“I could have a proper ankle” – a qualitative study of patients’ perceptions of total ankle replacement and ankle fusion surgery

We employed a qualitative research design to gain an in-depth understanding of patients’ perceptions of TAR and ankle fusion surgery. The study was informed by phenomenology as we sought to explore the meaning and shared understanding of participants’ lived experiences [11]. The study is reported in line with the consolidated criteria for reporting qualitative studies (COREQ) framework [12].

Ethical approval was obtained from the North West – Greater Manchester East Research Ethics Committee (18/NW/0819).

Our research team included researchers with expertise in qualitative research, rheumatology, orthopaedics, podiatry, physiotherapy and biomechanical engineering, and two Patient and Public Involvement (PPI) contributors (SW, JK). Both PPI contributors have lived experience of being offered TAR and ankle fusion, with one having undergone TAR and the other opting to continue conservative management.

Individuals who met the criteria specified in Table 1 were eligible to participate in the study. We purposively selected participants based on type of operation, length of time post-operatively and gender to maximise diversity of perspectives [11].

Participants were recruited from a single orthopaedic centre within a teaching hospital in northern England. Potential participants were identified from clinic lists, approached in person at a routine clinic appointment or via an invitation letter and given the opportunity to discuss the study with a researcher at their clinic appointment or via telephone. All individuals who were screened and met the eligibility criteria agreed to participate in the study, resulting in a sample of seven people. This is in line with typical sample sizes for phenomenological studies, which focus on exploring the experiences of a relatively small number of participants in depth [11, 13].

Semi-structured interviews were conducted between 8th March 2019 and 12th March 2020 using a topic guide (Additional File 1), developed in conjunction with a PPI contributor (SW). Questions were open-ended and structured around the research aim. Each participant completed a single interview and provided written informed consent prior to their interview. One or two female researchers who were previously unknown to the participants, conducted each interview. When two researchers were present, one led the interview and the other mainly observed but was able to ask questions as appropriate. The interviewers included a physiotherapist working as a healthcare researcher (AMA), a biomechanical engineer with expertise in ankle joint interventions (CLB), and an experienced healthcare researcher (AMK). Interviews were conducted face-to-face in participants’ own homes (n = 4) or a hospital-based research centre (n = 3). One participant was accompanied by a family member during their interview. Interviews were digitally recorded, transcribed verbatim and supplemented with field notes. The interview duration ranged from 25 to 74 minutes.

All interview transcripts were uploaded into NVivo 12 (QSR International; 2018) and analysed using reflexive thematic analysis [14, 15]. Initial coding was undertaken collaboratively by two researchers (AMA, CLB). Both researchers read and re-read the transcripts, collaboratively generated initial codes and collated similar codes together, holding regular discussions to review the coding. This was undertaken over a period of approximately eight months. The coding was inductive and focused on manifest content [14, 15]. A third researcher (DA) verified the coding for two randomly selected transcripts to assure credibility. Codes were grouped into provisional themes through a team discussion, then reviewed against the whole data set by two other researchers (HJS, LSC). Discrepancies in coding were settled by group consensus between all coders. Findings were discussed with all research team members, including our PPI contributors (SW, JK).

We employed multiple strategies to address the trustworthiness criteria proposed by Lincoln and Guba [16], including:

Member checking is another strategy that has been proposed to enhance credibility [16]. However, its appropriateness has been questioned due to the interpretative nature of qualitative data analysis and its potential to raise ethical issues [17, 18]. Due to these potential issues, and the fact that we discussed the findings with our PPI contributors, we did not incorporate member checking in the study design.

Our findings must be viewed in light of this study’s limitations. Whilst our sample size of seven is in line with typical sample sizes for phenomenological studies [11, 13], interviewing further participants may have provided additional insights. One option to help address this would have been to continue interviewing participants until we reached data saturation. However, the concept of saturation is inconsistently defined, challenging to assess and arguably inappropriate for the reflexive thematic analysis approach we employed [29]. An alternative to saturation is the concept of information power proposed by Malterud et al. [30]. This suggests that the more relevant information that can be obtained from a sample to address the study aim, the fewer participants are necessary. The seven interviews we conducted provided detailed and relevant information that directly addressed our study aim, supporting the appropriateness of our sample size.

Overall, our purposive sampling strategy captured a broad range of patients regarding gender, age, type of surgery, and duration since surgery. However, we did not interview any patients who had undergone ankle fusion revision surgery. In addition, we recruited participants from a single orthopaedic centre; therefore, our findings may not be transferable to other patients due to variations in service provision. We excluded patients unable to understand and speak English to ensure that participants could provide true informed consent and participate fully in the interviews. In addition, we did not record the ethnicity of participants to help preserve their anonymity, as relatively low numbers of patients undergo TAR. This meant we could not explore the impact of differing cultures and ethnicities on patients’ perceptions. This is an important limitation as previous research has highlighted racial disparities in utilisation of TAR and ankle fusion [31, 32]. Another potential limitation is that the participants were aware that members of the research team were based at the recruitment site, which may have discouraged them from criticising their care. This did not appear to be a substantial issue as the participants expressed both positive and negative perceptions.

Our findings demonstrate that education and clinical support for patients undergoing TAR/ankle fusion is not always perceived as adequate by patients. Identifying and addressing such inadequacies is vital to inform patients’ decision-making, guide their preparations for surgery and support their post-operative recovery. We found that patients’ individual circumstances and beliefs had a strong influence on their decision-making and perceptions of their post-operative outcomes, highlighting the need to personally tailor support. Given the lack of definitive evidence in this area, developing consensus-based guidelines on pre- and post-operative support for patients undergoing TAR/ankle fusion would be valuable to guide service provision. There is also a need for further qualitative research to address areas not covered in our study. For example, further research is required to explore which outcomes are most important to patients undergoing TAR/ankle fusion and other stakeholders to inform the development of a standardised core outcome set for ankle surgery.