[Behavioral pain scales assessment in neonates]

To establish the sensitivity and specificity of two behavioral pain scales in different gestational ages (GA).

133 clinically stable neonates, <72 h of life, without diseases, analgesic/sedative use or Apgar<7 at 5' were randomly assigned to receive capillary puncture--P or alcohol swab friction--F. Patients were divided according to GA (28-33 wk=A; 34-37 wk=B; 38-41 wk=C ) and procedure: Group A-P (n=17, BW 1.5 +/- .4 kg); A-F (n=18, BW 1.5 +/- .4 kg); B-P (n=25, BW 2.5 +/- .5 kg); B-F (n=25, BW 2.4 +/- 0.6 kg); C-P (n=23, BW 3.3 kg +/- 0.5 kg); C-F (n=25; BW 3.3 +/- 0.4 kg). A neonatologist, blind to P or F, evaluated the Neonatal Facial Action Coding System (NFCS 0-8 pts, pain>3) and the Neonatal Infant Pain Scale (NIPS 0-7 pts, pain>3). All evaluations were performed prior to P or F, without (pr) and with foot heating (h), during (0), 1' and 3' after P or F. Reliability was assessed in 20% of the sample. Agreement rate was NFCS-97% and NIPS-95%.

During the procedure, median NFCS and NIPS P score were greater than F (M. Whitney), p< or =0.0001) for A, B and C groups. At 1', NCFS and NIPS P scores were greater than F (MW, p< or =0.04) for A and C groups. For P patients, NFCS and NIPS scores were similar among the 3 GA groups at all study periods (K. Wallis). For F patients only at 0; NIPS scores were different (K. Wallis, p<0.03). During the procedure, NFCS>3 had sensitivity of 88-91% and specificity of 80-84%, NIPS>3 presented sensitivity of 77-87% and specificity of 68-83%.

Both behavioral scales are sensitive tools for pain assessment in premature infants, although NFCS seems superior to NIPS.