Galen Research is the market leader in the development of PRO measures. Our expertise has been gained from working on over 30 disease-specific measures spanning the period from the 1970s (Nottingham Health Profile) through to the present day. Several of these measures are recognised as the measure of choice for clinical trials and even used by payers to determine which patients should be treated with expensive products.

Galen applies a highly scientific, academic approach to developing new measures. This ensures that they meet the highest critical standards for PROs and that they are able to determine differences in effectiveness of alternative interventions.

All our measures are based on a stated model or theory of the construct being measured, an approach generally overlooked by other measure developers. The rigorous development process of a new PRO measure involves four key stages:

  • Qualitative Interviews
    In-depth qualitative interviews are conducted with relevant patients based on a purposive sampling frame to ensure a representation of age, gender and disease severity. From these interviews, statements are identified relating to the impact of the condition on the lives of the patients.  These are used to develop draft versions of the measure.  
  • Cognitive debriefing
    The draft measure is field-tested with a new sample of patients to assess face and content validity, comprehensiveness and relevance. 
  • Validation survey
    The measure is then tested via a full validation survey with 100-200 patients at two time points two weeks apart. The results are used to create the final version of the measure.
  • Data analysis
    The psychometric properties of the measure are carefully assessed. Test-retest reliability, internal consistency and construct validity are all assessed in this stage. In addition to these standard classical assessments of the quality of the measure Item Response Theory (Rasch) analyses are also undertaken to ensure that the measure has good scaling properties. These include item fit, local item dependence, item-stability over time and differential item functioning by age and gender. Only if data collected with the scale fit the Rasch model can it be concluded that the scale in unidimensional. Only where a scale is unidimensional can item scores be added to give a valid total score and parametric statistical tests applied to trial data. This, in turn, allows meaningful minimum important difference scores to be calculated and disease-specific utility values too be generated from responses to the measure.

All Galen’s measures reflect the views of and the language used by patients. They are very practical having very simple response formats and taking fewer than 10 minutes to complete. They are equally easy to administer and score - reducing the risk of data errors and the need for trained administrators.

Galen Research has an experienced research team who have well established links with international experts and investigators, key opinion leaders, regulatory experts, key researchers in several major Universities as well as links with patient support groups. We have established research links with groups around the World who are experienced with our methodology and work to our own high standards. In this way we are able to ensure that required language adaptations are produced to a very high standard.

Please contact us if you wish to discuss the development of a new measure.