Risk, trust and patients’ strategic choices of healthcare practitioners

Research on patients’ choice of healthcare practitioners has focussed on countries with regulated and controlled healthcare markets. In contrast, lowand middleincome countries have a pluralistic landscape where untrained, unqualified and unlicensed informal healthcare providers (IHPs) provide significant share of services. Using qualitative data from 58 interviews in an Indian village, this paper explores how patients choose between IHPs and qualified practitioners in the public and formal private sectors. The study found that patients’ choices were structurally constrained by accessibility and affordability of care and choosing a practitioner from any sector presented some risk. Negotiation and engagement with risks depended on perceived severity of the health condition and trust in practitioners. Patients had low institutional trust in public and formal private sectors, whereas IHPs operated outside any institutional framework. Consequently, people relied on relational or competence-derived interpersonal trust. Care was sought from formal private practitioners for severe issues due to high-competencebased interpersonal trust in them, whereas for other issues IHPs were preferred due to high relationship-based interpersonal trust. The research shows that patients develop a strategic approach to practitioner choice by using trust to negotiate risks, and crucially, in lowand middle-income countries IHPs bridge a gap by providing accessible and affordable care imbued with relational–interpersonal trust.

Healthcare decisions require patients to make choicesthey must choose when to seek care, where to seek care and, in most instances, how long to be under care. Current scientific understanding of patient choice, along with the narrower field of patients' choice of healthcare practitioners, is predominantly shaped by research conducted in rich, industrialised and western societies. To illustrate, all 118 studies on patients' choice of healthcare providers examined by Victoor et al. (2011) were conducted in the USA, Canada, Australia, the UK or other western European countries. Evidence from these countries suggests that healthcare choices of patients are shaped by a range of factors including cost, accessibility, waiting time, continuity of treatment, patient and provider demography, and availability of relevant information to make decisions. However, two issues must be noted with regard to the current understanding of patient choice.
First, patient choices are constrained and limited by a range of factors. Policy ideas emphasising the importance of choice often tend to assume that patients carefully consider all available choices and make 'rational' choices (Victoor et al., 2011). However, burgeoning evidence contradicts this assumption and suggests instead that patients seldom possess all pertinent information, unlimited resources (time, money, cognitive bandwidth, etc.), capacity to weigh their options or even have stable preferences when presented with choices (Hibbard et al. 1997, Kooreman and Prast 2010, Moser et al. 2010). An unqualified promotion of choice must also be questioned as evidence suggests that it is often not always valued highly by patients (Anell et al. 1997, Lambrew, 2005, and choice does not necessarily enhance patients' agency (Bryan et al. 2006).
Second, it is crucial to consider and theorise the issue of patient choice outside rich and developed economies. In developed industrialised economies, patients choose between healthcare providers that are qualified, trained, and accredited professionals working in either private clinics or the public health systems like the National Health Service (NHS) in the UK. While complementary and alternative medicine (CAMS) may be provided by non-accredited practitioners, the biomedical sector remains well-regulated (see, for example, NHS, 2020). In contrast, healthcare markets of low-and middle-income countries are messy and often less strictly regulated. These healthcare markets also present patients with an additional choice of seeking care in the third sector comprised of untrained, unlicensed and unqualified informal healthcare providers (IHPs). Past research studies have noted that IHPs provide a large share of health services in both urban and rural areas of many developing countries (Ahmad et al., 2011, Cousins, 2019, Crabbe et al., 1996, Diamond-Smith et al. 2016, Kruk et al. 2011, Memon, 2006, Peltzer and Pengpid 2015, Saha and Hossain 2017, Thac et al., 2016. Scholarly interest in this third sector has largely been limited to exploring the quality of the care provided by IHPs (Das et al. 2015, May et al. 2014, Pulla, 2016). Yet, there is a need to understand how the IHPs fit in the puzzle of patients' choice of practitioners in low-and middle-income countries.
A key starting point in understanding how patients choose practitioners in pluralistic healthcare markets of low-and middle-income countries is in exploring how patients frame the risks associated with seeking care in each of the three sectors. The next section outlines why patients' framing of risk and their trust-based negotiation of risk are appropriate starting points in exploring practitioner choices.

Risk, trust and practitioner choices
Risk is intertwined with living in modern societies (Beck, 1992), and when it comes to health care, patients' decisions are shaped by construal and negotiation of risks. The very decision to seek care involves a recognition of risk to well-being or survival. However, being under care is risky too, for example, due to the possibilities of incorrect diagnosis, unsuccessful treatment and adverse reactions to medicines. Such healthcare decisions involve deliberate, situation-driven engagement with risk, akin to what Tulloch and Lupton (2003) describe as voluntary risktaking. Indeed, when commencing any health treatment, patients are generally aware of the possibilities of harm associated with it (Zinn, 2015). Thus, in the context of this research, when patients choose a healthcare provider from public, formal private and informal sectors to manage their health, they also choose the risks associated with their choice. Therefore, a key route to understanding patients' choice of practitioners is in examining how they negotiate the risks associated with their choice. Understanding how people negotiate healthcare risks requires a sociocultural approach that takes into account the local knowledge, culture and the overall social context within which people make decisions (Douglas, 1992).
Within sociocultural approaches, trust is regarded as playing a crucial role in mitigating perceptions of risk and shaping people's engagement with uncertainty. Patient-practitioner trust is necessarily bidirectional (Petrocchi et al., 2019); however, this research requires examination of patients' trust in practitioners. Accordingly, trust is conceptualised as the manifestation of patients' belief that a practitioner can guide them towards desired health outcomes without deliberately harming them. Like all trustor-trustee relationships, the origin of patients' trust in practitioners rests in institutional and interpersonal bases (Birungi, 1998, Gilson, 2003, Pearson and Raeke 2000, Rowe and Calnan 2006. There are significant overlaps between institutional and interpersonal bases of trust and they seldom operate independent of each other; yet, there are important distinctions between them (Campos-Castillo et al., 2016, Zheng et al. 2017. Interpersonal trust emerges from personal relationship between trustees and trustors and develops through social interactions-encounters that Giddens (1990) calls 'facework'. With interpersonal trust, the reason for trusting is primarily relational. On the other hand, institutional trust stems from people's trust in the abstract systems that support institutions in modern societies. Giddens (1990) conceptualises it as sytems based trust that enables people to trust strangers. Arguably, complex modern societies are possible only because institutional trust mitigates the inherent risk of engaging with strangers. To illustrate, if one did not trust the quality of medical schools in a country (an institution supported by numerous systems), would one trust the competence of doctors (strangers) trained there? Institutional trust 'rubs-on' to agents that represent a trusted institution and lends them familiarity and individual-level trustworthiness. Any engagement with uncertainty involves trust of some kind (Luhmann, 2017), but its importance is amplified in the case in patient-practitioner relationships. Trustee-trustor relationships in health care are asymmetric due to practitioners' greater knowledge, power and control over the direction and outcome of the relationship and patients' markedly higher vulnerabilities (Lewis and Weigert 1985). Therefore, understanding how patients draw upon trust will provide crucial insights about how practitioner choices are made while negotiating risks inherent in the choice.
To conclude, this research considers patient choices and decisions to be embedded in sociocultural contexts, shaped by marketisation of healthcare, and neoliberal ideologies (Gabe et al. 2015). In addressing the neglected issue of patient choice in low-and middle-income countries with pluralistic healthcare markets, it examines how patients choose between public practitioners, formal private practitioners, and IHPs. In doing so, the paper focuses on how choices are shaped by patients' understanding of general risk associated with the three sectors and how trust shapes the engagement with the risks inherent in practitioner choices.

The context of practitioner choices in Bholi, India
The study was conducted in the village of Bholi in Bihar, India. Bholi (pseudonym) is a village of about 350 households, located approximately 150 kilometres from the state capital Patna, and 16 kilometers from the administrative centre of the district. An agrarian economy where most land was owned by few families, most of Bholi populace lived in deprivation. The nearest public health facilities were 6 kilometres away in the district town and charged a nominal consultation fee of 5 rupees (5 pence). In India, public health system promises the most affordable care which is either free or incurs a nominal cost to patients. However, as one of the lowest spenders on public health globally, India's public health institutions are ill-equipped and poorly resourced (Balarajan et al. 2011, Dreze and Sen 2013, GOI, 2001.
As a result, private sector health care has grown rapidly from providing 10 per cent of services at the time of Indian independence, to an astounding 80 per cent in the previous decade (GOI, 2016, Sengupta andNundy 2005). Not surprisingly, the district town closest to Bholi had over 400 registered private practitioners charging between Rs 300 and Rs 700 (£3 to £7) per consultation. Formal private sector is expensive, and patients mostly pay directly for services (Bhat and Jain 2006, Doorslaer et al., 2005, Gupta, 2007, Karan et al. 2017. In Bholi and elsewhere, out-of-pocket healthcare expenses are a common pathway into poverty and create a trap through debt and asset losses (Krishna, 2006, 2010, Marmot, 2002, Phipps, 2003. To combat this, a national health insurance up to Rs 30,000 (£300) in the form of Rashtriya Swasthya Bima Yojana (RSBY), and more recently, Ayushman Bharat Yojana, is available, but their effectiveness remains unclear (Gupta et al. 2017, Karan et al. 2017. In India and other developing countries, the space between affordable but unreliable public health system, and the more expensive formal private sector is occupied by IHPs (also called quacks, non-degree allopathic providers, unqualified private healthcare providers). Over 56 per cent of active healthcare workers in rural India are without formal qualifications or training (Rao et al. 2016) and equally high numbers are observed in other developing countries (Adeleke et al. 2019, Chandra and Bhattacharya 2019, Ecks and Basu 2009, Morgan et al. 2016, Naidu et al. 2003, Rao and Peters 2015, Rao et al. 2016, Wahed et al. 2010. These practitioners lack formal training and qualifications, tend to collect payments directly from users in cash or kind, and work outside the influence of regulatory bodies (Sudhinaraset et al. 2013). It has also been noted that without proper medical knowledge, they tend to overprescribe drugs and engage in risky and non-evidence-based procedures, and their intervention often delays patients' access to life-saving treatment (Gautham et al., 2014). Bholi had four IHPs who lived in the village itself and provided a wide range of services including first aid, antibiotic injections and even surgeries, as reported by people in the village.
Bholi provided an ideal site for this study because it had IHPs living in the community and its proximity to the district town also provided people with the option of visiting qualified and accredited practitioners in public and formal private sectors (cf. methodology used by May et al., 2014).

A qualitative approach to understanding risk and trust-based practitioner choices
The use of a qualitative approach reflects the study's commitment to stepping away from behavioural models of healthcare access (original iteration in Anderson, 1995) and developing a more contextualised sociocultural perspective on patients' choices. Behavioural models are extremely well utilised by scholars in developed countries (for a review, see Babitsch et al. 2012), but their suitability to pluralistic and complex healthcare markets of the developing world has been questioned (Adhikari et al. 2013, Brinda et al. 2015, Thind et al. 2008. A qualitative approach is also valuable in capturing the nuances that shape patients' choice of healthcare practitioners, especially as decisions are often based on partial and incomplete information, while dealing with ambiguity and operating under constraints (Damman et al. 2009, Fasolo et al. 2010, Haynes et al. 2003, Hibbard et al., 1997, Lubalin and Harris-Kojetin 1999, Moser et al., 2010. The study was funded and approved by the ethics committee of the London School of Economics and Political Science (LSE). Data were collected using individual (n = 19) and group (n = 8) interviews from a total of 58 participants. As the first author had been visiting the community for five weeks prior to formal data collection, participant recruitment was an organic processsome participants were approached for an interview, and others learnt about the study and contacted the researcher. No participant expressing an interest to participate was refused. Participants were not a priori categorised on their SES. Instead, a self-identification strategy was used. 41 participants self-identified as 'poor' and 17 did notwhile crude, the paper refers to the former as 'poor' and the latter as 'affluent'. Participants were not asked to provide reasons for their self-identification. Interviews were conducted in the native dialect of the region by the first author and covered themes such as recent healthcare episodes in participants' family, perspectives on healthcare providers in the three sectors, and overall experience of seeking health care. Interviews and consent were audio recorded as many participants were unable to read or write. Interviews incorporated elements of narrative interviewing insofar as participants were explicitly encouraged to share stories about their own life and lives of others. However, unlike a strict narrative technique, following Chase (2003), interruptions were made to seek clarification. Therefore, the method is best understood as semi-structured interviews where participants were invited to share stories of their experiences.
Data analysis followed principles of thematic analysis (Boyatzis, 1998). The coding strategy was inductive; yet, the research goals framed the process of analysis and naturally introduced a deductive aspect to the process. As DeSantis and Ugarriza (2000) note, themes did not 'spontaneously fall out or emerge' but were developed for their explanatory power. Identification of themes was guided by an iterative process of examining data considering the goals of the research, and while they are well-substantiated by the data, no criteria of repetition were used.

Where was care sought? A preliminary note
Bholi was a strongly collectivist culture, evident in its day to day life, practices and espoused codes of conduct (Chauhan, 2015). This reflected in the consultative nature of most decisionmaking issues faced by people and families, including those pertaining to healthcare needs. The decision of where people should seek care was almost always a collective decision. Whether going to the IHPs, formal private practitioner or government hospitals, people would talk to their friends, relations and neighbours, even during emergencies. This is strongly indicated in the data in statements such as 'I then asked Ramkripal: What should we do?', or 'Nanhe's mother heard about it and suggested going to see Dr [omitted] in town. She had a similar pain and the doctor was very good.' During both individual and group interviews, participants were asked to recall the last healthcare episode in their family and where care was sought. The specific health issues recalled by participants were not used in the analysis as very few participants provided a named diagnosis. Educated participants that sought care from qualified medical practitioners too tended to describe general symptoms or used broad terms such as ghao (wound). Similarly, treatments were generally described in broad terms as those involving goli (tablets), marham (ointment) or injection. Diagnostic tests too were described broadly as khoon ka jaanch (blood report) and photograph (X-ray/CT scan/ultrasound). As a preliminary analysis, the choices made by participants provide several insights. First, affluent people of Bholi did not visit a public practitioner even once. Second, monetary capacity was not associated with either avoidance or patronage of IHPsover half of both 'poor' and 'affluent' participants received IHPs care. Table 1 provides the details.
Thematically, data show that accessibility and affordability of care, and perceptions of quality play an important role in practitioner choices. While these formed the bedrock, practitioner choices were eventually shaped by trust-based negotiations and engagements with risk. As a result, the thematic areas are intertwined and influence each other and only collectively they build the overall picture of how practitioner choices were made.

Accessibility and affordability
Easy availability of care is a well-recognised determinant of healthcare utilisation (Bose andDutta 2015, Mahapatro, 2016). Peters et al. (2008) unpack accessibility in terms of geographic accessibility, financial accessibility, availability of care as and when needed, and acceptability of patients' health beliefs. The first three of these factors framed accessibility in Bholi. Compared to both public and formal private sector practitioners in the town, IHPs lived in the village and provided the most geographically accessible care. All four IHPs (locally called 'jhola chaap doctors'; literally, sling-bag doctors) took a biomedical approach (called allopathy locally) and were usually the first port of call for poor and affluent people alike. In terms of affordability, public practitioners provided the cheapest option. However, government care involved hidden costs such as travel and charges for diagnostic tests in private facilities. As Table 1 shows, while seven out of 17 affluent participants made private practitioners their first port of call, only one poor participant did the samethe high cost of private care presents a financial barrier for the poor. In comparison, the consultation fee of local IHPs was 25 rupees. The most organised IHP in the village maintained a stock of prescription drugs, antibiotics and injectable medicines eliminating the need to visit a pharmacy in town. IHPs also made care more affordable by keeping specific patients' monetary capacity in mind while developing treatment plans. Ethical and quality issues aside, such personalised approach gave poor patients a greater sense of being in control. Overall, both public and private practitioners had low geographic accessibility and additional latent costs. IHPs were more accessible both geographically and financially and were often the first port of call for poor as well as affluent people. Interestingly, of the 17 affluent participants, nine made IHPs their first port of call. However, when patients from Bholi did go to the district town for their health needs, both poor and affluent preferred relatively much more expensive private practitioners. This was linked to the perceptions of quality of care.

Perceptions of quality of care
Patients made assessments of the quality of care provided by practitioners in the three sectors on the basis of four issues: perceived medical competence, infrastructure and emergency care, invasiveness of treatment and friendliness of consultation.
In line with findings of Russell (2005), participants regarded both public health and formal private practitioners as skilled, knowledgeable and competent. There was a clear awareness of their proper training and qualifications. However, participants' perceptions of quality of public sector were undermined by poor infrastructure, understaffing, overcrowding, intermittent supply of medicines, ill-equipped emergency departments, long waiting periods, short consultations and apparent apathy amongst practitioners. Formal private practitioners were evaluated to be significantly better on these issues and were, therefore, preferred by both poor and affluent people whenever they decided to go to the town for consultation. IHPs were regarded as trying 'do their best', but simultaneously, there was an unequivocal understanding of their limited medical competence.
I took him to a quack in the village. He gave some injections and some medicines, and my child was all right. These are the things that kill poor people. We go to people who have no license to treat my buffalo.
(AM, poor male) If things get worse, they [government doctors] raise their hands and say, 'Take the patient to a private hospital'. So it is better to start treatment at a private clinic. They take money, but they treat you well. They have all the machines. (SW, poor female) Research on doctor-patient consultation has routinely evidenced that high-quality interactions are related to higher patient satisfaction with treatment (Ahmed andBates 2016, Ward, 2018). While people had familiarity and friendships with the local IHPs, establishing an interpersonal relationship with public health practitioners was difficult as patients often saw different doctors in separate visits. In terms of interpersonal experience of seeking care, formal private practitioners were between the two extremes. Working in a 'for-profit' model, they were perceived to invest more effort in building personal relationships with patients and being friendly. Participants' idea of 'good treatment' was incumbent on greater invasiveness, multiple diagnostic tests and lengthy drug prescriptions. As reported in previous studies (Patel et al., 2015), this research also found that formal private practitioners prescribed arguably unnecessary treatment plansfor example, in quote below RRA was put on month-long antibiotic treatments for diarrhoea. Due to the symmetry between participants' idea of good treatment and the approach adopted by formal private practitioners, quality of care in the sector was perceived to be high.
The [public health] doctor said that it was not a snake bite. [The doctor said] 'He has stepped on a nail or something. Take him home' We then went to a private hospital. There the doctor said that it was a scorpion bite. . .the boy was lucky to not have died on the way. He gave him many injections, put him on saline drips. I think he was given 4 or 5 bottles of saline over two days and the boy lived. At the government hospital, he certainly would have died.
(SD, poor female) In summary, people of Bholi regarded formal private practitioners to be the best option in terms of quality of care received and were clearly aware of the limited medical skills of their local IHPs. However, as demonstrated in Table 1, IHPs provide a large share of medical care to both poor and affluent people in the village. This requires an examination of patient frame and mitigate risks associated with their practitioner choices.

Negotiating risks and the dynamics of trust
The choice between public, formal private or informal practitioners required choosing between different risks. Data show that participants associated the choice of public practitioners with the risk unsatisfactory care, apathy and disinterest in patient welfare, lack of infrastructure and even outright refusal. High cost of care provided by formal private practitioners imbues this choice with the risk of financial drain, asset loss and debt. IHPs, on the other hand, presented the risk of incorrect and even dangerous treatment. Quotes presented below illustrate participants' framing of these risks.
We had some land. All the land we had was sold during the days of my father. People negotiated the risks of choosing practitioners by drawing upon their trust in making their choices. IHPs worked without an institutional framework and offered the possibility of interpersonal trust only. Practitioners in the other two sectors could, however, provide both interpersonal and institutional trust. Data show that participants had low institutional trust in both public and formal private sectors. People in the village lacked trust in government and its institutions (Author, 2016). Consequently, institutional trust in public health system was low, and as proximal agents of the institution, public practitioners themselves were not deemed trustworthy. Institutional trust in formal private sector was low due to a number of issues including practitioners receiving kickbacks from pharmacies and diagnostic centres. Most importantly, the for-profit nature of this sector eroded institutional trust as evident in several stories that patients narrated about getting defrauded of the insurance cover that RSBY insurance scheme provided to the poor.
The medicines are all fake. The government medicines never work. [. . .] The Government sends bogus medicines to the hospital just so that we remain fooled.
(KT, poor male) Government gives [RSBY] card worth 30 thousand but I think that money is robbed and never benefits the poor. If they get a simple fever and go to Dr [omitted] or one of the private hospitals participating in the scheme, [the hospital] puts the card in the computer and takes all 30 thousand. They rob all the money.
(NML, affluent male) Interpersonal trust draws upon dyadic social interactions and remains tied to specific people (Zheng et al., 2017). In doctor-patient relationships, interpersonal trust derives from perceived competence, caring, empathy, clear communication and honesty (Mayer et al. 1995, Poppo et al. 2016, Thom and Campbell 1997. Two sources of interpersonal trust with practitioners were evident in the data. The first source was based on medical competence of practitioners and was calculative-cognitive component of trust people place in expert individuals. In contrast, the second source of interpersonal trust was anchored in prior familiarity and deep continuing relationships with practitioners. It formed the basis of a more stable and thick interpersonal trust. Patients' interpersonal trust in practitioners in the three sectors can be understood in terms of these two bases.
Participants regarded public practitioners as corrupt, dishonest and with low empathy and respect for patients and patients develop little relational-interpersonal trust in them. Additionally, patients seldom had the opportunity to see the same public practitioners over a substantial period for relationships to develop. Interestingly, with regard to medical competence, public practitioners were considered competent. However, this did not foster patients' interpersonal trust as they doubted if public practitioners would use their competence sincerely in treating them.
[At] government hospital the doctors do not examine you properly. They do no tests. . .they will just ask you what the problem is and would write those medicines that have 'special' prices.
(AM, poor male) [Government doctors] sell all the good medicines and then the poor people go there are they get all the rejected and expired medicines.
(KT, poor male) In comparison, formal private practitioners were regarded as medically competent and also demonstrated friendliness during consultations. This facilitated greater interpersonal trust than public practitioners to develop from both bases and allowed patients to undertake the risk of financial drain and debt when a health issue was deemed serious or complicated. Nevertheless, as also noted by Smith (2005), patients were cognizant of the monetary motivation and transactional nature of consultations with these practitioners'their situational empathy was a transactional symbolic act with no deeper relational basis. Thus, perceived medical competence was the primary source of patients' interpersonal trust in formal private practitioners but their for-profit nature concurrently undermined the development of relational-interpersonal trust.
It doesn't matter that one has seen the [private] doctor 20 times before. If you do not have money for their fee, you won't be allowed to enter their premises. . .forget about meeting them. Because they don't treat us. . .we are customers. They treat the money that we pay.
(MP, poor female) In contrast, interpersonal trust in IHPs had its origin in the second source: familiarity and continuing relationships. IHPs were a part of village community life outside any care relationships and were well known and respected members of the community. Fostered by familiarity and relationships, patients held a strong belief that IHPs will not knowingly cause harm and inform patients if a problem was beyond their ability to treat. This belief underlines the importance of familiarity and relationship-driven interpersonal trust between patients and their healthcare providers. Following Jones (1996), it can be said that interpersonal trust in IHPs was driven by patients' conviction of moral competence instead of medical competence.
My father suddenly got ill during the wedding festivities. [. . .] We took him to [village IHP]. We know each other since we were toddlers. He gave my father 'hard' antibiotics thinking that it will keep him safe in the interim and then after the wedding he can go to the town. He did that in 'good-faith' but when we took him to the town doctor, he said that the antibiotics are not working as they should. Eventually he died. . .the doctors couldn't save him (RJT, affluent male).

Reciprocity and conditionalities of interpersonal trust in IHPs
Trust is dyadic and requires both parties to acknowledge trustworthiness of the other. In this regard, unlike public and private practitioners, the trust-relationship between patients and IHPs was reciprocal. Indeed, IHPs also showed trusting behaviour towards their patients, for example by providing care on the promise of future payments. But perhaps most importantly, IHPs function illegally and therefore must trust the community to not report them to the authorities even in cases where their intervention produced disastrous results. Such reciprocal trusting relationships were absent from patients' relationships with practitioners in the other two sectors, as highlighted by MP's quote above. The presence of vulnerabilities and risks on both sides made patients' relationships with IHPs more symmetrical.
But the jhola chaap who are as poor as we are. . .they never turn us away because we cannot pay them on the same day. And they are economically no better than we are.
(BMP, poor female) The interplay between trust and risk was dynamic and participants' relationship-based interpersonal trust in IHPs was not absolute. Linked to the perceptions of competence and quality discussed above, when a medical issue was evaluated to be 'severe' or 'complicated', the community preferred seeking formal private practitioners. The present research cannot reflect on what factors contributed to people reaching an assessment of health issue being severe or complicated. Yet, the attachment of this conditionality on their interpersonal trust in IHPs appears to be the key strategy used by both poor and affluent people in Bholiindeed, it is a strategy that allows them to judiciously expend resources that are valuable to them (time, money and effort) while also limiting and managing the risks.
There are jhola chaap doctors who can look after small problems. They take 25-30 rupees plus the cost of medicines. But we don't go to them if the problem is serious or long term.
(PP, poor male) Researcher: In town, where do you go. . .government hospital or any private doctor? SD (poor female): That depends on the case. If someone breaks their leg or arm, they go to [government] hospital because it is free and they do it quickly. In complicated illnesses. . .fever, gastric, or something very serious, we always go to private.
To conclude, patients associate different risks and draw upon different bases of trust with practitioners in the three sectors. Pervasively low institutional trust in both public and formal private sector, and competence and relational basis of interpersonal inform participants' engagement with risk. These nuanced formulations of trust allow people to manage risks associated with choosing a practitioner in a constrained and pluralistic healthcare market.

Discussion and future directions: Strategy and leaps of faith in practitioner choices
The goal of this research was to develop an understanding of how people choose between public practitioners, formal private practitioners, and IHPs that occupy the pluralistic healthcare markets of low-and middle-income countries. This research in rural India found that affordability, accessibility of care when needed and low institutional trust in public present barriers that prevent both poor and affluent rural patients in this research to consistently seek care from trained and qualified practitioners either in public or private sector. In this light, the current research presents a strong case for understanding IHPs as providing care that bridges these barriers. For affluent people, they bridge geographic barriers, and for the poor, financial barriers. Considering IHPs in the developing world as plugging obvious gaps in the healthcare landscape integrates evidence from a range of studies on the informal healthcare sector of developing countries (Ghosh, 2014, May et al., 2014, Raza et al. 2013.
A second important contribution of this research is in developing a socioculturally informed framing of patient choices of practitioners by examining them with the lens of risk and trust dynamics at play. Risk can never be eliminated completely but only managed in healthcare choices and decisionschoosing between practitioners from the three sectors required management of inherent risks. This research has shown that managing this risk involved people making a crude, case-by-case assessment of whether the problem was 'serious' or notwhen labelled 'serious', formal private sector was preferred. In making this choice, poor people voluntarily undertake the risk of debt, impoverishment and loss of assets but also reduced the risk of harm to health. When the condition was not deemed 'serious', health risks were perceived to be lower and both affluent and poor people voluntarily took strategic risksthe former for convenience of access and the latter for their reduced financial burden also. Thus, the use of IHPs by the Bholi populace wasneither necessarily nor alwaysout of poverty, lack of options, or ignorance. In each of these decisions, strategic risks were embraced and this paper posits that patients' choices of practitioners involve deliberate and strategic risk-taking. However, understanding a community's predominant reliance on IHPs purely in terms of strategic risk-taking would only reinforce the rationalistic assumptions that behavioural models purport. Choices in Bholi are also shaped by India's closeted neoliberal policies on health. Taxes on private healthcare generate revenue for governments and, arguably, reduce the incentive of improving welfare-based public health infrastructure.
Patients' choices and strategic risk-taking are based on fuzzy trust-based decision-making. Trust has both emotional and cognitive dimensions. Its emotional dimension draws on familiarity, similarity, and interpersonal relationships but the cognitive dimension is predominantly calculative in nature (Lewis and Weigert 1985;Poppo et al. 2016;Reich-Graefe, 2014). People drew on both calculative and emotional dimensions to develop different kinds of interpersonal trust in practitioners in an environment where institutional trust was either low (public and formal private sector) or impossible (IHPs). The research provides evidence that patients derive calculative interpersonal trust from perceived medical competence of practitionersthis was highest for formal private practitioners and led people to seek their care for 'serious' issues. On the other hand, interpersonal trust based on familiarity, relationships and moral competence of practitioners was emotional in nature -IHPs were most trusted in this regard. Crucially, this research shows that patients' high relational-interpersonal trust in IHPs can compensate for the lack of trust based on medical competence, qualifications and training. Rowe and Calnan (2006, p. 5) note that trust is 'forward looking and reflects a commitment to an ongoing relationship'. In much of rural India and other developing countries, people's ongoing relationships of medical care are usually with local IHPs and people build strong, interpersonal trusting relationships with them. Trust provides patients what M€ ollering (2001) considers 'good reasons' for making their choice of practitioner. Indeed, patients' trust in any practitioner was not blind and all-encompassing. Instead, the research has shown the presence of calculative and cognitive considerations that established conditionalities of trust. In essence, the research demonstrates that when people in the developing world choose to seek care from IHPs, they do not merely make 'leaps of faith' (Brownlie and Howson 2005, p. 235), but instead, they make strategic decisions and voluntarily embrace some uncertainties and risks.
Looking ahead, this research has underlined the need to examine the issue of risk and trust further and data have highlighted several inter-related issues that were not examined in this study. For example, what factors lead people to consider a health issue as 'severe' or 'complicated' and when does a health issue or condition cross the threshold for seeking care in expensive formal private sector? What individual-specific factorssocioeconomic status, patients' gender or age, pre-existing conditions, comorbidities, or indeed, their scientific literacycontribute to it? To conclude, patient choices in the complicated markets of developing countries are based on complex decisions. This research demonstrates that both poor and affluent patients make trust-based strategic decisions and actively engage with risks associated with practitioner choices in these pluralistic healthcare markets.

Data Availability
The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.