Summary of main findings
This study has systematically examined both nurse (including specialist nurses) and GP perceptions of implementing case finding/screening for depression in DM and CHD patients using structured screening tools. Nurses’ views are particularly important given that this task is predominantly undertaken by practice nurses. We identified several features of the screening which may cause systematic under-detection of depression: namely difficulty incorporating screening into review consultations (including time pressures); replacing individualised with mechanistic assessment; a disconnection for nurses between physical and mental health, and uncertainties about care provision. Far from being a standardised process that encouraged detection of depression, we found participants describing screening as being conducted in a way which increases the chances of negative responses, and as an uncomfortable task when they lacked additional skills to provide immediate support to patients at the time of disclosure.
Strengths and limitations
Participants in this study were self selecting and may therefore represent a set of professionals with research interests or particular interests in depression and depression management. Study 1 participants were recruited to a pilot trial of a nurse led intervention for depression in people with long term conditions (which happened to include QOF case finding as a method of identifying potential patient recruits) and Study 2 participants were recruited to a quality improvement project which included improving early detection. Therefore, their participation is unlikely to have been influenced by wanting to share their concerns about depression screening per se. Both studies recruited a mix of primary care practices, reflecting the general demographic mix of practice populations across Scotland. The focus groups were conducted for different project aims. However, both studies included questions aimed at eliciting views on how depression screening was conducted in everyday practice. The consistency of findings across two independent studies adds to the strength of their validity.
Some of the focus groups in Study 1 involve small numbers of participants and may be better described as joint interviews. Nonetheless, these often ensured that all participants were able to input their views. There may also be limitations in focus groups where participants come from a single general practice and therefore the group dynamic is affected in a way which limits voicing of different opinions, particularly where the power dynamic in a group is unbalanced. Conducting mixed groups including GPs and practice nurses from the same practice can include such a dynamic. However, the practical aspects of running focus groups in primary care needed to be acknowledged (practices normally only offered one day and time that would be suitable for the majority, and most were unable to travel to another surgery during the day or in the evening). Despite all efforts, expected participant numbers were often reduced on the day as other commitments took priority. However, both GPs and nurses appeared open in their self criticism and the limitations they perceived. Two group facilitators in each focus group probed for alternative opinions and ensured that all participants were included in discussion.
Comparison with other studies
Two studies have shown case finding leads to improved depression outcomes in DM, but assessment of depression was carried out separately from the consultation (using diagnostic instruments) rather than within it [29, 30]. Evidence from populations without chronic illness suggests that screening for depression in routine primary care is not effective in improving patient outcomes .
One reason for the poor impact of screening may simply be that those delivering the screening, such as nurses in the UK, have not been adequately trained. Taylor et al. found that variation in depression detection rates was related to experience and training in screening procedures . A UK study of practice nurses found that 82% felt they lacked adequate knowledge and training in dealing with depression. Only a quarter had attended post-qualification mental health training and nurses rated mental health training a lower priority than physical illness in which they had already received more training . The experience of the nurses in our study was one of being given a new task with minimal training. Therefore, the ‘initiation, enrolment and legitimation’ work required for cognitive participation are clearly lacking within the process of transferring this task to practice nurses.
Participants also reported other reasons why screening may be ineffective, in keeping with other studies [17, 18]. Patients with depression work to maintain “face” in front of others  and both practitioners and patients are inclined to normalise distress in the face of the losses that come with a long term illness meaning that the possibility of depressive disorder may be missed or avoided . Drawing on Normalization Process Theory, particularly as applied by Gunn to the implementation of depression care, it is clear there can be a lack of coherence by nurses in understanding depression (and who is, or is not depressed) that is not addressed within the delivery of routinized screening.
The cognitive participation of frontline staff is impacted by this lack of coherence, but more importantly, the use of screening tools seems to reduce opportunities for cognitive participation. The studies by Dowrick et al. and Leydon et al., found that doctors were cautious about the validity and utility of depression tools incentivised within the QOF, and considered their clinical judgement to be more important [17, 18]. Our study findings more strongly indicate that the use of screening tools is viewed as actively restricting the use of clinical judgement and communication skills. Only Mitchell at al have previously included nurse perspectives on depression screening within chronic disease management reviews and they also reported a perceived lack of training, nurse discomfort in asking about mental health, and burden of screening, resulting in screening questions being avoided or not asked in full . There is a clear lack of ‘legitimation’ on the part of nurses in relation to depression work that acts to prevent adoption of screening practices.
Our findings of implementation barriers, including replacing human interaction with mechanistic processes, the disconnect between physical and mental health, and uncertainties about care provision suggest that there are current deficits within the implementation of depression screening which limit the development of coherence and cognitive participation in depression care, particularly for primary care nurses.
Implications for practice and policy and research
As several commentators have discussed, there can be negative consequences associated with incentivised activity [35, 36] and possibly by formalising depression screening without adequate skills, an attempt to improve overall care may actually reduce holistic care. Simultaneously, it is important not to ‘throw the baby out with the bath water’ as there is a danger that the limitations of screening instruments or their poor application in practice might signal the end of case finding in these ‘at risk’ populations.
This research has identified lack of experience, lack of confidence and lack of time in health checks where screening is conducted as the key reasons why primary care nurses are reluctant to engage with mental health issues. Given that practice and specialist nurses’ current role in depression screening is significant, and that this may be extended to other primary healthcare roles (such as medical assistants), it is important that mental health awareness training is made more widely available to improve knowledge and confidence. This might include role play in how to enquire about mental health problems in patients. It could also be supported by mentoring/peer support by mental health nurses or primary care nurses with more experience in mental health.
The current default for nurses is not to engage in helping patients manage often minor mental health problems but to refer the patient to the GP. This may inhibit the potential for roll-out of self-help treatments within the primary care setting, which is a key goal for enhancing self care and for shifting the balance of care within long term condition management . Additional support of the nurses by GPs until they become more comfortable with this work, and clear protocols for dealing with cases or concerns, including protocols for referral to the GP with indicative timescales (e.g. concern about someone who might be very depressed or suicidal) would also help to develop the relational work required for the adoption of screening practices. This support from GPs could also extend to providing and legitimising more time for nurse led health checks, even at the level of a ‘catch up’ slot at intervals during clinics.