Content and psychometric evaluations of questionnaires for assessing physical function in people with low back disorders. A systematic review of the literature

Abstract

Purpose: The purpose was to investigate how physical function is assessed in people with musculoskeletal disorders in the low back. Specifically:

• Which questionnaires are used to assess physical function in people with musculoskeletal disorders in the low back?

• What aspects of physical function do those questionnaires measure?

• What are the measurement properties of the questionnaires?

Materials and methods: A systematic review was performed to identify questionnaires and psychometric evaluations of them. The content of the questionnaires was categorised according to the International Classification of Function, Disability and Health, and the psychometric evaluations were categorised using the COnsensus-based Standards for the selection of health Measurement Instruments (COSMIN) checklist.

Results: The questionnaires measured disability or ability to cope in everyday life, rather than physical function as such. Different aspects of a person’s mobility and ability to attend to one’s personal care were most often included regarding activity and participation. For body functions, items about sleep and pain were most often included. The Oswestry Disability Index and the Quebec Back Pain Disability Scale showed adequate psychometric properties in most evaluations.

Conclusions: The extent of psychometric evaluations differed substantially, as did the items included. Focus of measurement was predominantly on activities in daily life.

Implications for rehabilitation

• Valid and reliable instruments that measure relevant aspects of low back disorders are needed to provide early diagnostics and effective treatment.

• Most questionnaires need more psychometric evaluations to establish the quality.

• The Oswestry Disability Index and the Quebec Back Pain Disability Scale showed adequate psychometric properties in most evaluations.

• The results may be useful when making decisions about which measurement instruments to use when evaluating low back disorders.

Keywords:

• Questionnaire
• measure
• low back
• ICF
• psychometrics

Introduction

Patient reported outcome measures (PROMs) are promising tools for diagnosing a patient’s condition prior to treatment, and for evaluating intervention effects, since they provide necessary information without the need and considerable cost for staff to perform examinations or interviews. In musculoskeletal research and practice, a variety of self-reported measures have been used for this purpose [1–4]. The lack of consistency in selecting self-reported outcome measures may partly be explained by the diversity of symptoms and function in patients with musculoskeletal disorders in different regions. Nevertheless, the use of different PROMs in studies on similar populations makes it difficult to compare results, and thereby draw conclusions about, for example, which interventions are most effective. This problem has previously been addressed in several studies attempting to promote a more standardised use of outcome measures [5,6].

A closer examination of the items included in some questionnaires also shows that it is unclear which components and aspects of function and disability are actually being measured. According to the International Classification of Functioning, Disability and Health (ICF) [7], it may be components related to body functions and body structures, i.e., impairment level, or components related to activities and participation. To correctly diagnose patients and provide specific, targeted interventions for them, we need to distinguish physical function per se (e.g., ability to bend forward) from the consequences of daily life (e.g., ability to do housework). ICFs view of disability builds on the biopsychosocial model, which incorporates different perspectives on health, both the medical (as body functions and structure) and the social (as activity and participation) [8]. The definitions used are “the term functioning refers to all body functions, activities, and participation, while disability is similarly an umbrella term for impairments, activity limitations and participation restrictions” [8,p.2].

In addition to content, the quality of the instrument is important when deciding which instrument to use. Previous research shows a varying quality among questionnaires and a need to further study their psychometric properties [5,6]. A standardised set of criteria for determining which properties have been evaluated for an instrument is the COnsensus-based Standards for the selection of health Measurement Instruments (COSMIN checklist) [9–12]. The checklist also provides an opportunity to evaluate the validity and reliability of an instrument [3,13]. More details on the checklist are provided in the Materials and methods section.

A previous review of questionnaires for measuring physical function in people with musculoskeletal disorders in the neck found that the questionnaires differed substantially in items as well as the extent to which their psychometric properties had been evaluated [3]. Further, it was concluded that the focus of measurement was on activities in daily life rather than physical function as such [3]. In this study, PROMs intended and evaluated for measuring physical function in people with MSD in the low back were investigated. To delimit the review to studies performed on a homogeneous sample, we excluded studies where the questionnaires have been used in populations with symptoms due to trauma, infection, systemic diseases, and congenital or acquired deformities.

The aim of this systematic review was to investigate how physical function is assessed in people with MSD in the low back. Specifically, we aimed to determine:

• Which questionnaires are used to assess physical function in people with MSD in the low back?

• What aspects of physical function do those questionnaires measure?

• What are the measurement properties of the questionnaires?

Materials and methods

Search for questionnaires

In this systematic literature review [14,15] we searched the data bases PubMed, Cinahl, Web of Science, and PsycInfo for studies published in English until 12 September 2017. The keywords used in the search (Table 1) were defined according to the PICO model [16]. Thus, keywords were selected for P (population), I (intervention), O (outcome), while C (comparison) was excluded as comparative studies were not the focus of this study.

Table 1. Keywords sorted in word groups.

O (outcome)	I (intervention)	P (population)	P (population)
Physical function*	Measure*	Pain (not acute)	Musculoskeletal
Function*	Scale	–	Neck
Disability	Questionnaire*	–	Low back
–	Outcome measure*	–	Back
–	Inventory	–	–
–	Evaluation	–	–
–	Psychometric*	–	–
–	Reliability	–	–
–	Validity	–	–

Excluded: rheumatoid arthr*, gonarthros*, gout, cancer, foot, ankle, hip, knee, elbow, wrist, clinical trial, effect*, jaw, head*, trauma*, fear avoidance, and kinesiophobia.

Note. This table is previously published in Ref [3].

The systematic literature review process is illustrated in a flow chart (Figure 1). The literature search resulted in a total of 1922 articles; in PubMed 1333 articles, in Web of Science 457 articles, in PsycInfo 91 articles, and in Cinahl 41 articles. After excluding duplicates, the selection of relevant articles to include was performed by reviewing the title and abstracts first. At this stage, 284 articles were included from PubMed, 139 articles from Web of Science, 0 articles from Cinahl, and 0 articles from PsycInfo. Thereafter the abstracts, and when needed the full text, of the articles were reviewed to find articles presenting the original version of questionnaires for measurement of physical function in MSD. For this article, questionnaires aimed to measure physical function in people with MSD in the low back were selected. Both authors performed the review separately and then discussed the selection, to reach consensus. Questionnaires regarding general health, spinal stenosis, scoliosis, anchylosing spondylitis, surgical procedures, sciatica, ratings of functional capacity, work disability, not PROMs, and questionnaires for use among specific occupational groups or adolescents/children were excluded.

Figure 1. Flow chart; systematic literature review process.

A total of 98 different questionnaires for measurement of physical function in people with MSD were identified. Of these, 16 questionnaires focused on physical function in people with MSD in the low back and fulfilled our criteria for inclusion. These questionnaires were classified according to the ICF [7] to determine their focus of measurement and coverage of items. The ICF has previously been evaluated in different groups with varying results, from a low reliability and need for modification when tested in geriatric care [17] to acceptable reliability and reliability among people with osteoarthritis [18] and children with autism [19]. In this study, the ICF-classification was performed independently by both authors, and was followed by a comparative discussion to reach consensus.

Search for quality assessment of questionnaires

To identify the measurement properties that have been evaluated for the questionnaires, a second literature search in the previous databases was performed using the key words “validity OR reliability OR responsiveness AND low back AND name of questionnaire” to find articles that had quality-assessed the questionnaires. Only articles that analysed the original version of the questionnaire were included. Studies that met our previously used exclusion criteria were excluded, as well as studies concerning translated versions of the questionnaires. The translated versions were excluded to reduce the extent and complexity of the study.

Quality assessment of the questionnaires

The methodological quality of the questionnaires was examined with respect to relevant quality indicators of validity and reliability using the COSMIN checklist [10,12]. The COSMIN checklist contains standards for design and statistical methods to assess psychometric properties of PROMs. For further information go to the COSMIN checklist [20]. Relevant psychometric properties defined by the COSMIN group can be found on p. 9. The examination of psychometric properties was performed by both authors, first separately, then together to reach consensus regarding which properties of methodological quality had been assessed for each questionnaire, and the level of quality of the questionnaires. We rated the findings of the measurement properties as “adequate” or “not adequate” according to guidelines in the COSMIN checklist. Our quality criteria were identical to those used in the previous review of questionnaires measuring physical function in people with MSD in the neck [3], and are listed below.

• Internal consistency (adequate = unidimensional (sub)scale or Cronbach’s alpha(s) ≥ 0.70; not adequate = otherwise or cannot be determined).

• Reliability (adequate = intraclass correlation/weighted Kappa ≥0.70 or Pearson’s r ≥ 0.80; not adequate = otherwise or cannot be determined).

• Measurement error (adequate = standard error of measurement, smallest detectable change or limits of agreement presented; not adequate = otherwise or cannot be determined).

• Content validity (adequate = the target population was involved in the development of the questionnaire; not adequate = otherwise or cannot be determined).

• Construct validity (adequate = factors explained at least 50% of the variance; not adequate = otherwise or cannot be determined).

• Criterion validity (adequate = correlation with criterion instrument(s) ≥ 0.50; not adequate = otherwise or cannot be determined).

• Responsiveness (adequate = correlation with anchor instrument ≥0.50 or area under receiver operator curve ≥0.70; not adequate = otherwise or cannot be determined).

• Interpretability was marked “all” if the distribution of scores, floor/ceiling effects, scores for relevant groups and minimal important change or minimal important difference were reported, and “some” otherwise.

For criterion validity, only questionnaires (i.e., not single scales such as visual analogue scales) were considered as criterion instruments. If an instrument (either evaluated instrument or criterion instrument) contained several scales, the methodological quality was determined by taking the lowest rating of any of the scales.

Results

We identified 16 questionnaires that are used to assess physical function among people with low back pain [21–36]. These questionnaires are listed in Table 2.

Table 2. Included low back questionnaires.

Bournemouth Questionnaire (BQ) [24]

Curtain Back Screening Questionnaire (CBSQ) [33]

Extended Aberdeen Spine Pain Scale (EASPS) [36]

Istanbul Low Back Pain Disability Index (ILBPDI) [27]

Japanese Orthopedic Association Back Pain Evaluation Questionnaire (JOABPEQ) [30]

Low Back Outcome Score (LBOS) [32]

Low Back Pain Rating Scale (LBPRS) [35]

Low Back SF-36 Physical Functioning (LB SF-36 PF) [25]

Lumbar Spine Outcomes Questionnaire (LSOQ) [22]

Maine-Seattle Back Questionnaire (MSBQ) [21]

North American Spine Society Lumbar Spine Outcome Assessment Instrument (NASS) [26]

Oswestry Disability Index (ODI) [28]

ProFitMap-Low Back (PFM-LB) [23]

Quebec Back Pain Disability Scale (QBPDS) [34]

Spine Functional Index (SFI) [31]

Subjective Complaints Questionnaire for low back pain (SCQ) [29]

Questionnaire content

The classification of questionnaire items according to the ICF shows that most questionnaires only included items regarding the components body functions and activity and participation, with an overweight on items assessing activity and participation (Table 3). The Extended Aberdeen Spine Pain Scale, the Low Back Outcome Score and the Lumbar Spine Outcomes Questionnaire differed by also including items about environmental factors, and the Istanbul Low Back Pain Disability Index differed by only including items about activity and participation.

Table 3. ICF classification of the included low back pain questionnaires.

	ICF-categories
	Function and impairment																Contextual factors
	Body functions (b)							Activity participation (d)									Environmental factors (e)
	b1	b2	b3	b4	b5	b6	b7	d1	d2	d3	d4	d5	d6	d7	d8	d9	e1	e2	e3	e4	e5
Bournemouth Questionnaire (BQ)	140	280	–	–	–	–	–	–	–	–	410	510	640	760	859	910	–	–	–	–	–
	152	289	–	–	–	–	–	–	–	–	450	540	–	–	–	920
	–	–	–	–	–	–	–	–	–	–	460	–	–	–	–	–
Curtain Back Screening Questionnaire (CBSQ)	–	280	–	–	–	–	–	–	–	–	–	–	620	760		920	–	–	–	–	–
	–	–	–	–	–	–	–	–	–	–	–	–	640	799	–	–
Extended Aberdeen Spine Pain Scale (EASPS)	134	265	–	440	–	640	710	–	–	–	410	510	640	–	859	920	110	–	–	–	–
	–	280	–	–	–	–	720	–	–	–	415	540	–	–	–	–	–	–	–	–	–
	–	–	–	–	–	–	730	–	–	–	450	–	–	–	–	–	–	–	–	–	–
Istanbul Low Back Pain Disability Index (ILBPDI)	–	–	–	–	–	–	–	–	–	–	410	510	–	–	–	–	–	–	–	–	–
	–	–	–	–	–	–	–	–	–	–	415	540	–	–	–	–	–	–	–	–	–
	–	–	–	–	–	–	–	–	–	–	430	–	–	–	–	–	–	–	–	–	–
	–	–	–	–	–	–	–	–	–	–	450	–	–	–	–	–	–	–	–	–	–
	–	–	–	–	–	–	–	–	–	–	455	–	–	–	–	–	–	–	–	–	–
	–	–	–	–	–	–	–	–	–	–	460	–	–	–	–	–	–	–	–	–	–
	–	–	–	–	–	–	–	–	–	–	470	–	–	–	–	–	–	–	–	–	–
Japanese Orthopedic Association Back Pain Evaluation Questionnaire (JOABPEQ)	126	280	–	–	–	–	–	–	–	–	410	540	640	–	859	–	–	–	–	–	–
	130	–	–	–	–	–	–	–	–	–	415	–	–	–	–	–	–	–	–	–	–
	134	–	–	–	–	–	–	–	–	–	450	–	–	–	–	–	–	–	–	–	–
	152	–	–	–	–	–	–	–	–	–	460	–	–	–	–	–	–	–	–	–	–
Low Back Outcome Score (LBOS)	134	280	–	–	–	640	–	–	–	–	450	540	640	–	859	920	110	–	355	–	–
	–	–	–	–	–	–	–	–	–	–	470	–	–	–	–	–	–	–	–	–	–
Low Back Pain Rating Scale (LBPRS)	126	–	–	–	–	–	–	–	–	–	410	540	640	760	845	–	–	–	–	–	–
	134	–	–	–	–	–	–	–	–	–	430	–	–	799	859	–	–	–	–	–	–
	–	–	–	–	–	–	–	–	–	–	450	–	–	–	–	–	–	–	–	–	–
	–	–	–	–	–	–	–	–	–	–	455	–	–	–	–	–	–	–	–	–	–
	–	–	–	–	–	–	–	–	–	–	460	–	–	–	–	–	–	–	–	–	–
	–	–	–	–	–	–	–	–	–	–	475	–	–	–	–	–	–	–	–	–	–
Low Back SF-36 Physical Functioning (LB SF-36 PF)	134	280	–	–	–	–	–	–	–	–	410	510	640	–	–	920	–	–	–	–	–
	–	–	–	–	–	–	–	–	–	–	415	540	–	–	–	–	–	–	–	–	–
	–	–	–	–	–	–	–	–	–	–	430	–	–	–	–	–	–	–	–	–	–
	–	–	–	–	–	–	–	–	–	–	449	–	–	–	–	–	–	–	–	–	–
	–	–	–	–	–	–	–	–	–	–	450	–	–	–	–	–	–	–	–	–	–
	–	–	–	–	–	–	–	–	–	–	455	–	–	–	–	–	–	–	–	–	–
	–	–	–	–	–	–	–	–	–	–	460	–	–	–	–	–	–	–	–	–	–
	–	–	–	–	–	–	–	–	–	–	470	–	–	–	–	–	–	–	–	–	–
Lumbar Spine Outcomes Questionnaire (LSOQ)	130	265	–	–	525	620	730	–	–	–	410	510	640	–	850	910	110	–	–	–	570
	134	280	–	–	–	640	–	–	–	–	415	540	–	–	–	920	–	–	–	–	580
	152	–	–	–	–	–	–	–	–	–	430	550	–	–	–	930	–	–	–	–	–
	–	–	–	–	–	–	–	–	–	–	450	–	–	–	–	–	–	–	–	–	–
	–	–	–	–	–	–	–	–	–	–	460	–	–	–	–	–	–	–	–	–	–
	–	–	–	–	–	–	–	–	–	–	470	–	–	–	–	–	–	–	–	–	–
	–	–	–	–	–	–	–	–	–	–	475	–	–	–	–	–	–	–	–	–	–
Maine-Seattle back questionnaire (MSBQ)	126	280	–	–	–	640	–	–	–	–	410	540	640	–	–	999	–	–	–	–	–
	134	–	–	–	–	–	–	–	–	–	415	–	–	–	–	–	–	–	–	–	–
	152	–	–	–	–	–	–	–	–	–	450	–	–	–	–	–	–	–	–	–	–
	–	–	–	–	–	–	–	–	–	–	460	–	–	–	–	–	–	–	–	–	–
North American spine society lumbar spine outcome assessment instrument (NASS)	134	280	–	–	–	640	–	–	–	–	415	540	–	–	–	920	–	–	–	–	–
	–	–	–	–	–	–	–	–	–	–	430	–	–	–	–	–	–	–	–	–	–
	–	–	–	–	–	–	–	–	–	–	450	–	–	–	–	–	–	–	–	–	–
	–	–	–	–	–	–	–	–	–	–	455	–	–	–	–	–	–	–	–	–	–
	–	–	–	–	–	–	–	–	–	–	470	–	–	–	–	–	–	–	–	–	–
OSWESTRY Disability Questionnaire	134	280	–	–	–	670	–	–	–	–	415	510	–	750	–	–	–	–	–	–	–
(ODI)	–	–	–	–	–	–	–	–	–	–	430	540	–	–	–	–	–	–	–	–	–
	–	–	–	–	–	–	–	–	–	–	450	–	–	–	–	–	–	–	–	–	–
	–	–	–	–	–	–	–	–	–	–	470	–	–	–	–	–	–	–	–	–	–
ProFitMap-Low Back (PFM-LB)	134	265	–	–	525	620	720	–	–	–	410	540	–	–	845	–	–	–	–	–	–
	152	280	–	–	535	640	730	–	–	–	415	–	–	–	859	–	–	–	–	–	–
	–	289	–	–	–	–	755	–	–	–	430	–	–	–	–	–	–	–	–	–	–
	–	–	–	–	–	–	760	–	–	–	445	–	–	–	–	–	–	–	–	–	–
	–	–	–	–	–	–	770	–	–	–	450	–	–	–	–	–	–	–	–	–	–
	–	–	–	–	–	–	780	–	–	–	455	–	–	–	–	–	–	–	–	–	–
	–	–	–	–	–	–	789	–	–	–	460	–	–	–	–	–	–	–	–	–	–
	–	–	–	–	–	–	–	–	–	–	465	–	–	–	–	–	–	–	–	–	–
Quebec Back Pain Disability Scale (QBPDS)	134	–	–	–	–	–	–	–	–	–	410	540	630	–	–	–	–	–	–	–	–
	–	–	–	–	–	–	–	–	–	–	415	–	640	–	–	–	–	–	–	–	–
	–	–	–	–	–	–	–	–	–	–	430	–	–	–	–	–	–	–	–	–	–
	–	–	–	–	–	–	–	–	–	–	445	–	–	–	–	–	–	–	–	–	–
	–	–	–	–	–	–	–	–	–	–	450	–	–	–	–	–	–	–	–	–	–
	–	–	–	–	–	–	–	–	–	–	455	–	–	–	–	–	–	–	–	–	–
	–	–	–	–	–	–	–	–	–	–	460	–	–	–	–	–	–	–	–	–	–
	–	–	–	–	–	–	–	–	–	–	470	–	–	–	–	–	–	–	–	–	–
Spine Functional Index (SFI)	130152	280	–	–	–	–	–	–	–	–	410	510	640	–	–	–	–	–	–	–	–
	–	–	–	–	–	–	–	–	–	–	415	520	–	–	–	–	–	–	–	–	–
	–	–	–	–	–	–	–	–	–	–	470	–	–	–	–	–	–	–	–	–	–
	–	–	–	–	–	–	–	–	–	–	475	–	–	–	–	–	–	–	–	–	–
Subjective Complaints	134	265	–	–	–	–	730	–	–	–	410	–	–	–	845	–	–	–	–	–	–
Questionnaire for low back pain (SCQ-LBP)	–	280	–	–	–	–	–	–	–	–	415	–	–	–	–	–	–	–	–	–	–
	–	298	–	–	–	–	–	–	–	–	430	–	–	–	–	–	–	–	–	–	–
	–	299	–	–	–	–	–	–	–	–	450	–	–	–	–	–	–	–	–	–	–
	–	–	–	–	–	–	–	–	–	–	470	–	–	–	–	–	–	–	–	–	–

Body functions (b) – b1: mental functions; b2: sensory functions and pain; b3: voice and speech; b4: cardio-vascular-, haematological-, immunological-, and respiratory functions; b5: digestive-, metabolic-, and endocrine functions; b6: gender-, urinary-, and reproductive functions; b7: neuromusculoskeletal and movement related functions.

Activity and participation (d) – d1: learning and applying knowledge; d2: general tasks and demands; d3: communication; d4: mobility; d5: personal care; d6: domestic life; d7: interpersonal interactions and relationships; d8: important areas of life; d9: community, social, and civic life.

Environmental factors (e) – e1: products and technology; e2: natural environment and human-made environmental changes; e3: support and relationships; e4: attitudes; e5: services, systems, and policies.

Note: If several items of a questionnaire concerned the same aspect, it was only categorized once. If an item included several aspects, all aspects were listed in the table.

Within the component body functions, most questionnaires included items assessing “mental functions”, “sensory functions and pain”, “gender-, urinary- and reproductive functions”, and “neuromusculoskeletal and movement related functions.” For mental functions, the most common item assessed was “sleep functions” (b134). For sensory functions and pain, “pain” (b280) was most commonly assessed. For gender-, urinary- and reproductive functions, “sexual functions” (b640) was most common, while items concerning “muscle power” (b730) were most frequently used to assess neuromusculoskeletal and movement-related functions.

Within the component activity and participation, items regarding mobility were most common, such as “walking” (d450), “maintaining a body position” (d415) and “changing basic body position” (d410). Also rather common were items regarding personal care, such as “dressing” (d540), followed by items regarding domestic life such as “doing housework” (d640), community, social and civic life as “recreation and leisure” (d920), and important areas of life as “work and employment” (d859). The questionnaires with widest coverage of items in different areas were the Lumbar Spine Outcomes Questionnaire [22], the Extended Aberdeen Spine Pain Scale [36] and the Low Back Outcome Score [32] (See Table 3). An overview of all ICF-codes that were included in the analysed questionnaires can be found in Table 4.

Table 4. ICF codes with explanations.

b: Body functions	d: Activities and participation	e: Environmental factors
1 MENTAL FUNCTIONS	4 MOBILITY	1 PRODUCTS AND TECHNOLOGY
126 Temperament and personality functions	410 Changing basic body position	110 Products or substances for personal consumption
130 Energy and drive functions	415 Maintaining a body position	3 SUPPORT AND RELATIONSHIPS
134 Sleep functions	430 Lifting and carrying objects	355 Health professionals
140 Attention functions	445 Hand and arm use	5 SERVICES, SYSTEMS, AND POLICIES
152 Emotional functions	449 Carrying, moving and handling objects, other specified and unspecified	570 Social security services, systems and policies
2 SENSORY FUNCTIONS AND PAIN	450 Walking	580 Health services, systems and policies
265 Touch function	455 Moving around	–
280 Sensation of pain	460 Moving around in different locations	–
289 Sensation of pain, other specified and unspecified	465 Moving around using equipment	–
298 Sensory functions and pain, other specified	470 Using transportation	–
299 Sensory functions and pain, unspecified	475 Driving	–
4 FUNCTIONS OF THE CARDIOVASCULAR, HAEMATOLOGICAL, IMMUNOLOGICAL AND RESPIRATORY SYSTEMS	5 SELF-CARE	–
440 Respiration functions	510 Washing oneself	–
5 FUNCTIONS OF THE DIGESTIVE, METABOLIC AND ENDOCRINE SYSTEMS	520 Caring for body parts	–
525 Defecation functions	540 Dressing	–
535 Sensations associated with the digestive system	550 Eating	–
6 GENITOURINARY AND REPRODUCTIVE FUNCTIONS	6 DOMESTIC LIFE	–
620 Urination functions	620 Acquisition of goods and services	–
640 Sexual functions	640 Doing housework	–
670 Sensations associated with genital and reproductive functions	7 INTERPERSONAL INTERACTIONS AND RELATIONSHIPS	–
7 NEUROMUSCULOSKELETAL AND MOVEMENT-RELATED FUNCTIONS	750 Informal social relationships	–
710 Mobility of joint functions	760 Family relationships	–
720 Mobility of bone functions	799 Interpersonal interactions and relationships, unspecified	–
730 Muscle power functions	8 MAJOR LIFE AREAS	–
755 Involuntary movement reaction functions	845 Acquiring, keeping, and terminating a job	–
760 Control of voluntary movement functions	850 Remunerative employment	–
770 Gait pattern functions	859 Work and employment, other specified and unspecified	–
780 Sensations related to muscles and movement functions	9 COMMUNITY, SOCIAL AND CIVIC LIFE	–
789 Movement functions, other specified and unspecified	910 Community life	–
–	920 Recreation and leisure	–
–	930 Religion and spirituality	–
–	999 Community, social and civic life, unspecified	–

Bold text = Main categories.

Methodological quality

In the search for studies evaluating measurement properties of the 16 questionnaires, we identified 27 articles that fulfilled the inclusion criteria [21–48]. Table 5 shows the extent to which each questionnaire has been evaluated. The methodological quality of the questionnaires is summarised in Table 6.

Table 5. Mapping of psychometric assessment of the included questionnaires.

	Reliability			Validity
	Internal consistency	Reliability	Measurement error	Content validity	Construct validity	Criterion validity	Responsiveness	Interpretability
BQ	[24]	–	[24]	[24]	–	[24]	[24,46]	[24, 46]
CBSQ	[33]	[33]	–	[33]	–	[33]	–	[33]
EASPS	[36]	[36]	–	[36]	[36]	[36]	[36]	[36]
ILBPDI	[27]	[27]	–	[27]	[27]	[27]	–	[27]
JOABPEQ	–	[30]	–	[30]	[40]	–	–	[30, 40]
LBOS	[43]	[43]	–	–	–	[32]	[47]	[32, 43, 47]
LBPRS	–	[35]	–	–	[35]	–	[44]	[35, 44]
LB SF-36 PF	[25]	[25,37]	[25,37]	–	–	–	[25,37,41]	[25, 37, 41]
LSOQ	–	[22]	–	–	–	[22]	[22]	[22]
MSBQ	[21]	[21]	–	[21]	–		[21]	[21]
NASS	[26,45]	[26]	–	[26]	[26,45,48]	[26]	–	[26, 45, 48]
ODI	–	[37,39]	[37,39]	–	–	[39, 42]	[37,39,41,42,44,47]	[37, 39, 41, 42, 44, 47]
PFM-LB	[23]			[23]	–	[23]	–	[23]
QBPDS	[34]	[34,37,39]	[37–39]	[34]	[34]	[39]	[37,38,39]	[34, 37–39]
SFI	[31]	[31]	[31]	[31]	[31]	[31]	[31]	[31]
SCQ	–	[29]	–	[29]	–	–	–	[29]

BQ: Bournemouth Questionnaire; CBSQ: Curtain Back Screening Questionnaire; EASPS: Extended Aberdeen Spine Pain Scale; ILBPDI: Istanbul Low Back Pain Disability Index; JOABPEQ: Japanese Orthopedic Association Back Pain Evaluation Questionnaire; LBOS: Low Back Outcome Score; LBPRS: Low Back Pain Rating Scale; LB SF-36 PF: Low Back SF-36 Physical Functioning; LSOQ: Lumbar Spine Outcomes Questionnaire; MSBQ: Maine-Seattle Back Questionnaire; NASS: North American spine society lumbar spine outcome assessment Instrument; ODI: Oswestry Disability Index; PFM-LB: ProFitMap-Low Back; QBPDS: Quebec Back Pain Disability Scale; SFI: Spine Functional Index; SCQ: Subjective Complaints Questionnaire for low back pain.

Table 6. Methodological quality of the included questionnaires.

		Reliability				Validity
	Article	Internal consistency	Reliability	Measurement error	Content validity	Construct validity	Criterion validity	Responsiveness	Interpretability
BQ	[24]	+	–	+	+	–	MSBQ − FABQ + Zung + PLC −	–	Some
	[46]	–	–	–	–	–	–	–	Some
CBSQ	[33]	+	–	–	+	–	SIP +	–	Some
EASPS	[36]	+	–	–	–	–	SF-12 −	–	Some
ILBPDI	[27]	+	+	–	+	–	QBPDS + ODI + Wadell +	–	Some
JOABPEQ	[30]	–	–	–	–	–	–	–	Some
	[40]	–	–	–	–	+	–	–	Some
LBOS	[32]	–	–	–	–	–	ODI + Wadell +	–	Some
	[47]	–	–	–	–	–	–	+	Some
	[43]	+	+	–	–	–	–	–	Some
LBPRS	[35]	–	–	–	–	+	–	–	Some
	[44]	–	–	–	–	–	–	+	Some
LB SF-36 PF	[25]	+	+	+	–	–	–	–	Some
	[37]	–	+	+	–	–	–	+	Some
	[41]	–	–	–	–	–	–	–	Some
LSOQ	[22]	–	+	–	–	–	ODI − SF-36 −	–	Some
MSBQ	[21]	+	+	–	–	–	–	+	All
NASS	[48]	–	–	–	–	+	–	–	None
	[26]	+	+	–	+	–	SF-36 −	–	Some
	[45]	+	–	–	–	+	–	–	Some
ODI	[37]	–	+	+	–	–	–	+	Some
	[44]	–	–	–	–	–	–	+	Some
	[42]	–	–	–	–	–	LCAT −	+	Some
	[47]	–	–	–	–	–	–	+	Some
	[41]	–	–	–	–	–	–	+	Some
	[39]	–	+	+	–	–	QBPDS +	+	Some
PFM-LB	[23]	+	–	–	+	–	ABPS + LBOS + Wadell + RMQ + SF36 −	–	Some
QBPDS	[34]	+	–	–	+	+	–	–	Some
	[38]	–	–	+	–	–	–	+	Some
	[39]	–	–	+	–	–	ODI −	+	Some
	[37]	–	+	+	–	–	–	+	Some
SFI	[31]	+	+	+	+	–	FRI +	–	Some
SCQ	[29]	–	+	–	+	–	–	–	Some

+: adequate; −: not adequate.

For further information, see “Quality assessment of the questionnaires” in the Methods section.

BQ: Bournemouth Questionnaire; CBSQ: Curtain Back Screening Questionnaire; EASPS: Extended Aberdeen Spine Pain Scale; ILBPDI: Istanbul Low Back Pain Disability Index; JOABPEQ: Japanese Orthopedic Association Back Pain Evaluation Questionnaire; LBOS: Low Back Outcome Score; LBPRS: Low Back Pain Rating Scale; LB SF-36 PF: Low Back SF-36 Physical Functioning; LSOQ: Lumbar Spine Outcomes Questionnaire; MSBQ: Maine-Seattle Back Questionnaire; NASS: North American spine society lumbar spine outcome assessment Instrument; OD: Oswestry Disability Index; PFM-LB: ProFitMap-Low Back; QBPDS: Quebec Back Pain Disability Scale; SFI: Spine Functional Index; SCQ: Subjective Complaints Questionnaire for low back pain.

While the majority of the questionnaires have been evaluated in few studies, the evaluations tend to cover most properties of the COSMIN checklist. The most frequently evaluated questionnaires were the Oswestry Disability Index and the Quebec Back Pain Disability Scale (in 6 and 4 studies, respectively). The latter questionnaire has also been evaluated for all properties in the checklist. Notably, test–retest reliability has been assessed for almost all questionnaires, while measurement error has been assessed for only 5 of the 16 questionnaires.

Internal consistency was found adequate in all of the studies investigating the property. For all but North American Spine Society lumbar spine outcome assessment instrument; however, it was only evaluated in a single study. Test-retest reliability varied between the questionnaires. For Low Back SF-36 Physical Functioning and Oswestry Disability Index, their reliability was confirmed in two studies each. Measurement error was reported adequately for the Bournemouth Questionnaire (in one study), Low Back SF-36 Physical Functioning (in two studies), Oswestry Disability Index (in two studies), Quebec Back Pain Disability Scale (in three studies), and Spine Functional Index (in one study). For half of the questionnaires, content validity was found adequate, and for 5 of the 16 questionnaires it was not reported in any of the studies investigated. When assessing construct validity, we found that factor analysis of the construct to be measured had been performed for few questionnaires, and generally only in one study. The exception was the North American Spine Society lumbar spine outcome assessment instrument, where it had been performed in three studies, and the results were inconsistent.

For criterion validity, the findings were rated for each of the criterion instruments listed in Table 6. The two most commonly used criterion instruments were Oswestry Disability Index and the Short Form 12 or 36 Item Health Survey (SF-12/SF-36). For Oswestry Disability Index, the results were inconsistent. Its correlation with Istanbul Low Back Pain Disability Index and the Low Back Outcome Score was high, while its correlation with Lumbar Spines Outcome Questionnaire and Quebec Back Pain Disability Scale was low. For SF-12/SF-36, the criterion validity towards other questionnaires was low. The Waddell questionnaire was used as criterion instrument in three studies, showing high correlations with Istanbul Low Back Pain Disability Index, Low Back Outcome Score, and ProFitMap-Low Back. Responsiveness has been evaluated for 10 of the 16 questionnaires, and was consistently found to be adequate for Oswestry Disability Index and Quebec Back Pain Disability Scale.

In addition to the ratings of quality, we assessed the degree to which important characteristics relating to the instrument’s interpretability was reported in the studies. Almost all studies reported some characteristics of the instrument, but only one study evaluating Maine-Seattle Back Questionnaire reported all characteristics considered important.

Discussion

Questionnaire content

Based on our findings, we can conclude that the questionnaires tend to measure disability or the ability to cope in everyday life, rather than physical function as such. The same result was found in our previous review of the content of neck questionnaires [3]. A few previous studies have reviewed the content of low back questionnaires and classified it according to ICF [6,49–51]. However, comparisons do not come straightforward, as there are differences in design between the studies. Due to different search strategies (e.g., different keywords, or selection of widely used questionnaires) and/or different target samples, different instruments are evaluated in the studies. In the review by Grotle et al. [6], 6 of the 36 included questionnaires were also identified in this study. For Wang et al. [49], 6 of the included 15 questionnaires overlapped with this study, for Oner et al. [50], only 2 of the included 17 questionnaires were included in this study, and for Sigl et al. [51] 2 of the 3 were included. The corresponding questionnaires were Low Back Outcome Score [6,49,50], Oswestry Disability Index [6,49–51], North American Spine Society lumbar spine outcome assessment instrument [6,49,51], Bournemouth Questionnaire, Quebec Back Pain Disability Scale [6,49,51], Low Back Pain Rating Scale [6], and Japanese Orthopedic Association Back Pain Evaluation Questionnaire [49].

When comparing the ICF categorisations that have been performed in the different reviews and this study, some similarities and differences in the results are noticeable. Among the questionnaires analysed in this study and by Wang et al. [49], the classifications of the Quebec Back Pain Disability Scale corresponded the most, as Wang et al. only had one category that was not included in our classification (d210 Undertaking a single task), we had only two that were not included in theirs (d460 Moving around in different locations, d630 Preparing meals), while the remaining 10 where the same. For the Japanese Orthopedic Association Back Pain Evaluation Questionnaire, the correspondence was less: nine categories were the same, while three were only identified in this study (b126 Temperament and personality functions, d460 Moving around in different locations, d859 Work and employment, other specified and unspecified) and four identified only by Wang et al. [49] (d230 carrying out daily routine, d455 moving around, d845 acquiring, keeping and terminating a job, e3 support, and relationships). The classifications of the Bournemouth Questionnaire corresponded for six categories, while there were eight categories specific to this study (b140 attention functions, b289 sensation of pain, other specified and unspecified, d410 changing basic body position, d450 walking, d460 moving around in different locations, d760 family relations, d859 work and employment, other specified and unspecified, d910 community life) and five to the study by Wang et al. [49] (d210 undertaking a single task, d430 lifting and carrying objects, d560 drinking, d845 acquiring, keeping and terminating a job, d850 remunerative employment).

In the three reviews considering the North American Spine Society lumbar spine outcome assessment instrument, eight categories corresponded between the classification of Wang et al. [49], Sigl et al. [51] and this study, while two categories were unique in this study (b640 sexual functions, d470 using transportation), and three in the study by Sigl et al. [51] (b265 touch function, b730 muscle power functions, and e115 products and technology for personal use in daily living). Unfortunately, it was not possible to compare ICF categories with the review by Grotle et al. [6], since their ICF-classification was not presented in sufficient detail. The best correspondence between our ICF-categorisation and that of previous reviews was obtained for the Quebec Back Pain Disability Scale [49] and the North American Spine Society lumbar spine outcome assessment instrument [49,51], while the largest differences were found for the Low Back Outcome Score [49,50] and the Oswestry Disability Index [49–51].

The differences in findings are somewhat surprising, since all studies have used the same ICF model. A closer look at the reviews showed that Wang et al. [49], Oner et al. [50], and Sigl et al. [51] refer to linking rules described by Cieza et al. [52,53]. Initially, 10 linking rules were recommended in an effort to systematise and standardise the linking of items in PROMs to the ICF [52]. Later, these linking rules were redefined and reduced to four [53]. It is possible that the use of the linking rules, or different versions of the linking rules, could explain some of the differences in findings between the studies. In this study, and in the study by Grotle et al. [6], the linking rules were not applied. Since one aspect of the linking rules is the categorisation of the cause (e.g., pain) as well as the impairment (e.g., ability to walk) in an item comprising both, and we aimed to identify primarily the aspect of physical function covered, e.g., limited ability to walk because of pain, we categorised only the impairment part of such items. On an item level, we believe it is a more valid use of the ICF categorisation to distinguish between items measuring pain only (which belong to body functions, sensory functions, and pain), from those measuring the impairment (e.g., ability to walk, which belong to activity and participation, mobility). On the other hand, additional categorisations of several aspects that constitute clarifying examples of a concept do not reduce the validity, as they provide more information regarding the same domain (e.g., activity and participation). Another noticeable difference between categorisations in different studies was the level of detail. For example, Wang et al. [49] identified the categories “undertaking a simple task” (d 210) and “carrying out a daily task” (d 230) in the Bournemouth Questionnaire, Japanese Orthopedic Association Back Pain Evaluation Questionnaire, Quebec Back Pain Disability Questionnaire, and Low Back Outcome Score. Since these are quite general categories, they may have been more precisely categorised in the other studies. More detailed investigations of how the formulation of questions in the questionnaires with the highest correspondence in ICF classification differs from those with the lowest may be warranted.

Methodological quality

Among the questionnaires investigated in this study, the most frequently evaluated questionnaires were the Oswestry Disability Index and the Quebec Back Pain Disability Scale. The responsiveness of these questionnaires was found adequate in repeated studies. For the Quebec Back Pain Disability Scale, measurement error was reported in the studies assessing responsiveness. Remaining quality aspects of the Oswestry Disability Index and the Quebec Back Pain Disability Scale were evaluated in no or few studies, or found inconsistent between studies. We recommend that they be evaluated for more quality aspects in more studies, to increase the utility of the questionnaires.

Unfortunately, the questionnaires with the widest coverage in items were evaluated for few quality aspects in few studies. In our literature search, we found only one study evaluating the Lumbar Spine Outcomes Questionnaire and the Extended Aberdeen Spine Pain Scale, respectively. Although the Low Back Outcome Score was evaluated in three studies, each of the quality aspects (i.e., internal consistency, reliability, criterion validity, and responsiveness) was evaluated in a single study. Thus, more studies of the psychometric properties of these questionnaires would be needed to appreciate the benefit of using them in research and practice.

Previous reviews have also found that the Oswestry Disability Index shows good responsiveness and, although evaluated to a lesser extent, test–retest reliability [6,50]. Further, others have concluded that the Low Back Outcome Score shows promising psychometric properties, but they need to be confirmed in more studies [6,50]. In this review, and in the review by Grotle et al. [6], the Quebec Back Pain Disability Scale showed good responsiveness. However, contrary to Grotle et al., we found inconsistent results with respect to test–retest reliability. This difference in findings may depend on the criteria used for determining adequate reliability. For the Bournemouth Questionnaire, we did not find acceptable responsiveness, as concluded by Grotle et al. [6]. Both reviews found good construct validity for the Low Back Pain Rating Scale, albeit evaluated in only one study in this review. In terms of test-retest reliability, however, our results differed. For the North American Spine Society lumbar spine outcome assessment instrument, internal consistency, and test–retest reliability were found adequate in this review as well as in Grotle et al. [6], while our findings with respect to construct validity were inconsistent. Overall, the number of studies evaluating the questionnaires was few. Hence, the findings and comparisons between reviews should be interpreted with caution.

Conclusions

Ultimately, the choice of questionnaire(s) depends on what you want to measure. This classification of content and mapping of psychometric properties of questionnaires used to measure physical function in people with MSD in the low back may be used as guidance in selecting the appropriate measure for a study. At present, the Oswestry Disability Index and the Quebec Back Pain Disability Scale are the most frequently evaluated questionnaires, showing adequate psychometric properties in most evaluations. Based on their coverage of items, they may be appropriate to use when attempting to assess activity and participation. Questionnaires with more weight on body functions, such as the Lumbar Spine Outcomes Questionnaire, the Extended Aberdeen Spine Pain Scale, and the ProFitMap-Low Back need to be evaluated in more studies before definite recommendations can be made. The COSMIN checklist [10,12] may constitute a useful tool when planning these evaluations.

Disclosure statement

The authors declare that they have no conflict of interest.