Systematic Review of Therapeutic Alliance Measurement Instruments in Physiotherapy (2024)

Abstract

Purpose: Interest in measuring the therapeutic alliance has grown in recent years in both the clinical field and the literature. Several instruments can be used to measure the therapeutic alliance in physiotherapy, and choosing the most appropriate one is challenging. The purpose of this systematic review was to (1) identify the instruments that have been used to evaluate the therapeutic alliance in physiotherapy and (2) appraise the methodological quality of studies of the psychometric properties of these instruments. Method: The researchers carried out a systematic search in MEDLINE, SciELO, PsycINFO, Theseus, Cochrane Library, and Open Grey. Only articles published in English and Spanish were included. The articles were evaluated by two independent reviewers in accordance with the COnsensus-based Standards for the selection of health Measurement Instruments (COSMIN) and Preferred Reporting items for Systematic Review and Meta-Analysis standards, using the four-point COSMIN checklist. Results: Four studies were included in this review, and four instruments evaluated the therapeutic alliance in physiotherapy. The methodological quality of the studies was fair for most of the psychometric characteristics analyzed. Conclusions: The Working Alliance Inventory is the best instrument to measure therapeutic alliance and the measure that had the greatest number of psychometric properties evaluated against the COSMIN standard. More studies of high methodological quality are required to evaluate the psychometric properties of the instruments used to assess the therapeutic alliance in physiotherapy.

Key Words: psychometrics, rehabilitation, therapeutic alliance

One cannot imagine physiotherapy without an interpersonal relationship between the patient and the professional.¹ This therapeutic alliance is jointly constructed by the patient and professional, and the expectations, opinions, and ideas that they develop about the work they are doing, as well as the relationship they establish and their idea of the other, are important to establishing it.¹^,² Of all the components of the therapeutic relationship, the therapeutic alliance has received the greatest amount of research attention.³ In physiotherapy, the closer the therapeutic alliance, the better the therapeutic results.⁴ Therefore, in recent years interest in measuring the therapeutic alliance has grown in both the clinical field and the literature.⁵^,⁶

Although several instruments measure the physiotherapy therapeutic alliance by assessing basic psychometric properties, no review has been published that rates these properties according to the COnsensus-based Standards for the selection of health Measurement INstruments (COSMIN),⁵^,⁷^,⁸ following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA).⁹ The COSMIN checklist is a standardized tool for assessing the methodological quality of studies on the basis of measurement properties. It contains nine boxes, each of which addresses one measurement property, with 5–18 items per box pertaining to design aspects and statistical methods.⁸ We carried out a systematic review of the literature to (1) identify the instruments that have been used to measure the therapeutic alliance in physiotherapy, (2) determine their reported psychometric properties, and (3) evaluate the quality of the validation studies according to the COSMIN and PRISMA standards.⁷^–⁹

Methods

The protocol for this review was registered in the International Prospective Register of Systematic Reviews (Register No. CRD42018091735).

Data sources and searches

We performed a systematic search of the literature in MEDLINE, SciELO, PsycINFO, Theseus, Cochrane Library, and Open Grey using the following key terms: instrument, measure, therapeutic alliance, working alliance, and physiotherapy. The search was limited to articles published in English and Spanish and was conducted in February 2020.

Study selection

First, two reviewers (DGS and DPC) independently screened titles and abstracts to identify relevant articles. Any discrepancies were resolved by discussing them with the third author until consensus was reached. Then those articles determined to be relevant underwent a full-text review. For this review, we used the following inclusion criteria:

Types of participant: patients and professionals
Types of result: therapeutic alliance in physiotherapy
Types of study: studies that had evaluated and described at least one psychometric property from the COSMIN checklist
Types of instrument: instruments that had been used to evaluate the therapeutic alliance in physiotherapy

In addition, we examined the reference lists of eligible studies to identify additional studies.

Data extraction

After the full-text review, those articles that met the inclusion criteria underwent data extraction. We developed a form to collect the data, and the following general characteristics of each study were extracted: study (year) and name of instrument, country of validation and size of sample (N), setting, participants, description of instrument, properties measured, and psychometric values.

Quality assessment

Two reviewers (DGS and DPC) evaluated the studies using the four-point COSMIN checklist for systematic reviews.⁸ The checklist consists of nine boxes that provide a methodological quality score for the following psychometric properties: internal consistency, reliability, measurement error, content validity, structural validity, hypothesis testing, cross-cultural validity, criterion validity, and responsiveness to change. Each property is assessed by means of multiple items (ranging in number from 5 to 18), and each property’s overall quality is determined to be excellent, good, fair, or poor by taking the lowest rating among all the items for each property.⁸

Results

Our search identified 836 articles and, after eliminating 102 duplicates, we screened 734 titles and abstracts. Screening resulted in the exclusion of 717 articles, leaving 17 articles to be assessed for eligibility. Of these, 13 articles were excluded after full-text review. Four studies, reporting on four instruments, met the inclusion criteria for data extraction and analysis. A flowchart of this review is provided in Figure 1.

Table 1.

Study Characteristics and Evaluation of Physiotherapy Instruments

Study	Country and sample size	Setting	Participants	Description of instrument	Psychometric properties measured	COSMIN score	Psychometric values
Schönberger et al.¹⁰	Denmark; N=98	Brain Injury Rehabilitation Center of University of Copenhagen	Patients with brain damage	Working Alliance and Compliance Scale; 6 items on 3- to 5-point Likert-type scale	1. Internal consistency 2. Reliability	1=poor 2=fair	Cronbach’s α=0.85; Cohen’s κ=0.3
Hall et al.⁴	Australia; N=206	Hospital	Patients with chronic low back pain	WATOCI (reduced version); 1-factor structure, 9 items rated on a 7-point Li kert scale ((neverto always); a high score indicates a strong alliance	1. Internal consistency 2. Structural validity	1=fair 2=fair	Cronbach’s α=0.64; unifactorial solution explained 63.6% of variance
Araujo et al.¹⁵	Brazil; N=100	Physiotherapy clinics	Patients with nonspecific low back pain	WAI (short version) WAI-P: 12 items rated on a 5-point Likert-type scale (rarelyto always); total score can range from 12 (weak alliance) to 60 (strong alliance) WAI-T: 12 items rated on a 7-point Likert-type scale (rarelyto always); total score can range from 12 (weak alliance) to 84 (strong alliance)	1. Internal consistency 2. Reliability 3. Measurement error 4. Hypothesis testing 5. Cross-cultural validity 6. Criterion validity 7. Sensitivity to change	1=poor 2=fair 3=fair 4=fair 5=poor 6=fair 7=fair	WAI-P: Cronbach’s α=0.52–0.62; ICC_2,1=0.74(0.66–0.81); SEM=3.3; MDC=9.14; ES=0.15,84%; IC: 0.29, 0.49; ρ=0.39 WAI-T: Cronbach’s α=0.58–0.81; ICC_2,1=0.85 (0.79–0.89); SEM=3.2; MDC=8.87; ES=0.37,84%; IC: 0.29, 0.49; ρ=0.39
Araujo et al.¹⁵	Brazil; N=100	Physiotherapy clinics	Patients with nonspecific low back pain	SRS: 4 items scored on a 10-point visual analogue scale; total score=0–40	1. Reliability 2. Measurement error 3. Hypothesis testing 4. Cross-cultural validity 5. Criterion validity 6. Sensitivity to change	1=fair 2=fair 3=fair 4=poor 5=fair 6=fair	ICC_2,1=0.79(0.71–0.85); SEM=1.83; MDC=5.05; ES=0.05, 84%; IC: 0.22,0.11; ρ=0.39
Karel et al.¹⁴	Holland; N=389	Primary care	Patients with elbow pain	WAI (flamenco version): 12 items rated on a 5-point Likert-type scale (rarelyto always); total score can range from 12 (weak alliance) to 60 (strong alliance)	1. Internal consistency 2. Reliability 3. Structural validity	1=excellent 2=good 3=excellent	Cronbach’s α=0.89; Pearson’s r=0.30; single-factor solution

Open in a new tab

COSMIN=COnsensus-based Standards for the selection of health Measurement Instruments; WATOCI=Working Alliance Theory of Change Inventory; WAI=Working Alliance Inventory; WAI-ρ=Working Alliance Inventory Patient; WAI-T=Working Alliance Inventory Professional or Therapist; ICC=intra-class correlation coefficient;

SEM=standard error of measurement; IC=internal consistency; MDC=minimal detectable change; ES=effect size; ρ=Spearman correlation coefficient;

SRS=Session Rating Scale.

The most frequently evaluated psychometric properties of these four instruments were internal consistency and reliability. The psychometric properties of the instruments measuring the therapeutic alliance in physiotherapy are shown in Table 1. The results of the quality assessment using the COSMIN checklist are shown in Table 2.

Table 2.

Quality Assessment Results using the Four-Point COSMIN Checklist

Instrument	Article (year)	Internal consistency	Reliability	Measurement error	Structural validity	Hypotheses testing	Cross cultural validity	Criteria validity	Sensitivity to change
Working Alliance and Compliance Scale	Schönberger et al.¹⁰	+	++	-	-	-	-	-	-
WATOCI (reduced version)	Hall et al.⁴	++	-	-	++	-	-	-	-
WAI (short version)	Araujo et al.¹⁵	++	++	++	-	++	+	++	++
	Karel et al.¹⁴	++++	+++	-	++++	-	-	-	-
SRS	Araujo et al.¹⁵	-	++	++	-	++	+	++	++

Open in a new tab

Notes:-=not evaluated; +=poor; ++=fair; +++=good; ++++=excellent.

COSMIN=COSMIN=COnsensus-based Standards for the selection of health Measurement Instruments; WATOCI=Working Alliance Theory of Change Inventory; WAI=Working Alliance Inventory; SRS=Session Rating Scale.

Working Alliance and Compliance scale

The Working Alliance and Compliance scale has been validated for use with patients with brain damage.¹⁴ Only one study assessed this instrument for internal consistency and reliability.¹⁴ Its internal consistency was poor; the main reasons were the lack of description of the percentages of missing responses per item and their management because a high number of missing responses can introduce bias in the results of the study.

Cronbach’s α for the scale was 0.85. Its reliability was fair, mainly because of the lack of description of the percentages and management of lost values. The Cohen’s κ was 0.3.

Working Alliance Theory of Change Inventory (reduced version)

The WATOCI has been validated for use with patients with chronic low back pain.¹¹ It was evaluated in only one study, which assessed its internal consistency and structural validity.¹¹ The methodological quality of both was only fair because the study lacked any description of the percentages and management of missing responses. A Cronbach’s α of 0.64 was obtained. Structural validity was based on the completion of factor analysis. A Rasch analysis was performed, and a unifactorial solution was obtained.

Working Alliance Inventory

The WAI has been validated for use in patients with non-specific low back pain and patients with elbow pain.¹²^,¹³ This instrument was assessed in two studies.¹²^,¹³

Araujo and colleagues assessed the WAI for internal consistency, reliability, measurement error, hypothesis testing, cross-cultural validity, criterion validity, and sensitivity to change.¹⁵ The researchers judged the methodological quality of two of the analyzed properties (internal consistency and cross-cultural validity) as poor and five (reliability, measurement error, hypothesis testing, criterion validity and sensitivity to change) as fair.¹² The lack of description of the percentages and management of missing responses contributed to a lower quality rating for the psychometric properties analyzed in this study.

Two versions of this instrument were assessed in this study:¹² the WAI Patient (WAI–P) and the WAI Professional or Therapist (WAI–T). For the WAI–P, Cronbach’s α ranged from 0.52 to 0.62; for the WAI–T, it ranged from 0.58 to 0.81. The intra-class correlation coefficient (ICC_2,1) for reliability was 0.74 for the WAI–P and 0.85 for the WAI–T. The standard error of measurement (SEM) was 3.3 and 3.2, respectively, for the WAI–P and WAI–T, and the minimal detectable change (MDC) was 9.14 and 8.87, respectively, for the WAI–P and WAI–T. Both versions of the WAI obtained Spearman’s rank order correlations (r) of 0.39 for criterion validity with the SRS.

Karel and colleagues also evaluated the WAI for internal consistency, reliability, and structural validity.¹³ Two properties (internal consistency and structural validity) scored as excellent, and one (reliability) scored as good. The main reasons for these scores were the adequate sample size and the use of factor analysis. Cronbach’s α was 0.89 for internal consistency, Pearson’s r was 0.30 for reliability, and a single-factor solution was obtained for structural validity. Although no specific gold standard was used as a criterion, the criterion used – comparing patients’ and therapists’ scores – was considered a valid comparator.

The WAI was the best instrument for measuring therapeutic alliance and had the greatest number of psychometric properties evaluated.

Session Rating Scale

The SRS has been validated for use in patients with non-specific low back pain.¹² It was evaluated in one study for reliability, measurement error, hypothesis testing, cross-cultural validity, criterion validity, and sensitivity to change.¹² The methodological quality was fair for all psychometric properties except for hypothesis testing, for which it was poor. The main reasons for the low score were the non-completion of confirmatory factor analysis and the lack of any description of the percentages and management of missing responses. The reliability showed an overall ICC_2,1 value of 0.79 (range 0.71–0.85), the SEM was 1.83, and the MDC was 5.05. This instrument obtained a r of 0.39 for criterion validity (compared with WAI). Although no specific gold standard was used as a criterion, the criterion used was considered a valid comparator.

Discussion

To the best of our knowledge, this is the first systematic review to identify the instruments used to assess the therapeutic alliance in physiotherapy and to evaluate the methodological quality of the validation studies of these instruments against the COSMIN checklist. It is also the first systematic review to identify, evaluate, and summarize the evidence for the instruments used to measure the therapeutic alliance in physiotherapy. We conducted this systematic review in accordance with the COSMIN and PRISMA standards; thus, our method followed expert recommendations.⁷^–⁹

Although an interest in measuring the therapeutic alliance began some time ago in the fields of medical care and psychotherapy, this interest is just beginning in physiotherapy. Like psychotherapy and general medicine, physiotherapy involves a high level of patient–clinician interaction; however, the characteristics of the patient population, as well as the interventions, differ between these disciplines.⁵^,⁷ Thus, instruments that have been validated in physiotherapy settings can increase our understanding of the therapeutic alliance in physiotherapy. This study identified four such instruments.

The therapeutic alliance is a multidimensional concept and difficult to measure; therefore, using valid and reliable instruments to measure it is of special relevance.⁷^,¹⁴^–¹⁵ In deciding which instrument is best, clinicians and researchers should consider different factors, such as the setting in which the instrument will be used, the population, the number of dimensions and items, and the methodological quality of the evaluation of each psychometric property. The language and culture of the setting in which the instrument will be used must also be taken into account.

The results of this study indicate that internal consistency and reliability were the two most evaluated psychometric properties. However, most of the psychometric properties in the COSMIN checklist were not evaluated for four of the five questionnaires (reported in four articles), and none of them were evaluated for content validity. The methodological quality of the studies was fair for most of the psychometric characteristics analyzed, although cross-cultural validity obtained the lowest score. The lack of description of the percentages and management of lost values contributed to lower quality ratings.

The WAI was the instrument most commonly used to measure the therapeutic alliance; it was used in two of the four articles included in this review. It is the most validated and widely published measure of therapeutic alliance;¹²^,¹³ thus, it is the measure of therapeutic alliance for which the largest number of psychometric properties have been evaluated. In this regard, and on the basis of its psychometric properties, the WAI is the best instrument to measure therapeutic alliance.

The WATOCI and SRS were evaluated in separate studies;¹²^–¹⁴^, therefore, additional studies are necessary to carry out a broader evaluation of the psychometric properties of these instruments. In this review, we identified no physiotherapy instrument that had been adapted to Spanish, although Spanish is one of the most widely spoken languages in the world.

Although we conducted a meticulous search in several databases, it is possible that we did not identify all the instruments available to measure the therapeutic alliance. Because the search strategy was limited to articles published only in English and Spanish, it is possible that articles written in other languages were not identified.

Conclusion

Few instruments have been validated to evaluate the therapeutic alliance in physiotherapy. However, evaluating this alliance can be useful in both the clinical field and the literature; it is an area in which new knowledge is needed, and it would allow physiotherapists to identify and evaluate interventions that improve care. If physiotherapists want to identify such interventions, they must have valid and reliable instruments. The lack of evaluation of psychometric properties is a limitation of the instruments currently available. The WAI is the best instrument available, but more studies of high methodological quality need to be carried out to analyze the psychometric properties of all instruments and how they are adapted to different cultures.

Key Messages

What is already known on this topic

Validated instruments for assessing the therapeutic alliance in physiotherapy can be useful in clinical and research settings, and selecting the most appropriate one is challenging. To date, no systematic review has been performed to compare the measurement properties of the instruments available with the COnsensus-based Standards for the selection of health Measurement Instruments (COSMIN) methodology.

What this study adds

This systematic review identifies, assesses, and summarizes the evidence for measures used to assess the therapeutic alliance in physiotherapy. Four studies were found, which used four instruments, and the WAI was found to found to have the strongest psychometric properties. This measure has also had the greatest number of psychometric properties assessed according to the COSMIN standard.

References

1.Safran JD, Muran JC. Negotiating the therapeutic alliance: a relational treatment guide. New York: Guilford Press; 2003. [Google Scholar]
2.Pizer SA. The negotiation of paradox in the analytic process. Psychoanal Dialogues. 1992;2(2):215–40. 10.1080/104818892095389309. [DOI] [Google Scholar]
3.Gelso CJ, Carter JA. Components of the psychotherapy relationship: their interaction and unfolding during treatment. J Couns Psychol. 1994;41(3):296–306. 10.1037/0022-0167.41.3.296. [DOI] [Google Scholar]
4.Hall AM, Ferreira PH, Maher CG, et al. The influence of the therapist-patient relationship on treatment outcome in physical rehabilitation: a systematic review. Phys Ther. 2010;90(8):1099–110. 10.2522/ptj.20090245. Medline:20576715 [DOI] [PubMed] [Google Scholar]
5.Corbella S, Botella L. La alianza terapéutica: historia, investigación y evaluación. An Psicol. 2003;19(2):205–21. Spanish. [Google Scholar]
6.Babatunde F, MacDermid J, MacIntyre N. Characteristics of therapeutic alliance in musculoskeletal physiotherapy and occupational therapy practice: a scoping review of the literature. BMC Health Serv Res. 2017;17(1):375. 10.1186/s12913-017-2776-0. Medline:29233138 [DOI] [PMC free article] [PubMed] [Google Scholar]
7.Mokkink LB, Terwee CB, Patrick DL, 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(4):539–49. 10.1007/s11136-010-9606-8. Medline:20169472 [DOI] [PMC free article] [PubMed] [Google Scholar]
8.Terwee CB, Mokkink LB, Knol DL, et al. Rating the methodological quality in systematic reviews of studies on measurement properties: a scoring system for the COSMIN checklist. Qual Life Res. 2012;21(4):651–7. 10.1007/s11136-011-9960-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
9.Moher D, Liberati A, Tetzlaff J, et al. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. PLoS Med. 2009;6(7):e1000097. 10.1371/journal.pmed.1000097. [DOI] [PMC free article] [PubMed] [Google Scholar]
10.Schönberger M, Humle F, Zeeman P, et al. Working alliance and patient compliance in brain injury rehabilitation and their relation to psychosocial outcome. Neuropsychol Rehabil. 2006;16(3):298–314. 10.1080/09602010500176476. Medline:16835153 [DOI] [PubMed] [Google Scholar]
11.Hersoug AG, Høglend P, Monsen JT, et al. Quality of working alliance in psychotherapy: therapist variables and patient/therapist similarity as predictors. J Psychother Pract Res. 2001;10(4):205–16. Medline:11696646 [PMC free article] [PubMed] [Google Scholar]
12.Degnan A, Seymour-Hyde A, Harris A, et al. The role of therapist attachment in alliance and outcome: a systematic literature review. Clin Psychol Psychother. 2016;23(1):47–65. 10.1002/cpp.1937. Medline:25445258 [DOI] [PubMed] [Google Scholar]
13.Heinonen E, Lindfors O, Härkänen T, et al. Therapists’ professional and personal characteristics as predictors of working alliance in short-term and long-term psychotherapies. Clin Psychol Psychother. 2014;21(6):475–94. 10.1002/cpp.1852. Medline:23813617 [DOI] [PubMed] [Google Scholar]
14.Karel Y, Thoomes-de Graaf M, Scholten-Peeters G, et al. Validity of the Flemish Working Alliance Inventory in a Dutch physiotherapy setting in patients with shoulder pain. Physiother Theory Pract. 2018;34(5):384–92. 10.1080/09593985.2017.1400141. Medline:29120251 [DOI] [PubMed] [Google Scholar]
15.Araujo AC, Filho RN, Oliveira CB, et al. Measurement properties of the Brazilian version of the Working Alliance Inventory (patient and therapist short-forms) and Session Rating Scale for low back pain. J Back Musculoskelet Rehabil. 2017;30(4):879–87. 10.3233/BMR-160563. Medline:28282795 [DOI] [PubMed] [Google Scholar]