Journal Logo

Review Article

Psychometric properties and domains of postural control tests for individuals with knee osteoarthritis: a systematic review

French, Helen P.a; Hager, Charlotte K.b; Venience, Annec; Fagan, Ryana; Meldrum, Darad

Author Information
International Journal of Rehabilitation Research: June 2020 - Volume 43 - Issue 2 - p 102-115
doi: 10.1097/MRR.0000000000000403
  • Free

Abstract

Introduction

Knee osteoarthritis (OA) is a common chronic joint condition, which can cause pain, limit everyday activities and reduce quality of life. It is associated with increasing age and lifetime prevalence is approximately 40% for men and 47% for women (Murphy et al., 2008). Individuals with knee OA have demonstrated impaired postural control compared to those without OA (Wegener et al., 1997; Hassan et al., 2001; Hinman et al., 2002; Hsieh et al., 2013; Khalaj et al., 2014), which may be a risk factor for falls (Manlapaz et al., 2019). Increased risk of falls in individuals with knee OA has been identified compared to those without OA (Doré et al., 2015) with higher levels of fear of falling reported (French et al. 2016). Various neuromuscular deficits occur in OA including decreased quadriceps strength (Hassan et al., 2001; Hunt et al., 2010; Slemenda et al., 1997; Sturnieks et al., 2004; Takacs et al., 2015), reduced proprioception (Hurley et al., 1998; Hortobagyi et al., 2004; Sturnieks et al., 2004), knee malalignment (Hunt et al., 2010), reduced range of motion (ROM) (Takacs et al., 2015) and increased levels of pain (Hall et al., 2006; Hunt et al., 2010), and these are associated with falls risk in people with OA (Sturnieks et al., 2004; Takacs et al., 2015). These OA-related impairments may be compounded by age-related changes in balance (Laughton et al., 2003), strength (Frontera et al., 2000) and proprioception (Petrella et al., 1997; Takacs et al., 2015). Other relevant factors for knee OA which can increase the risk of falls include co-morbidities (van Raaij et al., 2010), opioid use and antidepressants (Lo-Ciganic et al., 2017). As life-expectancy increases (Kontis et al., 2017), falls risk will rise with a resultant significant impact on healthcare costs, unless falls prevention becomes a focus of management in the clinical setting (Houry et al., 2016).

Postural control entails maintaining, achieving or restoring a state of balance during any posture or activity (Pollock et al., 2000) for stability and orientation (Shumway-Cook and Woollacott, 2001). Static postural control refers to the ability to maintain equilibrium in a static position such as sitting or standing, whilst dynamic postural control relates to the maintenance of stability during movement and response to postural perturbations that occur during everyday activity (Takacs et al., 2013). Postural control is a complex motor skill requiring interaction of numerous sensorimotor processes to stabilise the centre of mass with respect to the support surface, the external environment and internal references (Horak, 2006). Consequently, it is not easy to classify or assess postural control as it comprises multiple sub-components as described by Horak (2006) (Table 1). The exact demands on the postural control system are influenced by the task itself and the environment in which it is performed (Huxham et al., 2001) and requires integration of the visual, vestibular, somatosensory and neuromuscular systems (Pollock et al., 2000). A Systems Framework for postural control recognises its complexity as a motor skill and the range of sensorimotor processes involved (Horak, 2006). It describes six major domains required to maintain postural control (Horak, 2006) which were adapted into nine operational definitions (Table 1). A scoping review by Sibley et al. (2015) identified 66 different standardised measures of postural control developed for use in adult populations. This framework can be used to determine which domains are contained in measures of postural control, and thereby ascertain the suitability of the measures for evaluation of postural control (Sibley et al., 2015).

Table 1
Table 1:
Domains of postural control (Horak, 2006 ; Sibley et al., 2015)

Assessment of postural control

Comprehensive assessment of postural control is required to identify impairments, inform appropriate management strategies and evaluate treatment effectiveness. Many measurement methods of postural control exist, extending from laboratory-based measurement systems such as force plates to methods which can take place in a clinical setting without specialised equipment (Tinetti, 1986; Berg et al., 1989; Podsiadlo and Richardson, 1991; Duncan et al., 1992; Horak et al., 2009).

A systematic review in 2015 identified 15 randomised controlled trials (RCTs) that investigated physical therapies for improving balance and reducing falls risk in OA. Whilst strengthening, aerobic exercise and Tai Chi resulted in improved balance outcomes, there was heterogeneity of assessment tools used and the authors recommended a consensus on standardised assessment tools in measuring balance and falls risk to allow structured comparisons between studies (Mat et al., 2015).

Assessment tools of postural control should be reliable, valid, responsive to change and interpretable for use in clinical and research contexts. The COnsensus-based Standards for the selection of health Measurement Instruments (COSMIN) initiative, comprises an international multidisciplinary team of researchers, with expertise in development and evaluation of outcome measurement instruments. The standards state that content validity is the most important psychometric property as it should contain items that are relevant, comprehensive, and comprehensible with respect to the construct of interest and target population (COSMIN, 2019). Internal validity, which includes structural validity, internal consistency and cross-cultural validity, is deemed next most important. The final properties considered important by COSMIN are reliability, measurement error, criterion validity, hypotheses testing for construct validity and responsiveness (COSMIN, 2019).

Based on the COSMIN framework, our aim was to systematically appraise the evidence on psychometric properties of postural control assessment tools used in populations with knee OA. A secondary aim was to identify what postural control domains are measured by the tools, using the Systems Framework for Postural Control (Horak, 2006; Sibley et al., 2015).

Methods

A protocol for this systematic review was registered with the PROSPERO database (CRD42016034108). This review was conducted in accordance with the preferred reporting items of systematic reviews and meta-analyses (PRISMA) reporting guidelines (Moher et al., 2010).

Search strategy

The following databases were searched from inception to January 2019: PubMed, CINAHL, EMBASE and Web of Science. The search terms were constructed using the following three themes: osteoarthritis, postural control and measurement. Related terms were combined using the Boolean ‘OR’ and all themes were then combined using the Boolean ‘AND’. Google Scholar was searched for additional citations associated with final included studies. The Rehabilitation Measures database (www.rehabmeasures.org) was also searched for citations associated with relevant outcome measures. Reference lists of selected studies were hand searched for additional studies. A sample search strategy is shown in Appendix 1.

Eligibility criteria

Studies were included if (1) their aim was to evaluate one or more of the following psychometric propertiesreliability, validity, responsiveness, internal consistency or interpretability, as outlined in the COSMIN guidelines (Mokkink et al., 2010b), (2) the population was adults with knee OA based on recognised clinical or radiographic criteria (Altman et al., 1991) and (3) if they explicitly stated that the purpose of the tool was to evaluate postural control. Studies were excluded if they were published in languages other than English or were not full text. Initial screening by title/abstract was initially performed by two independent reviewers (H.F. and A.V.), followed by independent screening of full-text articles of remaining studies. Any disagreements on potentially eligible studies were discussed and if agreement could not be reached, a third (D.M.) or fourth reviewer (C.H.) arbitrated.

Data extraction

The following data were extracted from each included study using a pre-piloted data extraction form: author(s), year, country, sample size, balance measure(s), quantification measure, equipment required, psychometric properties assessed and construct(s) of postural control being measured.

Assessment of methodological quality

The COSMIN checklist was used to determine methodological quality of included studies. This checklist has demonstrated high inter-rater agreement (Mokkink et al., 2010a). It contains 10 sections, which assess a different psychometric property: internal consistency, reliability, measurement error, content validity, structural validity, hypothesis testing, cross-cultural validity, responsiveness and interpretability. The remaining two boxes contain a checklist that includes general requirements for articles in which Item Response Theory methods are applied (Mokkink et al., 2010b). Methodological quality of each included article was rated on a four-point scale (poor, fair, good, excellent) by three independent reviewers (H.F., R.F. and D.M.). Disagreements were resolved by discussion among the authors. An overall quality score for each psychometric property was determined by taking the lowest rating of any item within that section, in accordance with COSMIN guidelines. The COSMIN tool was developed for patient-reported outcome measures (PROMs) and defined a sample size of 100 as adequate and 30 as fair. The COSMIN developers suggested that the sample size item be used as a guideline rather than as one of the criteria on which to determine the ‘worst scores’ (Mokkink et al., 2010b). In line with other systematic reviews of physical outcome measures in OA where smaller sample sizes are more likely than with PROMs (Dobson et al., 2012a; Dobson et al., 2012b), the sample size item was excluded from the scoring system when it was responsible for the ‘worst score’. If the sample size had been calculated or sample size was considered reasonable (n ≥ 20), then the next lowest rating item was used (Dobson et al., 2012a; Dobson et al., 2012b).

Domains of postural control evaluated

We used the Systems Framework for Postural Control (Horak, 2006) to identify the domains of postural control in each of the included measures. Nine operational domains of balance (Sibley et al., 2015) which had previously been adapted from the six domains as described by Horak (2006) were used as outlined in Table 1. Two review authors (D.M. and C.H.) independently assessed the measures to determine which constructs were captured.

Quality assessment of the measurement property

In addition to assessment of methodological quality of the included studies, the quality of the findings was rated using a recognised checklist. The overall rating of each psychometric property was considered positive (+), negative (−), or indeterminate (?) following the criteria reported by Terwee et al. (2007) (Table 2).

Table 2
Table 2:
Quality criteria for rating results of psychometric properties (Terwee et al., 2007)

Data analysis and synthesis

Data were summarised in a best evidence synthesis using a scoring of ‘unknown’, ‘strong’, ‘moderate’, ‘limited’ or ‘conflicting’ (Terwee et al., 2007).

  • (1) Unknown: not investigated in any study or investigated in studies of exclusively poor methodology;
  • (2) Strong: multiple studies of good methodological quality;
  • (3) Moderate: multiple fair methodological studies or one study of good methodology;
  • (4) Limited; one study of fair methodological quality;
  • (5) Conflicting; contradictory findings.

Results

Description of included studies

The literature search yielded 2643 articles, of which six studies were considered eligible for inclusion. Google Scholar and reference list searches of all included studies identified no further studies. The screening and selection process is outlined in Fig. 1 (Moher et al., 2010).

Fig. 1
Fig. 1:
A PRISMA flow diagram outlining the screening and selection process. PRISMA, preferred reporting items of systematic reviews and meta-analyses.

Characteristics of the included studies are provided in Table 3. Inclusion criteria for participants varied across studies. Five studies (Takacs et al., 2014a; Takacs et al., 2014b; Takacs et al., 2017a; Parveen and Noohu, 2017; Pirayeh et al., 2018) used Kellgren-Lawrence radiological criteria to determine OA severity (Kellgren and Lawrence, 1957). Two studies included people with radiological severity of grade 1 and above (Takacs et al., 2017a; Pirayeh et al., 2018), whilst three included grade 2 severity and above (Takacs, et al., 2014a; Takacs et al. 2014b; Parveen and Noohu, 2017). One study used clinical criteria only to define OA (Kanko et al., 2019). One study included participants based on a pain severity of at least three on a 0–10 scale (Parveen and Noohu, 2017). The mean age of the participants in one study (Parveen and Noohu, 2017) was lower (50.5 years) than the other studies (56.2–63.3 years). Average BMI was in the obese range (25–29 kg/m2) in all studies, in keeping with the profile of people with knee OA (Felson et al., 1988). Two studies included people with unilateral and bilateral knee OA (Parveen and Noohu, 2017; Pirayeh et al., 2018).

Table 3
Table 3:
Characteristics of the included studies

Clinical measures

Four of the six included studies investigated three clinical measures of postural control. Two evaluated the Community Balance and Mobility Scale (CBMS) (Takacs et al., 2014b; Takacs et al., 2017a), one evaluated the Tinetti Performance Oriented Mobility Assessment (POMA) (Parveen and Noohu, 2017) and one evaluated the Star Excursion Balance Test (SEBT) (Kanko et al., 2019). CBMS and POMA are both multi-item measures, which include other items which were classified as mobility (rather than balance) items. The CBMS was evaluated in two separate studies by the same authors (Takacs et al., 2014b; Takacs et al., 2017a).

The CBMS (Takacs et al., 2014b) was originally developed and validated for evaluation of balance and mobility in younger people with traumatic brain injury (Howe et al., 2006). It contains 13 items including unilateral stance, tandem walking, 180° tandem pivot, lateral foot scooting, hopping forward, crouch and walk, lateral dodging, walking and looking, running with a controlled stop, forwards-to-backwards walking, walk, look and carry, descending stairs and step-up. Tasks are rated by a trained assessor from 0 (unable to perform) to five (proficient). Assessment takes approximately 15 minutes. Total score range is 0–65; higher scores indicate higher levels of postural control. It has also been validated for use in cardiac rehabilitation (Martelli et al., 2018), stroke (Knorr et al., 2010) and healthy adults (Rocque et al., 2005; Weber et al., 2018). It is considered suitable for those at higher levels of functioning as it does not exhibit ceiling effects demonstrated by other balance measures (Balasubramanian, 2015).

The Tinetti POMA is a 16-item scale developed and validated for use with older populations (Lin et al., 2004; Tinetti, 1986). It comprises a balance (POMA-B) and a mobility subscale (POMA-G). The balance subscale contains nine items including sitting and standing balance, rising from sitting, returning to sitting and turning 360°. The score range on POMA-B is 0–16, with higher scores indicating better postural control. The POMA-G scores 0–12. Both can be combined to produce a total score (POMA-T) from 0 to 28. Psychometric properties have also been evaluated in stroke (Canbek et al., 2013), Parkinson’s disease (Behrman et al., 2002; Park et al., 2018) and other neurological populations (Kloos et al., 2014).

The SEBT, which was developed as a measure of dynamic balance (Kinzey and Armstrong, 1998), requires the participant to reach along a marked line with one leg while standing on the other leg. This is done in eight different directions (anterior, posterior, medial, lateral, anteromedial, anterolateral, posteromedial and posterolateral) every 45° from 0° to 360°. Psychometric properties have been ascertained in populations with lateral ankle instability (Bastien et al., 2014), anterior cruciate ligament deficiency (Dobija et al., 2019) and femoroacetabular impingement (Johansson and Karlsson, 2016).

Laboratory-based measures

Two studies evaluated static balance using a force platform (Takacs et al., 2014a; Pirayeh et al., 2018). One study evaluated double leg stance for a 30-second duration with eyes open and closed, and single leg stance with eyes open for 10 seconds (Pirayeh et al., 2018), while the second study evaluated single leg stance for 10 seconds (Takacs et al., 2014a). Centre of pressure measures included mediolateral (ML) SD, anteroposterior (AP) SD, velocity, sway area (Takacs et al., 2014a; Pirayeh et al., 2018), path length (Takacs et al., 2014a) and SD of velocity and amplitude in AP and ML directions (Pirayeh et al., 2018).

Four studies examined test-retest reliability (Takacs et al., 2014a; Takacs et al., 2014b; Parveen and Noohu, 2018; Kanko et al., 2019), three examined construct validity (Takacs, et al., 2014b; Pirayeh et al., 2018; Kanko et al., 2019), one assessed criterion validity (Pirayeh et al., 2018), one performed factor analysis (Takacs et al., 2017a) and one assessed responsiveness (Kanko et al., 2019). A comparison control group of people without OA was recruited in one study (Takacs et al., 2014b).

Reliability and measurement error

Test-retest reliability was measured in four of the six studies (Table 4). The three studies which evaluated clinical measures (Takacs et al., 2014b; Parveen and Noohu, 2017; Kanko et al., 2019) reported intraclass correlation coefficient values >0.9, indicating excellent test-retest reliability, consistent with a positive quality rating. One laboratory study evaluated test-retest reliability of centre of pressure measures during single-leg stance on a force platform (Takacs et al., 2014a). Reliability for measures of static balance (mean path length, velocity, area and SD in ML and AP directions) varied between 0.54 and 0.87, resulting in a positive quality rating (Table 2). Two measures (path length and velocity) scored >0.70, giving a positive quality rating for these measures (Table 2). A study limitation was the requirement to be able to stand on one leg for 10 seconds. Five (20%) of the sample were unable to achieve this, whilst 15 (60%) had to make three attempts before succeeding. This indicates a significant floor effect for this measure and considering that 10 (50%) of this sample had K-L severity of Gr 3 or 4 may affect its usability in moderate to severe OA.

Table 4
Table 4:
Summary of results of the reliability-related studies (n = 4)

The standard error of measurement (SEM) and minimal detectable change (MDC) were reported in three studies (Takacs et al., 2014b; Parveen and Noohu, 2017; Kanko et al., 2019). The MDC was outside the limits of agreement (LOA) for the CBMS (Takac et al., 2014b), providing a positive quality rating (Table 2). The MDC for the POMA-B, POMA-M and POMA-T scales were all within the LOA (Table 4), resulting in a negative quality rating (Table 2). As LOA was not determined for the SEBT (Kanko et al., 2019), quality rating was indeterminate (Table 2).

Validity

Three studies measured construct validity (Table 5). The CBMS was compared against the Berg Balance Scale (BBS), Timed Up and Go (TUG), single leg stance and 10 m walk test at self-selected and fast walking speed (Takacs et al., 2014b). Convergent validity of the POMA-B scale was evaluated by comparing with BBS (Takacs et al., 2014b) and the POMA-G scale was compared with TUG (Takacs et al., 2014b) (Table 5). Concurrent validity of the SEBT was determined by comparing against the SEBT using a 12-camera system to measure maximal reach (Kanko et al., 2019). Criterion validity of double leg and single leg stance tests on a force platform was ascertained by determining their ability to discriminate between people with mild and moderate/severe radiological OA (Pirayeh et al., 2018). An indeterminate quality rating was applied in this study, as although two of the measures resulted in a AUC >0.70 (Table 2), the use of radiological OA only as a gold standard is questionable due to poor correlation between radiological changes and symptoms in knee OA (Dieppe et al., 2000).

Table 5
Table 5:
Summary of results for validity-related studies (n = 5)

Structural validity using factor analysis was assessed for the CBMS (Takacs et al., 2017a) (Table 5). Confirmatory factor analysis showed that a three-factor model comprising three factors of lower limb strength, mobility and balance was most optimal and explained 65% of the variance. This resulted in a positive quality rating (Table 2).

Responsiveness

One study evaluated responsiveness of SEBT following a 12-week neuromuscular program (Table 6). SEBT was compared against the 40 m walk test, Knee Injury Osteoarthritis Outcome Score and Numerical Rating Scale for pain (Kanko et al., 2019). As inappropriate measures of responsiveness including standardised response mean and MDC were used, but not responsiveness ratios or LOA as recommended by Terwee et al. (2007), and as the change did not lie within the MDC range, the quality rating was indeterminate.

Table 6
Table 6:
Summary of results for responsiveness studies (n = 1)
Interpretability

Two studies reported ceiling and floor effects. POMA-T and POMA-G had 25% and 24% ceiling effects, respectively, but no floor effects (Parveen and Noohu, 2017). CBMS had no ceiling effects in the knee OA group, 12% ceiling effect in the control group and no floor effects in either OA or control group (Takacs et al., 2014b). Whilst ceiling and floor effects were not reported in the study which measured SLS using a force platform (Takacs et al., 2014a), 20% of the sample were unable to complete the test, indicative of a floor effect.

Methodological quality assessment

The results of the COSMIN quality scores are shown in Tables 46. Studies were rated as either ‘fair’ or ‘poor’ based on the lowest COSMIN score, discounting the sample size item (Tables 46).

Reliability

As shown in Table 4, three studies had an overall rating of ‘fair’ (Takacs et al., 2014a; Takacs, et al., 2014b; Kanko et al., 2019). The ‘fair’ score was assigned due to failure to report if administrations were independent (Takacs et al., 2014a; Takacs et al., 2014b) or if patients were stable in the interim period (Kanko et al., 2019). One had an overall rating of ‘poor’ due to major methodological flaws in the study, which included a highly functioning sample. The external validity of these results was very questionable based on a number of factors including the low mean age of the sample and the high mean scores on the POMA-G, POMA-B and BBS, with ceiling effects of 20–28% observed for these outcomes (Parveen and Noohu, 2017).

Validity

Of the five validity-related studies (Table 5), the lowest COSMIN score for one study was ‘poor’ due to methodological flaws as highlighted related to the reliability domain (Parveen and Noohu, 2017). For hypothesis testing (construct validity) two studies had an overall rating of ‘fair. Kanko et al. (2019) provided a poor description of the comparator constructs and Takacs et al. (2014b) failed to formulate hypotheses a priori as did Pirayeh et al. (2018) who measured criterion validity. One study, which measured structural validity (factor analysis), had an overall rating of ‘fair’ (Takacs et al., 2017a) based on an unclear description of how missing items were handled.

Responsiveness

One study measuring responsiveness (Kanko et al., 2019) had an overall rating of ‘poor’ based on lack of description of the psychometric properties of comparator instruments (Table 6).

Best evidence synthesis

A best evidence synthesis was derived based on the COSMIN checklist scores, study findings and number of studies using recognised quality criteria (Terwee et al., 2007). The evidence for reliability of the CBMS, POMA and centre of pressure measures during a single leg stance test is limited. Evidence for validity of CBMS and POMA and centre of pressure measures during single and double leg stance tests and responsiveness of SEBT is also limited.

Domains of postural control

Static stability was the most common domain assessed in all studies, followed by underlying motor systems, anticipatory postural control and dynamic stability that were evaluated in the Tinetti, SEBT and CBMS measures (Table 7). None of the included studies used measures that evaluated all nine domains of postural control.

Table 7
Table 7:
Domains of postural controla assessed by the postural control measures in knee OA populations (adapted from Sibley et al., 2015)

Discussion

Although impaired postural control is a recognised dysfunction in people with knee OA (Wegener et al., 1997; Hassan et al., 2001; Hsieh et al., 2013; Takacs et al., 2013; Khalaj et al., 2014), very few studies have ascertained the psychometric properties of postural control assessments for knee OA. This review will assist researchers in identifying specific assessment tools which have undergone some psychometric testing in people with knee OA

Psychometric properties

We found only six studies which evaluated any psychometric properties of single (Takacs et al., 2014a; Pirayeh et al., 2018; Kanko et al., 2019) or multi-item (Takacs et al., 2014b; Parveen and Noohu, 2017; Takacs et al., 2017a) aspects of postural control for knee OA in clinical or laboratory settings. We applied rigorous methods using recognised standards to determine the extent and quality of the evidence. Based on the small number of studies and inadequate methodological quality, the psychometric properties of these tools for use in populations with knee OA have not been adequately established.

Domains of postural control

Postural control requires the integration of many physiological systems. In this review, eight of the nine domains outlined in the Systems Framework were evaluated across the measures tested (Sibley et al., 2015), with six domains measured in both the CBMS and POMA. Single leg stance does not appear to be appropriate for use in knee OA as it measures one non-functional component of postural control, along with difficulty in completing the test (Takacs et al., 2014a). This was confirmed in other studies of knee OA (Hassan et al., 2001; Hunt et al., 2010). There were some differences between Sibley et al.’s (2015) interpretation of domains for these measures and ours. Specifically, they determined that the POMA measured ‘functional stability limits’, in contrast to our evaluation, as no component of the test required the person to reach or lean which would assess the ability of the centre of mass to move as far as possible within the base of support.

Multiple contributions to postural control deficits in knee OA have been identified. Reduced quadriceps strength is the most consistent factor. Impaired knee joint proprioception (Hortobágyi et al., 2004) knee malalignment (Hunt et al., 2010 and limited knee joint ROM (Takacs et al., 2015) may also contribute. Although pain is the most significant symptom in OA, there is conflicting evidence regarding its contribution to reduced postural control (Jadelis et al., 2001; Hassan et al., 2002; Hunt et al., 2010). How radiographic severity and postural control are related is also not clear, with a positive relationship (Birmingham et al., 2001), inverse relationship (Hunt et al., 2010) and no relationship (Hall et al., 2006) demonstrated. Obesity may also have a modifying effect (Jadelis et al., 2001). The presence of unilateral or bilateral symptoms should also be considered, as bilateral symptoms have been associated with impaired static balance (Alghadir et al., 2016). In our review, two studies included people with unilateral or bilateral OA (Parveen and Noohu, 2017; Pirayeh et al., 2018).

The choice of a measure used may limit its interpretability if it does not include all domains of postural control. Of the assessments evaluated in this review, the CBMS appears to be most preferred based on the psychometric properties and postural control domains assessed. One assessment tool which has been identified containing all six components of postural control described by Horak (2006) using the Systems Framework (Sibley et al., 2015) is the BESTest (Horak et al., 2009), which has been evaluated in one study of knee OA. Tamura et al. (2016) established that five of the six domains of the BESTest, except for ‘stability limits and verticality’ were reduced in people with knee OA, compared with age-matched controls. This construct highlighted that this is the only section that contains sitting tasks that may not be problematic for a person with knee OA. However, the other two items of functional reach forward and functional reach lateral are both tested in standing which is relevant in this population. Other variables associated with impaired postural control in knee OA are not captured in the BESTest such as pain or obesity, and therefore it may not be the measure of choice in this population. The length of time required to administer a test in clinical settings should be considered. The BESTest can take between 20 and 60 minutes to complete (Padgett et al., 2012), which has resulted in the development of the Mini BEStest and Brief BESTest shorter versions (Franchignoni et al., 2010; Padgett et al., 2012), but no studies have evaluated psychometric properties of the full or shortened versions for people with knee OA.

Study limitations

We included studies only published in English, which may have introduced bias, and we also only included studies that clearly stated the purpose of the test under investigation was to assess postural control. Other performance-based tests may also be considered tests of postural control. A comprehensive systematic review has evaluated the measurement properties of performance-based assessments, including tests used to assess balance in hip or knee OA (Dobson et al., 2012b).

Recommendations for future research

The small number of studies highlights the limited evaluation of psychometric properties of outcomes used to assess postural control in this population. As more RCTs are evaluating balance programs in OA (Takacs et al., 2017b), responsiveness of postural control assessments for knee OA is an important property to ascertain if such tools are to be used to evaluate treatment effectiveness. Further evaluation of psychometric properties of existing clinical outcomes used to assess postural control in this population (Hatfield et al., 2016) is required to determine if these tests are appropriate for use in this population, using the COSMIN methodology. The BESTest has been identified as the assessment tool of choice to assess all components of balance (Sibley et al., 2015). Alternatively, consideration could be given to how existing measures such as the CBMS could be modified, or an OA-specific measure be developed and validated to ensure that all relevant constructs in this population can be assessed

Conclusion

This systematic review identified only three assessment tools of postural control for which some psychometric properties have been evaluated in knee OA. Based on our findings, the evidence for the reliability, validity and responsiveness of these measures is limited and lack comprehensive testing of all domains of postural control. Future high-quality research should investigate the psychometric properties of postural control tools which could be used in studies of knee OA, and the COSMIN framework can be used in such evaluation to provide methodological rigour.

Acknowledgements

The research was undertaken with funding support from the RCSI undergraduate summer student research program.

Conflicts of interest

There are no conflicts of interest.

Appendix 1: Sample search strategy – PubMed

  • (1) Search “Osteoarthritis, Knee”[Mesh]
  • (2) Search (osteoarthritis[Title/Abstract]) OR arthrosis[Title/Abstract]
  • (3) Search knee*[Title/Abstract]
  • (4) Search (((osteoarthritis[Title/Abstract]) OR arthrosis[Title/Abstract])) AND knee*[Title/Abstract]
  • (5) Search (“Osteoarthritis, Knee”[Mesh]) OR ((((osteoarthritis[Title/Abstract]) OR arthrosis[Title/Abstract])) AND knee*[Title/Abstract])
  • (6) Search “Proprioception”[Mesh]
  • (7) Search “Biomechanical Phenomena”[Mesh]
  • (8) Search (((postural control[Title/Abstract]) OR balance[Title/Abstract]) OR equilibrium[Title/Abstract]) OR stabilit*[Title/Abstract]
  • (9) Search ((“Proprioception”[Mesh]) OR “Biomechanical Phenomena”[Mesh]) OR ((((postural control[Title/Abstract]) OR balance[Title/Abstract]) OR equilibrium[Title/Abstract]) OR stabilit*[Title/Abstract])
  • (10) Search (((“Osteoarthritis, Knee”[Mesh]) OR ((((osteoarthritis[Title/Abstract]) OR arthrosis[Title/Abstract])) AND knee*[Title/Abstract]))) AND (((“Proprioception”[Mesh]) OR “Biomechanical Phenomena”[Mesh]) OR ((((postural control[Title/Abstract]) OR balance[Title/Abstract]) OR equilibrium[Title/Abstract]) OR stabilit*[Title/Abstract]))
  • (11) Search ((“Psychometrics”[Mesh]) OR “Reproducibility of Results”[Mesh]) OR “Observer Variation”[Mesh]
  • (12) Search ((((((((valid*[Title/Abstract]) OR reliability[Title/Abstract]) OR reliable[Title/Abstract]) OR responsive*[Title/Abstract]) OR accuracy[Title/Abstract]) OR accurate[Title/Abstract]) OR reproducibility[Title/Abstract]) OR reproducible[Title/Abstract]) OR agreement
  • (13) Search ((((“Psychometrics”[Mesh]) OR “Reproducibility of Results”[Mesh]) OR “Observer Variation”[Mesh])) OR (((((((((valid*[Title/Abstract]) OR reliability[Title/Abstract]) OR reliable[Title/Abstract]) OR responsive*[Title/Abstract]) OR accuracy[Title/Abstract]) OR accurate[Title/Abstract]) OR reproducibility[Title/Abstract]) OR reproducible[Title/Abstract]) OR agreement)
  • (14) Search ((((test[Title/Abstract]) OR tests[Title/Abstract]) OR assess*[Title/Abstract]) OR evaluat*[Title/Abstract]) OR measur*[Title/Abstract]
  • (15) Search (((((“Osteoarthritis, Knee”[Mesh]) OR ((((osteoarthritis[Title/Abstract]) OR arthrosis[Title/Abstract])) AND knee*[Title/Abstract]))) AND (((“Proprioception”[Mesh]) OR “Biomechanical Phenomena”[Mesh]) OR ((((postural control[Title/Abstract]) OR balance[Title/Abstract]) OR equilibrium[Title/Abstract]) OR stabilit*[Title/Abstract])))) AND (((((“Psychometrics”[Mesh]) OR “Reproducibility of Results”[Mesh]) OR “Observer Variation”[Mesh])) OR (((((((((valid*[Title/Abstract]) OR reliability[Title/Abstract]) OR reliable[Title/Abstract]) OR responsive*[Title/Abstract]) OR accuracy[Title/Abstract]) OR accurate[Title/Abstract]) OR reproducibility[Title/Abstract]) OR reproducible[Title/Abstract]) OR agreement))
  • (16) Search (((((“Osteoarthritis, Knee”[Mesh]) OR ((((osteoarthritis[Title/Abstract]) OR arthrosis[Title/Abstract])) AND knee*[Title/Abstract]))) AND (((“Proprioception”[Mesh]) OR “Biomechanical Phenomena”[Mesh]) OR ((((postural control[Title/Abstract]) OR balance[Title/Abstract]) OR equilibrium[Title/Abstract]) OR stabilit*[Title/Abstract])))) AND (((((test[Title/Abstract]) OR tests[Title/Abstract]) OR assess*[Title/Abstract]) OR evaluat*[Title/Abstract]) OR measur*[Title/Abstract])

References

Alghadir A, Anwer S, Iqbal ZA, Alsanawi HA. Cross-cultural adaptation, reliability and validity of the Arabic version of the reduced western Ontario and McMaster Universities osteoarthritis index in patients with knee osteoarthritis. Disabil Rehabil. 2016; 38:689–694
Altman R, Alarcón G, Appelrouth D, Bloch D, Borenstein D, Brandt K, et al. The American College of Rheumatology criteria for the classification and reporting of osteoarthritis of the hip. Arthritis Rheum. 1991; 34:505–514
Balasubramanian CK. The Community Balance and Mobility Scale alleviates the ceiling effects observed in the currently used gait and balance assessments for the community-dwelling older adults. J Geriatr Phys Ther. 2015; 38:78–89
Bastien M, Moffet H, Bouyer L, Perron M, Hébert LJ, Leblond J. Concurrent and discriminant validity of the Star Excursion Balance Test for military personnel with lateral ankle sprain. J Sport Rehabil. 2014; 23:44–55
Behrman AL, Light KE, Miller GM. Sensitivity of the Tinetti gait assessment for detecting change in individuals with Parkinson's disease. Clin Rehabil. 2002; 16:399–405
Berg K, Wood-Dauphine S, Williams JI, Gayton D. Measuring balance in the elderly: preliminary development of an instrument. Physiother Canada. 1989; 41:304–311
Birmingham TB, Kramer JF, Kirkley A, Inglis JT, Spaulding SJ, Vandervoort AA. Association among neuromuscular and anatomic measures for patients with knee osteoarthritis. Arch Phys Med Rehabil. 2001; 82:1115–1118
Canbek J, Fulk G, Nof L, Echternach J. Test-retest reliability and construct validity of the Tinetti performance-oriented mobility assessment in people with stroke. J Neurol Phys Ther. 2013; 37:14–19
COSMIN. The COnsensus-based Standards for the selection of health Measurement Instruments (COSMIN).. 2019. Retrieved from www.cosmin.nl. [Accessed 4 February 2019]
Dieppe P, Cushnaghan J, Tucker M, Browning S, Shepstone L. The bristol ‘OA500 study’: progression and impact of the disease after 8 years. Osteoarthritis Cartilage. 2000; 8:63–68
Dobija L, Reynaud V, Pereira B, Van Hille W, Descamps S, Bonnin A, Coudeyre E. Measurement properties of the Star Excursion Balance Test in patients with ACL deficiency. Phys Ther Sport. 2019; 36:7–13
Dobson F, Choi YM, Hall M, Hinman RS. Clinimetric properties of observer-assessed impairment tests used to evaluate hip and groin impairments: a systematic review. Arthritis Care Res (Hoboken). 2012; 64:1565–1575
    Dobson F, Hinman RS, Hall M, Terwee CB, Roos EM, Bennell KL. Measurement properties of performance-based measures to assess physical function in hip and knee osteoarthritis: a systematic review. Osteoarthritis Cartilage. 2012; 20:1548–1562
    Doré AL, Golightly YM, Mercer VS, Shi XA, Renner JB, Jordan JM, Nelson AE. Lower-extremity osteoarthritis and the risk of falls in a community-based longitudinal study of adults with and without osteoarthritis. Arthritis Care Res (Hoboken). 2015; 67:633–639
    Duncan PW, Studenski S, Chandler J, Prescott B. Functional reach: predictive validity in a sample of elderly male veterans. J Gerontol. 1992; 47:M93–M98
    Felson DT, Anderson JJ, Naimark A, Walker AM, Meenan RF. Obesity and knee osteoarthritis. The Framingham study. Ann Intern Med. 1988; 109:18–24
    Franchignoni F, Horak F, Godi M, Nardone A, Giordano A. Using psychometric techniques to improve the balance evaluation systems test: the mini-bestest. J Rehabil Med. 2010; 42:323–331
    French HP, Galvin R, Horgan NF, Kenny RA. Prevalence and burden of osteoarthritis amongst older people in Ireland: findings from the Irish longitudinal study on ageing (TILDA). Eur J Public Health. 2016; 26:192–198
    Frontera WR, Hughes VA, Fielding RA, Fiatarone MA, Evans WJ, Roubenoff R. Aging of skeletal muscle: a 12-yr longitudinal study. J Appl Physiol (1985). 2000; 88:1321–1326
    Hall MC, Mockett SP, Doherty M. Relative impact of radiographic osteoarthritis and pain on quadriceps strength, proprioception, static postural sway and lower limb function. Ann Rheum Dis. 2006; 65:865–870
    Hassan BS, Doherty SA, Mockett S, Doherty M. Effect of pain reduction on postural sway, proprioception, and quadriceps strength in subjects with knee osteoarthritis. Ann Rheum Dis. 2002; 61:422–428
    Hassan BS, Mockett S, Doherty M. Static postural sway, proprioception, and maximal voluntary quadriceps contraction in patients with knee osteoarthritis and normal control subjects. Ann Rheum Dis. 2001; 60:612–618
    Hatfield GL, Morrison A, Wenman M, Hammond CA, Hunt MA. Clinical tests of standing balance in the knee osteoarthritis population: systematic review and meta-analysis. Phys Ther. 2016; 96:324–337
    Hinman RS, Bennell KL, Metcalf BR, Crossley KM. Balance impairments in individuals with symptomatic knee osteoarthritis: a comparison with matched controls using clinical tests. Rheumatology (Oxford). 2002; 41:1388–1394
    Horak FB. Postural orientation and equilibrium: what do we need to know about neural control of balance to prevent falls?. Age Ageing. 2006; 35Suppl 2ii7–ii11
    Horak FB, Wrisley DM, Frank J. The balance evaluation systems test (bestest) to differentiate balance deficits. Phys Ther. 2009; 89:484–498
    Hortobágyi T, Garry J, Holbert D, Devita P. Aberrations in the control of quadriceps muscle force in patients with knee osteoarthritis. Arthritis Rheum. 2004; 51:562–569
    Houry D, Florence C, Baldwin G, Stevens J, McClure R. The CDC Injury Center’s response to the growing public health problem of falls among older adults. Am J Lifestyle Med. 2016; 10:74–77
    Howe JA, Inness EL, Venturini A, Williams JI, Verrier MC. The Community Balance and Mobility Scale –a balance measure for individuals with traumatic brain injury. Clin Rehabil. 2006; 20:885–895
    Hsieh RL, Lee WC, Lo MT, Liao WC. Postural stability in patients with knee osteoarthritis: comparison with controls and evaluation of relationships between postural stability scores and international classification of functioning, disability and health components. Arch Phys Med Rehabil. 2013; 94:340–346
    Hunt MA, McManus FJ, Hinman RS, Bennell KL. Predictors of single-leg standing balance in individuals with medial knee osteoarthritis. Arthritis Care Res (Hoboken). 2010; 62:496–500
    Hurley MV, Rees J, Newham DJ. Quadriceps function, proprioceptive acuity and functional performance in healthy young, middle-aged and elderly subjects. Age Ageing. 1998; 27:55–62
    Huxham FE, Goldie PA, Patla AE. Theoretical considerations in balance assessment. Aust J Physiother. 2001; 47:89–100
    Jadelis K, Miller ME, Ettinger WH Jr, Messier SP. Strength, balance, and the modifying effects of obesity and knee pain: results from the observational arthritis study in seniors (oasis). J Am Geriatr Soc. 2001; 49:884–891
    Johansson AC, Karlsson H. The Star Excursion Balance Test: criterion and divergent validity on patients with femoral acetabular impingement. Man Ther. 2016; 26:104–109
    Kanko LE, Birmingham TB, Bryant DM, Gillanders K, Lemmon K, Chan R, et al. The Star Excursion Balance Test is a reliable and valid outcome measure for patients with knee osteoarthritis. Osteoarthritis Cartilage. 2019; 27:580–585
    Kellgren JH, Lawrence JS. Radiological assessment of osteo-arthrosis. Ann Rheum Dis. 1957; 16:494–502
    Khalaj N, Abu Osman NA, Mokhtar AH, Mehdikhani M, Wan Abas WA. Balance and risk of fall in individuals with bilateral mild and moderate knee osteoarthritis. PLoS One. 2014; 9:e92270
    Kinzey SJ, Armstrong CW. The reliability of the star-excursion test in assessing dynamic balance. J Orthop Sports Phys Ther. 1998; 27:356–360
    Kloos AD, Fritz NE, Kostyk SK, Young GS, Kegelmeyer DA. Clinimetric properties of the Tinetti mobility test, four square step test, activities-specific balance confidence scale, and spatiotemporal gait measures in individuals with Huntington's disease. Gait Posture. 2014; 40:647–651
    Knorr S, Brouwer B, Garland SJ. Validity of the Community Balance and Mobility Scale in community-dwelling persons after stroke. Arch Phys Med Rehabil. 2010; 91:890–896
    Kontis V, Bennett JE, Mathers CD, Li G, Foreman K, Ezzati M. Future life expectancy in 35 industrialised countries: projections with a Bayesian model ensemble. Lancet. 2017; 389:1323–1335
    Laughton CA, Slavin M, Katdare K, Nolan L, Bean JF, Kerrigan DC, et al. Aging, muscle activity, and balance control: physiologic changes associated with balance impairment. Gait Posture. 2003; 18:101–108
    Lin MR, Hwang HF, Hu MH, Wu HD, Wang YW, Huang FC. Psychometric comparisons of the timed up and go, one-leg stand, functional reach, and Tinetti balance measures in community-dwelling older people. J Am Geriatr Soc. 2004; 52:1343–1348
    Lo-Ciganic WH, Floden L, Lee JK, Ashbeck EL, Zhou L, Chinthammit C, et al. Analgesic use and risk of recurrent falls in participants with or at risk of knee osteoarthritis: data from the osteoarthritis initiative. Osteoarthritis Cartilage. 2017; 25:1390–1398
    Manlapaz DG, Sole G, Jayakaran P, Chapple CM. Risk factors for falls in adults with knee osteoarthritis: a systematic review. PM R. 2019; 11:745–757
    Martelli L, Saraswat D, Dechman G, Giacomantonio NB, Grandy SA. The Community Balance and Mobility Scale: a valid assessment tool of balance in cardiac rehabilitation patients. J Cardiopulm Rehabil Prev. 2018; 38:100–103
    Mat S, Tan MP, Kamaruzzaman SB, Ng CT. Physical therapies for improving balance and reducing falls risk in osteoarthritis of the knee: a systematic review. Age Ageing. 2015; 44:16–24
    Moher D, Liberati A, Tetzlaff J, Altman DG; PRISMA Group. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. Int J Surg. 2010; 8:336–341
    Mokkink LB, Terwee CB, Gibbons E, Stratford PW, Alonso J, Patrick D L, et al. Inter-rater agreement and reliability of the COSMIN (consensus-based standards for the selection of health status measurement instruments) checklist. BMC Med Res Methodol. 2010; 10:82
    Mokkink LB, Terwee CB, Patrick DL, Alonso J, Stratford PW, Knol D L, et al. The COSMIN checklist for assessing the methodological quality of studies on measurement properties of health status measurement instruments: an international Delphi study. Qual Life Res. 2010; 19:539–549
    Murphy L, Schwartz TA, Helmick CG, Renner JB, Tudor G, Koch G, et al. Lifetime risk of symptomatic knee osteoarthritis. Arthritis Rheum. 2008; 59:1207–1213
    Padgett PK, Jacobs JV, Kasser SL. Is the bestest at its best? A suggested brief version based on interrater reliability, validity, internal consistency, and theoretical construct. Phys Ther. 2012; 92:1197–1207
    Park J, Koh SB, Kim HJ, Oh E, Kim JS, Yun JY, et al. Validity and reliability study of the Korean Tinetti mobility test for Parkinson's disease. J Mov Disord. 2018; 11:24–29
    Parveen H, Noohu MM. Evaluation of psychometric properties of Tinetti performance-oriented mobility assessment scale in subjects with knee osteoarthritis. Hong Kong Physiother J. 2017; 36:25–32
    Petrella RJ, Lattanzio PJ, Nelson MG. Effect of age and activity on knee joint proprioception. Am J Phys Med Rehabil. 1997; 76:235–241
    Pirayeh N, Shaterzadeh-Yazdi MJ, Negahban H, Mehravar M, Mostafaee N, Saki-Malehi A. Examining the diagnostic accuracy of static postural stability measures in differentiating among knee osteoarthritis patients with mild and moderate to severe radiographic signs. Gait Posture. 2018; 64:1–6
    Podsiadlo D, Richardson S. The timed “up & go”: a test of basic functional mobility for frail elderly persons. J Am Geriatr Soc. 1991; 39:142–148
    Pollock AS, Durward BR, Rowe PJ, Paul JP. What is balance?. Clin Rehabil. 2000; 14:402–406
    Rocque R, Bartlett D, Brown J, Garland S. Influence of age and gender of healthy adults on scoring patterns on the Community Balance and Mobility Scale. Physiotherapy Canada. 2005; 57:2852–2892
    Shumway-Cook A, Woollacott MH. Motor Control: Theory and Applications. 20012nd ed, Philadelphia: Lippincott Williams and Wilkins
    Sibley KM, Beauchamp MK, Van Ooteghem K, Straus SE, Jaglal SB. Using the systems framework for postural control to analyze the components of balance evaluated in standardized balance measures: a scoping review. Arch Phys Med Rehabil. 2015; 96:122–132.e29
    Slemenda C, Brandt KD, Heilman DK, Mazzuca S, Braunstein EM, Katz BP, Wolinsky FD. Quadriceps weakness and osteoarthritis of the knee. Ann Intern Med. 1997; 127:97–104
    Sturnieks DL, Tiedemann A, Chapman K, Munro B, Murray SM, Lord S R. Physiological risk factors for falls in older people with lower limb arthritis. J Rheumatol. 2004; 31:2272–2279
    Takacs J, Carpenter MG, Garland SJ, Hunt MA. The role of neuromuscular changes in aging and knee osteoarthritis on dynamic postural control. Aging Dis. 2013; 4:84–99
    Takacs J, Carpenter MG, Garland SJ, Hunt MA. Test re-test reliability of centre of pressure measures during standing balance in individuals with knee osteoarthritis. Gait Posture. 2014a; 40:270–273
    Takacs J, Garland SJ, Carpenter MG, Hunt MA. Validity and reliability of the Community Balance and Mobility Scale in individuals with knee osteoarthritis. Phys Ther. 2014b; 94:866–874
    Takacs J, Carpenter MG, Garland SJ, Hunt MA. Factors associated with dynamic balance in people with knee osteoarthritis. Arch Phys Med Rehabil. 2015; 96:1873–1879
    Takacs J, Krowchuk NM, Garland SJ, Carpenter MG, Hunt MA. Dynamic balance training improves physical function in individuals with knee osteoarthritis: a pilot randomized controlled trial. Arch Phys Med Rehabil. 2017a; 98:1586–1593
    Takacs J, Krowchuk NM, Goldsmith CH, Hunt MA. Factor analysis of the Community Balance and Mobility Scale in individuals with knee osteoarthritis. Physiother Res Int. 2017b; 22:e1675
    Tamura T, Otaka Y, Konno S, Sadashima K, Tomatsu T, Machida S. The impaired balance systems identified by the bestest in older patients with knee osteoarthritis. PM R. 2016; 8:869–875
    Terwee CB, Bot SD, de Boer MR, van der Windt DA, Knol DL, Dekker J, et al. Quality criteria were proposed for measurement properties of health status questionnaires. J Clin Epidemiol. 2007; 60:34–42
    Tinetti ME. Performance-oriented assessment of mobility problems in elderly patients. J Am Geriatr Soc. 1986; 34:119–126
    van Raaij TM, Reijman M, Brouwer RW, Bierma-Zeinstra SM, Verhaar J A. Medial knee osteoarthritis treated by insoles or braces: a randomized trial. Clin Orthop Relat Res. 2010; 468:1926–1932
    Weber M, Van Ancum J, Bergquist R, Taraldsen K, Gordt K, Mikolaizak AS, et al. Concurrent validity and reliability of the Community Balance and Mobility Scale in young-older adults. BMC Geriatr. 2018; 18:156
    Wegener L, Kisner C, Nichols D. Static and dynamic balance responses in persons with bilateral knee osteoarthritis. J Orthop Sports Phys Ther. 1997; 25:13–18
    Keywords:

    balance; osteoarthritis; postural control; psychometric properties; reliability; systematic review; validity

    Copyright © 2020 Wolters Kluwer Health, Inc. All rights reserved.