“We’re trained to trust our patients”: a qualitative study on the general practitioners’ trust in patients for colorectal cancer shared care Authors: Faith R Yong, Sundresan Naicker, Kerry Uebel, Maria Agaliotis, Christopher Chan, John D T Nguyen, Thanya Pathirana, Alexandra Hawkey, Kylie Vuong Published: 2023 ABSTRACT AbstractBackgroundIn a therapeutic partnership, physicians rely on patients to describe their health conditions, join in shared decision-making, and engage with supported self-management activities. In shared care, the patient, primary care, and specialist services partner together using agreed processes and outputs for the patient to be placed at the centre of their care. However, few empirical studies have explored physicians’ trust in patients and its implications for shared care models.AimTo explore trust in patients amongst general practitioners (GPs), and the impacts of trust on GPs’ willingness to engage in new models of care, such as colorectal cancer shared care.MethodsGP participants were recruited through professional networks for semi-structured interviews. Transcripts were integrity checked, coded inductively, and themes developed iteratively.ResultsTwenty-five interviews were analysed. Some GPs view trust as a responsibility of the physician and have a high propensity for trusting patients. For other GPs, trust in patients is developed over successive consultations based on patient characteristics such as honesty, reliability, and proactivity in self-care. GPs were more willing to engage in colorectal cancer shared care with patients with whom they have a developed, trusting relationship.ConclusionsTrust plays a significant role in the patient’s access to shared care. The implementation of shared care should consider the relational dynamics between the patient and health care providers.Keywords: cancer survivors, colorectal neoplasms, delivery of healthcare, integrated, general practice, physician–patient relations, trust Key messages Trust is key to the therapeutic partnership Trust is key to the therapeutic partnership We found different trust conceptualizations We found different trust conceptualizations Some GPs view trust in patients as an innate responsibility Some GPs view trust in patients as an innate responsibility Some GPs’ trust in patients were shaped over time Some GPs’ trust in patients were shaped over time GPs are more willing to engage in new care models with patients they trust GPs are more willing to engage in new care models with patients they trust The rollout of new care models should consider relational dynamics. The rollout of new care models should consider relational dynamics. Background More people are living with and beyond a cancer diagnosis. Shared cancer care arrangements involve partnering of the patient, primary care, and specialist services to support self-management using agreed processes and outputs.1 It offers the patient greater control over their healthcare2 and is the preferred care modality among both patients and physicians for delivering whole-person patient-centred care.3–7 However, cancer shared care remains largely specialist-led.4,8 Mutual physician–patient trust is an essential component of this therapeutic partnership, with positive impacts on diagnoses, satisfaction, adherence to treatment, continuity of care, and health care integration.9 Patients rely on physicians to be competent, dependable and to communicate well.10 Physicians rely on patients to provide accurate information, join in shared decision-making, and engage with supported self-management activities.11,12 Improving mutual trust provides one mechanism for operationalizing cancer shared care.7 There is a large body of literature on ways to improve the patients’ trust in physicians.9,13–15 However, few empirical studies focus on physicians’ trust in patients and its impact on care.16 This study aims to explore: (i) trust in patients amongst general practitioners (GPs) and (ii) the impacts of trust on GPs’ willingness to engage in new models of care, such as colorectal cancer shared care. Methods Design This qualitative study utilized semi-structured interviews.17 This study is reported in accordance with the standards for reporting qualitative studies.18 Setting The study was undertaken in Australia, where the 5-year survival rate for colorectal cancer has increased from 52% (1986–1990) to 71% (2014–2018).19–21 Australian GPs hold specialist registration22 and have important responsibilities in colorectal cancer prevention, detection, and long-term management.4 GP consultations are subsidized by Medicare, the universal health insurance scheme.23 Nearly all Australians consult with their GP at least once per year.24 The study was undertaken in 2021 during the COVID-19 pandemic, where strict lockdowns were implemented in both New South Wales (NSW) and Victoria. The research team is composed of practising GPs with research expertise in cancer survivorship and chronic disease management, and academics in pharmacy, health services research, and implementation science. Participants GP participants were recruited through professional networks via email, posts on closed members-only GP noticeboards, and Twitter. Eligible participants were practising GPs who were either vocationally registered, non-vocationally registered, or enrolled in an accredited training program. Participants were offered an AUD$120 shopping voucher for their time. Data collection Participants completed a short questionnaire covering demographic information and colorectal cancer experiences (Supplementary Material 1) before undertaking a semi-structured interview with a researcher (FY, MA, SN or KV) via telephone or Microsoft Teams. The interview guide (Supplementary Material 2) focused on the GP’s trust in the patient and implications for shared care. Where GPs had no experience with colorectal cancer shared care, they were asked to consider hypothetical colorectal cancer shared care or discussed examples of other shared care arrangements, such as antenatal shared care. The interviews were audio-recorded, professionally transcribed verbatim, verified and de-identified. Preliminary data were discussed successively with the research team; these discussions guided recruitment efforts, which continued until thematic saturation was achieved.25 Informed verbal consent was obtained from all participants and recorded separately. Data analysis The transcripts were analysed thematically26 in parallel with data collection to facilitate identification of data saturation. Two researchers (FY and AH) read through a subset of the interview transcripts independently to identify potential codes. Distinct codes were agreed upon and utilized to construct a preliminary coding framework, as described in codebook thematic analysis.26 Codes or categories were generated iteratively, and the framework refined. The coded data were then grouped into higher level concepts and categories to identify relevant themes. To achieve agreement, regular discussion among the research team further revised and refined themes. Consensus was reached for the final themes. QSR NVivo (version 12) software27 was used to facilitate data management. Results We conducted interviews with 26 GPs in September–December 2021. One interview recording was inaudible, leaving 25 interviews for analyses. The participants were aged between 28 and 65 years old, with an even gender distribution (Table 1). The majority worked in metropolitan NSW, reporting varied experience in general practice. Below, we describe the themes of how GPs developed trust in patients and their decision-making for engaging in colorectal cancer shared care. Table 1.Participant characteristics.GPAge rangeGenderCountry of birthYears in general practiceLocation of primary practiceaRelative socio-economic advantage at primary place of practiceb,c160–69Man/MaleAustralia0–4 yearsMetropolitan areaMedian advantage/disadvantage260–69Man/MaleAustralia25+ yearsRegional centreSome advantage320–29Man/MaleAustralia0–4 yearsMetropolitan areaMost advantaged430–39Man/MaleAustralia10–14 yearsMetropolitan areaSome advantage520–29Woman/FemaleAustralia25+ yearsMissing dataMissing data650–59Woman/FemaleAustralia20–24 yearsMetropolitan areaMost disadvantaged750–59Woman/FemaleAustralia0–4 yearsRemote communityMost disadvantaged830–39Man/MaleAustralia25+ yearsMetropolitan areaMost advantaged950–59Man/MaleTaiwan25+ yearsMetropolitan areaMost disadvantaged1030–39Woman/FemaleAustralia5–9 yearsMetropolitan areaMost disadvantaged1130–39Woman/FemaleAustralia5–9 yearsMetropolitan areaMost advantaged1260–69Man/MaleAustralia25+ yearsMetropolitan areaSome disadvantage1340–49Woman/FemaleAustralia10–14 yearsMetropolitan areaSome disadvantage1440–49Woman/FemaleAustralia15–19 yearsRural townSome advantage1550–59Woman/FemaleUK25+ yearsMetropolitan areaSome disadvantage1630–39Man/MaleAustralia5–9 yearsMetropolitan areaMost advantaged1740–49Woman/FemaleAustralia15–19 yearsMetropolitan areaSome disadvantage1830–39Woman/FemaleUK5–9 yearsMetropolitan areaMedian advantage/disadvantage1940–49Woman/FemaleAustralia15–19 yearsMetropolitan areaMost disadvantaged2030–39Woman/FemaleAustralia5–9 yearsMetropolitan areaSome advantage2140–49Man/MaleAustralia15–19 yearsMetropolitan areaSome disadvantage2240–49Woman/FemaleAustralia15–19 yearsMetropolitan areaSome advantage2340–49Man/MaleIndia5–9 yearsMetropolitan areaSome disadvantage2440–49Man/MaleAustralia15–19 yearsMetropolitan areaMedian advantage/disadvantage2540–49Man/MaleAustralia15–19 yearsMetropolitan areaMost disadvantaged aModified Monash Model: MM1 areas are described as a Metropolitan area, MM2 as Regional centre, MM3-5 as Rural town, and MM6-7 as Remote community. Each index has its own descriptor but we have summarized the classifications for readability. bIndex of Economic Resources (IER) chosen for seeing relative access to economic resources in general i.e. financial aspects of relative socio-economic advantage or disadvantage. Where there were multiple IER classifications for one postcode, all IER indices were recorded, and the median index value was reported in the table. cWhere two IER classifications existed for one postcode, the lower (i.e. more disadvantaged) index value was reported. The scale for IER does not have a description, but IER 1 represents most disadvantage, and IER 5 least disadvantage for relative access to economic resources in general. For the purposes of this paper, we have described IER 1 as most disadvantaged, IER 2 as some disadvantage, IER 3 as median advantage/disadvantage, IER 4 as some advantage, and IER 5 as most advantaged. Participant characteristics. aModified Monash Model: MM1 areas are described as a Metropolitan area, MM2 as Regional centre, MM3-5 as Rural town, and MM6-7 as Remote community. Each index has its own descriptor but we have summarized the classifications for readability. bIndex of Economic Resources (IER) chosen for seeing relative access to economic resources in general i.e. financial aspects of relative socio-economic advantage or disadvantage. Where there were multiple IER classifications for one postcode, all IER indices were recorded, and the median index value was reported in the table. cWhere two IER classifications existed for one postcode, the lower (i.e. more disadvantaged) index value was reported. The scale for IER does not have a description, but IER 1 represents most disadvantage, and IER 5 least disadvantage for relative access to economic resources in general. For the purposes of this paper, we have described IER 1 as most disadvantaged, IER 2 as some disadvantage, IER 3 as median advantage/disadvantage, IER 4 as some advantage, and IER 5 as most advantaged. Trust in patients Trust in patients was perceived to be important in the therapeutic partnership. Some GPs perceived it as an innate responsibility. “[…]as GPs, we’re trained to believe the patient and, and try to believe the patient as much as possible.” (GP22) “Look, I’m just really fascinated that there’s doctors that don’t trust that patients [laughter]. That’s horrible, like, I don’t understand. That’s just not the way I approach it at all.” (GP6) For these GPs, their trust continued regardless of patient behaviour. “When they say, ‘Oh no. They didn’t tell me anything about doing a blood test to check my INR before the scope’, I don’t think they’re lying to me. I think they forgot.” (GP4) “I said, ‘You need to stop drinking, you’re killing your liver’, that sort of stuff. But I told him, ‘Look, I want you to come back and see me, even if you’re drunk, okay, or you’ve been drinking, because I’m still going to be here for you.’” (GP1) Others described the therapeutic partnership as beginning with an implied trust which was shaped over successive consultations, based on their interactions with the patient. The patient’s return for subsequent consultations represented a reciprocation of trust and provided opportunities for the GPs to learn about the patient and tailor management support for them. “The fact that I’ve been seeing them over the years, it means they must trust me enough, otherwise why would they come back and see me? And it means in their minds, I’m already doing what they think a doctor should be doing.” (GP13) “… there is an extra level of, maybe, warmth and care if I trust them more because, and this is because when I trust them, I think usually they trust me as well.” (GP21) However, some GPs reported a lower propensity to trust patients based on concerns around complaints or litigation. “[…] if a patient complains, or […] maybe the patient had launched a legal action, you know, so if they’ve been involved in that sort of thing. And I think it probably erodes the GP’s trust in their patients. Especially the initial, you know, preconceived expectation that you have when you meet people for the first time.” (GP22) Lower patient trustworthiness was associated with perceptions of higher workload. “If I didn’t trust the patient as much I’d be, sort of, having to do a lot more work on the side […] to see, you know, what’s happening? So, yeah, I think it negatively impacts the dynamic because it just takes up more of my time.” (GP3) Patient characteristics The levels of trust in the patient varied, with GPs making assessments based on the patient’s engagement in the therapeutic partnership. Patient honesty, reliability, and proactivity in self-care were associated with trust. The patient’s honest reporting allowed GPs to assess the situation authentically and work with the patient to negotiate a management plan that is tailored to their personal circumstances. “[…] just being honest with each other, I think is the […] main thing that builds trust between physicians and patients.” (GP3) Perceptions of reliability were identified by GPs when patients completed ongoing self-management activities, such as further investigations, as agreed. “[…] they will do what they say they will do, and I trust that they will do it. If they say they don’t want to do it and they don’t, at least they’re consistent and I trust that too.” (GP21) Some GPs also noted higher trust in proactive patients who initiated self-care behaviours, which “[…] also increased my trust in them because I felt like […] this person […] has taken self-responsibility of their own health.” (GP21) While engagement in the therapeutic partnership was seen as important in building trust in the patient, it was not deemed as essential for effective care as the GPs were able to introduce additional supports, such as family members or carers into the therapeutic partnership and SMS reminder systems for appointments. “Sometimes some of my colorectal cancer patients have been quite elderly […] they’ve had family members step in, like the daughter or the […] carer, or the wife. And they will take a bit of ownership and help, but […] in my view, that’s part of the patient support network. That’s part of the patient’s trust.” (GP19) The impact of trust in the patient on engagement with colorectal cancer shared care Trust in the patient was described as important in shared care due to its collaborative nature. “I think that shared care is much easier when there’s a high level of trust in the patient. So, where patients really most basically engage the process, and they trust – there’s actually a real trust, then shared care is easier to facilitate.” (GP19) Most GPs had some experience with shared care arrangements, which were perceived to be associated with higher physician workloads and thus increased medico-legal risk, including those pertaining to professional liability (i.e. who would be responsible in a negligence case) and duty of care.28 “So I try and set longer appointments for anyone who has any kind of shared care […] so that you have enough time to communicate with the team, either via like a conference call, or a conversation with the patient and sending emails and letters.” (GP11) “[…] if you didn’t trust someone to either be going to their specialist appointments, that would be really hard because, yeah, the hospital might tell you but […] We’ve always got a degree of responsibility […]” (GP14) However, the GPs were more willing to engage in colorectal cancer shared care with longer term patients with whom they trusted (i.e. patients they had already established a working alliance with). “[…] we’ve got that background of relationship, that longitudinal relationship. […] it’s that ongoing therapeutic alliance, that ongoing relationship over time that allows you to go pretty quickly to getting down to what’s important for them. [...] it is a conversation, it is about understanding where that person is coming from […]” (GP6) “[…] once they’re onside and they realise that, you know, you value their time and they value your time, you can start encouraging or, you know, just really, you know, reminding them and saying, ‘Look, please do try and show up to your appointment on time.’” (GP11) A few GPs spoke of being unwilling to engage in colorectal cancer shared care arrangements for patients deemed less trustworthy, after considering the many possible consequences. “[…] it’s just a higher risk situation, because, you know, like, […] if there’s going to be any issues in follow up with the cancer service, then I don’t know […] how good the patient’s going to be in – or proactive the patient’s going to be in trying to arrange an appointment with them.” (GP22) Other reported barriers to the routine implementation of colorectal cancer shared care included poorly integrated communication systems between the primary and tertiary cancer services, the lack of role definitions, perceived higher workloads, and low renumeration. Discussion Our study provides important insights into the decision-making processes regarding GP trust in patients within therapeutic partnerships, and related impact on GP engagement with new models of care, such as colorectal cancer shared care. It was challenging for some GPs to discuss their trust in patients. Trust in patients was perceived to be important in the therapeutic partnership; however, there was variation among GPs in how trust was conceptualized. Some GPs viewed trust as an innate responsibility which stood independent to patient behaviour, while others reported an implied trust which was shaped by subsequent consultations. Honesty, described as “truthfulness, sincerity, or frankness,”29 was an important behaviour for assessing trustworthiness which was valued by the physicians, the patients30 and society generally.31 Similar to our previous work, there were higher levels of trust among GPs for patients who have demonstrated such reliability and proactivity over time.7 However, mutual trust is built within the wider ecology of the healthcare system,32–34 including other medical specialists, with an “implied trust across the whole network.”7 Like Skirbekk and colleagues,35 trust was not described as being discussed explicitly during GP consultations. Rather, increased patient trustworthiness was reported to be based on positive experiences with patient self-care, with each subsequent consultation within the therapeutic partnership symbolizing a reciprocation of trust. However, an explicit discussion of trust provides a strong foundation for negotiating ongoing management plans,35 and may be more apparent in long-term therapeutic relationships where expectations have already been established over time. Primary care has important responsibilities within the therapeutic partnership in the prevention and management of chronic and complex illnesses.3 This includes supporting patient self-management and multi-disciplinary team care coordination, which is essential to high-quality chronic healthcare. However, it also poses potential risks to the physicians who rely on the patients to describe their conditions, participate in shared decision-making and engage with self-care management. Previously, lung cancer patients were shown to have poor comprehension of their situation; yet, this was not recognized by their physicians.36 In our study, colorectal cancer shared care, due to its highly collaborative nature, was perceived to be associated with higher physician workloads and medico-legal risks. GPs were more willing to engage in colorectal cancer shared care with longer-term patients whom they trusted; with confidence gained through prior in-depth knowledge of the patient and their health behaviours. Other GPs were unwilling to engage in new care models, such as colorectal cancer shared care arrangements, with patients with whom they did not have an ongoing relationship and therefore did not trust. To facilitate the implementation of cancer shared care, a recent systematic review recommended more efficient processes, better role definition between health service providers, improved patient supports and engagement, and compensation for GPs.37 Strengths and limitations Few studies have investigated the construct of physician trust in their patients. To our knowledge, this is the first empirical study to investigate GPs’ trust in their patients in Australia. While there are low levels of research participation among GPs,38 we were able to recruit a diverse sample of GP participants, with input from GPs working in different socio-economic areas. Qualitative studies do not require sample sizes estimations for sufficient power, instead, data saturation is the gold standard for information redundancy.39 Our findings are not intended to represent all Australian GPs but are novel findings pertinent to understanding trust in patients. This study should be interpreted in the context of its potential limitations. The study was undertaken during the COVID-19 pandemic in 2021. During this period, there was rapid implementation of social distancing, strict infection control measures, and telehealth consultations by general practices.40 It is not known how the COVID-19 pandemic affected physicians’ trust in patients. Previous studies have reported increased general practice workloads and burnout and lower wellbeing, which may have influenced the GPs’ willingness to engage in newer care models.41,42 Patient characteristics, such as age, gender, occupation, or ethnicity, may be associated with the patient’s engagement within the therapeutic partnership; several GPs discussed general experiences with patients of different ages and socioeconomic contexts but stated this was a complex area. Further discussion of intrinsic patient characteristics were beyond the scope of our study, which was focused on the concept of how general practitioners trust and develop therapeutic partnerships with patients in shared care. Implications for clinical practice Colorectal cancer shared care arrangements require patients and health care providers in primary care and other specialist services to work collaboratively. GPs’ trust in patients impacts their willingness to engage in colorectal cancer shared care. Some GPs have a high propensity for trusting patients; for others, trust in patients is developed over successive consultations. To assist with implementation, future cancer shared care programs should consider the relational dynamics between the patient and health care providers, including trust. Supplementary material Supplementary material is available at Family Practice online. Acknowledgements We would like to thank the general practitioners for their participation in this research. Contributor Information Faith R Yong, School of Population Health, Faculty of Medicine and Health, University of New South Wales, Sydney, Australia; Faculty of Medicine, University of Queensland, Brisbane, Australia; Safe and Effective Medication Research Collaborative, School of Pharmacy, Faculty of Health and Behavioural Science, University of Queensland, Brisbane, Australia; Westmead Institute of Medical Research, University of Sydney, Westmead, Australia. Sundresan Naicker, Australian Centre for Health Services Innovation, School of Public Health and Social Work, Faculty of Health, Queensland University of Technology, Brisbane, Australia. Kerry Uebel, School of Population Health, Faculty of Medicine and Health, University of New South Wales, Sydney, Australia. Maria Agaliotis, Australian Institute of Health Service Management, College of Business and Economics, University of Tasmania, Rozelle, Australia. Christopher Chan, School of Population Health, Faculty of Medicine and Health, University of New South Wales, Sydney, Australia. John D T Nguyen, School of Population Health, Faculty of Medicine and Health, University of New South Wales, Sydney, Australia. Thanya Pathirana, School of Medicine and Dentistry, Griffith University, Southport, Australia. Alexandra Hawkey, Translational Health Research Institute, Western Sydney University, Campbelltown, Australia. Kylie Vuong, School of Population Health, Faculty of Medicine and Health, University of New South Wales, Sydney, Australia; School of Medicine and Dentistry, Griffith University, Southport, Australia. Author contributions KV provided input and oversight into all aspects of the study. FY led the data collection and analyses. FY and KV led the writing of the manuscript. All other authors contributed to the study design and ongoing project management as detailed in the following. Conceptualization: KU, MA and KV. Methodology: FY, KU, MA, AH, SN and KV. Data collection: FY, MA, SN and KV. Data verification: FY, TP, CC and JN. Data analysis: FY, TP, CC, AH and SN. Writing (original draft): FY, KU, MA, CC, SN and KV. Writing (revision): all. Supervision: FY, SN and KV. Project administration: FY, KU and KV. Funding acquisition: KU and KV. Funding The researchers gratefully acknowledge the Royal Australian College of General Practitioners (RACGP) Foundation and the UNSW School of Population Health for their support of this project. CC received a GP Synergy medical student scholarship to support his involvement. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Ethical approval Ethical approval was obtained from the University of New South Wales Human Research Ethics Committee (HC210424). Conflict of interest statement The authors declare no potential conflicts of interest. KV and KU are practicing GPs with research experience in cancer prevention and survivorship. JN is a GP trainee. CC is a medical student. TP is a medical educationist. FY is a pharmacist researcher in health systems with experience in qualitative and participatory action research in pharmacy, hospital and GP settings. MA is a mixed methods researcher with a background in clinical nursing, social science and health management. AH is a qualitative researcher with expertise in utilizing social science and participatory research approaches to address public health concerns, particularly in disadvantaged and migrant communities. SN is a public health implementation scientist. Data availability The data underlying this article cannot be shared publicly for the privacy of individuals that participated in the study. The data will be shared on reasonable request to the corresponding author. References Associated Data Supplementary Materials Data Availability Statement The data underlying this article cannot be shared publicly for the privacy of individuals that participated in the study. The data will be shared on reasonable request to the corresponding author.