Pyry A.J. Välitalo(1), Monique van Dijk(2), Elke H.J. Krekels(1), Sinno H.P. Simons(3), Dick Tibboel(4), Catherijne A.J. Knibbe(1,5).
(1)Division of Pharmacology, Leiden Academic Centre for Drug Research, Leiden University, Leiden, the Netherlands; (2)Department of Pediatric Surgery and Pediatrics, Erasmus MC-Sophia Children’s hospital, Rotterdam; (3)Department of Neonatology, Erasmus MC-Sophia Children’s hospital, Rotterdam; (4)Department of Intensive Care for Children, Erasmus MC-Sophia Children’s hospital, Rotterdam; (5)Department of Clinical Pharmacy, St. Antonius Hospital, Nieuwegein, the Netherlands
Objectives: It has proven difficult to identify and quantify analgesic efficacy of morphine in newborns. This study aimed to assess the efficacy of morphine in preterm neonates undergoing endotracheal suctioning, using item-level data from multiple pain scales and item response theory (IRT) modelling.
Methods: The data included 140 preterm and term mechanically ventilated neonates, who were randomized to receive either placebo or morphine for analgesia in a double-blind study design[1]. Open-label morphine was allowed for rescue analgesia in both groups. Pain was quantified with items from COMFORT scale (CMT), items from Premature Infant Pain Profile (PIPP), compound scores from Neonatal Infant Pain Scale (NIPS), and VAS scores.
A previously developed PK model [2] was used with sequential IPP estimation method[3] for PKPD modeling. Graded response models of the IRT framework were used to estimate pain as a latent variable[4] based on item-level data from CMT and PIPP, compound scores from NIPS and continuous scores of VAS. Uninformative items with discrimination values less than 1 were omitted from the PKPD analysis. The effect of morphine concentrations on the latent variable was tested with linear, power and (sigmoidal) EMAX models. Additionally, gestational age, postnatal age and bodyweight were tested as covariates on the latent variable.
Results: Only behavioural items of the CMT and PIPP scales had a discrimination value >1, and thus contextual and physiological items from these scales were omitted from the analysis. Morphine concentrations were found to reduce the latent variable pain before, during, and after endotracheal suctioning (p<0.001); the morphine concentration required to reduce pain by one between-subject standard deviation was 60 ng/mL. Moreover, a positive linear relationship was found between postnatal age and pain (p<0.001), but as patients were generally included on the first day of their life, it cannot be excluded that this relation was caused by for instance duration of mechanical ventilation.
Conclusions: As IRT based modeling weighs the information content of item scores dynamically, it is statistically more powerful than an analysis based on compound scores. Thereby the use of IRT reduces the risk of false negative findings. Based on these results, we conclude that there is a modest but definite analgesic effect of morphine in mechanically ventilated neonates.
References:
[1] Simons et al. Routine morphine infusion in preterm newborns who received ventilatory support: A randomized controlled trial. JAMA. 2003;290(18):2419
[2] Knibbe et al. Morphine glucuronidation in preterm neonates, infants and children younger than 3 years. Clin Pharmacokinet. 2009;48(6):371
[3] Zhang et al. Simultaneous vs. sequential analysis for population PK/PD data II: robustness of methods. J Pharmacokinet Pharmacodyn. 2003;30(6):387
[4] Samejima. Estimation of latent ability using a response pattern of graded scores. Psychometrika 1970;35(1):139
Reference: PAGE 24 (2015) Abstr 3388 [www.page-meeting.org/?abstract=3388]
Poster: Drug/Disease modeling - Paediatrics