Validating skin conductance for measuring acute pain in mechanically ventilated infants

Measuring pain in mechanically ventilated infants is challenging. The measurement of skin conductance (SC) is based on the sympathetic nervous system response to stress. This study purpose was to evaluate the validity of SC for measuring pain in mechanically ventilated infants. A prospective cross-sectional observational design was used to study SC and its relation to: the category of procedure (i.e., painful or non-painful); the phase of procedure (i.e., before, during and after), and referent pain measures (i.e., Premature Infant Pain Profile-Revised (PIPP-R) and Neonatal Facial Coding System (NFCS)). Eligible infants were those up to 12 months of age, in intensive care units, who were mechanically ventilated, and required painful and non-painful procedures. From October 2017 to November 2018, 130 eligible infants were identified, and 55 infants were studied. SC (number of waves per second) during painful procedures (median 0.27, interquartile range 0.2-0.4) was statistically significantly higher than those during non-painful procedures (0, 0-0.09). SC during painful procedures was statistically significantly higher than those before (0, 0-0.07) and after painful procedures (0, 0-0.07). SC showed moderate statistically significant positive correlations with PIPP-R (Spearman's rho=0.4-0.62) and the four-item NFCS (Spearman's rho=0.31-0.67) before, during and after painful or non-painful procedures respectively. SC had excellent performance (area under the receiver operator curve=0.979) with excellent sensitivity (92.31%), specificity (95.42%) and negative predictive value (99.21%) but only sufficient positive predictive value (66.67%) when used to discriminate moderate-to-severe pain. This present study is the first to evaluate different sources of validity evidence of SC for pain measurement in the same population of infants during the same time period. Specifically, SC showed good validity for pain measurement in mechanically ventilated infants in relation to the category of procedure, the phase of painful and non-painful procedures and referent pain measures. This study was funded by the Canadian Pain Society (CPS) Trainee Research Award (Clinical Science, 2018-2019) and Sigma Theta Tau International/Rosemary Berkel Crisp Research Award (2018-2019). The first author (J. Hu) was supported by the Ontario Trillium Scholarship (2015-2019) and International Doctoral Scholarship.