Abstract
Research on medical interactions shows how the discursive construction of the clinical case impacts diagnostic reasoning and treatment recommendations. Drawing on an ethnographic study in an intensive care unit, we illustrate how this process is at play in a ward that adopts an extreme, guideline-divergent policy as to the use of antibiotics. The article focuses on how physicians assemble the case as ‘treatable’ or ‘not yet treatable’, and how in doing so they ‘talk into being’ two contrastive policies on antibiotics and position themselves toward the one adopted in the ward. The analysis identifies the discursive resources displayed by physicians to both project an infectious disease diagnosis and resist this treatment-implicative trajectory. We argue that the physicians’ contentious discursive construction of the case has crucial consequences in the way the ward’s extreme policy is jointly accomplished as a highly reflexive process sensitive to the contingencies of any particular case.
Keywords
Introduction
Research on medical interactions shows how and to what extent the discursive construction of the clinical case impacts on diagnostic trajectories and treatment recommendations, two activities expected to be evidence-based (Timmermans and Angell, 2001) and dependent upon an ‘objective professional voice’ (Kovarsky et al., 2005: 119–180).
The construction of the ‘clinical object’ (Galatolo and Margutti, 2016; Heath, 2006) has been widely studied in doctor–patient interaction. This joint activity takes place in the initial phase of the visit and has a dramatic impact on the subsequent phases, namely diagnosis delivery and treatment prescriptions (see Maynard, 1992; Stivers, 2002b, 2007). Apart from studies in medical education – where an explicit learning process between experts and novices is at stake (Hindmarsh et al., 2014; Koschmann et al., 2011; Mondada, 2007, 2014; Rees and Monrouxe, 2008; Zemel and Koschmann, 2014) – the clinical case construction between physicians has been relatively underexplored. Few remarkable studies on doctor-doctor interaction show how physicians collectively construct the clinical case while reporting it to the team (Anspach, 1988; Atkinson, 1995, 1999; Cicourel, 1985; Ikeda and Okada, 2007), to other professionals (Måseide, 2007) or while taking notes (Hobbs, 2003). These studies converge in underlining that ‘presenting cases is not merely a way of depicting reality but a way of constructing it’ (Good, 1994: 80) that impacts the overall medical diagnosing and decision-making processes (Alby et al., 2015; Halvorsen, 2010; Halvorsen and Sarangi, 2015). According to Anspach (1988), reporting the patient’s condition is a discursive activity that not only incorporates the physician’s self-presentation, but also embodies the clinical stance toward the medical challenge at stake and its problematic options (see also Atkinson, 1999). Adding to research on the communicative constitution of the medical object (see also Alby et al., 2015; Antaki et al., 2005; Cicourel, 1987; Lymer et al., 2014), we consider the case presentation in doctor–doctor interaction as an activity where physicians construct the ‘assessable object’ (on the analytic concept of assessable, see Goodwin, 2003; Goodwin and Goodwin, 1987). As we will show, this activity is theory-implicative and praxis-constitutive: in crafting the features of the object they have to assess collectively, physicians index their relevant expert knowledge and channel the subsequent actions by projecting a clinical trajectory.
Drawing on an extensive ethnographic fieldwork in an intensive care unit (ICU), this article analyzes physicians’ construction of the assessable object as potentially being a case of infection. The analysis focuses on how physicians of the same specialty (a) assemble the clinical object as ‘treatable’ or ‘not yet treatable’ with antibiotics in the report phase of the team briefings, (b) ‘talk into being’ (Heritage, 1984: p. 290) different policies concerning antibiotic treatment and (c) position themselves toward them.
Participants’ discursive resources and territories of knowledge: The case of infectious diseases and antibiotic treatment
As Stivers’ (2005, 2007) outstanding research shows, the case of antibiotic prescription is particularly revealing of the kind of struggles that may occur during medical interaction. The growing evidence of the risks involved in over- or otherwise improper prescription of antibiotics dramatically changed the physicians’ clinical stance toward antibiotic prescription worldwide. Risks concern the selective pressure of antibiotics on the bacteria they are supposed to fight and the consequent increase of new multidrug-resistant bacteria. The spread of these germs – responsible for infectious diseases with high morbidity and mortality rates – is a public health concern all over the world (see Center for Disease Dynamics, Economics & Policy (CDDEP), 2015; World Health Organization, 2014). The gap between folk medical notions (e.g. antibiotics as the ‘miracle drug’ and the commonsensical belief ‘if infection then antibiotics’) and the updated medical expert knowledge partially explains why patients (or relatives) and physicians often (although mildly) conflict upon diagnoses and treatment prescriptions concerning conditions that might be identified as infections and therefore might be treated with antibiotics.
Drawing on Stivers’ (2002a, 2007) research on doctors’ and parents’ communicative resources to – respectively – project ‘no need for antibiotics’ and resist this discursive trajectory, our study focuses on doctor–doctor interactions when there is no ‘socially sanctioned authority to know’ (Heritage, 2012b: 5) concerning antibiotics, 1 and (a) physicians do not share the same diagnostic trajectory, (b) more than one policy on antibiotic treatment (ATBT) is locally available (Ten Have, 1995) and (c) they have to collectively decide in an accountable way whether to treat or not to treat yet. Surprisingly enough, our data show that some patterns of interaction typically identified in doctor–patient (or–parent) interaction are at stake even among hospital physicians with the same epistemic access to the relevant expert knowledge, common technical jargon and heuristics. We advance the hypothesis that the physicians’ discursive oscillation (see Cooren, 2010) between two locally available yet extremely different alternative policies toward antibiotics accounts for the way decisions are interactively accomplished.
Design and method of the study
The study is part of a larger quantitative/qualitative research project aimed at studying the nonclinical dimensions that possibly impact the spread of nosocomial infections. 2 The overall project involved 40 ICUs. This article reports data from the Central Italy ICU 3 (CEICU) chosen as an exemplar of those having a high rate of infectious diseases and a very low rate of multidrug-resistant bacteria infections. CEICU follows a policy toward infectious disease prevention and treatment that both team members and their colleagues working in other ICUs consider ‘off-label’. Data analyzed here consist of excerpts from a corpus of nine video-recorded daily team briefings. They have been analyzed using a conversation analysis approach (Jefferson, 2004; see Appendix). We focus on the discursive resources deployed by medical staff members during the case presentation to resist or pursue the ‘no problem/no treatment’ trajectory implied by the official policy of the ward. These resources make relevant alternative yet not equivalent courses of action: prescribing, or not prescribing antibiotics yet.
The institutional setting: Alternative policies for antibiotic treatment
Two policies concerning ATBT are part of the expert knowledge of CEICU’s members: ‘the empirical therapy for suspicion of infection’ and ‘the definite therapy for proven infections’ (see Eggimann and Pittet, 2001: p. 2074). The former is totally consistent with the guidelines and the policy generally adopted in most ICUs that apply ‘early empirical broad-spectrum antimicrobial coverage for critically ill patients in whom the development of a N[osocomial] I[nfection] is suspected’ (Eggimann and Pittet, 2001: p. 2074). This widely used and highly recommended approach consists in (a) collecting cultures before treatment, (b) treating the patients with first-line, broad-spectrum antibiotics at the first relevant symptoms of a suspected infectious disease (American Thoracic Society and Infectious Disease Society of America, 2005: p. 388) and (c) re-orienting and tailoring the ATBT as soon as test results are available.
The ‘definite therapy’ for proven infections consists in (a) avoiding the empirical therapy; (b) tolerating clues that may be but are not necessarily symptoms of infections; (c) capitalizing, as much as possible, on the patient’s endogenous resources; and (d) prescribing targeted ATBT only when laboratory exams detect the germ(s) responsible for the disease, the antibiogram reveals the germ’s sensibilities and resistances and the vital signs of the patient clearly reveal that she or he can no longer resist the infection. Both policies are accounted for by CEICU members as relying on extant literature and clinical studies that supposedly provide evidence as to their respective risks and benefits. While the empirical-therapy approach insists on the positive effects of the early, broad spectrum ATBT in reducing the seriousness of the patient’s first-order infections, the ‘watch and stay’ approach focuses on the positive effects of delaying the therapy and avoiding the use of large-spectrum antibiotics – preserving the ward ecology (and consequently the single patient) by preventing the occurrence of the rare but far more dangerous and even fatal infectious diseases caused by multi-resistant bacteria.
Within this typical clinical yet also social dilemma, the CEICU Responsible Clinician (RC) has adopted the definite therapy – amicably labeled by CEICU members the ‘watch and stay’ approach 4 – as the main therapeutic line of the ward. This approach is held accountable for an overall reduced use of antibiotics in the ward with respect to those adopting the empirical approach, and use of less expensive ATBs as well as a low rate of multidrug resistant (MDR) bacteria. A quantitative analysis of the decisions concerning patients who display symptoms of possible infection shows that 71% are made following the ‘watch and stay’ approach: antibiotics are not prescribed that day. The ‘watch and stay’ approach does not follow the guidelines widely accepted and strictly (defensively) followed by most ICUs worldwide (although it is coherent with some recent studies; see Hranjec et al., 2012). Doctors’ and nurses’ oscillating orientation toward these policies is traceable and analyzable in the ways they participate in the event institutionally provided for the purpose of taking diagnostic and treatment decisions: the morning briefing (on the theoretical necessity of implying meaning and culture to make sense of local interaction, see the Van Dijk–Duranti debate; Duranti, 2015).
Assembling the case: How physicians project treatability through reporting
CEICU morning briefings take place each morning from 8 a.m. to around 9 a.m. The overall scope of the event is to assess the patients’ status and plan the courses of action for the next 24 hours. As is the case for most institutional talk in health care settings (see Byrne and Long, 1976; Heath, 1992; Heritage and Maynard, 2006), CEICU morning briefings are sequentially organized: they consist of a number of macro sections corresponding to the number of inpatients. Each section consists of three main consecutive phases: the report (i.e. case presentation; Atkinson, 1999; Erickson, 1999), the assessment and the plan. Participants regularly attending the event are the RC in charge of the clinical line of the ward, the night physician (NP) who is in charge of reporting the inpatients’ status updated as of 8 a.m. daily, the case manager nurse (CM), the head nurse (HN), the physiatrist (PH), the physician responsible for the patients throughout the week (lunghista; L) and other physicians (AP) attending the morning shift.
Members of this team officially adopt a shared decision-making model of teamwork: Diagnosis and treatment decisions are said (during in-depth interviews and informal talk with researchers) to be collectively taken (i.e. as one member said, ‘We are a ward. One ward, one line’). The morning briefings are the institutionally provided locus for that. The frequent use of the ‘inclusive we’ (O’Grady et al., 2014: 76; for a broader use of the inclusive ‘we’ in medical talk, see also Brookes-Howell, 2006; Lindwall and Lymer, 2011) in their diagnostic and evaluative talk is one of the cues displaying their orientation toward this officially declared model of teamwork. 5
During the morning briefing, CEICU physicians have the problem of assessing the patient’s conditions as ‘not yet treatable’ (with ATB) or – if they wish to resist the official policy of the ward – as ‘treatable’ (for the interactive construction of doctorability and treatability, see Heritage and Robinson, 2006). The crafting of the assessable object is therefore crucial as it provides the foundation for both the assessment and the planning. It is in the report phase that this work begins. The following excerpts illustrate some discursive resources used in the report phase by participants to assemble the assessable object as ‘treatable’ or ‘not yet treatable’. 6 The aim of the analysis is to illustrate a typology of the discursive resources displayed by participants to both project an infectious disease diagnosis and resist this treatment-implicative trajectory (see Stivers, 2002a, 2007).
Partial assessments and mitigating no problem conditions
When physicians present a case in ways that are aligned 7 to the ‘watch and stay’ approach (no problem/no treatment today), they can do this in a mitigated way through partial assessment of specific data concerning the patient’s status.
(1) He is basically a-pyretic [CEICU_Im_9] White (bianchi): Elliptical expression for white blood cells. BAS: Elliptical expression for the value of the Bronchial Aspirated Secretions. MSSA: Methicillin-Sensitive-Staphylococcus Aureus. ICP: Intra-Cranial Pressure.
The NP, Gerlando S. (NP_G), provides a partial assessment of the patient’s pyretic status (‘basically a-pyretic’, line 3) even before giving the numerical information concerning his body temperature. He then continues reporting data about the white blood cells (no recent data are found), the results from lab tests that detected some (pneumonia inducing) bacteria in his bronchial aspirates (line 18) and his not being on ATBT (line 19). He then assesses radiological exam results: the x-ray is ‘basically negative’ (line 23) and reports information projecting the patient’s relatively good condition: he is awake (line 19) and ready to be progressively weaned from the mechanical ventilator (lines 24–26).
The case presentation is aligned with the policy of the ward: the simple presence of a bacterium – in the absence of other relevant conditions – is not enough to suspect an infectious disease or to prescribe antibiotics. However, the alignment is cautious: exploiting his epistemic right to assess clinical data, he mitigates two ‘no problem’ indexing conditions. In doing so, he blurs the diagnosis implied in the way he assembled the case (no-infection at stake): the patient is not a-pyretic, he is ‘basically a-pyretic’ (line 3); the chest x-ray is not negative, it is ‘basically negative’ (line 23).
This partial alignment provides NP with the opportunity to insert a comment in the planning phase (not transcribed) through which he will reopen the possibility that the infectious disease is already established: by stating that the spontaneous respiration can make it manifest, he implies that the infection is latent. Although the case presentation is consistent with the ‘no problem/no treatment’ trajectory and therefore aligned with the ‘watch and stay’ policy (no ATBT will be prescribed this day), NP manages to create some fissures that forewarn of a possible imminent problem.
Selecting and underlining relevant parameters
In the previous case, NP mitigates a (constructed as) not-so-problematic case. The next excerpt illustrates the minimal form through which physicians project an infectious disease: selecting and underlining information that could be relevant for a potential infectious disease diagnosis, or that at least may alert toward this possibility. This case construction orients the diagnostic reasoning toward treatability.
(2) Very very foul-smelling [CEICU_Mart_2]
The NP, Giovanni M. (NP_M), starts by typically providing information on blood temperature: the patient is hypo-pyretic (line 1). This status is not consistent with infection. However, he immediately reports a piece of information that is consistent with a possible infection: the quality of the secretions. The quality of secretions is routinely used to assess a potential pneumonia; here they are qualified as foul-smelling (line 2). The speaker self-repairs his previous utterance: he provides a new and more explicit version of the information in a more straightforward way. He then goes on by selecting from the patient’s history two pieces of information that may explain those secretions, and therefore he further points toward a suspected pneumonia diagnosis. First, the patient has been considered as having possibly inhaled liquids from the mouth because he has been found at home (unconscious) with some vomit (lines 3–4). This information from his clinical history is relevant: inhalation may cause endogenous contamination and lead to pneumonia due to the entry of bacteria from the throat into the respiratory district. Second, the patient is reported as not having been immediately intubated yet positioned with a laryngeal mask (line 6). The known and implied information here is that the mask does not protect the lungs from inhalation. All the premises for a possible pneumonia are there as well as a typical symptom: the malodorous secretions. The sequence is closed by a summary where the NP invites the audience to make a connection between this information from the patient’s history and his present bronchial secretions: the turn begins with the discourse marker ‘so’ (line 7) that resumes and ‘indicates to the hearer that some kind of inferential connection between the two propositions needs to be made’ (Bolden, 2009: 976). Right after this indication to make a connection, the secretions are referred to again and hyper-qualified: they are ‘very very foul-smelling’ (line 8). Through this instance of upgraded assessment, NP reiterates the information that projects a treatable condition and makes relevant a suspicious of infection.
‘Doing nothing’ with information 8
In Excerpt 2, RC manages to not provide relevance to or emphasize the NP’s first-hand information (Heritage, 2012a; Pomerantz, 1980): he does not exploit the pause (line 9) to acknowledge, comment upon, or ratify the marked information, adopting what we call the ‘doing nothing’ resource (for the sequential meaning of ‘doing nothing’, see Stivers, 2006: 288). Consequently, NP goes on reporting on other aspects of the patient’s status (acceptable Intra Cranial Pressure, lines 10–12).
Although NP has repeated the same information on the secretions three times, in the assessment phase (not transcribed) no reference will be made to this potential symptom. The projected possibility of pneumonia is suspended: not confirmed nor denied. This zone of indeterminacy is totally consistent with the ‘watch and stay’ approach (on postponing diagnoses, see Alby et al., 2015): that day, the patient presents only one possible cue of a hypothetical pneumonia. Although sufficient for the followers of empirical therapy, it is not enough for a ward committed to delaying the beginning of the ATBT as long as possible. We suggest that – thanks to the work of the NP to construct the assessable object as treatable – the possibility is now part of the ‘representational field’ (Heritage and Raymond, 2005) of the team. The day after, the patient will be assessed again. 9
Anticipating and confuting alternative candidate explanations
The following excerpt identifies a more complex case: NP gives more than qualified information to project an infectious disease problem; he narrows the diagnostic field by anticipating and confuting an alternative candidate explanation (on candidate diagnosis, see Stivers, 2007).
(3) It is not a withdrawal symptom [CEICU_Gug_6]
The NP, Fernando G. (NP_F), starts by giving the name and the bed number of the patient (line 1). He then provides the hospital day of this patient (lines 2–3): the patient is in his 24th day. The duration of the hospital stay is crucial: first, long duration is positively correlated with the increase of hospital-acquired infectious diseases, and second, it is a parameter used in this ward to evaluate the beginning of ATBT. Although patients in ICUs usually begin ATBT on the first/second day following admission, at CEICU, they start it on average on the sixth/seventh day. So NP’s discursive strategy here consists in selecting and underlining information that orients toward treatability: the patient is far beyond the CEICU average day for beginning ATBT.
In line 4, NP frames the incoming information as new. In doing so, he marks a difference with respect to what is already known about this patient and alerts the recipients to the relevance of the incoming information: the patient has been having pyretic spikes with tremors (lines 4–7). All participants know that ‘fever is one of the cardinal signs of infection’ (Young and Saxena, 2014: 1), yet they also know that it can be a consequence of other conditions.
The report of this objective status is followed by a first position assessment (lines 8–9). The assessment is designed in a negative format: this condition is not a withdrawal symptom. Drawing on his first-hand knowledge (Heritage, 2012a; Pomerantz, 1980) – the last spike occurred last night and last night he was there – he lists the patient’s concurrent symptoms (lines 10–13) that make him identify what this spike with tremors is not a symptom of. In doing so, he restricts the range of possible assessments and makes relevant a possible infection diagnosis.
However, the ways he designs his turn downgrades the assertiveness of his declaration and reduces the claimed compatibility of the condition ‘with the asserted state of affairs not being the case’ (Heritage and Raymond, 2005: 18). In delivering his claim about what the fever is not a symptom of, he suddenly produces a self-repair (the last one last night, it was not a-, line 7) that creates a slot to introduce an evidentially qualified preface (in- in my modest opinion it was not a withdrawal symptom, lines 8–9; see Heritage and Raymond, 2005) that frames the assessment as a personal opinion and qualifies the opinion as a modest one (on indicators of evidentiality in doctor–doctor interaction, see Atkinson, 1999; Hobbs, 2003). After this evidentially qualified assessment, NP comes back to the territory of knowledge he masters better than RC: the objective status of the patient (lines 10–13). He repeats the information about fever with shaking tremors and further underlines its critical features: the fever jumped from 38 to 39 and a half (lines 12–13).
NP accomplishes two relevant actions that are consequential as to how his contribution will be received: he makes relevant a suspicion of infection (the patient is in his 24th hospital day, his blood temperature is high and his shaking tremors are not a symptom of withdrawal) and downgrades his own right to do that. Not surprisingly then, RC receives the report and its marked information through the ‘doing nothing’ resource (see Excerpt 2, line 9): he provides a marked acknowledgment (right, line 15) that basically signals the closing of the sequence. This is exactly what NP does in line 16: he changes the topic and does not expand on peak fever (on ways to receive news with tokens that do not encourage further elaboration and ‘may mark the end of an informing sequence’; see also Maynard, 2003: 101).
Reporting test results and referring to antibiotic treatment
The next example (shown previously in Excerpt 1) illustrates two other discursive resources used as a mean to make relevant a diagnosis of infection and therefore orient toward the hypothesis that the patient might undergo an ATBT: ‘reporting test results’ and ‘referring to antibiotic treatment’. NP is presenting the case of Isola, a patient in the weaning phase.
(4) And anyway he has a BAS with gram positive cocci [CEICU_Im_9]
After having constructed the patient’s condition in a mitigated no-treatment implicative way (see Excerpt 1), NP makes the alternative diagnostic trajectory (infection at stake) relevant. By introducing his statement with a conjunction plus concessive adverb (‘and anyway’, line 18), he refers to the presence of some detected bacteria (‘BAS with gram positive cocci’, line 18) and markedly signals that the patient is not on antibiotics (line 19). In doing so, he continues to display the partial alignment with the ‘watch and stay’ approach already exhibited at the very beginning of the report phase (see Excerpt 1, line 3).
Discursive resources to project and resist diagnostic trajectories in the report phase.
Discussion
Notwithstanding contemporary pressure for evidence-based medical decision-making (Castel, 2009; Timmermans and Angell, 2001) and for practices relying on ‘objective professional voice’ (Kovarsky et al., 2005: 119), our study on medical talk in an ICU confirms the unavoidable constitutive role of discourse in shaping medical activities as crucial as diagnosing and treatment recommendation (Alby et al., 2015; Atkinson, 1999; Brown, 1995; Glenn and Koshmann, 2005). We have illustrated how physicians contentiously project diagnostic trajectories and treatment recommendations from the initial phase of the morning briefing where they are supposed to report the patient’s objective status as a foundation for the subsequent assessment and planning phases. In particular, we have illustrated how physicians differently construct the assessable object as ‘treatable’ or ‘not yet treatable’ and how, in doing so, they ‘talk into being’ (Heritage, 1984: 290) two markedly different policies concerning ATBT in ICUs, position themselves toward them, and pursue or resist the unwritten yet officially shared ‘off-label’ policy adopted in the ward: the ‘watch and stay’ approach.
Surprisingly, our study reveals that differences in orientation toward treatability and some patterns of interaction identified in doctor–patient interaction (e.g. offering candidate diagnosis, projecting treatability, resisting a ‘no problem diagnosis/no treatment decision’, pursuing different and often conflicting agendas; see Stivers, 2002a, 2006, 2007) are demonstrably at stake even among physicians of the same specialty, that is, having the same epistemic access to relevant expert knowledge and similar rights to claim it (see Heritage, 2012a, 2012b). Although it is reasonable to expect agreement when the diagnostic reasoning and treatment prescription fall within a domain of ‘special knowledge possessed and controlled’ (Heritage, 2006: 85) by all the participants, our study reveals that their relative symmetry in epistemic status does not guarantee such a shared consensus. On the contrary, it requires substantial discursive work to make diagnostic and treatment alternatives relevant.
Particularly, we analyzed six discursive resources used mainly by NPs to construct the assessable object as ‘treatable’, thereby resisting the implied and known ‘off-label’ policy of the ward: partial assessment (Excerpt 1), mitigating no problem condition (Excerpt 1), selecting and underlining relevant parameters (Excerpt 2), anticipating and confuting alternative candidate explanations (Excerpt 3), referring to ATBT (Excerpt 4) and reporting test results (Excerpt 4). We also illustrated one discursive resource used by the RC in receiving his colleagues’ report: ‘doing nothing’ with information (Excerpts 1, 2 and 3). We showed how in receiving information with tokens that did not encourage further elaboration, he pursued a ‘no-treatment yet’ agenda, weakened his colleagues’ assemblage and aligned to the ‘watch and stay’ clinical line of the ward.
The following summary reports the discursive resources analyzed above (see Table 1).
The resources here analyzed appear to be the primary ways through which physicians, since the report phase of the morning briefing, orient to and make ‘actionable through talk’ (Heritage, 1997: 222) the different policies concerning antibiotic prescription in ICUs. Their assemblage of the assessable object is therefore highly theory-implicative yet also praxis-constitutive: it channels the other participants’ contributions as well as the team decision-making process.
The known discrepancy between the international guidelines strongly recommending empirical therapy and the therapeutic line of the ward creates a clinical dilemma as well as a practical problem for CEICU members: they have to decide every day, for each and every patient and for ‘another next first time’ (Garfinkel, 2002: 182) to not follow the evidence-based guidelines and make this understandably controversial decision highly accountable.
We argue that physicians’ discursive oscillation (see Cooren, 2010) between two locally available yet extremely different clinical alternatives accounts for the way decisions are interactively accomplished: the interaction analysis reveals indeed that this diagnostic path is not taken for granted nor followed in a mechanical, protocol-like way. On the contrary, this extreme policy appears to be jointly and contentiously accomplished one interaction at a time (Garfinkel, 2002) by participants who recurrently manage to make relevant the alternative guideline-oriented empirical approach.
In an ICU where doctors work all the time on the border between life and death, displaying what we may call a disaffiliative alignment toward a radical, guideline-divergent policy leads team members to take a reflective stance toward their professional vision (Goodwin, 1994) and to follow their policy as a highly accountable, non-standardized process sensitive to the contingencies of each and any particular case.
Footnotes
Appendix
Acknowledgements
This article has been written within the framework of the national research project Phenomenology of Infectious Diseases in Intensive Care Units (PHENICE) financed by the Mario Negri Institute for Pharmacological Research (Milan, Italy) and coordinated by Dr Guido Bertolini, MD. The ethnographic study has been coordinated by Letizia Caronia (University of Bologna) and Luigina Mortari (University of Verona). Giuseppina Mesetti and Roberta Silva (University of Verona) and Marco Pino (Loughborough University) participated in data collection and analysis. We wish to thank all ethnographic team members for their collaboration in data collection and analysis. We wish to thank also the medical and nursing staffs of the intensive care units (ICUs) where we conducted the fieldwork for their essential collaboration in data collection and interpretation. We are most grateful to Kathy Metzger for revising our English text and an anonymous reviewer for insightful comments on the earliest version of the article.
Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship and/or publication of this article.