Does WHO clinical stage reliably predict who should receive ARV treatment?

Polly Clayden, HIV i-Base

In resource-limited settings with limited access to laboratory facilities it has been suggested that WHO eligibility criteria can be used to guide treatment initiation decisions in the absence of CD4 counts.

A South African study conducted by Neil Martinson and co-workers in and presented by Glenda Gray assessed the proportion of adults staged as WHO I or II but with CD4 counts <200 cells/mm3 [1]. The study also looked at predictors for being assessed as stage I or II, but having a CD4 count below 200 cells/mm3.

Current WHO criteria recommend initiating therapy in stages III and IV and in stage II if total lymphocyte count is <1200 [2].

The study was conducted at two primary healthcare facilities, an urban site in Soweto and a rural site in Limpopo Province. It was a cross sectional, operational study with a total of 2000 patients, 1500 from the urban site and 500 from the rural. The median CD4 count was 246 cells/mm3 (range 126-426 cells/mm3) across both groups. The median CD4 at the urban site was 233 cells/mm3 (range 111-411 cells/mm3) and 203 cells/mm3 at the rural site (range 91-375 cells/mm3). See Table 1.

Table 1: CD4 cells/mm3 and interquartile range by stage

WHO stage Urban (n) IQR Rural (n) IQR
I 339 (302) 199-525 413 (78) 267-615
II 194 (123) 123-370.5 270 (128) 143-420.5
III 154 (241) 67-279 147 (285) 56-279
IV 176 (8) 30-450 105 (33) 51-327

The investigators reported that 23.9% (urban 25.0% and rural 14.5%) of patients classified stage I had CD4 counts <200 cells/mm3. For stage II the total was 46.1% (urban 50.4% and rural 37.9%).

Dr Gray noted that wrong staging of patients was slightly more common in the urban site than the rural site but this was not found to be a predictor in a multivariate analysis. In both univariate and multivariate analyses male gender was a predictor for mistaging as stage I or II with CD4 count <200 cells/mm3 (gender female vs male: univariate and multivariate 0.6 [0.4-0.8]).

The investigators also looked at patients staged as III or IV with CD4 <350 cells/mm3. See Table 2.

Table 2: CD4 strata WHO stages III and IV

Stage <200 200-350 >350
III 62% 20% 18%
IV 59% 15% 26%

The investigators summarised: “38-50% of adults staged as II and 15-25% staged as I had CD4 counts under 200 cells/mm3. Approximately 80% of adults staged in WHO III and IV had CD4 <350 cells/mm3. Experience appears to reduce ‘mistaging’. Men are more likely to be staged as I or II but have CD4<200 cells/mm3


WHO staging is not a good instrument for primary care particularly to recognise advanced disease.

CD4 counts in all settings will ensure relatively well people with low CD4s won’t miss out on ARVs until they are really ill. The WHO staging needs to be reassessed.


  1. Martinson N, Heyer A, Steyn J et al. Does WHO clinical stage reliably predict who should receive ARV treatment? 3rd IAS Conference on HIV Pathogenesis and Treatment, Rio de Janeiro, 2005. Abstract WeFo0304.
  2. WHO guidelines. Download pdf file: (492 k)

Links to other websites are current at date of posting but not maintained.