Abstract
We read, with interest, the manuscript by Ahsan et al. [1] which highlights the importance of considering noxious stimuli in the vulnerable population of neonates requiring intensive care. We completely agree that attention to patient discomfort during any test or procedure should be considered a priority and appropriate comfort provided. We question the strength of the authors’ conclusions that targeted neonatal echocardiography (TnECHO) causes significant pain and physiological instability, however, based on several methodological concerns.
First, the authors do not provide sufficient data on the qualifications or scope of practice of those performing the ultrasound examinations. From the manuscript it is not possible to determine that TnECHO was, in fact, the examination performed as opposed to a limited point of care ultrasound (POCUS) assessment of the heart. The past 10 years has witnessed significant growth in the field of neonatal hemodynamics, which has led to enhanced mechanistic insights and precision in the selection of treatment. The expert model is based on the completion of a comprehensive and standardized TnECHO assessment in an efficient and patient-centric manner. Qualification as a practitioner of TnECHO requires completion of both basic and advanced training with a period of not less than 8 months dedicated to pediatric echocardiography and not fewer than 300 scans performed, in a setting with expert supervision [2]. Further, competency-based training emphasizing exposure to a critical mass and diversity of pathophysiology has been proposed as an essential training characteristic [3]. This is relevant because the magnitude of physiological disturbance may be directly related to the experience of the practitioner. Whereas a learner or operator who performs low annual echocardiography volumes (< 150/year) may take 60 minutes or more to obtain a complete study, an expert may be able to complete 80–100 images in less than 10–15 minutes. Cardiac POCUS practitioners, on the other hand, are less regulated, and depending on the training model, may have a widely varied skill set. Although we have confidence that the authors have undergone training appropriate to their scope of clinical practice, it is difficult to generalize the results of the study without some details as to what their typical practice and training consists of. Additionally, details about the scan duration and ambient conditions (light, stimulation) would be beneficial to provide context regarding the magnitude of noxious stimulus introduced. We also question the authors’ use of the term TnECHO in association with their stated aim of screening for critical congenital heart disease, which is not typically considered within the scope of practice either of a TnECHO practitioner [2] or cardiac POCUS provider.
Second, an alternative approach to evaluating pain may have conveyed a lower risk of bias to the readers. A non-blinded analysis comparing a subjective score taken in “a 30 second interval” prior to and after the procedure to “the highest score any time during the TnECHO examination in each domain” is suspect. Neither the number, nor the experience of PIPP-R raters was specified. Multiple raters may minimize bias but correlation among raters is poor, indicating that the lack of this data and any reliability testing performed, limits the interpretation of the scoring [4]. A random 30 seconds sample during the procedure by a blinded 3rd party or panel, as described by Taplak et al. [5] would have been less subject to bias. Additionally, the PIPP-R tool itself is validated only for acute pain (i.e. heel lance or venipuncture) which lasts for seconds and the infant’s behavior is evaluated for 30 seconds after the procedure. The need to perform evaluation during the ultrasound, the duration of which has not been specified as previously described, may not be a validated technique. A more appropriate tool to evaluate exposure to discomfort over the duration of an ultrasound might have been Neonatal Pain, Agitation and Sedation Scale (N-PASS), or the Neonatal Facial Coding System (NFCS), which have been evaluated for both acute and chronic exposure [6]. Similarly, the use of “the highest score at any time” is subject to the risk of inflating PIPP-R scores by choosing temporally dissociated “worst moments” for different aspects of the score. Continuous vital sign monitoring before, after, and for the duration of the procedure with analysis of trends in addition to absolute change may also improve insight into the true impact of the intervention.
It is also problematic that the authors did not acknowledge in their discussion the literature which shows that many interventions cause a change in pain score [7–9]. The authors imply by their methods and conclusion that TnECHO is unique; the provision of a comparison procedure with documented impact on the PIPP-R, such as endotracheal suctioning or heel lance, would have enabled a superior perspective on the degree of patient impact. While scores of < 7 have been taken to indicate no or minimal pain, there is no clear definition of mild, moderate, or severe pain based on PIPP-R scores alone [10]; thus quantification as “significant pain” based on score change alone may not be valid, especially without a comparator of the pain experienced by the infant from other interventions. As it seems from the data presented, the changes in heart rate, arterial pressure, saturation, and respiratory rate were minor deviations on the order of approximately 1%change from baseline and recovered within 10 minutes. These physiological changes would not be considered clinically relevant by most practitioners.
Third, the rationale for studying TnECHO as opp-osed to other forms of neonatal imaging is unclear. In conducting an evaluation of the impact of imaging it would have been ideal to compare TnECHO assessments performed by neonatologists with echocardiography assessments performed by sonographers, radiologists, or pediatric cardiologists. In addition, comparison of radiological assessments (chest or abdominal radiography or ultrasound assessments of the abdomen, chest, or head) would have been a useful comparison. Variance in scores between novice vs experienced operators would also provide additional insights. While we agree conceptually that it is important to appraise the impact of any test on patient well-being, it is imperative to conduct such studies in a balanced and unbiased manner.
Finally, we question the authors’ choice to compare PIPP-R score between term and preterm patients. The authors observe that the preterm patients had scores 1-2 points higher at all time points than the term patients. How did the authors account for the fact that the PIPP-R score adds 1 additional point for patients with a gestation between 32–35 + 6 weeks and 2 additional points for patients 28–31 + 6 points [11]? A more valid comparison might have been to use an illness severity score, such as the Score for Acute Neonatal Physiology with Perinatal Extension II (SNAPPE-II) to stratify patients by illness severity [12]. A comparison between low- and high- acuity patients may have clinical utility. The performance of any procedure should be justified by a risk/benefit analysis. A clinically unstable patient may experience both stimulus and benefit differently than a well patient when exposed to a potentially noxious procedure. Whereas, in this cohort, relatively mature patients seem to have been evaluated for ductal shunt despite a low risk of complications and an unclear burden of clinical illness. “Well preterms” may achieve higher PIPP-R scores with a lower likelihood of benefit.
In conclusion, we agree with the authors that consideration of pharmacological and non-pharma-cological methods to minimize the impact of handling on this vulnerable patient population is warranted. We also advocate for other avenues, such as excellence in procedural performance achieved via high-quality training and the concentration of expertise, regardless of whether TnECHO or cardiac POCUS are the ultrasound modalities in use. This, in combination with appropriate patient selection for testing and appropriate selection of ultrasound modality, are also essential tools to mitigate discomfort. We are firmly supportive of TnECHO guided hemodynamic care practiced within scope of training and with adequate systems in place for backup in the setting of unexpected congenital heart disease. TnECHO is a highly valuable clinical tool, especially in the setting of acute cardiovascular instability. On balance, when used appropriately, the clinical insights gained outweigh minor discomfort in a similar fashion as it would in the setting of a nursing assessment, chest radiography and other ultrasounds commonly accepted as standard of care such as of the head for intracranial bleeding. We agree with the authors, however, that this field would benefit from further work using methodology which minimizes the risk of bias and uses clinically relevant, ideally objective, outcomes.