Reliability And Validity Of The Critical-Care Pain Observation Tool(CPOT)and The Behavioral Pain Scale(BPS)for Pain Assessment In Critically Ill Patients

BackgroundPain is widespread in critically ill patients. Most patients admitted in the Intensive Care Units (ICU) suffer bad memories, and pain is one of the main sufferings during their ICU stays. A variety of routine nursing procedures can cause pain for patients, such as drain tube removal, wound care, deep breathing, coughing exercises, suctioning and changing position, which are very commonly seen in ICU. Unrelieved pain may lead to multiple organ complications, as well as the prolonged chronic pain states and anxiety. Accurate assessment is the basis for the effective pain management. Patient’s self-report is the gold standard for pain assessment, but the majority of critically ill patients cannot report the experiencing pain by themselves because of unconsciousness, endotracheal intubation and/or other factors. Behavioral pain assessment tools may help recognize pain in patients unable to self-report, recommended by the American Society for Pain Management Nursing (ASPMN).Facial expression, body movement, body posture, verbal response and ventilator compliance are common indicators to assess pain in patients unable to self-report. Among the sveral existing behavioral pain assessment tools for non-verbal critically ill adult patients, the Critical-Care Pain Observation Tool (CPOT) and the Behavioral Pain Scale (BPS) are the most widely studied. The CPOT was developed by Gelinas from Canada in2006to assess pain in cardiac surgery patients with good reliability and validity. The original BPS was designed by Payen from France in2001to assess pain in intubated patients, and revised for non-intubated patients by Chanques as the BPS for the non-intubated (BPS-NI) in2009. The CPOT and the original BPS had been used in Chinese population from the mainland and Taiwan, but no research was reported regarding the BPS and the subscale of the BPS-NI for the non-intubaed.The objectives of this study were to examine the reliability and validity of the Chinese versions of the CPOT and the BPS in pain assessment across painful and non-painful nursing procedures in critically ill adult patients, and to provide effective tools for pain assessment in critically ill patients.Methods1ParticipantsParticipants were recruited from patients admitted in the ICU at a university-affiliated Level Three general hospital from October2013to January2014.Inclusion criteria were age≥18years, admission to the ICU more than24hours, suctioning and non-invasive blood pressure (NIBP) measurement included in routine nursing procedures. Patients were excluded if they had diagnosis with myasthenia gravis or severe brain injury, received muscle relaxant or cockstailytic, were in deep coma or deep sedation, or hemodynamic instability.2Instruments2.1Pain assessment toolsThe CPOT includes four behavioral categories: facial expression, body movements, muscle tension, and compliance with the ventilator for intubated patients or vocalization for non-intubated patients. Items in each category are scored from0to2, and the total score of the CPOT ranges from0to8.The original BPS and the BPS-NI have facial expressions, upper climbs, and compliance with mechanical ventilation for intubated patients or vocalization for the non-intubated. Each subscale is rated from1to4, and the possible score of the BPS ranges from3to12.2.2Critical Condition EvaluationThe Acute Physiology and Chronic Health Evaluation Ⅱ (APACHE Ⅱ) was revised from the APACHE Ⅰ by Knaus in1985. It is calculated from three different parts:a) Acute Physiology Score (APS); b) Age Points; and c) Chronic Health Points (CHP). The total score ranges from0to71, and higher score indicates heavier illness and greater risk of death.2.3Sedation AssessmentThe Ramsay Sedation Scale (RSS) was developed by Ramsay in1974, and is one of the most frequently used to assess the level of sedation of a hospitalized patient. The RSS defines sedation for six different levels from conscious state of level1to unresponsive level6.3Study Procedures3.1Medical Record ReviewThe demographic and clinical information such as ID, gender, age, whether with ventilation, or with analgesics, sedation level, and RSS and APACHE Ⅱ scores, were collected from the Electronic Medical Records.3.2Observational Pain AssessmentA trained ICU nurse was recruited as research assistant to assess patients’ behaviors simultaneously but independently with the researcher. The two procedures of suction and blood pressure measurement in routine nursing care were considered as painful (P) and non-painful (NP) stimulus, respectively. Observational pain assessments with the CPOT and the BPS were conducted at rest before and during the two procedures without any other interventions. Each patient was observed with the CPOT and the BPS four times, namely at rest before painful procedure (P0), during painful procedure (P1), at rest before non-painful procedure (NP0) and during non-painful procedure (NP1). Patients recieved the two procedures by different charge nurses in convenient order. To facilitate implementation and ensure accuracy, the interval between the two procedures was more than half hour but less than4hours.After each observation, patients’ behaviors were defined to different degrees according to each item’s ratings of the CPOT and the BPS. Patients’ demographic and clinical information and the scores of observational pain assessments with the two scales were recorded.4Statistical AnalysisThe internal consistency of the two behavioral scales was evaluated by Cronbach’s a coefficient. Pearson correlation coefficient was used to assess the test-retest reliability between the scores before the painful and non-painful procedures. The interrater reliability was measured by weighted Kappa value (κ) between the two raters. The discriminate validity of the CPOT and the BPS was tested by comparing the scores at rest before and during the painful and non-painful procedures with Mann-Whitney test.The Pearson correlation was analyzed between the scores of the CPOT and the BPS, and Bland-Altman Plot was used to make actual estimates of their agreement. The agreement of the two scales was calculated with different cutoff points to judge patient’s pain presence.Results1Demographic and Clinical CharacteristicsOne hundred and twenty-six patients were enrolled and152behavioral pain assessment were completed in117patients. A total of608assessments were obtained with the CPOT and the BPS.The mean age was57.97±16.31years old (ranged from18to89). Eighty-four patients were male (71.8%) and33female (28.2%). During pain assessments,65patients (55.6%) were intubated with mechanical ventilation,40patients (34.2%) received analgesics, and24(20.5%) received sedatives. The average score of RSS was3.58±0.97(ranged from2to5), and APACHE Ⅱ score was23.28±6.63(ranged from7to39). 2Reliability2.1Internal ConsistencyThe overall Cronbach’ a coefficient of the CPOT was0.795, and0.785and0.812for the intubated and the non-intubated patients, respectively. The overall Cronbach’ a coefficient of the BPS was0.791, and0.781and0.812for the intubated and the non-intubated patients, respectively.2.2Intraclass Correlation CoefficientThe intraclass Correlation Coefficient across the four assessments of the CPOT was0.859, and of the BPS was0.803.2.3Test-retest ReliabilityThe CPOT scores at rest before the painful and non-painful procedures were significantly related (P<0.001), and the Pearson correlation coefficient was0.942. The BPS scores at rest before the painful and non-painful procedures were significantly related (P<0.001) with the Pearson correlation coefficient0.922.2.4Interrater ReliabilityThe weighted k coefficient between the two raters for the CPOT was0.973, and0.985and0.955used for the intubated and the non-intubated patients, respectively. The weighted k coefficient of interrater reliability for the BPS was0.955, and0.939and0.977used for the intubated and the non-intubated patients, respectively.3ValidityPatient’s score of observational pain assessment with the CPOT was1.66±1.87(ranged from0to8), and with the BPS scale was4.70±2.10(ranged from3to11).The scores of the CPOT and the BPS during the painful procedures were both significantly higher than those during the non-painful procedures (Z=-14.352, P<0.001; Z=-14.440, P<0.001).The scores with the CPOT and the BPS during the painful procedures were both significantly higher than those at rest (Z=-14.183, P<0.001; Z=-14.468, P<0.001). There was no significant difference between the scores at rest and during the non-painful procedures both with the CPOT and the BPS (Z=-1.207, P>0.05; Z=-1.845, P>0.05). 4Comparison of the CPOT and the BPSPatient’s scores of the CPOT and the BPS were strongly correlated by Pearson correlation analysis (r=0.949, P<0.001). In the Bland-Altman plot,95%of data lie within the agreement limits.5Cutoff pointsThe correlation of the two scales were analyzed when the CPOT were cut off by2and3and the BPS by4,5,6and7, respectively, to judge the presence of patient’s pain. Agreement of the CPOT and the BPS were the highest (95.9%) when2was the cutoff point for the CPOT and5for the BPS. There were259(42.6%) assessments were judged painful by the CPOT and244(40.1%) painful by the BPS.ConclusionsThe CPOT and the BPS have good reliability and validity when used to assess pain in critically ill Chinese patients. The CPOT and the BPS are strongly correlated and interchangeable. The CPOT and the BPS could be good option for pain assessment in critical care patients.