ORIGINAL ARTICLE

Theoretical Repeatability Coefficient of a 100 mm Visual Analog Scale in Children Benoit Bailey, MD, MSc,* Julie Gabbay, MD,* Raoul Daoust, MD, MSc,w and Jocelyn Gravel, MD, MSc*

Objective: The aim of this study was to determine the theoretical repeatability coefficient of a 100 mm visual analog scale (VAS) in children in different circumstances. Methods: A prospective cohort study was conducted using a convenience sample of patients aged 8 to 17 years presenting to a pediatric emergency department. Patients were asked how they liked a variety of foods (surrogate for stable pain stimulus) on a 100 mm VAS with 4 different sets of questions repeated 3 times: set 1—questions at 3-minute intervals with no specific instructions other than how to complete the VAS; set 2—same format as set 1 except for the duration of the interval (1 min); set 3—same as set 1 except patients were asked to remember their answers; set 4—same as set 1 except patients were shown their previous answers. For each, the repeatability coefficient of the VAS was determined. Results: A total of 100 patients aged 12.1 ± 2.4 years were enrolled. The repeatability coefficient for the questions asked at the 3-minute interval was 12 mm, whereas it was 8 mm when asked at the 1minute interval. When asked to remember their previous answers or to reproduce them, the repeatability coefficients for the questions were 7 and 6 mm, respectively. Discussion: The conditions of the assessments influence the repeatability coefficient of the VAS. Depending on different circumstances, the repeatability coefficient in children aged 8 to 17 years varies from 6 to 12 mm on a 100 mm VAS. Key Words: reliability, VAS, children

(Clin J Pain 2014;30:515–520)

T

he visual analog scale (VAS) is a common method to measure self-report pain intensity in clinical trials. Despite this, little is known about its reliability to measure acute pain. In adults, very few studies have addressed this issue and there is only one study addressing this in children.1–4 Reliability represents the reproducibility of a measure during different occasions and aims at minimizing sources of random error.5 The repeatability coefficient (RC) is a test-retest reliability measure that represents the maximum difference that is likely to occur between repeated measurements of the same outcome on a scale with a probability of 95%.6

Received for publication February 20, 2013; revised October 9, 2013; accepted August 2, 2013. From the *Department of Pediatrics, Division of Emergency Medicine, CHU Sainte-Justine; and wDepartment of Emergency Medicine, Hoˆpital du Sacre´-Coeur, Universite´ de Montre´al, Montre´al, QC. The authors declare no conflict of interest. Reprints: Benoit Bailey MD, MSc, Department of Pediatrics, Division of Emergency Medicine, CHU Ste-Justine, 3175 Chemin de la CoˆteSte-Catherine, Montre´al, QC H3T 1C5 (e-mail: benoit.bailey@ umontreal.ca). Copyright r 2013 by Lippincott Williams & Wilkins

Clin J Pain



Volume 30, Number 6, June 2014

In children, the RC was 12 mm for patients reporting that pain had not changed over a 3-minute interval, but it increased to 22 mm when patients for whom pain had changed over the assessment period were included.4 In adults, the RC of a 100 mm VAS was determined to be 20 mm when pain was assessed at 3-minute intervals in one study and 16 and 11 mm in 2 studies that assessed pain at 1minute intervals, all without knowing if pain had changed during the assessment period.1–3 Other factors could also modify this value. For example, to help patients quantify changes in pain intensity over time, certain researchers ask them to recall or show them their previous scores. However, the impact of recalling or seeing the previous scores on the RC is unknown. All these factors could impact results of studies depending on verbal cues and time intervals. Therefore, a better understanding of the reliability of the VAS is important to be known especially if we want to determine within what range a measurement represents the same score by an individual. In particular, the minimum clinically significant difference of a 100 mm VAS in children was previously determined to be 10 mm.7 It is unclear if the VAS is precise enough to conclude that a difference of 10 mm observed in an individual has any clinical meaning because of the imprecision of the VAS. Thus, to get a better understanding of the RC and the various factors that can influence it, we aimed to determine the theoretical RC of a 100 mm paper VAS under different circumstances in children aged 8 to 17 years. Assessments were carried out either at 3-minute or 1-minute intervals, when participants were asked to recall their scores or when they were shown their previous answers. We hypothesized that the RC for a stable stimulus would vary if assessed at 3-minute or 1-minute interval and if participants were shown or were asked to recall their initial score.

METHODS Study Design This was a prospective cohort study on a convenience sample of children presenting to a pediatric emergency department (ED). The institution review board approved the study. Written informed consent was obtained from a parent for all children. All children provided their verbal or written assent.

Study Setting and Population Patient enrollment took place in a pediatric ED at a tertiary care urban pediatric hospital with an annual census of >65,000 visits during daytime when a research assistant was present. All children between the ages of 8 and 17 years who presented to the ED and who spoke either French or English were eligible. The minimum age of 8 years was chosen because our experience suggests that children www.clinicalpain.com |

515

Bailey et al

younger than 8 years have difficulty using the VAS.8 We excluded patients with drug intoxication, altered level of consciousness, decreased visual acuity, or those who were clinically unstable, required admission to the intensive care unit, did not speak French or English, or had developmental delay.

Study Protocol and Measurements To insure that differences in VAS were solely explained by the tool’s reliability, participants were asked questions that were not expected to change over a short period of time (surrogate for a stable pain stimulus). Therefore, children were asked to rate how much they liked a different food on a 100 mm paper VAS. Participants were asked 4 different sets of questions 3 times in the same order.

Set 1 Patients were asked to grade how much they liked a specific food using the VAS. Three and 6 minutes later, they were asked to grade the same food in the same manner. Each time, no specific instructions other than how to complete the VAS were given, and patients did not have access to their previous answers.

Set 2 After the first set was over, participants were asked to grade a different food using the same VAS. One and 2 minutes later, they were asked to grade the same food the same way. Each time, no specific instructions other than how to complete the VAS were given, and patients did not have access to their previous answers.

Set 3 After the second set was over, they were then asked to grade another food using the same VAS and to remember where they had drawn their lines because they were going to be asked the same question a second and third time. Three and 6 minutes later, they were asked the same question and were asked to remember where they had drawn their line. Each time, patients did not have access to their previous answers.

Set 4 Finally, participants were asked to grade another food using the same VAS and to remember where they had drawn their line because they were going to be asked the same question a second and third time. Three and 6 minutes later, they were asked the same question and were shown their previous responses. The verbatim used was the following: on the line below, show me how you like X, here is if you hate it, here if it is your favorite food. X was either vanilla ice cream, chocolate ice cream, apples, green apples, strawberry, raspberry, celery, or cucumber. All set were completed within

Theoretical repeatability coefficient of a 100 mm visual analog scale in children.

The aim of this study was to determine the theoretical repeatability coefficient of a 100 mm visual analog scale (VAS) in children in different circum...
189KB Sizes 0 Downloads 0 Views