CHAPTER II
Transplant Data: Sources, Collection, and Research Considerations, 2004

Introduction

In chapters corresponding with this one in the OPTN/SRTR Annual Report of the two previous years, we have discussed a range of topics: the scope of transplant data available and the evolution of data collection mechanisms; how that data collection system is improving the quality of these data and reducing data collection burden; how additional ascertainment of outcomes both completes and validates existing data; and caveats that remain for researchers (1,2). This year, we build upon that foundation and focus on three areas: 1) a brief summary of the scope of data available; 2) an expanded discussion of data submission patterns both on the waiting list and after transplant, highlighting data collection improvements and feedback, and their implications for analysis; and 3) further examination of the impacts of extra ascertainment of outcomes, particularly for posttransplant kidney graft failure from Medicare data.

It is important that researchers using transplant data have an understanding of the scope and structure of available data, and are familiar with how these data are collected. Readers seeking more detailed background about the structure and source of available data should refer to (1), which also includes a more detailed discussion of initial multiple-source validation of mortality data. Readers seeking a more comprehensive description of the UNet^sm data collection system and recent improvements should see (2).

The description of data quality and timeliness presented here are in constant flux. Organ procurement organizations and transplant centers are increasingly familiar with new, more efficient data collection tools implemented by the OPTN, and data quality and timeliness have improved as a result. New procedures implemented by the SRTR for including additional ascertainment of outcomes — particularly mortality — may also have implications for transplant centers' ability and motivation to report these statistics. Feedback to the centers, initiated by both the SRTR and OPTN, include extra ascertainment of mortality and feedback about outcomes, and further encourage change in data submission patterns. These factors make it important for researchers to keep a keen eye on the most current measures of data timeliness in choosing cohorts, deciding on methods, and watching for potential biases in their analyses.

Overview of Data Scope

Figure II-1 shows a useful method of organizing transplant data into "units of analysis." These units of analysis are designed to be of most use to researchers asking questions about the events or outcomes that may follow a placement of a candidate on the waiting list, organ donation, or a transplant itself. The table entities in Figure II-1 relate to specific subjects of interest for research: candidacies, donors, transplants, and the components thereof. Also shown are some of the more specialized tables, ones from which researchers might analyze organ turndowns, immunosuppression medications used, or changes in status history.

Three tables in Figure II-1 are the entry points for individual persons into the transplant process: the candidate registration table (which includes registrants who become transplant recipients), and the living and deceased donor tables. Underlying these three individual level tables (and not shown in the figure) is a "Person Linking Table" (PLT) that is vital to the integration of multiple data sources discussed later. The PLT holds one record per person, establishes links on the basis of similarities in Social Security Numbers (SSNs), names, dates of birth, and other person-level information, while accounting for many of the mistakes in entering these fields. The maintenance of this identification roster, with aggregated identification information compiled from all data sources, facilitates a system of matching to both external data sources and other records within OPTN data, such as for persons who receive multiple transplants or even for donors who later become recipients.

In addition, this figure documents some of the primary and secondary data sources that may contribute to each table. Further detail regarding the specific data collection instruments, before the information is aggregated to records of interest, is shown in Figure II-2.

Waiting List Data

In Figure II-1, the "candidate registration" table holds records for potential transplant recipients: patients who are placed on the waiting list as well as patients who receive living donor transplants without having been wait-listed. Analytically, this table helps researchers describe the "demand" side of the transplant process, comparing characteristics of successful and unsuccessful transplant candidates and describing disease progression among prospective recipients while they are not transplanted. These candidates act as a useful comparison to those who do receive transplants; considering the consequences of not being transplanted can be helpful in evaluating the benefit of transplanting each type of patient. Because mortality plays such an important role in evaluating transplant benefit, the examination of the timeliness and accuracy of candidate data sources presented in this section focuses in particular on the reliability of mortality information.

Primary Sources

The primary source of information about candidates for transplantation is the OPTN database, which stores information about all persons on the national transplant waiting lists. Information in this table is taken from the waiting list maintenance and the Transplant Candidate Registration (TCR) record completed soon after registration. Transplant centers must continuously maintain their waiting lists by reporting on changes in severity of illness (for some organs) and other outcomes such as transplant or death.

Because the maintenance of the waiting list is continuous, researchers should be able to report upon waiting list outcomes soon after they happen. In actuality, this depends on the outcome. Removal from the waiting list for transplant is linked to generation of a transplant record, so reporting is nearly immediate. Reporting of death on the waiting list may display more lag in reporting, particularly among patients who are offered organs less frequently because of low severity of illness or accumulated waiting time, since turndown of offers often spurs waiting list maintenance.

Timing of Waiting List Maintenance

Table II-1 helps an analyst assess the currency of waiting list data for mortality analyses by showing the time between death and removal from the waiting list for death. The first three columns show evidence of improved timeliness of waiting list removal for death, though the figures for 2003 may overstate completeness at any point in time because not all deaths during 2003 have been reported yet. About three-quarters of the deaths that are reported by the centers are reported within two months of their occurrence. This profile of lag time in reporting guides the researcher in choosing appropriate cohorts for analyses of waiting list outcomes that include mortality, based on primary data sources.

Table II-1. Lag Time to Report of Death On Waiting List; All Deaths to Waiting List Registrants Reported By Center (Cumulative Percent Reported)

Reporting is less prompt among candidates for kidney transplant than for other organs: 65.9% versus 82.4% for livers and 90.2% for hearts (Table II-1). This difference is expected because of the longer waiting times and available alternative therapies that may make the contact between patient and transplant center less frequent. In 2002, nearly 30% of deaths among heart registrants were reported on the day of death, compared with less than 5% of kidney registrant deaths.

Extra Ascertainment Sources

A transplant center's reporting duties end upon each candidate's removal from the waiting list. However, events occurring in the months following removal — such as death or transplant at another center — are frequently interesting analytical endpoints to the researcher. Therefore, a candidate file may incorporate additional mortality sources; or waiting list, transplant, and follow-up information reported by other centers for the same person.

Many of the same additional sources of outcome ascertainment are used for both waiting list analyses and posttransplant analyses, particularly for mortality. Using the patient-linking table (PLT, described above) to match patients, results may be incorporated from three other "secondary" sources:

• Patient linking between OPTN records allows a researcher to tell that a transplant candidate at one center has had a death or transplant reported by a different center; or that a graft has failed on the basis of a retransplant at another center.

• The Social Security Death Master File (SSDMF), publicly available from the Social Security Administration (SSA), contains over 70 million records created from reports of death to the SSA, for beneficiaries and non-beneficiaries alike.

• The CMS-ESRD Database provides data primarily from Medicare records for ESRD patients, and helps provide evidence of start of dialysis therapy, posttransplant resumption of maintenance dialysis indicating graft failure, or death.

In addition, the National Death Index (NDI) is available for validation of the completeness of these sources, though its use is not permitted for most analyses. The NDI, based on death certificate information submitted by state vital statistics agencies, misses only about 5% of all deaths in the US.

In 2002, the SRTR and OPTN jointly obtained data from the NDI for a sample of transplant candidates and patients to evaluate the completeness of mortality reporting in the other existing data sources. Analyses presented in this forum in 2002 showed the importance of using extra sources of mortality (1). While the majority of deaths are reported by the main transplant center following the patient (68.9% for kidney and pancreas; 88.3% for other organs), significant fractions of all the deaths are reported by other available sources. Among kidney and pancreas recipients, for whom secondary sources were more important than for other organs, 6.0% of deaths were reported first by another transplant program, and an additional 22.9% were reported by the SSDMF. The reassuring news for researchers, though, is that after these three sources are considered, very few additional deaths are found in the NDI. The data sources available for analyses capture more than 99% of the deaths known to transplant recipients. It is also important to note that no one of these sources alone looks sufficient: the SSMDF captures the highest percentage of all the deaths, but at 82.2% of all the deaths a significant element of mortality would be missed if mortality analyses were performed on a time period when only the SSDMF were available and complete.

The contribution of secondary sources varies by age group. There is evidence that the SSDMF is less complete for pediatric patients, particularly recipients of non-renal organs (1, Table 6). However, for these younger patients it also appears that secondary OPTN sources of death may be more important, since younger patients may be better candidates for retransplant and are thus more likely to show up in further OPTN records. Even for younger patients, the portion of deaths identified exclusively in NDI remains less than 1%.

Measuring Completeness of Waiting List Mortality

Because the SSDMF is the most complete and timely, it is used as a benchmark in several comparisons to measure changes in reporting patterns over time in the OPTN data. The data reported to the Social Security Administration that make up this file are increasingly accurate and timely. Of deaths reported for 2003, 87% were available to the SRTR within one month, and 96% within two months, of the death itself. Further, in 1995, 2.3% of the deaths reported by the SSDMF had an unknown day (i.e., only month and year were reported); by 2003, that share dropped to 0.3%.

Measuring the completeness of waiting list mortality may be considered in two ways. First, to what extent are deaths that occur while the patient is on the waiting list actually underreported by the listing center? And second, what are the analytical implications of considering mortality events after the patient has been removed and the center is no longer required to report about the patient? Many analyses used in developing allocation rules are based on a comparison of outcomes with and without a transplant, for which the outcome of a patient who becomes too ill to be transplanted may be worthwhile to consider.

Using the three main secondary sources of mortality data outlined above, we examined patients who were recently removed (2001-2003) from the waiting list for reasons other than transplant or death. Table II-2 shows that a substantial fraction of these patients died either soon after, or even before, removal from the waiting list. Ten percent of these patients died either while still on the waiting list, or within one month after removal. While these cases are concentrated among candidates removed for "other" or "condition deteriorated" reasons, it is interesting that so many of these patients were not removed for reason of death, and particularly surprising that several were removed for "condition improved." Without extra ascertainment, many analyses might not account for the adverse results seen among these patients. The number of patients who die without the center's report — or perhaps even without the center's knowledge — represents an improvement over the 13% reported here last year and may be due to several recent developments. Implementation of the MELD/PELD system for liver allocation may have prompted significant clean-up of waiting lists by many liver programs, as new data were entered for to accommodate the new scoring system. In addition, in 2002 the OPTN and SRTR jointly implemented a system to help transplant centers maintain their waiting lists and posttransplant follow-up by informing them of patients who were found in the SSDMF.

Table II-2. Extra Ascertainment Outcomes for Waiting List Registrations Removed in 2001-2003

The additional deaths reported for wait-listed registrants described here also have implications for the measured "lag time" reported above in Table II-1. Table II-3 shows the time until any death reported in the SSDMF is also reported by the center at which a patient is wait-listed. Though the majority of deaths (60.3% in 2003) are reported by the center within one month, even after six months, reasonably complete reporting of waiting list mortality relies on the SSDMF for 15%-20% of candidates (84.3% center-reported, 2003, with the balance reported by the SSDMF). Expected differences are seen by organ, with kidney least complete.

Table II-3. Lag Time to Report of Death on Waiting List; All Deaths to Waiting List Registrants Found In SSDMF That Occur Before Removal (Cumulative Percent Reported)

Implications of Time Lag and Extra Ascertainment for Waiting List Analyses

In choosing cohorts for waiting list analyses, researchers need to be aware of the timeliness of reporting for each source of data as outlined here. To obtain unbiased results over time, cohorts should be chosen to minimize the effect of lag in reporting on important results.

For example, the impact of extra ascertainment can be plainly seen in annual mortality rates among waiting list patients presented in the Annual Report, tables X.3 in each organ-specific section. Table II-4 shows three methods of calculating deaths per 1,000 patient years on the waiting list; the middle one, "With Extra Ascertainment, Until Waiting List Removal" is presented in this year's report.

Table II-4. Change in Waiting List Death Rates, Extra Ascertainment and Post-Removal Ascertainment

The first set of death rates, calculated without the benefit of extra ascertainment, use only deaths reported by the transplant center maintaining a patient's waiting list. For each organ presented, a large drop in death rate is seen in 2003. This drop could reflect improved patient care or allocation system, but it is hard to gauge these effects because recent deaths have not yet been reported by the transplant center.

Once extra ascertainment is incorporated, as shown in the second set of rows, the drop in death rates in the last year is less drastic. These lower death rates in the most recent years also may be a more accurate reflection of actual changes in patient survival on the waiting list. These rates include deaths reported using extra ascertainment as long as they occur before the patient was removed from the waiting list. Here, an increase in death rate is seen even in the earliest year, 2001, reflecting different treatment not only of registrants who still have yet to be removed from the waiting list, but also of patients for whom external sources indicate death before removal for other reasons. The largest jump in deaths between the first two sets of rows is seen in 2003, when the effect of lag in reporting is most noticeable.

When outcomes (and time at risk) following removal are added, death rates increase for all time periods and organs, as presented in the third set of rows in Table II-4. This suggests that outcomes after removal may indeed be important for some analyses, as they are not "the same" as outcomes while on the waiting list (when censoring at removal might be appropriate). The choice of using ascertainment of events after removal depends upon the purpose of the analysis, and may also depend on each individual patient's reason for removal.

Transplant and Posttransplant Data

The transplants table shown in Figure II-1 provides a collected source of information about each transplant event, including information about the donor, recipient, operation, and follow-up information, summarized to facilitate easy analyses. This file is used by analysts to characterize trends in the characteristics transplant recipients, examine transplant outcomes, and provide an estimate of posttransplant survival for comparison to waiting list survival in allocation policy decisions.

Primary Sources

The data for the transplant table are primarily taken from the Transplant Recipient Registration (TRR) form collected by the OPTN. Additional characteristics are added from the donor and candidate files are added for ease of analysis, as are aspects of the interaction between donor and recipient characteristics (e.g., calculated HLA mismatch scores; ABO blood type compatibility; whether the organ was shared, based on the relationship between the OPO recovering the organ and the transplanting center).

The transplant follow-up data, collected primarily from the Transplant Recipient Follow-Up (TRF) record, may be summarized to the transplant level, creating indicators of death, graft failure, and time of follow-up. The expected — and actual — timing of the follow-up forms are very important to cohort choice in analyses. After each transplant, follow-up forms are collected at the six month (for non-thoracic organs) and yearly anniversaries of the transplant; these forms may also be submitted off-schedule to report such adverse events as graft failure or death. While transplant follow-ups may also be useful on their own — or in conjunction with their own sub-tables for immunosuppression or malignancies — for analysis of specific events that occur during follow-up, they are most widely used in the summarized form for death and graft failure analyses discussed here.

Timeliness of Follow-Up Forms

Just as with events on the waiting list, it is important to consider the time lag until follow-up forms are filed when determining cohorts for analysis of posttransplant events. What is a sufficient time lag to expect the center to have submitted the forms after they are requested on the transplant anniversary? Implementation of new data collection mechanisms and stricter rules have shortened the time until validation. Table II-5 shows that the time from the date of record generation until validation (when the form has been submitted and verified by the center) has grown shorter, but it is still until about four months after each anniversary until nearly four of five forms are submitted, and six months before nine of ten are completed. A balance must be struck between the need for recent data and the need for complete data. The SRTR typically allows for between three and seven months of lag time, depending on the need for analyzing data from the most recent cohort available.

Table II-5. Timing for Validation* of Follow-Up Forms

Timing of Follow-Up Forms

In addition to the lag time until validation of follow-up forms after transplant, the pattern of form submission — often clustered soon after transplant anniversaries — has important implications for avoiding biases when analyzing recent data.

"Routine" follow-up forms are generated at each transplant anniversary, yet deaths occur on a continuous basis throughout the posttransplant period. When a patient dies during follow-up, the transplant center may file an "interim" follow-up form off the regular reporting schedule for that patient. This allows for the possibility that centers might report mortality more quickly and continuously than they report on surviving patients, for whom they must wait until the transplant anniversary.

For example, consider a survival analysis performed on a cohort of patients transplanted 18 months ago. Patients who are currently alive will have a one-year follow-up form indicating their survival until the one-year point, with no information beyond that. Patients who have died, on the other hand, might have follow-up forms indicating death both during the first year and any interim follow-up forms filed between months 12-18. Therefore, all of the data reported during months 12-18 would be about patients who had died. If a researcher used the Kaplan-Meier method to take advantage of the most recent data available, and censored at last follow-up, the portion of the survival curve calculated after the first year would be based inappropriately on over-reporting about patients who had died. To obtain an unbiased sample of outcomes after one year, the researcher must wait until the living patients are reported on at the two-year anniversary. Similarly, one-month survival rates cannot be reliably calculated until at least six months after transplant (one year for thoracic organs), after the anniversaries have prompted reporting on all patients.

The examples given above are stark; it would be rare to choose a sample of patients who were all transplanted 18 months ago. However, by including these patients in a sample used for survival calculations, without appropriate censoring at transplant anniversaries, introduces the same bias into the average results. Further, these caveats are not limited to survival analyses: other analyses might overrepresent outcomes associated with death in the final six-month period.

The above example describes the case when transplant centers may report deaths as they occur. If this were a reliable pattern — reporting all deaths immediately rather than waiting for the next reporting anniversary — one analytical solution might be to assume that the patient is alive unless we know otherwise. This would be a particularly attractive approach given the reliability of multiple sources of mortality in capturing all deaths. Recall, however, that the completeness of multiple sources is reliable only during periods for which we reasonably expect all sources to be mostly complete. In fact, actual death reporting patterns by transplant centers fall somewhere in between: Many deaths are reported as they occur, many more are reported at the reporting anniversary.

Figure II-3 depicts when transplant follow-up forms are filed, comparing those filed for patients who have died to those for patients who have not. The actual time of the follow-up event (death in the top panel or reported as alive in the lower panel) is shown on the y-axis, and the time that the follow-up form was validated by the center is shown on the x-axis. If all events were reported as they occurred, points would fall only along the 45-degree diagonal dashed line. The horizontal distance, left to right, between this diagonal and each point represents the time lag between the event and notification to the OPTN.

The top panel shows this relationship for follow-up forms reporting deaths, and the clustering of reporting along the diagonal shows deaths that were reported near the time of death itself. (In the earlier example of using a cohort of transplants from 18 months ago to calculate a survival curve, it is this clustering continuously along the diagonal for dead patients that introduces a possible bias beyond the 12 month follow-up time.) Another clustering, following to the right of each vertical line at 6, 12, and 24 months after transplant, shows deaths that are reported with the timing of routine follow-up forms. The actual death dates are distributed vertically along the line, emphasizing the extent to which many centers wait until prompted by the reporting cycle to report mortality, no matter when the death actually occurred.

The lower panel of the figure shows a similar clustering after each reporting anniversary, but the vertical height of these clusters, close to the diagonal itself, indicates that the events being reported on — that the patient is alive — occurred more recently compared to the reporting date. This difference is also borne out in the median lag reporting times, shown by arrows of different sizes in the two panels, at 139 days for deceased patients and only 31 days for alive patients.

Figure II-3 is instructive for choosing a cohort for posttransplant survival analysis. It is important to choose a combination of survival endpoint (horizontal line) and lag time (vertical line) that allows for a reasonable capture of both deaths and survivors. The survival endpoint (12 months) and additional lag time (+4 months) used for the SRTR Center-Specific Reports (CSRs) of one-year posttransplant survival are shown on the graph. Events in the shaded box are captured from center reporting, and would also be available in non-mortality analyses such as graft survival. Some events to the right of the shaded box will be reported by the center, if the transplant occurred early enough in the cohort to afford more than four months of lag time; others will rely on extra ascertainment, since center-reporting occurs after this lag time allows.

Extra Ascertainment Sources

The importance of extra ascertainment of mortality is as great for posttransplant analyses as it is for the waiting list analyses described earlier. Patients are more prone to becoming lost to follow up (LTFU) after receiving transplant than they are while still on the waiting list. About 10% of recipients transplanted with kidneys, livers, hearts, or lungs were LTFU by the end of the third year after transplant; about two-thirds of these had been coded as LTFU by the transplant center, and the other third had no records completed for at least the last 1.5 years before the three-year anniversary. Above, we outlined arguments suggesting that even with LTFU, extra ascertainment of mortality makes it plausible to assume that all sources taken together provide reasonably complete ascertainment of death, such that less than 1% of deaths are missed.

Multiple Sources of Data and Lag Time

For patient survival analyses, the SRTR often adopts a technique of assuming a patient is alive unless known otherwise, allowing us to follow patients after they become LTFU. It is important to continue to choose cohorts carefully, because the assumption of "alive unless we know otherwise" holds most true during periods when we expect all sources to be complete and unbiased. Because the lag time for extra sources such as the SSDMF is similar, if not a little shorter, than that for transplant follow-ups, the SRTR uses cohorts based on the timing of OPTN follow-up forms even when additional ascertainment is included. Researchers must also bear in mind other reporting patterns within the additional sources, such as the possible loss of Medicare eligibility three years after transplant for CMS data.

Implications for Overall and CSR Death Analyses; Possibility of Relying On SSDMF

The SRTR Center-Specific Reports (CSRs) present an excellent example of the impact of extra ascertainment on survival. The assumption of "alive unless we know otherwise" allows us to add to survival calculations not only deaths, but also additional time at risk. Reported survival for many cohorts changes very little and may actually go up by using extra ascertainment. For smaller cohorts, such as at the center-specific level, differences introduced with extra ascertainment are more evident, though they may still go in either direction. Prior to the implementation of extra ascertainment in the CSRs, concern was expressed that low follow-up completion for some centers may bias results (3,4). The use of extra ascertainment helps ensure more complete and reliable information for patients, families, and administrators using the survival statistics in these reports, even if the overall average does not change.

Table II-6 shows the one-year survival presented in the CSRs released at www.ustransplant.org in July 2004, and compares results to what would have been calculated without the use of extra ascertainment. While the total number of observed deaths, of course, increased with extra ascertainment, the mean center-specific survival rate for these organs remained virtually unchanged. Many individual programs, and in fact the majority of heart programs, experience no change in their center-specific one-year survival, and many others had increased survival as a result of extra ascertainment. On the other hand, the majority of kidney programs (131) did experience a reduction in reported survival rate with extra ascertainment.

Table II-6. CSR One-Year Survival Difference Due to Extra Ascertainment, July 2004

Some of the center-specific changes in survival were sizeable. For heart programs, decreases in survival were as large as nearly 7 percentage points, with increases more than 9 percentage points; the ranges were smaller but still sizeable for the other organ programs, which likely have more stable center-specific outcomes due to larger sample sizes. These large changes at a center level point to the importance of extra ascertainment for these types of analysis. As we would expect, the largest changes occurred in centers that had a low percentage of their follow-up forms completed (not shown in this table).

Graft Failure Analyses and Extra Ascertainment

In contrast to the options available for examining patient survival data, there is no "completed" source of graft failure data against which to test the completeness of OPTN/SRTR data. For many organs, retransplant is the only alternative therapy, so examining the transplant data file for retransplants for the same patient is sufficient for assuming complete follow-up. However, for kidney recipients the alternative of dialysis increases the possibility that graft failure may occur without the knowledge of the original transplanting center or any new (retransplanting) center. Some additional failure data may be available using CMS-ESRD data.

In 2003, the SRTR began using extra ascertainment for kidney graft failure for many types of analyses. Initially, a study was conducted to explore the possibility of supplementing existing SRTR data with CMS graft failure data for kidney recipients followed by the OPTN. The CMS data may provide additional information on recipients that are lost to follow-up, because CMS can be notified about a graft failure event through several possible mechanisms in addition to the OPTN, such as medical claims indicating return to (or initiation of) chronic dialysis treatment, the CMS 2728 medical evidence form, and Standardized Information Management System (SIMS) ESRD Network reporting. One noteworthy limitation is that CMS data typically capture graft failure events only for those patients covered by Medicare, which accounts for about 65%-70% of all kidney transplant recipients. Because insurance status is not well documented for kidney recipients after transplant, determining which patients may have Medicare coverage during follow-up is difficult.

Graft Failure Reporting Agreement Between Data Sources

For deceased donor kidney transplants that occurred between 1998 and 2001, graft failure events reported through October 31, 2003 were compared between OPTN and CMS. CMS failures reported after the end of the prescribed OPTN follow-up were not included in the calculations. Graft failures reported by the OPTN that resulted in death or retransplant within 30 days were also excluded from the analysis, because they might not be eligible for detection by CMS via return-to-dialysis claims or the agency's other means of identifying graft failure. Furthermore, prior SRTR analyses of extra mortality ascertainment suggest that the OPTN data provide near-perfect ascertainment for graft failures that result in immediate retransplantation. The overall agreement between the two sources with respect to reporting of graft failure is shown in Table II-7. This analysis assumed, for CMS follow-up, that if there was no evidence of a graft failure, then the graft was still functioning. Based on these results, CMS was found to have missed 22% of OPTN-reported failures, and 20% of total failures reported by either source. The inclusion of CMS failure dates would increase the fraction found to have failed, compared to that based on OPTN reporting alone, from 12.9% to 14.5%.

Table II-7. Graft Failure Status Agreement for OPTN vs. CMS Data, All Patients (N=31,265)

Predictors of Better or Worse CMS Reporting of Graft Failure Dates

When considering use of these additional outcomes, it is important to ask: For which patients would we expect to find graft failure dates in the CMS data, if they existed? These are the patients whom we can likely assume are failure-free if no report of CMS (or OPTN) graft failure exists. Answering these questions is a crucial first step in attempting to incorporate additional graft failure/survival information from CMS into the SRTR follow-up database.

To identify specific factors associated with follow-up in the CMS data, we tested the reliability of the graft failures CMS reported by comparing them to known graft failures reported by OPTN. Graft failure dates from the two sources were considered to "matchb" if they fell within 90 days of one another. Logistic regression modeling was then used to identify what factors, if any, were associated with better or worse agreement between the two data sources.

Results shown in Table II-8 strongly suggest that certain subgroups of patients are more likely to have follow-up reporting in the CMS data than other subgroups, dominated by factors indicating prior Medicare coverage which carries over to the posttransplant period (for up to three years, based on Medicare coverage rules). Kidney recipients with Medicare coverage at time of transplant were nearly three times more likely to have a matching graft failure date than those without Medicare. Patients receiving preemptive transplants were less than one-third as likely to have a graft failure date reported in the CMS data where an OPTN graft failure date existed. A significant relationship between the time since transplant and the failure date matching was also found, but the effect was very small (approximately one percent increase in agreement rate per month).

Table II-8. Predictors of Graft Failure Agreement Between CMS and OPTN

Table II-9 confirms these results by showing the graft failure status agreement for the subpopulation of non-preemptive, kidney transplants originally paid for by Medicare. With these restrictions, the number of OPTN-reported failures not detected by CMS (n=410, 2.0% of total cases) was equivalent to the number of CMS-reported failures not detected by the OPTN (n=398, 1.9% of total cases); each represented about 12 percent of the total failures.

Table II-9. Graft Failure Status Agreement for OPTN vs. CMS Data, Non-Preemptive Transplants Covered By Medicare (N=20,929)

Conclusion

In the two previous editions of the Annual Report, this chapter focused on data collection and organization schemes for transplant data, and offered beginning insights into implications of their timing and completeness. This year's chapter focuses further on caveats related to cohort choice, timing of data submission, and potential biases in follow-up data, as well as documents the continuing improvements in data collection timeliness and scope. Taken together, these three chapters provide the researcher — either analyst or reader of existing studies — with an important understanding of the limits and possibilities of these data. This background helps to document and explain factors influencing choices of cohorts and data sources, and also helps new researchers make these choices.

References

1. Dickinson DM; Bryant PC, Williams MC, et al. SRTR Report on the State of Transplantation: Transplant Data — Sources, Collection, and Caveats. Am J Transplant 4 (Suppl 9):13-26, 2004

2. Dickinson DM, Ellison MD, Webb RL. SRTR Report on the State of Transplantation: Data Sources and Structure. American Journal of Transplantation 3 (Suppl. 4):13-28, 2003.

3. Marchione M. Transplant rate reports don't tell whole story. Milwaukee J Sentinel 2001; July 27:G1.

4. Cooper L. Survival Data: Do the numbers really mean anything? Transplant News and Issues 2001. 2(2): s9-s11, s13.

CONTRIBUTORS

The following individuals prepared this chapter: David M. Dickinson, MA^a; Dawn M. Dykstra^a; Gregory N. Levine^b; Shiqian Li^a; James C. Welch^a; Randall L. Webb^a. ^aScientific Registry of Transplant Recipients / University Renal Research and Education Association; ^bScientific Registry of Transplant Recipients / University of Michigan.