The World Health Organization (WHO) has identified osteoarthritis (OA) and rheumatoid arthritis (RA) as chronic musculoskeletal conditions projected to increase in prevalence with an ageing population (Woolf & Pfleger 2003). When conservative management of RA and OA is no longer effective and quality of life diminishes, primary total joint arthroplasty (TJA) for the hip and knee is commonly recommended as a successful treatment for advanced OA (Dunn 2012; O’Neill, McCabe & McBeth 2018).

In South Africa and other low- and middle-income countries (LMICs), however, waiting lists for TJA are long, with significant costs involved in managing these patients (Abera Abaerei, Ncayiyana & Levin 2017). As per the statistics kept at a tertiary hospital in Gauteng, in 2024, there were 2089 patients awaiting arthroplasty and the next available theatre booking date was May 2029. Similar waiting times are common across academic and regional hospitals throughout the province (Dunn 2012). These delays stem from limited resources, the high cost of joint replacements and the prioritisation of trauma cases, which restricts bed availability and operating theatre time for elective surgeries. Consequently, patients face prolonged periods of chronic pain and limited mobility. Furthermore, waiting longer than 6 months can negatively impact post-operative satisfaction and patient-reported outcomes 1 year after total knee arthroplasty (TKA) (Desmeules et al. 2009). In a study using the EuroQol five-dimension questionnaire, many patients reported that the wait for an arthroplasty felt ‘worse than death’ (Scott, MacDonald & Howie 2019).

Multiple factors contribute to extensive waiting lists and high costs, including limited application of outcome-based, multidisciplinary, enhanced recovery pathways (Plenge et al. 2018; Swenson et al. 2018). However, there are no current, accepted national or regional policies on TJA or care pathway guidelines that offer recommendations. There are some South African–based studies which suggest that prehabilitation could serve as a key component of enhanced recovery. Studies have shown a link between inpatient length of stay and cost reduction and recommend addressing patient-related factors before surgery (Plenge et al. 2018; Swenson et al. 2018).

In the authors’ experience, socio-cultural ideologies, perceived benefits of physiotherapy, fear of pain and other factors impact post-operative mobilisation and serve as inhibitors to progress and have been studied in this context but in a different population group (Dangor et al. 2021). Other anecdotal factors that may indirectly affect mobilisation include the presence of long-standing unmanaged foot deformities, poor biomechanics, reduced movement efficiency because of back pain, chronic pain in other areas and a lack of understanding regarding lifestyle management. An exploration of these factors was briefly analysed in a scoping review of prehabilitation intervention by the first author (Pillay-Jayaraman, Chetty & Maddocks 2025); however, linkage in this population has not been specifically studied. All these factors can hinder early post-operative mobilisation in this specific setting. Currently, there is no available literature addressing the presence and influence of these factors in this context or in this population group.

Addressing these issues requires careful consideration of the context and needs of patients, as well as the development of a context-specific programme. It is well recognised that public health challenges are often rooted in complex social, political and economic factors (Moffatt et al. 2006). Additionally, empowering patients in their recovery by patient education and improving information-seeking behaviour are suggested as one of the most critical determinants of improved post-arthroplasty outcomes because of the low literacy levels in this cohort (Plenge et al. 2018). Hence, stakeholder input is essential, particularly in South Africa, because of all the above reasons and the fact that practices from high-income countries cannot be implemented as is. Plenge et al. (2018) serve a rationale for this study. As an upper-middle-income country (The World Bank 2022), South Africa provides arthroplasty in both state-funded tertiary centres and private hospitals. However, approaches proven in well-resourced settings may not be applicable because of the unique socioeconomic landscape in South Africa, shaped by a history of inequality impacting numerous sectors including healthcare (Plagerson 2023).

To design an effective programme, it is essential to conduct a needs assessment. This involves identifying the ‘felt need’, or individual perception of healthcare needs, and determining whether these are expressed as demands for healthcare (O’Brien 2010). Thus, a comprehensive understanding of the relationship between felt needs, expressed needs, socio-cultural factors and context is critical (Jinks, Ong & Richardson 2007) for developing an impactful service.

In our study, we explored patients’ existing knowledge and perceptions of physiotherapy and prehabilitation using a mixed-methods approach, enabling a robust, in-depth analysis. Exploring existing knowledge is essential so the programme and its various elements can address the misconceptions, if any, and will also enable the programme to be pitched and built up from the current level of understanding. We also examined patients’ educational needs regarding physiotherapy and prehabilitation, with the aim of informing a tailored prehabilitation programme. This contextually grounded study, we believe, is an essential first step towards implementing programmes that can reduce waiting times and improve patient outcomes. This study is phase two of a larger study, and the protocol has been previously published (Pillay-Jayaraman et al. 2025). Phase one of the study was an extensive literature review which informed elements of the design of this study as well as the questionnaire design and has been published (Pillay-Jayaraman, Chetty & Maddocks 2023). Phase 3 involved comparing standard care pathway to the prehabilitation pathway. The standard care pathway was informally tested, and its efficacy reported in an unpublished manuscript and adopted as the standard pathway in a published manuscript which details the specific intervention and timelines (Ebrahim et al. 2021).

Our study was conducted at a large tertiary teaching hospital in Soweto, Johannesburg, South Africa. This hospital, the third largest globally and the largest in the Southern Hemisphere, serves a population of over 3.5 million people with a capacity of 3 200 beds. Patients with hip and knee pain are referred by a general practitioner in a primary healthcare setting if their condition warrants a higher level of care. On arriving at the tertiary hospital, patients are managed by orthopaedic specialists, and information regarding their condition, various aspects of care and primary arthroplasty operation as a treatment of choice is given during one-on-one consultation with the orthopaedic surgeon. Referral to relevant multidisciplinary healthcare professions is done by them who will refer patients as needed. Hence, it is the discretion of an orthopaedic surgeon to refer patients to physiotherapy or other multidisciplinary healthcare practitioners. Patients who had been on the arthroplasty waiting list for over a year and were attending the outpatient clinic were included in the study. Arthroplasty clinics were held twice a week on Tuesday and Thursday between 8:00 and 12:00 for new patients, repeat appointments and post-operative patients, all following the designate queue. Every second patient in the queue who met our inclusion criteria and was reasonably conversant in English was invited to participate. This method was chosen to preserve an element of randomness, as the outpatient arthroplasty clinic was the only place patients congregated. The quantitative prospective, single-centre study took place from January 2022 to May 2022. Data collection beyond this period was not done, as most of the patients had already been recruited, and no new patients consented to be in the study. The length of time to answer the questionnaire and be in the interview could have been a deterrent to further recruitment. In addition, most patients reported to be physically and mentally fatigued, as they had to leave their home early, travel a distance in a taxi, queue for extended periods of time standing to collect their hospital file and then queue to see the orthopaedic surgeon. Most patients were also reliant on family, friends, paid transport to travel to and from the appointment as well as to accompany them because of the extent of disability and distance to access transport and hence could not wait longer than absolutely necessary. Patient recruitment and data collection were managed by an independent researcher who was not a part of the study and who was conversant in the local language, and a total of 49 patients were recruited to answer the questionnaires, while seven patients agreed to in-depth interviews.

Those who consented were taken to a private room where they completed a questionnaire using the Research Electronic Data Capture (REDCap), a secure web-based platform for data collection and storage. The independent researcher facilitated completion on her smartphone and was available to clarify questions as needed.

The questionnaire (see Table 1 and Online Appendix 1) was divided into six sections, and knowledge was assessed with a scoring system: a ‘No Knowledge’ score of 0 was assigned to patients who responded ‘Don’t know’ or left answers blank; a ‘Fair’ score of 1 for 1–2 correct answers; ‘Good’ score of 2 for 3–4 correct answers and ‘Excellent’ score of 3 for all correct responses.

Data were exported to Excel and SPSS for analysis by a biostatistician. Descriptive data were analysed with percentages, and significance was set at p < 0.05. Pearson’s chi-square and Fisher’s exact tests were used to examine correlations. Open-ended responses were coded and organised into thematic categories.

The qualitative component used convenient sampling of seven patients awaiting arthroplasty, and the final number of seven patients was arrived upon because there was saturation of the information being collected. Hence, this was the deciding factor in considering the data collection process as complete, and further recruitment was deemed redundant.

Data were collected through individual, semi-structured interviews (Online Appendix 2), conducted by an experienced researcher fluent in IsiZulu, the patients’ primary language. Interviews were flexible, audio-recorded and later transcribed and translated into English by a professional linguist. Analysis followed the principles of conventional content analysis (Hsieh & Shannon 2005). Prior to coding and analysis, member checking was conducted through post-interview summary validation, whereby participants were provided with a brief summary of their interview responses and key themes shortly after transcription. They were invited to confirm, clarify or correct the interpretations to ensure accuracy and resonance with their intended meaning. This was done to ensure participant validation and to enhance credibility and trustworthiness.

The analysis began with repeated reading to achieve immersion and a comprehensive sense of the data. This was followed by a detailed, word-by-word examination to identify key concepts and ideas. Through this process, labels for codes were developed through a conventional content analysis approach, as described by Hsieh and Shannon (2005), whereby coding categories were derived inductively from the data without imposing preconceived categories capturing overarching thoughts. Codes were then organised into themes, reflecting the interrelations and patterns among them, and grouped into meaningful clusters for interpretation. Although coding was conducted inductively following conventional content analysis (Hsieh & Shannon 2005), a deductive element was introduced during the interpretive phase. Emergent themes were subsequently mapped onto established frameworks (International Classification of Function [ICF] and Patient-Reported Outcomes [PROMs]) to guide synthesis and ensure coherence across qualitative and quantitative strands, reflecting a hybrid analytic strategy appropriate to mixed-methods research. A flow diagram illustrating the integration of qualitative and quantitative data in a convergent mixed-methods design is depicted in Figure 1. Frameworks were applied to guide analysis and synthesis, with triangulation used to compare themes and enhance interpretive validity.

Ethical clearance to conduct this study was obtained from the University of KwaZulu-Natal and the Biomedical Research Ethics Committee on 16 August 2022 (No. BREC/00002141/2020).

Knowledge of physiotherapy, patients’ exposure and prior referral to physiotherapy was assessed by open- and closed-ended questions. While 100% of the respondents had been managed with pain medication, only 14.2% (7) were referred for physiotherapy.

In response to the question on the role of a physiotherapist (Figure 2), a significant (p < 0.001) 77.6% (38) of patients indicated that they were unsure of the role or function of a physiotherapist. Pearson’s Chi-square and Fisher’s exact test was conducted to ascertain the relationship between this knowledge and education level (Table 3: Educational level of the sample), that is, completed matric/not completed matric, and it was found that no significant relationship existed with p = 0.240.

A significant 67.3% (33) of respondents were unsure of the necessity for pre-operative exercises (p < 0.001). Only 8.2% believed exercises should not be done, while 24.5% (12) indicated that pre-operative exercises were necessary.

When asked about the type of information they would have liked regarding prehabilitation from the open-ended questions, patients expressed a need for guidance on exercise types, exercise indications and precautions, visual examples of positive patient outcomes and instructions on assistive device use. Specific areas of interest included exercises for muscle strengthening, gait improvement, pain relief and stiffness prevention. One patient noted:

Patients also enquired if physiotherapy could have prevented the need for surgery and expressed concerns about exercise safety and its impact on pain. One patient remarked:

Some respondents, however, stated that they would not have done exercises because of pain, and a few reported being advised by community clinic physiotherapists to avoid exercises until after surgery.

When asked if they would perform exercises pre-operatively if the need was explained, an overwhelming 93.9% (46) of patients indicated they would be willing to participate (p < 0.001). Additionally, 88% (43) reported that receiving an exercise pamphlet would possibly motivate them to exercise (p < 0.001), and 80% (39) felt that SMS reminders would encourage them to continue with exercise (p < 0.001).

However, 90% of respondents indicated that attending face-to-face therapy once a week was not feasible. Reasons for non-attendance included cost and transportation challenges. Open-ended responses highlighted the financial impact on family members who would need to accompany them, resulting in lost workdays and wages. Patients reported difficulties using public transport because of limited accessibility and expressed discomfort in navigating these spaces with mobility aids. One patient explained:

Another mentioned:

Regarding post-operative exercise timing, the unit’s practice recommends (Pillay-Jayaraman et al. 2025) beginning exercises in bed within a few hours after surgery. However, as shown in Table 4, a significant 83.6% of patients were unaware of this recommendation (p < 0.001).

In response to the question about when patients could start walking post-operation, only 12.2% (see Table 4) were aware that mobilisation could begin the day after surgery. However, a significant 85.7% were unaware of the timing for walking post-surgery, a statistically significant finding (p < 0.001).

Additionally, 75.9% (22) of patients did not know what functional abilities were required before discharge, a result that was also statistically significant (p < 0.001, see Figure 3). Specifically, 90% (18/20) of patients awaiting hip replacement were uncertain about the discharge criteria they needed to meet, which again showed statistical significance (p < 0.001).

The data collected from the semi-structured in-depth interviews are presented in the following. The demographic information of the seven patients interviewed is summarised in Table 5, and Table 6 contains a summary of the patient characteristics and baseline knowledge. The names contained herewith are pseudonyms. Participants’ ages ranged from 55 years to 78 years, with most having waited over 3 years for their surgery. Notably, most participants had not completed their matriculation and presented with associated comorbid conditions.

Three main themes were identified during analysis: patients’ preconceptions of physiotherapy, the perceived value of prehabilitation and patients’ recommendations for physiotherapy and prehabilitation. While these thematic areas were broadly informed by the interview guide, the specific content and subthemes were inductively derived from participants’ narratives, consistent with a conventional content analysis approach.

Participants exhibited uncertainty regarding the nature and scope of physiotherapy. For instance a 78-year-old woman awaiting a total hip replacement, remarked:

Many patients derived their understanding of physiotherapy from the experiences of family and friends. For example, a woman also awaiting a total knee replacement, recounted:

Five patients based their understanding of physiotherapy on their experiential knowledge of the profession and their expectations of health benefits. They believed that physiotherapy would facilitate improved mobility, expedite healing and enhance recovery post-surgery. A woman awaiting a total knee replacement, stated:

The theme of perceived value of prehabilitation highlighted that most patients lacked awareness of prehabilitation and its role in preparing them for surgical intervention. This lack of knowledge may stem from their absence of prior experiences with physiotherapy, either personally or through their social circles. One patient expressed this uncertainty:

One patient with prior exposure to prehabilitation noted that, while the programme did not alleviate her pain, the exercises helped prevent joint stiffness. A woman awaiting a total hip replacement, said:

Similarly, other participants reported experiencing benefits in range of motion as a result of prehabilitation. A 61-year-old woman awaiting a total knee replacement, stated:

The final theme identified was patients’ recommendations regarding essential information that should be included in a prehabilitation programme, particularly about surgical procedures and post-rehabilitative care. Generally, patients expressed a desire for detailed information about their options and how best to prepare for surgery. The patient articulated this need:

Patients also expressed eagerness to understand the implications of engaging in prehabilitation versus not participating. Another woman awaiting a total knee replacement, asked:

Moreover, patients highlighted the importance of general health management, including weight and fitness, for surgical preparation and post-operative recovery. The patient further expressed:

Although a limitation of qualitative small group studies is the reduced transferability of results to other contexts, this article employed a mixed-methods approach, which we believe strengthened our findings. A few questions in the questionnaire and method of implementation could have led to responders giving socially desirable answers. The study also presents sampling and recruitment bias, whereby recruiting every second patient in the queue maintained an element of quasi-randomness but still skewed representation. The results have limited generalisability, wherein findings from a tertiary hospital may not reflect broader trends across other regions or healthcare settings, especially rural or private sectors. Patients’ physical and mental fatigue, long travel times and reliance on others may have deterred participation, especially among those with more severe disability or socioeconomic constraints. The implementation of the questionnaires in English only limited participation and is an inherent bias. However, the study serves as a baseline feasibility investigation.

In order to ensure generalisability, it is recommended that the study be expanded to a multi-centre, multi-health system health frameworks to get equitable representation across different race, language and socioeconomic groups.