Assessment and management of pain in newborns hospitalized in a Neonatal Intensive Care Unit: a cross-sectional study 1

ABSTRACT Objective: to determine the frequency of pain, to verify the measures adopted for pain relief during the first seven days of hospitalization in the Neonatal Intensive Care Unit and to identify the type and frequency of invasive procedures to which newborns are submitted. Method: cross-sectional retrospective study. Out of the 188 hospitalizations occurred during the 12-month period, 171 were included in the study. The data were collected from the charts and the presence of pain was analyzed based on the Neonatal Infant Pain Scale and on nursing notes suggestions of pain. For statistical analysis, the Statistical Package for the Social Sciences was used, and the significance level was set at 5%. Results: there was at least one record of pain in 50.3% of the hospitalizations, according to the pain scale adopted or nursing note. The newborns underwent a mean of 6.6 invasive procedures per day. Only 32.5% of the pain records resulted in the adoption of pharmacological or non-pharmacological intervention for pain relief. Conclusion: newborns are frequently exposed to pain and the low frequency of pharmacological or non-pharmacological interventions reinforces the undertreatment of this condition.


Introduction
Painful experiences in the neonatal period may result in physiological and behavioral alterations, as well as changes in the development of the nervous system, which can provoke considerable damage in the future (1)(2) . However, several studies indicate that hospitalization in a Neonatal Intensive Care Unit (NICU) includes a high number of painful procedures (3)(4)(5) , most of them necessary for diagnosis and treatment.
Moreover, studies still show gaps in the knowledge of nursing professionals regarding assessment and management of pain (6)(7) . In addition, a research demonstrates that, in general, the use of the available analgesic treatments is inadequate and insufficient (7) .
Another study emphasizes that a considerable number of health professionals do not assess the level of pain based on scales developed for this purpose (8) . This result demonstrates the need to increase the use of the available evidence on effective measures for pain management, in order to improve the care provided to newborns (NBs).
Faced with this alarming reality regarding neonatal pain, the objectives of this study were to determine the frequency of pain, to verify the measures adopted for pain relief during the first seven days of hospitalization in a NICU and to identify the type and frequency of invasive procedures to which newborns are submitted.

Method
This is a retrospective and cross-sectional study carried out at the NICU of a medium-complexity public teaching hospital located in the city of São Paulo, Brazil. The sample population was composed by NBs hospitalized in this unit. In this study, the conditions that could be related to pain, such as the medical devices in use and the painful procedures to which the newborns were submitted to during the first seven days of hospitalization were considered.
Regarding the term "devices in use", all devices used in the NBs for therapeutic or monitoring purposes, such as catheters, drains, cannulas and rectal thermometers, were listed. Regarding the term "invasive procedures", it should be noted that the definition used is based on an previous study (9) . Therefore, were considered as invasive those procedures that affected the integrity of the skin or mucosa, as well as those in which there was insertion of devices in natural cavities.
The inclusion criteria adopted were: newborns admitted to the NICU between June 2013 and May 2014; and, at the moment of admission, the maximum postnatal age was 28 days for the full-term newborns (FTNB) and for the preterm newborns (PTNB) the maximum postmenstrual age (PMA) was 44 weeks. NBs who were hospitalized for less than 6 hours and those who, at admission, were older than 28 days or had a PMA of 45 weeks or more were excluded.
During the collection period, 188 hospitalizations occurred in the NICU. Of these, 17 were excluded: two because they were not located by the Medical Records and Statistics Service after several attempts; five because the hospitalization period was shorter than six hours; and 10 because the newborns were older than 28 days at the time of admission. It is worth noting that, of the total of 171 eligible hospitalizations, 21 consisted of rehospitalizations of NBs previously included in the study. These data were considered because each hospitalization resulted in a new clinical and care scenario, with its own painful events. Therefore, the data of this study refer to 171 hospitalizations corresponding to 150 NBs. and non-pharmacological measures were all the interventions described in the nursing notes, provided they were indicated as related to the record of pain.
The nursing team of the institution where the research was conducted uses the NIPS scale for daily pain assessment in NBs since 2011. The NIPS is an instrument created in 1993 to assess the level of pain in full and preterm newborns. Scores higher than 3 in the scale indicate the presence of pain (10) . However, for this study, pain was considered present when the score was above zero, since, according to the pain evaluation form used at the institution, a score between 1 and 2 indicates mild pain, between 3 and 5 it indicates moderate pain, and between 6 and 7, severe pain.
It should be noted that although the NIPS is part of the form adopted for all patients hospitalized in the pediatric sectors, there is no routine or pre-established flowchart for pain relief measures.
Despite the fact that the institution has a scale for assessing pain, the nursing notes that described the newborn as crying, agitated or with an expression of pain www.eerp.usp.br/rlae 3 Sposito NPB, Rossato LM, Bueno M, Kimura AF, Costa T, Guedes DMB.
were considered as an alternative source of information, since the presence of these conditions would minimally correspond to a score one on the NIPS, and due to the empirical knowledge that the application of the scale occurs less often than ideal. In addition, Odds Ratio (OR) was calculated to analyze the association between categorical variables and, for that, contingency tables were used.

Results
Out of the 171 hospitalizations, 134 were newborns who were admitted in the NICU only once, while the remaining 37 correspond to 16 newborns that were admitted between 2 and 4 times in the period assessed.
Therefore, a total of 150 NBs were included in the study.
Most of the participants (56%) were male, and the mean length of stay in the NICU was 9.12 days, as shown in Table 1, which displays the main characteristics of the newborns and their hospitalizations. The records of three NBs did not include Gestational Age (GA). However, for the 147 NBs for which this data was available, PMA ranged from 23 to 43 weeks and, similarly to GA, presented a median of 36 weeks.

Records of non-pharmacological interventions
related to the performance of the procedures were not found in the medical records and no pharmacological intervention was registered for more than 96% of the total number of procedures. Of the 172 (3.6%) procedures in which at least one analgesic or sedative was used, the most frequent interventions were the combination of midazolam and Fentanyl® (37.8%) and the administration of midazolam alone (33.9%).
There was a statistically significant association between the number of procedures and the number       European countries (11) . This study found that NBs required a longer time of mechanical ventilation when compared to the other patients (11) , which demonstrates the need for an in-depth evaluation of the prescription of these drugs, considering the cost-benefit relation for each patient. scale, only 24% reported using it at all times (12) .
Therefore, it is necessary to investigate the personal, and swaddling, should be highlighted (13)(14) .
In 50.3% of the hospitalizations, the newborns presented pain at least once during hospitalization. This frequency is higher than the value of 30% (15) obtained from pediatric patients' charts in another study.
However, it still is probably underestimated, since this study indicated the existence of documentation gaps, seeing that the frequency of pain was 72% when it was reported by the nurse, the patient, or the caregivers. According to Brazilian studies, crying and facial expression are the main parameters used for pain assessment (6)(7)16) . However, using crying as an indicator of pain is difficult, not only because it occurs in situations where there is no painful stimulus, but also because the NBs might be unable to cry due to the devices used or their health conditions (17) . Therefore, the fact that, in this study, the isolated record of crying resulted in the smallest number of interventions may be due to its nonspecificity or because it is not considered a condition that alone justifies an intervention.
Still on the low frequency of interventions for pain relief, it is worth noting that non-pharmacological interventions are recognized as effective when isolated or as measures complementary to pharmacological treatment (2,(12)(13)18) . Also, they are a potential field of action for nursing care, and have not been properly implemented yet, considering the frequencies described (3)(4)(5) . Consequently, it is necessary to overcome the existing barriers, promoting knowledge about the subject and autonomy for decision making. in both studies (6)(7) .
The discrepancy between identification and management of pain can occur for several reasons.
Professionals working in Canadian NICUs have identified three themes that influence pain-related practice: a culture of collaboration and support for evidencebased practice, threats to autonomous decisionmaking, and the complexities in care delivery (19) .
Inter-professional collaboration and trust, joint work with families and the incentive for professional development were considered favorable situations. On the other hand, hierarchical relationships, care based on personal preferences, patient-related complexities, and organizational culture were unfavorable factors to the quality of care (19) .
The lack of changes in the context of the culture of pain goes beyond borders and requires the participation and joint action of managers of health organizations, professionals at all levels of care, and family members (18) .

Conclusion
The data presented indicate undertreatment of pain and underutilization of the NIPS as an instrument to guide nursing care for pain management. The presence of pain was recorded in approximately half of the hospitalizations through the NIPS score or the nursing notes. Regarding the procedures, it was observed that the newborns are exposed to a large amount and diversity of invasive procedures during the hospitalization, especially heel stick and aspiration of airways.
A significant deficiency of pharmacological and non-pharmacological interventions for effective pain relief is noted, since more than half of the records of pain did not result in the adoption of any measure.
In the hospitalizations where they were adopted, pharmacological interventions were more frequent.
As limitations of the study, it is important to note that it was conducted in a single institution and it had a retrospective design. Therefore, health professionals' records probably do not accurately reflect the care provided. Thus, future studies should cover a larger number of institutions in order to allow the comparison of assessment and management of neonatal pain in different scenarios, and also use prospective designs as a way to minimize data loss.