Renal impairment in a rural African antiretroviral programme

There is little knowledge regarding the prevalence and nature of renal impairment in African populations initiating antiretroviral treatment, nor evidence to inform the most cost effective methods of screening for renal impairment. With the increasing availability of the potentially nephrotixic drug, tenofovir, such information is important for the planning of antiretroviral programmes

(i) Retrospective review of the prevalence and risk factors for impaired renal function in 2189 individuals initiating antiretroviral treatment in a rural African setting between 2004 and 2007 (ii) A prospective study of 149 consecutive patients initiating antiretrovirals to assess the utility of urine analysis for the detection of impaired renal function. Severe renal and moderately impaired renal function were defined as an estimated GFR of < or = 30 mls/min/1.73 m(2) and 30-60 mls/min/1.73 m(2) respectively. Logistic regression was used to determine odds ratio (OR) of significantly impaired renal function (combining severe and moderate impairment). Co-variates for analysis were age, sex and CD4 count at initiation.

(i) There was a low prevalence of severe renal impairment (29/2189, 1.3% 95% C.I. 0.8-1.8) whereas moderate renal impairment was more frequent (287/2189, 13.1% 95% C.I. 11.6-14.5) with many patients having advanced immunosuppression at treatment initiation (median CD4 120 cells/microl). In multivariable logistic regression age over 40 (aOR 4.65, 95% C.I. 3.54-6.1), male gender (aOR 1.89, 95% C.I. 1.39-2.56) and CD4<100 cells/ul (aOR 1.4, 95% C.I. 1.07-1.82) were associated with risk of significant renal impairment (ii) In 149 consecutive patients, urine analysis had poor sensitivity and specificity for detecting impaired renal function.

In this rural African setting, significant renal impairment is uncommon in patients initiating antiretrovirals. Urine analysis alone may be inadequate for identification of those with impaired renal function where resources for biochemistry are limited.