PURPOSE: Numerous sources of noxious stimuli exist in the NICU, many of which are painful. Research is ongoing to differentiate stimuli that cause acute pain versus chronic pain, the severity of the pain, and the variation in response across a variety of gestational ages. However, there has been little attention given to the variety of stimuli that premature infants are subjected to with a determination of which stimuli are most noxious. This study employed a Delphi survey strategy in an attempt to categorize a large number of commonly occurring stimuli in the NICU into 4 distinct groups: actual tissue damaging stimuli; insertion of a foreign body; manipulations; and physiologic change. The ultimate goal was to determine operationally which group of stimuli is most noxious.
METHODS: Neonatal clinical research literature from 2004 to 2006 was evaluated using PubMed and OVID databases and key search terms such as neonatal nurse researcher, neonatal research, neonatal intervention studies, and neonatal intensive care. The search engine Google was also used with the key terms.
PARTICIPANTS: These searches generated 28 doctorally prepared neonatal nurse researchers. A round 1 instrument, as well as a cover letter, was mailed to 26 participants (2 had a conflict of interest) requesting a 2-week response time. A postage paid self-addressed envelope was included. This initial instrument gave preliminary definitions for each of the 4 proposed categories and provided a list of 38 common nursing interventions. Comments were also solicited regarding the need for additional categories and essential nursing interventions to be added. Thirteen participants responded; 1 was unusable and 1 came in too late.
RESULTS: The analysis of round 1 instrument demonstrated a need to eliminate 1 intervention and add 2 more. Interventions were categorized on the basis of the respondents' no. 1 choice in each category. New instructions for round 2 asked that respondents select the best fit for each, since the categories were not necessarily mutually exclusive. The round 2 instrument provided respondents with data on which category was chosen most often by their peers for each intervention and a new Likert-type scale asking for their agreement with the group as a whole for each intervention. Final analysis of round 2 demonstrated a mean of better agreement of 4 or more (4-4.67) on the scale for 30 interventions. Several interventions were eliminated from the model owing to their infrequency. This analysis was based on 9 respondents.
CONCLUSIONS: This study provides a framework for classification of commonly occurring interventions in the NICU and a way to collapse data for analysis. In future research of painful stimuli, this model will allow for an enhanced interpretation of stimuli and their subsequent impact on the growing premature infant.