Study objective: To assess the accuracy of five creatinine clearance equations in predicting measured creatinine clearance in hospitalized patients with human immunodeficiency viral (HIV) infection.
Design: Prospective evaluation over a 6-month period.
Setting: Erie County Medical Center, a 550-bed teaching institution.
Patients: Forty-seven HIV-positive patients (39 men, 8 women) who were admitted for a variety of HIV-related illnesses and judged clinically to have stable renal function. Of the 47 original patients, 44 were evaluable based on exclusion criteria.
Interventions: Serum creatinine and 24-hour measured creatinine clearance were performed in each patient.
Measurements and main results: The estimated creatinine clearance from each of the equations (Cockcroft-Gault, two Jeliffe equations, Mawer et al, and Hull et al) was compared with the measured creatinine clearance. Statistical analysis of these comparisons was performed and all of the equations were found to overestimate the measured creatinine clearance (mean error 34-45%).
Conclusions: Many HIV-infected patients have a decreased creatinine clearance despite a serum creatinine concentration within the normal range. Each of the equations overestimated the measured creatinine clearance.