Presented at the Neonatal Society 2010 Spring Meeting.
El-Kafrawy U, Settle P (introduced by Dr MJ Robinson)
Neonatal Unit, Salford Royal Hospital, Stott Lane, Salford M6 8HD, UK
Background: Obtaining accurate weight measurement is an important part of the care of the extremely preterm infant, but can be problematic. Incubator scales are inaccurate if infants are connected to ventilator tubing, while disconnection from the ventilator to obtain weights on free standing scales risks destabilising the sickest infants. We developed an adjustable device, to use in conjunction with incubator scales, which we hoped would eliminate this problem.
Aim: To compare the accuracy, when weighing ventilated infants, of incubator scales alone and in combination with this novel device.
Methods: Infants were weighed while connected to the ventilator (3100A High Frequency Oscillator or VIP Bird Gold) using the incubator scales alone then in combination with the adjustable device. Infants were then disconnected from the ventilator for weighing on free standing electronic scales (this being taken as the true weight). Incubator weights and free standing weights were obtained by different members of staff who were blinded to the other measurement.
A feasibility study using manikins gave encouraging and statistically significant results and allowed us to fine tune the device to give results within 1-2% of the true weight.
We subsequently used the above methodology to obtain 35 sets of weights on 19 pressure ventilated infants (on VIP Gold ventilators), on our neonatal unit.
Ethical approval was obtained from the Cambridgeshire 2 Research Ethics Committee at Victoria House, Capital Park, Cambridge, CB21 5XB. MHRA approval was sought but not required for this in-house study.
Results: The data were analysed using Microsoft Excel and Stats direct, and displayed using the Bland and Altman method. With incubator scales alone the mean error in weight was 45.5g (limits of agreement ±76.7g). With the adjustable device this error improved to a mean of 2.7g (limits of agreement ±32.2g). An F-test for variance showed that this improvement in accuracy was highly statistically significant (p < 0.001).
Conclusion: The results of our study indicate that a simple adjustable device can be used to correct the error in weights obtained from incubator scales in ventilated neonates and obviate the need for disconnection from the VIP Gold ventilator. The data on the High Frequency Oscillator is not yet complete.