TABLE 1.

Reasons for item deletion or modification and supporting analyses.^*

ITEM PROPERTY	REASON FOR CHANGE OR DELETION	SUPPORTING ANALYSIS AND CRITERIA
Clarity or relevance	Reported as not relevant by a large segment of the target population	Percentage of missing data; <10% missing data points considered within acceptable criteria
	Generates an unacceptably large amount of missing data points	Rasch modelling provides a measure of how difficult or easy an item is relative to the subject and used to assess the relevance of an item to that population
	Generates an unacceptably large amount of missing data points	n/a
	Patients interpret items and responses in a way that is inconsistent with the PRO instrument’s conceptual framework	n/a
Response range	A high percent of patients respond at the floor (response scale’s worst end) or ceiling (response scale’s optimal end)	Percentage scoring zeros or maximum scores; <80% floor (option score of zero endorsed) or ceiling effects (maximum option score endorsed) considered within acceptable criteria
	Patients note that none of the response choices applies to them	Aggregate frequency of pairs of adjacent options; aggregate frequency of >10% considered within acceptable criteria
	Distribution of item responses is highly skewed	n/a
Variability	All patients give the same answer (i.e., no variance)	Frequency of option endorsement; <50% endorsement of one option considered within acceptable criteria
	Most patients choose only one response choice
	Differences among patients not detected when important differences are known
Reproducibility	Unstable scores over time when there is no logical reason for variation from one assessment to the next	Present data does not allow assessment of reliability; estimated by correlation of DID items with MADRS items that are assumed to assess the same construct; r>0.75 considered within acceptable criteria
Inter-item correlation	Item highly correlated (redundant) with other items in the same concept of Interest	Inter-item correlations coefficients; r<0.75 considered within acceptable criteria
Ability to detect change	Item is not sensitive (i.e., does not change when there is a known change in the concepts of interest)	Present data does not allow assessment of change to therapy; estimated by correlation of DID items with MADRS items that are assumed to assess the same construct; r >0.75 considered within acceptable criteria
Item discrimination	Item is highly correlated with measures of concepts other than the one it is intended to measure	Correlation of DID items with MADRS items that are assumed to assess the same construct; r >0.75 considered within acceptable criteria
	Item does not show variability in relation to some known population characteristics (i.e., severity level, classification of condition, or other known characteristic)	IRT used to determine relation to overall depressive severity (total MADRS score)
Redundancy	Item duplicates information collected with other items that have equal or better measurement properties	Inter-item correlations coefficients; r<0.75 considered within acceptable criteria
Recall period	The population, disease state, or application of the instrument can affect the appropriateness of the recall period	Not directly assessed; previous literature suggests that recall period is appropriate in this population

^*Adapted from references #17 and #19

n/a: not available; PRO: patient reported outcome; DID: depression inventory development project; MADRS: Montgomery–Åsberg Depression Rating Scale; IRT: Item Reponse Theory