A survey of UK general practitioners about depression, antidepressants and withdrawal: implementing the 2019 Public Health England report

Abstract

Background:

In 2019, a literature review indicated that more than half of people who try to come off antidepressants experience withdrawal effects. Both the National Institute of Health and Care Excellence and the Royal College of Psychiatrists updated their positions in line with that review, and Public Health England published a 152-page report called Dependence and withdrawal associated with some prescribed medicines: an evidence review. The report made several recommendations relevant to general practice.

Method:

In order to facilitate implementation of these recommendations, an online survey was designed to explore United Kingdom general practitioner (GP) experiences, opinions, knowledge and needs in relation to depression, ADs and withdrawal. A total of 66 GPs had completed the survey when COVID-19 occurred.

Results:

In keeping with previous findings, this small sample of GPs had a predominantly psycho-social perspective on the causes of, and treatments for, depression. They broadly considered ADs effective for moderate/severe depression and ineffective for minimal/mild depression, for which they preferred psychological therapies and social prescribing. There was a marked lack of consistency in GPs’ knowledge about the incidence and duration of withdrawal effects. Only a minority (29%) felt their knowledge about withdrawal was ‘adequate’ and fewer (17%) believed this about their ‘Ability to distinguish between withdrawal effects and return of the original problem (e.g. depression)’. Two-thirds (68%) would like more training on these matters.

Conclusion:

It is hoped that even this small sample will be helpful when designing, and seeking funding for, GP training programmes, and when implementing the PHE recommendations for support services, based in the primary care system, for the millions of people contemplating or initiating withdrawal from ADs every year in the UK.

Introduction

Annual antidepressant (AD) prescribing in the United Kingdom (UK) has doubled over the past 10 years. Over a 12-month period between 2017 and 2018, 7.3 million adults (17% of the adult population) were prescribed ADs in England alone; the rates for women, older people and people living in deprived areas were even higher.¹ In the United States (US), 8% of the population aged over 12 used ADs in any given month between 1999 and 2002, increasing to 13% (37 million adults) by 2011–2014.² High prescription rates also occur in Australia, Belgium, Canada, Iceland, Portugal and Sweden.³

These continual increases occur despite significant concerns about efficacy and safety. A recent network meta-analysis reported small benefits compared with placebo, but the trials involved had multiple methodological flaws, with 82% rated as having moderate or high risk of bias.⁴ It has long been established that less than half of trials find ADs superior to placebo.^5,6 This lack of difference between ADs and placebos is particularly frequent in properly blinded, non industry-funded studies.^7,8 One meta-analysis found that ‘the overall effect of new-generation AD medications is below recommended criteria for clinical significance’, with benefit compared with placebo only for a tiny minority of recipients ‘at the upper end of the very severely depressed category’.⁹ Another meta-analysis, of 131 placebo-controlled trials, concluded that the overall effect size does not reach ‘clinical significance’ and argued that ‘the harmful effects of SSRIs versus placebo for major depressive disorder seem to outweigh any potential small beneficial effects’.¹⁰

High rates of adverse effects have been identified, originally in the biological domain, including nausea, impotence, insomnia, diarrhoea, dry mouth, dyspepsia and sweating,^11–13 but more recently also in the personal and interpersonal domains, including emotional numbing, feeling not like oneself, agitation, reduction in positive feelings, caring less about others and suicidality.^14–17

In this context, attempts to understand the perpetually increasing prescription rates began to focus on increases in repeat prescriptions. For example, UK data on 189,851 general practice patients revealed that a doubling of prescribing over 8 years was explained not by increases in new prescriptions but a doubling of the number of prescriptions per patient.¹⁸

Such findings raised the issue of the withdrawal effects of ADs, until recently a somewhat taboo topic. In 2018, guidelines from the UK National Institute for Health and Care Excellence (NICE) stated that AD withdrawal symptoms ‘are usually mild and self-limiting over about 1 week, but can be severe, particularly if the drug is stopped abruptly’.¹⁹ Meanwhile US guidelines claim that symptoms ‘typically resolve without specific treatment over 1–2 weeks’.²⁰ Three recent systematic reviews have, however, indicated that these are gross underestimates.^21–23

The most recent of the three reviews was undertaken for the All Party Parliamentary Group for Prescribed Drug Dependence in the UK,²¹ to inform an enquiry by Public Health England (PHE).¹ Fourteen studies found that withdrawal incidence ranged from 27% to 86%, with a weighted average of 56%. Only four studies assessed severity; they produced a weighted average of 46% of those experiencing withdrawal effects endorsing the most extreme severity rating on offer. Seven of the ten studies that reported duration found that a significant proportion of people experiencing withdrawal do so for much longer than 2 weeks, and that it is not uncommon for it to last for several months or, more rarely, years. The reviewers concluded¹:

‘We recommend that UK and USA guidelines on antidepressant withdrawal be urgently updated as they are clearly at variance with the evidence on the incidence, severity and duration of antidepressant withdrawal, and are probably leading to the widespread misdiagnosing of withdrawal, the consequent lengthening of antidepressant use, much unnecessary antidepressant prescribing and higher rates of antidepressant prescriptions overall. We also recommend that prescribers fully inform patients about the possibility of withdrawal effects.’

In May 2019, the Royal College of Psychiatrists (RCPsych) published an updated, evidence-based ‘Position statement on depression and antidepressants’,²⁴ including:

‘Discontinuation of antidepressants should involve the dosage being tapered or slowly decreased to reduce the risk of distressing symptoms, which may occur over several months. . . The use of antidepressants should always be underpinned by a discussion about the potential level of benefits and harms, including withdrawal.’

In September 2019, PHE published its 152 page document entitled ‘Dependence and withdrawal associated with some prescribed medications: An evidence review’.¹ Having meticulously documented the extent of the problem, it made a range of important recommendations, including for services to assist people coming off ADs and other psychiatric drugs, better research and more accurate national guidelines.

In October NICE updated its guidelines²⁵ in line with the 2019 Davies and Read review, recommending that doctors

‘Advise people taking antidepressant medication that if they stop taking it abruptly, miss doses or do not take a full dose, they may have discontinuation symptoms such as: restlessness, problems sleeping, unsteadiness, sweating, abdominal symptoms, altered sensations (for example electric shock sensations in the head), altered feelings (for example irritability, anxiety or confusion). Explain that whilst the withdrawal symptoms which arise when stopping or reducing antidepressants can be mild and self-limiting, there is substantial variation in people’s experience, with symptoms lasting much longer (sometimes months or more) and being more severe for some patients.’²⁶

Among the many evidence-based recommendations in the PHE report¹ was:

GPs develop their knowledge of, and competence to identify, assess and respond to, dependence or withdrawal associated with some medicines and the support needs of people experiencing problems with withdrawal or dependence.’

The current study was designed to assess GPs’ experiences, knowledge, views, and needs (see Methods), so as to help effectively target efforts to implement this recommendation, in relation to ADs.

Methods

The study was approved by the University of East London’s Research Ethics Committee (Application ID: ETH1920-0048).

An online questionnaire was designed, based primarily on the research literature discussed previously,^{1,4–11,18–23} and later,^27–38 in order to address UK GPs’ beliefs, knowledge and needs in relation to ADs in general and withdrawal therefrom in particular. Questions were also asked about what GPs think cause depression and about the influence of drug companies. Most questions generated quantitative data from multiple choice questions, but several generated qualitative data via open ended questions (including an ‘other’ option after some multiple-choice questions).

The questionnaire was trialled on three GPs, and minor amendments made. The British Medical Journal published an article announcing the launch of the study in February 2020.³⁹ The survey was also advertised on social media, including the ‘Resilient GP’ Facebook group. When the COVID-19 pandemic occurred, participation ceased. It was subsequently decided to publish the findings despite the small sample size, with clear statements about the obvious limitations involved.

Data analysis

Quantitative data are presented as descriptive statistics (percentages etc.) without analysis by demographics, due to small numbers. Responses to open questions were reported in terms of numbers of similar/identical responses.

Results

Sample characteristics

Between 7 February and 10 March 2020, 66 GPs completed the survey, although 3 left some of the questions unanswered towards the end of the survey (see Tables 3, 4, 6 and 8). Of these 66, 46 (70%) were women. The average age of the sample was 48.9 years (SD 10.3) and they had worked as GPs for an average of 18.2 years (SD 10.9). Almost all (97%) worked in England, with one each from Scotland and Wales.

Table 3.

Treatment approaches used for ‘minimal/mild’ depression.

Prescription for antidepressants	Never (1)	Some times (2)	About half the time (3)	Most of the time (4)	Always (5)	Mean
Recommend self-referral to IAPT	6%	14%	19%	38%	22%	3.56
Social prescribing (exercise, nutrition, social activity, self-help books etc.)	3%	21%	17%	46%	13%	3.44
Active monitoring/watchful waiting	2%	25%	25%	41%	6%	3.25
Refer to counsellor/psychotherapist/psychologist	6%	41%	22%	22%	8%	2.84
Refer to computerised CBT (e.g. ‘Beating the Blues’)	25%	44%	13%	14%	3%	2.25
Provide psychological intervention yourself	30%	43%	8%	14%	5%	2.21
Prescription for antidepressants	14%	71%	13%	2%	0%	2.02
Refer to mental health services	44%	38%	13%	2%	3%	1.81
Refer to in-house mental health staff	63%	24%	8%	3%	2%	1.56
Refer to psychiatrist	76%	19%	3%	0%	2%	1.32

CBT, cognitive behavioural therapy; IAPT, Improving Access to Psychological Treatments.

Table 4.

Treatment approaches used for ‘moderate/severe’ depression.

Prescription for antidepressants	Never (1)	Some times (2)	About half the time (3)	Most of the time (4)	Always (5)	Mean
Prescription for antidepressants	0%	9%	21%	62%	8%	3.68
Recommend self-referral to IAPT	13%	16%	11%	30%	30%	3.49
Social prescribing (exercise, nutrition, social activity, self-help books etc.)	6%	21%	19%	29%	25%	3.46
Refer to counsellor/psychotherapist/psychologist	5%	29%	19%	31%	16%	3.24
Refer to mental health services	7%	52%	21%	15%	5%	2.59
Provide psychological intervention yourself	37%	33%	8%	13%	9%	2.25
Refer to computerised CBT (e.g. ‘Beating the Blues’)	37%	40%	11%	6%	6%	2.06
Refer to psychiatrist	18%	63%	14%	5%	0%	2.06
Active monitoring/watchful waiting)	33%	54%	5%	5%	3%	1.90
Refer to in-house mental health staff	63%	14%	14%	6%	2%	1.69

CBT, cognitive behavioural therapy; IAPT, Improving Access to Psychological Treatments.

Table 6.

Length of time thought necessary for successful withdrawal.

n = 63% of patients	What % ‘can come off antidepressants successfully within two months’	What % ‘need very small decreases in antidepressant dosages over many months to come off them successfully’
0	3%	2%
1–10%	9%	19%
11–20%	8%	16%
21–30%	9%	9%
31–40%	6%	8%
41–50%	19%	11%
51–60%	5%	9%
61–70%	11%	5%
71–80%	21%	9%
81–90%	6%	6%
91–100%	2%	5%

Table 8.

Factors contributing to prescription rates of antidepressants increasing annually for the past 20 years.

n = 63	Strongly agree (1)	Agree (2)	Neither agree nor disagree (3)	Disagree (4)	Strongly disagree (5)	Mean
Cuts to social services, benefits etc.	51%	32%	16%	0%	2%	1.70
People are less embarrassed about saying they are depressed	38%	49%	5%	5%	3%	1.86
More people just want to feel better without making changes in their lives	36%	38%	17%	8%	0%	1.97
GPs have less time to talk with patients	43%	33%	6%	9%	8%	2.06
People are no more depressed than they used to be, but more are treated	29%	48%	11%	11%	2%	2.10
Social media	29%	40%	27%	3%	2%	2.10
Other types of treatment are not funded or are too expensive	36%	40%	6%	9%	8%	2.13
Drug companies have successfully promoted an illness model of depression	32%	35%	16%	16%	2%	2.21
People are finding it difficult to come off their antidepressants	21%	36%	29%	14%	0%	2.37
Many people don’t want talking therapies	16%	41%	22%	14%	6%	2.54
More people are depressed these days	13%	32%	30%	14%	11%	2.79
Brexit	6%	27%	30%	19%	17%	3.14
antidepressants are the best treatment	0%	8%	35%	36%	21%	3.70
	Far too high	Slightly too high	About right	Slightly too low	Far too low
What is your opinion about the current rate of antidepressant prescribing (one in six adults in England) (n = 66)	43%	40%	16%	2%	0%

When asked to estimate how many of their patients ‘present with mood/depressive symptoms’ 26 (39%) ticked ‘21–30%’, followed by 17 (26%) ticking ‘11–20%’ and 9 (14%) who estimated ‘31–40%’.

Causal beliefs

The GPs were asked: ‘What do you think are the relative contributions of bio-genetic causes (e.g. chemical imbalance, genetic predisposition) versus social causes (e.g. stressful/traumatic events, loss etc.) for depression?’ The majority (53; 80%) felt that social causes contributed more than bio-genetic causes. The ratio most commonly endorsed (19; 29%) was ‘Bio 30% – Soc 70%; followed by ‘Bio 20 – Soc 80% (17; 26%). The most strongly endorsed specific causal factors were ‘Child abuse or neglect’ and ‘Violence/rape in adulthood’, and the least commonly endorsed were ‘Genetic predisposition’ and ‘Chemical imbalance’ (see Table 1).

Table 1.

Which factors are ‘causes of depression’?.

n = 66	Strongly agree (1)	Agree (2)	Nether agree nor disagree (3)	Disagree (4)	Strongly disagree (5)	Mean
Child abuse or neglect	86%	14%				1.14
Violence/rape in adulthood	82%	18%				1.18
Other childhood adversities	80%	20%				1.20
Isolation/loneliness	79%	21%				1.21
Drug/alcohol abuse	79%	18%	3%			1.24
Family stress	73%	26%	2%			1.29
Financial problems	71%	18%	2%			1.30
Relationship problems	64%	36%				1.36
Loss of loved one	67%	29%	5%			1.38
Work stress	64%	35%	2%			1.38
Medical conditions	58%	41%	2%			1.44
Genetic predisposition	45%	36%	11%	5%	3%	1.83
Chemical Imbalance	23%	54%	8%	11%	5%	2.18

Twenty four participants added 28 ‘other’ causes. The only causes mentioned by more than one GP were social media - 3; personal characteristics - 3 (‘poor coping skills’, ‘low resilience’, ‘personality traits’); and loss of control - 2 (e.g. ‘lack of control over many aspects of life eg poor housing, bad environment, high crime area etc.).

Perceived efficacy

Table 2 shows that the GPs thought ADs were far more effective for ‘moderate/severe’ depression than for ‘minimal/mild’ depression; but only slightly more effective in the first year of treatment than thereafter.

Table 2.

Perceived efficacy of antidepressants in various circumstances.

n = 65	Very effective (1)	Somewhat effective (2)	Slightly effective (3)	Not at all effective (4)	Mean
‘minimal/mild depression’	5%	28%	37%	31%	2.94
‘moderate/severe depression’	25%	68%	5%	3%	1.86
‘short-term treatment of depression (less than a year)’	29%	48%	15%	8%	2.02
‘long-term treatment of depression (a year or more)’	15%	62%	17%	6%	2.14

Information sources

The most commonly endorsed responses to ‘Which of the following have you used in the past 12 months to inform your decisions about the treatment of depression?’ were British National Formulary (76%) and NICE Guidelines (71%), followed by Research Articles/Reviews (33%), Maudsley Prescribing Guide lines (27%) and Training Programme (27%). None endorsed ‘Drug Company Reps’ or ‘Other Drug Company Information’. Half of the participants (33) cited 38 ‘other’ information sources, most commonly: local presentations/trainings (9); discussions with psychiatrists/mental health team (8); my own experience/learning from patients (6); and local guidelines (4).

Of the 60 who answered the question about contact with drug company reps, 83% reported no contacts in the past year, 8% reported 1 contact and 8% reported between 2 and 14 contacts. Of those with at least one contact, 78% reported that their clinical practice was ‘not at all’ influenced, 17% ticked ‘a little’ and 5% ticked ‘a moderate amount. Overall, only four GPs (7%) acknowledged being influenced. However, when asked how much other GPs were influenced, they reported that 82% of their colleagues had been influenced, with 25% ‘a moderate amount’ and 5% ‘a lot’.

Clinical practice

When asked ‘On average how long are you able to spend with a patient in the session at which you first prescribe antidepressants?’ most (69%) ticked ‘10–20 min’; 23% ticked ‘Less than 10 min’; and 8% ‘21–30 min.’ None ticked ‘31–45 min’ or ‘ More than 45 min’.

Table 3 records that the most preferred of 10 treatment options for ‘minimal/mild’ depression were: recommend self-referral to Improving Access to Psychological Therapies (IAPT); social prescribing; and active monitoring; with ADs the 7th most endorsed option. For ‘moderate/severe’ depression, ADs were the most preferred option, followed by referral to IAPT and social prescribing, with active monitoring relegated to 9th position (Table 4). Referral to a psychiatrist was 10th (last) for ‘minimal/mild and 8th for moderate/severe.

‘Other’ treatments used by more than one GP, for minimal/mild depression, were mindfulness/meditation/yoga (3) and socialising (3). For moderate/severe depression some GPs also addressed work issues (3) and used crisis/support numbers (2).

Of the 62 GPs who responded to the statement ‘Talking therapies should be as accessible as pharmacological treatments for depression’, 93% strongly agreed, 5% agreed and 2% (one GP) had no opinion.

The GPs were asked ‘When discussing possible prescribing of antidepressants, how often do you inform patients of the possibility of withdrawal effects when reducing or coming off?’ Of the 63 who responded, 52% ticked ‘Always’, 25% ticked ‘Most of the time’, 14% ‘About half the time’, 5% ‘Occasionally’ and 3% ‘Never’.

Participants were also asked ‘After patients have been on antidepressants for 3 months, approximately how often do you initiate discussion about when to come off them?’ The most frequently endorsed of the five options were ‘every 3 months’ (36%) and ‘every 6 months’ (36%), followed by ‘once a year’ (17%), once a month (11%) and ‘never’ (3%).

Withdrawal: beliefs, knowledge and training needs

Table 5 reports GPs’ estimates of how many people are ‘likely to experience withdrawal’ after being on ADs for three different time periods. Regardless of the time period, about one in four GPs (27%, 24%, 24%) believe that withdrawal effects are experienced by no more than 10%. Forty percent thought that even after being on antidepressants for 3 years, withdrawal is experienced by no more than 30% of their patients.

Table 5.

GPs’ estimates of how many people are likely to experience withdrawal effects after being on antidepressants for various periods of time (n = 63).

% of patients thought to experience withdrawal	After 3 months	After 1 year	After 3 years
0	5%	3%	6%
1–10%	22%	21%	18%
11–20%	21%	14%	5%
21–30%	17%	9%	11%
31–40%	8%	3%	11%
41–50%	8%	19%	6%
51–60%	5%	13%	5%
61–70%	3%	8%	8%
71–80%	11%	3%	10%
81–90%	-	5%	10%
91–100%	-	2%	11%

There was a broad range of responses to ‘What percentage of patients can come off antidepressants within 2 months successfully?’ (see Table 6). There was a similar lack of consensus when asked ‘What % need very small decreases in antidepressant dosages over many months to come off them successfully’.

Table 7 shows that just over half of the GPs thought their knowledge about withdrawal effects and their ability to distinguish withdrawal from relapse was ‘somewhat adequate’. Perhaps the most important finding of this study is that the majority (68%) said they would like more training or information. When asked what kind of training or information, 35 provided 47 suggestions. In terms of content, six wanted guidelines/protocol/flowchart/strategies on how to wean patients off, and three wanted information about the withdrawal effects, including what they were, incidence and differences between antidepressants. In terms of process, 17 wanted some form of online training (e-learning, webinar) and 8 wanted a local, face-to-face meeting/training session.

Table 7.

Knowledge and training needs.

n = 66	Adequate	Somewhat adequate	Not sure	Somewhat inadequate	Inadequate
‘Knowledge about the withdrawal effects of antidepressants?’	29%	54%	11%	5%	2%
‘Ability to distinguish between withdrawal effects and return of the original problem (e.g. depression)’	14%	56%	19%	8%	3%
‘Would you like more training or information about the withdrawal effects of antidepressants?’		YES 68%	16%	NO 16%

Prescription rates

The most endorsed of 13 factors explaining the increasing rates of prescribing were ‘Cuts to social services, benefits, etc.’ and ‘People are less embarrassed about saying they are depressed’ (see Table 8). The least endorsed was ‘Antidepressants are the best treatment’. Twenty-one GPs offered 32 other factors, including patient expectations (5), need for quick fix/magic pill (4), austerity (3), time-pressured lives (3), drug company pressure and misinformation (2) and limited mental health services (2).The majority (83%) think the prescribing rate is too high.

Services for people in withdrawal

When asked ‘What services, if any, should be provided for people when they experience withdrawal effects from antidepressants’ 42 GPs offered 56 suggestions. Table 9 summarises these recommendations. All 56 who answered the question ‘Who should provide these services/’ ticked ‘NHS’ (National Health Service), with 48% also endorsing ‘NGOs (non-government organisations)/voluntary sector’ and 11% ‘private sector’ (participants could tick more than one).

Table 9.

Services needed for people when withdrawing from antidepressants.

		Examples
Counselling/ talking therapies/ psychological support	12	Supportive psychological therapies for targeted support. Support and counselling.
Written information	7	More patient information on what to expect when withdrawing. At least an information leaflet with support from GP.
Telephone helpline	6	A help line and website. A dedicated helpline based in community mental health sector (and/or in primary care, specifically commissioned)
Access to pharmacist	6	Access to trained, experienced pharmacists. Easy access to liquid formulations to be able to make microscopic downward titrations with pharmacy supervision etc.
Online support/information	5	On line information and guidance. Online support where they can submit their side effects and receive tailored guidance about how to reduce safely.
Informed GP	5	There is no reason why a GP or a primary care mental health worker cannot deal with this. They just need time, which is what we do not have. Ease of access to a GP with knowledge about how to manage it. Someone with expertise (could be the GP with information or access to specialist advice)
Mental health services	4	Community mental health support. Competent mental health professional (and I don’t mean a randomly named minimally experienced ‘mental health support worker’).
Individualised plan	4	An individual plan for the person to come off slowly with clear explanations as what to expect and what to do if s/he experiences withdrawal symptoms.
Group support	2	Patient groups
Key worker	2	Named support worker

Recommendations

The GPs were asked ‘Do you have a message for Mental Health or Health Ministers about depression and/or its treatment?’ and 47 (71%) responded, with 69 messages. The most common theme was increasing mental health services (21), with specific references to inaccessibility of psychological/talking therapies (12) [including IAPT (6) and CBT (2)], and children’s services (6). Only one GP mentioned community psychiatric nurses (CPNs) and psychiatrists. Six GPs wanted to tell the Minister to tackle the social causes of depression. Five wanted increased focus on social prescribing. Four wanted an overall reduction in the medicalising and medicating of depression and other forms of distress. Some examples are presented in Table 10.

Table 10.

Examples of messages for Ministers about depression and/or its treatment.

We are overtreating depression with medication instead of improving access to talking therapies and tackling the causative issues of social isolation, social media and poverty as well as many other social problems

We are causing significant harm to our patients by continuing with the biological chemical imbalance model of depression and prescribing potentially harmful drugs which cause suffering in withdrawal and may actually contribute to chronic depression

People need help and advice on how to improve quality of their lives, for example, advice re. hobbies locally. I think every area should have updated lists of interests/hobbies for all age groups and contact details provided. Thinking of isolated people, single parent families, etc.

There is still a huge shortage of MH services across the board. CAHMS is woeful in most areas, the waits for IAPT ridiculous, the CMHTs in crisis and so as GPs we are left trying to sort these vary complex patients in 10 min appointments as well as deal with their physical health, carers and families.

The erosion of continuity of care in general practice leads to poorer, more expensive health care. A trusted relationship with a GP who can see you repeatedly over time is one of the cheapest interventions and likely to be equally as effective as medication

Social prescribing should be promoted to people before they get ill, for example, through schools, on the tv.

Please stop medicalising everything. Not every low mood is Depression which needs to see a GP. Clinicians cannot solve mental health issues caused by poverty, unemployment, poor education etc. Please address poverty.

Increase psychological services provision

Please ensure that GPs have a real and accessible ALTERNATIVE to prescribing antidepressants. It is heart breaking to feel this is all we have to offer.

I think it’s shameful that talking therapies and social prescribing are so inaccessible, and if available have such long waiting lists as to render them useless.

CAMHS, child and adolescent mental health services; CMHT, community mental health team; IAPT, Improving Access to Psychological Treatments; MH, mental health.

Finally, the GPs were asked ‘What needs to change to reduce levels of depression in society?’ The 44 who responded provided 62 recommendations. The most common theme was increasing social connectedness/reducing isolation (15), followed by reducing inequality (9), improving children’s wellbeing/safety (8) and improving work culture (7). See Table 11 for examples.

Table 11.

Examples of responses to ‘What needs to change to reduce levels of depression in society?’.

Improve social support, help reduce isolation and bring back communities.

1. Accommodation; 2. Benefits; 3. Employment (IPS); 4. Education & training; 5. Socialisation support (preventing isolation)

More social support - both within society and provision by local government/social services/NHS.

Societal change, more inclusion, local activities, greater sense of community, Less shift work, more focus on importance of rest, daylight, time to prepare healthy meals, exercise- requiring improved/safer cycle ways towns designed around people not cars and shorter working days to allow time for people to look after themselves.

Greater equity of income/wealth. More emphasis on health and wellbeing of children and support for their parents.

Less pressure in schools on results, support for young peoples services

Less inequality, less poverty, less loneliness, more social cohesion.

More work to address health inequalities, child poverty, in-work poverty, homelessness and alcohol/substance misuse.

Many jobs are working people increasingly harder. Many jobs are not secure. The cost of living has risen much quicker than wages, putting financial strain on people. Employers expect more and more from employees with little reward or consideration for their well-being. It seems that these days people think that many problems can be fixed with tablets (I think the drug companies are to blame for this, and worry that if we head towards a system like America, this will get worse with drug advertising). Many medical problems can be helped by eating well/exercising/sleeping well/socialising - but these all take time and effort, something which the modern day doesn’t seem to allow for easily. Better education in school about looking after yourself would be a start.

IPS, individual placement support; NHS, National Health Service.

Discussion

A psycho-social perspective

Some commentators blame the epidemic of AD prescribing on an overly biological approach towards human distress adopted by psychiatry, and the powerful influence of the drug companies on prescribers and consumers,^7,27–30 often exerted via biased, industry-sponsored websites.^31,32

Overall, however, this small sample of GPs adopts a predominantly psycho-social perspective on the causes of, and solutions to, depression. Most (80%) believe that psycho-social factors are more important than bio-genetic factors, with ‘Child abuse/neglect’ and ‘Violence/rape in adulthood’ the 2 most endorsed of 13 specific causes, and ‘Genetic predisposition’ and ‘Chemical Imbalance’ the 2 least endorsed. The most common recommendations for reducing societal depression levels are: increasing social connectedness/reducing isolation, reducing inequality, and improving children’s wellbeing/safety. These beliefs are consistent with previous studies of GPs,^33,34 and with the public’s causal beliefs,^40,41 including people taking ADs.^42,43

These GPs do believe there is a role for ADs, but only for moderate/severe depression. For minimal/mild depression, six other treatment approaches are preferred, most strongly psychological therapies and social prescribing. The least endorsed of 13 explanations for increasing prescription rates is ‘Antidepressants are the best treatment’ (with the most endorsed being ‘Cuts to social services, benefits, etc.’). Most (83%) think prescription rates are too high. All but one (98%) agreed that’ Talking therapies should be as accessible as pharmacological treatments’.

Two in every three (67%) believe that one of the factors for ever increasing prescribing rates is that ‘Drug companies have successfully promoted an illness model of depression’. Although few (7%) believe that they are influenced by drug company salespeople themselves, most (82%) believe that their colleagues are influenced. (This is, however, a common phenomenon, exemplified by most of us believing that we are better than the average driver).⁴⁴ Another potential explanation, not considered by our survey, is that GPs prescribing decisions are led by whether patients present their difficulties in terms of symptoms or psycho-social events.⁴⁵

Implementing PHE recommendations

This very small sample of GPs reports a very wide range of beliefs about how many people experience withdrawal symptoms (Table 5), and about how long people need to successfully come off ADs (Table 6), indicating that, for many, these are guesses rather than evidence-based assertions. This is understandable given the misinformation published by NICE, RCPsych, and other official bodies until very recently. One in four (24%) think that even after being on antidepressants for 3 years, no more than 10% of people will experience withdrawal symptoms when they try to come off. This contrasts with the 56% average rate identified (for all lengths of treatments combined) by the latest review.²¹

Nearly half (45%) think that most people can ‘can come off antidepressants successfully within 2 months.’ Evidence is emerging, however, that suggests that antidepressants, like benzodiazepines, should usually be tapered very slowly, often over several months or longer (but tailored to the individual) not 2–4 weeks as suggested by many guidelines.^35,36

Less than one in three (29%) believe their ‘Knowledge about the withdrawal effects of antidepressants’ is ‘adequate’; and only about one in seven (14%) think their ‘Ability to distinguish between withdrawal effects and return of the original problem (e.g. depression)’ is ‘adequate’. Two-thirds (68%) state they would like more training on these matters, particularly (but not exclusively) online training about strategies for weaning patients off.

There was also a clear lack of consistency in how often GPs ‘initiate discussion about when to come off them’. The finding that nearly one in three (32%) did so once a year or less often might be a focus of training.

The PHE recommendation that has been the focus of this paper is just one of several aimed at ‘Improving support available from the healthcare system’. The range of services identified by our GPs in Table 9 seem important, as is their unanimous message that such services should be provided by the NHS, with about half (48%) also endorsing ‘NGOs/Voluntary sector’. The focus on targeted psychological support during withdrawal is consistent with a recent systematic review on managing withdrawal from antidepressants’.³⁷

Information given to patients

Most (77%) GPs ticked ‘Always’ or ‘Most of the time’ when asked ‘When discussing possible prescribing of antidepressants, how often do you inform patients of the possibility of withdrawal effects when reducing or coming off?’ This is in stark contrast to the two largest surveys ever conducted, of over 1800 and over 1400 AD recipients,^16,38 in which less than 2% reported being told anything about withdrawal effects by the prescribing doctor.

If our GPs’ reports of their own practice are accurate, rather than the result of social desirability, this would lend support to the possibility that our small sample did, indeed, differentially include GPs with a high degree of knowledge, and good practice, about, AD withdrawal (see Limitations).

Telling people about adverse effects is not only a pre-requisite for meeting the essential ethical principle of informed choice, it can have unexpected beneficial effects. In a large online survey, self-reported efficacy was independently predicted, after controlling for a range of other psycho-social variables, by both the amount of information about ADs offered by the prescriber and the perceived quality of the relationship between prescriber and patient.³⁸

Sources of information for GPs

It seems that at least as many GPs consult the British National Formulary (BNF) (76% in the current sample) as NICE guidelines (71%). Changes in NICE need to be paralleled by updates to the BNF, which currently promotes the notion that ‘Patients with a history of recurrent depression should receive maintenance treatment for at least 2 years’. It also states that the frequency of ‘withdrawal syndrome’ is ‘not known’, and that ‘withdrawal effects are usually mild and self-limiting, but in some cases may be severe’.⁴⁶

Limitations

The obvious limitation to this study is the very small sample size, representing only about 0.15% of GPs in England, and effectively none from the rest of the UK. Under normal circumstances such a sample size would prohibit submission to a journal. In the current abnormal circumstances (COVID-19), however, with no further recruitment possible, or appropriate, we hoped our data, however limited, may be helpful to government officials, professional bodies, and researchers,⁴⁷ planning for implementation of the hugely important PHE Report once these circumstances abate.

The most likely bias resulting from the tiny sample is disproportionate inclusion of GPs with a particular interest in, and knowledge of, ADs and withdrawal therefrom. If this was the case (and we have no way of knowing) then the findings relating to the psycho-social perspective of GPs, for example, should be received with caution. The levels of perceived inadequacy of knowledge and the numbers with inaccurate beliefs about incidence and duration of withdrawal, might be even more pronounced in a more representative sample.