Computerized Monitoring of Valproate and Physician Responsiveness to Laboratory Studies as a Quality Indicator

Abstract

Using computerized pharmacy, laboratory, and hospitalization data from a large state psychiatric hospital system, this study examined physician responses to laboratory studies obtained in the course of therapeutic drug monitoring. Computerized monitoring modules based on physician-developed guidelines identified out-of-range laboratory values and searched for appropriate corresponding physician responses within clinically driven, mathematically adjusted time frames. Valproate monitoring in four metropolitan hospitals showed that appropriate physician responses were associated with shorter hospital stays for patients and were predictive of length of stay in a multiple regression analysis (p<.001). After physicians received didactic feedback, the percentage of appropriate responses to low serum valproate levels increased.

Valproate in bipolar disorder: 2000 onwards

“…During maintenance treatment patients whose serum levels were maintained in a range of 75–100 μ g/ml had significantly longer time to intervention for a developing mood episode than did patients with serum levels either lower or higher than this range (37). Appropriate physician response in adjusting dosage based on serum valproate levels were significantly associated with reduced time in hospital based on multiple regression analysis ( P = 0.0001; 38). It is desirable to rapidly and safely attain therapeutic blood levels of valproic acid in the treatment of acute mania, because studies have shown that rapid attainment of trough serum concentrations ≥50 mg/l reduces the time to onset of improvement (12).…”

Linking Laboratory and Pharmacy

“…down-loads more than 1 million drug and laboratory records annually from the 19 Illinois state psychiatric hospitals and links them to evaluate a series of quality indicators. 71 "Cleaning" the data to make it usable for such screening has required extensive programming, particularly for pharmacy data. An important insight emerging from this experience is that pharmacy data files are much more complex and unstandardized than laboratory data.…”