Validity of clinical information from a maternity survey

This research was designed to validate data collected through a survey--an inexpensive way to provide information for quality measurement. The survey was sent to health maintenance organization (HMO) enrollees who had given birth(s) between October 1, 1994, and May 31, 1995. The responses were compared with the medical records. A sample of 407 women was randomly selected from the completed surveys. Medical records were reviewed for 89.9% (362/407) of the sample based on medical record availability. Over 98% of responses agreed with the medical record information regarding whether there were cesarean sections for previous deliveries (kappa = 1.0), cesarean section for recent delivery (kappa = 0.95), and vaginal birth after cesarean section (kappa = 0.96). Over 99% of the mothers agreed with the information regarding whether the newborn birth weight was under 2500 g (kappa = 0.91). The findings strongly support the validation of this instrument. Using this validated instrument enables health plans to cost-effectively obtain crucial information.