The Effect of Gender, Age, and Geographical Location on the Incidence and Prevalence of Renal Replacement Therapy in Wales

Hugo C van Woerden; * Jane Wilkinson; Martin Heaven; Jason Merrifield


BMC Nephrology 

In This Article


This study demonstrates significant variation in the provision of RRT by gender, age, and geographical location. Differences in RRT rates were most marked by gender and age, and to a lesser extent by geographical location. Some of the gender difference in prevalence may be due to risk factors shared by cardiovascular disease and end stage renal failure, both of which are commoner in men. However, some of the gender differences in the very elderly, particularly the differences in haemodialysis, may be indicative of underlying inequality in treatment provision.

Implications for service provision can be drawn from the findings. Age-specific prevalence for peritoneal dialysis was relatively similar at all ages, and consequently, peritoneal dialysis services need to be designed to serve all age groups in the population. In contrast, support services for patients with renal transplants needs to be focused on those under 65 years. For example, attendance at outpatient clinics may significantly interfere with working age patients who have a transplant.

Patients in Wales under the age of 50 years have a greater than 50% chance of having a transplant (Figure 3), which is encouraging for younger patients waiting for a transplant.

The probability of being on peritoneal or haemodialysis varies depending on the LHB in Wales. This may represent the preferences of different clinicians for particular treatment modalities, and a wide range of socio-economic determinants, as well as random chance.

There are a number of possible reasons why the tertiary English NHS Trusts, which treat Welsh transplant patients, might have a higher number of patients recorded as being alive, when the NHS Administrative Register suggested that they were deceased. Firstly, there may be less accurate recording of death in this Trust. However, tertiary care centres may see patients less frequently, if joint care is undertaken with a locally based clinician. Consequently, if a patient dies in a local hospital, the tertiary centre may not be aware of the patient's demise. Our findings suggest that there is merit in linking databases held by different Trusts, in forming links between Trust databases, national renal databases and the NHS AR using patients NHS numbers. This would reduce inaccuracies around deceased patients, patients who have recently moved address, and patients attending multiple renal centres.

This study suggests that there may be a significant number of untreated elderly patients with end stage renal failure. Figure 1 indicates that the rate of RRT provision rises with age but falls after the mid-seventies. This is unexpected given the rising trend in morbidity up to this age. The true underlying rate of end stage renal failure is very unlikely to fall over the age of 75 years and this suggests that there may be a gap between need and supply in this very elderly age group. The discrepancy may be occurring because clinicians and patients are taking into account the personal, physical, social and psychological costs of treatment, and the very poor survival curves associated with current treatment over this age, and decide not to offer or accept RRT.

We can provide a crude estimate of the number of potentially untreated patients over the age of 75 years by linear extrapolation of the population rate for RRT from 1790 pmp at 75 years to 2500 pmp for the age group 100 years. Similarly, we can extrapolate the haemodialysis rate from 1000 pmp at 85 years to 1500 pmp at 100 years. Extrapolating in this way would suggest that an additional 442 individuals over the age of 75 years might wish to be treated, if RRT methods were developed that provided excellent rates of survival and low negative impact on patients' quality of life. This unmet "need" would represent a 17% increase on current demand for RRT.

These figures provide a crude estimate of the possible increase in future demand in the very elderly, without taking into account other factors such as demographic shift in the population, or increased provision of RRT to those under 75 years. In summary, these findings suggest that, as RRT technologies continue to improve, an increasing number of very elderly patients are likely to present for treatment.

Patients over the age of 75 years have a very high probability of being on haemodialysis. Population predictions for the next decade for Wales suggest that this population group will expand the most. Unless the current treatment pattern changes significantly, the high use of haemodialysis in old age (Figures 4 & 5) combined with the demographic bulge predicted in this age group suggests that the demand for haemodialysis in the very elderly will continue to rise at a higher rate than in younger age groups.

Limitations of This Study

Our study has a number of limitations. We have used a different method of calculating incidence from that used by the renal registry, which includes patients who started RRT in the year before the census date, but who also died in that year. Our figures would, therefore, be expected to indicate a lower rate than if these patients were included. Our incidence figures would have been improved if our data collection had included patients who died in the preceding year. Our data is cross-sectional: longitudinal data gives better estimates of incidence, although cross sectional studies such as ours, which ask for changes over the proceeding year, can be used.

A renal physician from every renal centre in Wales, and NHS Trusts in England that treat Welsh patients, was used as a point of contact. Data managers in each trust were also involved. As each trust keeps a database of patients receiving treatment for ESRF we were confident that all individuals receiving treatment had been identified. However, not all patients with ESRF present for treatment and not all patients with ESRF are offered or receive treatment. Prevalence estimates of ESRF based on treatment databases are consequently underestimates of the true prevalence.

Based on our discussion with renal physicians in Wales we are confident that we have identified all patients receiving treatment by a renal physician. However, very low acceptance rates reported in some areas (under 50 pmp) could theoretically be the result of local treatment and failure to refer by non-renal physicians.

Data on ethnicity was not collected in this study, as there are relatively low rates of non-white populations in most of Wales. However, this would have provided some useful additional information as end stage renal failure is affected by ethnic origin.

The peak rate for a given modality, shown in the graphs using a moving average, is around two to four years younger than the true value. This is because the moving average brings forward the peak in the graphs by between two to four years when calculated using a 'window' moving from left to right.

Co-morbidity was not assessed in this study; however, it is an important factor in the elderly. There is some evidence that the presence and severity of co-morbid conditions is more important than age as a predictor of survival after commencement of dialysis[14,15]. Further information on the range, variation, severity and relative influence of co-morbidity on survival would have been informative.

Conservative options have also not been addressed in this study, but are important to a proportion of patients with end stage renal failure. There is significant potential for improved prevention and earlier conservative treatment to slow deterioration in renal function and affect the number of patients requiring RRT.[16] A decision not to offer treatment may also be appropriate[15] although age alone may sometimes inappropriately be used as a criterion for withholding treatment. In one study, the decision to withhold dialysis increased at a rate of 12% for every 10 years of increasing age even after adjustment for conditions such as dementia and/or dependency.[17]

Comparison with Other Studies

Most of the significant epidemiological work on renal disease in the UK has been undertaken by the UK Renal Registry[4] and its Scottish counterpart.[18] The annual report of the UK Renal Registry in particular contains information submitted by all the renal units in Wales and has expanded year on year. The negative influence of geographical distance from a renal unit has also previously been demonstrated in Wales.[19] In Scotland, the effect of social class,[20] co-morbidity[21] and referral patterns of GPs and non-nephrologists[22] have been studied in detail. Co-morbidity and non-referral are often linked to age[21,23] but the epidemiological consequences for very elderly patients has rarely been examined.[24]


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.