Summary
GPs may not be familiar with the concept of being at risk of developing psychosis. Some GPs mentioned that they may not be asking the right questions and would benefit from more training on the early symptoms of psychosis. GPs also reported that mild or short-lived psychotic symptoms often occurred in the context of other mental health disorders, which made the identification of these patients difficult. However, there were GPs who recognised this patient group, but reported that potential patients with an at-risk mental state for psychosis rarely consulted in primary care. In addition, GPs also mentioned that patients did not always feel comfortable disclosing psychotic symptoms. Those GPs who worked in areas where secondary care services were commissioned to offer treatment to patients at risk of psychosis were more likely to recognise this patient group.
The challenges of working within a healthcare system where resource limitations impose restrictions on appointment availability and length of consultations, as well as a lack of continuity of care were mentioned as having a negative impact on identifying these patients . Yet, GPs felt that being open, non-judgemental, and being able to establish a good therapeutic relationship could facilitate their identification. In psychological therapies, establishing a good therapeutic relationship has been shown to account for approximately 30% of the variation in psychotherapy outcome.23
GPs reported that identifying and managing patients with an at-risk mental state for psychosis could improve patients’ outcomes. However, there may be potential disadvantages such as the issue of overlabelling and potentially creating unnecessary worry at a time when GPs had little to offer patients in terms of providing effective interventions or referral to specialist services.
Strengths and limitations
Both male and female GPs were interviewed, with a range of clinical experience and who worked in areas where secondary care services were or were not commissioned to work with patients at risk for psychosis. Interviewing continued until data saturation had been reached and efforts were made to recruit GPs from the catchment areas of all six EI teams. The authors recognise though that the 21 GP practices that originally expressed an interest in the study, and from which 16 GP practices were purposefully selected, were self-selecting. It may be that GPs with a special interest in mental health were more likely to respond to the study invitation to participate, and this might have biased the present results in terms of including GPs who were perhaps more aware than their peers of this patient group. No difference was noticed in the depth of discussion between the telephone and face-to-face interviews, and research has shown that telephone interviews can gather the same material as those conducted face-to-face.24
The original topic guide was revised and GPs given a definition as some GPs were unfamiliar with the concept of at-risk mental state for psychosis. This helped ensure that GPs understood the patient group of interest, but doing so may have sensitised participants to this concept. After changing the topic guide, most GPs said that they were familiar with this patient group. This might be because about half of the GPs interviewed with the second topic guide worked in areas where secondary care services were commissioned to offer treatment to patients at risk of psychosis, but it could also be because providing a definition helped GPs recall patients they had consulted, or encouraged them to give what they thought were more socially desirable answers. Internal bias was minimised in data analysis by double coding some interviews, and discussing results with other clinicians, but the authors recognise that it would have been beneficial to have involved a GP in data analysis.
Comparison with existing literature
Other studies have shown that GPs may not recognise symptoms of early psychosis.19,20 The presented study extends these findings by highlighting factors that facilitate or hinder the identification of this patient group. The only study that has so far investigated predictive factors of identifying people at risk of psychosis used a semi-structured discussion with GPs to inform the construction of a questionnaire that was later applied to GPs working across England.21 That study found that GPs’ subjective norms, that is, a GP’s perception of whether their colleagues identify people with an at-risk mental state for psychosis, and whether other health professionals would approve of them doing so, were the strongest predictor of identifying these patients. The current study used semi-structured interviews and found that the identification of this patient group is a complex process that arises from an interplay of factors related to patients, GPs, and the challenges of working within the NHS.
GPs in the present study reported that they rarely saw patients with mild psychotic symptoms, which is consistent with findings by Simon et al.20 GPs also mentioned that there was a tendency for patients to consult only after their symptoms had worsened, and potentially transitioned to psychosis. Some support for this comes from a population-based cohort study, which showed that 50% of individuals aged 18 years25 and 30% of those aged 24 years12 who met criteria for a psychotic disorder had not sought professional help. However, other studies have shown that people with schizophrenia visited their GPs 43% more than controls in the 6 years before their index diagnosis,26 and that increasing frequency of consultations in primary care was a strong predictor of psychosis.27 This indicates that many people at risk of psychosis are indeed consulting, but the non-specific nature of early symptoms of psychosis, and high comorbidity with anxiety and depression, may hinder their identification as patients with an at-risk mental state for psychosis.27–29
The identification of these patients could be further complicated by the fact that patients who consulted did not always mention their psychotic symptoms. People with an at-risk mental state for psychosis most commonly consult for depression or anxiety.30,31 Therefore, GPs asking patients with depression or anxiety about psychotic phenomena could help identify individuals at risk of psychosis. Short screening tools such as the Primary Care Checklist could guide GPs as to when a specialist assessment might be warranted.32
Presence of suicidal behaviour and a pattern of increasing frequency of consultation also appear to be potentially important markers of risk.27 Other risk factors associated with an at-risk mental state for psychosis, for example, being a young adult, male, unemployed, with a lower educational level, trauma history, cannabis use, and social isolation,33 might also guide GPs on when to screen for psychotic symptoms.
Even though the transition rates to psychosis are quite low (around 20% in the first year),34 patients with an at-risk mental state for psychosis have an increased risk of developing other poor outcomes.35 If GPs are to feel confident that identifying such patients will improve treatment outcomes, then specific interventions for managing these patients need to be identified.
In the meantime, GPs might be reassured by recent evidence that shows non-specific psychosocial interventions, for example, supportive psychotherapy focusing on social relationships, and assistance with accommodation and monitoring, could also improve patients’ outcomes.10,11 It is possible that some of these interventions may be delivered in primary care, and that shared-care models with input from secondary care services would be beneficial to patients at risk of psychosis. This would be especially relevant for GPs working in areas where secondary care services are not funded to work with these particular patients, and where the duty of care rests with the GP.
In the light of evidence showing that the onset of psychosis can be prevented, it is important that clinicians identify these patients and intervene early. However, the authors are aware that early identification and provision of treatment will be challenging where there are limited resources.
Implications for research and practice
Clinical guidelines recommend that people who may be at risk of developing psychosis should be referred without delay to specialist services.9 However, GPs may not be familiar with this concept, and need more training on how to identify this patient group. Continuity of care is likely to help identify people with an at-risk mental state for psychosis as it improves the therapeutic relationship and may encourage patients to disclose psychotic experiences. Therefore, where possible, GP practices should support continuity of care. At the same time, access to specialist services should be improved, so that once GPs identify potential patients with an at-risk mental state for psychosis, there is a pathway for them to be assessed by specialist services and offered treatment.
Future research should explore how GPs manage patients who are potentially at risk of psychosis, and develop programmes for a better identification and management of this patient group which would take into consideration the limitations of working in resource scarce environments.
