Does reporting estimated glomerular filtration rate affect ordering of timed urine collections?

Background: The National Kidney Foundation recommends reporting estimates of glomerular filtration rate (eGFR) rather than timed urine collections. When 2 of the 3 major hospitals in our region began reporting eGFR at different times we recognized a natural experiment.

Methods: We conducted a retrospective, observational study at the 3 major hospitals in Chattanooga, Tennessee. Data were collected on the frequency of timed urines during a 41/2-year period. Regression analysis was used to study the association of the rate of ordering timed urines adjusted for other factors.

Results: There was a marked drop in the rate of ordering of timed urines at all 3 hospitals from a mean of 21.8 per 1000 admissions in the first 4 quarter years of the study to 10.9 in the last 4 (15th-18th) quarter years. The drop began before the reporting of eGFR at any hospital. The reporting of eGFR had a small effect (-1.7 per 1000 hospital admissions) that was not statistically significant (P = 0.15).

Conclusions: There has been a marked drop in the ordering of timed urines in our region. The decline began before the reporting of eGFR at 2 of the hospitals and therefore is attributable to other factors.