Pain Assessment Tools
In 2001, Bijur et al tested the reliability of the VAS for the measurement of acute pain in the ED. 62 Patients were asked to rate their pain on a 100 mm marked scale with extremes labeled as least possible and worst possible pain at 1 minute intervals, blinded to their first response. This was a prospective, convenience sample of 96 patients. They found that 90% of ratings were within 9 mm of one another, concluding that the VAS is highly reliable for the measurement of acute pain in the ED. Extrapolating these findings further, changes in pain intensity less than 10 mm may represent an error of the method and should be interpreted with caution.
Several studies have tried to quantify the clinically significant change in the VAS in an attempt to assess pain relief. "The minimum clinically significant difference in pain was defined as the mean change in pain associated with a rating of Ďa little more pain' or Ďa little less pain'." 63 An initial study by Todd et al in 1996 found that 13 mm represents the minimally clinically significant difference. 64 This was validated by a study by Gallagher in 2001. 63 One additional study by Kelly in 1998 found 9 mm to be the clinically significant number; 35 however, the confidence interval overlapped with the prior two studies.
The Visual Analog Scale (VAS) has been used extensively in clinical research and has been found to be reliable and valid in the ED setting. 64 62 However, it requires the cognitive ability to translate a patient's level of pain into a distance measure and adequate levels of visual acuity and motor function. The Numerical Rating Scale (NRS) can be administered verbally and is a familiar clinical tool. In 2003, Bijur et al compared the NRS to the VAS in the evaluation of acute pain in the ED. They found that the two measures strongly correlated with each other and concluded that the NRS could be substituted for the VAS. 65 A Verbal Descriptor Scale (VDS) is useful for those patients unable to rate their pain on the NRS. NRS and VDS were validated in an ED population by Tanabe. 10
In 1998, Berthier et al compared the VAS, the Verbal Rating Scale (VRS) and the NRS in measuring acute pain intensity in the ED. 66 They found that the VAS and NRS closely correlated in both trauma and non-trauma patients. The VRS is less abstract, but had an 11% non-response rate and poor differentiation between severe and unbearable pain in trauma patients. The VAS had a high non-response rate (39%) in trauma patients. The NRS had a very low non-response rate, required only a verbal response, and was used successfully in 96% of patients. The authors concluded that the NRS is the preferable form of self-evaluation of pain in the ED.66
These various scales for pain assessment have been validated in the ED setting. Equally important is their value in assessing clinical improvement. An improvement of 13 mm on the VAS, as discussed above, has been shown to be the minimum change necessary to correlate to clinical significance. Likewise, a change of 1.39 on the NRS has been shown to correlate to the same minimum clinically significant change. In conclusion, while both scales are valuable in the initial assessment of pain, they are also valuable in the assessment of pain relief. 67
In 2005, Fosnocht et al reported that, while there was a moderate correlation between the change in VAS and a verbal descriptor of pain, there was a wide variability of change in the VAS so they cautioned against the use of a change in VAS as an indicator of pain relief for individual patients. 68 The VAS is not a true interval scale and a change in pain intensity at the lower range may not correlate with a similar change in the higher pain range. This was further evaluated by Bird and Dickson who reported that clinically significant changes in pain were not uniform along the entire VAS: "Patients with greater pain required a larger change in VAS score to effect a clinically significant reduction or increase in perceived pain." 69
Despite the availability of different assessment tools, it has been documented that they are underutilized. In 1995, Ducharme and Barber performed a prospective, blinded observational study on pain assessment in the ED. 9 They found that none of the patients in the study had their level of pain documented in the chart and that there was no use of an objective pain scale. Ten out of 42 patients received some intervention for pain, but only four received an analgesic. Patients in severe pain waited an average of 66 minutes to be seen, then an additional 74 minutes for medication. Tanabe found that pain scales were not routinely used, unless the patient was complaining of chest pain. 10
Various pain assessment tools are available for children who are old enough to communicate. Pain scales have been developed using numbers, colors, and facial expressions, see Figure 3. 70 In preverbal children, several pain scales have been validated. The CHEOPS (Children's Hospital of Eastern Ontario Pain Scale) 71 is a well-validated tool for the assessment of pain in children, see Figure 4. It was initially developed for postsurgical patients, but has been used broadly since.72
The complete assessment includes a rating on verbalization and complaints, but the other five behaviors are appropriate for preverbal children. It quantifies pain by rating five behaviors: crying, facial expression, verbalization, activity of torso, touching, and extremity response, such as drawing up legs or squirming.
The assessment of pain in children unable to communicate, either because of age or cognitive impairment, is still more challenging for the clinician. The Non-communicating Children's Pain Checklist assesses pain by scoring multiple parameters: vocal, social interaction, facial expressions, level of activity, body movements, or guarding and physiologic signs such as shivering, sweating, or breath holding. 73 As in adults, self-reporting is the preferred method for assessment of pain in children. While even very young children can give some indication as to the pain they are experiencing, it is important to assess for competence, especially between the ages of three and seven, to describe their pain accurately. Observation should always accompany self-reporting and may be the necessary alternative when selfreport is unavailable or unreliable. 32
The American Geriatric Society (AGS) has published broad guidelines on the issues of pain management in older adults. Older patients, like all other patients, should self-report their pain and needs which should be considered the most accurate assessment of pain. Clinical assessments and surrogate reports should only be used when patients are unable to express their needs themselves. Communication difficulties can exist, especially in elderly patients who suffer from dementia or limitations from cerebral infarcts. Therefore, various descriptions for pain should be used, i.e., aching, burning, discomfort, etc., when questioning patients. In non-vocal patients, indicators of pain, including moaning and crying, should be taken as indicators of pain. 74
The NRS has been in use in the ED for many years for the evaluation of chest pain. It is easy to administer and only requires a verbal response from the patient. Additionally, it does not require a special template to be reproduced on the chart. Given these facts and that it has been validated, 66, 67 it is the authors' preferred tool for assessment of pain in the ED. However, there are still some patients who will be unable to assign a number to their pain. Some patients require verbal prompting or even pictorial facial expressions to aid in their description of their pain. The key point is not to advocate one scale over the other, but to emphasize their consistent use in the ED and provide alternatives when a preferred method of assessment is not feasible for a given clinical encounter.