Improving Patient Outcomes: Effectively Training Healthcare Staff in Psychological Practice Skills: A Mixed Systematic Literature Review

Training is an important part of modern European healthcare services and is often cited as a way to improve care quality. To date, various training methods have been used to impart skills relevant to psychological practice in a variety of mental health professionals. However, patient outcomes are rarely used in evaluating the effectiveness of the different training methods used, making it difficult to assess true utility. In the present review, we consider methods of training that can effectively impact trainee and patient outcomes. To do so, PubMed, PsycNET, Scopus, CENTRAL and ERIC were searched for studies on training of healthcare staff in psychological practice approaches. In total, 24 studies were identified (16 quantitative and 8 qualitative). For the most part, group, individual, and web-based training was used. A variety of health professionals were trained in skills including ‘communication’, ‘diagnosis’, and ‘referral’ to name but a few. In the majority of studies staff skill level improved. These findings hold implications for the design, implementation, and evaluation of training for mental healthcare staff.


Method Search Strategy
Searches were made of databases PsycNET, Scopus, PubMed, CENTRAL (The Cochrane Library), and ERIC (Institute of Education Studies) for publications between 2009 and 2014. The same search terms and search structures (e.g. nested searches) were used in each database. Where possible, filters appropriate to pre-define inclusion and exclusion criteria were included. Identical search criteria were used to identify both qualitative and quantitative data from studies. By not specifying the research methodology, we minimised the number of appropriate studies missed through inadequate filters (Gorecki et al., 2010).
Added to these results were any relevant documents identified from the resultant studies' citations, which were subject to the same inclusion and exclusion criteria. Study quality was evaluated using the Critical Appraisal Skill Program Qualitative Research Checklist (CASP, 2013) for qualitative studies, and the Mixed Methods Appraisal Tool (Pluye & Hong, 2014) for all other study designs. All reviewers carried out the same search on one of the databases in order to check consistency between searches and quality ratings.

Search Terms
The search terms used in the current review are presented in Table 1, below. Nested searches are indicated by terms within brackets. 'Muscle' was used as an exclusion term to reduce the number of spurious results related to muscle training. Terms related to skills, training and their synonyms were identified from prior reading of related literature (Higgins & Green, 2011) and have been cross-referenced with Index Terms on PsycNET to generate a greater number of relevant results. The search process was evaluated through consultation with senior researchers involved in the project. Searches were also checked to see whether pre-identified papers of relevance were picked up by the strategy. Multiple trials of different search terms and strategies were used in order to get the most appropriate results. These trials suggested using broader search terms over more specific ones in order to pick up the greatest number of relevant studies.
In deciding terms for MHPs, we have been selective in adding common key words likely to target the largest proportion of workers. The sheer number of professionals involved in mental health alone has necessarily meant we have not included some specific MHP terms in order to keep search results manageable. The charity Mind for example lists 37 different professions in mental health (Challis, 2013), and therefore it has not been possible to include them all in this search. Table 2 lists the inclusion and exclusion criteria for studies in this review. These were designed to select highquality studies with both staff-and patient-outcome data. We have used staff-and patient-level outcomes to indicate effectiveness of interventions. A patient outcome was defined as: "a likely consequence of training that is meant to impact patient mental health", where that impact is referenced in the Department of Health's current policies on improving mental health outcomes (Department of Health, 2011aHealth, , 2011bHealth, , 2014Public Health England, 2013). This understanding was based on a general definition of 'outcome' ("Outcome," 2015) with the addition of a specific focus on mental health as informed by government policy. Staff outcomes were defined as: "a likely consequence of training meant to impact staff skill(s)", following our earlier definition of skills.

Results
The search strategy identified over two thousand papers across the five search engines (Table 3). To be included in the analysis a study needed to describe a training process, report a trainee outcome linked to training effectiveness, and include a patient outcome that could potentially be connected to effects of training. Most studies were excluded as they did not to describe the training of MHPs, and others were screened out at later stages because they did not report patient outcomes or detailed descriptions of training procedures. After application of all the in-

Review of Studies
For clarity of description, the selected studies have been grouped into quantitative or qualitative methodologies, followed by a consolidative summary. Following Cochrane guidelines (Higgins & Green, 2011), quantitative papers were reviewed using a narrative approach, while qualitative studies were examined using a form of thematic synthesis (Thomas & Harden, 2008). Finally, the conclusions of these analyses were combined into a united synthesis, following a method of Bayesian conversion (The Joanna Briggs Institute, 2014).
Most studies had a CASP or MMAT quality rating of 50% or higher (see Table 4). Qualitative studies on average scored more highly than quantitative methodologies, which may be because few randomised control trials adequately described their blinding process. A full summary of the review findings from both qualitative, quantitative, and synthesised analyses can be found in the Appendix.

Quantitative Studies
Sample -Sixteen studies described training experiences with MHPs, who were most commonly general practitioners (GPs) or nurses. There was great heterogeneity in outcomes and training methods, which has prevented a meta-analysis. Results are therefore presented as a narrative, and conclusions should be treated tentatively as generalisations to the population cannot be assumed.
The training received by MHPs can be classed into 'group', 'individual', or 'web-based' categories. While there were some differences in how these types of training were applied across studies, in general group training involved a number of MHPs receiving the same training, ranging in length from 1-hour to 4-days. This type of training was often based in interactive approaches, including discussions, debates, and role-plays. Individual training typically involved trainees independently acquiring new skills. This was achieved by a variety of learning resources, including pamphlets, information sheets, or feedback. Finally, web-based training typically taught new skills through an online medium. Web-based training most often involved video-conferencing as a means to give feedback and tutorials to MHPs about their performance.
Group Training -Group training was used to teach communication skills to nurses (Jansink et al., 2013), GPs (Tarn et al., 2013), hospital doctors (Fossli Jensen et al., 2011), and learning disability staff (Mackey & Dodd, 2011). Across all studies, communication improved post intervention, although there is evidence to suggest this is not maintained over time (Jansink et al., 2013) or only to a small degree (Fossli Jensen et al., 2011). These Effectively Training Psychological Practice Skills 540 group training sessions were interactive, as opposed to a traditional lecture format, including role-plays, discussions, etc.
Group training was also used to enhance GPs' diagnostic skills (Bakker et al., 2010;Feder et al., 2011;Shirazi et al., 2013), and again improvements were shown post intervention. Interestingly, interactive group training was superior to education alone when a direct comparison was made (Shirazi et al., 2013), and an exploratory analysis of these studies' effect sizes suggests such interactivity may more effectively teach diagnostic skill than group lectures alone.
Finally, improvements in skill were noted when group training was used to train mental health nurses in supervision skills (White & Winstanley, 2010) GPs in prescription skills (Bregnhøj et al., 2009). As such, it seems that group training may be efficacious in training MHPs in communication, diagnostic, supervision and prescription related skills.
Not all group training improved MHP skill level. When used to train referral skills in GPs (Feder et al., 2011;Lester et al., 2009), only one study detected an increase in ability (Feder et al., 2011). In the other study, the authors noted that educating GPs alone may be insufficient to make a difference in their skill level. They argue that as primary care is just one part of the care system; a whole-systems approach is required (Lester et al., 2009). Also, this study did not explicitly use interactive groups as a teaching method, unlike the study that found a significant difference.
Three other studies that found mixed results with non-interactive groups looked at training therapists ( The following comments on patient outcomes must be treated cautiously. It is not within the scope of this study to draw conclusions on the effectiveness of particular interventions on particular outcomes, as the number of studies is too small and too varied to facilitate strong conclusions. Given this, in general, patients seen by MHPs with skills obtained through group training were more likely to be referred for treatment (Bakker et al., 2010;Feder et al., 2011;Lester et al., 2009;Shirazi et  Although about one third of the above studies mixed group training with additional individual work, there is not enough evidence to suggest this has a specific impact on either trainee or patient outcomes. Anecdotally, there is evidence to suggest interactive groups may enhance the acquisition of some skills in trainees. However, no such tendency was found regarding patient outcomes. Individual Training -Studies using individual training showed less methodological variation than the group interventions. While group training would often involve a mixture of role-plays, discussions, presentations, etc.-with and without additional individual learning-the two studies using purely individual training tended to focus on only one or two approaches. This is to be expected, particularly as the above-mentioned activities require more than one person; individual learning is naturally limited here. Penner et al. (2013) used written information as an aide-mémoire to convey a sense of 'being part of a team', while Guldin and colleagues (2013) supplemented their leaflets on recognising complicated grief with patient risk assessments for the same. Neither study was able to sustain their trainee effects at follow-up, suggesting written material alone is not sufficient to produce significant change. However, patients' trust in their doctor and adherence to medical advice was maintained at 16 weeks follow-up (Penner et al., 2013), as were small reductions approaching significance in depression and grief scores after one year (Guldin et al., 2013). This may indicate the interesting possibility that, for certain skills, MHP learning does not need to be sustained in order to produce long-standing patient changes. This may be more likely under conditions where the MHP has a time-limited role in the early stages of the care pathway. For example, a GP trained to give the most appropriate referral at the first signs of mental distress can greatly impact future care and course of the illness (Ross & Hardy, 1999). Further research in this area would be interesting, especially if such change is found more generally, such as by eliminating other potential factors like patient group.
Web-Enabled Training -As opposed to the above-mentioned group and individual training methods, in which the majority of learning aids are based in physical reality, a study by van Weert et al.'s (2011) used digital methods to create and host learning material. They used web-enabled video to record and upload individual nurses' consultations with cancer patients, which the nurses could watch at home or in the hospital. Individual feedback on communication skills was provided, and supplemented with a reflection task that encouraged the trainees to review their own material. This was linked to a small, statistically significant improvement in nurses' communication skills and patients' recall of consultation information compared to controls. However, an exploratory comparison with effect sizes from other studies targeting communication skills would suggest web-enabled training may be less effective at teaching this particular skill.

Qualitative Studies
Data was extracted using a meta-ethnographic approach, wherein summary terms and concepts were identified from the authors' original findings and listed (Campbell et al., 2003). These were then used as codes and organised into themes and sub-themes, following a method of thematic synthesis (Thomas & Harden, 2008). These themes were re-examined and regrouped (as appropriate) in order to best represent the meanings in the code groups in the context of training methodology. Three other review authors appraised these to agree a final version, resulting in nine themes related to effective training.
Sample -A total of eight qualitative studies were selected for inclusion in the present review. Similar to the quantitative studies above, training involved: 'group' (Aiarzaguena et al., 2009;McLoughlin, 2010;Nilsen, 2011), Effectively Training Psychological Practice Skills 542 Nilsen, 2011) methods. In addition, training by means of 'peer presentations' was also identified (Suldo, Friedrich, & Michalowski, 2010).

Theme 1: Translation of Policy Into Practice -
The challenges of applying what has been learnt within an educational environment to the healthcare setting was made clear in the research by Monrouxe and Rees (2012).
Here, medical students reported that qualified staff members would often demonstrate actions inconsistent with their teaching, resulting in students experiencing contradictory formal and informal learning experiences. This reflects the quantitative work of Lester and colleagues (2009)  experience of the environment. As knowledge and trust in the trainee's abilities increased, they would gradually be given more responsibility by their supervisor. This exposed them to new and more complex interactions within the healthcare environment, allowing more advanced skill development.

Theme 4: Understanding Individual Patients' Needs -
The needs of patients should also be a primary concern when training is provided. For example, Aiarzaguena and colleagues (2009)  the consultation stage helped GPs feel closer to "difficult" patient groups, and also promoted better understanding of patient needs. Nilsen (2011) found that training within a multi-disciplinary team (MDT) promoted faster and more effective diagnosis. This meant that all professionals were able to contribute their expertise in considering different patient needs, therefore optimally delivering high-quality care. It was also found that specialists benefited from the collaboration, as it allowed for continuity both within the team and for the patient.
Theme 5: Patient Safety -In terms of the patient experience, Monrouxe and Rees (2012) also commented on how patients' dignity and safety can be compromised during student training. They note this compromise was partly due to the amount of unsupervised student involvement within patient care. In their study, it was accepted that students were bound to make mistakes, yet they were also asked to practice beyond their level of competency.
It was also found that consent was often not gained from patients for treatment by students, which pressurised students to take full responsibility for the care provided.
Finding the balance between learning through experience and reducing the potential harm for patients in healthcare settings should be an important concern, yet Monrouxe and Rees (2012)  Clear guidelines used within training for GPs were found to increase the application of skills to a variety of patients' needs using a structured intervention technique (Aiarzaguena et al., 2009). This demonstrated the importance of clarification for skill-sets.
Co-ordination between the multidisciplinary team meant that learning could be on-going within the working environment; for example, when reviewing the patients' needs on a daily basis using video-conferencing (Nilsen, 2011).
Fast and effective communication was a central feature of the qualitative research on learning within the workplace.
Communication increased the speed and efficiency of patient diagnosis and planning of care (Collin et al., 2012).
Finally, when applied to a context of constant change, collaboration between the qualified member of staff, student and patient reduced the discrepancies within the healthcare training environment and maintained that the patients' needs were the priority of the service provided (Collin et al., 2012).

Structured and Unstructured
Training -Delivery of training that did not follow a predetermined structure commonly took the form of learning within the 'here and now'; getting involved with all processes of patient care, often in collaboration with a senior member of staff (de Feijter et al., 2011). It appeared less common that training was provided within a pre-planned, structured manner for students. Aiarzaguena et al. (2009)  Effectively Training Psychological Practice Skills 544 patients with psychological symptoms. The importance of relevant and regular training to improve insight for more challenging symptoms was highlighted by GP's that completed the intervention.
Individual Learning -Individual learning was found to increase self-efficacy in cognitive, practical and affective learning dimensions, when students engaged in an international service-learning project for nursing (Amerson, 2012). Here, individual study, e.g. through reading assignments, improved cognitive understanding of cultural differences, which was complemented with the individual's experience of working in another country; "seeing makes it real" (Amerson, 2012, p. 12). These cultural insights translated into better communication skills, and thereby greater practical understanding of patient requirements. This improved patient outcomes as graduates were able to recognise culturally different responses to pain and were able to care for each individual's specific needs more accurately.
Amerson (2012) also found that learning independently during international trips was effective training for allowing the learner to develop 'a desire and passion to fulfil their job role to the best of their ability'. It was concluded in this study that exposing MHPs to a multitude of variables, such as: the treatment process, the form of treatment and the challenges of patient care-including exposure to inconsistencies within the practice-increased their competency and optimised learning interactions.
Collin and colleagues (2012) also reported on the use of individual learning involving all staff members working within an emergency unit for acute care. Errors occurring in the treatment chain were connected to inconsistent practice when working individually within a multi-professional collaboration process for patient care. For example, single staff roles overlapped without them even knowing about each other. However, within the study this was interpreted as learning through experience when mishaps were highlighted to employees, enabling them to learn from others' experiences and mistakes. Individual-level learning was found to be common within the emergency work setting for patient care, as healthcare roles were represented as single, separate processes within the treatment chain for emergency practice.
Group Learning -As in the quantitative studies, training within a group was found to be the most common form of learning for MHPs, used with both qualified clinicians (Aiarzaguena et al., 2009;McLoughlin, 2010) and students (Amerson, 2012;de Feijter et al., 2011;Monrouxe & Rees, 2012).
Collaboration lead to higher quality of care for patients by improving the communication between multi-disciplinary teams, allowing for MHP's to have better knowledge of patients.
An important feature of group learning was collaboration (Suldo, Friedrich, & Michalowski, 2010), which was linked to improvement in communication skills (Collin et al., 2012). This was most pronounced when working within a group at an early stage of contact with the patient, leading to a reduction in discrepancies when delivering practical instructions for patient care (Collin et al., 2012). Inter-professional collaboration facilitated learning as a natural part of care practices, whereby being a member of a multi-disciplinary team allowed for on-going learning within the workplace (Nilsen, 2011).
Video-Conferencing -Videoconferencing between general practitioners and hospitals in Norway was found to be most successful for efficient, collaborative work, with the ability to successfully adapt to the patient's condition as it changed over time (Nilsen, 2011). By using videoconferencing as a communicative tool, healthcare professionals were able to exchange information in order to overcome any knowledge gaps relating to patients' needs. Within this study, the regular use of videoconferencing meant a continuous follow-up of the patient's progress, and allowing staff members to mutually share their expertise. Nilsen (2011) described this as separate activity systems coming together; the patient acts as a shared object uniting the healthcare professionals, providing a common activity and allowing for multi-disciplinary learning.
The selected patient case used to describe this training technique was discussed by the team over several days, allowing for a repertoire of knowledge to be developed. This improved the outcome for the patient, as their complex needs could be discussed as time unfolded allowing for continuity and prompt decision making with their treatment.
Peer Presentations -When given by MHPs, teaching through colleague presentations enhanced skill-sets of individuals within the team who did not specialise in mental health (Suldo, Friedrich, & Michalowski, 2010). It was found that solutions were generated before problems occurred, acting as a strong preventative method for risks.
This would also improve the service-user access by providing interventions in more frequently used locations, such as at schools and within the workplace.
This finding was consistent with McLoughlin (2010), who found that when staff came together, it was possible to be constructive and creative, rather than work in a reactive way. In doing so, staff were able to listen to each other and take the time required to work through a complex situation (McLoughlin, 2010, p. 238).

Synthesis
The data thus far has been presented as two separate analyses of quantitative and qualitative information. To unite these into a single summary, a synthesis is required. This has been achieved using a Bayesian-style conversion of quantitative-into-qualitative information, based on the method suggested by The Joanna Briggs Institute (2014).
First, the results of the quantitative analysis were written as descriptive-i.e. qualitative-statements. Second, these statements were compared to the conclusions of the qualitative analysis for topical similarity. Where similarity was identified, the original texts were referred to to determine the appropriate context of the statements (The Joanna Briggs Institute, 2014). Finally, a single statement reflective of both analyses was produced. These have been phrased in such a way as to give indications for practice, based on the available evidence.
Effectively Training in Psychological Practice Skills -The points below summarise the main findings of the synthesised material from the qualitative and quantitative studies. They have been divided into different methods of training, which were most amenable to a combined analysis.

Individual Training:
a. Individual training offers fewer opportunities for feedback, increasing the chance of inconsistencies in practice.
b. It is therefore more effective when combined with feedback from team members in related disciplines, particularly in MDTs.
c. Professional change from this type of training alone is unlikely to be permanent with a short period of study, but it can foster a passion for work that may motivate future efforts.
d. This motivation can contribute to improving patient outcomes that are based on single professional contacts early in the care pathway, such as trust in a doctor after one consultation. e. As such, this training can be more efficient where short-term skill improvement is wanted, and can be expressed as written material.

Group Training:
a. Interacting within a group naturally, through exchanges, aids the development of communication skills and fosters learning, but requires an initial ability to collaborate effectively.
b. It is therefore most effective when a range of disciplines are represented, such as in MDTs, as this brings the biggest pool of relevant knowledge.
c. Outcomes that are affected by the availability of professional knowledge, such as referrals, diagnoses, care pathways, and (as a result) mental health symptoms, are thereby more effectively taught by such groups.
d. By the same token, group training has relatively little effect on outcomes like patient satisfaction, which do not depend as heavily on breadth of knowledge.
e. As more information and feedback is shared, practice becomes more consistent and knowledge-based skills improve with time.
3. Web-enabled Video Training: a. As with group training, web-enabled video allows professionals to share knowledge and improve communication skill-related outcomes through interaction.
b. However, digitally-facilitated interaction requires several different communication practices specific to that medium, so is therefore not as effective as physically-present groups in enhancing face-to-face interaction abilities.
c. This method of training may then be more effective in training groups of remote individuals in communication-and knowledge-based skills.

Discussion
Group, individual, and web-based training were used across quantitative and qualitative studies. Peer presentation was also used to train MHPs in the qualitative studies. Overall, each type of training positively impacted skills

Implications for Clinical Practice
The studies reviewed suggest most methods of training are effective at improving skills in mental health professions, although this effect can be enhanced when type of skill and the desired patient outcome are taken into consideration. In general, it is suggested communication skills are more likely to improve with interactive, face-to-face groups, and this is more likely to improve patient outcomes that can be linked to availability of professional knowledge. Membership of an MDT is a reoccurring example of how this can be achieved even in an unstructured training environment, given appropriate collaboration.
Individual learning can be a useful way of making short-term skill changes, but is unlikely to create lasting change in trainees without significant time investment. Interestingly, the research suggests such change may not be necessary to positively affect patients when made early on in the care pathway. This further highlights the need for including patient outcomes when evaluating the effectiveness of learning interventions. The potential limitations of individual training should be considered in relation to the increase in 'e-learning' in many NHS trusts (e.g. Gray, Plaice, & Hadley, 2009), which is a short-term experience and generally a solitary activity.
Lastly, web-enabled video learning can be effective in teaching communication-based outcomes to remote participants when part of a group or as reflection, although this may not be as effective as groups occupying the same physical space.

Research Implications
This review has contributed to the literature by examining what training can be effective at improving both patient and trainee outcomes in a wide range of studies. The large attrition rate from the initial literature search reflects the fact that very few studies report patient outcomes, which should be an important consideration when researching effectiveness in healthcare. The present review should be considered as a first step toward understanding MHP training in psychological skills, and it is important that future work considers the heterogeneity that existed across studies in the present review. Type and content of training, trainee profession, skills, length, type of patients and patient outcomes were varied throughout, and although this heterogeneity provides a much-needed broad overview, it also limits the direct comparison of study findings. As such, more targeted and controlled studies would provide more nuanced understandings.

Limitations of the Review
The strength of the conclusions of this review has been severely curtailed by the heterogeneity present within the studies, and they should be treated as tentative first steps towards a more solid evidence base. The lack of a sufficient number of controlled studies has prevented a meta-analysis and thus a more definitive conclusion on what variables form the active ingredient for effective learning. For example, many studies using group training combined this with individual learning activities, but it has not been possible to tease out potential effects. This in itself should be the subject of a much larger publication. The generalisability of the findings is also limited, as it is unknown to what extent the particular type of MHP may influence the effectiveness of a training method, or indeed the specific type of patient.
The wider application of some training methods may be difficult in the average mental health care setting, limiting their value beyond this paper. For example, Amerson (2012)  Effectively Training Psychological Practice Skills 548 change Programme has been encouraging similar activities since 1981 (European Hospital and Healthcare Federation, 2013), it is not common. In these cases, as with most training processes, adaptations would need to be made to accommodate individual services practices.

Conclusion
In sum, this review can be said to provide an important platform for skills training and development among MHPs.
The work shows a variety of training methods that may be used. It also points to important avenues for further and future considerations. Overall, it is hoped the this review will serve as an integral first step in the improvement of much needed skills training in the area of mental health, which, ultimately, is intended to benefit the wide range of individuals who engage with mental health services across Europe.

Competing Interests
Pavlo Kanellakis is one of the Associate Editors of Europe's Journal of Psychology but played no editorial role for this particular article.

Funding
The authors wish to thank the Mental Health Institute Local Education and Training Council in the West Midlands for their support and funding towards this research, following the Up-skilling the Workforce in Psychological Practice Skills project, hosted by Coventry and Warwickshire NHS Partnership Trust.
b. consider brief individual training if short-term staff change is required. This can be delivered through relatively cost-effective means, such as handouts.
c. consider group training for more sustained staff change. Where staff are likely to have prolonged contact with patients, such as through therapy, include long follow-up periods of training, such as every six months over a one year period.

About the Authors
Katherine Garzonis is training for a doctorate in Clinical Psychology at University College London. Her interests include the transmission of learning, improving clinical decision-making, and the relationship between anxiety and mental health.
Eryn Mann is currently an Assistant Psychologist for a charity providing specialist NHS healthcare. Her research interests include investigating the impact of vicarious traumatisation and she is motivated to reduce the stigma of mental health by increasing awareness.
Aleksandra Wyrzykowska is a psychology undergraduate at Coventry University. Her interests include the relationship between social media and underlying mental health and external validation.
Pavlo Kanellakis was the Principle Investigator for first phase of the Up-skilling the Workforce in Psychological Practice Skills project. His main interests are currently in implementation and translational research, particularly the dissemination and application of research findings in the 'real' world, and training health professionals.