|Home | About | Journals | Submit | Contact Us | Français|
Background. The World Health Organization and the World Organization of Family Doctors have called for ‘doable’ and ‘limited’ tasks to integrate mental health into primary care. Little information is provided about tasks GPs can undertake outside of guidelines that suggest to prescribe medication and refer to specialists.
Objectives. The reorder study aimed to gather diverse patient and community perspectives to inform the development of an effective system of depression care.
Method. Five hundred and seventy-six patients completed computer-assisted telephone interviews. Two hundred and seventy-six community stakeholders completed a modified two round Delphi. Responses were analysed to identify tasks and these were synthesised into a conceptual design.
Results. Fifteen core tasks were identified, 5 were agreed upon and a further 10 identified by each group but not agreed upon. Listen, understand and empathize, provide thorough and competent diagnosis and management, follow-up and monitor patients, be accessible and do not rush appointments and provide holistic approach and tailor care to individual needs were agreed on. Other tasks included: develop plans with patients, assess for severity and suicide risk, account for social factors, be well trained in depression care and offer a range of treatment options, appropriate and timely referral, support and reassurance, educate patients about depression, prescribe appropriately and manage medication and be positive and encouraging.
Conclusions. The tasks form the basis of a conceptual design for developing a primary care response to depression. They fit within three domains of care: the relational, competency and systems domains. This illustrates tasks for GPs beyond prescription and referral.
The World Health Organization (WHO) and the World Organization of Family Doctors (WONCA) have called for action to integrate mental health care into primary care. The 10 principles listed by WHO and WONCA include identifying primary care tasks that are ‘limited’ and ‘doable’ (see Box 1).1 Identification of GPs and primary care ‘tasks’ for depression treatment and management is not entirely clear. Guidelines provide some examples, but GPs have been criticized for not following these.2 Guidelines are often inflexible, difficult to tailor to individual needs and do not account for social context.3 Claims have been made that GPs under diagnose depression4 or they over diagnose it5 and when it is detected, it is not treated adequately.6 Recent evidence indicates that misidentification of depression outnumbers missed cases.7,8
|1. Policy and plans need to incorporate primary care for mental health.|
|2. Advocacy is required to shift attitudes and behaviour.|
|3. Adequate training of primary care workers is required.|
|4. Primary care tasks must be limited and doable.|
|5. Specialist mental health professionals and facilities must be available to support primary care.|
|6. Patients must have access to essential psychotropic medications in primary care.|
|7. Integration is a process not an event.|
|8. A mental health service coordinator is crucial.|
|9. Collaboration with other government non-health sectors, non-governmental organizations, village and community health workers and volunteers is required.|
|10. Financial and human resources are needed.|
There is increasing recognition that depression care is complex. Yet, tasks within guidelines usually amount to no more than prescribe antidepressant medications9 and refer to specialists.10 The collaborative care model suggests screening and case management by other non-GP professionals11,12 but this has been costly in the USA and definitive results are awaited from UK and European trials.13 The identification of depression care tasks is complicated by the spectrum of depression seen.14–16 Indeed, in primary care, depression rarely reflects the Diagnostic and Statistical Manual of Mental Disorder (DSM-IV-TR) (DSW) categories8 and is often a combination of depression, stress and life worries.15,17 Debate remains around when the label of depression should be applied and whether ‘depression’ is a useful diagnostic label for general practice.17
The aim of the reorder study (reorganizing care for depression and related disorders in the Australian primary health care setting) was to identify the tasks that GPs and primary care should be equipped with to deliver depression care. During 2006 and 2007, reorder undertook consultations with patients, academics, government, non-government organizations and health professionals. Our aim was to identify diverse perspectives on what GPs and primary care should do for depression. While there have been a number of studies to gather patient and stakeholder views about what is important in depression care,18–20 none have been to the scale of the reorder consultations reported on in this paper.
Results presented in this paper are based on in-depth consultations held with patients in Victoria, Australia, and community stakeholders (henceforth referred to as ‘stakeholders’) from Australia wide including some international representatives. Each consultation phase was conducted separately and is described below.
Patients were identified from an existing longitudinal study, the ‘diamond’ study (the diagnosis, management and outcomes of depression in general practice). Diamond is a cohort of primary care patients who scored ≥ 16 (indicating probable depression) on the Centre for Epidemiological Depression Scale at baseline.21
Community stakeholders were identified using complexity theory and the social determinants of health framework.22–24 Researchers (JG, CD, FG, HH, KH, RK, GB, CJ and EB) participated in a 2-day workshop facilitated by RH to determine the sampling frame. Representatives were sought from federal, state and local governments in Australia, the non-government sector, the health sector and academics in Australia and internationally, including those engaged in primary care and broader research issues like migration, social inequalities and concepts of health and illness. Adequate representation was sought across aspects of the depression experience such as economic issues, work, poverty, disability, access to education, technology, history of local communities, roles of families and friends, relationship issues, the use of health care services, biology of depression, the health care consultation, health care professional attitudes and training, access to services and the organization of health care systems (all parts of the social determinants of health framework). Organizations, groups and individuals were identified by websites and publicly available directories and then contacted by email. Two hundred and ninety three participants were identified for each group of the stakeholder consultation.
Patients completed computer-assisted telephone interviews (CATI) between March 2006 and April 2007 12 months after joining the cohort study. The structured interview guide consisted of 14 sections, 5 sections formed part of the reorder consultation. There were 54 questions, 41 open and 13 closed questions, and three open–ended questions are reported on in this paper (see Box 2). Interview questions included all three terms depression, stress and worries to capture the spectrum of depression seen in general practice. Interviews were 1 hour in length on average and trained interviewers typed responses verbatim and collated these within NVivo 7.25 Patients were asked about individual experiences when seeking GP care.
|Patient CATI Questions|
|(1) What do you think are the most important tasks that GPs can do for people experiencing depression, stress or worries?|
|(2) Thinking of the times you have spoken to a GP about depression, stress or worries, what was good about this?|
|(3) Thinking of the times you have spoken to a GP about depression, stress or worries, what could have been done better?|
|Community Stakeholder modified DELPHI Questions|
|(1) How should general practice/primary care respond to people experiencing depression?|
|(2) What are the barriers for best practice in general practice/primary when faced with people experiencing depression?|
|(3) How would we know if general practice/primary care is meeting the needs of people experiencing depression?|
Questions for stakeholders were piloted by the investigating team and then posed to each stakeholder group.26 A usual Delphi incorporates smaller numbers of experts with three rounds for ranking;27 this consultation was a modified Delphi process because numbers were large and two rounds occurred. Figure 1 outlines the modified Delphi process. Stakeholders were asked about the systems level perspective of what primary care as an organization should do for depression (see Box 2).
Patient responses to the questions were read by members of the multidisciplinary research team (academic general practice, sociology, anthropology, psychology, psychiatry and health promotion). Team discussions about assigning codes and meanings were held (RK, EB, JG, FG, HH and CJ) and 20 responses to individual questions were coded by individual members of the team to ensure consistency. Frequently repeated themes were grouped within agreed upon broader categories. A list of the 10 most mentioned themes was developed for each question.
Content analysis techniques28 and inter-coder consensus were used to analyse Round 1 data. RH and BS led the process of developing a list of the 20 most frequently mentioned responses. Time emerged as such a common issue that it was excluded in Round 2. Participants were informed of this exclusion. The top 20 list was emailed back to participants in random order and they were asked to rank the 10 most important items for each question (with one being most important) (An additional ranking process was undertaken by BS and RH for the modified Delphi, which sought to identify the top 10 responses across each group. This ranking exercise is not reported on within this paper because figures from the stakeholder’s initial selection of their top 10 responses from the list of 20 have been used to conduct a proportional comparison between the two groups).
VP and JG reviewed the top 10 responses provided by patients and stakeholders. Their review found that responses fell within two categories of tasks. These explain what is required in an effective model of care in the primary care system and how we would determine if the model of care and system was effective. These categories reflect the questions participants responded to (see Box 2). The tasks were seen to refer to three important domains of health care: the ‘relational’, ‘competency’ and ‘systems’ domains. Tasks explaining the relational domain referred to the more qualitative and subjective aspects of the consultation, the tacit aspects of the doctor–patient relationship and in particular values like trust, empathy and compassion. Tasks in the competency domain were oriented towards technical aspects of care, the skills, education, competencies and techniques needed to deliver care. The systems domain tasks identified organizational changes and infrastructure support required for effective depression care. Patients and stakeholders provided responses that identified tasks within all three domains across both categories of what is required and how could the system be assessed. The three domains provide the foundations of a conceptual design to inform a primary care response to depression.
Five hundred and seventy-six patients completed CATI Question 1, 475 completed CATI Question 2 and 474 completed CATI Question 3 (see Box 2). The 100 missing responses to Questions 2 and 3 were from patients who had not spoken to a GP about their DSW. Table 1 details patient characteristics.
Three hundred and thirteen stakeholders participated in Round 1 (32.9% response rate) and 276 in Round 2 (31.5% response rate). Table 2 details participant characteristics from Round 2 of the modified Delphi for which the top 10 tasks are reported.
Table 3 presents results from patients and stakeholders combined. It is important to bear in mind that each group volunteered responses: no tasks were pre-established or set out by the research team. Fifteen tasks are identified by patients and stakeholders for effective depression care. Five are agreed upon and an additional ten are mentioned by each group but not agreed upon. Agreed-upon tasks include: ‘listening, understanding and empathy,’ ‘competent and thorough diagnosis and management,’ ‘following up and monitoring of patients,’ ‘funding longer consultations, having more accessibility to GPs, appointments not being rushed’ and ‘taking an holistic approach and tailoring care to individual needs’. Table 4 provides a selection of patient and stakeholder quotes to further illustrate perspectives on the tasks, the improvements that could be made (perceived barriers) and the tasks completed effectively (including ways to measure these).
Nearly half the patients (44.2%; 210/475) suggest that listening, understanding and being empathetic is done well by their GP, and a smaller proportion (13.1%; 62/474) say GPs could improve this. Over half the stakeholders (64.5%; 178/276) identify that ‘overreliance on medication’ is a barrier to good listening. Surveying patients is suggested to measure for listening, being understanding and empathetic. Responses highlight that measures need to incorporate the multifactorial nature of care and complexity of patient experiences rather than trying for a ‘one size fits all’ approach. Combining quantitative surveys and qualitative interviews are ways suggested to capture the patient experience.
Patients and stakeholders also agree that ‘diagnosis and management’ is an important task. Similar proportions of patients (15.8%; 75/475) mention diagnosis and management as a task done well compared to those (14.8%; 70/474) who say it can improve. Nearly half the stakeholders (48.6%; 134/276) link the problem of diagnosis and management with ‘inadequate general practice/primary care competency and training to recognize, assess and treat depression.’ Stakeholders propose ‘measuring diagnosis rates in general practice’ to assess system effectiveness.
A third agreed-upon task is ‘follow-up and monitoring’. Relatively small proportions (between 4% and 6% of patients) comment on this as a task done well or as a task to improve. The ‘poor integration between general practice/primary care and other providers’ is a barrier nominated by a large proportion of the stakeholders (70.7%; 195/276). Almost two-thirds of stakeholders (62.0%; 171/276) suggest that ‘monitoring the quality and duration of follow-up’ could measure effectiveness as could ‘monitoring patient recovery’ (56.2%; 155/276) and ‘measuring functional outcomes like the capacity to function physically, socially and in the community’ (48.6%; 134/276). Functional outcome measures could also be used for ‘assessment of severity and suicide risk’.
Patients and stakeholders agree on the need for longer consultations. More patients (10.1%; 48/474) said that improvements in ‘accessibility and appointments not being rushed’ are needed compared with those (5.5%; 26/475) who said that this is done well. Stakeholders identified the need for funding for longer consultations. Both groups overwhelmingly cite ‘time available’ as a barrier to effective depression care. ‘Measuring patient satisfaction’ could determine improvements in accessibility and time.
The final agreed-upon task is ‘holistic assessment and tailored treatment for individuals’. A moderate proportion of patients (10.5%; 50/475) suggest that this task is done well with a smaller number (6.8%; 32/474) saying this could improve. ‘An overreliance on the medical model’ is a barrier to whole-person care. Stakeholders suggest ‘surveying consumers and carers’ to measure holistic assessment and tailored treatment, as well as ‘surveying patients’.
Results indicate that some additional tasks were nominated more highly than some of the agreed-upon tasks. For example, stakeholders ranked ‘developing a plan with patients’ very highly (76.8%; 212/276), patients supported this but it was not mentioned frequently. Patient results show that ‘appropriate and timely referral’ (30.0%; 173/576) and ‘support and reassurance’ (18.8%; 108/576) are mentioned more often than the agreed-upon task of ‘follow-up and monitoring’. A substantial number of stakeholders (72.1%; 172/276) identify the ‘lack of affordable referral options’ and ‘GP access to referral’ as a barrier to ‘timely and appropriate referral’. Stakeholders mention ‘assessment of severity and suicide risk’, ‘accounting for social factors’, ‘being well trained in depression care’ and ‘offering a range of treatment options’ more often than ‘funding longer consultations’. The final three tasks patients identify are to ‘educate patients about depression’ (10.4%; 60/576), ‘prescribe appropriately and manage medication’ (7.1%; 41/576) and ‘be positive and encouraging’ (4.4%; 25/576). ‘Lack of knowledge about depression within general practice/primary care’ and ‘negative community attitudes’ are barriers to education and positivity.
Figure 2 represents the synthesis of the 15 identified tasks outlined for GPs and primary to respond to depression this includes tasks that could be undertaken to measure effectiveness or demonstrate a reduction in the barriers. These are presented within the two categories of ‘what is required to develop an effective model and system of depression care’ (upper section) and ‘how would we determine if the model of care and system were effective’ (lower section). Figure 2 is a conceptual design to inform a primary care response to depression and related disorders.
The reorder consultations were deliberately broad to capture diverse voices yet the results show simple desires. Patients and stakeholders overwhelming agree on the tasks required for an effective model and system of depression care and time undoubtedly underpins all these tasks. Patients and stakeholders show most agreement on tasks within the relational and competency domains in terms of what is required to develop an effective model of care (see Fig. 2). The doable tasks for GPs within the relational domain are listening, understanding, empathy, support, reassurance, whole-person care, involvement of patients in planning, positivity and encouragement. The doable tasks identified within the competency domain are to undertake a thorough diagnosis and management of depression, develop a plan with patients, undertake an assessment of severity and suicide risk, tailor care to individual needs, take into account social factors, offer a range of treatment options, prescribe and manage medication, educate patients about depression and monitor patients for recovery. These tasks will need to be supported by a system that provides funding for longer consultations and allows GPs to be accessible and not rush appointments. Mechanisms for follow-up and monitoring are important as is providing timely and appropriate referral that is affordable, funding and supporting a range of treatment options and establishing ways to integrate primary care and other providers. This includes providing professional support to general practice.
To determine system effectiveness, a number of tasks within the relational and competency domains could be assessed. For example, surveying and interviewing patients, carers and consumer groups and GPs can provide evidence to determine the effectiveness of tasks undertaken within the relational domain. Identifying if there is a reduction on reliance on medication and medical model, monitoring recovery, diagnosis rates, monitoring functional outcomes and prescribing and monitoring the number of people with depressive symptoms could also assess the effectiveness of the competency domain. Finally, the systems domain could be measured for effectiveness by assessing referral options and rates, services available (particularly in regional areas for Australia), access and affordability, the number of patients with depressive symptoms and duration and quality of follow-up.
The three identified domains and tasks should not be seen in isolation of each other; these are interrelated concepts. For example, relational tasks are effective when GPs are competent communicators, and competent communication is dependent on systems level support through appropriate GP training and education. It is notable that patient and stakeholders agree on the importance of the tacit tasks within the relational domain like listening. However, much of the current focus within health care systems is on the technical and mechanical aspects of competent care to the exclusion of these important tacit dimensions. Establishing how to give equal weight to both the relational and competency domains is fundamental to developing an effective model and system of care as our results demonstrate. This is particularly the case as patients and stakeholders have identified that both the relational tasks like listening are as important as competent and thorough diagnosis in depression care.
One strength of this study is the diversity of viewpoints integrated in the conceptual design including non-medically trained and non-health professionals. While qualitative studies have previously elicited patient perspectives, none that we are aware of have been to the scale of these consultations and few have synthesized both patient and stakeholder perspectives. Given that Delphi response rates are known to be low,27 our response rate to the modified Delphi is reasonable due to our large numbers and diverse participants.
A potential limitation to our analysis is the mixed methods used for data collection for each group. Our view, however, is that the responses from both groups assist to identify some very doable tasks for primary care beyond the limited tasks within current depression care guidelines. Additional limitations include that the patient cohort is predominantly female, from English-speaking backgrounds, well educated and in employment. These demographics undoubtedly play a role in how this group conceive depression and its treatment and management. Future research with people from culturally and linguistically diverse backgrounds and those for whom access to health services may be difficult will be essential to determine the relevance of this conceptual design for their depression care.
Existing literature suggests GP care emphasizes the relational domains (human connection, accessibility, supportiveness and empathy) over technical or mechanical approaches where the focus is on checking technicalities of illness.2 In keeping with Osler’s work, medicine needs science but this cannot replace the Hippocratic tradition of working with people.29 Listening is central to human connection but needs to be also understood as critical to diagnosis and monitoring of patients.20 Empathy too enhances the doctor and patient relationship, yet can improve both patient and doctor satisfaction and diagnostic accuracy.30 Previous research has found that patient confidence in clinical competence increases with an increased perception of receiving quality of care.31 This further highlights the interconnection between the three domains. While concerns have been raised that good listening and regular review and monitoring of patients increases GP workload in a time-limited setting, these tasks are obviously fundamental to patients and stakeholders for mental health care.15 Addressing the barrier of time is critical to depression care patients should not be worried about taking up GP time with unnecessary or insignificant concerns like DSW.32
This study has identified a wider range of doable tasks for GPs than to simply ‘prescribe’ and ‘refer’. There is opportunity now for GPs and policy makers to consider how to operationalize these at an education and practice level. A key challenge is how to incorporate fundamental relational tasks within everyday practice limited by time and for this to be valued by a system heavily weighted in the direction of mechanical competency. While this paper has focussed on principle four of the WHO and WONCA call for integration of mental health into primary care, there are other tasks that patients and stakeholders support, which are covered in the 10 principles.
This study illustrates that patients and stakeholders value the tasks that occur largely within the relational domain, but it is clear that they want this to be combined with technically competent care. The problem in an evidence-based world is that competency often trumps the relational and tacit dimensions because of its measurability and increasingly competency is being presented as a mechanistic deliverable achieved through ticking check boxes. The relational is far more difficult to measure and as a result is undervalued and sidelined in government policy documents. Our greatest challenge for integrating mental health into primary care is to ensure that designs for an effective model and system of care give equal weight to the relational and competency domains. These domains need to be supported by a clearly articulated systems domain suited to each local context. This will require considered attention to identify the appropriate ways to measure all of the tasks required for depression care but particularly the relational tasks.
Funding: Australian Primary Health Care Research Institute, which is supported by a grant from the Australian Government Department of Health and Ageing, to reorder project (2005–2008). National Health and Medical Research Council project numbers (299869) (2004–2006); (454463) (2007–2009); (566511) (2009–2011) to diamond longitudinal study; beyondblue Victoria Centre for Excellence in Depression and Related Disorders to diamond.
Ethical approval: University of Melbourne Human Research Ethics Committee (HREC number: 120406) to reorder; University of Melbourne Human Research Ethics Committee (HREC number: 030613X) to diamond.
Conflict of interest: none.
The reorder study is grateful for the ongoing participation of the participants and investigating team for diamond. The time and valuable contributions of all stakeholders is appreciated, including the research support provided by Ella Butler and Bradley Shrimpton in their analysis and coordination of the modified Delphi. The authors also wish to acknowledge the research support and analyses provided to the reorder team by Senior Research Assistant Maria Potiriadis. The information and opinions contained in this paper do not necessarily reflect the views or policy of the Australian Primary Health Care Research Institute or Department of Health and Ageing.
Contributors: RH, EB, BS, JG, RK and CJ were responsible for identifying stakeholders for consultation. RK oversaw the CATI 2 data collection and analysis. RH and BS were responsible for the coordination of the modified Delphi. HH, CD, FG, KH, CJ and GB assisted in design of questions, data analysis and synthesis with EB, JG and RK. KJA assisted in the analyses of CATI responses. VP and JG developed the conceptual design for a model of depression care suitable for the primary care setting. All listed authors have contributed to the writing, analysis and synthesis of this paper and approved the final version.