Table 3 Key studies that support response scale selection used in PRO instruments based on select therapeutic areas

Sherbourne et al. 2012 [18]

Cross-sectional observational study, Indirect, C

5-point VRS

Develop asthma-specific quality of life items

A 5-point VRS for asthma quality of life assessment in adults was understood based on qualitative research with patients (cognitive interviews).

Liu et al. 2007 [2]

Cross-sectional observational study, Indirect, C

4-point VRS

Develop and validate the Childhood Asthma Control Test (C-ACT)

Children between the ages of 4 and 11 could understand and complete a 4-point VRS assisted by facial graphics.

Hagell and Knutsson 2013 [13]

Prospective, observational study, Direct, A

5-point VRS and VAS

Compare test-retest properties of 2 general health single item response formats among people with neurological disorders

Test-retest reliability assessments were similar for both formats, however patients preferred the VRS over the VAS format.

Preston et al. 2011 [19]

Cross-sectional observational study, Direct, A

4-point VRS and 5-point VRS

Evaluate the precision of the 5-point VRS response scale utilized in the emotional distress PROMIS item bank

The 5-point response options are not always equally spaced (i.e., do not meet the assumptions of an equal interval scale) and 4-point response categories were as precise as five.

Lasch et al. 2012 [30]

Cross-sectional observational study, Indirect, C

11-point NRS

Develop a content valid PRO measure for Major Depressive Disorder (MDD)

Cognitive interview demonstrated that an 11-point NRS was well understood and appropriate for evaluating concepts.

Hewlett et al. 2007 [31]

Review, Indirect, B

VAS and NRS

Systematic literature review to identify fatigue in rheumatoid arthritis scales; assess scale measurement properties

A VAS scale was the most frequently utilized scale to evaluate fatigue in rheumatoid arthritis and shows evidence of validity but there was no standardized VAS scale to evaluate fatigue in rheumatoid arthritis as scales were study specific. NRS used to evaluate fatigue in rheumatoid arthritis showed some evidence of construct validity but data on criterion validity, reliability, or sensitivity were not found.

Nicklin et al. 2010 [32]

Cross-sectional observational study, Direct, A

VAS and NRS

Develop and validate a patient reported outcome measure of fatigue in RA, the Bristol RA Fatigue- Multidimensional Questionnaire (BRAF-MDQ) and the Bristol RA Fatigue (BRAF) short scales (VAS/NRS)

The final wording for fatigue severity, effect, and coping VAS/NRS scales was based on focus group recommendations and required measurement properties. The VAS /NRS were understoodby all patients in the way they were intended by the authors. Vertical orientation of the scales enhanced comprehension (rather than horizontal).

The NRS and VAS scales were correlated between 0.68–0.78, and showed similar criterion and construct validity. The NRS produced slightly higher scores than the VAS and although the differences were not significant, the results demonstrate the scales are not interchangeable. Although the VAS and NRS performed in similar ways, the NRS was selected for use in evaluating fatigue in this population since some patients found the VAS difficult to understand and because the NRS is easier to score.

Khanna et al. 2008 [33]

Prospective, observational study, Indirect, C

VAS

Evaluate score interpretation (MID) for a fatigue VAS

Mean MID estimates ranged from −0.82 to −1.12 for improvement and 1.13 to 1.26 for worsening (range of 0–10) for a fatigue VAS. These results were similar to those see in RA clinical trials.

Koshy et al. 2004 [34]

Cross-sectional, observational study, Direct, A

VAS, VRS, Graphical rating scales

Determine patient preferences for pain assessment scale type

Most patients (56%) preferred the pain VAS, 30% preferred the graphical (coin) rating scale, 13% preferred the VRS, and no patients preferred the graphical (color) scale. Findings of statistically significant positive correlations between the VAS and VRS suggest both represent similar pain intensity, and both could be used as reliable pain assessment tools. A single item VAS was recommended for evaluating pain in oncology patients because it is reliable and well understood, and preferred by most patients in this study.

Anderson et al. 2007 [35]

Review, Indirect, B

VAS, VRS, and NRS

Review of pain assessment scales for us in an oncology population

Pain intensity ratings using the VAS, NRS, and VRS are highly inter-correlated. The NRS is easily understood by most patients, recommended in many pain treatment guidelines, and may be more reliable than the VAS in clinical trials, particularly with low literacy patients.

Pediatric cancer pain scales including color scales, pain thermometers, and Faces scales are suitable for evaluating cancer pain in children under 5 years of age. Children over the age of 5 years can typically complete NRS or VAS.

Rohan 2012 [36]

Review, Indirect, B

VRS and 11-point NRS

Review of distress screening measures used in oncology

A review of the multi-item Hospital Anxiety and Depression Scale (HADS) and the Brief Symptom Inventory- 18 (BSI-18) scale, and a single item Distress Thermometer (11-point NRS) concluded the Distress Thermometer was as discriminative as the multi-item HADS and BSI-18.

Sigurdardottir et al. 2014 [37]

Delphi-process, Indirect, D

NRS

Delphi process to obtain consensus on a basic set of core variables to describe or classify a palliative care cancer population

The 11-point NRS scale was recommended to evaluate important aspects of palliative care in cancer (e.g., appetite, depression, anxiety) and PRO instrument selection should always be undertaken with consideration of specific objectives, samples, treatments, and available resources.

King et al. 2014 [38]

Prospective observational study, Direct, A

11-point NRS and VAS

Determine optimal instrument to measure subjective symptom benefit in clinical trials of palliative

For an ovarian symptom PRO measure, the 11-point NRS was preferable over the VAS and VRS due to improved responsiveness, ease of use, and compliance.

Jacobs et al. 2013 [39]

Prospective observational study, Indirect, C

Faces scale

Psychometric evaluation of a pediatric mucositis scale in cancer patients

For a pediatric mucositis scale in cancer patients ages 8 to 18, a Faces scale was found to be reliable, valid, and responsive.

Ng et al. 2012 [40]

Cross-sectional, observational study, Direct, A

VAS, NRS, and Faces scales

Investigate correlations between, and patient preference for, pain assessment scales for use in an oncology population

The VAS, NRS, and Faces scale showed a high degree of association with intensity of pain making these scales appropriate for pain assessment in cancer. The Faces scale was preferred over the VAS and NRS and was superior to the NRS or VAS with cognitively impaired patients

Chordas et al. 2013 [41]

Prospective observational study, Direct, A

11-point NRS, VAS, VRS

Determine if a single item pain measure can accurately identify clinically significant pain in a pediatric brain cancer population

In a pediatric population of brain cancer patients, a multi-item measure with VRS was more precise than a single item disease thermometer (variation of 11-point NRS).

Banthia et al. 2006 [42]

Prospective observational study, Direct, A

VAS and VRS

Comparison of daily versus weekly, unidimensional versus multidimensional measures of fatigue in a breast cancer population

A single item cancer fatigue VAS daily and weekly had some discordance between the daily and weekly measurement, indicating they are not capturing the same information. The single item fatigue VAS showed greatest overlap with the general fatigue subscale of the multidimensional fatigue measure, suggesting the VAS item is a unidimensional measure of one aspect of fatigue. The decision to use a multidimensional or unidimensional measures of fatigue will depend upon the research question.

Grassi et al. 2013 [43]

Cross-sectional, observational study, Indirect, C

NRS with Graphical component and multi-item measures

Validation and acceptance of the Distress Thermometer in an Italian cancer population

A distress thermometer (NRS with graphical component) was as specific and sensitive as multi-item measures and was slightly preferred by patients.