A Questionnaire of Physiotherapeutic Specific Exercises of Scoliosis—QPSSE

Study design: Creating and psychometric testing of a new QoL Questionnaire about Physiotherapeutic Specific Exercises of Scoliosis (Questionnaire of Physiotherapeutic Specific Exercises of Scoliosis—QPSSE). Purpose: The purpose of this study is to create a reliable and valid questionnaire for patients suffering from mild and moderate adolescent idiopathic scoliosis (AIS) who have been treated with Physiotherapeutic Specific Exercises of Scoliosis (PSSE) in order to evaluate their quality of life. Materials and Methods: The developed questionnaire was based on a thorough literature review as well as on authors’ experience. It consists of 53 questions, of which 37 have a positive meaning, 15 have a negative meaning, and 1 is a multiple choice question; additionally, there are 6 “open” questions. Except for the multiple choice question, all other questions are answered on a Likert scale ranging from 1 to 5 points. Five represents a positive meaning or very positive one, whereas one stands for a negative meaning or none at all. Questions were developed by the authors who subsequently categorized the 53 questions into the following eight domains: physical functioning, self-image, Physiotherapeutic Scoliosis-Specific Exercises (PSSEs), psychosocial functioning, cognitive functioning, compliance, motivation, and pain. A pilot study was conducted so that we could calculate Cronbach’s Alpha based on the outcome. Due to the COVID-19 pandemic, the authors worked through the Zoom online platform to structure the questionnaire. Results: Pearson’s correlation coefficient was used for all correlations evaluated. P values of less than 0.05 were considered to be significant. Internal consistency was evaluated with Cronbach’s Alpha. Although there were very few missing values, accounting for 0.78% of the total values of the questionnaire, the expectation maximization likelihood algorithm was used to impute data. IBM® SPSS® Statistics Software v.25 was used for the analysis. Cronbach’s Alpha coefficients for the overall score were 0.84. Conclusions: This original QPSSE was found to be a reliable and valid tool for AIS treated conservatively with PSSE and for the patients’ clinicians.


Introduction
In the existing literature for idiopathic scoliosis (IS), the following is widely reported: the common type of it [1,2], its 3-D character [3,4], the incidence of adolescent idiopathic scoliosis (AIS) [4], and how an early diagnosis is made performing the forward bending test, called Adam's test [5].Also, the curve that is named scoliosis according to the SRS is discussed [6,7].Additionally, a plethora of publications report many possible genetic and epigenetic factors [4,8] as causes of its development.
Although scoliosis is considered as a harmless condition, evidence suggests that people with scoliosis are more prone to experiencing back and low-back pain [9,10].Scoliosis causes multiple dysfunctions and appears to be a burden on healthcare, particularly when patients require extensive surgical treatment [11,12].
The application of Physiotherapeutic Specific Exercises for Scoliosis (PSEE) alone or with bracing is currently one of the therapeutic models for mild and moderate IS.
As far as the definitions of the severity of scoliosis go, there is not full agreement on what is mild and moderate idiopathic scoliosis.Mild idiopathic scoliosis is characterized by a Cobb angle of more than 10 and less than 30 degrees [13], of more than 10 but less than 25 degrees [14], and of more than 10 but less than 20 degrees [4].Moderate IS is characterized by a Cobb angle of 25-40 degrees, which is indicated for non-operative treatment [2,15] and a Cobb angle greater than 21 to 35 degrees [4].We consider as mild curves those with a Cobb angle of greater than 10 but less than 20 degrees and as moderate those with a Cobb angle of greater than 21 to 35-40 degrees.
SOSORT uses the term Physiotherapeutic Specific Exercises for Scoliosis (PSSE) for all approved schools and methods.Each method and school incorporates the SOSORT guideline principles and shares a common goal, that of stabilization, arresting progression of IS, and improving the quality of life of patients.The methodology of the PSSE must be based on scientific evidence and adjusted according to the type of the deformity of each patient [16,17].Another protagonistic and high-priority common goal during the implementation of PSSEs, as recommended by SOSORT, is "three-dimensional selfcorrection" [4,18].Self-correction can be defined as the best possible trunk alignment that a patient with scoliosis can achieve in the three planes and axes [19].
The existing PSSE schools that were created involve the following [17,20]: • Schroth, Germany In Greece, many of the above methods are applied for the rehabilitation of IS.To our knowledge, a specific questionnaire for PSSEs has not yet been published.In this report, a questionnaire is presented, which aims to evaluate the quality of life of children and adolescents with diagnosed IS, who are undergoing treatment with one of the above PSSE methods.

The Questionnaire
For the development of the PSSEQ (Physiotherapeutic Scoliosis-Specific Exercises Questionnaire), we followed the recommended guidelines [21] for designing a questionnaire.It is based on a thorough literature review of the field of scoliosis and Physiotherapeutic Scoliosis-Specific Exercises (PSSEs), which are approved by the SOSORT organization [4], as well as on our experience regarding the needs of scoliotic children and adolescent patients in specialized physiotherapeutic centers, in our country.Permission and approval for conducting the research were requested and provided by the ethical committee of the Medical Association of Piraeus.In addition, because participants were minors, consent was provided by their parents for using data from their children who participated in the study.
The questionnaire consists of 53 questions.There are 37 with positive meaning and 15 with negative meaning and 1 multiple choice question.Furthermore, 6 open-ended questions were included related to the PSSEs and the questionnaire.In the scored items, a Likert scale of 1 -5 was used.More specifically, responses 1-5 correspond to 5 = Strongly agree and 1 = Strongly disagree.An additional edit was made to the wording of the questions so that they could correspond meaningfully to these responses.
The first sheet of the questionnaire includes questions about the date of data collection, demographic characteristics, first name and surname, father's and mother's names, gender, date of birth, address, contact details, height and weight in kilos to calculate BMI, right or left handedness, date of menarche for females, hair and eye color, and type of scoliosis.Data to be filled in by the researcher include the Scoliometer angle, Cobb angle, Risser sign, (Formetric-scoliosis angle), type of treatment (exercise or both exercise and bracing).
A pre-final PSSE questionnaire was developed and a pilot study was conducted, which was completed by 16 participants with AIS.We asked the 16 participants to complete this questionnaire so that we could calculate Cronbach's Alpha based on the results in their responses.The total sample responses to the questions were entered in SPSS and for each domain, the minimum and maximum effect (floor and ceiling effect), defined as the percentage of participants showing the minimum and maximum possible scores, respectively, were calculated.Results at the minimum and maximum effect exceeding 15% were considered statistically significant.The floor effect was 53/265 and the ceiling effect was 265/265, respectively.Cronbach's Alpha was calculated as >0.7, which is considered a strong correlation-standard-deviation result between responses.Internal consistency of the questionnaire was evaluated by calculating the Cronbach's Alpha.The initial number of questions was 64; thus, 11 were excluded from the study or combined with the existing questions just to become more specific and easier for the patients to understand.Thus, the final PSSE questionnaire consists of 53 questions, and 6 "open" questions; see Supplementary Materials.The original questionnaire for PSSEs was developed in the Greek language and was coined QPSSE.

Study Population
The inclusion criteria for participating in this study are the following: (1) patients should be from 10 to 18 years old, (2) patients should have been diagnosed with mild or moderate AIS, (3) patients should be able to speak and read in Greek, and (4) patients should be undergoing or have undergone in the past Physiotherapeutic Scoliosis-Specific Exercises (PSSEs) for a period of at least 2 months.
The exclusion criteria were mental health problems, a low level of communication, congenital, neurological, or another type of scoliosis, and/or having or had surgery for scoliosis.
Eighty patients qualified and were included in the study.For their characteristics, see below in Section 3.1.

Psychometric Evaluation
The PSSE questionnaire was evaluated for the following psychometrics: reliability, validity, and floor and ceiling effects.Reliability was assessed by analyzing internal consistency and test-retest reliability.Internal consistency was determined using the Cronbach's Alpha.The intraclass correlation coefficient (ICC) was calculated to measure the test-retest reliability.The patients completed the questionnaire twice at an interval of 4-7 days for the measuring of test-retest reliability.Convergent validity was compared against the critical value of Pearson's correlation.Divergent validity was evaluated by analyzing the answers of the patients of the PSSEQ and their characteristics (e.g., Risser sign, Formetric, Cobb angle, gender, etc.) using Pearson's or Spearman correlation coefficient.The IBM SPSS Statistics v.25 was used for the statistical analysis.

Formetric 4D-DIERS
The Formetric 4D DIERS is a scanning system with a light projector, which scans the back of the patient.This system is connected to a computer that analyzes the data and provides information about the posture of the body, spinal curves (frontal, lateral), pelvic position, vertebral rotation, and muscle imbalance [19].

BMI
Body mass index (BMI) involves a method that calculates body fat according to height and weight in females or males.A normal BMI range is between 18.5 and 24.9 [22].

Cobb Angle
The Cobb angle measurement is used for calculating spine curvatures in the frontal plane.A radiograph is necessary in order for the Cobb angle to be measured.This method is used so as to determine the upper/lower-end vertebras (UEV/LEV) on the radiograph; then, a vertical line at the, respectively, upper/lower-end vertebra endplate lines is necessary (UEVEL/LEVEL), and the included angle of the two vertical lines is the Cobb angle [23].

Risser Sign
The Risser sign is used by clinicians in order to assess the skeletal maturity of a human.The Risser sign is determined using the iliac apophysis from radiographs and is classified in 6 stages (0-5 Risser).It is often used for the evaluation of adolescent idiopathic scoliosis and for the selection of its treatment [24].

Demographics
Demographics are characteristics of a population and are often evaluated for a statistical analysis.Some of them are age, gender, ethnicity, education, geographic location, etc.In this study, age, gender, and color of hair and eyes were used in the statistical analysis.

Results
Time needed to complete the PSSEQ was about 10-11 min.Eighty patients were included in the study and 21 test-retests were completed in a period of 14 months.

Statistical Analysis
Based on the Kolmogorov-Smirnov goodness-of-fit test and Shapiro-Wilk test for normality, data did not follow the normal distribution; therefore, non-parametric tests were used for the statistical analysis.Pearson's correlation coefficient was used for all correlations evaluated.p values of less than 0.05 were considered to be significant.Internal consistency was evaluated through the Cronbach's Alpha method.Although there were very few missing values, accounting for 0.78% of the total values of the questionnaire, the expectation maximization likelihood algorithm was used to impute data.IBM ® SPSS ® Statistics Software v.25 was used for the analysis.

Factor Analysis
The results of the content validity analysis demonstrated excellent reliability and content validity for the questionnaire, as summarized in Table 1.

Internal Consistency Reliability
Cronbach's Alpha coefficients for the overall score were 0.84, exceeding the minimum recommended standard of 0.70 and indicating satisfactory internal consistency.

Item Convergent Validity
The criterion for item convergent validity was the correlation coefficient of each item of each domain with the domain scale variable.This value was compared against the critical value of Pearson's r equal to 0.219, taking into account 78 degrees of freedom for each comparison with a p < 0.05.

Item Divergent Validity
The criterion for item divergent validity was the correlation coefficient of each domain scale variable and the clinical continuous variables (age, BMI, Cobb angle).Except for the pain domain variable that was positively correlated with age, there was no other significant correlation found between the aforementioned variables, indicating the lack of relationship of the 53 measurements of the questionnaire with the clinical data.

Test-Retest Reliability
Test-retest reliability was assessed using Pearson's correlation coefficient, r.Twentytwo subjects were re-evaluated one week after the first interview.The results of the questionnaire were used for 53 discrete bivariate correlations, 1 for each variable of the questionnaire.The results showed that there was perfect test-retest reliability in terms of achieving r values equal to 1 in all correlations.

Floor and Ceiling Effects for the Overall Score
For the overall score, in the present study, 0% of patients scored at the floor and 0% scored at the ceiling.Therefore, there were no floor or ceiling effects for the overall score.Floor and ceiling effects for each domain are shown in Table 1.Floor and ceiling effects for each item are shown in Table 2.
The total values for all scale scores, as well as the range of possible scores, are shown in Table 3.

Discussion
The aim of this study was to develop a questionnaire that evaluates the quality of life in the Greek population with idiopathic scoliosis, who are undergoing Physiotherapeutic Specific Exercises for Scoliosis (PSSE), and the evaluation of its psychometrics.
As it is described below, many other good questionnaires are used to assess scoliosis but a specific treatment for a PSSE questionnaire was not developed until the present.This makes the difference from other questionnaires developed to assess the QoL for IS.
This treatment-specific questionnaire, the QPSSE, was created in order to determine how various parts of treatment of IS using PSSEs influence patients with AIS, and this is its strong point, which is not provided using other pertinent-to-IS questionnaires.
Previous studies have developed other questionnaires that evaluate similar aspects of patients with AIS after other treatments, either conservative or surgical.One questionnaire that evaluates the quality of life regarding AIS is the 22-item revised questionnaire of SRS (SRS-22R) that evaluates quality of life regarding AIS, especially after surgery treatment [25].Another one is the Brace questionnaire (BrQ) that evaluates quality of life in populations with IS treated with a brace [26].Also, Short-Form Health Survey 36 (SF-36) or Short-Form Health Survey 12 (SF-12) that evaluates Health-Related Quality of Life (HRQoL) [27,28] and the spinal appearance questionnaire (SAQ) aim to assess self-image in patients with AIS [29].
SRS-22-R is one of the most frequently used questionnaires for patients with scoliosis and includes 22 items that are divided in five domains: pain, self-image, function, mental health, and satisfaction with management [30].QPSSE was divided into eight domains.Some of them were, also, pain, self-image, and psychosocial functioning.As it was mentioned before, SRS-22R was mainly intended for patients with AIS treated with surgery.However, to date, no questionnaires have been developed to evaluate the quality of life for patients with AIS, so SRS-22R was often used for evaluating general quality of life for these patients regardless of their treatment.
BrQ is a self-administrated questionnaire that evaluates the quality of life for patients with AIS who are treated with a brace.There is a 34-item Likert scale that consists of eight domains such as QPSSE general health perception, physical functioning (physical functioning, also, in QPSSE), emotional functioning (psychosocial functioning in QPSSE), self-esteem and aesthetics (self-image in QPSSE), vitality (motivation in QPSSE), school activity, bodily pain (pain in QPSSE), and social functioning (psychosocial functioning in QPSSE) [26].
SF-36 is a 36-item self-reported questionnaire and one of the most widely used Health-Related Quality of Life questionnaires.It is divided into eight sections: (1) vitality or energy (motivation in QPSSE), (2) physical functioning (also in QPSSE), (3) bodily pain (also in QPSSE), (4) general health perceptions, (5) physical role functioning, (6) emotional role functioning, (7) social role functioning (psychosocial functioning in QPSSE), (8) mental health or emotional wellbeing (cognitive functioning in QPSSE).SF-12 is a smaller version of SF-36.These questionnaires have also been used for patients with AIS in order to evaluate their general quality of life [27,28].
The 32-item SAQ is a questionnaire based on the Walter Reed Visual Assessment Scale (WRVAS) and it evaluates perception of spinal appearance by patients with AIS.After validation, a 20-item SAQ was developed and divided into nine domains, and three textual items about the most distressing aspects of deformity.The items of SAQ are divided into the following domains: general (three items), curve (one), prominence (two), trunk shift (two), waist (three), shoulders (two), kyphosis (one), chest (two), and surgical scar (one), and an extra three textual questions as it was mentioned.The SAQ has been widely used and culturally adapted into many languages for the assessment of the appearance in patients with AIS.The original Greek-QPSSE (has) also included a domain about the self-image of patients with AIS (four items) [31].
The PSSEQ results showed that there was perfect test-retest reliability in terms of achieving r values equal to 1 in all correlations and an overall score for internal consistency of 0.84, exceeding the minimum recommended standard of 0.70 and indicating satisfactory internal consistency.The Greek version of SRS-22 was shown to have three domains with a very satisfactory Cronbach's α (pain, 0.85; mental health, 0.87; self-image, 0.83) and for two domains (function/activity, 0.72; satisfaction, 0.67), they were good.The intraclass correlation coefficient (ICC) was >0.70, demonstrating very satisfactory or excellent test/retest reliability [30].The initial Greek BrQ was shown to have satisfactory internal consistency with a Cronbach's Alpha of 0.82 [26].The Greek version of SF-36 was found to have a Cronbach's α > 0.70 [32].The original English SAQ had good-to-excellent reliability (Spearman's rho, 0.57-0.99)and high internal scale consistency (Cronbach's Alpha > 0.7) [29].
Divergent validity in QPSSE was found with no other significant correlation between the aforementioned variables, indicating the lack of relationship of the 53 measurements of the questionnaire with the clinical data.Concurrent validity of SRS-22 was evaluated through its correlation with SF-36 domains analyzing Pearson Correlation Coefficients, and all correlations were found to be statistically significant [30].Other cultural adaptations of SAQ evaluated convergent validity by correlating SAQ with the appearance domain of SRS-22R and divergent validity by correlating patients' answers in SAQ with their characteristics, demonstrating good-to-excellent results [31,[33][34][35][36] as in this study of the original Greek QPSSE.
The original Greek version of QPSSE showed similar results with the Greek versions or other studies of SRS-22, BrQ, SF-36, or SF-12 and SAQ.All these questionnaires evaluate the quality of life for patients with AIS generally or after a treatment or assess Health-Related Quality of Life in general.Our study assessed the intervention after Physiotherapeutic Scoliosis-Specific Exercises, BrQ after bracing, and SRS-22 especially after surgery.However, further study of the QPSSE is needed in order to evaluate more psychometric properties, such as convergent validity or responsiveness.These tests would be necessary so as to determine if the PSSQ is responsive to changes.In order to assess convergent validity, it would be essential for other similar questionnaires or tools about the PSSEs and quality of life to be developed.Furthermore, a further study with a greater number of participants would potentially have better results in psychometrics in a questionnaire evaluating the quality of life in the Greek population with AIS.
This questionnaire is a significant tool for the clinicians and physical therapists who work using PSSEs for adolescents with IS, in order to evaluate their patients' quality of life and interventions after exercises.This tool will provide clinicians with information about Greek patients with AIS so as to improve or change something in their treatment or intervention.It would be essential for this questionnaire to be translated into other languages too, so that other countries could have a tool available for evaluation of quality of life for patients who undergo PSSEs.
A limitation of the application of this QPSSE may be considered as the time period of scoliosis exercise treatment, which is not clearly agreed upon and recommended on any curve type, and consequently the outcomes based on the generated data will need carefully conducted further studies including long-term follow ups.

Conclusions
In conclusion, the PSSE questionnaire was found to be reliable and valid for clinical use for patients with AIS treated conservatively with PSSEs or both PSSEs and a brace in the Greek population.

Table 1 .
The questionnaire domains and the results of tests of item convergent validity, item consistency reliability, and floor and ceiling effects.

Table 2 .
Floor and ceiling effects (percentage of respondents with minimum/maximum scale scores) for each item of the questionnaire.

Table 3 .
Mean values and ranges for each item of the questionnaire.