KRAS gene mutation analysis is commercially available through several labs. However, the labs themselves have not provided information on analytic performance. 
A recent, industry sponsored study compared different KRAS testing assays from 5 labs (Agencourt, Gentris, Genzyme, HistoGeneX, and Invitek) against the Amgen DNA Sequencing Laboratory direct sequencing assay. KRAS was classified as either wild type or mutant. Techniques were in agreement if both assays identified wild type or a mutant. Agreement was assessed by κ statistics. Agreement for each assay were reported as: HistoGeneX (kappa=0.95), Genzyme (kappa=0.94), Agencourt (kappa=0.94), Gentris (kappa=0.75), and Invitek (kappa=0.13). 
Clinical Validity in Metastatic Colorectal Cancer:
Retrospective, subset analyses of tumor tissue samples from small clinical trials have demonstrated that tumor KRAS gene mutations are associated with lack of response to both of the EGFR-targeted monoclonal antibodies approved for use in colorectal cancer, cetuximab and panitumumab. ,,,,,,,,,,,,,,,,,,,,,,,,,,
The strength of this association has been substantiated in retrospective analyses of patients treated in six, large randomized studies. ,,,,,
A review of 8 studies (306 of 817 patients with tumors mutant for the KRAS gene) conducted by Linardou et al found that KRAS mutations were significantly associated with an absence of response to anti-EGFR monoclonal-antibody-based treatments (sensitivity=0.47 [0.43-0.52]; specificity=0.93 [0.83-0.97]; +LR=6.82; -LR=0.57). 
A recent meta-analysis of 22 studies including persons with metastatic colorectal cancer treated with cetuximab found that progression free and overall survival in persons with wildtype KRAS tumors was better compared to persons with mutated KRAS tumors. 
The data are mixed whether KRAS mutation is indicative of worse prognosis independent of therapy. At present, KRAS mutational analysis is not recommended for risk assessment.
NCCN and BCBS TEC conclude that tumor testing for KRAS mutation offers clinical utility, and testing is available outside of research settings in clinical practice. However, we did not identify studies reporting physician and patient acceptance or population-based health outcomes data from use in clinical practice.