Validation of a problem-focused nomenclature.

To test the value of a problem-focused nomenclature, problems were compared with DSM-II diagnoses as predictors of treatment and outcome. Patient problems, diagnoses, and treatment plans were coded from the medical records of 779 psychiatric inpatients. Problems were found to predict treatment plans better than did diagnoses. The combination of problems and diagnoses, however, was the best predictor of treatment with psychotropic medications. Problems of over 300 patients were compared with diagnoses as predictors of patient outcomes in overall functioning, subjective well-being, family functioning and mental status. At two months, problems and diagnoses predicted outcome equally well while the two in combination afforded the best prediction. At one year, however, the variance in outcomes explained by diagnosis was one third less than that explained by problems.