We found 4,456 studies, of which 11 studies were eligible (Figure 1). All but one study used interviews, with the other one using focus groups and participatory action . An additional three used focus groups alongside interviews [38-40]. The most common analytic approach was thematic [9, 40-44], alongside framework [9, 45], grounded theory [38, 39, 46], and narrative  methods. One study was conducted in Austria , one in Canada , one in Denmark , one in Malaysia , five in the UK [9, 39, 40, 46, 49], and two in the USA [43, 50]. Studies included African-American women , Latinos , Malay, Chinese, and Indian people , and adults >61 years .
The contextual focus of four studies was primary care [9, 41, 45, 46]. The remaining studies were related to secondary care , or views of mental health conditions and mental health care generally [40, 43, 44, 47, 48, 50]. Authors classified treatment-resistance via depression inventories (psychometric tests), including the International Classification of Disease . Treatment-resistance was also classed via diagnosis [38, 41, 43], evidence of recurrent depression [45, 47], continuous use of antidepressants [38, 41, 44, 46], seeking of secondary or tertiary care , and patient's self-description . No studies sampled people with obsessive-compulsive disorder, panic disorder, or post-traumatic stress. Table 1 summarises the characteristics of included studies.
[Insert Figure 1: PRIMSA Flow Diagram and Identification of Studies]
[Insert Table 1: Characteristics of Included Studies]
We judged all but one study to be of high quality using the CASP. One study was judged to be medium quality because it did not clearly report their research design or recruitment methods or consider ethical issues . Two studies were perceived as very high quality, scoring a 10  and a 9.5 . Other high-quality studies were perceived to be less clear on their recruitment strategies. For the full quality assessment results see Table 2.
[Insert Table 2: Results from Quality Assessment]
Descriptive Themes Summary
We developed four descriptive themes to reflect the content of included studies. These four themes describe a cyclic experience of primary care for people with TRMHCs. These stages are: barriers and crisis point; seeing a GP; treatment; and self-management (Figure 2). We focus on people's experiences with antidepressants because other depression medications (e.g., beta-blockers) were not mentioned in the primary studies. Primary quotations are presented to support our themes, and demographic information is included where provided in the original study. Table 3 shows the distribution of descriptive themes.
[Insert Figure 2: Cyclic Care for TRMHCs in Primary Care: Visualisation of the Descriptive Themes]
[Insert Table 3: Distribution of Descriptive Themes]
Stage One: Barriers and Crisis Point
This stage refers to the barriers that prevented people from seeing a GP for their mental health and the point people understood their mental health as declining. One barrier included people's preference to self-manage their mental health. Self-management activities included acupuncture, music therapy, exercise, prayer, aromatherapy, and dietary changes [39, 40, 42-44]. Other activities included late-night working, smoking, alcohol, and illicit drugs [9, 43]. Some participants described how they were engaging with these activities because they preferred to manage their mental health "on their own" and without the support of a GP [38, 39, 43]. These activities were not seen as cure but a form of respite [38, 39, 43].
It was evident that participants were self-managing their mental health because of perceived barriers to accessing primary care. For example, in five studies, participants mentioned that the perceived lack of emotional support from friends and family decreased their probability of booking an appointment with their GP [38, 39, 41-44, 47]. One participant spoke about how she hid her anxiety attacks from her family, who told her that her condition was "stupid" . Another participant explained how her mother downplayed her depressive symptoms as "growing pains" that she could "just walk off" . Participants with children felt that seeing a GP could cause unnecessary worry to other family members [38, 42, 47]:
"You don’t want to overburden your children... they’re young, and they have little children and busy lives, and as a parent, you don’t want to be the needy one." (Female, 67 years, Persistent Moderate to Severe Depression) .
Across most studies, it was evident that stigma (including self-stigma) acted as a help-seeking barrier [9, 39, 41-43, 46, 47]. Participants commonly described their poor mental health as a "trivial" problem , that "somebody's worse off than we are, so we just got to deal" , and that seeking GP support is "an admission of failure" . Some participants were told by others to "pull yourself together"  and that seeing doctors for mental health was a sign of being "loco (crazy)" . Many African-American women also spoke about how the "strong black woman" stereotype prevented them from seeking care . These women also perceived prejudice within their healthcare system when it came to supporting black people's mental health :
"My depression might not be like Suzie Ann’s depression, OK? Well, they’re going to call her name before they call my name. And they’re going to treat her just a little bit more different than me." (Female, African-American, Major Depression) 
Despite these perceived barriers, participants often felt that they reached a crisis point where the symptoms were so severe that they had no choice but to seek medical help. Many participants used metaphors to communicate these experiences: "anger ball" , "wanting to get out" , "I feel like I have something here [touching her chest], like a car" , "on edge" , "a volcano bursting" , and "a wall of pain" . Others were more direct with their descriptions, referencing the chronicity and severity of their symptoms :
"I had a lot of work stress going on as well, and it all got on top of me... I was massively overeating, oversleeping, permanent low mood, just generally unwell... At that point, I went to the doctor and said, ‘Look, this is what’s going on. I need some help with this". (Male, 52 years, Treatment-Resistant Depression) 
In contrast to earlier cases, some participants were motivated to seek medical support because of people in their personal network [38, 42, 47]. One participant mentioned seeking medical help after her eight-year-old daughter noticed her crying a lot . Another participant was forced to go to hospital during a mental health crisis by her neighbour and was told by her sons, who had not realised how severe her depression was, to seek help :
"One night I phoned my neighbours and asked them to come over, cause I said I was feeling so awful. She came over, and said, "I think I’ll call an ambulance". I said, "oh no, I don’t need an ambulance". She said, "Well then, I’m taking you to the hospital". I was there until the next afternoon. Then my sons came, and they hadn’t realised what rough shape I was in, and they said I needed help." (Female, 70 years, Persistent Moderate to Severe Depression) 
Stage Two: Seeing a GP
At the second stage, participants sought GP support for their mental health and appeared to feel confident that their GP could improve their depression [38-41, 47]. For example, one participant said that her GP had "never let me down" , and another said she had "absolute faith" in her GPs ability to treat her mental health . For these participants, GPs were an accountable person who could facilitate discussions around accepting their mental health condition and deciding on possible treatments [38-41, 47]:
"The doctor tells me. You have to accept your diabetes. You have to accept your high blood pressure. You have to accept....bad moods...you accept your problems, you have to accept your illness”... That is what I am trying to do, accept" (Female, Major Depression, Diabetes) 
However, in seven studies participants were disappointed by short consultations with their GP. They felt that short consultations did not provide enough opportunity to discuss their mental health [9, 38, 39, 42, 43, 46, 47], and in four studies participants mentioned a lack of therapeutic continuity, which impacted on motivation to continue with treatment [38, 41, 42, 47]:
“The doctor kept changing. If every time we see the same one, we would have more confidence in that doctor and will continue the treatment.” (Male, Indian, Major Depression) 
Participants were further disappointed in the GPs advice because it did not meet their expectations around adjunctive care [41, 47]. In two studies, for example, participants felt their GP ”gave up” because they did not offer follow-up appointments or suggest counselling or other non-medicinal therapies [41, 47]:
I: "Why do you say that your general practitioner gave you up?"
P: "He just wrote the prescriptions, and then he was ﬁnished with me. He didn't say that I should return; he didn't say that I should come for some counselling; he didn't say, “I'd like to keep track of you”. “You can come and get a renewed prescription, and we'll talk." (Female, Depression, Depressive Episodes) (Buss, 2004).
Participants in six studies felt that GPs did not have adequate mental health training and instead offered antidepressant treatment rather than providing a more detailed mental health assessment or opportunities to consider non-medicinal therapies [9, 41-44, 47]. Some participants felt like their consultations came off as robotic or like their GP was reading off a script :
"It’s as quick as they can get you out, write a script and out you go again." (Female, 51 years, Treatment-Resistant Depression) 
Stage Three: Treatment
At stage three, participants engaged with antidepressant treatment. However, some also self-referred to psychotherapy [9, 38], secondary care services [9, 41], or were participating in a trial testing the acceptability of a mindfulness course prescribed in primary care . The views of the mindfulness course were generally positive, with most describing improvements in their mental health, sleep, and reduced isolation due to practising in a group setting . Only a few mentioned how the course did not fit with their schedule and disappointment that the course did not have the desired immediate impact on their mental health :
"I am able to deal with my emotions...I am not scared of things any more...I don't want to turn about and walk away from things...I'll take the time out to sit down and face up to it..." (Recurrent Depression/ Recurrent Depression and Anxiety) 
Views of antidepressants were considerably more diverse. Participants in four studies reported a compromise between ambivalence towards the efficacy of antidepressants and hesitancy to discontinue due to possible relapse [9, 41, 42, 46]:
"I didn’t wanna get involved in taking tablets for 6, 9, 12 months. I’m already 6 months into taking them now, which is longer than I thought... I thought, ‘Oh, I’ll get rid of it. I’ll be OK. I’ll have a few months, or I’ll have a couple of months off. I’ll be back to my normal self,’ but it hasn’t worked like that. Whether another antidepressant would help, I really don’t know." (Male, 55 years, Treatment-Resistant Depression) 
In this way, antidepressants were viewed as preventative against further mental health decline, rather than treatment in itself [9, 38-40]. Participants felt antidepressants could support them to achieve a “baseline” level of functioning, and this could equip them to work on the social and psychological causes of their depression [9, 38-40]:
"I’m on quite a low dose really, 20 mg of citalopram, and I think it was doing the job it needed to do…to get me to point where I could look at some issues." (Female, 39 years, Treatment-Resistant Depression) 
Subsequently, many participants felt antidepressants did not address the 'true cause' of their depression [9, 39, 44, 45]. One participant described antidepressant discontinuation exacerbated the psychological causes of their depression (e.g. all-or-nothing thinking)  while other participants emphasised the importance of talking therapies to support more holistic recovery:
"I just felt as if I wasn't in control anymore. They made me feel different [the antidepressants]. The same problems were there. So when I stop taking the tablets, I still had the emotional baggage and everything that I had stopped feeling when I started taking the pills... I've dealt with everything myself and at the end of the [mindfulness] course the feelings are still there, but I can deal with them so I would definitely feel that this [the mindfulness] is an alternative." (Recurrent Depression/ Recurrent Anxiety and Depression) 
Stage Four: Self-Management
At the fourth stage of the cycle, some participants stopped taking their antidepressants, mostly without the knowledge or support of a GP [9, 39, 41, 44, 46, 47]. Participants perceived a trade-off between the mental health benefits of antidepressants and the reported side-effects that affected other parts of their physical and mental health (worsened depressive symptoms, fatigue, affectless and apathy, sexual dysfunction, weight gain) [9, 39, 41, 44, 46, 47]. Under these circumstances, some participants followed advice from people in their personal networks who believed that long-term antidepressant treatment was not healthy [42, 43]:
“My family members told me not to take this medicine [antidepressants]. They said it’s not good to take so many medications especially for long term....so I don’t take it.” (Male, Chinese, Major Depression) 
Other participants experimented with antidepressants to counteract adverse side effects [9, 38, 39, 41-43]. Self-management included changing the dosage of the antidepressant, irregularly taking the antidepressants, and discontinuing antidepressants altogether [9, 38, 39, 41-43]. These self-management activities were thought to alleviate some of the perceived negative side-effects. Often participants had not told their GP about their experiments because they believed that they would not listen to their concerns about the side-effects and advise against it [9, 38, 39, 41-43]:
I: "How come you took this decision [to self-medicate] without talking to the doctor?"
P: "It was probably because the doctor would be against it. I think I have an appointment in about a month from now. I thought that if I stopped them, I could see if it reduced my tiredness, and if there are no problems, then there is no reason to take them." (Depression, Depressive Episode) 
However, often self-management of antidepressants was unsuccessful and caused participants to relapse with their depression [8, 41, 42, 45, 46]. Many of these participants eventually reached another crisis point and needed to see a GP for more support [8, 41, 42, 45, 46]:
"I tried to come off medication months ago, and I had a couple of little wobbles and stuff, so I went back on it." (Female, 26 years, Treatment-Resistant Depression) 
"The only reason I still take them now is because a) I haven’t actually technically been told not, you know to come off them, and b) I just think it’s not a good idea to just suddenly stop them like I did last time." (Female, 26 years, Treatment-Resistant Depression) 
High-Order Theme Summary
So far, the results describe each stage of a mental health cycle that people with TRMHCs can experience in primary care. We mind-mapped the descriptive themes to develop a high-order theme [34, 36], not observed by the original authors, to explain how the cycle could be broken. We show how this high-order theme was created in Figure 3.
[Insert Figure 3: Creation of the High-Order Theme: Breaking the Cycle]
Breaking the Cycle
Participants described opportunities to optimise the management of mental health conditions in primary care, which we interpreted as opportunities to break the cycle of care. In the first stage, support networks including family and friends encouraged participants to seek support from their GP [43, 44, 47, 48]. Seeking support early for a mental health condition helped people establish a relationship with their GP whereby they could ask questions about their depression including treatment [43, 44, 47, 48]:
‘… you actually feel worthless and as though you can never, ever have a normal life again. And I know that’s ridiculous because you do come out of it, but it’s … you, you just want to know why. And I mean probably a doctor or whatever can’t tell you why but at least they can ask the questions which might make you think about it, why". (Female, 56 years, Persistent Depression) 
Having an established and continuous relationship with a GP allowed conversations about whether to change/or stop treatment helped participants instil confidence in their care plan and prevented a transition to stage four: self-management [9, 40, 41, 44, 46, 48]:
"She’s [GP] been keeping quite a close eye on how I am and listening to me, she’s very good like that...Each time I see her, she says, come back and see me, and we’ll see how you’re going, and we’ll discuss again if you want to come down off the tablets." (Female, 53 years, Treatment-Resistant Depression) 
At all stages, participants appreciated shared-decision making. Participants wanted to talk about their depression and treatment options [9, 40, 41, 44, 46, 48, 49]. This could break the cycle by building people's knowledge and confidence and subsequent engagement in the treatment option:
“We must be very clear that we have a health problem now, and we need medication to recover from the illness. If we are sick and have to take a lot of medication, we have to take it. If we want to get well, we have to take medication.” (Male, Malay, Major Depression) 
"Well, since they treat me every six months… we hardly have talked, I only come and they look me over, and they say to me “where does it hurt, if it hurts”. They only prescribe me the medication and that is it." 
Generally, participants were aware of psychological therapies like counselling and mindfulness, and would have welcomed opportunities to talk with GPs about referrals to such services [9, 40, 41, 44, 46, 48, 49]:
"I think it works [counselling]... Like you said, you don’t have to take the meds, you know, you don’t have to take the meds—just meet with your counsellor once a week for an hour." (Recurrent Depression/ Recurrent Anxiety and Depression) 
"I think the type of support I would have wanted was somebody just to talk to me or tell me I could make it ... more so than “here’s medication”. (Female, 28 years, Pakistani, Currently Depressed) 
Participants who continued with antidepressants wanted to see their GPs regularly about remission/ relapse and whether to change their antidepressants [9, 40, 41, 44, 46, 48, 49]. This could break the cycle by giving people regular opportunities to talk to the same GP about treatment effectiveness and jointly plan treatment changes:
"I was taking sertraline for quite a while; about four months or so. I felt they hadn’t made enough difference. So, I went back to the doctor, and I was really worried, and I said... ‘Either up this medication or change it, do something" (Female, 49 years, Treatment-Resistant Depression)