outcomes of organ donors per capita and transplantations using difference-in-differences regression models. Difference-in-differences models have been commonly used to study the effects of policies adopted at different times by treating the passage of a particular policy as a quasi-experiment.20 We used these models to estimate the differential change in organ donation and transplantation rates (outcomes) between states that passed policies (treatment group) and states that had not yet passed policies (control group). States that adopted policies during the period were classified as being in the control group until passage and thereafter in the treatment group.
Each donation-related policy variable was represented by time-varying dummy variables for the passage of a policy in a given year. We included state and year fixed effects, which control for time-invariant state-level confounders and nationwide secular trends in donation and transplantation rates, respectively, as well as state-specific linear time trends, which help capture state-level changes in policies, attitudes toward transplantation, and developments in medical technology across time. State and time-varying controls for the logs of state gross domestic product per capita, total population, population with end-stage renal disease, and percentage of the population without insurance were also included. All models were weighted by state population. Outcome variables were log transformed to represent a normal distribution. Coefficients were interpreted as the percentage change in the outcome between the year preceding and following the passage of a particular policy. Similar to prior quasi-experimental studies using difference-in-differences models, standard errors were clustered at the state level to account for correlation in the outcome variables.14,21,22 Finally, we performed subgroup analyses to assess whether policy effects differed for organ donation or transplantation from living vs deceased donors.
We performed several sensitivity analyses. First, we restricted our sample windows from 1998 to 2010 (when most states passed their respective policies) and from 1988 to 2008 (when most policy adoption had occurred). Second, we included state-specific quadratic time trends in the models. Both analyses are commonly used sensitivity checks that address additional unobserved, state-level, time-varying factors that may jointly affect policy adoption and donation and transplantation rates—the main sources of bias in difference-in-differences models.20 Third, to account for possible delayed policy effects, we allowed for up to 5-year lags of each policy enactment variable. Fourth, although ordinary least-squares regression is often suitable regardless of the distribution of the dependent variable,23 we also estimated models using fixed-effects negative binomial regression. Fifth, we recoded the main policy variable as a count of the total policies enacted in a given state and stratified analyses by organ type to account for possible composite policy effects. Finally, we performed a subset of analyses in which we modeled policies that were most directly tied to a specific outcome, such as modeling the effects of registries and first-person consent for donation from deceased donors only.
Donation rates and number of transplants were obtained for 41 states and the District of Columbia and 47 states and the District of Columbia, respectively, which represent the full set of states for which information is available in the United Network for Organ Sharing and Organ Procurement and Transplantation Network databases. Information on state policies was obtained for all 50 states and the District of Columbia. The number of states passing at least 1 donation-related policy increased from 7 (14%) to 50 (100%), with 6 states adopting all 7 policies by 2010 (Figure). Most policies were passed within the past decade, with all states adopting at least 2 policies as of 2010. The most commonly adopted policy during the study period was the passage of first-person consent laws (50 states and the District of Columbia), while the least common was granting leaves of absence from private sector employment (6 states). During the same period, the total number of organ donors increased nationally by 245.4% (from 5909 to 14 504), while the number of transplants increased by 140.3% (from 9617 to 13 490) (eFigure in the Supplement). The mean (SD) number of donors per state, per year was 4.6 (5.5) per 100 000 residents, and the mean (SD) number of transplants was 7.6 (6.8) per 100 000 (Table 1).
The passage of revenue policies was associated with a 5.3% increase in the number of transplantations (P = .03) (Table 2). This increase is equivalent, on average, to 15 additional transplants per state, per year. There was a trend toward an increase in donation rates and transplantations associated with leave-of-absence policies for public employees, but this increase did not reach statistical significance (4.1% increase in donation rates: P = .05; and 7.1% increase in transplantations: P = .06). No other policies had a statistically significant association with either donation rates or transplantations. The full set of coefficient estimates, including the covariates, is provided in eTable 2 in the Supplement.
Revenue policies were associated with a 4.9% increase in the number of deceased donors per capita (P = .01) (Table 3), as well as an 8.0% increase in the number of transplants of organs from deceased donors (P = .001). This increase represents an additional 6.5 deceased donors and 8 transplants from deceased donors per state, per year. Policies allowing for leaves of absence from public sector employment were associated with a 9.8% increase in transplants from living donors (P = .01), but there was no significant effect among deceased donors. This policy also exhibited a trend toward increasing the number of living donors, although it did not reach statistical significance (7.7% increase; P = .05). Using a Bonferroni correction for multiple comparisons and applying a significance threshold of P < .007, the only statistically significant effect remains the association between the revenue policy and transplants from deceased donors (Table 3).
Regressions using a narrow time frame showed similar effects of the revenue policies on overall outcomes (7.7% increase in number of transplantations: P = .05 [eTable 3 in the Supplement]; 9.1% increase in number of deceased donors: P = .01; and 8.8% increase in the number of organ transplants from deceased donors: P = .01 [eTable 4 in the Supplement]). The associations between public leave policies and living donations and transplants were no longer significant in these specifications. This pattern of results from the first sensitivity analysis was consistent in models using flexible state-specific time trends (eTables 5 and 6 in the Supplement). Regression models inclusive of up to 5-year time lags showed no evidence for lagged policy effects (eTable 7 in the Supplement). Results were also consistent in models that used a negative binomial specification (eTable 8 in the Supplement), with the main variable as a count of the total policies passed (eTable 9 in the Supplement), stratified by organ (eTable 10 in the Supplement), and with a limited set of policy variables that was most closely related to each outcome (eTable 11 in the Supplement).
In this national study of organ donation and transplantation, we found that state policies have had a modest effect during the past 2 decades. First-person consent laws, donor registries, public education programs, paid leave, and tax incentives had no robust, significant association with either donation rates or number of transplants, even after allowing for prolonged delays for policies to take effect. Policies that establish state-based revenue pools were associated with increases in the absolute number of transplants, specifically among deceased donors. However, the magnitude of the effect was small, with revenue policies contributing to an additional 6.5 deceased donors and 8 transplants from deceased donors per year for an average state.
This largely negative effect may be explained by multiple factors. First, state-by-state differences in policy implementation may have attenuated overall effects such that existing positive effects in certain states may have been of no effect by less successful efforts in others. For example, there is anecdotal evidence that the implementation of first-person consent laws involves a fundamentally new approach to procuring consent and required retraining staff, addressing families whose wishes conflict with those of the potential donor, and decisions on how to create the consent process.24,25
Second, existing incentives may not be sufficient to motivate the behavioral change required to become an organ donor. For instance, the maximum cash value of tax deduction policies under existing policies is approximately $600,14 which is markedly less than the suggested $10 000 threshold to motivate donation of a solid organ, as described in prior studies.26 Similarly, the loss of wages associated with leave-of-absence policies may be too great to forgo for potential donors: living kidney donors assume an average cost of $3650 for loss of income and other costs associated with donation,27 with 15% of donors incurring costs of greater than $7500. Future policies may consider incentives that offer sufficient benefit to outweigh the various costs associated with any act of donation. Finally, the act of donation may be most strongly motivated by individual altruism, and existing policies may not be sufficiently appealing to this aspect of donor characteristics.28,29
The small observed benefits from the institution of revenue pools—the only policy that had a statistically significant effect—may be owing to several possible mechanisms. Funding from revenue pools may be used for recruitment activities not covered under public education, such as community outreach campaigns and worksite campaigns. Funds may also be used to educate physicians, lawyers, funeral directors, and other professionals who may affect the public’s willingness to donate. Other states may choose to use revenue funds to strengthen other policies, such as supplementing the costs of running registries. These policies were particularly associated with increased rates of deceased donation, which may indicate that such funds are being targeted toward the improvement of clinical practices to optimize the organ procurement process at the time of deceased donation. Existing revenue policies rarely stipulate exactly how funds will be used in a state, allowing states to fund activities that may be particularly meaningful to the local context. We further explored these ideas by examining whether the effects for the other state policies were magnified in states that had passed revenue policies. However, these analyses, which involved interaction terms between the policies, were underpowered to draw meaningful policy conclusions.
These results offer the first examination of the national effect of a variety of policies on organ donation. Prior studies have been limited to fewer policies30 or single donor types.3 Previous work has examined policy promotion in isolated contexts but did not assess the effect on eventual downstream outcomes of donation or transplantation.31 Our findings are consistent with prior studies showing a minimal effect of tax benefits and paid leave policies on solid-organ donation.13- 15 Our small effect sizes are also consistent with research on repealed motorcycle helmet laws, which led to a 0.7% increase in the number of total donors in these states, representing an additional 0.5 donors per year.32 Furthermore, this study clarifies an important differential effect between donation and effective transplantation and offers a possible mechanism of action of policies leading to changes specific among deceased donors and transplants.
Our study has limitations. First, our use of state-level data may have limited the evaluation of policy efficacy. For instance, our ability to assess the effectiveness of donor registries may have been restricted owing to the lack of data on the number of registered donors, which could be more sensitive to these policies. Second, although we performed sensitivity analyses allowing policies up to 5 years to take effect, policy ramifications may simply take longer. Educational programs or registry policies may have led to increases in the donor pool, but these effects will not be identifiable until the eventual date of organ procurement (most likely many years after the date of policy passage).
Third, we used dates of policy passage to determine changes in outcomes before and after passage; however, passage of legislation does not always correlate with immediate policy implementation. We explored details on the timing of policy enactment and implementation using additional legal and policy briefs, but they were incompletely obtained for all states. Analyses were repeated using the subset of states for which dates of policy enactment were available, and results were consistent. Fourth, our study is limited to policies that directly seek to increase organ donation rates. Other policies aimed at motorcycle helmet use, health system transformation, public health spending, smoking rates, and diabetes mellitus could conceivably effect donations by affecting the health of the donor pool. Finally, our study was observational in that state policies were not randomly assigned. Thus, even with a careful difference-in-differences research strategy, our estimates remain open to bias in which states that opted to implement policies may have different characteristics than those that did not. Causal claims, therefore, should be made with caution.
We found that state policies designed to increase organ donation and transplantation during the past 2 decades had little to no effect. The only policy with a statistically significant effect was the establishment of state-based revenue policies. This policy was associated with national increases in the number of transplants, specifically among deceased donors, during the past 2 decades. However, state-by-state variation in how these funds are used is large, limiting the generalizability of this finding. These findings suggest that new policy designs may be needed to increase donation rates. Future information on the successful use of revenue pools on a state-by-state basis will be beneficial to policymakers in providing crucial insights into how state and national policies can best use the current policies in place and help resolve critical organ shortages. More generally, using county-level data on donations and transplantations, as well as more granular information on specific policies passed in each state, can greatly advance the research in this area and provide deeper insights into best practices in organ donation policy.
Accepted for Publication: December 18, 2014.
Corresponding Author: Erika G. Martin, PhD, MPH, Department of Public Administration and Policy, Rockefeller College of Public Affairs and Policy, University at Albany–State University of New York, 1400 Washington Ave, Milne 300E, Albany, NY 12222 (email@example.com).
Published Online: June 1, 2015. doi:10.1001/jamainternmed.2015.2194.
Author Contributions: Drs Chatterjee and Martin had full access to all the data in the study and take responsibility for the integrity and the accuracy of the data analysis.
Study concept and design: All authors.
Acquisition, analysis, or interpretation of data: All authors.
Drafting of the manuscript: Chatterjee.
Critical revision of the manuscript for important intellectual content: All authors.
Statistical analysis: Chatterjee, Venkataramani.
Administrative, technical, or material support: Vijayan, Wellen, Martin.
Study supervision: Venkataramani, Martin.
Conflict of Interest Disclosures: None reported.
Previous Presentations: This work was presented at the Keeneland Conference for Public Health Service and Systems Research; April 7-10, 2014; Lexington, Kentucky; and at the Annual Research Meeting for Academy Health; June 8-10, 2014; San Diego, California.