Purpose. The goal of this study was to develop and field test a classification of diagnoses for use by pediatric nurse practitioners (PNPs). A secondary goal was to describe the clinical practice of PNPs in a new way using the classification.

Design. A conceptual framework was derived from theories of nursing. It identified and defined three domains of diagnoses--Developmental Problems, Pediatric Diseases, and Daily Living Problems. Diagnoses from the North American Nursing Diagnosis Association's list, the International Classification of Diseases (ICD-9-CM), and the Diagnostic and Statistical Manual, 3rd revision (DSM-III) were selected. The diagnoses were sorted for content validity into the domains using scores from a panel of expert nurses. Diagnoses assigned to the Daily Living Problems domain were then sorted by the same panel into the eleven Functional Health Patterns described by Gordon (1982). Reliability was measured using Proportions of Agreement and Kappas for all pairs of judges on test and a retest several months later. The validity of the groups created was measured using Indices of Content Validity and Average Congruency Percentages.

Six PNPs working in general primary care settings in Oregon then used the classification to code diagnoses they identified and managed for 240 consecutive cases each. They used open numbers if unable to find suitable diagnoses and evaluated the system for usefulness.

Results. The experts sorted the diagnoses in a new way which decreased overlaps among the domains. The Developmental and Pediatric Disease domains were judged reliable and valid. The Daily Living domain demonstrated marginally acceptable validity by acceptable interrater and intrarater agreement. Six Functional Health Patterns were judged reliable and valid; mixed results were found for the remaining five categories. The PNPs used more than 3000 diagnoses with 1450 cases. Their scope of practice then included 1 percent Developmental problems, 72 percent Pediatric Diseases which included 51 percent illnesses and 21 percent disease prevention, and 27 percent Daily Living problems. The open numbers were used for less than five percent of diagnoses indicating few gaps in the system. The evaluation of the system was rated quite positively by the PNPs.