Can a quantitative means be used to predict flow patterns: Agreement between visual inspection vs. flow index derived...

Area of Science:

• Pediatric Urology
• Diagnostic Imaging
• Medical Device Technology

Background:

• Uroflowmetry is crucial for evaluating pediatric lower urinary tract symptoms.
• Current interpretation of uroflowmetry curve shapes (bell, tower, plateau) suffers from significant intra- and interobserver variability.
• Previous research indicated a correlation between the flow index (FI) and uroflow curve shapes in children.

Purpose of the Study:

• To test the hypothesis that FI-defined uroflowmetry curve shapes are as accurate or superior to current subjective grading systems.
• To introduce objectivity and enhance comparability in uroflowmetry assessments by reducing observer bias.

Main Methods:

• A review of consecutive uroflowmetry studies from two centers in children evaluated for lower urinary tract symptoms.
• Comparison of visually assessed curve shapes (bell, plateau, tower) with quantitative FI-derived patterns (FI = Qmax/estimated Qmax).
• Experienced readers at each institution performed curve shape assessments.

Main Results:

• Analysis included 591 males and 1039 females; 409 males and 819 females had curves classified as bell, tower, or plateau.
• Substantial agreement (kappa = 0.71) was observed in males using a 3x3 matrix.
• Moderate agreement (kappa = 0.52) was found in females using receiver operating characteristic cutoffs, with similar results for 1 SD cutoff (kappa = 0.51).

Conclusions:

• The FI method demonstrates substantial to moderate agreement for defining uroflowmetry curve shapes, offering a more quantitative approach.
• While FI-derived shapes show accuracy comparable to or exceeding visual inspection, cutoffs remain somewhat arbitrary.
• This quantitative system has the potential to significantly decrease ambiguity in uroflowmetry interpretation, particularly in transitional curve zones.