Premium
ESTIMATING EFFECTS OF SYSTEMATIC TREATMENT ON RENAL FAILURE AND DEATH WITHOUT A PARALLEL PLACEBO CONTROL GROUP
Author(s) -
We Hoy,
P Baker,
Z Wang,
A Cass,
Jd Mathews,
P. Van Buynder
Publication year - 2000
Publication title -
nephrology
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.752
H-Index - 61
eISSN - 1440-1797
pISSN - 1320-5358
DOI - 10.1046/j.1440-1797.2000.abs122.x
Subject(s) - medicine , hazard ratio , albuminuria , disease , dialysis , psychological intervention , kidney disease , intensive care medicine , placebo , clinical endpoint , randomized controlled trial , confidence interval , alternative medicine , pathology , psychiatry
Background: Chronic disease programs are poorly developed in most Aboriginal communities. Much disease is unrecognised or inadequately treated, although appropriate interventions profoundly reduce morbidity and mortality in nonAboriginal populations. Programs of improved management must aspire to best practice for all, so that maintaining parallel untreated control groups is unethical. This poses challenges for evaluating effect. Methods: We identified a large burden of chronic disease in a 1990‐1995 screening program in one community, and started a renal & cardiovascular‐protection program in Nov 1995. This centred around use of ACE inhibitors, rigorous BP control, better control of glycemia and lipids, & health education. By late 1999 about 275 people, or 30% of all adults had enrolled. The courses of BP, albuminuria and GFR was compared with those in the pre‐program era (ANZSN, 1999). Treatment effects on renal failure & natural death were estimated in 3 ways. 1) Comparison of these endpoints in the “intention to treat” group with those in persons potentially eligible for treatment on their 1990‐1995 screening results, ‘controls’. There was 50% overlap between the groups, & controls were younger and had less severe disease than the treatment group. 2.Community‐based trends in endpoints. 3. Comparison of these trends with those in other NT Top End communities. Results: 1. Risk ratios of rates, Kaplan Meier survivals, and Cox hazard ratios all showed better survival of the treated group over controls, with estimates of 41%‐64% reductions in endpoints, after accounting for disease severity. 2. Dialysis starts in the entire community have fallen by at least 38% and natural deaths by 32%. 3. In contrast dialysis continue to increase at 11% per yr in other communities and deaths have not fallen. These results all suggest a marked benefit from the treatment program. Similar methods might be used where truly controlled observations are not feasible.