Visual Interpretation of Neonatal Growth Charts

BACKGROUND Optimization of growth is essential for preterm infants. Optimum post-natal growth has been shown to improve long-term developmental outcomes. Close monitoring of weight gain, including early identification of growth faltering, is an important part of appropriate nutritional management in the Neonatal Intensive Care Unit.

Current practices in neonatal intensive care units involve the routine use of growth charts to track and identify patterns in the growth of preterm infants. Perhaps the most popular growth chart in use today is the Fenton Growth Chart. To interpret this chart, a patient's growth curve is plotted over a set of standard curves (typically representing the 10th, 25th, 50th, 75th and 90th weight-for-age centile). The curves are then visually compared to identify patterns in growth (adequate, slow, or catch-up growth). Patterns in neonatal growth may be subtle and current growth curves may be prone to misinterpretation. This may be due to biases inherent in the human visual perception system, including the difficulty in visually estimating slopes and angles between two curves.

An improved growth chart would use horizontal lines rather than curves, and would have no oblique lines to skew visual perception. Z-score growth charts accomplish this. Z-score growth charts are created using the same standardized data used to create the Fenton Growth Chart. Z-scores represent standard deviation scores. Rather than using a set of standard curves to represent average healthy growth, the z-score chart uses a single horizontal line. This allows for easier comparison when plotting a patient's weights over time. Such charts may lead to more accurate interpretation of overall growth patterns.

This study will compares the accuracy of interpretation of the growth data using the same standard - once as presented using traditional weight-vs.-age charts, and again using the same data displayed as Z-score vs age charts.

INCLUSION AND EXCLUSION CRITERIA

It will include UC Davis Medical Center healthcare personnel who interpret growth charts for preterm infants. This may include the following:

• Attending Neonatologists
• Attending Pediatricians
• Dietitians
• Medical Students
• Neonatal Fellows
• Neonatal Nurse Practitioners
• Pediatric Residents This study will exclude persons who do not routinely interpret growth charts for preterm infants as part of their clinical responsibilities.

NUMBER OF SUBJECTS This will be a single center study. The sample size needed for 80% power with an alpha of 0.05 is estimated to be at least 40. Approaching 45-50 providers should be sufficient to achieve this number.

RECRUITMENT METHODS Healthcare providers who meet the inclusion criteria will be approached on an individual and/or group basis. We may attempt to hold a meeting (perhaps at lunch time) when volunteers may participate in the study.

COMPENSATION TO SUBJECTS If there is a meeting held at lunchtime for volunteers to participate in this study, there may be a small lunch provided (such as sandwiches, pizza and/or soft drinks). When subjects are approached on an individual basis, there will be no compensation offered.

STUDY TIMELINES

• The duration of an individual subject's participation in the study: less than one hour.
• The duration anticipated to enroll all study subjects: three to six months.
• The estimated date for the investigators to complete this study (complete primary analyses): Less than six months

STUDY ENDPOINTS Primary endpoint: after 40-50 persons have completed the study. Primary or secondary safety endpoints: there are no foreseeable safety issues with participating in this questionnaire based study.

PROCEDURES INVOLVED

Study design:

• Study subjects will be asked to view a brief presentation. This will describe Fenton Preterm and Z-score growth charts, and review how they are interpreted.
• Hypothetical (i.e. dummy data) for 10-20 infants will be displayed both as weigh vs. age charts and as Z-score vs age charts. The data will consist of 2 points approximately 1-2w apart. Subjects will be asked whether the growth between the two points is "normal" (paralleling centile lines on the weight vs age charts, or horizontal on the Z-score vs age charts), "slower than normal" or "faster than normal"). The order of presentation of the charts will be randomized
• Subjects will record their interpretation of the growth data on a sheet (marking a box for "normal", "faster than normal" or "slower than normal".
• At the end of the presentation subjects will be asked to record their level of training (medical student, resident, fellow, NNP, neonatal attending, or other) on the sheet.
• A space will be provided on the datasheet for subjects to record any impression of the charts they may wish to share.
• Subjects will be provided unmarked envelopes to put their datasheets in to return them to the investigators (so investigators will be unaware of the answers, or whether the subjects completed the datasheet)