Annual renal replacement therapy (RRT) initiation rates per million persons in Canada between 1981 and 1996.
Probability of receiving a kidney transplant among Canadian patients undergoing renal replacement therapy between 1981 and 1986. Male and female cohorts (n = 13,255 each) were matched on age, calendar period, province, and diabetes status.
Schaubel DE, Stewart DE, Morrison HI, Zimmerman DL, Cameron JI, Jeffery JJ, Fenton SSA. Sex Inequality in Kidney Transplantation Rates. Arch Intern Med. 2000;160(15):2349-2354. doi:10.1001/archinte.160.15.2349
Copyright 2000 American Medical Association. All Rights Reserved. Applicable FARS/DFARS Restrictions Apply to Government Use.2000
Men in the United States undergoing renal replacement therapy are more likely than women to receive a kidney transplant. However, the ability to pay may, in part, be responsible for this finding.
To compare adult male and female transplantation rates in a setting in which equal access to medical treatment is assumed.
Using data from the Canadian Organ Replacement Register, the rate of first transplantations was computed for the 20,131 men and the 13,458 women aged 20 years or older who initiated renal replacement therapy between January 1, 1981, and December 31, 1996. Poisson regression analysis was used to estimate the male-female transplantation rate ratio, adjusting for age, race, province, calendar period, underlying disease leading to renal failure, and dialytic modality. Actuarial survival methods were used to compare transplantation probability for covariable-matched cohorts of men and women.
Men experienced 20% greater covariable-adjusted kidney transplantation rates relative to women (rate ratio, 1.20; 95% confidence interval, 1.13-1.27). The sex disparity was stronger for cadaveric transplants (rate ratio, 1.23) compared with those from living donors (rate ratio, 1.10). The 5-year probability of receiving a transplant was 47% for men and 39% for women within covariable-matched cohorts (P<.001). The sex disparity in transplantation rates increased with increasing age. The sex effect was weaker among whites and Oriental persons (Chinese, Japanese, Vietnamese, Cambodian, Laotian, Filipino, Malaysian, Indonesian, and Korean) and stronger among blacks, Asian Indians (Indian, Pakistani, and Sri Lankan), and North American Indians (aboriginal).
Since survival probability and quality of life are superior for patients who undergo transplantation relative to those who undergo dialysis, an increased effort should be made to distribute kidneys available for transplantation more equitably by sex among patients undergoing renal replacement therapy.
THE PUBLIC health importance of end-stage renal disease (ESRD) derives from the high mortality rate and impaired quality of life among patients and the high cost of renal replacement therapy (RRT).1 Incidence rates and the prevalence of ESRD are increasing in Canada,2 the United States,3 Europe,4 and Australia,5 with further increases projected for the future.6 Kidney transplantation is the preferred form of therapy for survival probability and quality of life. Unfortunately, the number of kidneys available for transplantation has not kept pace with demand.2- 5
Studies7- 13 from the United States have shown that male patients undergoing RRT are more likely to receive a transplant than female patients. However, sex-specific differences in income, supplementary insurance, or both, necessary to cover additional costs associated with transplantation, could explain the disparity. This investigation assesses whether male patients undergoing RRT are more likely than female patients to receive a kidney transplant in a setting in which a patient's ability to pay therapy-related costs should not distort the results.
Data were obtained from the Canadian Organ Replacement Register, a population-based, nationwide organ failure registry2 operated by the Canadian Institute for Health Information. Clinical history data (dialytic modality assignments and switches, transplantations, and transplantation failures) are submitted annually by each of the 86 renal centers. Deaths are reported along with the other follow-up information. Data on demographics (date of birth, sex, province of residence, and race), comorbid conditions, and the underlying cause of renal failure are collected from all patients at RRT initiation by the centers.
Data were available on all patients initiating therapy between January 1, 1981, and December 31, 1996. Since the focus of this investigation was on transplantation rates for adults, patients aged 0 to 19 years were excluded. The total study population (N = 33,589) consisted of 20,131 men and 13,458 women. Patients began follow-up on the date of RRT initiation, and were followed up until death, loss to follow-up, date of first transplantation, or December 31, 1996 (ie, the earliest thereof). Transplantation rates were computed as the ratio of first transplantations to person-time undergoing dialysis. Patients were classified by sex, primary renal diagnosis (PRD) (the underlying disease that resulted in renal failure), province of therapy initiation, and race. Race was categorized as follows: Oriental (Chinese, Japanese, Vietnamese, Cambodian, Laotian, Filipino, Malaysian, Indonesian, and Korean); Asian Indian (Indian, Pakistani, and Sri Lankan); North American Indian (aboriginal); black; white; and other or unknown. Person-years within these categories were classified by age and calendar period, which were time-dependent variables in that patients could contribute to more than 1 category during the follow-up period. Comorbidity data were available only for patients initiating RRT during the 1988 to 1996 period, and were only used for supplementary analysis.
Actuarial survival analysis14 was used to estimate the 5-year probability of transplantation for male and female cohorts (n = 13,255 each) matched by age, diabetes status, province, and calendar period. Poisson regression analysis15,16 was used to estimate the effect of sex on transplantation rates, adjusted for all other covariables. That is, the log (transplantation rate) was modeled as a linear function of parameters, representing each variable. Separate models were fitted by age and race to evaluate the degree of interaction between these covariables and sex.
Figure 1 displays RRT initiation rates per million persons for men and women aged 20 years or older. Great increases in ESRD incidence were experienced by men and women, as the incidence rate rose from 85.3 to 194.5 per million among men and from 50.7 to 119.1 per million for women. For every year between 1981 and 1996, the incidence rate was greater for men than for women, the difference tending to increase over time from 34.6 per million in 1981 to 75.4 per million in 1996.
Demographic and clinical characteristics of the study population are presented by sex in Table 1. The proportion of men initially undergoing hemodialysis was higher than that for women, whereas women were more likely to undergo peritoneal dialysis. A greater percentage of men underwent transplantation relative to women. Equal proportions of men and women had died by the end of the observation period.
As presented in Table 2, among underlying diseases leading to ESRD, men were more likely to have glomerulonephritis than women. Women were more likely to have diabetes as a PRD than men, while a greater percentage of men had renal vascular disease than women.
In Table 3, predialysis comorbidity prevalence is examined by sex among patients initiating RRT between 1988 and 1996. Men were more likely to have cardiovascular disease, chronic obstructive lung disease, malignant neoplasms, and any "other illness" expected to significantly reduce 5-year survival probability.
In Table 4, transplantation rates by age group and calendar period are presented. Generally, transplantation rates increased during the 1986 to 1990 period, relative to the 1981 to 1985 period, then decreased during the 1991 to 1996 period. Within each age or period stratum, men had higher transplantation rates than women.
Figure 2 displays the probability of kidney transplantation among patients undergoing RRT. Male and female cohorts (n = 13,255 each) were matched by age, calendar period, province, and diabetes status. The 5-year probability of receiving a kidney transplant was estimated at 47% (95% confidence interval [CI], 46%-49%) for men and at 39% (95% CI, 38%-41%) for women (P<.001).
Rate ratios (RRs) from the multivariate analysis are presented in Table 5. The covariable-adjusted male-female transplantation RR was estimated at 1.20 (95% CI, 1.13-1.27), indicating that male patients undergoing RRT were 20% more likely to receive a kidney transplant relative to female patients after adjusting for age, calendar period, province, race, PRD, and dialytic modality. Results did not change on adjustment for comorbid conditions. The sex disparity was stronger for cadaveric transplantation rates compared with those for living donors.
The sex effect strengthened monotonically with increasing age, from an RR of 1.01 (95% CI, 0.88-1.17) among patients aged 20 to 29 years to an RR of 1.37 (95% CI, 1.18-1.59) among patients aged 60 years or older (Table 6). An important variation in the sex effect by race was also observed. The effect was weakest among whites and Oriental persons and strongest among Asian Indians, blacks, and North American Indians.
Canadian men undergoing RRT between January 1, 1981, and December 31, 1996, experienced significantly greater kidney transplantation rates relative to women after adjusting for other medical and social factors related to transplantation suitability. The male-female discrepancy in transplantation rates increased with increasing age, and was more prominent among blacks, Asian Indians, and North American Indians and less pronounced among whites and Oriental persons. Since survival probability and quality of life are generally better with a functioning transplant relative to dialysis, it would appear that women are being denied a superior lifesaving therapy. We estimate that between 1981 and 1996 in Canada, 588 women died while undergoing dialysis who would have survived until the end of our study had women experienced the same covariable-adjusted transplantation rates as men.
Our investigation features certain strengths rarely observed simultaneously in previous studies. The study population was, in effect, a nationwide census of patients with ESRD, as opposed to a sample. The period of observation spanned 16 years. An incident patient cohort was used, and no follow-up time after RRT initiation was excluded. Sex-specific rates of cadaveric and living donor transplants were compared in a setting in which the ability to pay therapy-related costs should not be a factor.
The increases in kidney transplantation rates for men relative to women observed in our investigation are consistent with, although less extreme than, results generated by US studies. In one US study of 14,721 patients undergoing RRT from 1981 to 1985, Held et al13 found a male-female transplantation RR of 1.34, adjusted for age, race, PRD, income, and dialytic modality (P<.01). In a more recent national US study of a random sample of 4118 patients with ESRD from 1986 to 1987, Gaylin et al10 found that women were 70% as likely to receive a transplant as were men, adjusting for age, race, PRD, dialytic modality, income, and comorbidity (P<.01). In a recent study by Bloembergen et al,12 women were 10% less likely to receive a living-related transplant than men (P<.001) between 1991 and 1993, a finding quite similar to ours for living donor transplants. Also consistent with our results, the sex effect in the study by Bloembergen et al was not significant among younger adults.
In the United States, the reduction in Medicare eligibility after transplantation17 may have served as a greater deterrent to transplantation for women receiving RRT relative to men because of sex differences in income and insurance coverage. Men may have been more able to afford additional costs related to transplantation, because of either increased income or supplemental insurance plans obtained through their employer. Certain past studies10,13 used data on income. However, median ZIP code income can provide a poor estimate of a patient's actual household income. Moreover, household income itself is, at best, a surrogate measure of a patient's ability to cover transplantation-related costs, an entity that would indeed be difficult to quantify. All Canadians have equal access to health care, at least in theory. Thus, the more pronounced sex effect observed in US studies relative to ours might be explained by lack of (or incomplete) adjustment for income.
Because of the unavailability of waiting list data, we are unable to determine whether women undergoing RRT are less likely to be placed on a waiting list for transplantation or whether their waiting times are longer than those of men after being placed on a waiting list. A recent US study8 that compared the likelihood of being placed on a waiting list with cadaveric transplantation rates among patients placed on waiting lists by sex found that sex differences occurred at both stages of the process.
Sex inequalities among patients undergoing RRT are not restricted to transplantation. Canadian men have been shown to be more likely than women to undergo hemodialysis as opposed to peritoneal dialysis.18 Sex discrepancies may even extend to receipt of RRT itself among populations with ESRD, since RRT initiation rates in Canada and the United States3 are significantly higher for men than for women. Sex inequality in RRT rates may simply reflect true discrepancies in underlying ESRD incidence. However, based on a review of death certificates, Kjellstrand19- 21 and colleagues hypothesize that sex bias on the part of clinicians is, in part, responsible for the difference, given that women in the United States and Sweden were more likely than men to die of ESRD without undergoing RRT.
Sex inequalities exist in other areas of medicine as well. For example, for coronary heart disease, the leading cause of death among women, male patients are reportedly treated more aggressively than female patients, independent of disease severity.22 Women experiencing chest pain reportedly wait longer than men in emergency departments before they are examined.23 Sex bias in therapy selection has been suspected in each of the previously listed scenarios.
However, physician sex bias is not the only possible reason why women are less likely than men to receive a kidney transplant. Women might be poorer transplantation candidates with respect to variables not explicitly measured in our investigation. A biological rationale for preferentially selecting men over women for kidney transplantation relates to immune function (ie, panel-reactive antibody sensitization) and thus the propensity for graft rejection. Although panel-reactive antibody sensitization status is obtained by the Canadian Organ Replacement Register, at the time of this investigation, the high percentage of missing values precluded its usefulness for analysis. The frequency of occurrence and the level of panel-reactive antibody sensitization are higher in women than men.24 In a study of midwestern US patients undergoing renal transplantation, Kjellstrand7 observed a significantly elevated percentage of patients who underwent transplantation and RRT among men relative to women, which persisted on exclusion of patients who had cytotoxic antibodies against 90% or more of random donors, but disappeared when the cut point was lowered to 50% or more.
As discussed by Sabatini,25 an alternative hypothesis is that women are less likely to desire transplantation because of increased risk aversion, or decreased willingness to tolerate the adverse effects of immunosuppression, relative to men. Also, men might be less agreeable than women to the inconvenience and time constraints characteristic of long-term dialysis. To our knowledge, little research has been conducted on sex differences in patient preference of transplantation over dialysis. The idea that women may be more risk averse in medical decisions and, hence, in deciding whether to be placed on a waiting list for transplantation may be supported by a US study26 of patients with severe heart failure, in which women were found to be more likely than men to refuse cardiac transplantation. However, Kjellstrand et al27 observed no difference in those wanting renal transplantation among Swedish men and women undergoing long-term dialysis. Alternatively, psychosocial or cultural factors may play a role and result in women being more likely to refuse transplantation and less likely to request, accept, or be offered organ donation from their families.12,25
If sex bias exists independently of medical suitability and desire for transplantation, women are either actively or unintentionally discriminated against, perhaps due to overestimation of medical risks. It has been conjectured that such unintentional discrimination might preferentially affect elderly women,28 consistent with our observation that the male-female transplantation RR increased with increasing age.
The male-female RR for transplantation was greatest among blacks, Asian Indians, and North American Indians and was less pronounced among whites and Oriental persons. Reasons for such variability in the sex effect by race are unclear, but may relate to cultural differences that influence the probability of refusing transplantation or being offered organ donation from family members.
We provide strong evidence that sex inequalities exist for kidney transplantation among patients undergoing RRT. Because health care in Canada is, at least in theory, equally accessible to all, differences between men and women with respect to income, insurance coverage, or both should not underlie our findings. We cannot rule out the possibility that men are preferentially referred for transplantation over women for valid medical reasons for which we have no data, particularly immunologic sensitivity. A future study in which data were collected on comorbidity severity and panel-reactive antibody sensitization status would provide an invaluable contribution to the literature. Ideally, the incorporation of waiting list data would permit the determination of where in the process the difference in kidney transplantation rates arises.
Recently, sex inequality in medicine has received increased attention in the scientific literature.29 Further research is clearly needed in this area, particularly since physicians are often reluctant to even consider the possibility that sex bias exists. Sex differences in assertiveness, motivation, attitudes, and beliefs on the part of the patient may also be contributory. Whether sex bias in kidney transplantation rates is existent or nonexistent, deliberate or unintentional, inequality with respect to delivered therapy is evident and warrants further investigation and possible intervention.
Accepted for publication February 1, 2000.
The collection of data and the maintenance of the Canadian Organ Replacement Register are made possible by the collaboration of the 86 individual renal programs across Canada. The contribution of the current and past full-time staff assigned to the register at the Canadian Institute for Health Information (formerly, Hospital Medical Records Institute) also has been essential to the success of the register. The Canadian Society of Nephrology, the Canadian Transplant Society, the Canadian Association of Transplantation, and the Canadian Association of Nephrology, nurses and technicians, and their constituent members also have made an essential contribution to the register since its inception in 1981. The Canadian Organ Replacement Register is funded 15% by the federal department of health and 80% by the provincial and territorial departments of health based on population. The Kidney Foundation of Canada and the health care industry provide approximately 5% of the funding.
Reprints: Stanley S. A. Fenton, MD, FRCPI, Division of Nephrology, Toronto General Hospital, University Health Network, 200 Elizabeth St, Eaton Wing North, 13th Floor, Room 232, Toronto, Ontario, Canada M5G 2C4 (e-mail: email@example.com).