Quality of life defined Quality of life (QOL) ality of life (QOL) - PDF document

6/9/2014 Psychometric Properties of Quality of Life and Health ‐ Related Quality of Life Assessments in People with Multiple Sclerosis Learmonth, Y. C., Hubbard, E. A., McAuley, E. Motl, R. W. Department of Kinesiology and Community Health, University of Illinois at Urbana ‐ Champaign, Urbana, IL, USA Quality of life defined • Quality of life (QOL) ality of life (QOL) – Subjective w bjective well-being ll-being or or Satisfaction with tisfaction with life life 1, 2 1, 2 • Health r alth related-QOL (HR lated-QOL (HRQOL) QOL) – Physical & psy Physical & psychological hological aspects of evaluating aspects of evaluating one one’s health status s health status 3,4 3,4 • Both ar Both are lo e lower in MS r in MS – Compar Compared with healthier population ed with healthier population 5 – Compar Compared with other ed with other disease population disease population 6-8 6-8 1

6/9/2014 • 9 9, 10 Measurement of QOL & HRQOL • Ge Gene neric sc scales – 12-item 12-item Shor ort F t Form rm Health S alth Survey (SF-12) ey (SF-12) 11 11 – Satisfaction tisfaction With Life Scale (SWLS th Life Scale (SWLS )2, )2, • MS scales MS scales – Leeds M Leeds Multiple Sclerosis Q ltiple Sclerosis Quality ality of Life of Life (LMSQOL) (LMSQOL) 12 12 – 29-item 29-item Multiple Sclerosis I ltiple Sclerosis Impact pact Scale Scale (MSIS-29 (MSIS-29 )13, )13, 2

6/9/2014 Measurement of Psychometric properties Reliability • Test-r st-retest r etest reliability; liability; – Temporal stability mporal stability • Phenomenon Phenomenon (i.e., QOL and HR (i.e., QOL and HRQOL) QOL) • Measur asurement ement (i.e., SF-12 and SWLS) (i.e., SF-12 and SWLS) Measurement error • All measur All measures ar es are vulnerable e vulnerable to err to error – Standar andard error error of of measur measurement ement (SEM) (SEM) – Coefficient of Coefficient of variation (CV variation (CV) Interpretability • Smallest detectable change (SDC) allest detectable change (SDC) 3

6/9/2014 Is the change meaningful? No change SEM or SDC Maximal CV change ≥ SEM/CV <SEM/CV ≥ SDC Change cannot be Change statistically Change statistically distinguished from significant, but NOT significant and MAY be measurement error important important Study Purpose • Determine the test-r termine the test-retest etest reliability liability, measurement err measur ement error, and interpr and interpretability of tability of QOL (i.e., SWLS QOL (i.e., SWLS and LMSQOL) and LMSQOL) and and HR HRQOL (i.e., SF-12 and MSIS-29) QOL (i.e., SF-12 and MSIS-29) measur measures es over six months er six months in people with in people with MS. MS. • Interpr terpret the r t the results sults of inter of intervention ention effectiv effectiveness eness 4

6/9/2014 Participants Initial interest 511 •124 lost interest with initial explanation 387 screened •27 did not meet inclusion •16 lost interest 344 •44 did not return Sent demographic, baseline QOL and HRQOL documents questionnaires •8 discontinued participation 292 at baseline •18 did not return documents at 6 month follow ‐ up Final analysis 274 Outcomes • SWLS SWLS – 5 items, 7-point scale. H 5 items, 7-point scale. Higher gher scor scores = higher es = higher QOL. QOL. • LMSQOL LMSQOL – 8 items, 4 point 8 items, 4 point scale. H scale. Higher scor gher scores es = worse QOL = worse QOL • SF12 SF12 – 12 12 items, composite point items, composite point scale. P scale. Physical ysical composite composite (PCS) & (PCS) & Mental composite ntal composite (MCS). H (MCS). Higher scor gher scores es = = higher higher HR HRQOL QOL • MSIS-29 MSIS-29 – 29 29 items, 4 point items, 4 point scale. P scale. Physical ysical and and psy psychological hological components. H components. Higher gher scor scores = worse es = worse HR HRQOL QOL 5

6/9/2014 Data analysis • Reliability liability ICC CC analys analyses es (2,1 (2,1 mix mixed model) d model) – ≥0.6=moderate reliability ≥0.8=good reliability • Measur asurement err ement error – SEM = SEM = SD SD ba ne x √(1-I x √(1-ICC) CC) (SD-of each (SD-of each outcome) outcome) basel seline – CV CV = = dividing s dividing sample SD mple SD of the diffe of the difference rence betw between the een the two time-points two time-points, by by the the mean mean difference betw difference between the een the time time points points x100 x100 • Interpr terpretability tability – SDC =1.96 SDC =1.96 x √(2) x √(2) x SEM x SEM – SDC % SDC % = % = % of of baseline baseline mean mean • Va Validity – Spearman corr earman correl elat ations ions • ≥0.5=good v ≥0.5=good validity lidity Results; Sample description (n=274) Variable Sex (N, % female) 229 (84) Age (years) Mean (SD) 48.0 (10.4) Range 20 ‐ 84 Type of MS Relapsing Remitting N (%) 222(81) Secondary Progressive N (%) 33(12) Primary Progressive N (%) 12(4) Benign 6(2) Disease duration (years) Mean (SD) 10.3 (7.8) Range 1 ‐ 37 PDDS score Median (IQR) 3 (3) Range 0 ‐ 6 6

6/9/2014 Change over time Measure Baseline mean Six month mean Change p ‐ (SD, SE) (SD, SE) value SWLS 21.8 (8.0, 0.5) 22.1 (8.2, 0.5) 1.38 .306 SF ‐ 12 PCS 41.9 (9.0, 0.5) 41.3 (9.5, 0.6) ‐ 1.43 .182 SF ‐ 12 MCS 41.5 (9.2, 0.6) 43.3 (7.5, 0.5) 4.34 <.001 LMSQOL 19.3 (4.8, 0.3) 19.0 (4.9, 0.3) 1.55 .133 MSIS ‐ 29 39.9 (27.8, 1.7) 39.0 (29.3, 1.8) 2.26 .339 Physical MSIS ‐ 29 43.0 (29.85, 1.8) 39.5 (28.7, 1.7) 8.14 .004 Mental Reliability Measure ICC 95% CI ICC SWLS .772 .720 ‐ .816 SF ‐ 12 PCS .741 .682 ‐ .790 SF ‐ 12 MCS .669 .598 ‐ .730 LMSQOL .812 .767 ‐ .849 MSIS ‐ 29 Physical .883 .853 ‐ .906 MSIS ‐ 29 Mental .768 .715 ‐ .813 Moderate ( ≥ 0.6) & good ( ≥ 0.8) reliability 7

6/9/2014 Measurement Error Measure SEM %SEM CV (%) SWLS 3.8 17.4 13.4 SF ‐ 12 PCS 4.6 11.1 9.3 SF ‐ 12 MCS 5.3 12.5 9.7 LMSQOL 2.1 10.9 8.9 MSIS ‐ 29 Physical 9.5 24 28.7 MSIS ‐ 29 Mental 13.2 30.7 31.2 Accuracy Interpretability Measure SDC 95 % SDC 95 SWLS 10.5 48.1 SF ‐ 12 PCS 12.7 30.6 SF ‐ 12 MCS 14.7 34.7 LMSQOL 5.8 30.1 MSIS ‐ 29 Physical 26.4 67 MSIS ‐ 29 Mental 36.7 89 8

6/9/2014 Validity Measure SWLS SF ‐ 12 SF ‐ 12 LMSQOL MSIS ‐ 29 MSIS ‐ 29 PDDS PCS MCS Physical Mental SF ‐ 12 PCS .355** .410** .071 SF ‐ 12 MCS LMSQOL ‐ .674** ‐ .411** ‐ .623** MSIS ‐ 29 ‐ .489** ‐ .671** ‐ .386** .571** Physical ‐ .561** ‐ .326** ‐ .581** .696** .669** MSIS ‐ 29 Mental PDDS ‐ .309** ‐ .681** ‐ .107 .360** .704** .350** ‐ Convergent Convergent Construct ≥ 0.5 good Validity Validity Validity In summary • Moderate derate (to good) r (to good) reliability liability • Suppor pport past past findings for LMSQOL findings for LMSQOL & MSIS- & MSIS- 29 29 12,13. 12,13. • Novel r l reliability r liability results sults for SWLS & SF-12 in for SWLS & SF-12 in MS MS • Stability o ability over six er six months is months is impor important ant 9

6/9/2014 Interpreting change (LMSQOL) 10% 30% No change SEM or SDC Maximal CV change <10% ≥ 30% ≥ 10% Change cannot be Change statistically Change statistically distinguished from significant and MAY be significant, but NOT measurement error important important Validity • First study r rst study repor porting ing relationships betw lationships between een QOL and HR QOL and HRQOL in MS QOL in MS • Good ood validity of all four lidity of all four measur measures es – Constr Construct uct – Convergent Convergent 10

6/9/2014 Discussion • Ov Overall, our data suggest erall, our data suggest that the that the phenomenon phenomenon (HR (HRQOL & QOL) & all four QOL & QOL) & all four measur measures hav es have acceptable acceptable measur measurement ement stability stability, as indicated as indicated thr through the r ugh the reliability liability estimates. estimates. • Power calculations r calculations • Interpr terpret clinical et clinical scor scores es • Limitations Limitations • Distribution & criterion method stribution & criterion method recommended. commended. Recommendations • Resear search r ch recommendation commendation – Consider all psy Consider all psychometric hometric proper properties ties • QOL r QOL recommendation commendation – LMSQOL LMSQOL • HR HRQOL r QOL recommendation commendation – SF-12 SF-12 11