To our knowledge, this is the first national survey evaluating current opinions of healthcare professionals in Germany regarding placement of an UVC or IO access in an emergency setting in the DR. In accordance with current guidelines, responders preferred an UVC over IO access during transition at birth.
Only a narrow majority of 60% was in favour of an UVC in emergency situations in the DR, which could be due to the fact that establishing an IO access was classified as (very) easy by 67% of respondents, even though only 50% had ever implemented one themselves.
Respondents to our survey rated the level of difficulty according to their own level of training and experience, which might be a reason why, contrary to current recommendations, with less experience the affinity to IO access increased. While many extremely preterm infants born in tertiary centres need a central venous line access during their subsequent neonatal intensive care [10], the UVC is often placed either in the delivery room or shortly afterwards in the NICU [11], as it provides a painless and reliable vascular access for preterm infants avoiding the skin punctures needed for other forms of vascular access [12].
However, as long as the UVC remains the recommend access in DR management in international guidelines [3] and as long as there is a lack of a device that simplifies the inserting procedure, consistent training should be enforced [4].
An umbilical cord simulator may offer a realistic training with real human cords [13] and should be preferred to manikins with an artificial and more unrealistic umbilical cord [14].
Another reason for placing an UVC is that high plasma levels of epinephrine can be reached faster and more reliably via a centrally positioned UVC than via the endotracheal route [15] and, according to the 2015 ERC Guidelines, drugs should be applied this way [3]. It remains unclear whether the same is true for an IO access [16]. Initial studies (excluding neonatal patients) showed no significant interaction between the access route and study drug outcomes [16]. However, IO access in neonates has not yet been investigated in detail, only case series, post-mortem studies and simulation studies could be identified and showed a lack of evidence in this patient group [8].
However, successful placement of an UVC took 46 s longer than application of an IO access in a simulation study [17]. In an emergency situation in the DR, this delay may be responsible for the increasing preference of an IO access during the resuscitation of term neonates, as confirmed by our survey. Therefore, such IO access should be available, trained and taken into consideration on all neonatal units if other access routes have failed [18].
Previous experience with IO access significantly reduced reluctance and increased the willingness to use an IO access as the first choice for emergency vascular access [19].
Besides that, the IO access is a non-natural access route with a complication rate of 13% [9] such as fractures, limb ischaemia and need for amputation [20, 21]. The complications are higher in smaller infants due to the small margin of error when inserting an IO device [22].The “risk of a bone injury” and “causing pain” were the main reasons cited in our survey why respondents would not apply an IO access.
A recent trial in 16 stillborns showed success rates in newborns between 40% and 60%, depending on the needle used [22]. However, checking the correct position of an IO access with a CT-scan is difficult to accomplish during neonatal resuscitation.
Moreover, a major problem of UVC placement is malposition, which is associated with a higher risk of thrombosis [5] and NEC [6]. In our survey, a risk of malposition was mentioned 72 times in the free-text option when respondents were asked to mention most common problems during UVC insertion.
This leads to the first limitation of our study: only few multiple-choice responses were used, thus “malposition” was not offered as a response. Furthermore, we did not compare blunt hollow cannulas with umbilical venous catheters, this might have been interesting especially in the context of resuscitation in the DR.
Additionally, we included fewer open questions to minimize the risk of having a low participation rate.
Another limitation is that distribution of the questionnaire was random and therefore not all neonatal centres might have been reached. This was due to the fact that there is no general email list for specified healthcare professionals in neonatology, which may have introduced a selection bias. In addition to that, GDPR (General Data Protection Regulation) rules dictated anonymity of the e-mail responses received, prohibiting us from gathering data on the number of participants and refusals to participation.
Nevertheless, a strength of our study is a high number of respondents during the 3-month period so that we can still present a broad picture of opinions.