Effects of paternal and maternal voice on procedural pain and physiological responses during arterial blood sampling in preterm infants in a neonatal intensive care unit: a three-arm randomized controlled trial

Objectives To evaluate the effects of paternal and maternal voice interventions on procedural pain, comfort, heart rate, oxygen saturation, and behavioral state during arterial blood sampling in preterm infants, and to compare their effects with routine care and with each other. Methods A three-arm randomized controlled trial was conducted in China between May 2024 and June 2025. Of 135 randomized preterm infants, 119 completed the study (paternal voice, n = 39; maternal voice, n = 39; routine care, n = 41). Parental voice recordings were played from 10 min before until 10 min after arterial blood sampling. Outcomes were assessed at five time points. Generalized estimating equations were used for repeated-measures analyses. Results GEE analysis showed significant group and time effects for pain ( P = 0.027 and P < 0.001), with lower pain scores in the paternal voice group than in routine care (mean difference = −0.51; 95% CI −0.96 to −0.06; P = 0.021). Paternal voice was also associated with lower heart rates during sampling, and better comfort scores at 1 and 5 min after sampling, although the overall group effect for comfort was not significant. Maternal voice intervention did not differ significantly from routine care. No between-group differences were observed for oxygen saturation or behavioral state. Conclusions Paternal voice was associated with favorable pain and heart rate responses in preterm infants during and after arterial blood sampling, with time-specific differences in comfort during early recovery. Its overall effect on comfort and relative effectiveness compared with maternal voice require further investigation. Implications for Clinical Practice Paternal voice may offer a simple, safe, and feasible adjunct to procedural pain management and family-centered care in preterm infants.