CLINICAL BOTTOM LINE & AVAILABLE LITERATURE

...ABOUT THE COMPARISON OF DIFFERENT PAIN SCORES IN THE EMERGENCY DEPARTMENT: IS ONE BETTER?

1998

Berthier F, Potel G, Leconte P, Touze MD, Baron D. Comparative study of methods of measuring acute pain intensity in an ED. Am J Emerg Med 1998;16:132-6.

​

Study design

Prospective comparative cohort study

​

Setting           

ED of university hospital in France

Population  

Patients had to be 15 years or older with acute pain

Exclusion criteria: functional alteration; skin laceration with NRS pain less than 5; refusal to participate; inability to perform pain scores

Objectives 

To determine the most effective method for self-evaluation of acute pain intensity in the Emergency Department, comparing NRS, VRS and VAS

Methods   

Patients were divided in trauma and non-trauma patients and during 2 described periods of 3 months, patients were included at random (divided in patients with traumatic injuries and non-trauma patients). Patients were recruited by investigators, not involved in clinical care. In all patients first NRS was asked, then VRS and then VAS.

Outcomes

- Correlation and discriminant powers of NRS, VRS and VAS

- Ease of use of these pain scales

Results           

A total of 49076 patients presented in the ED during the year that the study ran (1993).

Of these patients, 290 were included. No baseline differences, however, trauma patients were significantly younger (35.7) than non-trauma patients (49.9).

​

NRS and VAS were closely correlated for trauma (r=0.795) and non-trauma (r=0.911) and for both under and over 65 years (r=0.851 and r=0.905, respectively).

Mean VAS and mean NRS were significantly different from one VRS description to another.

Non-response was equal among NRS and VRS.

​

Authors’ conclusions

NRS and VAS show good discriminative power for indicating acute pain intensity and are closely correlated both for trauma and for non-trauma patients.

The non-response rate of VAS was higher for trauma and discriminant power was greater for non-trauma patients. This was especially in discriminating between severe and unbearable pain (as did the other 2 scores).

​

The low non-response rate of NRS potentially makes this pain score preferable. 

​

Pros

Patients included not by treating staff

Limitations

Level of training research team not described. Patients who were excluded were not accounted for. Patients were not recruited consecutively.

Patients were asked pain scores, always in the same order, potentially increasing effective answering of the latter two pain scores (which were VRS and VAS). 

Level of evidence 4

2008

Daoust R, Beaulieu P, Manzini C, Chauny JM, Lavigne G. Estimation of pain intensity in emergency medicine: a validation study. Pain 2008;138:565-70.

​

Study design

Prospective validation study

Setting

Emergency Department in an academic hospital in Canada

Population

Adult (18 years and older) presenting to the ED with acute pain (<24 hours). Patients were recruited consecutively during 6 months.

Exclusion criteria: alteration of level of consciousness, inability to understand or use the different scales and the possibility that participation could delay urgent treatment.

Objectives

To estimate validity of 11-point VNRS and a plasticized VAS (pVAS)

​

Methods

Bland-Altman and Intra-Class Coefficient of Correlation were used for analysis and the paper VAS was utilized as gold standard. Treating physicians or research students included patients (all were trained). Pain was scored always in same order: VNRS, pVAS and VAS. VNRS was multiplied by 10 to compare with the other scores.

Limits of agreement between scales was set at 20 U.

​

Outcomes

Agreement and correlation of VNRS and pVAS with regard to VAS

​

Results

A total of 10231 patients were evaluated of whom 1218 were included. An additional 42 patients were excluded, because VAS was missing, leaving 1176 patients for analysis.

​

Patients who were not included were analysed regarding baseline criteria and were found to be not different from the included patients.

Regarding pain severity: most (34.4%) had VAS 60-79.9; 23.8% VAS 40-59.9; 20.2% 80-100 and 17.3% 20-39.9 and the remainder 0-19.9. Average pain was 64 U.

Sites of pain were abdomen in 31.5%; extremity in 22.4% and chest in 15.3%.

​

Intra-Class coefficient of correlation was excellent (>0.75) for both VNRS and pVAS, however limits of agreement were -16.5 to 18.3 (+/- 0.9) for pVAS versus VAS and -24.7 to 16.9 (+/- 1.1) for VNRS versus VAS.

There were small levels of bias along the spectrum of severity of pain, especially for VNRS.

​

Authors’ conclusions

Both pain scores showed excellent correlation with VAS and small level of bias.

VNRS is practical for use in the ED, but has its limitations with variable bias and wide limits of agreement, in the 40-100 U range it is mostly reliable and repeatable.

pVAS might be more practical and more valid than VNRS.

Pros

Trained research team recruited patients, who were included consecutively.

VAS was reviewed by a researcher blinded for the other scale results. Excluded patients were accounted for and analysed whether different than included patients.

​

Limitations

No sample size, crowding might have played in role in patient selection as might have the extremes of pain.

Level of evidence 2b

2010

Mohan H, Ryan J, Whelan B, Wakai A. The end of the line? The Visual Analogue Scale and Verbal Numerical Rating Scale as pain assessment tools in the emergency department. Emergency medicine journal : EMJ 2010;27:372-5.

Study design

single center prospective observational study

Setting

ED in Ireland

​

Population

Convenience sample of patients 18 years and older presenting with pain and speaking English.

Exclusion criteria: speaking English insufficiently; altered mental status or to acutely unwell to participate; patients who gave only one pain reading.

​

Objectives

- To test agreement between VAS and vNRS in the ED

- To determine whether agreement was altered by patient characteristics

​

Methods

Patients’ pain was recorded using 100mm VAS and with vNRS 0-10 and was reassessed every 30 minutes up to 2 hours. The order of pain scores was alternated between patients and was consistent within patients.

​

Outcomes

1. Agreement between VAS and vNRS

2. - Patient utility

    - Effect of demographics on pain scores

​

Results

A total of 123 patients were included, with 531 paired measurements of VAS and vNRS. Causes: trauma in 44.7% and location of pain: limb 50.4%; abdomen 24.4%; chest 8.9%; headache 5.7%; orofacial 4.9%; back 4.1% and testicular 1.6%.

Correlation between VAS and vNRS was significant (rs=0.93 with 95%CI 0.92-0.94). Variability was tested using Bland-Altman method: difference between means VAS – vNRS was -7.1mm (95%CI -7.9 to -6.2). Differences between VAS and vNRS were significant for educational level, gender and increasing age.

​

Majority (59.4%) of patients found that using a pain scale helped them describe pain severity; 17.9% found VAS easier and 35% found vNRS easier; 47.2% no difference.

​

The minimum clinically significant difference for VAS was 11.9mm (95%CI 9.7 to 14.9); for vNRS 1.25 (95%CI 1.05 to 1.45). MCSD not influenced by patient demographics.

​

Authors’ conclusions

There is an almost linear correlation between both pain scores, however, their use is not interchangeable as there is no perfect agreement. Age, gender and level of education affected this agreement, but did not affect the individual measurements or the MCSDs. Therefore self-reported pain does not differ based on age, gender, cause or level of education. vNRS is a good alternative for VAS in the ED.

​

Limitations

The study population was a convenience sample

Level of evidence 4

2015

Göransson KE, Heilborn U, Selberg J, von Scheele S, Djärv T. Pain rating in the ED-a comparison between 2 scales in a Swedish hospital. Am J Emerg Med 2015;33:419-22.

Study design

Cross-sectional hospital based cohort study during approximately 6 weeks

Setting

The ED of a level 1 trauma center and receiving 80000 patients annually

​

Population

Age older than 18 years with one of the following reasons of presentation: chest pain; abdominal pain; orthopedic related chief complaint.

Exclusion criteria: not able to give IC, no understanding of oral or written info due to cognitive reasons or linguistic difficulties; participation leading to delay in treatment (in patients with highest triage level)

​

Objectives

- To investigate the distribution and correlations between values on NRS and VAS

- To assess patients’ preferences of scale

​

Methods

Patients were recruited on weekdays between 8AM and 9PM by research assistants during a study period of six weeks. Patients rated their pain using VAS and NRS and a study-specific questionnaire was used. Patients were randomized to use VAS or NRS first. Scatter-plots and Bland-Altman analysis were used. Each patient rated their pain only once.

​

Results

403 patients were eligible; 9 not inclusion criteria, 84 exclusion criteria and 80 were missed due to other patients at that specific moment, leaving 230 patients for participation, but 13 declined without reasons: 217 included for analysis.

38% had abdominal pain; 36% orthopedic related pain and 26% chest pain.

Initial pain scores were correlated with a difference of 0.41. However, only the results of the NRS was presented: orthopedic median 6; abdomen median 5 and chest median 3.

​

61% found NRS easier to use than VAS (22%), p<0.001

53% found that NRS better reflected their pain (vs 26% VAS; p<0.01)

71% preferred to use NRS and 21% VAS (p<0.001).

No differences regarding age, sex of cause of pain

​

Authors’ conclusions

Both VAS and NRS pain scales had similar distribution and correlated strongly.

Patients preferred to use NRS as this was easier to use and better reflected their pain.

​

Pros

Randomization which scale to use first (only study that did this); solid statistical analysis

​

Limitations

No sample size, non-consecutive patient recruitment. Pain scores on admission were relatively low and we know that in moderate pain both scales score the same (and on extremes they deviate from each other). Low scores on admission might be due to high proportion of exclusion of high-priority patients.

Level of evidence 4

2015

Bahreini M, Jalili M, Moradi-Lakeh M. A comparison of three self-report pain scales in adults with acute pain. J Emerg Med 2015;48:10-8.

​

Study design

Cross sectional study

​

Setting

Emergency Departments of two university hospitals

Population

A convenience sample of adult patients 18 years or older presenting with acute pain. Exclusion criteria: altered mental status, cognitive impairment, unable to understand explanation and commands in Persian language and patients whose pain was relieved at study onset, noncooperative patients and unwilling to give IC, visual problems and motor abnormalities

​

Objectives

- To determine the correlation and agreement between pain scores using 3 different pain scales (NRS, VAS and CAS) in adult patients presenting to the ED with acute pain

- To to find out which pain scale was preferred by patients

​

Methods

Authors write that patients were included during all shifts, but later on, a convenience sample was described. Patients approached and included by an independent researcher. Patients assessed pain using 3 different pain scales: VAS, CAS and NRS and the order was randomly varied among patients. Correlation was measured with Spearman and agreement with Bland-Altman with maximum limit of agreement of 20mm

​

Outcomes

Equivalence of methods of pain measurement in adult patients who are in acutely painful conditions

​

Results

242 patients approached of which 157 met inclusion criteria and 7 refused. These were analysed (but 8 were LTFU).

Included patients were relatively young (mean age 32.3) with pain at extremities in 32%; abdomen in 30.7%; torso in 18.7%; head/neck in 17.3% and generalized pain in 1.3%.

At baseline median pain scores were NRS 5.0 (4-8); VAS 5.1 (2.6-7.6) and CAS 6.0 (3.7-8.0). Pearman's p showed strong correlation with correlation coefficients between NRS-CAS 0.95; NRS-VAS 0.94 and CAS-VAS 0.94 (p<0.001).

​

Although correlated well, the scores were not equivalent. Bland-Altman showed mean differences of -0.4 between VAS/CAS; 0.3 VAS/NRS and -0.1 CAS/NRS.

​

Patient satisfaction was not different in 38% between the scales and different in 62%, of which 32% preferred CAS. The illiterate group (16.6%): 60% no difference and none chose VAS as most preferable.

​

Authors’ conclusions

These three scores are interchangeable due to their strong correlation and their close agreement for use in the ED.

​

Pros

Very solid statistical methodology; random assignment of order of applied pain scores; patients recruited by independent researcher

​

Limitations

Convenience sample of patients (bias minimized by including patients during all shifts). Causes of pain not described.

Level of evidence 2b

2017

Leigheb M, Sabbatini M, Baldrighi M, et al. Prospective analysis of pain and pain management in an emergency department. Acta Biomed 2017;88:19-30.

​

Study design

Prospective observational study

​

Setting

ED in Italy

​

Population

Patients with pain

Exclusion criteria: under the age of 18; trauma more than 24 hours prior to presentation; cognitive deficits; language barrier; unable to describe intensity of pain; gynaecologic pain; life-threatening injuries

​

Objectives

To evaluate:

- intensity and location of pain experienced by patients in the ED

- time to analgesia

- patient satisfaction

​

Methods

Patients were only included in case the data-collection nurse was on duty, non-consecutively

​

Outcomes

1. Pain intensity measured with VAS and NRS on admission and disposition

2. Patient satisfaction measured with  a questionnaire

​

Results

137 patients were included, patients not included were not accounted for. Most pain from musculoskeletal, abdomen, chest and headache.

​

Median pain score at admission: NRS 8 and VAS 7. At discharge: NRS 7 and VAS 5.

These differences were not statistically significant.

Changes in both scores were not significant after pain treatment: both scores showed a median reduction of -2.

When pain was stratified into categories of mild, moderate and severe; then severe pain was present at higher frequency when using NRS compared to VAS (p<0.0001).

The opposite was true for mild and moderate pain.

​

Authors’ conclusions

Pain intensity decreased during the ED stay. NRS pain scores were higher than VAS pain scores at admission, but not at disposition.

​

Pros

Treating health care staff were blinded for the study purpose; power analysis was performed

​

Limitations

Inclusion criteria not clearly defined; convenience sample; results regarding differences in scores and causes of pain were not clearly and supposedly inconsistently reported.

Level of evidence 4

Remarks

These results suggest that utilizing the NRS overestimates pain when severe and underestimates pain when pain is mild.