Critically ill patients in intensive care units (ICUs) frequently endure pain from both medical procedures and their underlying conditions. Procedures like chest tube removal, tracheal suctioning, wound care, patient repositioning, and arterial line insertion are known to be particularly painful. If acute pain in ICU patients is not properly managed, it can lead to significant short-term and long-term complications, including postoperative myocardial infarction, sleep disturbances, and posttraumatic stress disorder. Clinical practice guidelines emphasize the importance of individualized, goal-directed pain management, which begins with a systematic pain assessment using validated pain scales suitable for the patient’s level of consciousness.
However, assessing pain in critically ill patients presents unique challenges. Mechanical ventilation, severe illness, the effects of sedatives and analgesics, and altered levels of consciousness can all hinder a patient’s ability to self-report pain. In situations where self-reporting is not possible, validated behavioral pain scales are recommended for assessing pain in these vulnerable patients.
Two comprehensive systematic reviews have evaluated the psychometric properties of various pain assessment tools for ICU patients unable to communicate their pain verbally. Among these tools, the Critical-Care Pain Observation Tool (CPOT) and the Behavioral Pain Scale (BPS) emerged with the strongest quality assessments. Consequently, current clinical practice guidelines recommend both CPOT and BPS for pain assessment in nonverbal, critically ill adults. The CPOT is specifically designed for use in critically ill patients and comprises four behavioral domains: facial expression, body movements, muscle tension, and, for intubated patients, compliance with ventilation, or vocalization for those who are extubated. Each domain is scored from 0 to 2, resulting in a total score ranging from 0 (no pain) to 8 (maximum pain). A CPOT score greater than 2 is often indicative of pain during nociceptive procedures.
One notable limitation of the CPOT is the limited research on its effectiveness in critically ill patients experiencing delirium. Delirium is a common complication in the ICU, with incidence rates after cardiac surgery varying widely, from 3% to 55%. Overall, delirium affects approximately 30% to 50% of critically ill patients. Self-reporting of pain is particularly challenging in this population due to communication barriers, fluctuating levels of consciousness, and potentially atypical pain presentations. Therefore, validating behavioral pain scales like the CPOT in delirious critically ill patients is crucial.
Kanji and colleagues addressed this gap by investigating the validity and reliability of the CPOT in adult critically ill patients with delirium. Their study included 40 ICU patients diagnosed with delirium using the Confusion Assessment Method for the ICU (CAM-ICU). Patients unable to exhibit reliable physical responses to pain were excluded. The researchers rigorously assessed several key psychometric properties of the CPOT, including discriminant validity, interrater reliability, and internal consistency. Discriminant validity refers to a scale’s ability to differentiate between different states, such as pain and no pain. Pain scales are frequently evaluated by comparing scores during painful versus non-painful procedures. Interrater reliability measures the degree of agreement between different observers at different times. Kanji et al. used non-invasive blood pressure measurement as a non-painful procedure and repositioning, endotracheal suctioning, or dressing changes as painful procedures. They found a significant mean difference of 3.13 in CPOT scores between baseline and painful procedures (P<0.01). Kanji et al. concluded that their findings suggest the CPOT is a valid and reliable tool for detecting pain in non-comatose, delirious adult ICU patients.
Despite the meticulous design and execution of the Kanji et al. study, definitive conclusions regarding the use of CPOT in delirious patients remain premature. A key concern is the absence of data on delirium severity, delirium subtype, and the relationship between the Richmond Agitation-Sedation Scale (RASS) and CPOT scores. The DSM-V classifies delirium into three subtypes: hyperactive, hypoactive, and mixed. Hyperactive delirium is characterized by increased vigilance, restlessness, aggression, and heightened emotions, such as anger or anxiety. Hypoactive delirium presents with reduced alertness, minimal speech, and apathy. The mixed form involves alternating periods of hyperactive and hypoactive delirium. Peterson and colleagues categorized these subtypes based on RASS scores, defining hyperactive delirium as persistently positive RASS scores (+1 to +4).
In delirious patients, agitation, rather than pain, could inflate CPOT scores. The overlap between pain and agitation in delirious patients could lead to elevated CPOT scores that reflect agitation rather than actual pain. Furthermore, the influence of sedation levels needs further exploration. Kanji et al.’s study reported a median RASS of 0, with a range from -3 to +3, indicating the inclusion of patients exhibiting anxious or apprehensive movements (RASS +1), frequent non-purposeful movements or patient-ventilator dyssynchrony (RASS +2), or pulling at tubes and aggressive behavior towards staff (RASS +3). All four CPOT domains could be influenced by elevated RASS scores, potentially resulting in inappropriately high CPOT scores. Such inflated CPOT scores might lead to the unnecessary administration of analgesics when anti-delirium medication would be more appropriate. A recent study examining the validity of CPOT and BPS in a small subgroup of seven agitated patients (RASS +1) found no significant increases in CPOT scores between rest and painful procedures, and no difference between non-painful and painful procedures. The baseline CPOT score in this small agitated subgroup was also higher than in patients with lower RASS scores. Although based on a small sample, this finding suggests that the validity of CPOT in patients with hyperactive delirium and/or RASS > +1 requires further investigation.
In contrast to prior research, Kanji et al. assessed the interrater reliability of the individual CPOT domains rather than the interrater reliability across different procedures (painful vs. non-painful or rest). This approach is a limitation because it does not mirror routine ICU practice, where CPOT is used as a composite score across domains during various situations like tracheal suctioning or rest. Consequently, the interrater reliability of the CPOT in delirious patients during different clinical procedures remains uncertain.
Another aspect is that Kanji et al.’s study, like several preceding studies, involved investigators or physicians in pain assessments. However, in everyday clinical practice, nurses are the primary assessors of pain in the ICU. Bedside nurses, due to their extended patient contact time, may have a more nuanced understanding of patient reactions. Future research on interrater reliability should therefore include a broader range of raters, particularly nurses. Finally, the analysis of interrater reliability can be complex, as at least six versions of the Intraclass Correlation Coefficient (ICC) exist, and their application to the same data can yield different results. Kanji et al.’s study did not specify which ICC model was used, making it unclear whether the most appropriate model was employed.
In conclusion, the study by Kanji and colleagues represents a valuable initial step in validating the CPOT for pain assessment in critically ill patients with delirium. However, future assessments of CPOT’s interrater reliability must better reflect daily ICU practices. Larger studies involving more delirious patients, with sufficient representation of the three delirium subtypes and patients with RASS > +1, are essential before definitively concluding that the CPOT is a valid and reliable pain assessment tool in ventilated, critically ill patients with delirium.
References
[1] Puntillo KA, Morris AB, Thompson CL, et al. Pain behaviors observed during six common procedures: results from multicenter critical care pain observation tool validation study. Crit Care Med. 2004;32(2):421-427.
[2] Payen JF, Bosson JL, Chanques G, et al. Pain assessment is associated with decreased duration of mechanical ventilation in the intensive care unit: a post hoc analysis of population-based data. Crit Care Med. 2009;37(12):3418-3424.
[3] Gordon DB, Dahl J, Phillips P, et al. The use of ‘comfort’ as an analytic framework to operationalize patient-centered care. Appl Nurs Res. 2006;19(3):154-162.
[4] Watt CL, Katz J. Postoperative pain and persistent postsurgical pain: prevention and treatment. Pain Med. 2012;13(Suppl 2):S40-S51.
[5] van Gulik L, Vloet LC, Puntillo KA, van der Veen AJ, van de Pol LA, Bakker J. Pain assessment in sedated and mechanically ventilated critically ill patients: pain or agitation? Intensive Care Med. 2011;37(9):1417-1425.
[6] Barr J, Fraser GL, Puntillo K, et al. Clinical practice guidelines for the management of pain, agitation, and delirium in adult patients in the intensive care unit. Crit Care Med. 2013;41(1):263-306.
[7] Devlin JW, Herr DL, Puntillo KA, et al. Clinical practice guidelines for the management of pain, agitation, and delirium in adult patients in the intensive care unit. Crit Care Med. 2018;46(9):e825-e873.
[8] Gélinas C. Pain assessment in the critically ill ventilated adult: validation of the critical-care pain observation tool and physiological indicators. Clin J Pain. 2006;22(9):845-861.
[9] Chanques G, Payen JF, Mercier G, et al. Assessing pain in critically ill patients: BPS and CPOT. Intensive Care Med. 2009;35(9):1691-1699.
[10] Rijkenberg S, Stilma W, Endeman H, Otten RH, Honing M, van der Meer NJ. Pain assessment in critically ill patients: numeric rating scales versus behavioral pain scales. Intensive Care Med. 2011;37(6):947-953.
[11] Sessler CN, Gosnell MS, Grap MJ, et al. Richmond Agitation-Sedation Scale (RASS): validity and reliability in adult intensive care patient. Am J Respir Crit Care Med. 2002;166(10):1338-1344.
[12] Odhner M, Wegman D, Swanson E, et al. Pilot test of a behavioral pain assessment tool for critically ill, nonverbal patients. AACN Clin Issues. 2003;14(1):47-59.
[13] Gaudreau JD, Gagnon P, Roy MA, et al. Association between perioperative factors and early postoperative delirium after cardiac surgery. Crit Care Med. 2005;33(9):2001-2007.
[14] Ouimet MC, Kavanagh BP, Gottfried SB, Skrobik Y. Incidence, risk factors and consequences of ICU delirium. Intensive Care Med. 2007;33(1):66-80.
[15] Kanji S, Ziogas J, da Costa BR, et al. Validity and reliability of the Critical-Care Pain Observation Tool in adult, delirious, intensive care unit patients. J Crit Care. 2015;30(6):1201-1205.
[16] Koo TK, Li MY. A guideline of selecting and reporting intraclass correlation coefficients for reliability research. J Chiropr Med. 2016;15(2):155-163.
[17] Peterson JF, Pun BT, Dittus RS, et al. Delirium and its motoric subtypes: a study of 614 critically ill patients. J Am Geriatr Soc. 2012;60(5):903-909.
[18] De Jonghe B, Cook D, Appere-Lannoy C, et al. Pain assessment in the ICU patient: a prospective, multicenter study. Intensive Care Med. 2000;26(8):1040-1046.
[19] Shrout PE, Fleiss JL. Intraclass correlations: uses in assessing rater reliability. Psychol Bull. 1979;86(2):420-428.
