In a full factorial study design, we investigated the influence of insertion angle, needle type, bevel orientation and tissue on needle visibility in sonographic image sequences.

Our findings indicate that the SonoPlex Stim Sprotte, SonoTAP Facet and SpinoStar PencilPoint were the needles with best ultrasound visibility and that the bevel down orientation was associated with the best results all statistically significant. Furthermore, an increase of the insertion angle decreased the probability of good needle visibility, also with statistical significance. The tissue resembled by the different cadavers, has significant influence on the needle visibility as well.

Since the reference variables for groups of categorical variables were not necessarily the best or the worst in their category and because the estimate for angulation is given per degree, it is important to compare the total effect range of every group of variables. This allows to interpret by how much each measure impacts the needle visibility overall. Therefore, the effect ranges (from minimal to maximal) are presented on the log-odds scale in relation to the curve for probability in Fig. 2. The figure shows that the angulation affected needle tip and shaft visibility with a range of 6.33, respectively 7.36 on the log-odds scale. This resembles a large difference in the probability of good or poor visibility practically anywhere on the curve. The tissue conditions affected the probability of good or poor needle visibility with a smaller range of 3.76 and 3.96. The needle type and bevel orientation however, had a surprisingly small overall effect on the probability with ranges of 1.25 and 1.86 (needle type tip and shaft) and 1.45 and 0.95 (bevel orientation tip and shaft).

This indicates that despite certain echogenic needles may improve visibility with statistical significance compared to other needles, the effect of good angulation is still a multiple of that effect which can be expected from switching to a better, most likely more expensive needle. The tissue is a factor that cannot be controlled for in the clinical setting, but it also affects the needle visibility much more than the choice of a needle does. While the importance of limiting needle angulation for better ultrasound visibility is common knowledge, this relationship has not been quantified before and the effect of the needle type on needle visibility was much smaller than expected.

In the scientific literature echogenic needles have shown to improve the ultrasound visibility with statistical significance, especially at steeper insertion angles.

For better comparability with other studies on this topic, we have presented bar plots of our data, showing the visibility of needle tip and shaft of echogenic and non-echogenic needles in the supplemental content (Supplemental Fig. 1). Despite statistical significance at higher angles (Chi-Square-Test), the proportion of poor visibility remained high. At angles of 40–45° needle tip visibility was poor in 89–95% with echogenic needles and in 99–100% with non-echogenic needles. Needle shaft visibility was poor in 71–81% with echogenic needles and in 99–100% with non-echogenic needles. On other words, this would mean that non-echogenic needles are expected to almost never show good visibility at 40–45°, whereas 10–25% of echogenic needles are expected to have good visibility at these angles. In our opinion, this effect is smaller than expected, but can be of relevance in some cases, if it translates to clinical practice.

Studies that investigated the visibility of tip or shaft of echogenic needles compared to non-echogenic ones regularly show a significant benefit of the echogenicity, especially at higher insertion angles of 30–75°. [1, 3, 9] This statistical significance is also present in our data. Unfortunately, these studies did not relate this effect to other measures like the actual effect of angulation or tissue.

The systematic review of Hovgesen et al. is the largest review on the topic and the advantages of echogenicity have been shown to be most apparent when using steeper insertion angles in relation to the transducer. [10] However, a quantification of needle visibility via scoring could not be done from the inhomogeneous reporting of the individual studies.

Guo et al. assessed needle visibility scores in a human Thiel Cadaver [1]. Primary endpoint in this study was needle visibility assessed by two independent reviewers.

Although, needle visibility could be improved using echogenic needles, in-plane-technique and spatial compound imaging, a high percentage of needles had shown only poor visibility [1, 3].

Maecken et al. did an investigation with visibility at 0- versus 45-degree angles in the animal model and concluded that visibility was severely limited at 45 degrees and that only few of the investigated echogenic needles had an “acceptable” visibility at that angle. [11].

To our knowledge, the relationship between angulation, tissue, bevel orientation and choice of an (echogenic) needle has not been quantified as precisely as we have done in this analysis.

This study was performed in a non-clinical setting using embalmed human cadavers. Needle visibility in embalmed cadavers is not identical, yet comparable with that in human. Other models (meat, artificial gel) are more problematic as they provide significantly higher needle visibility [12]. Therefore, this is only a minor limitation.

The visibility was subjectively rated by a single observer, which however, resembles clinical routine in most cases.

Taking the independent variable “tissue”, represented by the different cadavers, into the regression analysis is debatable, since in clinical practice, the patient itself cannot be altered. The estimates for the different cadavers have therefore no direct implications on how to improve needle visibility in general, but our aim was to present a statistical model with high explanatory value, which we achieved by taking the tissue into the analysis. Not considering it or not knowing about the impact of the tissue condition may also lead to frustration when trying to optimize visibility via controllable factors or to overambitious attempts to perform blocks in these situations. Having this knowledge on the other hand may take pressure off clinicians who abide from undertaking risky blocks under poor conditions. These aspects however are hypotheses, that are derived from the presented statistical model from data of embalmed cadavers and conclusions may not necessarily translate into the clinical setting.