Variations in compliance with documentation using computerized obstetric records

Abstract

Objective: To explore factors that affect documentation completeness using an electronic medical record with a decision support system for obstetrics.

Methods: Two thousand eight hundred nine consecutive deliveries were analyzed and data were obtained from structured fields in the decision support system. The decision support system was customized to deactivate the system's repetitive prompts and reminders for documentation completeness for the chosen study parameters. Completion of documentation for estimated fetal weight, pelvic adequacy, and fetal position were selected as outcome variables. One point was given for each missing item. Data were analyzed using general linear univariable models. Tukey's honest difference method was used to adjust the P values for potential multiple comparison biases.

Results: Midwives had fewer missing items (score 1.42) than both attendings (1.87) and residents (1.74), P<.01. When comparing the following groups, the mean scores differed significantly: vaginal birth after cesarean and repeat cesarean delivery, 1.95 and 1.83 (P<.04); neonatal intensive care unit admission and regular nursery, 1.96 and 1.82 (P<.05). Patients experiencing normal and abnormal labors were similar in documentation completeness, but patients who lacked enough data to have their labors classified were significantly less likely to have complete documentation for the chosen outcome variables.

Conclusion: Compliance with documentation in electronic medical record is very low when the reminders for documentation completeness are deactivated and varies with type of provider, as well as with some clinical aspects of the patient.