SEARCH

SEARCH BY CITATION

Keywords:

  • Kidney transplantation;
  • medicare;
  • pregnancy after transplantation;
  • pregnancy and fertility;
  • USRDS

Abstract

  1. Top of page
  2. Abstract
  3. Background
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References
  9. Appendix

Fertility is one of the potential benefits for women undergoing kidney transplantation; however, population-based information about the likelihood of pregnancy and successful fetal outcome is not available. In this observational study of 16 195 female kidney transplant recipients aged 15–45 years in the United States between 1990 and 2003, we determined the pregnancy rate and live birth rate using Medicare claims data from the first three posttransplant years. The pregnancy rate was 33 per thousand female transplant recipients between 1990 and 2003 and progressively declined from 59 in 1990 to 20 in 2000. The live birth rate between 1990 and 2003 was 19 per thousand female transplant recipients and declined in parallel with the pregnancy rate. Despite a decrease in therapeutic abortions over time, the proportion of pregnancies resulting in fetal loss (45.6%) remained constant during the study due to an increase in spontaneous abortions and other causes of fetal loss. The pregnancy rate in kidney transplant recipients was markedly lower and declined more rapidly than reported in the general American population during the same period. The live birth rate was substantially lower than reported in voluntary registries of transplant recipients, and the proportion of pregnancies resulting in unexpected fetal loss increased over time.


Background

  1. Top of page
  2. Abstract
  3. Background
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References
  9. Appendix

Kidney transplantation is the preferred treatment for end-stage renal disease (ESRD). Compared to patients treated with dialysis, kidney transplant recipients live longer, have improved quality of life and consume fewer health care resources (1–3). In addition, hypothalamic gonadal dysfunction in females with ESRD may be reversed within the first few months after kidney transplantation (4). Therefore, the ability to have children is a potential benefit of kidney transplantation.

Despite thousands of successful pregnancies in kidney transplant recipients, there is limited information about the likelihood of pregnancy and anticipated fetal outcomes in these patients (4). Current data about pregnancy after kidney transplantation are derived from voluntary registries (5–8), case reports (9,10) or single-center case series (11) rather than population-based studies, which may be less prone to selection bias (12,13).

In the United States, kidney transplant recipients remain eligible for ESRD Medicare insurance for a period of 3 years after transplantation. We utilized inpatient and outpatient Medicare claims data to determine the pregnancy rate and the live birth rate among Medicare-insured kidney transplant recipients between 1990 and 2003 in the United States. Our objective was to provide population-based estimates of the pregnancy rate and live birth rate after kidney transplantation.

Methods

  1. Top of page
  2. Abstract
  3. Background
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References
  9. Appendix

Study population and data sources

Data from the United States Renal Data System (USRDS) were used for this study, which was approved by our hospital ethics review board. Females aged 15–45 years who received a kidney transplant between January 1, 1990, and December 31, 2003, and were insured only by Medicare at the time of transplantation were studied. Repeat kidney transplant recipients, and recipients of multiorgan transplants were excluded.

Patient follow-up

Study patients were followed from the date of transplantation until death, transplant failure (defined by repeat transplantation or resumption of chronic dialysis treatment) or December 31, 2003, the last date for which pregnancy-related claims were available.

Identification of pregnancies

Pregnancies and pregnancy-related events were identified using inpatient and outpatient Medicare claims during the first three posttransplant years. Medicare claims were available in the USRDS until December 31, 2003. We used International Classification of Diseases, 9th Revision, Clinical Modification (ICD-9-CM) codes to identify pregnancy-related clinical encounters (see the Appendix). A single inpatient claim was considered sufficient evidence of pregnancy. We assumed that the birth of live or stillborn infants would normally require hospitalization. Therefore, outpatient claims specifying delivery of a live or stillborn infant that were not corroborated by inpatient claims (n = 56) were excluded. We accepted single outpatient claims (n = 50) specifying an abortion-related event (for example a therapeutic abortion), as these events may only require a single outpatient clinical encounter. In all other cases where only outpatient claims indicated a pregnancy or pregnancy-related event, we required at least two outpatient claims to define an event (n = 131).

Statistical analysis

We compared the characteristics of study patients and patients who otherwise met inclusion criteria but who were excluded because they had an insurer other than Medicare using the chi-square test for categorical variables and the t-test for continuous variables. The chi-square test or ANOVA were used to compare the characteristics of patients transplanted during three different time periods defined by the predominant maintenance immunosuppressant medications used at the time of transplantation: 1990–1994 (cyclosporine and/or azathioprine prescribed to 96% of recipients), 1995–1999 (a transition period when a calcineurin inhibitor (either cyclosporine or tacrolimus) was used in combination with either azathioprine or mycophenolate mofetil (MMF) in 84% of recipients) and 2000–2003 (a calcineurin inhibitor [either cyclosporine or tacrolimus] in combination with MMF prescribed to 72% of recipients).

Pregnancy rate:  Among patients transplanted before December 31, 2000, we determined the pregnancy rate (number of pregnancies per 1000 patients) by transplant year. Pregnancies were identified using Medicare claims from the day of transplantation until the date of transplant failure or 3 years after the date of transplantation whichever came first. Separate results were determined for patients in different age and race groups, as well as among patients with different types of kidney disease, and patients with deceased and living donors. Comparisons with pregnancy rates in the general population were made using published information provided by the United States Division of Vital Statistics (14). The pregnancy rate was also determined separately during the first, second and third posttransplant years. In addition, the pregnancy rate was determined in a variety of patient subgroups including patients with different median within-neighborhood household incomes (determined by linkage of patient zip codes to data from the 2002 U.S. Census), and by the extent to which the residence location of each patient was rural or urban using the rural–urban commuting area (RUCA) code (15). RUCA codes are assigned to each US zip code based on markers of population density, with values ranging from 1.0 (most urban) to 10.6 (most rural). Information on population density is supplemented by data on employment commuting to ensure that suburban areas with low population density in which many residents work in nearby large urban centers are classified as urban. As in previous work (15), we classified each patient in the current analysis as belonging to one of three mutually exclusive RUCA groups: metropolitan (RUCA 1.0–3.9; cities with population of >50 000 and their associated suburban areas); micropolitan (RUCA 4.0–6.0; towns or cities with population of 10 000–50 000); and rural (RUCA 7.0–10.6; towns with population of <10 000).

Timing of pregnancy after transplantation:  Because pregnancies were identified by a pregnancy-related diagnostic code, we estimated the date of conception using the following assumptions regarding gestational age: term pregnancy, 37 weeks; preterm delivery, 33 weeks; therapeutic abortion, spontaneous abortion, abortion not otherwise specified, ectopic pregnancy and abnormal products of conception, 12 weeks. Stillbirths were assumed to occur at 30 weeks.

Although spontaneous abortions, induced abortions and ectopic pregnancies often occur before 12 weeks, we estimated that all such events occurred at 12 weeks as this represents the end of the first trimester, beyond which time such events would be infrequent. Similarly, we estimated stillbirths to occur at 30 weeks as this is the midpoint between 20 weeks gestation (the earliest point at which a nonlive fetus is considered a stillbirth) and 40 weeks gestation (full-term pregnancy). The results of multivariate models were unchanged when these estimates were varied in sensitivity analyses.

Factors associated with pregnancy after transplantation:  Multivariate Cox regression analysis was used to determine factors associated with pregnancy. The following factors were included in the multivariate model: age at transplantation, selfreported race, cause of ESRD, transplant kidney function (glomerular filtration rate (GFR) estimated using the abbreviated equation derived from the Modification of Diet in Renal Disease Study (16) using serum creatinine measurements measured approximately 6 months after transplantation), donor type (living or deceased), duration of dialysis prior to transplantation, transplant era (1990–1994, 1995–1999, 2000–2003), the type of maintenance immunosuppressant medications prescribed at the time of transplantation, median household income and the RUCA group. The proportional hazards assumption was tested using log-negative-log plots of the within-group survivorship probabilities versus log-time.

Fetal outcomes:  Fetal outcomes were determined separately for pregnancies during the first, second and third posttransplant year, and during the three different time periods (1990–1994, 1995–1999, 2000–2003). Comparisons of the live birth rates in study patients with those in the general population were made using published information provided by the United States Division of Vital Statistics (17). A multivariate logistic regression model was used to determine associations with fetal loss. Fetal loss included spontaneous abortion, therapeutic abortion, abortion not otherwise specified, stillbirth and ectopic pregnancy. This analysis was repeated excluding therapeutic abortions as a cause of fetal loss. Pregnancies in which fetal outcomes were not reported (n = 77) were excluded from these analyses. Covariates included in the multivariate model are as described above with the addition of the year of pregnancy. Analyses were performed with SAS software, version 9.1 (SAS Institute, Cary, NC).

Results

  1. Top of page
  2. Abstract
  3. Background
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References
  9. Appendix

Patient characteristics

Among the 30 078 women aged 15–45 years at the time of first kidney-only transplantation between January 1, 1990, and December 31, 2003, in the USRDS data files, 16 195 had Medicare recorded as their only source of medical insurance and were included in the study, while 13 883 had an alternative insurer and were excluded. Included and excluded patients were of similar age (33.4 ± 7.6 vs. 33.4 ± 7.9 years, p = 0.14), but study patients had longer exposure to dialysis prior to transplantation (2.8 ± 2.8 vs. 1.2 ± 2.0 years, p ≤ 0.0001), were more likely to be Black (29.6 vs. 18.0%, p < 0.0001), and were more likely to have renal failure because of diabetes (23.2 vs. 21.8%, p < 0.0001) or glomerulonephritis (39.6 vs. 36.4%, p < 0.0001). Study patients were also more frequently recipients of deceased donor transplants (70.8% vs. 48.8%, p < 0.0001).

The majority of study patients were white, and glomerulonephritis was the most frequent cause of ESRD (Table 1). Over time, the proportion of Black and living donor transplant recipients increased, as did the duration of dialysis exposure prior to transplantation. There were marked changes in the use of immunosuppressant medications over time: between 1990 and 1994, almost all patients were prescribed cyclosporine and azathioprine; however, by 2000–2003, cyclosporine and azathioprine were prescribed to only 33.3% and 4.0% of patients, respectively, with the majority receiving tacrolimus and mycophenolate instead.

Table 1.  Patient characteristics and comparison of patients transplanted during three time periods: 1990–1994, 1995–1999, 2000–20031
 All patients (N = 16 195)Transplanted 1990–1994 (N = 5862)Transplanted 1995–1999 (N = 5702)Transplanted 2000–2003 (N = 4631)p-Value
  1. 1Percentages or mean (standard deviation) are shown as appropriate unless otherwise indicated.

  2. 2Prescribed at the time of hospital discharge after transplantation.

  3. 3Estimated using the abbreviated equation from the Modification of Diet in Renal Disease Study using serum creatinine values recorded approximately 6 months after the time of transplantation. Nine hundred and twenty-six patients (6.3%) had missing serum creatinine measurements at 6 months and were coded as missing. There were 314 (6.2%), 327 (6.3%) and 285 (6.6%) of patients missing GFR during 1990–1994, 1995–1999 and 2000—2003, respectively.

  4. 4RUCA rural–urban commuting area code—used to classify the extent to which the residence location of each patient was rural or urban (see the text). ESRD = end-stage renal disease; Q1 = first quartile; Q3 = third quartile; GFR = glomerular filtration rate.

Age at transplantation
 Mean (years)33.4 (7.6)33.0 (7.6)33.6 (7.5)33.7 (7.7)<0.0001
   15–19 5.5 5.8 5.0 5.8 0.005
   20–2410.410.810.210.2 
   25–2916.817.916.615.5 
   ≥3067.365.568.268.5 
Race
   White64.067.963.859.2<0.0001
   Black29.626.529.933.0 
   Other 6.4 5.6 6.3 7.8 
Cause of ESRD
   Glomerulonephritis39.636.739.643.2<0.0001
   Diabetes23.224.124.021.2 
   Other37.239.236.435.6 
Duration of dialysis before transplantation
 Median (Q1–Q3) in years2.3 (1.2–3.9)1.9 (1.0,3.2)2.3 (1.2,3.8)3.0 (1.4,4.7)<0.0001
   No dialysis 6.2 6.0 5.7 6.9<0.0001
   <1 year19.623.319.015.7 
   1 to <3 years39.444.940.431.4 
   ≥3 years34.825.834.946.1 
Donor type
   Deceased70.877.270.263.5<0.0001
   Living29.222.829.836.5 
Immunosuppressant medications2
   Cyclosporine71.196.676.333.3<0.0001
   Tacrolimus28.23.423.764.3 
   Azathioprine49.397.338.54.0 
   Mycophenolate46.62.459.684.7 
   Sirolimus 7.4 1.8 4.217.9 
 GFR after transplantation355.4 (20.7)50.7 (18.8)55.9 (20.8)60.1 (21.5)<0.0001
   (mL/min/1.73 m2)     
   <308.911.68.76.1<0.0001
   30–5954.261.253.247.2 
   ≥6036.927.238.146.7 
RUCA group4
   1.0–3.979.479.779.379.20.85 
   4.0–6.010.910.611.210.9 
   >6.0 9.7 9.7 9.5 9.9 
Median household income
 $0–28 99921.520.422.122.10.16 
 $29 000–35 99924.424.024.724.5 
 $36 000–45 99926.927.826.626.3 
 ≥$46 00027.227.826.627.1 

Pregnancy events

There were 530 pregnancies identified in 483 women. The majority of women (N = 439, 90.9%) had a single pregnancy, 41 (8.5%) women had two pregnancies and 3 (0.01%) women had three pregnancies. The pregnancy rate during the first three posttransplant years was markedly lower among women transplanted in more recent years (59 per 1000 in 1990 vs. 20 per 1000 in 2000), with a large decline evident in women transplanted after 1995 (Figure 1). The decrease in the pregnancy rate among women transplanted between 1990 and 2000 was evident in analyses stratified by patient age, race, cause of ESRD, donor type, RUCA group and the median household income (Table 2). In comparison, the pregnancy rate in the general population was greater than 100 per 1000 in each calendar year between 1990 and 2000 (Figure 1), and declined by 10% from 116 per 1000 in 1990 to 104 per 1000 in 2000 (14). The pregnancy rate in the first, second and third posttransplant year was 10 per 1000, 14 per 1000 and 15 per 1000, respectively.

image

Figure 1. Pregnancy rate in transplant recipients and the general population. Pregnancy rates in transplant recipients were determined by the year of transplantation and included pregnancies in the first three posttransplant years while patients retained transplant function. Pregnancy rates were determined in the 12 690 women transplanted before December 31, 2000, who had the potential for 3 years of posttransplant follow-up for ascertainment of pregnancy-related events. Pregnancy rates in the general population are for women in the same age range (15–45 years) as transplant recipients, but are determined by the calendar year (14).

Download figure to PowerPoint

Table 2.  Pregnancy rate per 1000 women in select patient subgroups during three time periods
VariablesOverall1990–19941995–19992000–2003p-Value
  1. 1RUCA rural–urban commuting area code—used to classify the extent to which the residence location of each patient was rural or urban (see the text).

Age at transplant
 <15–1954.7102.3 42.1 7.5  0.0006
 20–2493.5154.3 72.038.2<0.0001
 25–2963.191.859.126.4<0.0001
 ≥3013.924.011.1 5.4<0.0001
Race
 White31.148.026.912.0<0.0001
 Black34.563.726.413.7<0.0001
 Other41.494.827.9 5.6<0.0001
Cause of end-stage renal disease
 Glomerulonephritis41.574.833.215.0<0.0001
 Diabetes20.733.216.8 8.20.0008
 Other30.949.326.510.9<0.0001
Deceased donor30.146.425.811.9<0.0001
Living donor39.183.029.414.2<0.0001
RUCA group1
 1.0–3.934.355.629.913.1<0.0001
 4.0–6.030.060.620.8 4.3<0.0001
 >6.027.351.022.4 7.10.0006
Median household income
 $0–28 99942.672.237.614.9<0.0001
 $29 000–35 99927.443.527.0 8.1<0.0001
 $36 000–45 99931.948.726.410.1<0.0001
 ≥$46 00030.653.019.016.3<0.0001

Factors associated with pregnancy

Compared to women aged 25–29 years, pregnancy was more likely in women aged 20–24 years and less likely in women aged 30–45 years (Table 3). Compared to white women, Black women had a similar likelihood of pregnancy. Pregnancy was less likely in women with ESRD due to diabetes compared to women with glomerulonephritis. The likelihood of pregnancy did not differ between deceased and living donor transplant recipients; however, the likelihood of pregnancy in women with pretransplant dialysis duration ≥3 years was higher compared to preemptively transplanted women. The adjusted likelihood of pregnancy was markedly lower among women transplanted in more recent years. Women treated with tacrolimus had a higher likelihood of pregnancy than those treated with cyclosporine, and women treated with mycophenolate had a lower likelihood of pregnancy than women treated with azathioprine. Women with GFR ≥60 mL/min/1.73 m2 posttransplantation (compared to GFR <30 mL/min/1.73 m2) showed a trend in the increased likelihood of pregnancy.

Table 3.  Factors associated with pregnancy
FactorHazard ratio95% CIp-Value
  1. 1Calculated using serum creatinine measurements recorded approximately 6 months after transplantation using the Modification in Diet and Renal Disease Study formula (GFR = 186 × app−1.154×[patient age]−0.203× 0.742 ×[1.212 if black]) (16). GFR estimates were not available for N = 2504 (15.5%) of patients and these patients were coded as missing.

  2. 2RUCA rural–urban commuting area code—used to classify the extent to which the residence location of each patient was rural or urban (see the text).

  3. CI = confidence interval.

Age at transplantation (years)
 15–190.850.60–1.190.34
 20–241.421.12–1.810.004
 25–29Reference
 30–440.260.20– 0.32<0.0001
Race
 WhiteReference
 Black1.150.93–1.430.20
 Other1.340.96–1.890.09
Cause of end-stage renal disease
 GlomerulonephritisReference
 Diabetes0.630.48–0.840.001
 Other0.770.63–0.940.01
Living donor (reference deceased donor)1.070.88–1.310.48
Duration of dialysis treatment before transplantation
 No dialysisReference
 <1 year1.330.82–2.150.24
 1 to <3 years1.440.90–2.300.13
 ≥3 years1.641.01–2.660.05
Year of transplantation
 1990–19941.00
 1995–19990.570.45–0.73<0.0001
 2000–20030.380.25–0.57<0.0001
Immunosuppressant medication treatment
 Tacrolimus (reference cyclosporine)1.481.10–2.000.01
 Mycophenolate mofetil (reference azathioprine)0.460.33–0.63<0.0001
 Sirolimus (reference no sirolimus)1.370.77–2.430.29
Glomerular filtration rate (mL/min/1.73 m2)1
 <30Reference
 30–591.300.85–1.980.22
 ≥601.480.96–2.280.07
RUCA group2
 1.0–3.9Reference
 4.0–6.00.820.58–1.150.25
 >6.00.810.81–1.160.25
Median household income
 $0–28 9991.361.04–1.770.03
 $29 000–35 9990.920.69–1.220.56
 $36 000–45 9991.100.85–1.410.48
 ≥$46 000Reference

Fetal outcomes:  Information regarding fetal outcomes was available for 453 (85%) of 530 pregnancies (Table 4). There were 251 (55.4%) of 453 live births, among which 82 (32.7%) were preterm and 127 (50.6%) were cesarean deliveries. Spontaneous and therapeutic abortions accounted for 47.5% and 8.4% of the (n = 202) fetal losses. The majority (55.3%) of pregnancies during the first posttransplant year ended in fetal loss, while 40.2% and 37.6% of pregnancies during the second and third posttransplant year ended in fetal loss (Table 4).

Table 4.  Fetal outcomes by posttransplant year of conception1
Fetal outcomeAny year (N = 453) N (%)Year 1 (N = 152) N (%)Year 2 (N = 184) N (%)Year 3 (N = 117) N (%)
  1. 1Fetal outcomes could be determined in 453 of 530 pregnancies.

Live births251 (55.4)68 (44.7)110 (59.8) 73 (62.4)
Fetal loss202 (44.6)84 (55.3)74 (40.2)44 (37.6)
Therapeutic abortion17 (3.8)7 (4.6)6 (3.3)4 (3.4)
Other fetal loss185 (40.8)77 (50.7)68 (37.0)40 (34.2)
 Spontaneous abortion 96 (21.2)36 (23.7)38 (20.7)22 (18.8)
 Abortion not otherwise specified 65 (14.3)32 (21.1)22 (12.0)11 (9.4) 
 Stillbirth 7 (1.5)2 (1.3)4 (2.2)1 (0.9)
 Ectopic pregnancy12 (2.6)3 (2.0)4 (2.2)5 (4.3)
 Abnormal product of conception 5 (1.1)4 (2.6)01 (0.9)

The overall live birth rate was 19 per 1000, and the live birth rate fell in parallel with the decrease in the pregnancy rate between 1990 and 2000 from 28.5 per 1000 to 6.2 per 1000 (Figure 2). In comparison, the live birth rate in the general population decreased from 70.9 per 1000 to 65.9 per 1000 between 1990 and 2000 (Figure 2) (17).

image

Figure 2. Live birth rate per 1000 female transplant recipients. Live birth rates in transplant recipients were determined by the year of transplantation and included pregnancies in the first three posttransplant years while patients retained transplant function. Live birth rates were determined in the 12 690 women transplanted before December 31, 2000, who had the potential for three years of posttransplant follow-up for the ascertainment of pregnancy-related events. Live birth rates in the general population are for women in the same age range (15–45 years) as transplant recipients, but are determined by the calendar year (14).

Download figure to PowerPoint

There was no change in the proportion of pregnancies resulting in a live birth during the study, and therefore the decrease in the live birth rate between 1990 and 2000 was related to the falling pregnancy rate rather than an increase in fetal loss during the study. However, the cause of fetal loss changed markedly over time. Between 1990–1994 and 1995–2000 the proportion of pregnancies ending in therapeutic abortion decreased from 18.7 to 6.2%, while the proportion of pregnancies ending due to all other causes increased from 25.0 to 36.7%.

Factors associated with fetal loss

In the multivariate analysis that included adjustment for recipient age, race, cause of ESRD, donor type, duration of dialysis treatment prior to transplantation, transplant year, immunosuppressant medication treatment, posttransplant year of pregnancy, within-neighborhood median household income, RUCA score and glomerular filtration rate after transplantation, only patient race was clearly associated with fetal loss. Compared to white women, Black women had a higher odds of fetal loss (odds ratio (OR) 1.91, 95% confidence interval (CI) 1.16–3.14, p = 0.01). Other factors that were weakly associated with fetal loss included the cause of ESRD (compared to women with glomerulonephritis, diabetic women had a higher odds of fetal loss (OR 1.87, 95% CI 0.92–3.79, p = 0.08)) and the median with neighborhood household income (compared to patients with income <$29 000, patients with income ≥$46 000 had a lower odds of fetal loss (OR 0.53, 95% CI 0.28–1.00, p = 0.05)). Of note, although there was a borderline increase in the risk of fetal loss during the first transplant year compared with subsequent years, there was no overall association between the timing of pregnancy after transplantation and fetal loss. Compared to pregnancies in the third posttransplant year, the odds of fetal loss for pregnancies during the first posttransplant year was 1.65 (95% CI 0.93–2.91), while the odds for pregnancies in the second posttransplant year was 1.13 (95% CI 0.66–1.96). Similar results were obtained when therapeutic abortions were excluded as a cause of fetal loss (data not shown).

Discussion

  1. Top of page
  2. Abstract
  3. Background
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References
  9. Appendix

This study provides the pregnancy rate and live birth rate among Medicare-insured kidney transplant recipients in the United States based on data from a mandatory population-based registry. Previous data regarding pregnancy after kidney transplantation are primarily derived from voluntary data sources or single-center case series. In this study, the pregnancy rate in kidney transplant recipients was markedly lower and declined more rapidly than in the general American population.

Only 55.4% of pregnancies ended in a live birth, a proportion far lower than the 74–79% previously reported in studies from voluntary registries including the National Transplantation Pregnancy Registry in the United States (which included pregnancies between 1991 and January 2005, as well as other) and the UK Transplant Pregnancy Registry (which included pregnancies between 1994 and 2001) (5,8). These registries may also have included pregnancies outside the specified years because of retrospective data entry. The similar reported live birth rate in these registries involving different patient populations suggested that the different reporting methods used in the NTPR and UK registries had little impact on the capture of pregnancy-related events. Our findings suggest that both registries may have underestimated fetal loss. Our findings regarding fetal loss are more in keeping with the conclusions from a review paper by Davison that estimated that just under 40% of conceptions do not go beyond the first trimester, but of those that do, greater than 90% end successfully (18).

Of potential concern, the proportion of pregnancies ending in fetal loss did not decrease in parallel with secular decreases in the number of therapeutic abortions, and Black and diabetic women as well as women with low median household incomes were at increased risk of fetal loss. In multivariate analysis, no overall association was found between the timing of pregnancy after transplantation and the risk of fetal loss. However, the majority of pregnancies during the first posttransplant year ended in fetal loss, and analyses examining the timing of fetal loss showed a borderline increase in risk during the first year after transplantation as compared with the third year. In addition, the risk of fetal loss during the second and third year following transplantation appeared comparable, although the number of pregnancies during each year was relatively small.

These findings support current recommendations regarding the timing of pregnancy after transplantation, including the suggestion that a 2-year waiting period between transplantation and attempts at conception may be overly conservative (12). The high overall risk of fetal loss that we observed also validates the opinion-based recommendation that pregnancies after transplantation should be managed by transplant professionals and specialists in high-risk pregnancies whenever possible (12). Although prospective studies are required to confirm and understand the basis of these results, our findings may be valuable to clinicians providing counseling to women wanting to have children after kidney transplantation.

There are a number of possible explanations for the low pregnancy rate in kidney transplant recipients, including incomplete restoration of fertility after transplantation, an effect of immunosuppressant medications used either before or after transplantation, or deliberate avoidance of pregnancy. Alternatively, the capture of pregnancies from claims data may be incomplete since our use of Medicare claims to identify pregnancies and pregnancy-related events is not validated. Of note, induced abortions are not insured under Medicare and thus some patients may have obtained abortions outside of the Medicare system potentially leading to underestimation of the therapeutic abortions and pregnancies reported in our study. We are unaware of changes in Medicare claims procedures for pregnancy and related events or of a change in the management of pregnancy in the United States that could explain the marked drop in the pregnancy rate among transplant recipients during the study period.

The reasons for the dramatic decline in pregnancy during the study cannot be determined. Indeed we cannot determine if the decrease in pregnancy over time reflects fewer unwanted pregnancies or a decrease in fertility. A number of factors including volitional and sociological changes may be important contributing factors. Importantly, during the study period, educational initiatives supported by the American Society of Transplantation and other professional societies were introduced to increase the use of contraception after transplantation in order to avoid unwanted and or high-risk pregnancies. It is also possible that pretransplant factors may contribute to the observed decrease in pregnancy over time. Although we did not find any evidence that women with longer duration of pretransplant dialysis duration had a decreased likelihood of pregnancy, it is possible that increased utilization of immunosuppressants such as cyclophosphamide to treat glomerulonephritis prior to transplantation may have contributed to the decline in pregnancy after transplantation among the subgroup of patients with glomerulonephritis as their primary renal disease.

The greater decrease in the pregnancy rate in transplant recipients compared to the general population suggests a role for transplant-specific factors. It is plausible that uncertainty regarding the safety of newly approved immunosuppressant medications in pregnancy may be an important factor contributing to the dramatic decline in pregnancy during the study. The precipitous decrease in pregnancy in 1995 coincided with the introduction of mycophenolate mofetil (MMF). Between 1995 and 1996, the use of MMF in new kidney transplant recipients increased from 12.2 to 55.3%. The adjusted likelihood of pregnancy in MMF-treated patients was 56% lower than in azathioprine-treated patients (Table 2). Like most immunosuppressant medications, MMF was approved for use with only limited information regarding its safety in pregnancy, and it is likely that many physicians counseled MMF-treated patients to avoid pregnancy. Because MMF is associated with a lower risk of acute rejection than azathioprine (19), physicians may have been reluctant to switch patients considering pregnancy from MMF to azathioprine, even if this precluded pregnancy.

The Food and Drug Administration classification of the safety of immunosuppressant medications in pregnancy categorizes most immunosuppressant drugs as ‘human risk not ruled out’. Information regarding the safety of immunosuppressant drugs in pregnancy accumulates slowly and is hampered by the absence of comprehensive postmarketing ascertainment of pregnancy outcomes in women treated with immunosuppressant medications. For example, definitive information regarding the increased risk of birth defects with MMF only became available more than a decade after this drug was approved for use in kidney transplantation (20). Given these realities, appropriate education and counseling regarding the risk of rejection and available immunosuppressant drug treatment for women wishing to become pregnant after transplantation is clearly essential. Failure to provide this information may be an important remediable factor contributing to the dramatic drop in pregnancy observed in this study.

Our study has limitations that should be considered when interpreting results. First, the study only included Medicare-insured patients and provides information regarding pregnancy only during the first three posttransplant years while patients maintained Medicare insurance coverage. Information regarding the use of immunosuppressant medications was obtained at the time of transplantation and we do not have information regarding immunosuppressant drug use at the time of pregnancy. Further, we do not have information regarding the incidence of birth defects. No conclusions regarding the safety of individual immunosuppressant medications should be drawn from this study and published guidelines should be used to guide the use of immunosuppressant drugs in pregnancy (12). Ascertainment of the timing of pregnancy in the current study was based on pregnancy-related claims and assumptions were made regarding the time of conception based on the type of pregnancy-related claim (see the Methods section). Importantly, the strength of our study is that it provides information about the likelihood of pregnancy and pregnancy outcomes in the population of Medicare-insured transplant recipients and thus avoids the potential shortfalls of registries dependent on voluntary reporting or patient recall. However, our study cannot provide the detailed information necessary to identify the impact of immunosuppressant medications on maternal and fetal outcomes. Our data reinforce the continued need for detailed information to be submitted to registries such as the National Transplantation Pregnancy Registry (NTPR).

In conclusion, the pregnancy rate after kidney transplantation was far lower and declined more rapidly than reported in the general American population. In addition, the proportion of pregnancies ending in a live birth was far lower than that previously reported in studies of voluntary registries. These findings have implications for female patients considering pregnancy after transplantation, as well as their partners and physicians.

Acknowledgments

  1. Top of page
  2. Abstract
  3. Background
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References
  9. Appendix

Dr Gill was supported by the Michael Smith Foundation for Medical Research. Dr Tonelli was supported by a Population Health Investigator Award from the Alberta Heritage Foundation for Medical Research, and by a joint initiative by Alberta Health and Wellness and the University of Alberta. Dr Tonelli was also supported by a New Investigator Award from the Canadian Institute of Health Research.

The data reported here have been supplied by the USRDS. The interpretation and reporting of these data are the responsibility of the authors and in no way should be seen as offical policy of the US government.

References

  1. Top of page
  2. Abstract
  3. Background
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References
  9. Appendix
  • 1
    Wolfe R, Ashby V, Milford E et al. Comparison of mortality in all patients on dialysis, patients on dialysis awaiting transplantation, and recipients of a first cadaveric transplant. N Engl J Med 1999; 341: 17251730.
  • 2
    Laupacis A, Keown P, Pus N et al. A study of the quality of life and cost-utility of renal transplantation. Kidney Int 1996; 50: 235242.
  • 3
    Winkelmayer WC, Weinstein MC, Mittleman MA, Glynn RJ, Pliskin JS. Health economic evaluations: The special case of end-stage renal disease treatment. Med Decis Making 2002; 22: 417430.
  • 4
    McKay DB, Josephson MA. Pregnancy in recipients of solid organs—Effects on mother and child. N Engl J Med 2006; 354: 12811293.
  • 5
    Armenti VT, Radomski JS, Moritz MJ et al. Report from the National Transplantation Pregnancy Registry (NTPR): Outcomes of pregnancy after transplantation. Clin Transpl 2004: 103114.
  • 6
    Armenti VT, Radomski JS, Moritz MJ et al. Report from the National Transplantation Pregnancy Registry (NTPR): Outcomes of pregnancy after transplantation. Clin Transpl 2005: 6983.
  • 7
    Briggs JD, Jager K. The first year of the new ERA-EDTA Registry. Nephrol Dial Transplant 2001; 16: 11301131.
  • 8
    Sibanda N, Briggs JD, Davison JM, Johnson RJ, Rudge CJ. Pregnancy after organ transplantation: A report from the UK Transplant pregnancy registry. Transplantation 2007; 83: 13011307.
  • 9
    Davison JM, Lind T, Uldall PR. Planned pregnancy in a renal transplant recipient. Br J Obstet Gynaecol 1976; 83: 518527.
  • 10
    Meier PR, Makowski EL. Pregnancy in the patient with a renal transplant. Clin Obstet Gynecol 1984; 27: 902913.
  • 11
    Areia A, Galvao A, Pais MS, Freitas L, Moura P. Outcome of pregnancy in renal allograft recipients. Arch Gynecol Obstet 2009; 279: 273277.
  • 12
    McKay DB, Josephson MA, Armenti VT et al. Reproduction and transplantation: Report on the AST Consensus Conference on Reproductive Issues and Transplantation. Am J Transplant 2005; 5: 15921599.
  • 13
    European best practice guidelines for renal transplantation. Section IV: Long-term management of the transplant recipient: IV.10. Pregnancy in renal transplant recipients. Nephrol Dial Transplant 2002; 17(Suppl 4): 5055.
  • 14
    Ventura SJ, Abma JC, Mosher WD, Henshaw SK. Estimated pregnancy rates by outcome for the United States, 1990–2004. Natl Vital Stat Rep 2008; 56: 125, 28.
  • 15
    Axelrod DA, Guidinger MK, Finlayson S et al. Rates of solid-organ wait-listing, transplantation, and survival among residents of rural and urban areas. JAMA 2008; 299: 202207.
  • 16
    Levey A, Greene T, Kusek J, Beck G. A simplified equation to predict glomerular filtration rate from serum creatinine. J Am Soc Nephrol 2000; 11: 155A.
  • 17
    Martin JA, Hamilton BE, Sutton PD et al. Births: Final data for 2005. Natl Vital Stat Rep 2007; 56: 1103.
  • 18
    Davison JM. Renal transplantation and pregnancy. Am J Kidney Dis 1987; 9: 374380.
  • 19
    The Tricontinental Mycophenolate Mofetil Renal Transplantation Study Group. A blinded, randomized clinical trial of mycophenolate mofetil for the prevention of acute rejection in cadaveric renal transplantation. Transplantation 1996; 61: 10291037.
  • 20
    Black Box Warning Mycophenolate Mofetil in Pregnancy. 2007. Available at: http://www.fda.gov/medwatch/safety/2007/CellCept_dearhcpoct07.pdf. Accessed December 16, 2008.

Appendix

  1. Top of page
  2. Abstract
  3. Background
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References
  9. Appendix

Appendix: ICD9-CM Codes

(a) Diagnosis codes:

Ectopic and molar pregnancy: 630, 631, 632, 633, 633.x, 633.xx

Pregnancy with abortive outcome: 634, 634.x, 634.xx, 635, 635.x, 635.xx, 636, 636.x, 636.xx, 637, 637.x, 637.xx, 638, 638.x, 639, 639.x

Complications related to pregnancy: 640, 640.x, 640.xx, 641, 641.x, 641.xx, 642, 642.x, 642.xx, 643, 643.x, 643.xx, 644, 644.x, 644.xx, 645, 645.x, 645.xx, 646, 646.x, 646.xx, 647, 647.x, 647.xx, 648, 648.x, 648.xx

Normal delivery and other indications for care in pregnancy, labor and delivery: 650, 651, 651.x, 651.xx, 652, 652.x, 652.xx, 653, 653.x, 653.xx, 654, 654.x, 654.xx, 655, 655.x, 655.xx, 656, 656.x, 656.xx, 657, 657.x, 657.xx, 658, 658.x, 658.xx, 659, 659.x, 659.xx

Complications occurring in the course of labor and delivery: 660, 660.x, 660.xx, 661, 661.x, 661.xx, 662, 662.x, 662.xx, 663, 663.x, 663.xx, 664, 664.x, 664.xx, 665, 665.x, 665.xx, 666, 666.x, 666.xx, 667, 667.x, 667.xx, 668, 668.x, 668.xx, 669, 669.x, 669.xx

Complications of the puerperium: 670, 670.x, 670.xx, 671, 671.x, 671.xx, 672, 672.x, 672.xx, 673, 673.x, 673.xx, 674, 674.x, 674.xx, 675, 675.x, 675.xx, 676, 676.x, 676.xx

Supervision of pregnancy and postpartum care: v22, v220–2, v23, v23.0–3, v23.4, v23.4x, v23.5, v23.7, v23.8, v23.8x, v23.9, v24, v24.0–2,

Outcome of delivery: v27, v27.0–7, v27.9

Antenatal screening: v28, v28.0–6, v28.8, v28.9

(b) Obstetrical procedure codes: 72, 72.0, 72.1, 72.2, 72.2x, 72.3, 72.3x, 72.4, 72.5, 72.5x, 72.6, 72.7, 72.7x, 72.8, 72.9, 73.0, 73.0x, 73.1, 73.2, 73.2x, 73.3, 73.4, 73.5, 73.5x, 73.6, 73.8, 73.9, 73.9x, 74, 74.0-4, 74.9, 74.9x, 75.0-2, 75.3, 75.3x, 75.4, 75.5, 75.5x, 75.6, 75.6x, 75.7, 75.8, 75.9, 75.9x