The fact that neonatal abstinence syndrome is common enough to allow, and require, formal study in this era is a very sad commentary on the resurgence of opiate addiction in the United States. What follows is a commentary not just on 1 article, but on 2 paired articles plus an editorial, all in the same issue of Journal of Pediatrics. Collectively they provide some valuable lessons about healthcare research and delivery.
Source: Hall ES, Meinzen-Derr J, Wexelblatt SL. Cohort analysis of a pharmacokinetic-modeled methadone weaning optimization for neonatal abstinence syndrome. J Pediatr. 2015;167(6):1221-1225; doi:10.1016/j.jpeds.2015.09.038. See AAP Grand Rounds commentary by Dr. Johnathan Wagner (subscription required).
PICO Question: Among infants with neonatal abstinence syndrome, does treatment with a revised methadone weaning protocol result in improved treatment outcomes relative to a standard protocol?
Question type: Intervention
Study design: Pre-post
The main focus of Dr. Wagner's AAP Grand Rounds commentary is a cohort analysis using historical controls. In a group of hospitals in Cincinnati, they looked at management of neonatal abstinence syndrome (NAS) before and after a change in their standard of care. The first study interval of 18 months in 2014 and 2015 utilized frequent low doses of methadone according to a standard protocol, while the new standard of care had fewer doses but starting at higher individual dosing based on weight. The investigators found that the new standard of care resulted in a shorter weaning period as well as shorter hospital stay. However, as Dr. Wagner pointed out in his commentary, this was not a randomized controlled trial. Use of a historical control group opens the door for variables other than the methadone weaning regimen that might be different between the 2 groups, perhaps a subtle but important change in general care not easily detected from chart review. The authors stated that randomization was "not feasible," but they didn't elaborate. I presume the only reason it wasn't feasible is because they decided to change the standard of care for NAS in their hospitals rather than perform a randomized controlled trial. Also, the main driver for clinicians using either protocol is a NAS scoring tool, the modified Finnegan tool, originally developed in 1975. It has been well accepted and used for decades, but it is also somewhat subjective, making it more prone to bias than are scoring tools that do not require a high degree of clinician interpretation. This adds additional concerns in interpreting a study with historical controls.
Equally interesting to me is a partner article preceding this one in the same edition of the journal. It describes the pharmacokinetic analyses that led to development of the revised NAS management protocol. It is based on the concept of population pharmacokinetics. Rather than drawing numerous blood samples from every infant to model drug kinetics, this approach uses relatively few samples from each infant, and then applies computer simulation techniques to broadly describe drug pharmacokinetics for a population. This is now a very common method of analysis and is particularly useful for neonates where issues regarding volume of blood and frequency of blood draws obtained for research can raise clinical and ethical problems. The researchers used the information from this study to develop their new standard for NAS care.
One final note on the "triple aim" concept. I was reminded of this in the excellent editorial accompanying these articles. The term was proposed by Berwick, et al and denotes the goal of healthcare research and delivery: 1) better health for individuals; 2) better health for populations; and 3) reduction of healthcare costs. This new NAS protocol examined all of these aspects and demonstrated some success, with the caveat that it is only a pilot study with many questions remaining to be answered.
