key: cord-0948928-k8b7g29t
authors: Menichetti Delor, Julia Paola; Borghi, Lidia; Cao di San Marco, Eugenia; Fossati, Ivan; Vegni, Elena
title: Phone follow up to families of COVID‐19 patients who died at the hospital: families' grief reactions and clinical psychologists' roles
date: 2021-01-28
journal: Int J Psychol
DOI: 10.1002/ijop.12742
sha: c129397225233a0ced5d2853706987fc6275bd18
doc_id: 948928
cord_uid: k8b7g29t

A proportion of persons affected by coronavirus disease‐19 (COVID‐19) die and do so in extraordinary circumstances. This can make grief management extremely challenging for families. The Clinical Psychology unit of an Italian hospital offered a bereavement follow‐up call to such families. This study aimed to explore the families' experiences and needs collected during these calls, and the role that the psychologists played through the call. A total of 246 families were called over 3 months. Multiple qualitative methods included: (i) written reports of the calls with relatives of patients who died at the hospital for COVID‐19; (ii) qualitative semi‐structured interviews with psychologists involved in the calls; (iii) observation of psychologists' peer group discussions. A thematic analysis was conducted. Six themes emerged: without death rituals, solitary, unexpected, unfair, unsafe, coexisting with other stressors. Families' reactions were perceived by psychologists as close to a traumatic grief. Families' needs ranged from finding alternative rituals to giving meaning and expressing different emotions. The psychologists played both a social‐institutional and a psychological‐human role through the calls (e.g., they cured disrupted communication or validated feelings and choices). This study highlighted the potential of traumatic grief of families of COVID‐19 victims, and provided indications for supporting them within the space of a short phone call.

Indeed, loss, especially unexpected loss, can start a process of rebuilding and transformation, where "many people come to realise their own strengths, appreciate the impact of their relationships, and have new spiritual insights" (Calhoun et al., 2010, p. 125) . However, some individuals might manifest complicated grief, also known as persistent complex bereavement disorder in the Diagnostic and Statistical Manual of Mental Disorders, 5th ed. (American Psychiatric Association, 2013) . In this case, they experience severe and disabling responses to loss, which differ from the normal bereavement for type of symptoms, duration and intensity (Crunk et al., 2017; Lichtenthal et al., 2004; Stroebe et al., 2007) .

Literature reviews have pointed out some risk factors for complicated grief (Mason et al., 2020; Stroebe et al., 2007) . In addition to an individual (personality and attachment traits, psychiatric history, poor physical health, spirituality, coping style, difficulty accepting death) and relational factors (poor quality of the relationship with the deceased, family conflict at end-of-life), the type of death (e.g., violent, unexpected or untimely), the place of death (the hospital, and especially the intensive care unit, compared to death in hospice care or at home), and the lack of emotional and social support have been found to be key risk factors for complicated grief (Mason et al., 2020; Stroebe et al., 2007) .

COVID-19 bereaved families may be simultaneously exposed to multiple risk factors (Carr et al., 2020) , which, as suggested by some theoretical contributions (Bertuccio & Runion, 2020; Eisma et al., 2020; Gesi et al., 2020) are expected to affect the grief process, for example, by delaying a person's ability to adapt, heal and recover. The COVID-19 deaths are often rapid and unexpected. Moreover, due to the pandemic regulations for healthcare environments (especially at the beginning of the pandemic), people sometimes cannot be with their loved ones when they die and/or cannot physically mourn together with their family and friends (Wallace et al., 2020) . Pandemic restrictions also may lead to limitations in funeral ceremonies. This unusual absence of traditional rituals is claimed to undermine the grieving process of these families (Burrell & Selman, 2020) . Indeed, funeral ceremonies mark a transition in which the irreversibility of the death can be fully realised, offering a moment for expressing emotions and a starting point for recovery (Mitima-Verloop et al., 2019; Romanoff & Terenzio, 1998) . Without funeral ceremonies, families may miss a crucial occasion to give meaning to the loss and find social and emotional support (Burrell & Selman, 2020) .

In such a unique situation, there are no precedents in history that can help organising the support to families in managing their losses and prevent complicated grief (Carr et al., 2020) . Especially at the beginning of the COVID-19 pandemic, the possibility of intervention through clinicians-relatives communication or hospital support services was limited by COVID-19 safety restrictions, limited time and available resources (Gesi et al., 2020) . As an example, the large number of patients who needed hospitalisation and intensive care treatments, as well as the high mortality rate, required hospitals to reorganise their services to prioritise the medical treatment and care of COVID-19 patients. The possibility of emotional support and psychological intervention by directly interacting with families has been also affected by safety restrictions (Sun et al., 2020) .

Recent contributions have proposed literature-based recommendations for supporting COVID-19 bereaved families and preventing dysfunctional grief (Carr et al., 2020; Selman et al., 2020; Wallace et al., 2020) . Among the many recommendations, all these studies highlighted the importance of following up families by calling them after the loss (i.e., phone follow-up), as a mean to assess more serious symptoms that may require additional support. In general, a phone follow-up to families is considered a key element for bereavement support (Bromberg & Higginson, 1996; Hudson et al., 2012) . However, literature provides only generic indications for tasks to be accomplished in this type of contact (e.g., providing information, sharing the grief, assessing the need for further support) (Parkes, 1998) . It is unclear what tasks and actions need to be specifically accomplished to meet families' needs. Recent reviews highlighted that there is no evidence for specific interventions, models of delivery, or timing for bereavement support (Efstathiou et al., 2019; Pattison et al., 2020) . A scoping review underscored the need to better understand bereaved families' experiences to develop appropriate interventions (Coombs, 2015) . Furthermore, bereavement follow-up is usually made by nurses or other healthcare professionals worldwide (Bromberg & Higginson, 1996; Stephen et al., 2009) . Given the COVID-19 circumstances, such contact may need the expertise of a psychologist, in order to properly assess psychological risks, support families in their normal healing processes, and reduce the burden that hospital professionals already experience within this pandemic. The scientific literature offers few examples of effective early psychological interventions for family members who lost a relative in a hospital unit, such as elicitation of feelings and information provision (Wang et al., 2010) . These interventions usually require dedicated staff and time, which are limited resources in a pandemic. Looking at the literature on psychological support during traumatic events and emergencies, a systematic review proposed the "screen and treat" model as potentially able to provide effective follow-up support and assessment or the Psychological First Aid Model where any discussion, support or referral is tailored to the person's needs (Rose et al., 2002) . Basically, the screen and treat model highlights the importance of assessing psychosocial needs and resources and bases any action on them, while the psychological first aid model represents an approach to reduce the initial distress caused by traumatic events by stabilising psychological functioning, supporting people to adapt to the situation, and helping them to access further care if indicated. Even if some studies have suggested that this type of action can have greater benefit immediately after the event than after more than 48 hours (Campfield, 2001) , evidence about the best timing for these interventions is still unclear (Currier et al., 2008) .

Within this context, during the first-wave of the COVID-19 pandemic, the Clinical Psychology Service of a big public hospital in Lombardy, Italy, decided to offer a phone follow-up to all family members of patients with COVID-19 who died at the hospital. This call was considered as a psychological bereavement follow-up led by families' emerging needs and experiences and aimed to: (i) be the last act of hospital care, thus, serving as a place to answer questions, address urgent issues, or facilitate connection with other services; (ii) identify relatives who may show highly-complex/at-risk situations (and/or limited protective factors) and refer them to further psychological support; (iii) support normal grief processes. In setting up this particular service, given the lack of indications about the best timing for this type of intervention and given the aims of the follow up, it was decided to contact the relatives 48-72 hours after they received the death announcement.

This study aimed to explore the contents and the functions of this early psychological phone follow up, from the perspectives of the clinical psychologists involved in the calls. In particular, it aimed:

• to describe the bereaved families' experiences and needs as they emerged in the calls; • to identify the main functions that psychologists fulfilled within the space of the call; • to clarify the roles played by an early phone bereavement psychological follow up from the perspectives of the psychologists involved.

The findings are explorative and from a local sample of psychologists working in a hospital in Northern Italy; however, they might help in drawing recommendations for how to support bereaved relatives after a death in a hospital setting during a pandemic.

A qualitative approach was chosen to fit the exploratory nature of the study (Morse, 1994) . We adopted multiple qualitative methods (Morse, 2009) , in order of collection: phone-related documents review, observation of psychologists' peer group discussions and interviews. These were combined in a triangulated approach during the data analysis, in order to ensure completeness and confirmation (and therefore internal validity) of the information provided (Kimchi et al., 1991) . Such a methodological triangulation was performed to cross-check data from the three sources of methods, capture different dimensions of the contents and functions of the calls, and test validity of findings through the convergence of information from different sources. Two different methods to collect psychologists' perspectives (observation of peer group discussions and interviews) were chosen to explore both the psychologists' social group perspectives and each individuals' inner experiences with the calls, and these were combined with psychologists' written reports of the calls to embed the "objective" accounts of what happened in the calls.

The study was conducted in accordance with the Declaration of Helsinki Ethical Principles and the broader research which included the present study was approved by the Ethical Committee of the University of Milan. All the materials (documents, transcripts, and notes) were self-reported by the psychologists involved in the study and anonymized before analysing them. Written informed consent was obtained from the participants.

The Azienda Socio Sanitaria Territoriale (ASST) Santi Paolo and Carlo is a large public healthcare organisation in Milan (Lombardy region, Italy) covering hospital and community care services. It provides more than 150.000 emergency services per year and has a capacity of 400 emergency beds. The Clinical Psychology unit within the hospital offers psychological assessment, support and therapy to patients with any kind of psychological sufferance .

At the beginning of March, the Clinical Psychology unit reoriented part of its activities to address the emerging needs of the hospital caused by the COVID-19 emergency. The Unit's main goal has been to provide psychological support to healthcare professionals, hospitalised COVID-19 patients and their relatives (Cao di San Marco et al., 2020; Leone et al., 2020) . In particular, for healthcare professionals, two types of psychological support were organised: a safe room where to "decompress" (e.g., reflect, relax, get in touch with emotions) and the small group sessions; COVID-19 patients have been offered private sessions with the psychologists about their ongoing situation; and bereaved relatives of persons who died in the hospital for COVID-19 have been offered support through a phone call.

The Clinical Psychology unit decided to call the families of the deceased patients, about 48-72 hours after the family received notification of the death from the hospital 4 MENICHETTI ET AL. (Campfield, 2001) . The main purposes of such phone calls were: to support the family offering a space to express the loss-related emotions; to verify and sustain spontaneous psycho-emotional resources; and eventually, if the psychologist observed highly-complex/at-risk situations, to refer for further psychological support. A brief starting speech outline was shared among the clinical psychologists: introducing him/herself with name and role (i.e., psychologist employed in the hospital); asking for a confirmation of the interlocutor's identity as a family member of the deceased patient; presenting the call as being part of the hospital care pathway; and asking for consent to proceed. The calls presented some unique features for a psychological intervention: no prior referral, unclear multiple purposes, lack of setting, lack of face-to-face contact, and critical timing (i.e., a single conversation in proximity of the loss).

Twenty psychologists (14 females) conducted the phone calls. From March 19th to June 15th, they called 284 families and had a conversation with 246 family members (38 family members were unreachable). Calls involved a wide range of family members (Table 1) . On average, each psychologist conducted 12 calls (range 2-42), with an average duration of 18 minutes each (range 1-60 minutes).

Data were collected by a female research psychologist (JM), expert of qualitative research methods, external to the ASST and living in a different Country (so to approach participants and the context with "naïve" eyes). Multiple methods were employed to collect call-related data including: (i) reports written by psychologists after the call, leaflets explaining the call service, psychologists' personal notes; (ii) qualitative semi-structured individual interviews; (iii) non-participant observation with field notes of peer group discussions among the psychologists involved in the calls.

A total of 246 anonymized written reports of the phone calls that psychologists filled after each conversation were collected. Such reports summarised the contents of the call, with additional information on timing and on the position of the interlocutor within the family. Other phone-related documents were collected and used as materials for analysis (e.g. leaflets, psychologists' notes). Overall, these data comprised 93 pages of written documents.

Individual semi-structured interviews were conducted using the Zoom platform with the eight psychologists who conducted at least 10 calls, so to include information-rich cases for an in-depth exploration of their experiences and views. Purposive sampling with a criterion of intensity of the experience was used to identify eligible participants, given the need to ensure all participants were knowledgeable experts able to competently and efficiently inform study outcomes (Patton, 1990; Tongco, 2007) . They represented a wide range of psychotherapeutic approaches. They included five females; their average age was 50 years (range 36-63); and they worked as psychologists 22 years in average (range 10-40). Interviews lasted about 45 minutes. They followed a semi-structured question route organised in a way that concrete, more objective facts (the contents of the calls and the actions performed in the calls) were touched before moving to abstract and self-reflective aspects (the roles played by the calls). The main areas investigated were: (i) families' experiences and needs that emerged during the calls; (ii) strategies in place in the families to manage the loss; (iii) actions performed by the psychologist during the call; (iv) roles played by the calls according to the interviewed.

Interviews were audio-recorded and transcribed verbatim.

Four peer group discussions among the psychologists who conducted the phone calls were observed and written notes of the discussion were taken by one of the authors (JM). These discussions were conducted by a psychotherapist. On average, seven psychologists participated at each group discussion, which lasted about 90 minutes. Such discussions had the clinical purpose of sharing experiences of the calls, exploring difficulties and reflecting on the role taken on during the calls. Written notes of the non-participant observation of these discussions were used as materials for the data analysis. Such notes included all the verbatim statements from psychologists, as well as the researcher's personal notes on the organisation, interpretation and mood of the discussion.

Data were integrated to develop a comprehensive view of what were the contents of the calls (and specifically the families' experiences and needs discussed in the calls) and the functions of the calls in the perspectives of the psychologists involved. In particular, what was combined were psychologists' interactive cognitions (as from the peer group discussions), their knowledge and beliefs (as from the interviews), and the descriptive report of the families' behaviours in the calls.

The analysis used an inductive approach and followed the principles of thematic analysis (Braun & Clarke, 2006) . Two research psychologists (JM, LB) (i) read and familiarised with the data, with a focus on the contents and functions of the calls, (ii) wrote notes on paper and extracted first recurrent themes linked to relevant quotes (preliminary themes were written on an Excel sheet with linked source and quotes, and reflected as much as possible the words of participants), (iii) discussed preliminary themes, clustered them and extracted super-ordinate themes (using the Excel sheet and organising themes for similarities/differences), (iv) checked themes with data (going back from the Excel sheet to the extended documents) and (v) reached consensus on final themes (by discussing selected themes and their organisation in meetings). Two psychologists involved in the calls were asked to constantly discuss the on-going emerging findings, and the final draft was shared with all participants to qualitatively validate findings.

Such a process was performed in three key steps of data triangulation, where data from the three methods were recursively analysed and included.

The first step of the data analysis was the use of written reports and documents of the calls as the database to extract key contents of the calls (i.e., families' experiences of bereavement, families' needs and spontaneous or supported grief coping strategies). This was a preliminary stage of analysis, where the goal was to establish preliminary categories to describe the data and to serve as a basis for the analysis.

The second step was the triangulation and enrichment of the first themes resulting from the written reports with the observational notes from the peer group discussions. Two authors (JM, LB) first read through the observational notes and compared the preliminary categories about the calls' "contents" to the observational data to assess their validity, reformulated them or generated new categories from the data, returned to the data to assess their validity and went on doing so until the categories suited the data satisfactorily. This was in particular performed by creating joint display tables in Excel, and by going back and forth from the Excel summaries to the Word documents. Observational notes also provided information on the roles played by the calls (this information was not displayed in the written reports), which was inductively summarised by the same two researchers in preliminary themes as for the contents. Preliminary findings emerging from the analysis were also used to inform and refine questions in the individual interviews, which were performed at the end of the data collection process.

Finally, as last step, themes related to contents/needs collected in the calls and functions of the calls were again integrated and enriched with data from individual interviews, by adding additional themes, confirming those already emerged, and providing interpretation for organising them. Interviews particularly provided materials for confirming themes related to the contents, enriching and validating those related to the functions, and interpreting and combining the different themes in a comprehensive picture. Final themes resulting from the integration of data from the interviews were cross-checked with the overall amount of data. A similar process of analysing multiple methods with a procedural approach was described in Meijer et al. (2002) .

During the research, we employed several strategies to improve validity, reliability and generalizability of findings. First, researchers were engaged in a constant reflexive process, critically self-evaluating each positionality in relationship to the data and the context (Mitchell et al., 2018) . For example, the two researchers analysing the data (JM, LB) had very different positions ("external" vs. "internal," in terms of workplace, relationship with the participants, involvement in the COVID-19 situation), and constantly discussed how their position was eventually interacting with the data during the data collection and analysis. A reflective journal was also maintained to document decisions. Participants were also invited to self-reflect on their experiences, and this resulted in some of them producing personal reflective notes which were included in the analysis. Second, there was a prolonged engagement with participants, through the observation of the peer group discussions, the interviews and the participants checking of findings and themes. This enabled a clarification of findings as an ongoing process, true to participants' accounts. As an example, in the last interview emerged unprompted the final organisation and interpretation of findings. Fourth, we adopted a triangulation process during the data collection and analysis. Fifth, participants were all those who performed the calls in the defined period (which spanned from lockdown to phase two measures with lifted mobility restrictions and reopening of companies with social distancing measures in place), and written reports covered all the calls done in that period. We finally provided verbatim extracts from materials to connect themes with participants' accounts. These strategies are known to ensure rigour of findings in qualitative research (Buetow, 2019; Morse, 2015; Noble & Smith, 2015) .

The thematic analysis of the data yielded several themes within two main areas: (i) the contents of the call, that include the families' experiences and needs; (ii) the perceived functions and roles of the calls. An overview of these themes is presented in Figure 1 and an in-depth description will follow, along with exemplificative quotes taken from the written reports ('Report'), the group discussions ('GD'), and the interviews ('Int').

During the calls, psychologists collected the families' perspective on the death of their loved one hospitalised for COVID-19. For relatives, the dying processes were: (i) without death rituals, (ii) solitary, (iii) unexpected and fast, iv) unfair, (v) unsafe, (vi) coexisting with other stressors. These themes are described in Table 2 .

From the psychologists' perspective, because of such extraordinary circumstances, the regular grieving process was at risk. During the interviews and the group discussions, they expressed concern for the possibility of a grief possibly complicated by a traumatic, violent experience of loss during the emergency: Different needs emerged from families during the calls in response of their challenging experience of loss (Table 3) : (i) to give meaning, (ii) to give voice to the full range of emotions, (iii) to find symbolic ways to say the last goodbye, (iv) to remember, (v) to solve practical issues.

Psychologists conducting the calls faced a unique scenario, full of challenges. On the one hand, families expressed grieving needs and a traumatic experience of loss, in a unique combination due to the unprecedented emergency circumstances. On the other hand, the calls were new for psychologists in terms of setting, referral, timing and limited possibility of intervention. As one of the psychologists shared in the first GD: Physical isolation was an element featuring not only how victims died, but also how families lived the grief. The physical distance from the patient made families feel like they had not done enough, creating feelings of guilt and deprivation. Social distancing also did not allow family members to mourn together with others. Nevertheless, the presence of social support, even if not physical, was an important protective factor.

The lady's biggest concern is the impossibility of staying physically close to the brother during the hospitalisation: she imagines him very scared (Report) On the other hand, there is the theme of the social resources, it was evident that those who had this support were much less in need (Int2 

The perception of unsafety had different shapes. There was an unexpected lack of medical knowledge and a mistrust in the health institutions (which should heal, but instead became one of the places of the contagion). Then, there was the loss of the idealisation of the human body: even a healthy person can be turned down by this virus. Lastly, there was an overall lack of stability and certainty in the daily life.

where people should have felt safe were a contagion site (Int8)

Experiencing the death of a dear one for COVID-19 was often a part of a more complex scenario in which stressors tended to pile up. Sometimes there was more than one sick relative, maybe hospitalised or dying, within the same family. People struggled to contain the spreading of the virus into the family. Some relatives had the possibility to visit their loved one for the last goodbye, but they had to deal with the risk of taking the virus themselves and spreading it to others. This kind of dilemmatic choices left feelings of regret and guilt.

The lady is quarantined too, and this makes it difficult for her to access the first grieving phase (Report) The theme of their own health, the surprise for not having taken the virus themselves, the quarantine, the worries for their own health (Int8) we are fostering a grieving process, but these persons are alone so how can we foster a grieving that cannot rely on essential steps? Like the human physical consolation, or the attachment. We risk activating the grief, without the relative having the possibility to continue grieving. We need to revise our meanings, positions, roles. What is our role here? We are facing classical themes for the psychologist (grief, death, trauma, trust) 

In this unique scenario, "performing a call as psychologist" was a challenge. The calls were perceived as a delicate equilibrium among an institutional mandate, a human mission, and a psychological intervention, where each psychologist had his/her own epistemological approach and working praxis on how to approach patients. In managing this delicate equilibrium, a perceived risk in their view was to have the role of an "encyclopaedia salesman," that is, of someone knocking at the door and selling pills of knowledge. To avoid this, psychologists found opening/closure formulas, structures, meanings and positions to make the calls a psychological action. It emerged that psychologists enriched the original shared opening speech outline and 8 MENICHETTI ET AL. Psychologists used labels like "putting the pieces together" or "shouting their stories", to express this need of meaning-making. When the victim was sick or old, this was done by families by seeing death as a natural course of events, sometimes even a relief from painful conditions. Families were trying to normalise death in the current extraordinary circumstances. Similarly, addressing faith or hope, with thoughts like "it was destiny", was an effort to come to terms with the loss.

They needed to tell, to scream their stories, they were isolated so they needed to talk, and we know that talking helps in processing and listening to ourselves (Int7 Some families expressed the need to solve the practicalities required by the virus. These were perceived by psychologists as the expression of the natural oscillation between confronting the painful reality and avoiding it by doing something else.

She needed a trusted person to ask for the procedures on how to go out (Report) There were also concrete needs, like finding a way to go to the cemetery or solve bureaucratic issues, but this was a residual part of needs, functional for not being stuck (Int8)

The structure of the phone call (1) Opening, with formulas to express caution: Check of respondent, introduction with name and role (i.e., psychologist employed in the hospital), reason of the call (2) Offer of a space to talk (3) Active listening of family member experiences, emotions, and needs (4) Need-based psychological action (e.g., information giving, education on stages of grief, emotional validation, or small therapeutic actions like relaxation pills) (5) Eventual referral to further psychological support (6) Closure: emphasis on closeness/warmness structured the call more or less in the same way (Table 4) , with opening formula to show caution (as a way to manage the feeling of intrusiveness, the embarrassment of not knowing what to say, and the fear related to not knowing who is the respondent and his/her story) and closing formula to express closeness (as a way to make the other feel cared, but also as a way to deal with not knowing what to say and not having the possibility to use non-verbal communication).

For example, one psychologist reported: I always started by mentioning, after having checked who was on the other side of the phone, my affiliation with the Clinical Unit of the hospital. Then, based on what I was perceiving from their tone of voice, I always tried to enter on tiptoe and saying that there was this listening space for all those who lost a relative at the hospital due to the COVID-19 and if there was something that the relative wanted to say, I always also asked if I was disturbing (Int1). Another psychologist said: at the end, what surprised me is that I decided to say "a hug from us and from the hospital"; I said it the first time because I did not know what to say and to stress my closeness, but then I implemented it when I saw that families were extremely grateful for this, so it became a sort of ritual to relief the pain, to testify that in that hospital where their beloved ones died there is a psyche, it was a way to close the circle (Int4).

The functions of such calls were multiple but transversal to the psychologists' experiences. Different themes of their role and function emerged, which can be positioned along two lines: social-institutional and psychological-human.

In the social-institutional line, the first function played by psychologists with the calls was to substitute the ritual of death itself, in different ways. First of all, psychologists felt the call as a possibility to open to symbolic burials which help families saying goodbye to their dead through a ritual (GD4). Secondly, the act of collecting the memories of these deaths may open the possibility to activate a collective ritual, as if society is participating, which is what usually happens in the public funeral (GD3). Moreover, the act of presenting ourselves as part of an institution embodies the coral moment of the funeral rite; in this way, we activate the relationship with the social context (GD1). When covering this social function, psychologists felt like they had the symbolic role of "a flower" or "the incense" at the funeral (our role is mostly to substitute the act of the hospital of placing a flower on the tomb, Int4).

From the same social-institutional perspective, specifically from the hospital as an institution rather than from society itself, psychologists view the call as the closing act of the hospital caring process, humanising it. The call reassured the family that the human part of clinical care was fulfilled. In their words, it was meaningful because families have been able to feel that the hospital has a psyche, not only tools, that it has nurses with a name who call families every day; and our phone call completed the institutional process of care (Int4). Within this function, psychologists perceived their role as being "a closing mark" or a "customer satisfaction service."

A final institutional function was to act as "a bridge" and to have the role of putting bandages to a non-harmonic communication process (GD1). In a situation where relational bridges had been limited or interrupted, such calls had the function of curing communication bridges between patients, families, clinicians and others involved in the final care of victims (e.g. undertakers), seeking to stimulate dialogue and undoing the knots of interrupted communication.

Along with these institutional functions, psychologists felt that the calls were covering also more psychological and human functions. First of all, they felt to be like "ferryman," like a companion going with them into the woods of grieving. They thought that the call played a role in helping families starting the grieving, processing the event, by uncovering the silence surrounding those deaths and enhancing the translation of what happened into words. This was perceived as an important task in a situation where the socially co-constructed meaning-making processes were not available. The psychologists felt like they had the role of pulling out the frozen pain or to encourage them finding the words to communicate the loss to the young nephews (Int3), but also, on the opposite side, simply being with them in their silence and solitude, with compassion, and, if it was the case, with honest words to express closeness without being commercial or consolatory (Int4).

In addition, all psychologists felt that the calls were having the important function of validating and giving legitimacy to the entire range of emotions that the families were feeling, and lifting families from feelings of guilt, and help them doing a reality assessment about the impossibility of doing something different from what they did (GD1). In other words, the call helped families reassessing the reasoning behind the decisions made (especially the dilemmatic ones), in order to prevent complicated grief. In this regard, the psychologists reported that it was crucial to maintain a position of acceptance and neutrality regarding the different feelings and emotions of the relative, without being judgmental and moral (GD2).

Another role that they felt like they have taken on through the calls was of "educators": they answered practical questions (e.g., about how to collect the personal belongings), but they also educate families about the normal stages of grief and the emotions experienced.

Furthermore, the psychologists felt like "assessors," because one of the key actions they performed during the calls was assessing the mental health status of the relative, checking for protective and risk factors, and eventually orienting to further psychological support.

Finally, psychologists reported that they took on the role of a "mirror" for the experiences and the emotions of the relative. Within this role, they felt like they were echoing what the relative was feeling. This was one of the most challenging aspects for psychologists, because of the consequent personal involvement: there was this strong resonance, I was particularly empathic; it was of course useful and important, but I was very sensitive and very personally engaged in their stories (Int7). Another psychologist reported that it was like swimming in the same water, which was good because I knew their situation, but also challenging because you are in like a peer; we were all on the same boat in this dramatic and new experience (Int2).

This study highlighted unique aspects and needs of families' mourning right after the loss of a loved person for COVID-19 at the hospital. Moreover, it described the different functions that a bereavement follow-up call played according to the psychologists involved, along with their connected roles.

Regarding the families' experience, the findings showed an intersection of loss and trauma first reactions, complicated by factors related to the current emergency scenario, such as the limited protective factors (e.g., social support, life opportunities) and the presence of common precipitating/perpetuating factors (e.g., isolation, feeling of guilt, lack of traditional rituals). The psychologists involved in our study reported a wide range of emotional reactions among families: avoidance and anaesthetic reactions, hyperarousal and preparedness/already in place acceptance of the loss (especially in cases where the deceased was already sick and/or old). It is demonstrated that grief and coping with grief may take many forms (Bonanno & Kaltman, 2001) . The most common anaesthetic reactions resembled the first stage of grieving, that is, numbness (Payne, 2007) , but were described as particularly intense. This could be due to the proximity to the loss but also to the COVID-19 scenario. Indeed, several features of the COVID-19 deaths' experiences (e.g., isolation, impossibility to see the victim's body and to hold funeral rites, unexpectedness, rapidity, neither clear information nor preparation, coexisting dilemmatic choices and other stressors that reduce opportunities to redefine life and relationships) added to the grieving process. These aspects are risk factors for complicated grief as identified by psychological literature (Burke & Neimeyer, 2013; Love, 2007) , which usually characterise situations of traumatic loss like disasters and wars (Neria & Litz, 2004) . Therefore, findings from this study suggest that the COVID-19 pandemic may bring an additional traumatic burden on families, similar to other mass violence situations, and that families may confront a new, more subtle and hidden type of trauma, together with the loss, which can possibly slower the recovery trajectory. In addition to the risk factors, most of the protective factors to well adapt to bereavement (Lyons & Chamberlain, 2006) , like family involvement, physical presence, clear information (Dose et al., 2015; Virdun et al., 2017) , were negated in the COVID-19 situation. Social relationships, and in particular the physical aspects like hugs and touch, which are basic resources for grieving and finding meaning in the loss (Bradbury, 2012; Parkes, 1998) , were also hindered. In models and theories considering the social determinants of grief (Bradbury, 2012; Pressman & Bonanno, 2007; Stroebe & Schut, 1999) , grief is viewed as a phenomenon dependent on other people who facilitate the process of making sense to the loss. Relationship with others also fosters post-traumatic growth (Johnsen & Afgun, 2020) , which could be an unexpected outcome of violent, traumatic losses. In the case of the families in this study, the possibility to ease the focus to the loss with other activities with the help of family and friends, as suggested by the dual-process model of coping with grief (Stroebe & Schut, 1999) , was at stake. Thus, our findings, although focused on a particular moment of families' reactions very close to the loss, give credit to suggestions from other researchers (Bertuccio & Runion, 2020; Eisma et al., 2020; Shear, 2012) : families of COVID-19 victims may possibly deal with a complicated grief in the future or limited opportunity for post-traumatic growth because of the co-occurrence of risk factors and limited protective factors.

In addition to the understudied combination of trauma and grief (Neria & Litz, 2004) , other challenges were linked to the extraordinary circumstances dictated by COVID-19. The families presented both common needs of grieving, such as meaning-making, memories' expression and narratives' employment, expression of (negative or positive) emotions (Park, 2010; Stroebe et al., 2007) ; and others specific needs due to the COVID-19 situation, like the need to hold ceremonies with what was possible or to solve communication, information and bureaucratic stubs.

Considering the mentioned limitations dictated by COVID-19, our findings suggest that psychologist might fulfil both a social-institutional purpose and a psychological-human one, in order to support the first loss reactions of these families in the complexity of the loss situation, in the space of a phone call. These phone calls can represent a sort of primary preventive intervention for bereavement, that the literature suggested as important especially when it is accessible and free of charge (Stroebe et al., 2007) . Based on existing literature in similar situations to COVID-19 pandemic (Regehr & Sussman, 2004; Rose et al., 2002) , the treatment procedures (e.g., psychoeducation, mental health assessment, small relational/psychodynamic actions like dialectical thinking, active and compassionate listening, reminiscence, meaning-making, and cognitive-behavioural actions like cognitive restructuring of ruminative or maladaptive thoughts)) might be useful for providing early support to a normal grief process (and to prevent complications) in a stressful, traumatic, and violent situation of loss for COVID-19.

Some unique roles taken on by the psychologists emerged from this specific work with COVID-19 bereaved families, such as covering social-institutional functions or the expression and validation of new emergent feelings, in a delicate equilibrium between non-judgmental hospitality of all experiences and the intrusiveness of the call, carefulness and personal involvement into families' stories. Working on the line of carefulness and hospitality while covering a social-institutional function may have been a way to avoid the risk of intervening in an early stage and cause a disruption in the course of grief.

Bereavement follow-up services are usually delivered by bereavement coordinators, nurses or volunteers across the globe (Bromberg & Higginson, 1996; Stephen et al., 2009) . However, the multiplicity of functions and roles played by the calls and the unique emergency, almost traumatic circumstances of the loss, may suggest the need to have a professional with the necessary expertise to manage the range of families' loss reactions and to appropriately evaluate and refer to further support cases that may need specific psychological help. In particular, being the family member in an extremely vulnerable moment, such an intervention may require particular caution and professional care.

Some of the findings of this study might reflect Italian cultural aspects, for example, the relevance of (the lack of) funeral rituals, the perceived violence of cremation, or the importance of the social relationships in sharing and processing grief. However, funeral rituals and the social environment are commonly reported factors in the international grieving literature (Bradbury, 2012; Pressman & Bonanno, 2007; Stroebe & Schut, 1999) ; the perceived violence of the cremation highlights the importance of considering (and, when possible, respecting) religious beliefs when managing the loss and supporting the bereaved family (Shaw et al., 2005) . In general, findings may reflect, more than cultural aspects, the specific circumstances of the COVID-19 situation during restrictive preventive measures. Some of these circumstances may differ from one country to another or from one period to another (e.g. the possibility to see the beloved one or other family members/friends, the possibility to hold a funeral ritual), thus, reducing or enhancing the grief risk/preventive factors. The perception of emergency may also soften during time, as people can get used to living with the pandemic. Therefore, findings may be relevant for psychologists or hospital professionals around the globe who are dealing with following up families who lost a loved one under stressful circumstances. In particular, the actions performed during the calls and the functions enacted by the psychologists may provide practical indications about the multiple tasks that a similar phone follow up may require, thus, helping psychologists (or other healthcare professionals) orienting their professional actions based on the roles they may need to play. Findings from this qualitative study involving a local sample of hospital psychologists can also provide indications for healthcare organisations around the globe about how to use psychologists' expertise in addressing bereaved families' needs and preventing eventual psychological implications of the COVID-19 crisis, as part of the hospital care process. Indeed, the qualitative nature of the study allowed the generation of new insights on how psychologists can address bereaved families' early-stage needs within the short space of a phone call, under the new emergency circumstances of COVID-19, and as part of the hospital caring process. Further studies should study the implementation of such an intervention in other healthcare institutions and countries, its impact on families' grieving outcomes in the long-term, and longitudinally assess short-and long-term grieving reactions together with other psycho-social outcomes like resilience, post-traumatic growth, complicated grief, and social support.

The present study has some limitations. First, experiences and needs of families were not directly collected from them but were reported by the psychologists. Moreover, even if all the calls to the families reached during the research period were included in the analysis, more than half of the family members were sons and daughters. Therefore, findings may mostly reflect the needs, experiences and actions required from this particular group. At the same time, psychologists did not report to have tailored the intervention based on the family member's relationship to the deceased, and this aspect did not emerge as relevant for delivering the calls. Finally, our results only described the families' needs and the functions of psychological bereavement follow-up at 2-3 days from the loss, therefore, they reflect reactions in a very early stage of grief and we do not know how the mourning process developed. Findings do not also contribute to the debate around utility of psychological or psychotherapeutic interventions in the grief process (Hoyt & Larson, 2010) , as the intervention proposed was a bereavement follow-up as part of the hospital service more than continuous psychological grief counselling and the study did not focus on the effect of this intervention.

As far as we know, this is the first study reporting COVID-19 bereaved families' themes and needs in the first days after the loss, along with the main psychological functions and roles that psychologists might play during a hospital phone follow-up. The multiple and extensive qualitative data, the focus on first grieving reactions after COVID-19 losses, and the in-depth exploration of what psychologists can do to support families within the brief and limited space of a phone call, represent points of novelty of the study. Findings have several implications for healthcare practice, policy and future research.

They highlight a complexity and multiplicity of needs of bereaved families grieving during emergency circumstances like COVID-19. Such families may need specific psychological follow up. Healthcare organisations may need to activate psychological resources, that can support the hospital in the process of managing losses. They also provide concrete indications for psychologists about how to manage a phone follow up to such families, in terms of actions, functions and roles played. Finally, findings show the need for future research on the understudied area of intersection of grief and trauma, and on first, early-stage reactions to challenging losses that may require further support.

Diagnostic and statistical manual of mental disorders (5th

Considering grief in mental health outcomes of COVID-19

The varieties of grief experience

Prospective patterns of resilience and maladjustment during widowhood

The psychological consultation in hospital setting: A model of intervention through the clinical practice

Representations of death: A social psychological perspective

Using thematic analysis in psychology

Bereavement follow-up: What do palliative support teams actually do

Apophenia, unconscious bias and reflexivity in nursing qualitative research

Prospective risk factors for complicated grief: A review of the empirical literature

How do funeral practices impact bereaved Relatives' mental health, grief and bereavement? A mixed methods review with implications for COVID-19

Positive outcomes following bereavement: Paths to posttraumatic growth

COVID-19 emergency in the hospital: How the clinical psychology unit is responding

Bereavement in the time of coronavirus: Unprecedented challenges demand novel interventions

A scoping review of family experience and need during end of life care in intensive care

Excess deaths from COVID-19, community bereavement, and restorative justice for communities of color

COVID-19 Dashboard by the

Complicated grief: An evolving theoretical landscape

The effectiveness of psychotherapeutic interventions for bereaved persons: A comprehensive quantitative review

Dying in the hospital: Perspectives of family members

The state of bereavement support in adult intensive care: A systematic review and narrative synthesis

Prolonged grief disorder following the coronavirus (COVID-19) pandemic

Complicated grief: What to expect after the coronavirus pandemic

Guidelines for the psychosocial and bereavement support of family caregivers of palliative care patients

Complicated grief and post-traumatic growth in traumatically bereaved siblings and close friends

Triangulation: Operational definitions

Psychological interventions in hospital during the first-wave of CoViD-19: An overview of the experiences of the units of clinical psychology in

A case for establishing complicated grief as a distinct mental disorder in the DSM-V

Progress in understanding grief, complicated grief and caring for the bereaved

Health psychology: A critical introduction

Complicated grief: Risk factors, protective factors, and interventions

Who do we think we are? Disrupting notions of quality in qualitative research

Facilitating grief: An exploration of the function of funerals and rituals in relation to grief reactions

Designing funded qualitative research

Mixing qualitative methods. Qualitative Health Research

Critical analysis of strategies for determining rigor in qualitative inquiry

Bereavement by traumatic means: The complex synergy of trauma and grief

Issues of validity and reliability in qualitative research

Making sense of the meaning literature: An integrative review of meaning making and its effects on adjustment to stressful life events

Coping with loss: Bereavement in adult life

Bereavement in critical care: A narrative review and practice exploration of current provision of support services and future challenges

Qualitative evaluation and research methods

Contemporary views of bereavement and the experience of grief

With whom do we grieve? Social and cultural determinants of grief processing in the United States and China

Intersections between grief and trauma: Toward an empirically based model for treating traumatic grief

Rituals and the grieving process

Psychological debriefing for preventing post-traumatic stress disorder (PTSD)

Bereavement support on the frontline of COVID-19: Recommendations for hospital clinicians

Religion, spirituality, and posttraumatic growth: A systematic review

Grief and mourning gone awry: Pathway and course of complicated grief

Mental health disorders related to COVID-19-related deaths

Bereavement and bereavement Care in Health and Social Care: Provision and practice in Scotland

The dual process model of coping with bereavement: Rationale and description

Health outcomes of bereavement

Addressing mental health care for the bereaved during the COVID-19 pandemic

Purposive sampling as a tool for informant selection

Dying in the hospital setting: A meta-synthesis identifying the elements of end-of-life care that patients and their families describe as being important

Grief during the COVID-19 pandemic: Considerations for palliative care providers

A study of psychological crisis intervention with family members of patients who died after emergency admission to hospital