Mistrustful Dependency: Mistrust as Risk Management in an Italian Emergency Department

ABSTRACT Mistrust is increasingly a daily reality of healthcare delivery worldwide. Yet it remains understudied as a form of relationship and a force in its own right. I address this gap through the ethnography of an Italian Emergency Department (ED), where conflicts have increased since the 2008 financial crisis. I show how mistrust does not result in a breakdown of healthcare interactions. Rather, mistrust is used in ambivalent care relationships to negotiate the roles, the risks, and the power that patients and staff are willing to entrust to others. Mistrust manifests in risk management strategies within relationships of “mistrustful dependency.”


Context: A place of friction
The Emergency Department in Italy, where I carried out the ethnographic research for this article, offers a unique observational vantage point through which the productive effects of mistrust can be accounted for. Since 2010, the Italian National Health Care Service has suffered crippling budget cuts of about 37 billion Euros (about 45 billion USD), leaving Italian healthcare expenditures at 6.2% of the GDP, dangerously below OECD countries' average rate of 9.5% (Maciocco 2019). The cuts were made as a consequence of the 2008 economic crisis and neoliberal policies that focused on cutting costs. They were broadly, even haphazardly, applied, and they resulted in a severe decrease in the number of hospital beds, nurses, and medical specialists. Budget cuts hit Italian EDs particularly hard, while a growing number of people who cannot afford to turn to private health practice increasingly relies on the ED -a public healthcare service that is always open, and available to everybody (Pasquini 2022;under contract).
Easily accessible, the ED is also a healthcare service that individuals find convenient. It is often preferred to General Practitioners, whose limited office hours (around 10 to 20 hours per week in Italy) are incompatible with long working days (Johannessen 2018;Pasquini 2022). A wide variety of media alarmingly report that one in every three people in Italy visits the ED at least once a year. Though, the Italian Association of Emergency Medicine (SIMEU 2016), affirms that health care staff in Italian EDs assigns to the 70% of all those patients either low-priority or non-urgent care codes.
Overall, patients find it increasingly difficult to access public health care services in Italy. Waiting times have become epically lengthy since the 2008 economic crisis. The pandemic furthered this trend where perfunctory medical examinations became the norm. People who could afford to do so began going to private practices. Growing short of available alternatives, a rising number of people in Italy are resorting to the ED as a kind of safe house for receiving medical attention. People frequently return to this safe house -some individuals go back to the ED up to 60 times a year. At the same time, the Italian population has been aging and socio-economic inequalities have been rising (Comodo and Maciocco 2011). Requests for health care skyrocketed due to the growing presence of chronic illnesses and the absence of a systematic welfare response (Muehlebach 2012). In the midst of such upheaval, widespread malcontent toward public health care initiated a nationwide torrent of lawsuits that targeted hospital staff (2018 MEDMAL report, Marsh research agency).
Mistrust took many shapes in a context marked by the uncertainties of understaffing and overcrowding like the Italian ED where I did my fieldwork. For instance, sensationalistic media reporting on malpractice cases invited patients to be suspicious of medical staff. Many patients had heard about the reported cases of infections after surgery due to bandages and surgical tools being forgotten in patients' abdomens. In 2018, the Italian Institute for Monitoring Insurance Activities (Istituto Italiano di Vigilanza sulle Assicurazioni, IVASS) reported 17,000 medical errors in the public healthcare system (above the average rate of OECD countries). Unsurprisingly, the annual increase of documented medical errors (up 3% between 2017 and 2018) went hand in hand with the progressive underfinancing of the public health system.
The mistrust that health care practitioners experienced from patients was reciprocal. In the ED where I worked, practitioners routinely mistrusted many of the patients they saw. ED staff constantly suspected people of attempting to skip long waiting lists for specialist consultations or in-depth diagnostic examinations in hospital wards. Health care providers often questioned, openly or silently, the reasons why people had come to the ED. To evaluate those reasons, practitioners cross-referenced patients' accounts with the documents that they could access on the computer database. Health care practitioners also increasingly preferred to base their assessment of a patient's state on bodily signs rather than the patient's account of suffering. Health care providers got suspicious and even dismissive when patients reported suffering that was not accompanied by any obvious medical signs. Mistrust in the ED was mutual, and it significantly structured healthcare interactions.

Method
Fieldwork upon which this article is based was conducted between 2017 and 2018 in a large university hospital structure in northern Italy. The hospital is an important emergency hub of the Emilia Romagna region, an area known for its civic tradition and political activism (the city of Bologna is its regional capital). The ED serves a provincial area of around 700,000 inhabitants, and it receives between 40-110 patients per day. The staff, which works eight to twelve-hour shifts, comprises two to four physicians and five to six nurses (these variations depend on whether it is a night or daytime shift).
The scope of my fieldwork was to ethnographically explore, through a negotiated interactive observation, the way mistrust unfolded among patients and ED staff and what forms it took (cf. Brown 2012;Hull 2017;Mulla 2014;Wind 2008). I asked how mistrust practically impacted the way healthcare was delivered and the possibility of building meaningful care relationships. To find this out, I followed emergency nurses during their entire working shifts (eight to twelve hours). For a year, between 2017 and 2018, I spent five to six days per week working morning, afternoon, and night shifts.
My fieldwork in the ED was organized into two key ethnographic approaches. First, I shadowed emergency nurses and their in-patient procedures (meaning I followed them around with a notepad and a digital recorder). I observed, discussed, and asked for explanations, which most health professionals gladly gave me. To disambiguate my presence as much as possible, I always took the time to explain who I was and what I was doing: not least because while following nurses and doctors as a nonmedical researcher, I had to wear a lab coat for hygiene purposes.
Second, I spent time in the external waiting room with patients and caregivers. There, I discussed with them their suffering and experience with Italian healthcare more generally. I conducted 86 interviews inside and outside the ED venues, 21 of which were with nurses, 7 with physicians and 57 with patients. The topics discussed ranged from overcrowding, to trust and mistrust. The material I present here is derived by the same ethnographic work I have described in another paper in which I address the effects of violence against healthcare workers in the ED (Pasquini 2022; see also 2023).
During fieldwork, I deployed in-site drawing as an ethnographic method in order to map out the movements of people and technologies within the ward space. I was also allowed to audio record care conversations (94 recordings overall). The final corpus of data gathered about 2000 pages of interview transcripts and field notes. Data analysis was carried out with NVivo 12 through an open, axial and selective coding, drawing from grounded theory (Bryant et al. 2010). I detailed 90 complete case studies concerning people's visits to the ED. I followed each one of them throughout both nurses' and doctors' examinations. Analyzing the interactions that unfolded within cases, I identified recurrent conversational patterns between patients and the ED staff (Sidnell and Stivers 2014). These methods enabled a thorough description of the developing frictions within ED daily practice.

The daily encounter with mistrust
Mistrust became particularly visible when dealing with situations of great uncertainty. During a busy afternoon in the ED, when crowding in the external waiting room was intense, signora Emma, 1 a Woman of about 80, arrived in an ambulance. After conducting a brief evaluation of her condition upon arrival, nurse Luciano placed the elderly lady on a stretcher. I was shadowing him during triage, the assessment that nurses first, and doctors later, make of people's suffering when they come to the ED. During triage, nurses assign priority according to four color codes (from the least urgent to the most: white, green, yellow, and red). Codes have specific consequences for patients so-labeled: consequences that are both temporal (less urgent codes mean longer waits), and financial (less urgent codes require the patient to pay more as a contact-fee).
Signora Emma was wrapped in a white sheet with her bare legs poking out, revealing intricate webs of swollen veins. Both rails of the stretcher were pulled up to prevent her from falling off. Signora Emma was contracting her left arm and lower lip unnaturally. The left side of her body appeared to be sliding downward. The ambulance crew who delivered her to nurse Luciano reported that she had probably fallen out of her bed at the nursing home where she lived. The nursing home staff claimed they did not know how it had happened: a not improbable case of negligence given the severe understaffing that nursing homes often endure. Signora Emma might have suffered a concussion, but the doctor who worked at the nursing home was not sure. Doctors who work in nursing homes are often very young -nursing homes tend to be doctors' first temporary assignments. In the view of many health care providers at the hospital, this meant that they were too inexperienced to be trusted.
Nurse Luciano called signora Emma's name, touching her gently on the shoulder. She remained silent, totally absorbed in what looked like pain but could also have been the distraction of dementia. Her eyes were not reddish, which meant that no evident sign of concussion was detectable. Was she just an unresponsive elderly woman with dementia?
"Wonderful!" said nurse Luciano sarcastically to me after we had moved away from signora Emma. "If we don't know how she looked before the fall, how can we know what has changed?" The nursing home staff did not know exactly, and they did not have any prior CAT scans to use as a reference for comparison. The ambulance crew reported on signora Emma's past history of neurological issues (she had had a minor stroke a year previously) and on her use of benzodiazepines to treat depression. Signora Emma had neurological difficulties of all sorts: she could not speak properly, her mouth was twisted, her left arm was held in an unnatural position. But none of these difficulties could conclusively determine an emerging acute state of either stroke or concussion. All the signs nurse Luciano observed could have resulted from the previous year's minor stroke.
Since there clearly was no way of knowing what signora Emma looked like before her presumed fall, nurse Luciano decided to wait for her relatives to arrive, in the hope that they might provide a clue. In the meantime, he assigned a yellow code (a high priority code) to her, so as not to take any chances in case she was suffering from a concussion or a stroke. She needed a priority CAT scan.
What nurse Luciano really feared was that signora Emma had been sent to the hospital to die. It was far from unknown for elderly-care institutions to attempt to avoid medico-legal troubles by sending critically ill or aging people to the ED as emergency cases, even though they were suffering from chronic conditions that could not be treated there. This was another reason why nurse Luciano and his colleagues in the ED tended to mistrust reports coming from nursing homes. The fall could have been a mere excuse to admit a dying old lady.
After two hours, signora Emma's family appeared at the reception. It had been a month since they had last seen her, so they were unsure as to whether she looked any different. Nurse Luciano concluded that they could not be trusted to help him decide whether the old woman had had a stroke or a concussion.
Soon after he spoke to signora Emma's relatives, the CAT scan revealed no concussion. At that point, doctor Roberto, a young but experienced practitioner, had to decide whether to admit signora Emma to the hospital or discharge her back to her nursing home. It was not possible to improve her medical situation. On the other hand, in his conversations with signora Emma's relatives, doctor Roberto later told me that he had got the impression that they did not trust him and would protest in the event that he refused to admit her to the hospital. Doctor Roberto decided to avoid the problem by finding a bed for the elderly woman to die in, in the solitude of a hospital ward.

Mistrustful dependency: Not a breakdown
Signora Emma's example allows us to elaborate over the role of trust and mistrust in healthcare interactions. As political theorist Russell Hardin phrased it, when we trust another person, we expect the other not to have any reason to betray our interests (Hardin 2002). In the extensive literature on trust, diverse kinds of trust can be sketched out as broad-brush stroke generalizations (Carey 2017;Cook 2003;Corsín 2011;Debaise and Stengers 2022). We may trust others (1) on the basis of our functional relationship with them; that is because of institutional roles (e.g. as patients, as nurses, and as doctors, Birkhäuer et al. 2017;Cook 2003;Gille et al. 2015;Kramer and Cook 2007;Shore 2006); or our contextual interaction with them (e.g. the way a person behaves, looks, or displays competency (Bruun et al. 2020;Coates 2019;Gambetta 1990;Kleinman and Van der Geest 2009). We may also trust others (2) out of a sheer rational calculation of advantage, i.e. our interest is convenient for others as well (Hardin 2002(Hardin , 2009. Last, our trust of others may remain simply unquestioned: (3) a basic trust that we do not even think about, like for instance when we ask a stranger for directions or we download a file to our computer from an unknown source (Debaise and Stengers 2022;Lagerspetz et al. 2015;Seligman 2000).
Being neither exhaustive nor exclusive, these understandings of trust as (1) relational, given by social relations and positionality, as (2) a rational assessment of interests, or as (3) the basic groundwork for interactions, are three major threads of enquiry in the immense scholarship on trust (Gilson 2003;Hardin 2009).
A relational approach to trust is best captured in practice by the work on care by the philosopher Annmarie Mol, sociologist Ingunn Mol et al. (2010). To them, care is an imperfect attempt, constantly being tinkered with. People and health care providers persist in a constant effort to do better, working toward an ideal of good care, and always having each other's interests in mind. As noticed by anthropologist Catherine Trundle (2020), such understanding of trust amid care, though, cannot account for the power asymmetries existing within healthcare relationships.
Ambivalent by nature, trust implies the risk of making oneself vulnerable to others' intentions, which may work against our expectations. Such ambivalence is exemplified by the cautious, mistrustful attitude of all the actors involved in signora Emma's care. The nursing home staff entrusted end of life palliative care to the ED to avoid lawsuits. Nurse Luciano decided he could not trust any of the parties involved in her care. Her family immediately made it clear to doctor Roberto that they would not accept her dismissal from the hospital. Finally, doctor Roberto who, to avoid a potential fight with Emma's relatives, admitted her to hospital. All of them try actively to delimit others' power, by avoiding taking risks which would expose them to others' intentions and potential betrayal.
To elaborate on such ambivalence, recent anthropology works have detailed the diverse intersections of trust, power, and risk (Cook and Trundle 2020;Lukšaitė 2022;Palmberger 2019;Rubaii 2020). A relevant example is anthropologist Sylvie Fainzang's account of patients and doctors lying to each other in primary care in France. She analyzes lying as a way for patients to deal with power in an asymmetric relationship (Fainzang 2015). Whereas for doctors, lying is a way to exercise power in such a context (Van Dongen and Fainzang 2005;Fainzang 2015; see also Van Dongen 2002. Another instance is the work of anthropologist Cristiana Giordano. Analyzing mental health interventions with migrant women in the northern Italy city of Turin, Giordano describes how women have no choice but to undergo medical and psychiatric evaluation in order to be "recognized" with a particular diagnostic categorization within the healthcare system (2014). But this process is not without conflict. Indeed, the women that Giordano worked with -like the people resorting to the ED in the absence of alternatives (Pasquini under contract) -attempt to influence care decision-making with the means at their disposal.
Following philosopher Harald Grimen, I suggest that the node of trust (a), the risk of betrayal entailed within it (b), and power (c), as intrinsic in asymmetric relationships of care, stand at the very basis of unsettled care relationships within healthcare systems (Grimen 2009).
Keeping this node of trust, risk, and power at heart, I would modify it to include mistrust within the equation. When understood as the opposite of trust, mistrust is often conceived as the complete breakdown of relationships of trust (Baker 2020). Whereas trust relationships would foster collaboration and allow both patients and professionals to work -by taking calculated risks and asserting power over patients' bodies, or among colleagues to coordinate a cooperative endeavor -mistrust would be a paralyzing force that dissolves the social glue represented by trust (Gilson 2003;Brown 2008;Caronna 2011). Within much literature on healthcare systems, mistrust is seen as a destructive force and thus as the opposite of trust capable of fostering collaborative relations (Baker 2020;Birkhäuer et al. 2017;Gille et al. 2015;Shore 2006). Though this extreme perspective of mistrust might be true in some cases, most of the time mistrust in healthcare relations looks quite different.
As critical medical anthropologists studying power relations within healthcare systems have widely proven, asymmetric relations are central to healthcare (Closser et al. 2022;Dao and Nichter 2016;Giordano 2014;Manderson and Wahlberg 2020). Emma's case is a perfect example of how rare it is for both patients and professionals working in the healthcare system to do away with care relationships all together. Their dependency on reaching out for medical care, or on collaborating with colleagues in order to face complex issues is usually not negotiable.
For example, the shortage of resources and staff for managing end of life and palliative care in elderly-care institutions encourages staff working in these facilities to avoid potential legal charges of misconduct and mistreatment by trusting the ED facilities to admit people like signora Emma to die. But, in this case, mistrust does not produce a systemic breakdown in healthcare relationships; instead, it changes them. It reshapes the trust, the risks, and the arbitrary power people are willing to bestow on others.
Whereas betrayal in relationships of trust causes a breakdown (Farrell 2004), mistrust, on the other hand, does not develop into a full rupture between patients and professionals involved with care. Unable to completely overcome dependency, actors instead try to negotiate the very terms by which they depend on one another: that is the power that can be asserted in their relationship and the kind of risk participants in care are willing to undergo. Unable to completely avoid relying on others, actors develop risk management strategies. One such risk management strategy is doctor Roberto's decision not to argue for the discharge of signora Emma, to avoid potential legal confrontation. In this case, risk management manifests his mistrust toward signora Emma's family.
But, as the second ethnographic example will show, those very same risk management strategies can in turn become the basis for new-found relationships of trust. Rather than a destructive force, relations of "mistrustful dependency" illustrate a mutually constitutive relation between trust and mistrust. Mistrustful dependency can radically transform healthcare interactions.

Mistrustful dependency as a basis for trust
Signora Emma's example illustrates how mistrust manifests in risk management strategies in asymmetric power relations of "mistrustful dependencies." But mistrust in such a case is reductive of the power that can be asserted by professionals and ends up with a negative care outcome. This is not always the case. Relations of "mistrustful dependency" -where people cannot completely get away from ambivalent, asymmetric care relations -might instead covertly facilitate people's access to medical care.
The case of signora Diana, a woman in her early 70s, illustrates this point. This example is also representative of the many chronic patients seeking aid from the ED. During an unusually warm and sunny morning in mid-November, signora Diana was escorted by two ambulance volunteers to the inner waiting area. Dressed in a threadbare pink dress, and looking around searchingly with blue eyes nestled in a mat of wrinkles, signora Diana slowly limped her way up to the reception desk, holding her lower back with her left hand. She explained to me and nurse Patrizia -who I was shadowing that day -that she had fallen over on a bus the previous day.
She had gone home and taken some paracetamol but, she said, "Even if I took a thousand pills, non conta [it wouldn't make any difference] at all. The only thing che conta [that works, that makes a difference] is Contramal [a powerful opioid]. But my idiot General Practitioner took those away from me. What should I do? Should I be suffering all my life?" Signora Diana explained further that she had got abituata a (used to) Contramal because she suffered from vertebral fractures and spinal cord compression.
"I just want to stop suffering," she repeated. "You know, I am also diabetic!" Nurse Patrizia nodded and assessed signora Diana as a low urgency, green code. Upon receiving her code and a brochure that explained what it signified, signora Diana kept standing near the reception desk, anxiously peering at us. "I could have given her a white code [not an urgency]" nurse Patrizia whispered to me. "She is always here asking for painkillers!" What did we really know about signora Diana? Triage interactions ranged from between three to fifteen minutes, and were almost completely focused on evaluating a short-term temporal scenario (the emergent, the "here" and the "now"). The capacity of the ED to meet people's needs, to understand che conta, i.e. what matters, was subordinated to its capacity to make sense of people's states of immediate vulnerability.
A clinical line of reasoning and an approximate timeline can be drawn from the triage interview in order to portray signora Diana's state of vulnerability. The encounter produced six factors that nurse Patrizia illustrated to me as relevant to an evaluation of signora Diana's health needs.
First, vertebral fractures and spinal cord compression underscored signora Diana's history of suffering and her past use of painkillers.
Second, diabetes was another clinical risk factor. It indicated signora Diana's dependency on care and a long history of involvement with health services.
Third, signora Diana's dismissive comments about her General Practitioner (GP) suggested that she found it difficult to access what she considered to be appropriate care.
Fourth, her reported fall on the bus confirmed that she was an elderly woman who had difficulties balancing. This in itself clinically excluded the allocation of a white code. A trauma reported immediately after the fact was considered to be at least a green code, since there was a major risk of it worsening within the first 48 hours. Signora Diana's case therefore required that she be given at least a green code, without any payment due.
Fifth, having taken paracetamol suggested that signora Diana was a responsible person who did not just run to the ED on the slightest pretext.
And finally, sixth, signora Diana's limp and her nervous lingering around the reception desk indicated that she felt anxiety and was suffering.
All these factors helped shape a linear narrative of events that decided the urgency evaluation criteria. Clinically, nurse Patrizia did not consider it feasible to give signora Diana some light palliatives. As signora Diana was already used to powerful painkillers such as Contramal, the drugs available to the nurse for infusion (ketoprofen or paracetamol) were of no use. The ED staff would say that giving someone like signora Diana ketoprofen or paracetamol would have been "like giving her acqua fresca (fresh cold water)." The only way to alleviate Diana's pain was for her to see the doctor quite soon, since the strong opioids she apparently was used to were allowed only under medical prescription.
Because the ED was not very busy on the day signora Diana came, she ended up waiting only half an hour before the nurse invited her to see a doctor.
When she entered the office of doctor Carmen, the expert emergency physician on duty that afternoon, signora Diana sketched a different narrative from the one she thus far had revealed. She started repeating her story about falling while on the bus, but then she suddenly turned to her real concern: the Contramal. When doctor Carmen asked for further information, she explained that her psychiatrist (she had not mentioned psychiatric treatment previously) had advised her GP to cut off her strong painkillers because signora Diana was possibly genetically predisposed to dementia. The psychiatrist had informed her GP that opioids could trigger Diana's genetic predisposition. To this, signora Diana added that she had fibromyalgia and an abdominal hernia which her GP had not taken care of.
"So, while I am here, can I also ask you for an X-ray for my hernia?" she wondered. Doctor Carmen replied, "You know, we cannot use the ED for routine checkups. Let's start with an X-ray of your lower back, which is where you got hurt, then we will see." At that point, signora Diana left the office and went to the inner aisle to wait for the X-ray. Doctor Carmen turned to me.
"She is here because she is addicted to opioids," she said with a sigh. "Her GP lied to her about the genetic predisposition to scare her off with the painkillers' side effects. She probably takes them a lot. I asked doctor Carmen if she was going to prescribe any painkillers for her. Yes, because of the fall," doctor Carmen said, giving me a knowing look. "I will not leave her without them. But I will change the type of opioid, so that her GP will not complain.
While she was waiting for her X-ray, I interviewed Signora Diana in a secluded corner of the inner corridor. She explained to me what Contramal meant to her. She told me about her difficult childhood, where back pain was a constant concern. Then she suffered a major depression after the death of her first son, a few months after he was born. After her first episodes of anxiety and depression, she started drinking and taking Valium. Right after that, she was diagnosed with fibromyalgia and started taking Contramal. Around this time, her father died too. Then, she had vertebral fractures and spinal cord compression. Now, I am 144 cm tall. Before the compression, I was 158! I was so beautiful back then; I was only forty [years old]!
The only thing that could control her suffering was Contramal, signora Diana insisted. She admitted to having had dizziness and other side effects but, in the end, she said, "It is my life and I should decide. It is the only thing that works on me (che conta). Should I have to suffer my whole life? Is that right?" Signora Diana continued, saying, "I come here because I trust them [the ED]. So, every time I have a problem, I come here to solve it." Signora Diana disregarded the fact that the ED existed only to deal with medical emergencies. She trusted the ED to address her routine health issues. By personalizing her care regime, she considered the ED as taking the caregiving place of her GP, who she mistrusted. But even as she professed trust for the ED, signora Diana nonetheless clearly felt it necessary to omit some parts of her story to better fit the urgency criteria of the ED, and to use the doctors at the ED to bypass her GP and obtain a new prescription for the Contramal medication she so desired.
In signora Diana's story, trust and mistrust are in a mutually constitutive relationship. Disbelieving her GP, signora Diana turned to the ED, where she misled the nurse who conducted triage in order to be admitted. She seemed aware of the fact that her GP had a point in preventing her from overdosing on Contramal. As she herself admitted, she felt dizziness and other side effects of the painkiller. But she still wanted it, and so, when coming to the ED, signora Diana omitted the reason why her GP had stopped her prescription for painkillers, instead insisting that she was there because of a fall on the bus. Having had extensive experience of the ED, signora Diana created a narrative in line with urgency allocation criteria, by disclosing concordant signs and hiding discordant ones.
Mistrust structured signora Diana's foregrounding of her fall, triage nurse Patrizia's initial suspicion, and doctor Carmen's realization that signora Diana was indeed at the ED for the opioid Contramal.
An important point here, though, is that this scenario of mutual mistrust did not hinder the establishment of a shared ground. In this case, it actually enabled it. The lack of interprofessional communication (between for instance the GPs, nursing homes, the ER, and the specialists) and the absence of appropriate funding to ensure care continuity, create a context in which mistrust becomes necessary to navigate around such disabling blockages. Mistrust in signora Diana's example connects the pieces of a fragmented system in order for care to happen in a meaningful way. The relationship of mistrust dependency between the actors involved in signora Diana's care allowed the formation of an alternative care pathway to be followed.
By performing a clinically relevant display of vulnerability, signora Diana described her situation without directly asking for a new prescription of Contramal, which, if expressed explicitly at that point, would likely have been denied. This strategy set the scene for a silent understanding between signora Diana and doctor Carmen, concerning the doctor's responsibility to somehow address her patient's chronic suffering. Doctor Carmen immediately recognized this necessity, even though she was skeptical of signora Diana's narrative about the fall. Given signora Diana's desire to tailor her care, doctor Carmen also engaged in a double agenda: addressing signora Diana's narrative as though she believed it, while actually pursuing the secondary aim of mediating between signora Diana's longing for Contramal and doctor Carmen's medical diagnosis.
Did signora Diana really fall on the bus? Who knows? The timeline she recounted during triage evaluation vanished as soon as she was granted entrance to the doctor's office. She turned the ED into a place for routine checkups and requests. She affirmed her own idea of wellbeing, without openly challenging the health care providers' reference frame. Through a game of the said and not-said, ambivalence on both sides facilitated a meta-communication to flow between two very different ways of understanding vulnerability.
This does not mean that a shared agreement about what was urgent was reached. Indeed, the opposite is true. Within a space of duplicity, two competing understandings of what was urgent were able to coexist without being reduced by one another, or developing into a full-fledged conflict. In signora Diana's case, the relationship between a lay person's needs and expert knowledge was at once iterated and subverted. Doctor Carmen's capacity to decide over treatment priority remained intact, and signora Diana obtained her painkiller.
Within signora Diana's story, as in many others I witnessed in the ED, "mistrustful dependency" developed into a form of trust. Relationships of mistrustful dependency did not lead to a breakdown of care interactions. They rather connected the fragmented pieces of the healthcare service and developed a workplace ethic that avoided conflicts through risk management.

Conclusion: Mistrust dependency in ambivalent care relationships
Recent global public health literature on the pandemic maintains that widespread mistrust is a crisis, an exceptional event, a break within everyday life (Baker 2020;Fukuyama 2020). Mistrust is often reported to be initiated by patients and their families who do not trust healthcare authorities (e.g. Manderson and Levine 2020).
This paper opposes these two ideas. It does not contest the urgency of the current situation of mistrust within healthcare all over the world. Instead, it proposes to address mistrust within daily health care relations as a structural phenomenon, inbuilt within the very core of the asymmetric power relationships that make up healthcare systems.
Signora Emma's and signora Diana's stories provide us with practical instances of how conflictual unequal relationships during care do not prevent the formation of productive links within healthcare organizations. Rather, they represent the norm for many health care professionals and patients, dependent on one another while navigating fragmented healthcare systems.
This article thematizes such conflictual normality as relationships of "mistrustful dependency," where healthcare links do not fall apart because of conflicts, but tensions never fully settle and continue to matter within negotiations regarding which kind of discretionary power should be granted to others and which kind of risks patients or health care staff are willing to run within care relationships. Such negotiation of power and risk pinpoints a relationship of mistrustful dependency that manifests in risk management strategies amid ambivalent care relationships.
"Mistrustful dependency" expands our understanding of the node of power, risk and trust at the basis of contemporary healthcare relationships (Cook and Trundle 2020;Grimen 2009). Differing from how Carey conceives of mistrust as shaping personhood (Carey 2017), and how Mühlfried elaborates upon it in relation to uncertainty (Mühlfried 2019), "Mistrustful dependency" provides an understanding of mistrust that facilitates the analysis of power dynamics. It is a way to conceptualize ambivalence within and beyond clinical care; where power, risks, and trust are not fixed but rather renegotiated in context through silent strategies of risk management that pinpoint attitudes of mistrust.
Following anthropology's engagement with ambivalence, studying the practical effects of mistrust in care relationships contributes to providing a complex, grounded view of care in practice (Brown 2012;Cook and Trundle 2020;Giordano 2014;Livingston 2012;Mol et al. 2010;Mulla 2014;Street 2014;Varma 2020). As this paper shows, "mistrustful dependency" brings down essentializing oppositional distinctions between trust and mistrust in healthcare situations, underscoring their interdependence, and highlighting how mistrust manifests as risk management strategies in asymmetric care relationships.
The concept of "mistrustful dependency" in ambivalent care relationships invites further anthropological work examining mistrust as a productive force in situated contexts. This would demonstrate how mistrust possesses a specific dynamic that actively, if often imperceptibly, shapes healthcare at all levels: from the granular individual level of self-perception as someone in need of healthcare (or as someone providing it), through interactional exchanges between people, to institutional structures, and professional investments.
As anthropologist Saiba Varma explains in her work about a psychiatric clinic in the military occupied territory of Kashmir, when care practices are only considered for their capacity to build trust and do good, we risk losing their context-specific complexity; which means that we risk losing sight of the setbacks and hidden or explicit frictions of care situations (2020:11-13). Within an international healthcare landscape increasingly saturated with mistrust, the stakes of such loss are an impoverished understanding of what really happens when mistrust unfolds in practical care situations.

Note
1. The Italian language has three basic honorific forms that convey politeness and deference when addressing people: Signor (Mr. or Sir), Signora (Mrs. or Madam) and Signorina (Miss -anyone familiar with the patriarchal bias of Italian culture will be unsurprised to discover that a marriage-neutral honorific equivalent to the English "Ms." still hasn't made it into Italian). These honorifics are titles, and as such, they are capitalized in Italian where, in formal contexts of address, they are commonly abbreviated to Sig., Sig.ra and Sig.na, respectively. Because the abbreviations will be unfamiliar to most English-language readers, I have chosen not to use them in this article. Instead, I write out the words whenever I use them. But note that the words signor, signora, and signorina (in lower case) are also used as nominals that specify gender, age, and marital status. This is how I use them throughout this article.