Reference intervals developed from data for hospitalized patients: computerized method based on combination of...

Area of Science:

• Clinical Biochemistry
• Laboratory Medicine
• Hematology

Background:

• Establishing accurate health-related reference intervals is crucial for clinical diagnostics.
• Conventional methods often rely on healthy volunteer cohorts, which may not fully represent diverse patient populations.
• Hospital information systems offer vast datasets for alternative reference interval determination.

Purpose of the Study:

• To determine health-related reference intervals for hemoglobin concentration, mean corpuscular volume (MCV), and erythrocyte count using hospitalized patient data.
• To compare reference intervals derived from hospitalized patients with those obtained conventionally.
• To assess the applicability of a diagnosis-based selection method for establishing laboratory reference values.

Main Methods:

• Utilized hospital information system databases to select patients based on discharge diagnoses.
• Calculated nonparametric 2.5-97.5% reference intervals for key hematologic parameters.
• Excluded patients with diseases potentially affecting erythrocyte variables, analyzing 1786 women and 1450 men (ages 20-65).

Main Results:

• Health-related reference intervals were established for hemoglobin, MCV, and erythrocyte count in adult men and women.
• The upper reference limits for MCV in hospitalized patients were found to be higher than conventionally reported values.
• Age-specific reference intervals were determined for four age groups.

Conclusions:

• A diagnosis-based patient selection method can yield clinically relevant reference intervals.
• This approach may provide more accurate upper reference limits for certain analytes, like MCV.
• The method is potentially applicable to various laboratory analytes using existing hospital databases.