Participants

Consecutive sampling was used to recruit individuals actively enrolled in a clinical trial of DBS for TRD who were awaiting surgery (Australian New Zealand Clinical Trials Registry: ACTRN12613000412730) (Clinical trial details, including inclusions/exclusion criteria, participant characteristics, surgical information, and full psychometric outcomes, will be presented in a forthcoming publication. Any correspondence regarding the clinical trial should be directed to paul.fitzgerald@anu.edu.au.). Participants were recruited by the first-author after permission to be contacted was granted via the clinical trial co-ordinator. Verbal and written study information was provided to the candidates and their respective caregivers. All who were approached chose to participate, including: six DBS candidates (5 women, 1 man, M_age = 52 years, range = 26–73 years) and five caregivers (4 spouses, 1 parent, 2 women, 3 men, M_age = 59.5 years, range = 45–75 years). One candidate lived alone and chose to participate independently. One DBS candidate who completed a preoperative interview did not complete a postoperative interview as it was deemed too burdensome for the patient. The current sample represents the majority, but not complete clinical trial sample, as the current study commenced after the trial began. DBS candidates had a long history of TRD (time since diagnosis: M = 18 years, range = 8–42 years) and met Stage V criteria of the Thase and Rush [28] classification for TRD. None of the DBS candidates were engaged in paid employment. The DBS candidates and caregivers had long-term relationships, spanning decades (M = 37 years, range 24–50 years).

Procedure

One-on-one, semi-structured interviews were conducted between 2016 and 2019 with the first-author, a female psychologist with training in qualitative methods and experience interviewing DBS patients, families, and clinicians. Interviews were conducted face-to-face (at participants’ home, research centre) or via telephone or video-conference for participants living interstate. Interviews were conducted separately with patients and caregivers to allow open discussion. Pre-surgery interviews occurred 3 to 15-weeks prior to surgery (M = 46 minutes, range = 34–58 minutes) and explored participants’ knowledge of DBS and beliefs about how the patient’s personality, self, and relationships may be affected (see S1 File for interview schedules). Participants were not provided with a definition or description of ‘personality’. We were interested specifically in their personal experiences of personality and identity change. This was done to avoid restricting participants’ responses to a particular conceptual framework. Participants then underwent DBS surgery and were implanted with Medtronic Activa 3389 electrodes in the BNST. After a recovery period, a pseudorandomised schedule of active or sham (control condition) stimulation commenced, with participants blinded. Over five months, five stimulation settings were trialled: one inactive, two low-level (2 volts), and two moderate-level (4 volts). Following this, stimulation continued in an open-label manner, with parameters optimised according to individual responses. Any existing treatments (medications/psychotherapy) were kept constant during the blinded phase with participants able to make changes during the open-label follow-up. Post-surgery interviews occurred 9 to 11-months after stimulation initiation (M = 55 minutes, range = 36–86 minutes), roughly 3-months into the optimisation phase. These explored experiences living with DBS and the perceived personal and relational impact. Field notes were maintained during the interview process. Audio-recordings were transcribed verbatim by a professional transcription service, then reviewed for accuracy (by CT) and de-identified using pseudonyms.

Data analysis

Interview transcripts were entered into NVivo 12 software (QSR International) and analysed using thematic analysis [29]. This is an iterative and inductive process that allows novel themes and patterns of meaning to develop from the data. Data was analysed according to the six-step process outlined by Braun and Clarke [29]. All interviews were coded by CT. Data saturation, the stage at which no novel themes were evident (commonly used to determine sufficient sample size in qualitative research), was reached at interview 9 of both the pre- and post-DBS interviews. Approaches to increase the quality and credibility of the results included cross-coding a subset (6) of interviews with the research team (AC, RS) to ensure consistency in coding. Any disagreements in coding were discussed until a consensus was reached. Triangulation, where a phenomenon is examined from different perspectives, was achieved by inclusion of a diverse research team (clinical psychology, neuropsychology, neuroethics, social science).

Impact of mental illness and treatment on self-concept

Prior to DBS, participants expressed how severe and persistent depression had obscured defining aspects of the patient’s self and personality (Table 1). These included traits such as being outgoing, confident, and a “bit of a joker”. Three patients had previously enjoyed fulfilling careers in health care. These caring professions were integral to patients’ identities and their inability to perform them due to mental illness was a significant loss. Patients’ interest and ability to engage in meaningful and rewarding activities, such as intellectual and creative pursuits, were profoundly impaired and detrimental to self-worth.

Years of depression had created a sense of an absent person and life, with patients merely existing. This was pronounced in patients receiving regular ECT, which caregivers characterised as a means of keeping the patient alive rather than a treatment that remediated their depression. When asked if they thought DBS could change the patient’s personality, most participants did not expect or were unconcerned that it would. Some hoped regained wellness would restore their premorbid personality. Others expressed concern DBS may alter personality in an undesirable way, but felt comfortable that adjusting or ceasing stimulation would resolve this, or otherwise they would simply “have to cope”. These concerns were minor compared with participants’ primary preoccupation: would the intervention be effective and how would they cope if it wasn’t?

Caregivers expressed hope that the person they had known prior to depression would be restored, but acknowledged that after long-term mental illness they were unlikely to be exactly the same. Caregivers also identified qualities they valued in their loved-one that they hoped would remain after DBS, such as being a loyal partner and parent, a gentle and caring person, and sharing a close bond. However, some were willing to lose these if it meant the patient could achieve relief from depression and improved quality of life. When imagining an ideal picture of their future self or loved-one, participants desired normality and the ability to simply “do things”. Some patients envisaged achieving professional goals or travelling, while others wished to feel more confident and at ease amongst people.

With DBS, patients’ perspectives of themselves were greatly influenced by the procedure’s perceived benefit. Those who experienced a subjectively meaningful antidepressant benefit (i.e., described a pronounced improvement in mood that was sustained over multiple months) conveyed a sense of restored self (n = 2) (Table 2). Those who experienced modest or transient benefits (n = 3) (i.e., described subtle improvements in mood and/or notable lifts lasting only 2–3 days duration) recognised encouraging moments of their prior self (e.g., increased curiosity, interest in work activities). Caregivers recognised aspects of their loved-one previously, but were less emphatic about this restoration, noting the persistence of qualities that had developed with depression (e.g., sympathy-seeking, self-focused).

Decreased depression was accompanied by some transformation in self-concept; however, patients remained aware of the substantial distance between their idealised self (e.g., physical appearance, having friends, working etc.) and actual self. Novel qualities, including irritability and anger (n = 3) were reported by patients and caregivers. These were considered reactions to either the challenges of social reintegration, the procedure’s perceived lack of effect, or a side-effect of a particular stimulation setting.

Patients who experienced no benefit felt fundamentally unchanged by DBS (n = 2), but transient experiences of acting uncharacteristically and feeling unlike oneself were reported in the context of stimulation-induced manic episodes (n = 1).

Device acceptability and usability

In preparation for DBS, patients considered the prospect of having an implanted device, their relationship with it, and how this may affect their view of themselves and their body (Table 3). Most anticipated it having little impact and being easily adapted into their self-view. With the device encased within the body and not directly visible to others, most patients predicted being rarely conscious of it. Many likened it to other implanted devices for medical conditions, such as pace-makers and insulin pumps. Unlike some of these static devices, one patient highlighted the dynamic relationship with DBS, requiring setting adjustments and battery recharging. Another patient expressed mild disgust imagining electrodes embedded in their brain and the IPG in their chest. They anticipated feeling physically self-conscious, especially while recovering with a shaved head and scarring. Whether the intervention was successful was considered likely to dictate how the device was viewed (i.e., with pride or disgust).

Participants had considered implications of having a DBS device, including not flying after surgery, navigating airport security scanners, and needing reliable access to electricity, none of which were considered particularly restrictive. Participants were expecting to receive a rechargeable battery (to increase battery longevity) and were aware of the need to regularly recharge. Some compared this process with charging a mobile phone–an everyday part of life. Those comfortable with technology felt confident they could learn to manage the device independently. Less confident patients expressed fear of “pressing the wrong button” and wanted control left in the professionals’ hands. The need for device maintenance and specialist care in the distant future played on some patients’ minds.

After surgery, most patients felt conscious of their physical appearance (shaved head, bandages, scarring) and how others perceived them (Table 4). For one patient, only once the stimulation was turned on and exerting some effect did they become conscious of their appearance. Patients avoided discussing the procedure or its purpose with others unless asked directly (“I just don’t want people to think I’m a freak”). Some experienced pain, discomfort, and tightness around their IPG and subcutaneous wiring in certain positions or postures. For some, comfort and freedom of movement came with time, while others made adaptations to alleviate device-related discomfort (sleeping on one side, cushioning pillows). Patients tended to see themselves no differently with DBS and accepted the device as part of their body. For some, this acceptance had taken time with their perspective influenced by whether they believed DBS was having an effect.

One aspect all patients struggled with to varying degrees was battery recharging. The recharging experience had been one of frustration and annoyance, even for those who considered themselves easy-going or tolerant. Frustration was associated with difficulty establishing a reliable connection, inconsistency in how long and often recharging was required, recharging taking longer than originally expected, and the growing need to prioritise recharging in daily life. Recharging requirements varied depending on each individuals’ settings (greater recharging burden associated with higher voltages). Some were still establishing an optimal recharging routine. Many had made contact with their device company representative to receive reassurance and guidance around recharging technique (e.g., understanding how body temperature, posture, and tension influence connection).

Within the open-label phase, arrangements for adjustment of stimulation settings varied for each individual. They were influenced by the patient’s geographical location, accessibility of support people (caregiver, psychiatrist), size of the intended adjustment, plus the patient’s preference. Patients’ level of comfort controlling the device (turning on/off, small setting adjustments) were consistent with their pre-DBS comments around technological confidence. Some preferred all adjustments be made by the DBS specialist, but this was not always possible or practical, due to geographical distance. This occasionally required patients to make adjustments with caregiver or psychiatrist support. One patient with no caregiver had all adjustments made while in hospital. They described experiencing cognitive side-effects with one setting and how these effects would have impacted their ability to re-adjust the settings independently. Outside the specialist clinical trial team, participants’ experience with medical professionals broadly was that DBS, particularly for depression, was not well-understood. Device company representatives who provided patients with device education peri-operatively, were also engaged to guide medical professionals around appropriate device management during medical procedures unrelated to DBS.

Relationships and connection

In pre-surgery interviews, patients and caregivers discussed how their relationship had been impacted by depression (Table 5). Those in spousal relationships had experienced a shift in dynamics, with caregivers performing and identifying with the ‘carer’ role to varying degrees (“pseudo” to “full-time”). All expressed their commitment to and support of one another, but acknowledged their relationship was not reflective of a true husband and wife relationship. Couples described their inability to plan due to the uncertainty of the illness. Plans for socialising or holidays were avoided, with couples living a day-to-day existence and keeping their world small (family, close friends only). Caregivers had freedom to do activities independently, but expressed their preference to share and enjoy these experiences with their partner. In addition to intense sadness and suicidal ideation, some patients experienced moments of aggression and intolerance, which could be directed towards caregivers. A decrease in open, honest communication was described. Some caregivers were “walking on eggshells”, while some patients withheld details about the depths of their depression or suicidal thoughts. In certain cases, suicide attempts or completions had been prevented by caregivers, either in a physical sense or due to patients not wanting to cause them anguish. Despite difficulties posed by depression, patients and caregivers maintained strong and meaningful bonds. For patients with children, discomfort was associated with their children seeing them unwell and patients avoided allowing them too much insight into their illness. Social relationships more broadly were considered effortful to maintain and the prospect of social interaction was anxiety-provoking for most patients.

Both patients and caregivers hoped aspects of their relationship would change after DBS. Patients hoped caregivers would not have to regularly bear witness to their intense sadness or anger. Some felt indebted to their caregiver and hoped to be able to repay them. Others desired greater physical intimacy with their spouse and hoped DBS would improve their sex drive. Caregivers hoped for a more equal relationship and to be able to enjoy “small things, but meaningful things” together (e.g., having a coffee, going for a walk). Many caregivers experienced worry and guilt when leaving the patient at home alone for fear of what they would return to and hoped DBS would alleviate this heavy burden. All participants wished for greater ease when interacting with others and that time spent with family and friends would be enjoyed rather than endured.

Some participants reflected on how society tends to respond to individuals with mental illness. It was felt that people distance themselves when depression is involved and rarely show the same compassion and support to families that they would if it were a different illness, such as cancer.

Post-DBS, those who had experienced notable benefits described how their relationships had both improved and become more complicated (Table 6). Changing dynamics and re-negotiating roles occasionally led to increased conflict (e.g., patient more interactive and independent, caregiver needing to relinquish control). Despite improvements, some dynamics remained ingrained (e.g., patient not contributing to housework, being waited on by caregiver). Patients’ increased energy and motivation to do things was generally considered beneficial for the couple. However, caregivers occasionally tempered patients’ impatience and enthusiasm to jump back into things quickly (e.g., overseas travel, return to work). These appeared driven by patients’ desire to make up for lost time or re-establish their independence and identity. Caregivers felt able to be more open and expressive in their communication, although would still withhold or filter certain information. Patients’ risk of self-harm had reduced, which was a considerable relief for caregivers. Relationships with family tended to improve, with patients more engaged and nurturing with children and grandchildren. Patients expressed interest in developing or re-establishing friendships, but were also hesitant due to out-of-practice social skills and fear of judgement. Mental health support groups had been considered as avenues for building social connections.

For those who experienced little-to-no benefit from DBS, their relationships remained mostly unchanged and functioned as before. However, the process of undergoing surgery and participating in the trial resulted in some changes to relationship quality. For one patient, the time and energy they had available for others prior to DBS was depleted through the process of regularly travelling interstate for stimulation adjustments. This was difficult for some friends to understand and accept, resulting in more strained relationships. The tendency for some patients to display irritability after DBS led to caregivers becoming more irritable themselves. Instances of stimulation-related adverse events (e.g., mania, suicidal ideation), had been confronting experiences for caregivers and close-others and left lasting impressions. The experimental procedure had prompted more discussions about depression amongst families (including children), but patients still avoided the topic so as not to burden others.

Caregivers discussed the relationships they had available to support their own wellbeing. Most caregivers had friends they could speak openly with or adult children they could seek support from. Some caregivers took antidepressant medication to help manage their daily stresses but none were engaged in psychotherapy, although a number felt perhaps they should be. A lack of support available to carers of people with mental illness, both in society generally and within community services (e.g., support groups), was noted.

Personality and identity in treatment-resistant depression

Clinical and ethical concerns have been voiced about the potential for neurointerventions such as DBS to result in personality change [30, 31], yet in the case of psychiatric disorders some researchers and clinicians argue this is the intended outcome [32, 33]. Synofzik and Schlaepfer [34] suggest that if personality–that is “a dynamic and organised set of characteristics in a person that uniquely influences his or her cognitions, motivations, and behaviours in various situation” (p. 1514)–remains unchanged, the intervention has failed. The hope many have, including the current participants, is that DBS will restore premorbid personality and self. This is a reasonable desire; however, it should not be expected that individuals who have endured years or decades of depression (as most patients had) will go completely unchanged by this experience and be restored unmarked [35]. In the present sample, onset of depression for most was during adulthood (n = 5) after they had established distinct identities and personalities. The degree to which patients felt their old personality or self had been restored was closely associated with the degree of benefit they experienced with DBS. Still, even patients who felt much restored identified ways in which they were far from who they were previously, particularly regarding functional capacity to perform meaningful activities and roles (working, travelling, socialising). Similarly, caregivers recognised elements of their loved-one previously, but continued to see persistent illness traits (e.g., sympathy-seeking, self-focus).

Previous DBS clinical trials have investigated the relationship between antidepressant effect and personality, as measured using the NEO Five-Factor Inventory (NEO-FFI). Psychometric personality scales, such as the NEO-FFI, often appear homogenous from individuals with severe depression. They are characterised by high Neuroticism and low Extraversion, regardless of age [36]. These profiles tend to reflect the nature of the illness more than the individual themselves [37], suggesting that if depression is alleviated through DBS it should result in a change in NEO-FFI scales. Bewernick et al. [38] found no difference in Five-Factor scales at 6-months, 2 years or 5 years after DBS of the supero-lateral branch of the medial forebrain bundle (n = 21), despite patients obtaining an antidepressant response. The authors suggest this may reflect a ‘scar effect’ in which lengthy depressive episodes result in long-term, irreversible changes to personality. In contrast, Ramasubbu et al. [37] found a significant decrease in Neuroticism and increase in Extraversion following DBS of the subcallosal cingulate (n = 22), with patients assessed regularly up to 15-months. The personality changes were associated with clinical improvement in depression, with the direction of these changes aligning with normative personality data. The contrasting outcomes across the two studies may be explained by differences in target nuclei, assessment timelines, and patient populations. Results from the current study could lend support to either finding, given those who experienced antidepressant effect described a sense of a restored personality, yet the lingering depressive qualities support Bewernick et al.’s suggestion of a ‘scar effect’.

While an exact restoration of premorbid personality or self is improbable, if not impossible, they serve as helpful guideposts for recovery. The process is more complex where mental illness has emerged in adolescence, during important stages of personality and identity development [39, 40]. This is often the case in OCD and regularly occurs in TRD. Following DBS, rather than attempting to pick up where one left off, these patients must attempt to forge a new identity and personality that is not structured around mental illness [33, 41]. Identity challenges and poor psychosocial adjustment seen post-operatively in neurological populations has been described as the burden of normality [42, 43] and has been applied in psychiatric DBS cases [44]. This phenomenon where patients and caregivers experience difficulties transitioning from chronically ill to suddenly well, highlights the need for preparatory work and psychosocial rehabilitation in DBS to assist with regaining wellness and transitioning self-concept. Not all of the current patients were engaged in psychotherapy after DBS, but for one patient who gained benefit from DBS this support was vital for their psychosocial adjustment. In OCD samples, cognitive behavioural therapy (CBT) has been shown to augment DBS treatment effects [45] and is recommended for best-practice post-operative care [46]. In TRD samples, there is no evidence indicating that CBT augments the DBS treatment effect [47]; however, psychological therapy is likely to play an important role in relapse prevention. DBS is only a starting point to living well, as the scaffolding around the patient will still reflect someone with a chronic and severe mental illness (e.g., limited social connections, ingrained dependency on caregiver).

Relationship with and acceptance of the device

Patients did not experience sustained fundamental changes in self-perception or feel dehumanised by the implanted electrical device, as has been reported in some DBS studies [24]. However, certain factors impacted patients’ relationship with and acceptance of the device. These included: the noticeability of scarring, intermittent pain associated with IPG and wiring, DBS’ perceived benefit (or lack of), and recharging needs. Recharging was a frequently difficult and frustrating experience. If an implanted device functions well, it typically goes unnoticed and is easily incorporated into one’s body schema [48]. However, if it requires frequent attention and creates frustration (through regular, lengthy recharging), this impacts agency and may reinforce internalised feelings of defectiveness, particularly if no therapeutic benefit is gained. Patient satisfaction with rechargeable DBS products has been evaluated, but primarily in movement disorders [49, 50]. The recharging needs for individuals with psychiatric conditions typically differs from those with movement disorders (higher voltages resulting in faster battery depletion), as does their psychological and cognitive profiles. In OCD, post-operative education has been shown to assist with device-related anxiety [51], but further investigation is needed. Notable memory difficulties, which the majority of current patients reported (n = 4), can affect retention of device-related information. For this reason, perioperative device education may require modification for TRD purposes. This may include consideration of when device education is offered (e.g., not immediately after surgery) and conducting multiple sessions. Those who had received support from their device company representative in the months following surgery had found this helpful for improving their recharging process (achieving a more reliable recharging connection). A reduction in how long and how often recharging is required would increase the tolerability of the process. This is dependent on the capacity of the current technology. As DBS for psychiatric indications progresses, so too should device technology and research evaluations of patients’ device-related needs [17].

Relational adjustment

For patients who experienced therapeutic benefit, renegotiating the patient-caregiver relationship to a more egalitarian dynamic was a challenge. Even positive changes (e.g., patient exerting independence, eager for daily outings) were occasionally difficult for caregivers, as they broke predictable patterns. In De Haan and colleagues’ [52] OCD sample, patients reported that both they and their partners had to “get used to” the new them. As OCD had always been present in their relationship, couples had created lives that worked around the illness. For the married couples in the current sample (n = 4), their relationships were established well before the onset of depression. Therefore, partners were experiencing varying degrees of ‘reacquaintance’ with their loved one (depending on benefit gained from DBS). Regardless of outcome, DBS created challenges for marital relationships. These couples, however, had weathered similar storms across their long-term, committed relationships. So too was the case for the parent-child pair.

The caregiver participants in the study possessed valuable knowledge of the patient and were well-positioned to provide observational feedback. This is one reason caregivers should routinely be involved in patients’ DBS clinical care. Caregivers should also be included in clinical research to assess the impact DBS has on their own lives. DBS can be a transformative experience involving considerable adjustment for both parties [53]. In PD, there is growing recognition of caregivers’ vital role in patients’ DBS management, as well as the procedure’s impact on their own wellbeing. A lack of support to assist caregivers with the implications of DBS has been noted. A number of pilot studies have attempted to address this issue, including an 8-session psychoeducation program designed for patients and caregivers [54], and an 8-session individual CBT program for caregivers post-DBS [55]. Although small samples, these promising interventions contribute to an area of unmet need. Given the unique challenges created by DBS, programs designed specifically to meet these needs are recommended to support caregiver wellbeing.

Limitations

The patient sample (n = 6) was small and reflects the limited number of patients undergoing DBS for TRD in Australia. When collecting in-depth, longitudinal qualitative information, samples of this size can be sufficient and data saturation was achieved in the analysis of pre- and post-DBS data [56, 57]. The triangulation of caregiver perspective, in addition to multiple time points, added to the richness and depth of the analysis. Qualitative research strives for naturalistic generalisability (e.g., resonance with personal experiences and observations) rather than statistical-probabilistic generalisability, and is achieved through the select presentation of rich and in-depth subjective data [58]. It remains important, however, to acknowledge the influence of specific contextual features (e.g., target location, lead placement, research trial protocol, patient characteristics, geographical location) upon the experiences and outcomes described. While a mixed-method approach could have produced findings with both naturalistic and statistical-probabilistic generalisability, this approach was not feasible as the small experimental sample prohibited the conduct of robust quantitative analyses on psychometric data. These patients were extreme in depression and treatment resistance compared with others who have undergone DBS for depression globally. The majority of the patient sample were female (5 of 6), which may reflect the higher representation of females in depressed populations. This gender imbalance is likely to have influenced the nature of the observed findings, potentially reducing the transferability of these findings to males with depression. Few patients experienced sustained meaningful benefit at the time of follow-up (n = 2); therefore, our insights into the process of adjusting to wellness are from limited cases. There is a crucial obligation, however, to learn as much as we can from the few DBS for TRD cases that exist, regardless of outcome [59]. Repeat interviews in subsequent years when patients have had more time with the device would have been valuable to see how their experience evolved over time.