crit care nurse 2013 stites 68 78

7/29/2019 Crit Care Nurse 2013 Stites 68 78

1/13

http://ccn.aacnjournals.org/cgi/external_ref?link_type=PERMISSIONDIRECTPersonal use only. For copyright permission information:Published online http://www.cconline.org 2013 American Association of Critical-Care Nurses

doi: 10.4037/ccn20138042013;33:68-78Crit Care NurseMindy StitesObservational Pain Scales in Critically Ill Adults

http://ccn.aacnjournals.org/subscriptions/Subscription Information

http://ccn.aacnjournals.org/misc/ifora.xhtmlInformation for authors

http://www.editorialmanager.com/ccnSubmit a manuscript

http://ccn.aacnjournals.org/subscriptions/etoc.xhtmlEmail alerts

by AACN. All rights reserved. 2013ext. 532. Fax: (949) 362-2049. Copyright

101 Columbia, Aliso Viejo, CA 92656. Telephone: (800) 899-1712, (949) 362-2050,Association of Critical-Care Nurses, published bi-monthly by The InnoVision GroupCritical Care Nurse is the official peer-reviewed clinical journal of the American

by guest on July 27, 2013ccn.aacnjournals.orgDownloaded from
http://ccn.aacnjournals.org/cgi/external_ref?link_type=PERMISSIONDIRECThttp://ccn.aacnjournals.org/cgi/external_ref?link_type=PERMISSIONDIRECThttp://ccn.aacnjournals.org/subscriptions/http://ccn.aacnjournals.org/subscriptions/http://ccn.aacnjournals.org/subscriptions/http://ccn.aacnjournals.org/misc/ifora.xhtmlhttp://ccn.aacnjournals.org/misc/ifora.xhtmlhttp://ccn.aacnjournals.org/misc/ifora.xhtmlhttp://www.editorialmanager.com/ccnhttp://www.editorialmanager.com/ccnhttp://www.editorialmanager.com/ccnhttp://ccn.aacnjournals.org/subscriptions/etoc.xhtmlhttp://ccn.aacnjournals.org/subscriptions/etoc.xhtmlhttp://ccn.aacnjournals.org/subscriptions/etoc.xhtmlhttp://ccn.aacnjournals.org/http://ccn.aacnjournals.org/http://ccn.aacnjournals.org/http://ccn.aacnjournals.org/subscriptions/etoc.xhtmlhttp://www.editorialmanager.com/ccnhttp://ccn.aacnjournals.org/misc/ifora.xhtmlhttp://ccn.aacnjournals.org/subscriptions/http://ccn.aacnjournals.org/cgi/external_ref?link_type=PERMISSIONDIRECT


2/13

Pain is an important problem in the intensive care unit (ICU), and inadequate pain

assessment and management have been linked to increased morbidity and mortal-

ity.1

The physiological response to pain is almost universally adverse, causing poten-tially fatal unstable hemodynamic status, alterations in immune system functioning,hyperglycemia, and increased release of catecholamine, cortisol, and antidiuretic hor-

mones.2 Moreover, uncontrolled pain has been implicated in a variety of psychosocial effects,including depression, anxiety, delirium, posttraumatic stress disorder, and disorientation.3 Despitethe acknowledgment that pain is a common stressor in the ICU,4 high rates of uncontrolled pain incritically ill patients remain common.5 This situation can be attributed, in part, to circumstances,such as mechanical ventilation or unstable hemodynamic status, that preclude the assessment ofpain by self-report. Despite strong evidence that documentation of pain assessment improves painmanagement and decreases patients pain, no pain assessment instrument has been universallyrecommended for use in critically ill patients incapable of self-reporting.6

Observational Pain Scalesin Critically Ill AdultsMindy Stites, RN, MSN, APRN, ACNS-BC, CCNS, CCRN

2013 American Association of Critical-Care Nurses doi: http://dx.doi.org/10.4037/ccn2013804

Pain Management

Pain is a common and distressing symptom in critically ill patients. Uncontrolled pain places patients at riskfor numerous adverse psychological and physiological consequences, some of which may be life-threatening.A systematic assessment of pain is difficult in intensive care units because of the high percentage of patientswho are noncommunicative and unable to self-report pain. Several tools have been developed to identify objec-

tive measures of pain, but the best tool has yet to be identified. A comprehensive search on the reliability andvalidity of observational pain scales indicated that although the Critical-Care Pain Observation Tool was supe-rior to other tools in reliably detecting pain, pain assessment in individuals incapable of spontaneous neuro-muscular movements or in patients with concurrent conditions, such as chronic pain or delirium, remains anenigma. (Critical Care Nurse. 2013;33[3]:68-79)

This article has been designated for CNE credit. A closed-book, multiple-choice examination follows this article,which tests your knowledge of the following objectives:

1. Describe the impact of uncontrolled pain on critically ill patients2. Identify 2 validated pain scales for critically ill patients3. Describe instrument reliability and validity when considering pain scales for critically ill patients

CNE Continuing Nursing Education

68 CriticalCareNurse Vol 33, No. 3, JUNE 2013 www.ccnonline.org

http://ccn.aacnjournals.org/http://ccn.aacnjournals.org/http://ccn.aacnjournals.org/


3/13

Background and SignificanceThe pathophysiology of pain is complex, but involves

the release of inflammatory mediators after tissue injury.The mediators stimulate nociceptors, which transmitpain impulses to the dorsal horn of the spinal cord. Thepain impulses arrive at the somatosensory cortex, where

the localization and meaning of pain occur.7

Pain can beclassified as acute or chronic, depending on duration, andas peripheral or central, depending on location. Furtherclassifications are used to describe the pain source, suchas from injury to the skin (cutaneous), nerves (neuro-pathic), muscles and bones (somatic), or organs (visceral).8

Nearly 5 million patients are admitted to the ICU eachyear,9 and an estimated 71% of those patients rememberexperiencing pain during their stay.10 Painful procedures,such as turning and tracheal suctioning, are common inthe ICU and precipitate acute pain.11 In addition, manycritically ill patients have a history of chronic pain, whichcomplicates assessment and treatment.12 Pain is one ofthe most common symptoms in critically ill patients andis experienced by each patient in a unique manner. 13

The routine use of an appropriate assessment ofpain has been mandated by the Agency for HealthcareResearch and Quality and the Joint Commission.14 Pro-fessional organizations such as the American Associationof Critical-Care Nurses, the American College of ChestPhysicians, the Society for Critical Care Medicine, and

the American Society for Pain Management agree. Allof these organizations advocate for implementation ofstandardized pain assessment tools that include behav-ioral indicators in patients who are sedated and receivingmechanical ventilation and incapable of self-reportingor whose self-reports may be unreliable.15 Furthermore,studies have indicated that appropriate treatment ofpain and anxiety is associated with decreased duration ofmechanical ventilation (8 vs 11 days;P< .01),16 decreasedrate of nosocomial infections,4 and increased patient sat-isfaction with pain control.14 Additionally, systematic

pain assessment in critically ill patients can decreaseICU length of stay (13 vs 18 days; P< .01).16

OVID, PubMed, Medline, and CINAHL databaseswere used to search the literature. The key words pain,assessment, tool, instrument, scale, intensive care, criti-cal care, and critically ill were used. Articles published

between 2000 and 2010 and available in English wereincluded in the review. Additional inclusion criteriaincluded publication in a professional, scholarly, peer-reviewed nursing and health care journal. Articles thatfocused oninfants andchildren,addressedpain assess-ment outside the ICU, or focused on the evaluation ofchronic pain were excluded. A total of 1120 articles werefound; of these, 17 met the inclusion criteria. A fewstudies are included that did not meet every inclusioncriterion but did contain important information.

Review of the LiteratureThe review focuses on tools developed for and tested

in adult critical care patients. These instruments includethe Pain Assessment and Intervention Notation (PAIN)algorithm, the Nonverbal Pain Assessment Tool (NPAT),the Adult Nonverbal Pain Scale (NVPS), the Behavioral

Pain Scale (BPS), and the Critical-Care Pain ObservationTool (CPOT).

PAIN Algorithm

The PAIN algorithm,17 developed in 2001, consistsof 3 parts: pain assessment, the ability of the patient totolerate opioids, and guidelines for analgesic treatmentdecisions and documentation. The pain assessment sec-tion of the tool contains both behavioral (movement,facial cues, posturing) and physiological (increased heartrate, respiratory rate, and blood pressure and perspirationor pallor) dimensions.

The initial testing of the tool included a study sampleof 11 nurses who used the instrument to assess 31 post-operative patients in the ICU or postanesthesia care unit.17

The algorithm was helpful in providing a systematicapproach to pain assessment and guidance of analgesicadministration. However, 4 of the 11 nurses thought thatthe PAIN tool was not helpful: it was too long and cum-bersome to be used in everyday practice. Reliability

Author

Mindy Stites is a critical care clinical nurse specialist at the Universityof Kansas Hospital, Kansas City, Kansas, and a recent graduate of theUniversity of Missouri, Sinclair School of Nursing, Columbia, Missouri.

Corresponding author: Mindy Stites, RN, MSN, APRN, ACNS-BC, CCNS, CCRN,231 EdgewoodDr, Wellsville KS, 66092 (e-mail: [email protected]).

To purchase electronic or print reprints, contact The InnoVision Group, 101 Columbia,Aliso Viejo, CA 92656. Phone, (800) 899-1712 or (949) 362-2050 (ext 532); fax, (949)362-2049; e-mail, [email protected].

www.ccnonline.org CriticalCareNurse Vol 33, No. 3, JUNE 2013 69

Of the nearly 5 million patients admitted to the

ICU each year, an estimated 71% remember

experiencing pain during their stay.



4/13

and validity of the tool were not determined. Althoughthe researchers found that the instrument could be auseful training technique for beginning ICU nurses, thelength of the tool has limited its clinical usefulness, andno further testing has been done.17

Nonverbal Pain Assessment Tool

The NPAT consists of 5 domains: emotion, movement,verbal cues, facial cues, and positioning/guarding.10 Twoseparate scoring systems are provided on the instru-ment for use in both verbal and nonverbal patients.Scores range from 0 to 10 points, with higher numbersindicating higher severity of pain. The initial testing ofthe NPAT consisted of 3 phases of validity testing and 2revisions of the tool. In the third phase of validity test-ing, the investigators analyzed the correlation betweenthe NPAT score and the patients self-reported painscore in a sample of 50 general medical-surgical, post-operative patients.

Interrater reliability was high (concordance coefficient,

0.72; 95% confidence interval), and criterion validity(Table 1) was moderately strong when the NPAT was com-pared with the gold standard of self-report (concordancecoefficient, 0.66; 95% confidence interval).10 Althoughdesigned for noncommunicative critically ill patients,the NPAT has not been validated in this population ofpatients; the initial testing of the third phase was done inverbal medical-surgical patients. In addition, the inclu-sion of a verbal cues domain in a nonverbal pain

assessment tool is perplexing. Previous versions of theNPAT studied in the ICU showed weak validity (concor-dance correlation coefficient, 0.31; 95% confidence inter-val and concordance correlation coefficient, 0.21; 95%confidence interval, respectively). Since the initial testingin 2010, no further reliability or validity testing of theNPAT has been completed.19

Adult Nonverbal Pain Scale

The NVPS (Table 2) was based on the Faces, Legs,Activity, Cry, Consolability (FLACC) scale.20 Like the PAINalgorithm and the NPAT, the NVPS contains behavioraldimensions (facial expression, activity, and guarding)and physiological dimensions (heart rate, blood pressure,and respiratory rate) that are graded in severity. The finaldomain of the NVPS includes autonomic indicators suchas dilated pupils, diaphoresis, flushing, or pallor. Eachdomain is ranked from 0 to 2, with a total score between0 (no pain) and 10 (maximum pain).

The initial testing of the NVPS was conducted in a

burn trauma unit on 59 patients, with a total of 100paired assessments performed. Internal consistency washigh (=.78;P


5/13

In a study by Kabes et al,21 a revised version of theNVPS was compared with the original NVPS for reliabil-

ity and validity testing. The revised NVPS includes a newrespiratory category that replaces the physiological IIdimension of the original scale.22 The dimension includesan assessment of the amount of deviation from the base-line respiratory rate, as well as oxygen saturation as meas-ured by pulse oximetry and level of compliance with theventilator. Using a nonexperimental design, nurses in atrauma-surgical ICU assessed patients before, during, andat rest after a painful nursing procedure.

The levels of interrater reliability (90% in 94.7% withoriginal NVPS and 90.8% with the revised NVPS) andinternal consistency (Cronbach , 0.36 before, 0.72 dur-ing, 0.71 after) were high. Both the original and the revisedNVPS showed significant differences between ratingsbefore, during, and after the painful procedure (original,135.86;P30 mm Hg* HR>25/min

RR >20 above baseline,or 10% SpO2 severe

asynchrony withventilator




6/13

NVPS education sessions that wereprovided to the staff before the datacollection period.

Behavioral Pain Scale

The BPS26 (Table 3) was based

on the work of Puntillo et al27

thatidentified unique behaviors presentin patients undergoing a noxiousstimulus. The BPS is composed of 3observational items (facial expression,upper limbs, and compliance withventilation) that are scored from 1to 4, with higher numbers indicat-ing higher levels of discomfort. Thetotal BPS score can range from 3 (nopain) to 12 (most pain). The BPS was initially tested by

using a quasi-experimental design and a sample of 30patients receiving mechanical ventilation.26 The BPS scoreswere assessed at rest and then during a nonpainful pro-cedure (application of compression stockings or dressingchange of a central venous catheter) or during a nocicep-tive procedure (endotracheal suctioning or turning).Patients who had painful procedures had higher BPSscores (range, 4.6-5.2) than did patients who had non-painful procedures (BPS score range, 3.3-3.7). A test-retestprocedure was then performed in 31 cases, resulting in

BPS scores that increased from 3.0 to 3.3 at rest to 4.0to 4.8during apainfulprocedure,establish-

ing construct validity. Interrater agreement was high( = 0.94), and 24 of the 28 evaluators reported thatthe BPS was easy to use and took a mean of 2 to 5 min-utes to complete.

Assaoui et al28 tested the BPS in 30 general ICUpatients in a total of 360 observations. All patientsincluded in the study were sedated and were receivingmechanical ventilation. In paired, independent observa-tions, BPS scores were obtained at rest and during painfulprocedures (endotracheal suctioning or peripheral veincannulation). The BPS had good internal consistency(Cronbach = 0.72), and interrater agreement was high(intraclass correlation coefficient, 0.95). Validity wasestablished by the change in BPS scores (mean, 3.9; SD,

1.1 at rest and mean, 6.8; SD, 1.9 during procedures).

The elevated BPS scores at rest are interesting and mostlikely support hypotheses provided in previous researchthat critically ill patients may have moderate levels ofpain even at rest.29

In another study, Young et al30 used a descriptive,repeated-measures design to analyze the reliability andvalidity of the BPS. A total of 44 patients receivingmechanical ventilation whose hemodynamic status wasstable were assessed before and after a nonpainful (eyecare) and a painful (repositioning) procedure. The mean

BPS score increased significantly after the repositioningprocedure (score, 3.36 before vs 5.02 after). In contrast,the mean BPS score did not change significantly afterthe eye care procedure (score, 3.23 before vs 3.38 after),supporting discriminate validity of the BPS. However,interrater agreement varied widely, with good agreement(82%-91%) during rest states, but a marked decrease inagreement after a painful procedure (36%-46%). Theseresults indicate that raters more easily agreed on the levelof pain when the level was low and that variance betweenraters occurred when pain levels were high.

Ahlers et al31 obtained 371 independent determinationsof the BPS and NRS (when possible) in 113 critically illpatients. The BPS was completed for all patients receiv-ing mechanical ventilation; the NRS was either estimatedby the bedside nurse or obtained from the patient if thepatient was able to respond. Having a bedside nurseestimate the NRS score is a questionable practice, becausethis study31 indicated that the NRS score as reported bythe patient correlated with the score estimated by the

Table 3 Behavioral Pain Scale

Item

Facial expression

Upper limb movements

Compliance with mechanicalventilation

Scorea

1234

1234

12

34

Description

RelaxedPartially tightened (eg, brow lowering)Fully tightened (eg, eyelid closing)Grimacing

No movementPartially bentFully bent with finger flexionPermanently retracted

Tolerating movementCoughing but tolerating ventilation for

most of the timeFighting ventilatorUnable to control ventilation

a Score ranges from 3 (no pain) to 12 (maximum pain).


Although designed for noncommunicative

critically ill patients, the NPAT has not been

validated in this population of patients.



7/13

nurse only 73% of the time. Despite the problematic studydesign, the interrater reliability of the BPS was good( = 0.67; 95% CI, 0.54-0.80). Of note, NRS scores were0 in only 5% of the observations, whereas BPS scoreswere 3, indicating no pain, in 68%. These data illuminatean interesting question in the study by Payen et al26 inwhich the proportion of low BPS scores at rest was high.Possibly a BPS score of 3 is not truly representative of apain-free state as previously postulated. In addition, theBPS equates lack of body movement with a pain-free state,which conflicts with research32 that indicates a correlationbetween decreased movement and increased pain levels.

In a follow-up study, Ahlers et al33 examined the reli-

ability and validity of the BPS in both deeply sedatedpatients and patients undergoing conscious sedation fora painful procedure. Nurses made 175 observations in aconvenience sample of 80 medical-surgical ICU patientsat 4 times: at rest, during a nonpainful procedure, dur-ing a painful procedure, and after the painful procedure.The assessments were done independently, and commu-nicative patients were asked to provide a report of theirpain intensity by using the Verbal Rating Scale 4 29 if ableto do so after the BPS scoring was completed. Interrater

reliability of the BPS was excellent (=0.80 for deeplysedated patients and 0.83 for conscious sedated patients).Internal consistency was moderate (=0.63 in deeplysedated patients and 0.66 in conscious sedated patients).BPS scores were higher during painful procedures thanat rest in both deeply sedated patients (mean, 5.1; 95%CI, 4.8-5.5 vs mean, 3.4; 95% CI, 3.3-3.5) and conscioussedated patients (mean, 5.4; 95% CI, 4.9-5.9 vs mean, 3.8;95% CI, 3.5-4.1), thus establishing construct validity.The BPS and the Verbal Rating Scale 4 had a strong pos-itive correlation during painful procedures (r=0.67;

P< .001), indicating that the BPS is a valid tool to be usedin both deeply sedated and moderately sedated patients.

Critical-Care Pain Observation Tool

The CPOT34 (Table 4) was developed on the basis ofretrospective chart reviews to determine common painnotations and findings with focus groups of ICU clinicians.This instrument is designed for use in both intubated andnonintubated critical care patients. Four domainsfacialexpressions, movements, muscle tension, and ventilatorcomplianceare scored from 0 to 2; total scores rangefrom 0 (no pain) to 8 (most pain). The CPOT contains

Indicator

Facial expression

Body movements

Muscle tensionEvaluation by passive flexion and

extension of upper extremities

Compliance with the ventilator

(intubated patients)

OR

Vocalization (extubated patients)

Total, range

Description

No muscular tension observedPresence of frowning, brow lowering, orbit tightening,

and levator contractionAll of the above facial movements plus eyelid tightly

closed

Does not move at all (does not necessarily meanabsence of pain)

Slow, cautious movements, touching or rubbing thepain site, seeking attention through movements

Pulling tube, attempting to sit up, moving limbs/thrashing, not following commands, striking atstaff, trying to climb out of bed

No resistance to passive movementsResistance to passive movementsStrong resistance to passive movements, inability

to complete them

Alarms not activated, easy ventilation

Alarms stop spontaneouslyAsynchrony: blocking ventilation, alarms frequently

activated

Talking in normal tone or no soundSighing, moaningCrying out, sobbing

Score

Relaxed, neutral 0Tense 1

Grimacing 2

Absence of movements 0

Protection 1

Restlessness 2

Relaxed 0Tense, rigid 1Very tense or rigid 2

Tolerating ventilator or movement 0

Coughing but tolerating 1Fighting ventilator 2

Talking in normal tone or no sound 0Sighing, moaning 1Crying out, sobbing 2

0-8

Table 4 Critical-Care Pain Observation Tool




8/13

evidence-based descriptors that are operationally definedfor each domain, and the content validity index of allindicators was 0.88 to 1.0, according to an analysis of theresults of a questionnaire provided to physicians andcritical care nurses. The CPOT was originally developedin French and was tested in a convenience sample of 105

cardiac surgery patients.Glinas et al35 used a repeated-measures design andtrained data collectors to obtain the CPOT score of 105patients at 3 times: at rest, immediately after reposition-ing (nociceptive procedure), and at recovery (20 minutesafter repositioning). After completion of the assessmentby 2 observers, the patient was asked to indicate thepresence or absence of pain by nodding the head yes orno. Patients who had delirium (as indicated by the Con-fusion Assessment Method for the Intensive Care Unit)were excluded from the study. Interrater reliability was

moderate to high (weighted coefficient=0.52-0.88)when tested between the same 2 data collectors. Crite-rion validity was established by comparing the patientsself-report of pain and the concurrent CPOT score. Inpatients reporting pain, the mean CPOT score was 1.62to 3.65; in patients reporting no pain, the mean score was

0.49 to2.11.Althoughthe mean

CPOTscores were lower in the patients reporting no pain, theCPOT appeared to provide an overestimation of pain insome instances, especially in the immediate period aftera nociceptive procedure (mean, 2.11; SD, 0.90). Discrim-inate validity was supported by the finding that the CPOTscores were significantly higher (P< .001) during posi-tioning (t= -15.96) than at rest (t= -9.01). Although thestudy sample was large, interrater reliability and thehomogeneity of the sample were acknowledged limitations.

In a post hoc data analysis, Glinas et al36 evaluatedthe sensitivity and specificity of the CPOT. They foundthat the tool had a sensitivity of 86% and a specificity of78% during painful procedures. However, the sensitivitydecreased to 83% before a painful procedure and was 63%after the procedure. Specificity remained high at 83%and 97%, respectively. A cross-over observational designwas used to test the first English version of the CPOT ina group of 30 conscious and 25 unconscious general ICUpatients.37 Additional data were collected in this study,37

including physiological data and patients self-report ofpain via use of the Faces Pain Thermometer. Interraterreliability was acceptable (intraclass correlation coeffi-cients, 0.80-0.93) among the 51 nurses who collected thedata. Discriminate validity was established by detectionof elevations in heart rate and blood pressure that occurred

in accordance with elevated CPOT scores. Yet, physicalmotion is known to increase heart rate and blood pressureto compensate for increased oxygen demand, makingthis method of establishing validity problematic. How-ever, when patients self-reported pain values were com-pared with the observer-derived CPOT scores, the positivepredictive value of the CPOT was high (85.7%). All of thenurse respondents reported that the CPOT directiveswere clear and that the tool was easy to use. The major-ity of the nurses also reported that the CPOT was quickto use (78%) and that they would recommend use of the

CPOT routinely in practice (72.7%).38 Marmo and Fowler23

tested the CPOT in patients after heart surgery andfound that the tool had high reliability ( = 0.89). Theseresearchers23 were also the first to report the internal con-sistency of the CPOT (56%-100% agreement). The CPOTwas also included in the previously discussed study byWibbenmeyer et al,24 who reported a high internal con-sistency (Cronbach , 0.71) and good discriminate valid-ity (mean scale scores = 0.27 at rest to 0.56 after noxiousstimulation). As with the NVPS, Wibbenmeyer et al

reported that the interrater reliability of the CPOT waspoor (Pearson correlation coefficient, 0.63;P


9/13

was high (95%-100%), and agreement during turning wasacceptable (73%-91%). After implementation of the CPOT,interrater agreement improved to 86% to 100%. Whenpain assessment practices were analyzed by using descrip-tive statistics, reports of pain assessments were 3 to 4times more frequent after implementation than they

were before implementation. Interestingly, implementa-tion of the CPOT was associated with decreased frequencyof administration of sedatives and analgesics. Glinas etal40 provide 2 possible explanations: increased ability ofnursing staff to discern pain from other symptoms (suchas anxiety) or decreased number of trauma patients inthe group after implementation because of a change inthe centers trauma designation.

Analysis of the LiteratureThe published evidence indicates that more research

to find the ideal instrument to assess pain in critically illnonverbal adults is needed. Although multiple tools areavailable, few have been reliable and valid across a multi-tude of patient populations and settings (Table 5). How-ever, some tools clearly have been tested much morethan others have. The PAIN algorithm and NPAT haveeach been used in only a single study, and the findingsin both studies were of limited value. The original test-ing of the NVPS is of limited value because of the studysnonexperimental design and the use of the FLACC scale

as the gold standard for comparison. Although the NVPSis the only tool that includes dimensions of physiologicaldata, these indicators have been some of the least sensi-tive markers for the presence of pain.41

An overwhelming majority of the studies providesupport for the reliability and validity of both the BPSand the CPOT in detecting pain in noncommunicativecritically ill adults. The internal consistency and interraterreliability of both the BPS and the CPOT have been shownin more than 1 prospective study. However, comparisonsof study findings indicate that the CPOT has been moreinternally consistent than the BPS (Cronbach =0.71-0.89 vs Cronbach =0.63-0.66, respectively). Whetherthese data translate into a true implication for practiceremains to be seen; high levels of internal consistency maysimply indicate redundancy within the tool.42 Moreover,the range of statistical analyses used in the data analysisfor the CPOT was broader than that used for the BPS.

The results of construct, criterion, and discriminatevalidity analyses in a variety of patient populations have

been favorable for both the BPS and the CPOT; both toolsindicated a statistically significant increase in scores aftera painful procedure was performed. However, the stud-ies by Payen et al26 and Young et al30 also indicated anincrease in the BPS score after a nonnociceptive proce-dure, making the specificity of the instrument a concern.

Although the sensitivity and specificity of the CPOThave been investigated,32 no data on sensitivity or speci-ficity of the BPS have been reported. An interesting find-ing is the mean at-rest scores, which were elevated withboth tools (BPS score, 3.7; CPOT score, 0.27). These find-ings suggest baseline pain is present in a majority of crit-ically ill patients, consistent with the results of a previousstudy29 on characteristics of pain in the ICU.

An important finding of the validity testing of bothtools is a lack of strong correlation of the observationalpain tools with patients self-reports. The correlation was

especiallypoor whenpatientswere scoredas having nopain (BPSscore, 3; CPOT score, 0), and only a few patients had nopain on self-report scales. These findings suggest thatalthough scores increased during nociceptive proce-dures, indicating the presence of increased pain, the BPS

and the CPOT may not be useful tools for establishing theseverity of pain. Attempts to validate the tools as a methodof assessing pain severity have not been successful.41

Summary and Recommendations forFurther Research

In summary, internal consistency and interrater reli-ability in several studies have indicated the reliability ofboth the CPOT, and to a lesser extent, the BPS. Constructvalidity of both tools has been indicated in multiple pop-ulations of patients. However, the CPOT has been testedin both verbal and nonverbal patients in ICUs, a charac-teristic that may extend its applicability to practice. Inaddition, the evidence indicates that the BPS may lackspecificity in the detection of pain in noncommunicativecritically ill adults, a phenomenon not found in the test-ing of the CPOT. A comparison of the evidence support-ing the use of the BPS with that supporting use of theCPOT suggests that the CPOT may be a more specific,reliable, tool in nonverbal critically ill patients.


Although multiple observational pain scales

are available, few have been reliable and

valid across a multitude of patient popula-

tions and settings.



10/13

Research study

Behavioral Pain Scale (BPS)Payen et al,27 2001

Assaoui et al,28 2005

Young et al,30 2006

Ahlers et al,31 2008

Ahlers et al,33 2010

Critical-Care Pain Observation Tool (CPOT)Glinas et al,14 2004

Glinas et al,35 2006

Glinas and Johnson,37 2007

Marmo and Fowler,23 2010

Wibbenmeyer et al,24 2011

Vzquez et al,39 2011

Gelinas et al,40 2011

Nonverbal Pain Scale (NVPS)Odhner et al,20 2003

Kabes et al,21 2009 (revised version)

Marmo and Fowler,23 2010

Wibbenmeyer et al,24 2011

Reliability findings

Interrater reliability: 0.94 (Cronbach )

Internal consistency: 0.72 (Cronbach )Interrater agreement: 0.95 (ICC)

Interrater agreement: 82%-91% at rest,36%-46% during procedure

Interrater reliability: 0.67 (95% CI,0.54-0.80)

Interrater reliability: = 0.80-0.83

Internal consistency: Cronbach

=0.63-0.66

Interrater reliability: weighted coefficient=0.52-0.88

Interrater reliability: 0.80-0.93 (ICC)

Interrater reliability: 0.89 (Cronbach )Interrater agreement: 56%-100%

Interrater reliability: 0.63, P


11/13

Pain assessment in noncommunicative critically illadults remains a work in progress. By design, observa-tional pain scales require the presence of a spontaneous,neuromuscular-mediated physical response that can beobserved by a third party. Therefore, patients who arephysically unable to produce such a response, such as

patients who have quadriplegia or neuromuscular disor-ders or are receiving neuromuscular blocking agents,cannot be assessed with these tools. Patients with com-promised hemodynamic status, who account for a largeproportion of critically ill adults, are also at risk; thesepatients were often excluded in the aforementionedstudies.28,30,39 Therefore, the validity of observational painscales in patients with hemodynamic instability needsto be further explored.

Few investigations have addressed the assessmentand treatment of chronic pain in the ICU. Furthermore,

delirium and agitation are common in the ICU; 28% to87% of ICU patients experience one or both of these con-ditions at least once during their stay in the unit.13 Thepresence of delirium or agitation makes ascertaining apain level difficult. Currently, no pain assessment toolshave been validated in patients who are critically ill orare being treated with mechanical ventilation who areconcurrently delirious or agitated.

Finally, pain is a complex, personal experience thatcannot easily be adequately described and treated with-

out a patients self-report. Pain management providersoften use describing words such as burning or hotto indicate specific types of pain (eg, neuropathic pain).12

Without the benefit of that type of subjective data,patients may be at risk for receiving a less effective phar-macological agent (eg, opioids for the treatment of neu-ropathic pain).43 Future studies on strategies that promoteself-report of pain in critically ill patients who are sedatedand/or being treated with mechanical ventilation wouldcontribute much to the advancement of pain assessment.

ConclusionThe rates of uncontrolled pain in critically ill patients

remain unacceptably high. A systematic assessment ofpain should be done routinely, and self-report by thepatient should be the primary basis for pain evaluationwhenever possible. The routine assessment of pain withan observational pain assessment instrument can decreaseICU length of stay; decrease the duration of mechanicalventilation; and increase the satisfaction of patients,

patients family members, and health care providers.4,25

Of the available observational pain scales, the CPOT hasshown superior reliability and validity when used innonverbal critically ill adults. However, the CPOT shouldbe used cautiously in evaluating patients who have chronicpain or concurrent delirium. More research is indicated

in the assessment of pain in patients incapable of spon-taneous neuromuscular movement and in those withchronic pain. CCN

Financial DisclosuresNone reported.

References1. Shannon K, Bucknall T. Pain assessment in critical care: what we have

learnt from research.Intensive Crit Care Nurs. 2003;19:154-162.2. Puntillo K, Morris A, Thompson C, Stanik-Hutt J, White C, Wild L.

Pain behaviors observed during six common procedures: results fromthe Thunder Project II. Crit Care Med. 2004;32(2):421-427.

3. Jacobi J, Fraser G, Coursin D, et al. Clinical practice guidelines for thesustained use of sedatives and analgesics in the critically ill adult. Crit CareMed. 2002;30(1):119-141.

4. Chanques G, Jaber S, Barbotte E, et al. Impact of systematic evaluationof pain and agitation in an intensive care unit. Crit Care Med. 2006;34(6):1691-1699.

5. Campbell G, Happ M. Symptom identification in the chronically criti-cally ill.AACN Adv Crit Care. 2010;21(1):64-79.

6. Arif-Rahu M, Grap MJ. Facial expression and pain in the critically illnon-communicative patient: state of science review.Intensive Crit CareNurs. 2010;26(6):343-352

7. Porth C.Essentials of Pathophysiology: Concepts of Altered Health States.2nd ed. Philadelphia, PA: Lippincott Williams & Wilkins.

8. Siedlecki S. Pain and sedation. In: Carlson KK, ed.AACN AdvancedCritical Care Nursing. St Louis, MO: Elsevier; 2009:4 4-87.

9. Pronovost P, Goeschel C. Improving ICU care: it takes a team.Healthc Exec.2005;20(2):14-20.

10. Klein D, Dumpe M, Katz E, Bena J. Pain assessment in the intensivecare unit: development and psychometric testing of the Nonverbal PainAssessment Tool.Heart Lung. 2010;39(6):521-528.

11. Cade C. Clinical tools for the assessment of pain in sedated critically illadults.Nurs Crit Care. 2008;13(6):288-297.

12. Drew D, St. Marie B. Pain in critically ill patients with substance usedisorder or long-term opioid use for chronic pain. AACN Adv Crit Care.2011;22(3):238-254.

13. Puntillo K, Smith D, Arai S, Stotts N. Critical care nurses provide theirperspectives of patients symptoms in intensive care units.Heart Lung.2008;37(6):466-475.

14. Glinas C, Fortier M, Viens C, Fillion L, Puntillo K. Pain assessment andmanagement in critically ill intubated patients: a retrospective study.Am J Crit Care. 2004;13(2):126-135.

15. Pun B, Dunn J. The sedation of critically ill adults, I: assessment.Am JNurs. 2007;107(7):40-48.

16. Payen JF, Bosson JL, Chanques G, Mantz J, Labarere J; DOLOREA Inves-tigators. Pain assessment is associated with decreased duration of

Now that youve read the article, create or contribute to an online discussionabout this topic using eLetters. Just visit www.ccnonline.org and click Submit aresponse in either the full-text or PDF view of the article.

To learn more about using pain scales in critically ill patients, readPatient Satisfaction and Documentation of Pain Assessments andManagement After Implementing the Adult Nonverbal Pain Scaleby Topolovec-Vranic et al in theAmerican Journal of Critical Care,

July 2010;19:345-354. Available at www.ajcconline.org.




12/13

mechanical ventilation in the intensive care unit: a post hoc analysis ofthe DOLOREA study.Anesthesiology. 2009;111(6):1308-1316.

17. Puntillo K, Stannard D, Miakowski C, Kehrle K, Gleeson S. Use of a painassessment and intervention notation (P.A.I.N.) tool in critical carenursing practice: nurses evaluations.Heart Lung. 2002;31(4):303-314.

18. Polit DF, Beck CT.Nursing Research: Generating and Assessing Evidence forNursing Practice. Philadelphia, PA: Lippincott Williams & Wilkins; 2008.

19. Li D, Puntillo K, Miaskowski C. A review of objective pain measures foruse with critical care adult patients unable to self-report.J Pain. 2008;9(1):2-10.

20. Odhner M, Wegman D, Freeland N, Seinmetz A, Ingersoll G. Assessingpain control in nonverbal critically ill adults.Dimens Crit Care Nurs.2003;22(6):260-267.

21. Kabes A, Graves J, Norris J. Further validation of the Nonverbal PainScale in intensive care patients. Crit Care Nurse. 2009;29(1):59-66.

22. Wegman D. Tool for pain assessment. Crit Care Nurse. 2005;(25)1:14-15.23. Marmo L, Fowler S. Pain assessment tool in the critically ill post-open

heart surgery patient population.Pain Manag Nurs. 2010;11(3):134-140.24. Wibbenmeyer L, Sevier A, Liao J, et al. Evaluation of the usefulness of

two established assessment tools in a burn population.J Burn Care Res.2011;32(1):52-60.

25. Topolovec-Vranic J, Canzian S, Innis J, Pollmann-Mudryj MA, McFarlanAW, Baker AJ. Patient satisfaction and documentation of pain assessmentsand management after implementing the adult Nonverbal Pain Scale.Am J Crit Care. 2010;19:345-354.

26. Payen J, Bru O, Bosson J, et al. Assessing pain in critically ill sedatedpatients by using a behavioral pain scale. Crit Care Med. 2001;29(12):2258-2263.

27. Puntillo K, Miaskowski C, Kehrle K, Stannard D, Gleeson S, Nye P.Relationship between behavioral and physiological indicators of pain,critical care patients self-reports of pain, and opioid administration.Crit Care Med. 1997;25(7):1159-1166.

28. Assaoui Y, Zeggwagh AA, Zekraoui A, Abidi K, Abouqal R. Validationof a behavioral pain scale in critically ill, sedated, and mechanically ven-tilated patients.Anesth Analg. 2005;101(5):1470-1476.

29. Arroyo-Novoa CM, Figueroa-Ramos MI, Puntillo KA, et al. Pain related totracheal suctioning in awake acutely and critically ill adults: a descrip-tive study.Intensive Crit Care Nurs. 2008;24(1):20-27.

30. Young J, Siffleet J, Nikoletti S, Shaw T. Use of a behavioural pain scale toassess pain in ventilated, unconscious and/or sedated patients. IntensiveCrit Care Nurs. 2006;22(1):32-39.

31. Ahlers S, van Gulik L, van der Veen A, et al. Comparison of differentpain scoring systems in critically ill patients in a general ICU. Crit Care.2008;12(1):R15.

32. Blenkharn A, Faughnan S, Morgan A. Developing a pain assessmenttool for use by nurses in an adult intensive care unit.Intensive Crit CareNurs. 2002;1(6)8:332-341.

33. Ahlers S, van der Veen A, van Dijk M, Tibboel D, Knibbe C. The use ofthe Behavioral Pain Scale to assess pain in conscious sedated patients.Anesth Analg. 2010;110(1):127-133.

34. Glinas C, Fillion L, Puntillo K. Item selection and content validity ofthe Critical-Care Pain Observation Tool for non-verbal adults.J AdvNurs. 2008;65(1):203-216.

35. Glinas C, Fillion L, Puntillo K, Veins C, Fortier M. Validation of theCritical-Care Pain Observation Tool in adult patients.Am J Crit Care.2006;15(4):420-427.

36. Glinas C, Harel F, Fillion L, Puntillo K, Johnston C. Sensitivity andspecificity of the Critical-Care Pain Observation Tool for the detection ofpain in intubated adults after cardiac surgery.J Pain Symptom Manage.2009;37(1):59-67.

37. Glinas C, Johnston C. Pain assessment in the critically ill ventilated adult:validation of the Critical-Care Observational Tool and physiologic indi-

cators. Clin J Pain. 2007;23(6):497-505.38. Glinas C. Nurses evaluations of the feasibility and the clinical utilityof the Critical-Care Pain Observation Tool.Pain Manag Nurs. 2010;11(2):115-125

39. Vzquez M, Pardavila MI, Lucia M, Aguado Y, Margall M, Asiain MC.Pain assessment in turning procedures for patients with invasivemechanical ventilation.Nurs Crit Care. 2011;16(4):178-185.

40. Glinas C, Arbour C, Michaud C, Vaillant F, Desjardins S. Implementationof the Critical-Care Pain Observation Tool on pain assessment/manage-ment nursing practices in an intensive care unit with nonverbal criticallyill adults: a before and after study.Int J Nurs Stud. 2011;48(12):1495-1504.

41. Pasero C, McCaffery M. No self-report means no pain-intensity rating.Am J Nurs. 2005;105(10):50-53.

42. Boyle GJ. Does item homogeneity indicate internal consistency or itemredundancy in psychometric scales?Pers Individ Dif. 1991;12(3):291-294.

43. Bennett M, Attal N, Backonja M, et al. Using screening tools to identifyneuropathic pain. Pain. 2007;127(3):199-203.




13/13

CNE Test Test ID C1333: Observational Pain Scales in Critically Ill AdultsLearning objectives: 1. Describe the impact of uncontrolled pain on critically ill patients 2. Identify 2 validated pain scales for critically ill patients3. Describe instrument reliability and validity when considering pain scales for critically ill patients

Program evaluationYes No

Objective 1 was met q qObjective 2 was met q qObjective 3 was met q qContent was relevant to my

nursing practice q qMy expectations were met q qThis method of CNE is effective

for this content q qThe level of difficulty of this test was:

q easy q medium q difficultTo complete this program,

it took me hours/minutes.

1. Which of the following statements best describes the pain response?a. Inhibition of nociceptors transmits pain impulses to the ventral horn of the spinal cord.b. Chronic and acute pain pathophysiology is related to direct tissue injury transmitted

by peripheral nerves.c. After tissue injury, inflammatory mediators stimulate pain impulses to be sent to

the dorsal horn of the spinal cord.d. The inhibition of the somatosensory cortex results in the pain localization.

2. Which of the following statements describes the Adult Nonverbal Pain Scale(NVPS)?a. Both behavioral and physiologic data demonstrate validity of the NVPS.b. The domains of NVPS include behavioral, physiologic, and autonomic scoring

between 0 and 10.c. The basis of the NVPS scale is well-studied in adults and children.d. An NVPS score of 2 in the physiologic category indicates a change in heart rate of

crit care nurse 2013 stites 68 78

Documents