Psychometric Properties of the Photograph Series of

Psychometric properties of the Photograph Series of Daily Activities - Short Electronic Version (PHODA-SeV) in patients with chronic low back pain according to COSMIN checklist. Crystian Bitencourt Oliveira, PT, MSc1, Marcia Rodrigues Franco, PhD2, Samantha Janaina Demarchi, PT1, Rob Johannes E. M. Smeets, PhD3,4, Ivan P. J. Huijnen, PhD3,5, Priscila Kalil Morelhão, PT, MSc1, Thalysi Mayumi Hisamatsu, PT, MSc 1, Rafael

Journal of Orthopaedic & Sports Physical Therapy® Downloaded from www.jospt.org at La Trobe University on May 27, 2018. For personal use only. No other uses without permission. Copyright © ${year} Journal of Orthopaedic & Sports Physical Therapy®. All rights reserved.

Zambelli Pinto, PhD6. ¹ Department of Physical Therapy, Faculdade de Ciências e Tecnologia, Universidade Estadual Paulista (UNESP), Presidente Prudente, Brazil. 2

Department of Physical Therapy, Centro Universitário UNA, Contagem, Brazil

3

Department of Rehabilitation Medicine, Research School Caphri, Maastricht University 4

Libra Rehabilitation and Audiology, Eindhoven/Weert, The Netherlands

5

Adelante Centre of Expertise in Rehabilitation and Audiology, Hoensbroek, The Netherlands 6

Department of Physical Therapy, Universidade Federal de Minas Gerais (UFMG), Belo Horizonte, Brazil

Corresponding author: Rafael Zambelli Pinto, Av. Pres. Antônio Carlos, 6627, Pampulha, Belo Horizonte, MG, Brazil T: +55 18 3229 5534 E: [email protected] W: www.unesp.br/international/

1

Funding: This research did not receive any specific grant from funding agencies. C.B.O., T.M.H. and M.R.F. were supported by São Paulo Research Foundation (FAPESP/grant numbers: 2016/03826-5, 2016/03826-5, 2015/07704-9). P.K.M. was supported by Capes Foundation, Ministry of Education of Brazil. Register protocol number in the clinicaltrials.gov: NCT02398760. Research Ethics Committee: CAAE36332514.0.0000.5402 (Sao Paulo State


University Ethics Research Committee).

2


1

Abstract

2

Study Design: A prospective cohort study.

3

Background: The Photograph Daily Activities Series - Short Electronic Version

4

(PHODA-SeV) assesses perceived harmfulness of daily activities in patients with low

5

back pain (LBP). Although there is some evidence that the PHODA-SeV is a reliable

6

and valid tool, its psychometric propertieshave not been fully investigated.

7

Objectives: To investigate the test-retest reliability, measurement error, interpretability,

8

construct validity, internal and external responsiveness of the PHODA-SeV in patients

9

with chronic LBP.

10

Methods: Ninety-one patients were included in the analysis. For reliability purposes,

11

the PHODA-SeV was administered twice with 1-week interval before commencement

12

of the treatment. Pain, disability and measures of pain-related fear (i.e. PHODA-SeV,

13

Fear Avoidance Beliefs Questionnaire (FABQ) and Tampa Scale of Kinesiophobia

14

(TSK) were collected before and after the 8-week treatment period.

15

Results: PHODA-SeV showed an excellent reliability (ICC2,1=0.91) without evidence

16

of ceiling and floor effects. The construct validity analysis demonstrated fair

17

correlations (i.e. r between 0.25 and 0.50) of PHODA-SeV with FABQ, but no

18

correlation with TSK (i.e. r lower than 0.25). For the internal responsiveness, PHODA-

19

SeV showed an effect size of 0.87 and standardized response mean of 0.92, interpreted

20

as large effect (i.e. greater than 0.80). For the external responsiveness, the correlation

21

between PHODA-SeV and changes in TSK and FABQ were considered low and the

22

Receiver Operating Curve analyses revealed an area under curve lower than the

23

proposed threshold of 0.70.

24

Conclusion: The PHODA-SeV is a reliable tool and able to detect changes overtime in

25

pain-related fear of patients with chronic LBP undergoing a physical therapy treatment.

26

This tool, however, failed to identify those patients who improved or not according to

27

other pain-related fear measures.

28

Key-words: chronic low back pain; pain-related fear; reliability; validity;

29

responsiveness;

3


30

Introduction

31

Low back pain (LBP) is a highly prevalent musculoskeletal condition and imposes a

32

challenge for health care systems10. Psychological factors are believed to contribute to

33

the transition from acute to chronic pain and disability27, 41, 42. The fear-avoidance model

34

postulates that factors, such as fear and expectation of adverse consequences from

35

increasing activities, are associated with the development of avoidance behavior,

36

resulting in disability, depression and disuse8,

37

specific fear that physical activities will cause (re)injury is the central concept in the fear-

38

avoidance model, which has been found to be associated with the maintenance of chronic

39

LBP39 and with poor outcomes9,

40

properties that measure pain-related fear may help clinicians to identify those patients

41

with poor prognosis.

42

The Tampa Scale for Kinesiophobia (TSK) and the Fear Avoidance Beliefs Questionnaire

43

(FABQ) are commonly used tools to evaluate pain-related fear in patients with LBP. The

44

TSK assesses the beliefs that a painful activity will cause damage, suffering, or functional

45

loss4. The FABQ was developed to measure levels of fear and avoidance beliefs about

46

work and physical activity40. Although both tools have demonstrated good psychometric

47

properties7,

48

patients would fear or avoid during daily activities30. The Photograph Series of Daily

49

Activities Scale (PHODA) is an innovative tool14 to measure to what extent movements

50

or activities are regarded as harmful by patients with back pain, using pictures to rate

51

harmfulness of a range of specific spinal movements and daily activities. Two versions

52

of the PHODA are available: the original version consisting of 100 photographs14 and the

53

PHODA- Short Electronic Version (PHODA-SeV) with 40 photographs16. The PHODA-

54

SeV has the advantage of being easy to administer, which makes it suitable to be used

55

routinely in clinical practice. Although the PHODA-SeV can be used to identify harmful

56

activities to be addressed using a specific intervention (e.g. graded exposure), this tool

57

could also be implemented as an evaluative measure after a full investigation of its

58

psychometric properties.

59

Leeuw et al. 200716 investigated the psychometric properties of the PHODA-SeV in

60

adults with chronic LBP. The PHODA-SeV showed good internal consistency, excellent

61

reliability and consistent relationship with related constructs such as self-report measures

25, 34

35

15, 18, 38

. Pain-related fear or the more

. Hence, instruments with good psychometric

, these tools fail to evaluate specific spinal movements or activities

4

62

of fear of movement/(re)injury, pain catastrophizing, functional disability, and current

63

pain intensity. However, the previous study selected the participants from a previous trial

64

with high levels of fear of movement and referred from physicians in Netherlands. Given

65

that not all measurement properties are evaluated yet and to increase the implementation

66

of this instrument to measure pain-related fear, the aim of this study is to fully investigate

67

the measurement properties of the PHODA-SeV, including internal consistency,

68

reliability, measurement error, construct validity, responsiveness and interpretability, in

69

patients with nonspecific chronic LBP.


70 71

Methods

72

The protocol of this prospective cohort study has been previously registered in the

73

clinicaltrials.gov (NCT02398760).

74

Participants

75

Patients with non-specific chronic LBP were recruited via local press and social media

76

advertisements, and from two outpatient physical therapy university clinics in Presidente

77

Prudente. To be included, patients aged between 18 and 60 years had to be diagnosed

78

with non-specific chronic LBP, defined as pain not attributable to a specific cause

79

localized below the costal margin and above the inferior gluteal folds, with or without leg

80

pain, for at least three months. In addition, patients had to report at least moderate-

81

intensity LBP (i.e. at least 4 points in the 0-10 pain Numerical Rating Scale [NRS]) or

82

moderate interference with function as measured by items 7 and 8 of the SF36

83

questionnaire5. Patients were excluded if presenting: suspected or diagnosed serious

84

spinal pathology (i.e. inflammatory spondyloarthropathy, fracture, malignancy, cauda

85

equina syndrome, or infection), nerve root compromise (i.e. at least 2 of the following

86

signs: weakness, reflex change, or sensation loss, associated with the same spinal nerve),

87

history of spinal surgery, fibromyalgia, any other musculoskeletal condition which may

88

affect activity and movement, pregnancy, illiterate, insufficient understanding of the

89

Portuguese language, or any contraindications to exercise according to American College

90

Sports Medicine (ACSM)2. Initial examination was performed by a physical therapist.

91

The university ethics research committee at the Sao Paulo State University approved the

92

study (CAAE36332514.0.0000.5402) and all participants provided informed consent.

5


93

Procedures

94

For those patients considered eligible, we collected socio-demographic and

95

anthropometric data, symptoms duration, pain intensity, disability and measures of pain-

96

related fear (i.e. PHODA-SeV, TSK and FABQ). The questionnaires were administered

97

by two trained assessors but the order of administration was not randomized. For

98

reliability purposes, the PHODA-SeV was administered again 1 week after the initial

99

assessment before commencement of the treatment. The time interval was adopted to

100

prevent recall of previous assessment and ensure that patient’s condition remain

101

unchanged. After initial assessments, all the included participants underwent an

102

individual program for eight weeks, twice a week, delivered by physical therapists. The

103

intervention was not protocolled but the physical therapists who administered the

104

treatment were instructed to follow the clinical practice guidelines for the management

105

of chronic LBP (i.e. supervised exercise therapy)13. Pain intensity, disability, and

106

measures of pain-related fear were collected after 8-week treatment period (i.e. 9 weeks

107

after the initial assessment).

108

Instruments

109

A standardized assessment form was used to collect demographic and anthropometric

110

data (e.g. age, sex, body mass index, employment status [i.e. unemployed/employed]),

111

duration of symptoms, pain intensity, disability and levels of pain-related fear and

112

patient’s perception of recovery (only post-intervention). The Brazilian-Portuguese

113

version or an adaptation of the questionnaires were used in the present study. The NRS

114

was used to assess the average pain intensity during the last 24 hours. The NRS ranges

115

from 0 (“no pain”) to 10 (“worst pain possible”). The Roland Morris Disability

116

Questionnaire (RMDQ) was used to measure disability26, 31. The RMDQ consists of 24

117

yes-or-no items with a total score ranging from 0 (“no disability”) to 24 (“maximum

118

disability”).

119

Measures of pain-related fear assessed in this study were FABQ, TSK and PHODA-SeV.

120

The FABQ is a self-reported questionnaire consisting of 16 items grouped in two

121

subscales: physical activity (FABQ-PA) and work (FABQ-Work) subscales. Of these, 11

122

items (i.e. 7 items from FABQ-Work and 4 items from FABQ-PA) are scored from 0

123

(“totally disagree”) to 6 points ("totally agree")1, 40. The FABQ have a total score ranging

124

from 0 to 66 points, with higher scores indicating higher fear-avoidance beliefs. The 6


125

Brazilian-Portuguese version of the FABQ has shown good reliability (Intraclass

126

Correlation Coefficient (ICC)(2,1) = 0.96, 95% Confidence Interval (CI): 0.94 to 0.98)7.

127

The TSK is a self-reported questionnaire consisting of 17 items. Each item in a scale from

128

1 (“totally disagree”) to 4 points (“totally agree”)22, 32. TSK total scores ranges from 17

129

to 68 points, with higher scores indicating greater levels of fear of movement. The

130

Brazilian-Portuguese version of the TSK demonstrated high reliability (ICC(2,1) = 0.93

131

95% CI: 0.87 to 0.96)7.

132

The PHODA-SeV is a self-administered tool used to assess the perceived harmfulness of

133

movements performed during daily activities. This tool uses pictures to describe 8

134

possible movements (lifting, bending, turning, reaching, falling, intermittent load

135

unexpected movement, and long-lasting load in stance or sit with limited dynamics), set

136

against 4 areas of daily occupations (activities of daily living, housekeeping, work and

137

sports and leisure time) and converted into recognizable and frequent daily activities

138

instead of in terms of their biomechanics16. The original version of the PHODA was

139

proposed with 100 photographs of daily activities14. In the current study, we used a

140

shortened

141

(https://ppw.kuleuven.be/ogp/software/phodasev-en.zip), known as the PHODA-SeV,

142

which contains 40 selected photographs16. Considering that the PHODA-SeV is an

143

electronic instrument, each patient was familiarized with how to rate score the items using

144

a laptop computer. The participants rate the activities using a scale ranging from 0 (“not

145

harmful at all”) to 100 (“extremely harmful”). The respondent is asked to look at the

146

photographs and try to imagine himself performing the shown activity. Then, for each

147

photograph the participant answers the following question: “How harmful do you feel

148

that this activity would be for your back?”. The total score is calculated by averaging the

149

score of each of the 40 photographs.

150

Statistical Analysis

151

Data are presented as frequency (proportion), mean (standard deviation) and median

152

[interquartile range]. Paired T-test was used to compare the differences between baseline

153

and 8-week follow up.

154

The measurement properties of the PHODA-SeV investigated in this study were:

electronic

version

of

the

PHODA

7


155

Internal Consistency: Internal consistency investigates the level of interrelatedness

156

between the items of the instrument. The Cronbach’s α was calculated to analyze the

157

internal consistency of the PHODA-SeV which values ranging from 0.70 to 0.95

158

indicating sufficient homogeneity36.

159

Test-retest reliability. Test-retest reliability is the extent to which the measures for the

160

same patients remain unchanged after repeated measurements. Intraclass correlation

161

coefficient (ICC(2,1)) was used to calculate the intra-rater reliability using the total score

162

and each item of the PHODA-SeV. ICC values greater than 0.70 indicates excellent

163

reliability36.

164

Measurement error. Measurement error is the error of a patient’s score that is not

165

attributable to true changes in the construct measured. The Standard Error of

166

Measurement (SEM) and Smallest Detectable Change (SDC) were calculated using the

167

following formulas, respectively: SEM= SD × √1 − 𝐼𝐶𝐶; and SDC= 1,96× √2 × 𝑆𝐸𝑀,

168

where SD is the Standard Deviation of the baseline. The SEM reflects the possibility that

169

the change in the measure may be attributable to random error. The SDC is defined as the

170

smallest within-person change in the measure that can be interpreted as a real change,

171

above the measurement error36.

172

Construct Validity. The construct validity measures the degree to which the instrument is

173

consistent with the hypotheses based on the assumption that the instrument validity

174

measures the construct intended to measure. Pearson coefficient (r) was used to analyze

175

the correlation of the PHODA-SeV with other measures of pain-related fear (i.e. TSK and

176

FABQ). The levels of construct validity are interpreted as follows: weak validity, if r

177

lower than 0.30; moderate validity, if r between 0.30 to 0.60; good validity, if r higher

178

than 0.6036. Our hypothesis for construct validity was that the PHODA-SeV would show

179

positive and at least moderate validity with at least one measure of pain-related fear, TSK

180

or FABQ.

181

Responsiveness. There are two major aspects of responsiveness: internal and external

182

responsiveness. Internal responsiveness refers to the ability of a measure to change over

183

a pre-specified time frame and external responsiveness reflects the extent to which change

184

in a measure relates to a corresponding change in a reference measure of clinical or health

185

status11. Internal responsiveness was calculated using the Effect Size (ES) (𝐸𝑆 =

8

186


187

Mean changes

) and the Standardized Response Mean (SRM) formulas (𝑆𝑅𝑀 =

SD of baseline

Mean changes

), where mean changes were calculated as the difference between baseline

SD of the change

188

and post-treatment. Considering that the TSK and FABQ also assess pain-related fear

189

constructs, we divided the sample in “improved” and “not improved” using the changes

190

in TSK and FABQ as external anchors to investigate the external responsiveness. For the

191

TSK, patients with change scores equal or more than 12 points were categorized as

192

“improved” and patients scoring less than 12 as “not improved”7. For the FABQ, patients

193

with change scores equal or more than 17 points were categorized as “improved” and

194

patients scoring less than 17 as “not improved”. The cutoff points of the TSK and FABQ

195

were determined using the SDC. We calculated the SDC using the intraclass correlation

196

coefficient reported in previous studies investigating the psychometric properties of these

197

questionnaires1, 32. The ES and SRM were interpreted as follows: 0.20 represents a small

198

effect; 0.50 a moderate effect; and 0.80 a large effect6. For external responsiveness, we

199

correlated the change in the PHODA-SeV (i.e. baseline minus follow-up score) with

200

change scores of the TSK and FABQ and conducted the receiver operating characteristics

201

(ROC) analysis. If the ROC analysis shows an area under the curve (AUC) equal or

202

greater than 0.70, the instrument was considered to have an appropriate external

203

responsiveness. An AUC of 0.70 means that the instrument can correctly identify the

204

subject condition, “improved” or “not improved” in 70% of the sample.

205

According to the COSMIN checklist24, the responsiveness must be examined using a

206

priori-formulated hypotheses. The criterion used to assess the responsiveness is that the

207

results should be in accordance with at least 75% of the hypotheses23. Previous studies

208

showed the performance of pain-related fear measures to vary from low to medium effect

209

size (ES0.5)

210

after multidisciplinary treatment3,

211

hypotheses were that the ES and SRM of the PHODA-SeV would be at least 0.5

212

(Hypotheses 1 and 2). Considering the changes in TSK and FABQ, we also hypothesized,

213

that the ES and SRM for those patients categorized as “improved” would be large (i.e. >

214

0.80) (Hypotheses 3) and for those categorized as “not improved” would be at least

215

medium (i.e. > 0.05) (Hypotheses 4). We considered an appropriate internal

216

responsiveness for the PHODA-SeV if at least three of the total four hypotheses (75%)

217

were supported.

17, 25

. Hence, for internal responsiveness, our

9


218

For external responsiveness, given that a previous study found a moderate correlation (r

219

= 0.37) between TSK and PHODA-SeV16, we hypothesized that the change in PHODA-

220

SeV would show at least moderate correlation with the changes scores of the TSK

221

(Hypothesis 1) or FABQ (Hypothesis 2). The magnitude of the Pearson correlation

222

coefficient was interpreted as: 0.00 to 0.25 little or no correlation; 0.25 to 0.50 fair

223

correlation; 0.50 to 0.75 moderate to good correlation; and more than 0.75 excellent

224

correlation29. We also hypothesized AUCs using the TSK (Hypothesis 3) and FABQ

225

(Hypothesis 4) as an external anchors will be equal or greater than 0.70. We considered

226

an appropriate external responsiveness for the PHODA-SeV if at least three of the total

227

four hypotheses (75%) are supported.

228

Interpretability. Interpretability is the degree to which one can transform a qualitative

229

meaning to quantitative scores. For the interpretability, we calculated the floor and ceiling

230

effects and the change scores for pain, disability and measures of pain-related fear. Floor

231

and ceiling effects were calculated through the percentage of participants that answered

232

the maximum score (ceiling effect) and the minimum scores (floor effect) of the

233

questionnaire. Ceiling and floor effects were considered to be present when more than

234

15% of the respondents answer the maximum and/or minimum score, respectively36.

235

All statistics analyses were performed using IBM SPSS version 20.0 (IBM corporation,

236

Somers, 205 NY, USA).

237

Results

238

From August 2014 to December 2015, 104 patients with chronic non-specific LBP were

239

recruited. Patients who not completed the 8-week treatment were excluded from the

240

analysis. Therefore, ninety-one patients completed the 8-week treatment period and were

241

included in the analyses. No missing items were identified in the questionnaires from

242

those patients included in the analyses. Nearly two thirds (n=62) of the sample was

243

female, with mean (SD) age of 37.8 (12.4) years and a median duration of pain of 18

244

months (interquartile range: 6.0 to 48.0). Table 1 describes the sample characteristics.

245

>

246

Internal Consistency, test-retest reliability and measurement error

10


247

The Cronbach’s α of the total score of the PHODA-SeV was 0.95, indicating homogeneity

248

among the items of the instrument. The total scores of PHODA-SeV demonstrated an

249

ICC(2,1) of 0.91 (95% CI: 0.86 to 0.94), indicating excellent reliability. The SEM and SDC

250

were of 5.46 and 15.13 points on a scale from 0 to 100, respectively. The reliability of

251

each activity of the PHODA-SeV ranged from 0.56 (i.e. Ironing while sitting) to 0.86 (i.e.

252

Walking up stairs), indicating good to excellent reliability. Table 2 describes the

253

reliability for each item of the PHODA-SeV.

254

>

255

Construct Validity

256

Table 3 shows the construct validity results. Our findings suggest a moderate validity for

257

the PHODA-SeV as there was a moderate correlation with one measure of pain-related

258

fear (i.e. FABQ).

259

>

260

Responsiveness

261

Table 4 shows the findings for the responsiveness analyses. For internal responsiveness,

262

large ES and SRM were found for the whole sample and for the subgroup of patients

263

categorised as “improved” and “not improved”. The only exception was found for

264

“improved” subgroup defined according to the changes in TSK which showed a medium

265

ES. All four priori-formulated hypotheses related to internal responsiveness were

266

supported. For external responsiveness, we found a low correlation between the changes

267

in TSK and FABQ and changes in the PHODA-SeV. In addition, the ROC curve analysis

268

considering the patients who improved and not improved in the TSK and FABQ showed

269

an AUC lower than the cut-off point of 0.70. None of the priori-formulated hypotheses

270

related to external responsiveness were supported.

271

>

272

Interpretability

273

Floor effects and ceiling effects were not detected for the PHODA-SeV. None of the

274

participants reported the highest and lowest score. Furthermore, table 5 presents the

275

comparison between before and after the 8-week treatment period for all measures.

11


276

>

277

Discussion

278

Our findings revealed that PHODA-SeV is a reliable tool to assess pain-related fear in

279

patients undergoing physical therapy treatment. Furthermore, our findings supported all

280

priori-formulated hypotheses related to internal responsiveness suggesting that this

281

instrument is able to detect change after an 8-week physical therapy treatment. The

282

PHODA-SeV showed, however, fair construct validity as well as poor external

283

responsiveness. Based on these results, we would recommend the PHODA-SeV as a

284

reliable tool to evaluate pain-related fear and detect change over time in patients with

285

chronic LBP in primary care settings.

286

The PHODA-SeV was found to provide consistent and reproducible results and showed

287

good validity with one of the pain-related fear measures investigated (i.e. FABQ). These

288

findings are consistent with previous studies investigating the psychometric properties of

289

PHODA-SeV in adults16 and adolescents37, which also found high reliability as well as

290

correlation with a pain-related fear measure. In regard to the construct validity of the

291

PHODA-SeV, our objective was to investigate which construct of pain-related fear would

292

be correlated with the instrument. Our results revealed that PHODA-SeV showed good

293

validity with FABQ (r = 0.39, p

Psychometric Properties of the Photograph Series of

Psychometric Properties of the Photograph Series of

Suggest Documents

PSYCHOMETRIC PROPERTIES OF HAWLEY'S

Psychometric properties of the Specific

Psychometric properties of assessment

PSYCHOMETRIC PROPERTIES of the PROCESSES of CHANGE ...

Psychometric properties of the Spanish version of

Assessment of the Psychometric Properties of E

Evaluation of the psychometric properties of the

Psychometric properties of the Spanish version of

Assessment of psychometric properties of the

Psychometric Properties of the Persian Version of

Psychometric properties of the Persian version of

The Psychometric Properties of the Italian ...

The psychometric properties of the English language

Examining the Psychometric Properties of the Antidepressant ...

The Psychometric Properties of the Korean

Revisiting the Psychometric Properties of Market Orientation ...

Psychometric Properties of the MacArthur ...

The Psychometric Properties of Menstrual Attitude ... - IJIRSET

Psychometric properties of the Looming Maladaptive ... - Psicothema

Psychometric Properties of the Multidimensional Pain

Psychometric Properties of the Montreal Cognitive ...

Psychometric Properties of the Wechsler Abbreviated

Psychometric Properties of the Social Skills

psychometric properties of the general self-efficacy