My Jun 30 blog post discussed the rapid rise in the number of transplants involving unrelated live kidney donors. Today’s post will explore this trend in finer detail by appending two additional pieces of data. First, it looks at the distribution of transplants by transplant center. Then it will look at the distribution by source of donor. That is, did the donor come directly to the hospital (which I call Internal) or did the transplant involve an organ procurement organization (OPO) or a kidney exchange.

About the data

The counts of transplants by donor relation, transplant year (1988-2010), and transplant center name is available from the UNOS website. For each transplant center I appended the city, state, OPO region (11 total), and OPO name (48 total) using data from the Association of OPOs. I appended a hospital type (for profit, non-profit, teaching, children’s, or VA/military) based on hospital name. I appended the name(s) of kidney exchange(s) it participates in using data from the four kidney exchanges that I am aware of (National Kidney Registry, Alliance for Paired Donation, Paired Donation Network, and New England Program for Kidney Exchange). I also assigned all transplant centers in the Washington DC area to a presumed kidney exchange run by the Washington Regional Transplant Community OPO. I did this based on recent newspaper articles describing multi-hospital kidney trades occurring there.

Then for each record, I assigned a donor source using the following criteria. If the donor was deceased, I set the source to be the OPO. If the donor was live and known to the recipient or nondirected, I assigned it to Internal. (This introduces a small error since a few nondirected donors, like me, enter into exchanges.) If the transplant involved a paired exchange in 2008 or later, I assigned the donor to a kidney exchange if there was one. (This introduces an error if the hospital did not join the exchange until a later year or if the trade was conducted outside the exchange.) If the center is not a participant in an exchange, I assigned it to Internal. If the transplant involved a paired trade prior to 2008, I assigned the donor to Internal. (I did this because most hospitals have only joined exchanges recently. However, this introduces a small error since a few kidney exchange mediated transplants did occur prior to 2008.)

I deleted all records with a count of 0, cleaned the transplant center names and appended the new data using the vlookup function in Excel. The final Excel xlsx file contains a total of 27,960 records and is posted on SkyDrive if you want to download it to create your own pivot tables.

The counts in the table are annual averages based on the three most recent full-year data available, 2007 to 2009. The counts are rounded to the nearest whole number, but the rankings are based on the raw fractional data (so two centers with the same count may have different ranks).

I want to thank the UNOS, the Association of OPOs, and the kidney exchanges for the use of their data. Any errors in the data append or in the analysis are mine alone. Please contact me if you find any errors.

Total transplants

From 2007 to 2009, there were a total of 246 transplant centers in the U.S. that performed at least one transplant (the lowest count is three), for a total of 49,984 transplants. The ten highest volume kidney transplant centers are shown in the table below.

Rank Transplant center Avg. annual transplants
  1 UCSF Medical Center 339
  2 UCLA Medical Center 298
  3 Univ Wisconsin Hospital and Clinics 277
  4 Univ Alabama Hospital 273
  5 New York-Presbyterian/Columbia Univ 257
  6 Northwestern Memorial Hospital 247
  7 Univ Maryland Medical System 240
  8 Jackson Memorial Medical Center 235
  9 Univ Michigan Medical Center 233
10 Clarian Health-Methodist/Indiana Uni./Riley 217
  – Total for 246 centers 16,661

Data from UNOS for 2007-2009

Live donor transplants

About 37% of all kidney transplants involve a live donor. All transplant centers performed at least one living donor transplant. However, there is a surprisingly large variation in live donor rates between individual centers. Six out of the ten hospitals with the highest live donor rates are children’s hospitals. (Some of the lowest rates are also at children’s hospitals.) Among the centers that performed at least 60 transplants in the past three years, Rochester Methodist Hospital (Mayo Clinic) had the highest proportion at 80% while the Univ Mississippi Medical Center had the lowest at 3%. The table below highlights some of this data. Given the superior outcomes for live donor transplants, this difference can have a significant impact on the outcome by hospital. Plotting this data geographically may reveal if the differences have a regional or urban/rural bias. That might be the subject of a future blog post.

Rank Transplant center Avg. annual live donor transplants Live donors as % of all transplants
    4 Rochester Methodist Hospital (Mayo Clinic) 127 80%
    8 Children’s Hospital of Pittsburgh of UPMC 5 68%
    9 Univ Cincinnati Medical Center 38 67%
  16 Northwestern Memorial Hospital 152 62%
  66 UCLA Medical Center 132 44%
   Average for 246 centers 6,126 37%
112 Univ Wisconsin Hospital and Clinics 100 36%
114 UCSF Medical Center 122 36%
242 Harper Univ Hospital Detroit Medical Center 10 9%
243 Hahnemann Univ Hospital 10 5%
245 Univ Mississippi Medical Center 3 3%

Data from UNOS for 2007-2009

Unrelated donors

A survey described in Amer. J. Transpl. Oct 2007 indicates that possibly all transplant centers permit unrelated persons to be a live donors. However, of the 246 centers that perform live donor transplants, there are 14 that did not perform any with unrelated donors in the past 3 years. All but two are children’s hospitals. Overall, children’s hospitals tend to have much lower rates of unrelated donors. Among transplant centers that performed at least 60 transplants in the past three years, Green Hospital of Scripps Clinic had the highest proportion of unrelated donors at 67%. Rochester Methodist Hospital (Mayo Clinic) performed the highest absolute number of unrelated donor transplants, at 68 annually. The table below highlights some of this data.

Rank Transplant center Avg. annual unrelated donor transplants Unrelated donors as % of live donor transplants
   4 Green Hospital of Scripps Clinic 18 67%
   5 Johnson City Medical Center Hospital 6 67%
   6 Virginia Mason Medical Center 17 65%
   7 Shands Jacksonville 3 63%
   8 Swedish Medical Center 18 61%
  15 Johns Hopkins Hospital 51 54%
  16 Rochester Methodist Hospital (Mayo Clinic) 68 53%
  23 Univ Michigan Medical Center 53 52%
  25 Methodist Specialty & Transplant Hospital 56 52%
  58 UCSF Medical Center 59 46%
  85 Univ Alabama Hospital 52 43%
Average for 232 centers 2,528 41%
103 UCLA Medical Center 54 41%
116 Northwestern Memorial Hospital 61 40%
232 Children’s Hospital of Pittsburgh of UPMC 3 7%
233 tie* Children’s Hospital Los Angeles 0 0%

*There are 14 hospitals that did not perform any unrelated donor transplants.
Data from UNOS for 2007-2009

Kidney trades

A kidney trade involves an incompatible donor-recipient pair that trades kidneys with another incompatible pair or for a deceased donor kidney. Both type of trades started in mid-1990s, though they are still rare events at most hospitals. Although 159 transplant centers (just under two-thirds) have performed at least one, only 59 have performed 5 or more in the past three years. (This implies that awareness of trades is high and that the practice meets the ethical standards of the hospitals. However most hospitals perform a few opportunistically and are not systematically pursuing them as a strategy. This is a great opportunity for the organized exchanges to offer their services.)

As the table below shows, 111 transplant centers have participated in at least one live pair trade in the past three years, resulting in 665 transplants. 94 transplant centers have participated in a living/deceased donor trades, resulting in 317 transplants (there is overlap as 46 centers perform both). Note that both transplants in a live pair trade is counted as a trade while only one is for a live/deceased trade. The recipient surgery in the live/deceased pair is counted as a deceased donor transplant.

The table below shows the growth in trades from 2007 to 2009 for hospitals that performed only live pair trades, only live/deceased trades, or both. The top two rows show the number of live pair trades and the number of transplant centers involved (in parentheses) while the bottom two rows show the number of live/deceased trades (and centers involved). It appears that both types of trades are becoming more popular and all three groups of hospitals are growing.

  Total*  2009 2008 2007 
Participate in live pair trades only 381 (65) 193 (42) 131 (44)   57 (23)
Participate in both (counts for live pair trades) 284 (46) 111 (30) 109 (29)   64 (21)
Participate in both (counts for live/deceased trades) 185 (46)   75 (30)   50 (21)   60 (22)
Participate in live/deceased trades only 132 (48)   52 (25)   49 (25)   31 (19)

*Number of transplants (number of centers)
Data from UNOS for 2007-2009

As I stated in my Jun 30 blog post, I dislike living/deceased donor trades because the recipients receive a poorer quality kidneys than if they were to enter into live pair trades. Both types of trades are growing, even at hospitals that perform both, despite the advantages of live kidney transplants. I think kidney exchanges should actively assist hospitals that practice living/deceased trades to rely more on live exchanges.

Among transplant centers that performed at least 60 transplants in the past three years, Methodist Univ Hospital in Memphis, Tennessee has the highest proportion of unrelated donor transplants that involve a trade at 79%. All of them were live pair exchanges that appear to have been arranged internally. Methodist Specialty and Transplant Hospital in San Antonio Texas performed the largest number of trades, at 18 per year.

Note that the most active transplant centers perform very few kidney trades. This includes UCSF and UCLA Medical Centers in California, the state with the longest waiting list for organ transplants. This is unfortunate since entering incompatible donor-recipient pairs into an exchange can dramatically shorten wait times. Also note that several hospitals performing the most trades are not members of any of the regional or national kidney exchanges. This indicates there is plenty of opportunities to improve matching. Finally, note that since participation in live pair trades is growing rapidly, the use of a 3-year averages and excluding 2010 data may obscure trends.

Rank Transplant center (Paired exchange) Avg. annual live pair + live/ deceased trades Trades as a % of unrelated donor transplants
  5 Methodist Univ Hospital (Internal) 7  +   0 79%
12 Tufts Medical Center (Internal and NEPKE) 7  +   0 57%
13 Froedtert Mem. Lutheran Hospital (APD and PDN) 2  +   7 51%
20 Medical College of Virginia Hospitals (Internal) 0  +   7 49%
21 Clarian Health-Methodist/Indiana Univ/Riley (APD and PDN) 16  +   0 45%
29 Emory Univ Hospital (Internal and NKR) 4  +   4 36%
41 California Pacific Medical Center (Internal and NKR) 7  +   4 33%
42 Methodist Specialty & Transplant Hospital (Internal) 18  +   0 31%
45 Johns Hopkins Hospital (Internal and NEPKE) 15  +   0 29%
52 Banner Good Samaritan Medical Center (Internal and APD) 8  +   0 27%
  Average for 159 centers 222 + 106 13%
101 UCLA Medical Center (Internal and NKR) 6  +   0 12%
148 Rochester Methodist Hospital (Mayo Clinic) (Internal) 2  +   0 3%
155 UCSF Medical Center (Internal and NKR) 1  +   0 2%
160 tie* Univ Wisconsin Hospital and Clinics 0  +   0 0%

*There are 87 hospitals that did not perform any transplants that involved kidney trades
Data from UNOS for 2007-2009 and from kidney exchanges

Since my counts of transplants mediated by each kidney exchange is not clean, this blog post will not report this data. If I can obtain this data, I will try to make it the subject of a future post.

Nondirected stranger donors

As mentioned in a Sep 2009 blog post, the consensus regarding the ethical acceptability of nondirected and directed stranger organ donations is still developing. The 2007 Amer. J. Transpl. survey shows about 60% of all transplant centers allow nondirected donations. The UNOS data shows 110 centers have completed at least one transplant with a nondirected donor in the past three years.

Only recently have hospitals even encouraged the general public to consider becoming a nondirected donor. For instance see a May 2010 blog post describing a new pay-it-forward program launched by Loyola Univ. Medical Center after 4 nondirected donors walked in with another 21 undergoing evaluation so far this year. Compare that with the 3 or 4 per year recorded by the top hospitals in the table below. As mentioned above, some nondirected stranger donor participate through exchanges.

Rank Transplant Center Avg. annual nondirected donor transplants Nondirected donors as % of unrelated donors
  1 New York-Presbyterian/Weill-Cornell/Rogosin 4 9%
  2 tie Pinnacle Health System at Harrisburg Hospital 4 23%
  2 tie Massachusetts General Hospital 4 17%
  2 tie Univ Minnesota Medical Center 4 8%
  2 tie New York-Presbyterian/Columbia Univ 4 8%
  6 Montefiore Medical Center 4 23%
  7 tie Univ Utah Medical Center 3 25%
  7 tie Banner Good Samaritan Medical Center 3 12%
  7 tie Rochester Methodist Hospital (Mayo Clinic) 3 5%
10 tie Virginia Mason Medical Center 3 17%
10 tie Univ Wisconsin Hospital and Clinics 3 7%
10 tie St. Barnabas Medical Center 3 7%
  Total for 110 transplant centers 115 5%

Data from UNOS for 2007-2009

Directed stranger donors

A directed stranger donation is one that the recipient has solicited though publicity generated by news, advertising, or the web.(See a Sep 2009 blog entry for information about a site called MatchingDonors.) The 2007 Amer. J. Transpl. survey shows that only 30% of all transplant centers allow directed stranger donations, though acceptance is growing.  The UNOS database doesn’t distinguish between unrelated known and stranger donations, so no counts are available. Again, if I am able to find this data, I will try to include it in a future blog post.