Physical Therapists’ Choices, Views and Agreements Regarding Non-Pharmacological and Non-Surgical Interventions for Knee Osteoarthritis Patients: A Mixed-Methods Study

Objective/Aims: The aims of this study are to collect the most common non-pharmacological and non-surgical interventions used by the Portuguese physical therapists in their knee osteoarthritis patients, and to deeper understand the factors associated to their intervention choices. Methods: This study incorporated a mixed-methods design. For the quantitative data it was choose an e-survey (with 25 close-end questions, plus general information of the study and a clinical vignette), retrieving sociodemographic and self-reported practice on knee osteoarthritis information. It was analysed response frequencies and associations between variables with logistic regression analyses. For the qualitative data, it was chosen to perform semi-structured interviews in purposefully selected physical therapists to include different sociodemographic factors and survey responses regarding the physical therapists’ interventions chosen. After the interviews, the audios were collected, anonymised, transcribed verbatim, and the texts explored by the thematic approach. Results: From the 277 PTs that shown interest in participating in the study, 120 fully completed the questionnaire and, from those, 10 participated in the interviews. The most chosen interventions included Resistance Exercise, Manual Therapy, Nutrition/Weight Loss, Self-care/Education, Stretching and Aquatic Exercise. Furthermore, it seems that PTs’ individual characteristics (age, experience, and clinical reasoning), patient’s characteristics (clinical findings and preferences), and work-related factors (facility type, work environment and available resources) are the main actors responsible for an intervention chosen. Conclusions: In the Portuguese PTs context the most important interventions are Exercise, Manual Therapy, Nutrition/Weight Loss, and Self-care/Education; these interventions chosen may be influenced by PT, patient and work-related factors.


INTRODUCTION
Osteoarthritis (OA) is the most common form of arthritis and, from all joints, the knee OA is the most prevalent. 1,2 Current knee OA rehabilitation strategy is a complex process, where it may be used surgical and non-surgical interventions. [3][4][5] There are several non-pharmacological and non-surgical interventions that can be used to manage patients with knee OA, the majority physical therapy related. [6][7][8][9][10] Despite being widely used to manage patients with knee OA, physical therapy practice has been subjected to decades of criticism for its lack of research, and is often perceived as a profession that bases its practice largely on anecdotal evidence, using treatment techniques that have little scientific support. 11 This was identified, as early as 1969, to be a significant issue for the physical therapy profession. 12 Over the years, many efforts were made to increase physical therapy research 13 and to shift from the traditional models of practice (guided on the therapist tacit knowledge and opinion) to a more evidence-based practice (EBP) overtime. 11,14,15 So, the aims of this study are to collect the most common non-pharmacological and non-surgical interventions used by the Portuguese physical therapists (PTs) in their patients with knee OA, and deeper understand the factors associated to their intervention choices.

MATERIAL AND METHODS
This study incorporated a concurrent mixed-methods design [16][17][18][19] and followed the Ethical Principles of the Helsinki Declaration (2013). 20 Additionally, it was approved by the ethics committee (CEFADE24-2019) and all PTs enrolled were informed and signed the individual inform consent form.

Sample
In an attempt to ensure the correct population sample, the national physical therapy professional association (APFISIO) e-mail database was requested for the Portuguese PTs working class recruitment. Also, in order to increase the number of enrolled participants, the e-mails of past students from all physical therapy national schools were requested.

Design -Quantitative
For the quantitative data, it was chosen to apply a self-administered e-survey. The e-survey was evaluated, designed, administered, conducted and collected according to established guidelines. [21][22][23] The e-survey was initially e-mailed all voluntary PTs in the APFISIO database in the regular online newsletter and to past PT students as a formal e-mail with a cover letter containing the study's information (background, justification and aims). Additionally, after reading the study's information, the participants were invited to click in the e-survey link (https://pt.surveymonkey.com/r/PBE2019FADEUP). When clicking the link, the participants were then connected to the SurveyMonkey and forwarded to the e-survey. Before initiating the e-survey, the informed consent, the data protection rights, and how the results will be used (analysed anonymously and confidentially, the data gathered was only used for statistical information in an academic environment), the criteria for selecting the participants and the reasons for non-participation, the possibility to stop the e-survey at any time, the information that no incentives will be provided, instructions how to fill and complete the e-survey, and e-mail address for possible clarifications, were explicitly stated. The e-survey included 25 close-ended questions, divided into 2 main stages (the e-survey may be found in the supplemental data): 1. Sociodemographic information. At this stage, in addition to collecting sociodemographic information, the participants' eligibility was also analysed with the inclusion and exclusion criteria: • Inclusion: have an active physical therapy license; obtained at least the physical therapy bachelor's degree; work or have worked as a PT in the past 6 months in Portugal; be able to read, write, and speak Portuguese. • Exclusion: do not have an active physical therapy license or have another profession than PT; obtained the physical therapy bachelor's degree in a foreign country; does not work in Portugal; is not be able to read, write or speak Portuguese; be a physical therapy bachelor student. 2. Most frequently used non-pharmacological and non-surgical interventions applied in patients with knee OA. The respondents were invited to rank by preference 5 non-pharmacological and non-surgical interventions for managing patients with knee OA, from 31 available interventions options. The interventions options were achieved after a preliminary literature search. In order not to bias the PTs interventions choices, the interventions appeared in a random order, not repeating its order from e-survey to e-survey. For helping to contextualise, a knee OA clinical vignette was provided (translated to Portuguese from the Holden et al. 24  Before sending the e-survey by e-mail, the e-survey was pre-tested by the authors and evaluated in its completion time, design, questions order, attractiveness, syntax, clarity, logic, correct question types, and response format. Also, it was permitted to the respondents to review and change their answers. The sample size goal for this study was 373 responses, based in a 95% confidence level, a margin of error of 5% and a 50% response distribution. 25 To ensure that the sample size goal was achieved, after two, four and six weeks respectively, a thank you note and a reminder containing the e-survey link was e-mailed. In an attempt to avoid duplication filled questionnaires, only responses were accepted for each IP address.

Design -Qualitative
For the qualitative data collecting, it was chosen to apply semi-structured interviews with open-ended questions on the PTs. The interviews were conducted by 1 PhD and methodological experienced author, blinded to the PTs characteristics and prior questionnaire answers, using Skype (Microsoft Corporation, Rives de Clausen, Luxemburg). Only audio-recorded was performed -excluding any face-to-face or written contact. There was no relationship between the interviewer and the PTs prior to the study, and the interviewees were blind regarding the interviewer' characteristics (an "anonymous" e-mail and Skype account were created). The interviewees were recruited by completing the study during previous stages where, following a review of questionnaire responses, the sample was purposefully selected to include different sociodemographic factors and interventions responses for patients with knee OA. To ensure a high participation rate, after one, two, and four weeks respectively, a thank you note, a reminder containing the interview objectives, and a request to provide their most convenient dates/times for the interview, were e-mailed. The semi-structured interviews were performed according to Leech et al. 26 guidelines. The questions in the interview script were constructed according to Qu et al. 27 The interview script was properly validated by an external expert panel (of 2 independent and methodological experienced PhDs), where there were able to comment and suggest improvements. Before initiating the "core" questions, an introductory section with the purpose of the study, the protection rights, how the data will be used and some warm-up questions were included in order to build empathy and comfort. The "yes" or "no" answers were avoided. At the end of the core questions, it was given the opportunity for the interviewees to add information and opinions that they found to be relevant. Additionally, the interview script was tested on the first participant who, after the interview, was asked for feedback on the interview conduction, structure, design and phrasing of questions. The script may be consulted in the supplemental data.

Data Analysis -Quantitative
Response frequencies were analysed using Microsoft Excel and IBM SPSS 26.0 software. After examining the response frequencies, the variables categories were collapsed. In the interventions choices, the "1 st ", "2 nd ", "3 rd ", "4 th ", and "5 th " were combined so that a 2-category response was obtained: "Present" (if the PT chooses 1 st , 2 nd , 3 rd , 4 th or 5 th ) or "Absent" (no intervention choice). Additionally, in sociodemographic data where subsamples were smaller, we collapsed categories in an effort to derive stable models. The Certificate and Baccalaureate degrees into the same category (Baccalaureate) -as in Portugal they are the minimum required professional entry-level -and our sample included only 1 PT who indicated a professional Post-Doctorate degree, so we included him/her with others PhD degrees. After item categories were collapsed, logistic regression analyses were conducted to examine the associations with the PTs' characteristics. An alpha level of 0.05 was used to determine whether a model was to be reported. Odds ratios (OR) and their 95% confidence intervals (CIs) were determined for each level of the independent variables in those models that were significant. 28 Data Analysis -Qualitative The data was analysed with a Computer Assisted Qualitative Data Analysis Software, namely the NVivo v12 (QRS International, Doncaster, Victoria, Australia). 29 The audios collected in the interviews were anonymized and verbatim transcribed. Then the texts were explored by 3 authors with the thematic approach. 30 The original classification tree was analysed and further discussed with an external expert panel of 3 methodological experienced PhDs,, where some categories were collapsed, eliminated, or renamed. Quotations were identified to report the findings and illustrate the content, and were translated from Portuguese to English. To ensure complete and transparent data reporting, the methodology was conducted according to established guidelines. 31-35

Quantitative
From the 227 PTs that shown interest in participating in the study, only 120 (52.9%) fully completed the questionnaire (Figure 1). The descriptive statistics of the PTs personal and practice characteristics are presented in Table 1.
Regarding the statistically significant associations between PTs' characteristics and the most used knee OA interventions, Aerobic and Resistance Exercises were more likely to be chosen by the least experienced therapists, in comparison to more experienced PTs (OR  . For a more detailed information, consult the word cloud provided in the supplemental data. With the interviews, the main themes identified were: Interventions (applied, eventually applied, and not applied); Intervention plan rationale; Physical therapy sessions frequency; and Principal and secondary knee OA symptoms. For a more detailed information, the classification tree and codes are in the supplemental data. The summary of the qualitative results is described in Figure 4. More detailed information is included in the supplemental file texts and quotations.

DISCUSSION
In the present Portuguese context and after all the data gathering, the most important interventions to manage patients with knee OA are: Exercise, Manual Therapy, Nutrition/Weight Loss, and Self-care/Education.

Exercise
In the Exercise group, the most important interventions were: Aquatic exercises, Balance exercises, Resistance exercises, and Stretching. From those, more emphasis needs to be given to Resistance exercises. Resistance exercises was the most chosen intervention in Exercise group (73%), being in 2 nd and 3 rd place in the general level of importance. This was also found in other countries. 24 Although all patients should be advised to perform exercises to improve both physical and psychological outcomes, they should be patient appropriate. 41,42,54,61-63 As confirmed in the qualitative data, the exercise movements should be pain free and should respond to the patients' preferences and clinical findings. So, firstly, preliminary pain relief interventions can be used to allow pain free exercises practice, secondly, patients' exercises preferences and pain tolerance should be addressed and, thirdly, the exercises should be adapted and adjusted to their individual physical, physiological, social, and emotional characteristics, kinesiophobia, co-morbidities, and other clinical findings. All this will ensure a better exercise plan adhesion and participation. 39,47,54,61-65

Manual Therapy
Although a substantial decrease in its use was found compared to previous Portuguese studies, [66][67] Manual Therapy was still the most important intervention for the Portuguese PTs. In fact, 31% of the PTs chose it in 1 st place. This importance may be explained by: (1) Therapy related factors; and (2) Profession related factors. Manual Therapy is a very versatile intervention, has a highly interventions variation, and could be easily combined with other interventions in the PT daily practice (for example, in our study, 25 interactions were found in a sample of 28 interventions). 10  TITLE patients, even if the Manual Therapy is considered placebo, feel more comfortable and more enthusiastic if there is a direct contact between PT-patient. 10 and, as showed in the qualitative data, their choice will depend on: patients' preferences; adhesion and individual characteristics; clinical findings and patients' co-morbidities; signs and symptom types and severity; and workplace and PTs' characteristics.

Self-care/Education
In relation to Self-care/Education, this intervention achieved mixed results. Although in the quantitative data it is not the most chosen intervention, in the qualitative data PTs considered it as the most important. One reason for this discrepancy is that PTs considered it as a mandatory intervention and should be present in all patients "since day 1". In fact, although not often prescribed, PTs naturally performed it. As so, many of the PTs could not choose it in our e-survey, as they almost see it as a moral duty and not so much as an intervention. Nevertheless, this intervention was integrated in the 4 th most chosen interventions combination, being easily associated with other interventions (21 interactions in a 28 interventions sample) and used in all signs and symptoms approached. To proper perform it, it is important to adapt the information to the patients' health literacy and provide different information supports (oral and written). 42,89 If the PT do not adapt the information to the patient's health literacy or provide it in just one way, the information transmitted could be lost or misunderstood. To ensure that the patient truly understand the information given, a simple test could be performed, the so-called "Kieran O'Sullivan test". This test suggests that the PT should ask the patients to describe how they will explain the information given to their family (or significant other) when they come back to home. Evidence highly recommends its use in these patients OA disease progression; knee anatomy; pathophysiology; joint protection; home exercises and self-care techniques; and overall lifestyle changes. The objective is to promote hope, optimism, and a positive expectation of the benefits of the intervention plan. 39,41,42,45,46,53,64,89 Furthermore, during PT-patient communication, PTs should avoid using "wear and tear", "it's your age", "nothing can be done for you", or "give up" expressions, as they could result in negative feelings in the patients regarding the intervention plan and the OA progression. 90 Other As shown, other interventions were applied due to personal, patient, and work-related factors. Similar factors were found in other countries and health care professions, showing that they could condition the interventions choice. 81,91,92 In the personal factors, PTs' age and experience may have an important role in the intervention plan design. In our study, Exercise modalities were more chosen in young and less experienced PTs. This may be explained by evidence access and given importance. 93 Younger PTs could be more technologically advanced and could access evidence quicker compared to their older peers. Furthermore, in their intervention plan rationale, they cannot balance evidence and clinical experience equally, they have experience deficits. In other hand, older and more experienced Portuguese PTs may have less ability to reach evidence and tend to follow their clinical experience even more. 67 In fact, it is expected that only half of the PTs use databases to aid in clinical decision-making. 93 Additionally, personal doubts about evidence and treatment effectiveness may also exist. 91 Explanation for this may include 93,94 : poor quality evidence; contradictory clinical practice guidelines (CPGs) recommendations; poor quality in the information transmission; PTs inability to understand statistical data; lack of skills in searching and critically appraise evidence; lack of data generalisation for the patient; and not enough explored OA factors, such as economic aspects of recommendations or the patients' co-morbidities influence. Facilitators may include 94-95 : regular clinical cases and evidence peers discussion; higher quality studies; CPGs concordance; better information reaching with an user-friendly format; CPGs should become patient-focused rather than disease-driven. Academic degree, belonging to a practice-oriented organisation, and participate in continuing educational courses may also influence the Portuguese PT practice, however further studies are needed to understand their true importance. Patient was a central piece on the decision-making process puzzle. Almost all PTs reported that the interventions choice was from the patients' signs and symptoms, co-morbidities, and other clinical findings (such as pain, ROM limitations, muscular weakness, and activities of daily living restrictions). There were similar to evidence-reported most important factors. [96][97][98] As the interventions are applied in the patients, the PTs also though that their preferences have an important role. Nevertheless, in a deeper analysis, the PTs used it in their clinical-making intervention plan more as a way to decide between two equal effective interventions, or as "bargaining chip" to introduce more evidence-based interventions. Therefore, a better PT-patient communication and interaction is mandatory, as well as more importance needs to be given to their preferences, providing them with a more active participation in the intervention plan decision. 79,99 Patients are often septic and pessimist about interventions and OA progression. 63,99 So, other factors may also be important to increase the knee OA patients' optimism, satisfaction and security, such as 79,99 : good PT accessibility, deviation, convention, prioritising therapeutic over financial consideration; PT competence; feeling that their opinions and preferences are taken into account; and their intervention plan is individualised.
The structure of the system in which PTs worked influenced their knee OA treatment approach. In this factor, two main issues raised; money and time. As in Portugal the salaries are low and the PTs services are considered as cheap, sometimes the PTs have to give in to the patients demands (even if the PT do not agree with the intervention efficacy) as they could lose a client and consequently money (since most of the small health care units are client-financially dependent). In other hand, there are bigger health care units that are stated-financially dependent, so many times the PTs have to do what is medically prescribed and stated funded. In fact, 88-90% of the Portuguese patients with knee OA reach physical therapy after general practitioners consultation and prescription. 66,67 Lack of money could also result in a lack of resources (such as, technological clinical equipment or computers), influencing the interventions choice. 93 Similar to what was found in other studies, 93 time was one of the largest work-related barriers. In our study, the PTs needed time to evaluate, review and treat patients, and for extra work activities such as evidence or skills improvements. Comparable concepts were found in UK PTs. 54 Also, the workplace environment itself could be a barrier. 93 One of the most important barrier to the Portuguese PTs is that in the workplace it is not given enough importance if they do (or not) an EBP. 67 Other barriers found in the literature include 93 : lack of support from the employer; and colleagues not favourable to EBP.

Limitations
One limitation of this study was the number of valid questionnaires. The sample size goal of 373 was not reached. Therefore, the results could not truly represent the Portuguese PTs practice. Another limitation was found in the qualitative data, where the instruments used in the patients' TITLE evaluation and follow-up were not fully explored and understood. Finally, it would also be interesting to have conducted the study with different clinical vignettes to understand how patients' characteristics, the level of pain, joint range of motion, functionality, physical activity, or other clinical findings influence the choice of intervention.

CONCLUSION
In conclusion, in the context of Portuguese PTs, the most important interventions are Exercise (specially, Resistance Training), Manual Therapy, Nutrition/Weight Loss and Self-care/Education. PTs individual characteristics (age, experience, and clinical reasoning), patient's characteristics (clinical findings and preferences), and work-related factors (facility type, work environment, and available resources) are the main actors responsible for the use (or not) of an intervention.

AUTHOR CONTRIBUTIONS
RMF, RSG and NP were responsible for the manuscript design. RMF and RSG were responsible for the sample gathering. PNM, RSG, and NP were responsible for the data analysing. All authors read and approved the final manuscript.

ACKNOWLEDGMENTS
The authors would like to thank all PTs that participated in the study.

CONFLICTS OF INTEREST
The authors declare that they have no conflicts of interest.

FUNDING
The authors declare that they have no funding source.   Over how many pages was the questionnaire distributed? The number of items is an important factor for the completion rate.

ETHICS APPROVAL AND CONSENT TO PARTICIPATE
-

Completeness check
It is technically possible to do consistency or completeness checks before the questionnaire is submitted. Was this done, and if "yes", how (usually JAVAScript)? An alternative is to check for completeness after the questionnaire has been submitted (and highlight mandatory items). If this has been done, it should be reported. All items should provide a non-response option such as "not applicable" or "rather not say", and selection of one response option should be enforced. -

Review step
State whether respondents were able to review and change their answers (eg, through a Back button or a Review step which displays a summary of the responses and asks the respondents if they are correct). Count the unique number of people who filled in the first survey page (or agreed to participate, for example by checking a checkbox), divided by visitors who visit the first page of the survey (or the informed consents page, if present). This can also be called "recruitment" rate.

10
Completion rate (Ratio of users who finished the survey/users who The number of people submitting the last questionnaire page, divided by the number of people who agreed to participate (or submitted the first survey page). This is only relevant if there is a separate "informed consent" page or if the survey goes over several pages. This is a measure for attrition. Note that 10

Checklist Item Explanation Page Number Describe survey design
Describe target population, sample frame. Is the sample a convenience sample? (In "open" surveys this is most likely.) 6 IRB approval Mention whether the study has been approved by an IRB. 5

Informed consent
Describe the informed consent process. Where were the participants told the length of time of the survey, which data were stored and where and for how long, who the investigator was, and the purpose of the study? 6 Data protection If any personal information was collected or stored, describe what mechanisms were used to protect unauthorized access. -

Development and testing
State how the survey was developed, including whether the usability and technical functionality of the electronic questionnaire had been tested before fielding the questionnaire. "completion" can involve leaving questionnaire items blank. This is not a measure for how completely questionnaires were filled in. (If you need a measure for this, use the word "completeness rate".) Cookies used Indicate whether cookies were used to assign a unique user identifier to each client computer. If so, mention the page on which the cookie was set and read, and how long the cookie was valid. Were duplicate entries avoided by preventing users access to the survey twice; or were duplicate database entries having the same user ID eliminated before analysis? In the latter case, which entries were kept for analysis (eg, the first entry or the most recent)?
-IP check Indicate whether the IP address of the client computer was used to identify potential duplicate entries from the same user. If so, mention the period of time for which no two entries from the same IP address were allowed (eg, 24 hours). Were duplicate entries avoided by preventing users with the same IP address access to the survey twice; or were duplicate database entries having the same IP address within a given period of time eliminated before analysis? If the latter, which entries were kept for analysis (eg, the first entry or the most recent)? 8 Log file analysis Indicate whether other techniques to analyze the log file for identification of multiple entries were used. If so, please describe. -

Registration
In "closed" (non-open) surveys, users need to login first and it is easier to prevent duplicate entries from the same user. Describe how this was done. For example, was the survey never displayed a second time once the user had filled it in, or was the username stored together with the survey results and later eliminated? If the latter, which entries were kept for analysis (eg, the first entry or the most recent)? -

Handling of incomplete questionnaires
Were only completed questionnaires analyzed? Were questionnaires which terminated early (where, for example, users did not go through all questionnaire pages) also analyzed? 10 Questionnaires submitted with an atypical timestamp Some investigators may measure the time people needed to fill in a questionnaire and exclude questionnaires that were submitted too soon. Specify the timeframe that was used as a cut-off point, and describe how this point was determined. -

Statistical correction
Indicate whether any methods such as weighting of items or propensity scores have been used to adjust for the non-representative sample; if so, please describe the methods. Review step State whether respondents were able to review and change their answers (eg, through a Back button or a Review step which displays a summary of the responses and asks the respondents if they are correct).

7
Unique site visitor If you provide view rates or participation rates, you need to define how you determined a unique visitor. There are different techniques available, based on IP addresses or cookies or both. Count the unique number of people who filled in the first survey page (or agreed to participate, for example by checking a checkbox), divided by visitors who visit the first page of the survey (or the informed consents page, if present). This can also be called "recruitment" rate.

10
Completion rate (Ratio of users who finished the survey/users who The number of people submitting the last questionnaire page, divided by the number of people who agreed to participate (or submitted the first survey page). This is only relevant if there is a separate "informed consent" page or if the survey goes over several pages. This is a measure for attrition. Note that 10 Suppl.

Electrical Stimulation Therapy
Interferential All included PTs pursue a higher academic degree, participate in continuing education courses, and belong to a practice-orientated organisation.
Suppl. Table 4. Classification tree. The stated applied interventions to manage knee OA patients were: Self-care/Education; Electrical Therapy; Exercise; Kinesio Tape (KT); Manual Therapy; Thermal Agents; US; and Vertical Bed. The PTs' choose to use these interventions because: (1) after the patient assessment, they are the interventions that best respond to the patients' signs and symptoms, as well as the treatment objectives created; (2) are in accordance to an EBP; and, (3) give priority to more active interventions (Quotations 1, 2 and 3).  [FT_7] In relation to Self-care/Education, it was performed with the aim of: (1) informing the patient about its condition, OA related issues and intervention plan; (2) ending OA related myths; (3) explaining how to manage their problems alone, including giving some home interventions that could help improving their condition; (4) promoting an healthy life style, such as weight loss; and, (5) during all treatments sessions, explaining how to perform the intervention properly and why they are doing it (Quotations 4, 5, 6, 7 and 8). As a note, a PT added that it is important to adapt the information to the patient's health literacy/academic qualifications, for him/her to receive and better understand the transmitted information (Quotation 9). In the Electrical Therapy group, TENS, Iontophoresis, IFC and NMES were used. Except for NMES, all Electrical Therapy group interventions were applied to reduce pain (Quotations 10 and 11). Between the different analgesic techniques, their choices depended on: (1) what the interventions that health insurers companies contributed financially were; and, (2) which medical devices were present at the health care units (Quotations 12 and 13). Despite this, two PTs were against their use, since their benefits were limited and there is some evidence non-recommendation (Quotation 14). In other hand, NMES was the least stated intervention in this group and it was mainly used as a complementation of active exercises, or as an initiation stage for more debilitated patients (not able to perform active exercises) to gain some strength (Quotations 15 and 16). Exercises were the most frequent intervention used for managing knee OA patients. The interventions stated in this group were: Aerobic exercises; Active mobilizations/Rage of Motion (ROM) gain/Stretching exercises; Resistance exercises; and Functional training/exercises. From all, Resistance exercises were the most used, mainly because of OA progression characteristics (namely, strength loss). Within these exercises, the PTs preferred to start performing isometric exercises, only progressing to other contractions types and exercises after the patients reveal a good strength standard base (Quotation 17). Regarding the Aerobic exercises, they were performed to promote knee movement, reduce impacts, as well as maintain the physical condition, being then a form of preparation for more intense exercises (Quotation 18). With similar objectives were the Active mobilizations/ROM gain/Stretching exercises. The PTs used these exercises not only to increase (or at least maintain) a "normal" knee ROM, but also to decrease the muscular and articular tensions (Quotations 19, 20 and 21). Lastly, they also reported the use of Functional training/exercises. These exercises were applied to train the activities of daily life (such as, stand and sit, stability training and gait training), equipping/educating the patient with a range of strategies to reduce some daily constraints (Quotation 22). The intervention choices in the Exercise group depended, not only on the patients' condition and the OA degree (such as, obesity, ROM limitation and knee inflammation), but most importantly pain. PTs evaluated their patients' pain level and in which exercises the pain intensified more. The aim was always to do pain free exercises, giving to the patients a good movement experience and sensation (Quotation 23). The use of KT was only indicated by 2 PTs. In both cases, although it was used by them, none considered it to be a core intervention. The intervention was considered as complementary. Additionally, in both situations, they were somewhat reticent and septic about its clinical efficacy. The main reason for this mistrust was due to the results found in evidence, which often discredited its use and deconstructed some efficacy related-myths. Nevertheless, its use was mainly for pain relief and knee biomechanical corrections (Quotations 24 and 25). Likewise Exercise, Manual Therapy was one of the most intervention groups mentioned and used by the PTs. Within this group, the most common interventions applied were: Massage; Passive mobilizations; and, Soft tissues mobilization/ manipulation techniques (Quotations 26 and 27). Additionally, Maitland and Mulligan techniques were also referred in this group, however just one PT confirmed their use (Quotation 28). The overall objectives of these interventions were soft tissue relaxation (harmonize muscular tensions, stabilize muscular tonus and release soft tissues), drainage, pain decrease and ROM improvement (Quotation 29). Nevertheless, one PT did not recommend the use of Massage on these patients, because it will not offer any of the overall intended benefits, which are providing mobility and functionality (Quotation 30). Additionally, another PT do not recommend the use of Passive mobilization. He/She preferred to spend most of the time doing more active interventions (Quotation 31). Although the PTs use Thermal Agents, this intervention group was far from gathering consensus. In this group, the two interventions stated were Heat and Ice. All PTs considered Thermal Agents to be a complementary intervention. In fact, most of them just use it because the patient asks for it (Quotation 32). Their use depended on the patient's condition, nonetheless the aim was to decrease pain (Heat and Ice), promote muscular relaxation (Heat) and reduce the inflammatory process (Ice) (Quotations 33, 34 and 35). Despite these advantages, some PTs did not recommend their use. For the Heat interventions, they did not use them because it could increase the knee inflammation process (Quotation 36).

Classification tree
Regarding the Ice interventions, the PTs said that it could increase some of the patients' signs and symptoms (Quotation 37). Likewise, US were used by the PTs, however there was some reluctance in its use. In fact, some stated that they only use it because they have to follow the medical prescription (Quotation 38). The main objective was pain reduction (Quotation 39). Vertical Bed use was only stated by one PT. The intention of its use is for lower limb stretching, load stimulus and imposing knee extension. This could be important for patients who cannot perform exercises or have an excessive knee flexion (Quotation 40).  (3) there is a medical prescription and an obligation to perform them (Quotations 56 and 57). Nevertheless, sometimes it is important to use some of these interventions as they could work as an "open door" to other interventions. As the PT gives in on these less evidence-based interventions desired by the patient, the patient-PT confidence rise and then he/she can negotiate with the patient with other more evidence-based interventions, integrating them into the treatment plan and making more effective sessions.

Other
The PTs reported that, whenever possible, they used an EBP to define the knee OA patients' intervention plan. So, to accomplish it, they have to rely on the information gathered in the evidence, do a proper patients' clinical signs and symptoms evaluation, integrate the patient's preferences, collect the patients' socioeconomic status information and co-morbidities, listen to the patient's main issues and goals, and delimit by their clinical experience what are the most mastered interventions (Quotations 58 and 59). However, it was also not uncommon to find that many of them feel more limited in their practice because they have to follow the physiatrist's intervention plan prescription (Quotation 60 and 61). Regardless of how the intervention plans are designed, to combine and order the various interventions during the treatment session, the rationale will depend on the effects that the PT defines for each intervention. Usually, to order the treatment plan interventions the previous intervention, in addition to have a specific effect, will also help in carrying out the succeeding interventions. For example, a PT referred that in addition to the effect of reducing stiffness and muscle tension (which itself can be beneficial for the patient), the use of Heat as a first intervention can facilitate the performance of other techniques, such as Massage and afterwards Exercises (Quotation 62).
Quotation 62: "… what we usually do is, first, a component of tissue muscle relaxation (we often use some kind of heat, which helps with stiffness, helps to reduce muscle tension, and facilitates the work that we do afterwards), then we use manual therapy (massage, physiological and accessory mobilizations), and exercise or balance training and gait training to patients that need it." [FT_3] Interventions are applied (more often) 2 to 3 times per week (Quotation 63). However, there are still patients who are treated daily (Quotation 64). The number of sessions applied depends on the workplace and the treatment session "recovery" (patients' signs and symptoms easing) (Quotations 65 and 66 As showed, the patient's signs and symptoms have a very important role in both treatment plan interventions selection, and the treatment sessions per week decision and progression. There are several signs and symptoms related with OA stated by the PTs (Quotations 68, 69 and 70). The most important OA sign and symptom was knee pain, followed by activities of daily living limitations, lower limb muscular weakness and knee ROM limitations. By far, knee pain was the most important OA sign and symptom (referred 58 times in the PTs discourses) (Quotation 71). Although PTs believe that many of the signs and symptoms are related to pain, they also point out that most of the times, in a superficial analysis, pain may seem as the main patient problem, but when doing a deeper analysis to the patient's speech at the anamnesis, his/her real problem is not being able to perform his/her activities of daily living (Quotations 72, 73, and 74). Therefore, while they still consider pain as the main sign and symptom, the importance given to activities of daily living limitations should be reviewed. First of all, we appreciate your interest in participating in our questionnaire.
This questionnaire is being carried out at Faculdade de Desporto da Universidade do Porto, as part of a doctoral dissertation, in the Physical Therapy area.
The aim of this study is to characterize the Physical Therapists working class in Portugal and to know which are the most common interventions used in knee osteoarthritis patients.
For that, we are looking for Physical Therapists who: can read and write Portuguese; have completed in Portugal the enter-level degree for working as Physical Therapists; work or worked recently (in the last 6 months) as Physical Therapists in Portugal; have treated a knee osteoarthritis patient in the last 6 months.
During the questionnaire there will be no right or wrong answers, we are simply interested in your options and opinions on knee osteoarthritis. Furthermore, throughout the questionnaire, your participation will be anonymous and the data collected will be confidential, non-transferable, only used statistically in the academic context.
If you think that you will fit the population type and that you agree with the data treatment method, we would be very grateful for your participation in our questionnaire. Your feedback would be very important to us! 1. I declare that I have been informed of the questionnaire objectives and that I have been guaranteed that all participants identification data are confidential and that anonymity will be maintained. I know that I can refuse or interrupt participation in the study at any time, without any type of penalty. I understood the information that was given to me, I had the opportunity to ask questions and my doubts were clarified.
In this way, I also authorize the results obtained to be reveled in the scientific community, guaranteeing confidentiality and anonymity.
I agree to participate freely in the study mentioned above □ I do not agree to participate in the study mentioned above □ Suppl. Table 7. Informed Consent Statement -Questionnaire.