PRINCIPALS OF DESIGN AND EVALUATION OF AN INFORMATION-SYSTEM FOR A DEPARTMENT OF RESPIRATORY MEDICINE

Objectives--To evaluate a departmental computer system. Design--a. Dir ect comparison of the time taken to use a manual system with the time taken to use a computer system for lung function evaluation, loan of e quipment and production of correspondence. b. Analysis of the accuracy of data capture before and after the introduction of the computer sys tem. c. Analysis of the comparative running costs of the manual and co mputer systems. Setting--Within a department of respiratory medicine s erving a hospital of 1323 beds. Main Outcome Measures--a. Time taken t o perform functions with the assistance of computerised methods, in co mparison to the manual method used alone. b. Accuracy of data capture. c. Relative running costs. Results--a. The computer system (CS) was s ignificantly faster than the manual system (MS) for lung function eval uation (CS=7.63 min/test, MS=12.25 min/test), loan of equipment (CS=0. 40 min/loan, MS=2.07 min/loan), and checking for overdue equipment (CS =0.49 s/record, MS=9 s/record). The production of correspondence was s lightly slower with the computer (CS=9.30 min/letter, MS=8.54 min/lett er). b. All patient episodes, but only 43 of 65 (66%) of inpatient epi sodes, were captured. Lung function and managerial report data were ac curate using both manual and computerised methods. The manual system f or equipment loans was inefficient, and use of the computer resulted i n the recovery of 221 nebulisers. c. Development costs for 1988-1990 w ere high (pound 72178). Only pound 1200 to pound 1845 per year was rec overed directly from staff time saved by the computer but larger savin gs resulted from changes in work practice (pound 4049-4765). After 10 years the projected deficit is pound 10000 per annum in running costs. Conclusions--In comparison with the manual methods, the computer syst em has shown significant advantages which provide accurate information , with significant favourable effects on working practices. In evaluat ing computer systems used in clinical practice it is essential to ensu re that the projected work practice benefits are achieved without unac ceptable costs in staff time, inaccurate data and high financial outla y. (C) 1997 by John Wiley & Sons, Ltd.