Real world evaluation of kidney failure risk equations in predicting progression from chronic kidney disease to kidney failure in an Australian cohort.

BACKGROUND: Chronic kidney disease progression to kidney failure is diverse, and progression may be different according to genetic aspects and settings of care. We aimed to describe kidney failure risk equation prognostic accuracy in an Australian population. METHODS: A retrospective cohort study was undertaken in a public hospital community-based chronic kidney disease service in Brisbane, Australia, which included a cohort of 406 adult patients with chronic kidney disease Stages 3-4 followed up over 5 years (1/1/13-1/1/18). Risk of progression to kidney failure at baseline using Kidney Failure Risk Equation models with three (eGFR/age/sex), four (add urinary-ACR) and eight variables (add serum-albumin/phosphate/bicarbonate/calcium) at 5 and 2 years were compared to actual patient outcomes. RESULTS: Of 406 patients followed up over 5 years, 71 (17.5%) developed kidney failure, while 112 died before reaching kidney failure. The overall mean difference between observed and predicted risk was 0.51% (p = 0.659), 0.93% (p = 0.602), and - 0.03% (p = 0.967) for the three-, four- and eight-variable models, respectively. There was small improvement in the receiver operating characteristic-area under the curve from three-variable to four-variable models: 0.888 (95%CI = 0.819-0.957) versus 0.916 (95%CI = 0.847-0.985). The eight-variable model showed marginal receiver operating characteristic-area under the curve improvement: 0.916 (95%CI = 0.847-0.985) versus 0.922 (95%CI = 0.853-0.991). The results were similar in predicting 2 year risk of kidney failure. CONCLUSIONS: The kidney failure risk equation accurately predicted progression to kidney failure in an Australian chronic kidney disease population. Younger age, male sex, lower estimated glomerular filtration rate, higher albuminuria, diabetes mellitus, tobacco smoking and non-Caucasian ethnicity were associated with increased risk of kidney failure. Cause-specific cumulative incidence function for progression to kidney failure or death, stratified by chronic kidney disease stage, demonstrated differences within different chronic kidney disease stages, highlighting the interaction between comorbidity and outcome.

Making the economic case for prevention--a view from Wales.

BACKGROUND: It is widely acknowledged that adverse lifestyle behaviours in the population now will place an unsustainable burden on health service resources in the future. It has been estimated that the combined cost to the NHS in Wales of overweight and obesity, alcohol and tobacco is in excess of £540 million.In the current climate of financial austerity, there can be a tendency for the case for prevention efforts to be judged on the basis of their scope for cost savings. This paper was prompted by discussion in Wales about the evidence for the cost savings from prevention and early intervention and a resulting concern that these programmes were thus being evaluated in policy terms using an incorrect metric. Following a review of the literature, this paper contributes to the discussion of the potential role that economics can play in informing decisions in this area. DISCUSSION: This paper argues that whilst studies of the economic burden of diseases provide information about the magnitude of the problem faced, they should not be used as a means of priority setting. Similarly, studies discussing the likelihood of savings as a result of prevention programmes may be distorting the arguments for public health.Prevention spend needs to be considered purposefully, resulting in a strategic commitment to spending. The role of economics in this process is to provide evidence demonstrating that information and support can be provided cost effectively to individuals to change their lifestyles thus avoiding lifestyle related morbidity and mortality. There is growing evidence that prevention programmes represent value for money using the currently accepted techniques and decision making metrics such as those advocated by NICE. SUMMARY: The issue here is not one of arguing that the economic evaluation of prevention and early intervention should be treated differently, although in some instances that may be appropriate, rather it is about making the case for these interventions to be treated and evaluated to the same standard. The difficulty arises when a higher standard of cost saving may be expected from prevention and public health programmes.The paper concludes that it is of vital importance that during times of budget constraints, as currently faced, the public health budgets are not eroded to fund secondary care budget shortfalls, which are more easily identifiable. To do so would diminish any possibility of reducing the future burden faced by the NHS of lifestyle-related illnesses.

Free breakfasts in schools: design and conduct of a cluster randomised controlled trial of the Primary School Free Breakfast Initiative in Wales [ISRCTN18336527].

BACKGROUND: School-based breakfast provision is increasingly being seen as a means of improving educational performance and dietary behaviour amongst children. Furthermore, recognition is growing that breakfast provision offers potential as a means of addressing social inequalities in these outcomes. At present however, the evidence base on the effectiveness of breakfast provision in bringing about these improvements is limited. METHODS/DESIGN: This paper describes the research design of a large scale evaluation of the effectiveness of the Welsh Assembly Government's Primary School Free Breakfast Initiative. A cluster randomised trial, with school as the unit of randomisation was used for the outcome evaluation, with a nested qualitative process evaluation. Quantitative outcome measures included dietary habits, attitudes, cognitive function, classroom behaviour, and school attendance. The study recruited 111 primary schools in Wales, of which 56 were randomly assigned to control condition and 55 to intervention. Participants were Year 5 and 6 students (aged 9-11 years) in these schools. Data were collected for all 111 schools at each of three time points: baseline, 4 month and 12 month follow-up. This was achieved through a repeated cross-sectional survey of approximately 4350 students on each of these occasions. Of those students in Year 5 at baseline, 1975 provided data at one or both of the follow-ups, forming a nested cohort. The evaluation also included a nested process evaluation, using questionnaires, semi-structured interviews and case studies with students, school staff, and local authority scheme coordinators as key informants. DISCUSSION: An overview of the methods used for the evaluation is presented, providing an example of the feasibility of conducting robust evaluations of policy initiatives using a randomised trial design with nested process evaluation. Details are provided of response rates and the flow of participants. Reflection is offered on methodological issues encountered at various stages through the course of the study, focusing upon issues associated with conducting a randomised trial of a government policy initiative, and with conducting research in school settings.

Cluster-randomised trial of risk communication to enhance informed uptake of cervical screening.

BACKGROUND: Women overestimate both population and individual risk of cervical cancer. This may contribute to the recognised excess screening frequency for low-risk women. AIM: To investigate whether an individualized risk communication package could affect stated preferences for screening interval and actual screening behaviour. DESIGN: Pragmatic, practice-based cluster randomised controlled trial. SETTING: Twenty-nine practices (15 intervention, 14 control) in North Wales recruited 1890 women attending for cervical smears. METHOD: A risk communication package containing visual material was compared with normal practice. Practice nurses received training in its delivery. The short-term primary outcome was stated preference for screening interval; the long-term primary outcome was actual screening behaviour. RESULTS: In the short term, intervention arm women were significantly less likely to prefer a shorter than recommended interval (odds ratio [OR] = 0.51, 95% confidence interval [CI] = 0.41 to 0.64; P < 0.0001). At the five-year follow-up, fewer women in the intervention arm had attended for screening sooner than their recommended recall. The magnitude of difference in excess screening interval preference and behaviour was similar, but behaviour had a wider confidence interval and a marginally non-significant P-value (OR = 0.61, 95% CI = 0.36 to 1.03; P = 0.063). Better knowledge and more accurate risk perceptions were demonstrated, with an improvement in measures of anxiety. The extra cost per woman receiving the intervention was 6 Pounds. CONCLUSIONS: Women's perception of risk contributes to determining screening intervals in addition to practice factors. Simple risk information delivered in primary care affected women's stated preferences for tests. The impact on actual screening behaviour was more equivocal. Overall, the intervention showed a substantial benefit and any disbenefit can be ruled out. This approach to providing risk information could, at low cost, benefit other screening programmes and may relieve anxiety.