Item response theory and factor analysis as a mean to characterize occurrence of response shift in a longitudinal quality of life study in breast cancer patients

Anota, Amélie; Bascoul-Mollevi, Caroline; Conroy, Thierry; Guillemin, Francis; Velten, Michel; Jolly, Damien; Mercier, Mariette; Causeret, Sylvain; Cuisenier, Jean; Graesslin, Olivier; Hamidou, Zeinab; Bonnetain, Franck

doi:10.1186/1477-7525-12-32

Table 4 Main recalibration profiles highlighted by a multiple correspondence analysis performed on the EORTC QLQ-C30

From: Item response theory and factor analysis as a mean to characterize occurrence of response shift in a longitudinal quality of life study in breast cancer patients

Dimension (items)	Time points	N	Percentage of recalibration (number of patients)	Recalibration category 1 to category 2	Recalibration category 2 to category 1	Recalibration category 3 to category 4	Recalibration category 4 to category 3	Categories 3 and 4 dispersed
Physical functioning	T1 - T2_R^a	100	37% (272)	Q1-Q5	Q1-Q5			Q1-Q5
(Q1, Q5)	T1 - T3_R^b	139	51% (274)	Q1-Q5	Q1-Q5			Q1-Q5
	T3 - T4_R^c	201	76% (266)	Q1-Q5	Q1-Q5			Q1-Q5
Role functioning	T1 - T2_R	84	31% (272)	Q6, Q7	Q6, Q7			Q6, Q7
(Q6-Q7)	T1 - T3_R	118	43% (274)	Q6, Q7	Q6, Q7	Q6, Q7	Q6, Q7
	T3 - T4_R	164	63% (261)	Q6, Q7	Q6, Q7			Q6, Q7
Emotional functioning	T1 - T2_R	180	68% (263)	Q21-Q24	Q21-Q24
(Q21-Q24)	T1 - T3_R	208	79% (263)	Q21-Q24	Q21-Q24
	T3 - T4_R	196	77% (255)	Q21-Q24	Q21-Q24			Q21-Q24
Cognitive functioning	T1 - T2_R	103	39% (266)	Q20, Q25
(Q20, Q25)	T1 - T3_R	129	48% (268)	Q20, Q25	Q20, Q25	Q20, Q25	Q20, Q25	Q20, Q25
	T3 - T4_R	130	51% (256)	Q20, Q25	Q20, Q25	Q20, Q25	Q20, Q25
Social functioning	T1 - T2_R	75	29% (260)	Q26, Q27	Q26, Q27			Q26, Q27
(Q26, Q27)	T1 - T3_R	101	38% (268)	Q26, Q27	Q26, Q27	Q26, Q27	Q26, Q27
	T3 - T4_R	142	56% (255)	Q26, Q27	Q26, Q27	Q26, Q27	Q26, Q27	Q26, Q27
Fatigue	T1 - T2_R	141	54% (259)	Q10, Q12, Q18	Q10, Q12, Q18			Q10, Q12, Q18
(Q10, Q12, Q18)	T1 - T3_R	160	61% (261)	Q10, Q12, Q18	Q10, Q12, Q18	Q12	Q10, Q12, Q18	Q10, Q12, Q18
	T3 - T4_R	183	73% (251)	Q10, Q12, Q18	Q10, Q12, Q18		Q10, Q12, Q18	Q10, Q12, Q18
Nausea and vomiting	T1 - T2_R	37	13% (280)	Q14	Q14, Q15			Q14, Q15
(Q14, Q15)	T1 - T3_R	65	24% (275)	Q14	Q14		Q14
	T3 - T4_R	98	37% (265)	Q14	Q14
Pain	T1 - T2_R	96	36% (267)	Q9	Q9			Q9, Q19
(Q9, Q19)	T1 - T3_R	124	46% (268)	Q9, Q19	Q9, Q19			Q9, Q19
	T3 - T4_R	148	58% (253)	Q9, Q19	Q9, Q19		Q9, Q19
Insomnia	T1 - T2_R	115	42% (277)	Q11	Q11		Q11
(Q8)	T1 - T3_R	135	49% (274)	Q11	Q11	Q11	Q11
	T3 - T4_R	147	55% (266)	Q11	Q11	Q11	Q11
Dyspnea	T1 - T2_R	50	18% (280)	Q8	Q8		Q8
(Q11)	T1 - T3_R	55	20% (279)	Q8	Q8	Q8	Q8
	T3 - T4_R	94	35% (269)	Q8	Q8	Q8	Q8
Appetite loss	T1 - T2_R	61	22% (280)	Q13	Q13		Q13
(Q13)	T1 - T3_R	85	30% (280)	Q13	Q13		Q13
	T3 - T4_R	90	34% (267)	Q13	Q13		Q13
Constipation	T1 - T2_R	71	26% (276)	Q16	Q16	Q16	Q16
(Q16)	T1 - T3_R	89	32% (277)	Q16	Q16	Q16	Q16
	T3 - T4_R	93	35% (264)	Q16	Q16	Q16	Q16
Diarrhea	T1 - T2_R	39	14% (278)	Q17	Q17
(Q17)	T1 - T3_R	64	23% (277)	Q17	Q17
	T3 - T4_R	61	24% (263)	Q17	Q17		Q17
Financial difficulties	T1 - T2_R	22	8% (264)	Q28	Q28
(Q28)	T1 - T3_R	33	12% (269)	Q28	Q28		Q28
	T3 - T4_R	47	18% (264)	Q28	Q28		Q28

Only patients with a recalibration of 5-point at least between a pre-test and a then-test measure are incorporated in these analyses. Items with observed recalibration are listed.
^aT1 → T2_R: comparison of baseline HRQoL assessment and retrospective measure performed after surgery.
^bT1 → T3_R: comparison of baseline HRQoL assessment and retrospective measure performed three months later.
^cT3 → T4_R: comparison of HRQoL assessment at three months and retrospective measure performed three months later.
As example, 100 patients presented a significant recalibration of physical functioning among the 272 patients with all the five items of the dimension answered at both measurement time. The graph representing the response categories highlight some recalibration profile:
- patients who had chosen response category 1 at the prospective measurement time for all the 5 items and who had chosen response category 2 to the same items at the retrospective measurement time.
- patients who had chosen response category 2 at the prospective measurement time for all the 5 items and who had chosen response category 1 to the same items at the retrospective measurement time.
- no recalibration profile is highlighted for response categories 3 and 4 and few patients had chosen these categories of response (these response categories are dispersed).

Back to article page

ISSN: 1477-7525

Contact us

Submission enquiries: journalsubmissions@springernature.com

Health and Quality of Life Outcomes

Contact us