Treatment of depression in the first primary care consultation: A qualitative study

ABSTRACT

Introduction:

The first primary care consultation for patients with depression can have long-term consequences for patients, but little is known about treatment decisions at this visit. The aim of this study was to explore the treatment of patients presenting in primary care with a new episode of depression and the drivers behind GPs’ treatment decisions at the initial consultation.

Materials and Methods:

A random sample of GPs in Auckland was invited to participate. A qualitative study was undertaken using semi-structured interviews. Interview transcripts were analyzed using a general inductive approach.

Results:

Twenty-one GPs were interviewed. We identified three themes as drivers of treatment decisions at the first visit: characteristics of GPs, characteristics of patients, and characteristics of treatment options. Drivers for prescribing were severe depression and time constraints. A driver for non-pharmacological treatment was a strong doctor–patient relationship. Limited time, skill, and training were associated with low confidence using talking therapies. Access to counseling was reported as poor. There was a very wide range of approaches taken. GPs described preferring antidepressants less and talking therapies more with Māori patients. Behavioral activation was used least despite its ease of use and it being one of the most effective treatments for depression.

Conclusion:

Treatment of depression at the first visit varies widely between practitioners. GPs report multiple barriers to the provision of talking therapies. A move to a more standardized approach may lead to more equitable care. This is the first study to report findings about the initial primary care consultation for depression.

Introduction

Mental disorders, such as major depressive disorder (MDD), are a significant health issue in primary care. Over a third of adults attending primary care are likely to have met the criteria for a DSM-IV diagnosis within the past year.[1] The wide-ranging impact of poor psychological health must be considered, including the mind–body interface and the complexities that comorbidity raises[2] regarding management and the burden of disease.[3] The first consultation where a patient presents with symptoms of depression is important as decisions made at this time may have long-term consequences.[4] Early prescribing of antidepressants without discussing other options such as behavioral activation or problem-solving therapy risks creating an impression that pharmacological therapy is the optimum or only treatment. This may lead to difficulty in stopping antidepressants and potentially long-term use of unnecessary medication.[5]

A recent systematic review of treatments for depression in primary care reported that psychotherapies have comparable effects to pharmacotherapies and appear to be as effective in treating depression when compared to waitlist controls and treatment as usual.[6] Combined treatment is signaled as potentially being better than either psychotherapy or pharmacotherapy alone.[6]

Antidepressant prescribing seems to be increasing.[7] One study reported a 68% increase in the number of annually dispensed antidepressants for people aged 1–24 years in New Zealand from 2007/2008 to 2015/2016.[8] Similarly, antidepressant use is increasing at an average of 10% per year in England.[9] Given the high prevalence of depression and the possible downstream effects of initial treatment decisions, research in this area is crucial. However, we are not aware of empirical evidence on the first visit to primary care. The purposes of this study were to investigate how GPs treat patients presenting with a first or new episode of depression and to explore the drivers for treatment decisions, including ethnicity, GPs confidence with talking therapies, and antidepressant prescribing.

Materials and Methods

This study was conducted using a general inductive approach.[10] As the management of depression in the first consultation is under-researched, an exploratory qualitative study is appropriate to identify key themes. It has been described using the Standards for Reporting Qualitative Research (SRQR) guidelines.[11]

Participants were eligible if they were currently practicing GPs in the Auckland region, selected from a directory of all general practices in the region, and publicly available through a website (healthpoint.co.nz). A random sample of 45 GPs was specified to approximate the spectrum of GPs in practice in Auckland. Invitations to participate were made via email, followed by phone calls. These were made by a research assistant (LB) and by the Head of Department of General Practice (BA), both of whom had no relationship with participants. Participating GPs were given £UK35 as thanks for their participation, given the nature of private practice in New Zealand.

Semi-structured interviews were undertaken by an independent interviewer (LB). Two practicing GPs (BA and VM) drafted the questions, informed by findings from earlier research.[4] A mixture of closed and open-ended questions was used, with participants able to elaborate on key points at the interviewer’s discretion. Interviews were transcribed by hand contemporaneously and later typed out, considered by the researchers to be the optimal approach for these short, relatively structured conversations, and within the project budget. We used the general inductive approach for qualitative analysis of the questionnaire transcripts.[10] This involves a deductive approach with research objectives being considered during data analysis, alongside an inductive approach with findings arising from the data. Transcripts were repeatedly read until researchers gained familiarity with their content.

The researcher (DL) grouped segments of text that conveyed similar ideas into emerging codes using NVivo 12; then, the researchers (DL and BA) grouped related codes independently into categories. Overlapping and redundant categories were identified and merged or removed; then, the remaining categories were grouped into themes. Collaborators (DL, EW, and BA) agreed on the final themes and categories as these three researchers encompassed a broad range of perspectives of primary care, with their different levels of experience. An aspect of the qualitative dataset was analyzed similarly to quantitative data, an approach that is appropriate for reporting patterns such as frequencies. We recorded responses to highly structured questions where limited answer options were provided, created a spreadsheet to determine the distribution of responses, and reported the results in a table.

Ethical approval was from the University of Auckland Human Participants Ethics committee (UAHPEC) Ref # 021192 May 10, 2018.

Results

A random sample of 45 GPs was obtained, of whom six were retired or not working in standard general practice. Out of the remaining 39, 21 agreed to participate, giving a 54% (21/39) response rate. Of those who declined, 66% (12/18) were female and 56% (10/18) were New Zealand graduates, with an average number of 25 years since graduation. Of those who agreed to participate, 43% (9/21) were female and 52% (11/21) were New Zealand graduates, with a range of 5–47 years since graduation, giving an average of 26.7 years. Therefore, the actual participants are similar to the targeted sample, and the level of participation is unlikely to affect the findings.

The three top-level categories (themes) and nine sub-level (categories) that were developed are shown in Table 1, with the categories then described in-depth below.

Table 1.

Themes, categories, and descriptions

Theme	Category	Description
Characteristics of GPs	Confidence using talking therapy	GPs confidence using talking therapies as individuals and as a collective.
	Limitations of GPs	Factors that limit a GP’s ability to provide optimal treatment, such as time constraints or limited training.
	Relationship with patient	The influence of the GP’s relationship with the patient on the treatment of that patient.
Characteristics of patients	Severity	The influence of the severity of the patient’s depression on the treatment of that patient and how severity is measured by GPs.
	Ethnicity	How GPs consider the ethnicity of the patient to affect, or not affect, their treatment of that patient.
	Patient preferences	The influence of patients preferences and expectations of their treatment
Characteristics of treatment options	Effectiveness	Perceived effectiveness of the treatment option and what influences this
	Ease of delivery	Ease of GPs administering the treatment including resources.
	Accessibility	Availability, cost, funding options, and waiting time associated with treatment options.

Confidence using talking therapies

This category included the participants’ own level of confidence in this area along with their perceptions of “the profession’s” level of comfort with the use of talking therapies.

The majority of respondents (15/21) felt between “very” and “somewhat” confident with personally using talking as a therapy for their patients, with the acknowledgment that it was common practice:

“Yes, it is a big part of their job.” (female, 30 years, NZ).

The remaining six respondents did not feel confident using this approach, including some experienced GPs:

“It is my weak area, so I am developing my skills as much as possible by attending conferences.” (female, 8 years, NZ).

The majority of respondents (13/21) believed that GPs as a whole were not confident using talking therapies. A further five found it difficult to comment because they did not know what other GPs were like, and one GP felt it would not be appropriate for some GPs to offer talking therapy.

Only two respondents felt that GPs were confident in this area:

“I think most people would be confident that they are talking to people. In terms of saying a specific intervention, I am not sure what levels of confidence would be.” (male, 16 years, NZ)

Limitations of GPs

Low confidence using talking therapy was linked to limitations that GPs face, such as time constraints and lack of appropriate skills and training:

“Don’t have time, don’t have training.” (male, 29 years, overseas).

When asked if GPs prescribe too quickly, respondents identified limited time as a possible driver of this behavior:

“Yes I don’t want to judge them but I think people overprescribe because the system for most doctors does not allow for time to talk for longer with patients.” (male, 46 years, overseas).

While the GPs felt that they were limited, they did not view specialists as facing the same restrictions:

“I refer patients for CBT as I am not trained to do it. I do not regard myself as a therapist.” (female, 32 years, NZ).

Relationship with patient/rapport

The respondents’ confidence in delivering non-pharmacological treatments was associated with the rapport they had with their patients:

“A benefit of working in general practice is the relationship you have with patients. Most of them I have known for a long time, some for 20 years, and thus relationships are meaningful, powerful, and helpful.” (male, 31 years, NZ).

Severity

Respondents were more likely to prescribe for patients with severe depression:

“… more severe, then more likely to prescribe.” (male, 9 years, NZ).

A reason for this was a reduced effect in patients with moderate to mild symptoms:

“In most mild to moderate cases, it is not proven to have a huge effect.” (female, 11 years, overseas).

There were a variety of responses regarding treatment for patients who do not meet the MDD criteria. One GP indicated that they would not start treatment for patients who do not meet the MDD criteria:

“We discuss the level of depression and inform them that they are depressed but not to the stage where we need to start treatment or counseling. We tell them some coping strategies, for example, stress management, and ring them after 6 weeks and see how they are doing.” (female, 35 years, overseas).

Another experienced GP had an opposing outlook, viewing pharmacological treatment as important for most levels of severity.

“Most of the people who I see aren’t able to work through it on their own even if mild, and moderate need an SSRI to work their way up.” (female, 35 years, overseas).

Some respondents said they had no standardized approach, and tailored management according to each patient and situation.

Ethnicity

There was a mixed response from GPs about ethnicity. Many respondents indicated that they would not treat people differently based on ethnicity.

“There are diversities within ethnic groups that is far greater than the diversity between ethnicities.” (female, 35 years overseas).

Respondents also recognized that ethnicity might influence a patient’s own treatment preference:

“It may affect their preference of what they would want.” (female, 35 years, overseas).

Patient preferences

Responses indicated that patient preference was taken into consideration when treating:

“… depends on patient preference and making an informed decision together as to what they wanted to do.” (female, 11 years, overseas).

Effectiveness

GPs viewed antidepressants as effective for some patients, but not all. Effectiveness was frequently linked to the severity of the patient’s depression (theme above).

A few respondents framed antidepressants as a short-term solution:

“I would say that they are likely to be effective but don’t expect a miracle. They do not help with long-term issues or on a spiritual level.” (male, 29 years, overseas).

Ease of delivery

Time constraints were noted as limiting the delivery of behavioral activation:

“We do not have that much time so will tell them little things such as breathing exercises, meditation, but not extensively because of time constraint. We usually refer to counseling.” (female, 35 years, overseas).

Accessibility

Counseling was deemed difficult to access. Reasons for this were the costs of private treatment and the waiting time associated with fully-funded public treatment.

“if it is newly diagnosed depression we have to wait 6 weeks before referring to the counseling.” (female, 35 years, overseas).

For CBT, this waiting period contributed to patients preferring medication-based treatment:

“Some want to start on something straight away as they don’t want to wait for CBT.” (male, 13 years, NZ).

The cost of psychological treatment for patients was a barrier to access, which some GPs were able to overcome by referral to funding programs:

“I try to explain that psychology works as well as the medication and I think cost is a huge barrier, hence referring for the depression program.” (male, 13 years, NZ).

Pharmacological treatment was viewed as the cheaper treatment option:

“It is the cheapest option, and in some ways, if they aren’t going to do talk therapy, it is the only option, unless they qualify for programs and they don’t happen immediately.” (male, 46 years, overseas).

Characteristics of treatment options

Seventeen of 21 participants indicated that GPs tended to prescribe antidepressants at the first visit 50% of the time or less frequently. A range of treatments were offered [Table 2], with medication and physical activity being the most common. Many respondents used methods not listed in the questionnaire prompts such as websites, apps, books, CBT, online tools, (e.g., for mindfulness), alcohol and drug advice, seeking support from family and friends, and engaging more in art or music.

Table 2.

Treatment options offered by respondents for MDD at the first visit

Treatment option	Respondents who offer option	Respondents who sometimes offer option	Respondents who do not offer option
Physical activity	17	1	3
Medication	17	1	3
Sleep advice	16	1	4
Pleasurable activities	15	0	6
Time off work	13	7	1
Problem-solving	10	2	9
Behavioral activation	10	0	11

Discussion

This study is novel as we are not aware of other research exploring treatment issues for the first consultation for depression in primary care. Key findings are that GPs offered a wide range of treatment options at the first visit for depression. Time off work, medication, physical activity, and sleep advice were the most highly offered options, while behavioral activation and problem-solving were the least offered. We identified three major themes each with three categories that drove the treatment offered.

With regards to GP characteristics, most respondents reported confidence using talking therapies as high for themselves but low for GPs overall, while some GPs found it difficult to comment on the behaviors of other GPs. Good rapport was helpful for the management of depression, especially for providing talking therapies. Limitations GPs face, such as time constraints and limited skills, were identified as barriers to the provision of talking therapies. Respondents viewed specialists as not facing these same limitations, which drove GPs to refer their patients for talking therapies, instead of undertaking this themselves.

Many respondents reported considering patient preferences in treatment decisions. There were mixed views on the influence of patient ethnicity on treatment decisions. However, GPs drew distinctions about Māori patients, favoring psychological treatment over pharmacological treatment for Māori. The severity of the patient’s depression was used as a guide for prescribing. As antidepressants are more effective in the treatment of severe depression than mild or moderate depression, most respondents would offer medication if the depression was severe. However, there were mixed views on managing patients who do not meet the MDD criteria. Prescribing was viewed as quicker, easier, and cheaper than talking therapies, which were associated with barriers of cost and waiting times. While these factors drove antidepressant prescribing, GPs also raised issues that medication caused for patients such as side effects and delayed onset of action.

Random sampling ensured a diverse range of responses, including a range of experience levels of both New Zealand and overseas-trained clinicians. The sample of GPs was fairly representative of GPs in New Zealand in terms of years of experience, gender, and location of training. Although this random sample is representative of Auckland GPs, results may not be generalizable to GPs practicing in a rural environment.

Because the trustworthiness of the findings cannot be assessed by comparison with previous research, triangulation within this study could have been used, such as observation of consultations. Stakeholder checks on the transcripts or on data interpretation could also have been done to enhance credibility. However, a self-critical approach to data analysis was undertaken with multiple author discussions, and the sampling strategy enabled different participant voices to be heard, which enhanced authenticity.

A further limitation of this study was that our findings were based on self-reporting by GPs, which may not reflect actual behavior. The patient perspective was not explored, and a recording device could have improved data collection. However, reliability was strengthened by the use of semi-structured questionnaires, enabling a consistent approach during interviews. The wording of the questionnaire may also have limited our findings. For example, the use of the term “MDD” may have led respondents to view depression through a biomedical lens rather than a social one.

It is known there are low follow-up rates following a depression diagnosis,[12] so the first consultation may be the only opportunity for GPs and patients to discuss interventions. Differential treatment based on severity, a strategy used by many respondents in the present study, aligns with multiple guidelines on the management of depression in primary care.[13] A study of Māori and non-Māori patients found that Māori patients have a preference against antidepressant medication.[14] This was consistent with views expressed in the current study.

The findings of this study are consistent with previous data describing limited resources and time constraints in primary care as drivers for prescribing antidepressants instead of providing psychological treatments.[15,16] Time constraints are also cited as a barrier to optimal treatment in studies from New Zealand, the UK, and Canada.[16,17,18] Cost is commonly seen as a barrier to counseling.[17] The perceived lower cost of prescribing compared to talking therapies has been noted.[16] Recent developments in online therapy, which has comparable efficacy to in-person therapy,[6] may provide a cost-effective pathway to increased access. However, there is evidence increasing access to therapy has little influence on antidepressant prescribing rates, with a large-scale intervention to increase access to psychological therapies in England failing to curb increases in antidepressant prescribing.[19]

Many patients with depression improve without treatment, and some patients do not respond to treatment at all,[6,20] necessitating caution about prescribing too quickly. In this study, behavioral activation was not offered frequently despite being one of the most effective treatments for depression, with a number needed to treat of 2.5. In contrast, numbers needed to treat with antidepressants are 4 for severe depression and 16 for mild to moderate.[21,22]

Regarding implications for practice, provider training in non-pharmacological strategies may be useful. However, there is little evidence that efforts focused on provider training result in positive patient outcomes,[19] and education on depression detection and treatment alone does not improve outcomes.[13]

Future researchers can gain a more accurate depiction of treatment at the first consultation by using an observational approach incorporating audio-visual recording of consultations to capture non-verbal communication. Recording-assisted recall interviews of practitioners and patients may also aid in understanding conversations from both perspectives. The patient perspective, missing from the current study, will be imperative to the complete understanding of the first consultation. Future research should consider the age and perhaps gender or years of practice of GPs in terms of prescribing antidepressants. Further research into depression treatment in primary care can also assess both the type of treatment given and at which visit treatment was administered. A definition of “the first depression visit” may need to be established for this while considering the many ways that depression or distress can initially present.

In conclusion

• • There is considerable variation between individual GPs in how they manage patients with symptoms suggestive of depression at the initial consultation, and the drivers behind different management approaches.

• • A move toward a more standardized approach in primary care may be required for patients to receive equitable care that is evidence-based.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.