\begin{minipage}{\columnwidth}
\centering
\includegraphics [width=1.1\textwidth]{figures/volumevsprotocol.png}
\label{volumeofadmissions}
\captionof{figure}{Relationship between volume of admissions and presence of a oxygen weaning protocol. The result was normalised for the units with an admission volume of less than 500 cases a year. The largest units were more likely …show more content…
Of the rest, 21 clinicians (42\%) targeted between 8.1 and 10 kPa. Only 8 (16\%) aimed for the normal range (10.1-13 kPa).(Figure …show more content…
The $PaO_{2}$ range from $< $8 kPa in the top panel to $> $13 kPa in the bottom panel within each scenario. $FiO_{2}$ ranges from 0.21 (blue) through to 0.81-1 (purple). The three scenarios in the top panel show a pattern of more restrictive $PaO_{2}$ targets with increasing $FiO_{2}$. The 2 scenarios in the lower panel show that higher normal $PaO_{2}$ ranges are targeted irrespective of increasing $FiO_{2}$.}
\end{minipage} \\
The response to no improvement after 24 hours of intensive care followed a similar trend as the initial management. A larger proportion targeted higher \textit{$PaO_{2}$} in children who suffered from pulmonary hypertension and traumatic brain injury. The majority of respondents were permissive with the children suffering from acute respiratory distress syndrome. Almost 58\% of the respondents will aim for a \textit{$FiO_{2}$} of 0.4-0.6 to the amended \textit{$PaO_{2}$} targets.
An equal number of respondents felt there was a need for a randomised control trial with tight arterial oxygenation targets. A large number were not fully convinced of the need for a trial.(Figure 3.8) This suggests that clinicians are not yet convinced with the evidence of harm from hyperoxia in critical illness. A small proportion of the respondents felt it was not ethical to perform such a trial