SEE RELATED ARTICLE, P. 610 .

Procedural guidance or assistance using ultrasonography allows visualization of anatomic landmarks. One example is for central venous access, in which ultrasonographic guidance improves procedural success and reduces complications. Before using ultrasonography, clinicians placed central lines by using landmark approaches, with up to 14% of procedures resulting in a complication and up to 25% being unsuccessful. Recognized as a best safety practice by the Agency for Healthcare Research and Quality in 2001, ultrasonographic guidance or assistance is now widely accepted for central line placement.

As with central line placement, lumbar puncture is a common procedure in the emergency department (ED) and is part of the routine evaluation of the febrile young infant. The proportion of failed or traumatic lumbar punctures in pediatric patients can be as high as 40%, and procedural success is not entirely dependent on clinician expertise. There are likely intrinsic factors, notably, patient anatomy, that influence lumbar puncture success. In this context, Neal et al designed a randomized trial investigating the utility of ultrasonographic assistance for infant lumbar puncture as measured by a successful lumbar puncture. The intervention incorporated point-of-care ultrasonographic markings made on the patient’s skin, which identified the optimal needle insertion site from sonographic landmarks, as well as the optimal depth of insertion. The point-of-care ultrasonographic imaging and site markings were performed by a small, select group of sonologists, and the actual lumbar puncture procedure was conducted by the treating clinician. The authors analyzed several combinations of outcomes. Using a definition of first-attempt success and RBC count less than 1,000/mm ³ (primary outcome), success was 58% in the ultrasonography arm compared with 31% in the traditional, landmark arm (absolute risk difference 27% [95% confidence interval {CI} 10% to 43%]). A more expansive definition of success within 3 attempts and RBC count less than 10,000/mm ³ yielded 81% success in the ultrasonography arm compared with 56% in the traditional arm (absolute risk difference 25% [95% CI 10% to 41%]). In addition to being statistically significant, even the lower bounds of these 95% CIs are clinically significant.

Although a large effect size existed in this trial, there are some features of the study that might preclude current widespread adoption of this technique for infant lumbar punctures. First, the failure rate (44% to 59%, depending on which outcome was used) in the control group (lumbar punctures without ultrasonographic assistance) is a higher estimate than in most previous studies and may have magnified the effect sizes observed in this study. The study setting was an academic children’s hospital with various levels and types of trainees performing procedures. As such, the effect of ultrasonography on lumbar puncture success may not be as large at nonteaching settings. Although the authors found no difference in success according to clinician experience, the study was not powered to detect a difference based on this outcome; additionally, there was a greater effect of ultrasonographically-assisted site marking for less experienced providers. Therefore, more data are needed on the benefit of ultrasonographic assistance for lumbar punctures when performed by experienced providers. Ultrasonographically-assisted lumbar puncture may have its highest utility in certain situations, such as after a single failed attempt or as an adjunct to simulation for the novice proceduralist.

Other design threats exist. In the study by Neal et al, individuals performing the lumbar puncture could not be blinded to the intervention. The study protocol used essentially only 2 sonologists, raising concern over the reliability of point-of-care ultrasonography for this purpose. Specifically, although the authors claim the study sonologists achieved competency with ultrasonographically-assisted site marking with minimal training, whether this translates to other providers, particularly those with little or no ultrasonographic training, remains to be studied. None of these threats make the observations invalid, but they do offer potential concerns about the impact others might see in deployment during routine clinical practice.

The addition of ultrasonographic guidance adds time to a procedure; in the current trial, there was a median of more than 5 minutes needed to perform the site markings for individuals in the ultrasonography arm. This represents the minimum additional time needed for ultrasonographic assistance, in addition to the time needed to retrieve the machine and set it up in the room before beginning the procedure. In a busy ED, even 5 minutes can alter care opportunities for the collection of patients. Because success happens more often with the intervention, perhaps this overall effect is positive (ie, less wasted effort), but we need a different data analysis to address that opportunity.

By improving the success rates of infant lumbar puncture, there is potential to reduce pain and discomfort, hospitalizations, ED and hospital lengths of stay, antibiotic use, and costs. As secondary aims, Neal et al sought to answer whether hospital length of stay and duration of intravenous antibiotic therapy changed with the use of ultrasonographic assistance. The premise here is that unsuccessful lumbar punctures in the ED would lead to prolonged hospitalization to obtain adequate cerebrospinal fluid or presumptively treat with intravenous antibiotics. In their study, neither the median hospital length of stay nor the median duration of antibiotics was statistically different between the 2 groups; however, the study was not powered according to these outcome measures. It is patient-centered outcomes such as these that are ripe for future study.

Previous studies of ultrasonographically-assisted lumbar puncture, primarily focused on adult patients, demonstrate various results, with some showing a reduction in failure and others showing no benefit to ultrasonographic assistance compared with landmark methods. One small study evaluating ultrasonographically-assisted site marking for infant lumbar puncture found improved success with ultrasonographic assistance. Although it is encouraging that we now have 2 trials showing a benefit of ultrasonographic assistance for infant lumbar puncture, there remains some ambiguity over for whom and how much benefit exists.

Ultrasonographically-assisted lumbar puncture may offer an advantage over the landmark method and lead to improved patients outcomes. Perhaps with more evidence, similar to what we observed with ultrasonographically-guided central venous access, ultrasonographic assistance will become standard care for infant lumbar puncture. Right now, this technique shows much promise.