Premium
Measured versus Estimated Creatinine Clearance in a High‐Functioning Elderly Sample: MacArthur Foundation Study of Successful Aging
Author(s) -
Malmrose Lynda C.,
Gray Shelly L.,
Pieper Carl F.,
Blazer Dan G.,
Rowe John W.,
Seeman Teresa E.,
Albert Marilyn S.
Publication year - 1993
Publication title -
journal of the american geriatrics society
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.992
H-Index - 232
eISSN - 1532-5415
pISSN - 0002-8614
DOI - 10.1111/j.1532-5415.1993.tb07459.x
Subject(s) - creatinine , medicine , renal function , estimating equations , urine , linear regression , stepwise regression , regression analysis , urology , gerontology , demography , statistics , mathematics , maximum likelihood , sociology
Objective: To assess the validity of several equations for estimating creatinine clearance in a large sample of high‐functioning, community‐dwelling elderly. Design: Serum and 12‐hour urine samples were collected and assayed for creatinine using the Jaffe total chromagen method. Fifteen clearance‐estimating equations were evaluated for bias, accuracy, correlation with measured clearance values, and frequency of erroneous placement into renal function categories. Stepwise regression modeling and reliability testing were performed on a split sample to construct and assess a novel creatinine‐clearance‐estimating equation. Setting: New Haven, Connecticut, East Boston, Massachusetts, and a five‐county region in and around Durham, North Carolina. Participants: A subsample of community‐dwelling men and women (age range 70–79 years) from the Established Populations for Epidemiological Studies of the Elderly was screened for physical and cognitive functioning and placed into high‐, medium‐, and low‐functioning groups ( n = 1354). High‐functioning respondents who provided blood and complete urine samples ( n = 762) were included in the present study. Results: In general, estimated creatinine clearance was more closely correlated to measured values in males than in females. Most equations underestimated creatinine clearance, with average bias ranging from—33.1 mL/min to +19.6 mL/min. Predictive accuracy ranged from 18.2 mL/min to 38.0 mL/min. Equations were variable in their erroneous placement of individuals into renal function categories. Regression modeling yielded an equation which contained novel components but failed to provide better estimates of creatinine clearance than those already available. Conclusions: The equations evaluated here provide unacceptable predictions of creatinine clearance in normally aging individuals. We advocate the use of serum drug concentration measurements when available and encourage investigation into timed urine collections of short duration as alternatives to clearance‐estimating equations in the elderly.