Construct validity of the OCTOPuS stratification algorithm for allocating patients with knee osteoarthritis into subgroups

Knoop, Jesper; Ostelo, Raymond W. J. G.; van der Esch, Martin; de Zwart, Arjan; Bennell, Kim L.; van der Leeden, Marike; Dekker, Joost

doi:10.1186/s12891-021-04485-1

Table 1 Overview of a-priori hypotheses (n = 63) and scores to accept these hypotheses

From: Construct validity of the OCTOPuS stratification algorithm for allocating patients with knee osteoarthritis into subgroups

RESEARCH QUESTION I: SIMILAR SUBGROUP PROPORTIONS
Subgroup proportion in one cohort is similar to subgroup proportion in total sample	Deviation
*‘Low muscle strength subgroup’*
AMS-OA [5] vs. total sample	−10, + 10%¹
STABILO [17] vs. total sample	−10, + 10%¹
NEXA [18] vs. total sample	−10, + 10%¹
CBT [19] vs. total sample	−10, + 10%¹
VIDEX (De Zwart AH, Dekker J, Roorda LD, van der Esch M, Lips P, van Schoor NM, et al.: High-intensity resistance training and vitamin D supplementation for knee osteoarthritis: a randomized controlled trial, Under review) vs. total sample	−10, + 10%¹
*‘High muscle strength subgroup’*
AMS-OA [5] vs. total sample	− 10, + 10%¹
STABILO [17] vs. total sample	−10, + 10%¹
NEXA [18] vs. total sample	−10, + 10%¹
CBT [19] vs. total sample	−10, + 10%¹
VIDEX (De Zwart AH, Dekker J, Roorda LD, van der Esch M, Lips P, van Schoor NM, et al.: High-intensity resistance training and vitamin D supplementation for knee osteoarthritis: a randomized controlled trial, Under review) vs. total sample	−10, + 10%¹
*‘Obesity subgroup’*
AMS-OA [5] vs. total sample	− 10, + 10%¹
STABILO [17] vs. total sample	−10, + 10%¹
NEXA [18] vs. total sample	−10, + 10%¹
CBT [19] vs. total sample	−10, + 10%¹
VIDEX (De Zwart AH, Dekker J, Roorda LD, van der Esch M, Lips P, van Schoor NM, et al.: High-intensity resistance training and vitamin D supplementation for knee osteoarthritis: a randomized controlled trial, Under review) vs. total sample	-10, + 10%¹
RESEARCH QUESTION 2: CHARACTERISTICS IN LINE WITH UNDERLYING PHENOTYPES
Characteristic in one subgroup that is in line with proposed underlying phenotype is different from other subgroups	p-value
*‘Low muscle strength subgroup’ (‘age-induced phenotype’)*
Higher age, compared to:
‘high muscle strength subgroup’	P < 0.05²
‘obesity subgroup’	P < 0.05²
Lower muscle strength, compared to:
‘high muscle strength subgroup’	P < 0.05²
‘obesity subgroup’	P < 0.05²
‘*High muscle strength subgroup’ (‘post-traumatic phenotype’)*
More history of knee surgery, compared to:
‘low muscle strength subgroup’	P < 0.05²
‘obesity subgroup’	P < 0.05²
Higher muscle strength, compared to:
‘low muscle strength subgroup’	P < 0.05²
‘obesity subgroup’	P < 0.05²
More males, compared to:
‘low muscle strength subgroup’	P < 0.05²
‘obesity subgroup’	P < 0.05²
Younger age, compared to:
‘low muscle strength subgroup’	P < 0.05²
‘obesity subgroup’	P < 0.05²
Higher K/L grade, compared to:
‘low muscle strength subgroup’	P < 0.05²
‘obesity subgroup’	P < 0.05²
Less comorbidities, compared to:
‘low muscle strength subgroup’	P < 0.05²
‘obesity subgroup’	P < 0.05²
Less severe knee pain, compared to:
‘low muscle strength subgroup’	P < 0.05²
‘obesity subgroup’	P < 0.05²
Less impaired physical function, compared to:
‘low muscle strength subgroup’	P < 0.05²
‘obesity subgroup’	P < 0.05²
*‘Obesity subgroup’ (‘metabolic phenotype’)*
Higher BMI, compared to:
‘high muscle strength subgroup’	P < 0.05²
‘low muscle strength subgroup’	P < 0.05²
More comorbidities, compared to:
‘high muscle strength subgroup’	P < 0.05²
‘low muscle strength subgroup’	P < 0.05²
Lower muscle strength, compared to:
‘high muscle strength subgroup’	P < 0.05²
‘low muscle strength subgroup’	P < 0.05²
More severe knee pain, compared to:
‘high muscle strength subgroup’	P < 0.05²
‘low muscle strength subgroup’	P < 0.05²
More severe impaired physical function, compared to:
‘high muscle strength subgroup’	P < 0.05²
‘low muscle strength subgroup’	P < 0.05²
RESEARCH QUESTION 3: EFFECTS OF USUAL EXERCISE THERAPY IN LINE WITH HYPOTHESIZED EFFECTS
	Effect size/ % with MIC
*Large effects in ‘low muscle strength subgroup’*
Large effect size on knee pain	0.8 ± 0.2
Majority with MIC on knee pain	> 67%
Large effect size on physical function	0.8 ± 0.2
Majority with MIC on physical function	> 67%
Large effect size on muscle strength	0.8 ± 0.2
Majority with MIC on muscle strength	< 67%
*Medium effects in ‘obesity subgroup’*
Medium effect size on knee pain	0.5 ± 0.2
Half with MIC on knee pain	33–67%
Medium effect size on physical function	0.5 ± 0.2
Half with MIC on physical function	33–67%
Medium effect size on muscle strength	0.5 ± 0.2
Half with MIC on muscle strength	33–67%
*Small effects in ‘high muscle strength subgroup’*
Small effect size on knee pain	0.2 ± 0.2
Minority with MIC on knee pain	< 33%
Small effect size on physical function	0.2 ± 0.2
Minority with MIC on physical function	< 33%
Small effect size on muscle strength	0.2 ± 0.2
Minority with MIC on muscle strength	< 33%

MIC = minimal important change; ¹ difference in subgroup proportion (%)in one cohort compared to subgroup proportion in total sample; ² p-value for differences between subgroups; ³ isokinetic knee extensor strength measure as outcome; ⁴ 30-s chair stand test as outcome; *significant finding in the opposite direction as expected, therefore hypothesis not accepted

Back to article page

ISSN: 1471-2474

Contact us

Submission enquiries: bmcmusculoskeletaldisorders@biomedcentral.com
General enquiries: ORSupport@springernature.com

BMC Musculoskeletal Disorders

Contact us