DISCREPANCIES BETWEEN SURVEY-BASED AND ACTUAL DECISIONS IN DISCHARGING AND FOLLOWING UP WITH NEWBORNS Stacia A. Finch, Richard C. Wasserman, Cathie Spino, Eric J. Slora, Heidi C. Schwartz, Hank H. Bernstein. PROS, Ctr for Child Health Research, Amer Acad of Peds, Elk Grove Village, IL; PROS, Univ of Vermont, Burlington, VT; Ann Arbor, MI; Div of General Peds, Children?s Hosp, Boston, MA.

BACKGROUND: Surveys that use global and scenario-based questions may not accurately capture clinicians? real-world decision-making.

OBJECTIVE: To determine the degree to which office-based clinicians discharged and followed up newborns as their pre-study survey responses would have predicted.

DESIGN/METHODS: Prior to beginning Life Around Newborn Discharge (LAND) study enrollment, participating Pediatric Research in Office Settings (PROS) clinicians completed a survey (PS), addressing: 1) beliefs on optimal length of hospital stay (LOS) for healthy term newborns; and 2) scenario-based question about routinely scheduled contacts during the first four weeks after discharge for breastfed (BF) and non-breastfed (N/BF) newborns. A total of 1,514 mother/infant pairs were enrolled in LAND by 192 clinicians. Of those, 654 BF and 207 N/BF newborns matched those described in the PS (ie, non-jaundiced, feeding well). Discharge practices were studied with respect to chosen LOS timeframes (eg, <12, 12-24 hours) and scheduled follow-up plans for home visit, office visit, and timing of first office visit. Agreement on discharge practices was scored when a clinician indicated a LOS on the PS that matched that of comparable enrolled newborn(s). Follow-up agreement was scored when a clinician?s report on the PS that a routinely scheduled contact would or would not be made matched what was done for comparable enrolled newborn(s).

RESULTS: The following agreement percentages were found between survey responses and actual practices:

Clinicians discharged 1/4 of enrolled newborns earlier than they indicated on PS as optimal. Maternal responsibilities/preference was cited most often (71%) as the reason for shorter LOS, with insurance mentioned in only 9% of responses. Of the 46% discharged later than optimal, infant (51%) and maternal (43%) health status were the most common reasons.

CONCLUSIONS: The PS predicted the timing of discharge poorly, but did better at predicting the timing and need for follow-up visits in BF and N/BF infants. Of note, maternal responsibilities and preferences, rather than insurance pressure, appear to account for most cases of shorter than optimal LOS.