| Popis: |
SHOSHANA HABERMAN, KATERINA PERLMAN, ISRAEL THALER, JOSEPH FELDMAN, Maimonides Medical Center, Obstetrics and Gynecology, Brooklyn, New York, TechnionIsrael Institute of Technology, Obstetrics and Gynecology, Haifa, Israel, Israel, SUNY Downstate Medical Center, Preventive Medicine, Brooklyn, Hawaii OBJECTIVE: With the growing use of decision support systems (DSS) in medicine, it is essential to ensure completeness of documentation, as missing data can greatly jeopardize the accuracy of those systems. This study was aimed at exploring the leading factors that affect documentation, using an Intelligent Patient Record (IPR) system in a busy obstetrical setting. STUDY DESIGN: Completeness of documentation was analyzed using data obtained from structured fields in an IPR for Obstetrics, encompassing 2764 term singleton pregnancies in women who underwent labor between 6/03-12/03. Three clinical items, prompted by the system as essential for documentation were selected: estimated fetal weight, fetal position and pelvic adequacy. A score for completeness of documentation ranging from 0 to 3 was assigned, reflecting the number of items that were missing from the medical record, with lower scores indicating better documentation. The predictor variables were: provider type, time of day, day of week, mode of delivery, patient education, use of oxytocin, parity, NICU admission, oxytocin to delivery time, second stage duration, length of stay in labor room, pattern of labor. Analysis of covariance was used to compare the frequency of missing clinical data. RESULTS: The more complicated cases had less documentation. When comparing between the following groups, the mean scores significantly differed: C/S vs. vaginal delivery: 2.43,1.70 (P ! .05), VBAC vs. none: 1.95,1.83 (P ! .04), NICU admission vs. none:1.96,1.82 (P ! .05). Midwives had fewer missing items (score-1.49) than both attendings (1.94) and residents (1.82), P ! .01.There was no correlation between all other parameters tested and completeness of documentation. Documentation level remained the same for abnormal labor cases in spite of specified prompting. CONCLUSION: Data integrity is of primary importance in DSS, yet the quality of documentation in complicated cases was worse. Results varied significantly S128 SMFM Abstracts |
Full Text from ScienceDirect