Pain assessment in critical patients using the Behavioral Pain Scale

BACKGROUND AND OBJECTIVES : The sensation of pain is essential for life, and its assessment in critical non-contacting patients can be performed using validated scales. The Behavioral Pain Scale is a highly accurate tool that has been widely used in this group of patients. This study aimed to describe and charac-terize pain and the use of analgesia in the emergency or intensive care service. METHODS : This was a cross-sectional study with a quantita-tive approach with 67 critically ill patients unable to verbalize their pain perception, who were hospitalized in the emergency service or Intensive Care Units of a public hospital in Vitória da Conquista, Bahia from April to July 2017. Clinical and epide-miological data were collected using the medical record and then applied to the Behavioral Pain Scale for pain assessment. RESULTS : There was a predominance of male patients (47/70.1%). Three groups were identified based on the use of sedatives and analgesics: patients taking sedatives and analgesics combined, only analgesia, and those without any sedation or analgesia. We observed ascending Behavioral Pain Scale scores in all groups during tracheal aspiration, but the same did not occur with the physiological parameters. CONCLUSION : The study proposes the adoption of pain assessment scales in critical patients, such as the Behavioral Pain Scale, as well as the use of protocols for analgesia management, and con-sequently improve the quality of care and patient’s recovery.


INTRODUCTION
The sensation of pain is essential for life.Its perception is the result of multidimensional and personal experiences in the face of the various stimuli that result or not in tissue injury 1 .Therefore, the protocols of care recommend the evaluation of pain by health professionals during the assistance.In individuals who verbalized and have preserved cognition, pain measurement can be more easily reported because the individual is able to describe the pain he/she feels 2 .However, in critically ill patients who are under adverse conditions that prevent them from verbalizing the presence or absence of pain 3,4 either by changes in the level of consciousness, the effects of sedative agents and/or the use of mechanical ventilation 5 , the measurement of pain can only be indirect.Some studies report that the observation of changes in physiological parameters may be a quick and simple method to infer pain [6][7][8] .However, the use of physiological data alone is debatable, since several factors such as fear, anxiety and psychological stressors can influence this measurement 6,7,9 .Also, the absence of changes in the vital signs does not necessarily indicate the absence of pain 10 .
As the inability to report pain does not deny its existence and does not discard the right to adequate treatment 11 , when it is impossible to obtain the patient's self-report about the pain it is recommended to use observational scales 12 that are based on the individual's physiological parameters and body expressions.Among the available scales to measure the pain in non-responsive patients, the most used by the health services is the Behavioral Pain Scale (BPS) because it is highly accurate and easy to apply to patients with severe pain 13,14 .Knowing the level of pain of patients, whether critical or not, is essential to optimize comfort and minimize suffering 15 .In addition, effective and adequate pain control is associated with a reduction in mechanical ventilation time, shorter patient length of stay, and lower morbidity and mortality rates in intensive care units (ICU) 16 .However, despite these benefits, pain assessment has been performed inadequately (or not performed) in some health services that provide care to critical patients, making it difficult to manage pain in these patients adequately 17,18 .
Given the above, the present study aimed to evaluate the pain and the use of analgesia in critically ill patients admitted to the emergency and intensive care services of a public reference institution in southwest Bahia.

METHODS
This is a cross-sectional and descriptive study with a quantitative approach, referring to critically ill patients admitted to the red area of the emergency room or to one of the two ICUs of a public reference hospital in Vitória da Conquista, Bahia, Brazil, between April and July 2017.This hospital is located 519 km from the capital, Salvador, and it is a reference for 73 smaller cities with a population of approximately 1.7 million inhabitants 19 .The sampling was non-probabilistic due to adequacy, with an estimated sample size of about 60 -65 patients.The calculation was based on an accuracy of 0.95±0.05 of the Cronbach's alpha coefficient for a scale with three subscales.All critical patients admitted during the study period, older than 18 years of age, of both genders, using mechanical ventilation, sedated and unarticulated, who were unable to report pain, and with a maximum stay time of 48h were included.Patients in neurological protection, quadriplegic, who had received a neuromuscular blocker, who had peripheral neuropathy or suspicion of brain death, were excluded.These exclusion criteria were used not to include patients whose diseases or drugs could compromise the expression of pain behaviors.After the written consent was signed by a responsible family member, duly trained research assistants, using the previously prepared data collection instrument, obtained the clinical and demographic data from the patient's medical records.The demographic information included age and gender.The clinical data included prior comorbidities, diagnosis, pharmacological prescription (use of analgesics and continuous infusion sedatives, given at regular intervals or if necessary).In addition, the information on the neurological assessment of each patient was obtained using the Glasgow Coma Scale, FOUR (Full Outline of Unresponsiveness), and the Richmond Agitation Sedation Scale (RASS), which are routinely used by professionals who work in the field of study.
The research assistants also collected information on the vital signs of each patient using a multimodal monitor during three moments of the study: at rest, during eye cleansing (EC) with gauze moistened in saline (considered a non-painful procedure) performed by the nursing technician and during the tracheal suction (TS) (considered a painful procedure) 20 performed by the assistant physiotherapist.These procedures are already part of the patient care routine, so no additional procedure is required.Simultaneously, the researchers applied the BPS validated in Brazil by Morete et al. 14 (Table 1).The BPS has a total of 12 descriptors, distributed in 3 items (1.Facial expression, 2. Upper limbs, 3. Adaptation to mechanical ventilation) 7,14 with results varying from 3 (absence of pain) to 12 (unbearable pain) 13 .A score >3 indicates the presence of pain and ≥5 indicates significant pain 21 .

Statistical analysis
All information obtained was coded and inserted into a database.Then, we performed an exploratory analysis of the data through the calculation of absolute and percentage simple frequencies for the categorical variables.The normal distribution of the data set was checked using the Kolmogorov-Smirnov test.The ANOVA test was used to check the fluctuation of the

RESULTS
Sixty-seven patients were included in the study.Each of them was evaluated in three moments: a) at rest, b) eye cleansing (EC) and c) tracheal suction (TS); totaling 201 observations (67 patients versus three observations each).Patients were predominantly male (47/70.1%),with a median age of 56 years (IIQ: 36-74), urban residents (51/76.1%),with reports of pre-existing comorbidity (41/61.2%).Most patients had a clinical diagnosis (49/73.1%),followed by trauma (18/26.9%)(Table 2).After collection, three groups of patients were identified based on the use of sedatives and analgesics: G1, patients undergoing sedation and analgesia; G2, only using analgesia; and G3, without sedation or analgesia.The majority of the patients were using analgesia associated with sedation (31/46.3%),and midazolam and fentanyl were the most commonly used drugs.Patients undergoing sedation and analgesia were evaluated by RASS and had an average score of -4.5±1.29.Eighteen (26.9%) were using analgesia alone, with fentanyl and dipyrone being the most prescribed analgesics.For these individuals, the neurological evaluation was performed using the FOUR scale in 28 patients with a mean of 6.2±3.63; and Glasgow coma scale in six patients with a mean of 4.3±2.16.
Sixteen patients (23.9%) were without analgesia or sedation.Of the 201 observations, in 70 (34.8%)patients had a score of ≥5 on BPS (Table 2).Table 3 shows the variation of the physiological parameters in the three evaluation moments for the three groups identified.Variation was observed in the three evaluation groups in all the physiological parameters with the interventions, except in temperature.
Table 4 shows the variation of the BPS scores.In all three groups of patients, a significant fluctuation was observed in all scores on the scale, especially between at rest and tracheal suction.parameters within the same group between the three moments of measurement of the values.The level of significance of the analysis was 5% (p<0.05).The data were analyzed using the SPSS software version 20.0.

DISCUSSION
Pain control, even in critically unarticulated patients, is vital.However, despite the technological advances in the care of critical patients in emergency or intensive care units, pain assessment and its proper management have been poorly addressed.This study found that even for patients undergoing analgesia and sedation, there was variation in the physiological parameters and BPS scores when they underwent painful procedures, especially TS, a routine technique in hospital units.This implies flaws in the process of pain assessment and the adequacy of analgesia in patients in intensive care.This ineffective control of pain is the result of a series of factors indicated in literature such as choosing an inadequate method of pain measurement, insufficient professional training or improper management of the pain without scientific evidence 9,17,22,23 .In addition, the resistance to change the routine of many professionals is also an important cause of inadequacies in the control of the pain in critical patients 17 .Some studies have also pointed out the lack of knowledge of the professionals about scales with considerable accuracy to assess pain in unarticulated patients 9,24 .However, since it is not possible to obtain the patient's verbal report about his/her pain, several observational scales have been recommended 12 .Among them, the BPS stands out for its high accuracy, easy application and for being adapted to the Brazilian reality 8,25 .
Like other studies conducted in Brazil 6,17,25 , this study showed that the BPS was adequate to evaluate pain in unarticulated patients.The comparison of the scale scores at rest and during eye cleansing, considered as a non-painful procedure, showed no variation.In this study, eye cleansing simulates other situations or procedures that are performed by the professional, such as dressing changes and measure temperature, but which do not necessarily correspond to painful stimuli 8,25,26 .However, the results showed a significant variation in the scale scores during TS -considered a painful process for the patient -and the highest scores of the instrument were observed, regardless of the form of analgesia (or absence).
Although the change in physiological parameters is not necessarily an indication to assess the pain in unarticulated patients, a significant variation in heart rate, respiratory rate, oxygen saturation, and systemic blood pressure was observed in the present study.These results are similar to those reported in other studies that analysed whether changes in BPS were accompanied by physiological changes in patients 13,17 .However, it is worth mentioning that the physiological parameters can be sensitive to several factors besides the presence of pain, such as fear, anxiety and psychological stressors 6,7,9 and other clinical conditions.Thus, the monitoring of physiological parameters alone as a way to assess pain has not been recommended 9,12,[27][28][29][30][31][32] , making necessary the use of properly validated scales/instruments and with proved accuracy, such as the BPS 9,33 .
The lowest BPS scores were found in group 1.However, even these patients showed significant variations in the scores of the pain scale, which implies that even in them the pain was being underestimated.Also, it was not possible to establish whether those in group 1 were experiencing less pain or if they were unable to present it 8 , as it would be unethical to conduct research involving the manipulation of sedation or analgesia levels since the patients would be exposed to a higher possibility of feeling pain 17 .Indeed, from the results of this study, it can be inferred that adjustments in pain control should be performed even in individuals with sedation and analgesia, which can be obtained with an accurate assessment of the pain in these patients.
Other studies show that behavioral indicators are more sensitive and present more adequate data than the hemodynamic parameters in the assessment of pain in critically ill patients 29,33,34 .However, the use of observational scales should not be considered as the most reliable or the only evaluation necessary since they do not reflect the intensity or the location of the pain, and they can be masked by deep sedation or the use of blocking agents.Likewise, these instruments should not replace the self-report of pain, when possible 25,34 .

CONCLUSION
Considering the observed aspects, we noticed an intense pain during TS, visualized by the elevation of the BPS scores in all the groups observed, confirming their responsiveness.Despite the alteration of the physiological parameters during the observation of groups 1 and 2, the same did not happen in group 3, indicating that the hemodynamic alterations should not be used as valid precursors to measure the pain.

Table 1 .
The Brazilian version of the Behavioral Pain Scale

Table 2 .
Demographic and clinical data of the patients included in the study.Vitória da Conquista, April to July/2017 BPS = Behavioral Pain Scale; TS = tracheal suction; FOUR = Full Outline of Unresponsiveness; RASS = Richmond Agitation Sedation Scale; IQI = interquartile interval; Group 1 = sedation and analgesia; Group 2 = analgesia; Group 3 = no sedation or analgesia.Categorical data presented quantitatively and percentage, numerical data on average and standard deviation.* 201 observations, ** 67 observations.

Table 3 .
Variation of the physiological parameters in the three moments of evaluation with the Behavioral Pain Scale in patients hospitalized in a regional hospital of Vitória da Conquista, Bahia, Brazil, 2017 Continue...

Table 3 .
Variation of the physiological parameters in the three moments of evaluation with the Behavioral Pain Scale in patients hospitalized in a regional hospital of Vitória da Conquista, Bahia, Brazil, 2017 -continuation Group 2 = analgesia; Group 3 = no sedation or analgesia; HR = heart rate; RR = respiratory rate; SpO 2 = peripheral oxygen saturation; SBP = systolic blood pressure; DBP = diastolic blood pressure; MAP = mean arterial pressure.Data expressed as mean and standard deviation.*ANOVA test.

Table 4 .
Behavioral Pain Scale scores during the three times of scale application in patients hospitalized in a regional hospital, Vitória da Conquista, Bahia, Brazil, 2017