Psychiatrists' use an outcome centered prescribing model

Abstract

Psychiatrists are said by critical psychiatrists to use a disease-centered prescribing model in which medication is used to reverse an underlying disease-causing mechanism. Consultant psychiatrists were invited to complete an online questionnaire exploring their rationale in prescribing medication for specific indications to discover their prescribing model(s). In most clinical situations psychiatrists were more likely to choose medication based on an outcome-centered model where medication has been shown to be effective in research but the mechanism of this effect is unclear. An exception was the use of thiamine to prevent Wernicke-Korsakoff syndrome in alcohol withdrawal where a disease-centered model was commonest. In contradiction to the assertion by critical psychiatrists, a disease-centered model is rarely used except when there is clear evidence for this model. In most clinical situations prescribers acknowledged uncertainty in the evidence base.

Full text

1
Psychiatrists’ use an outcome centered prescribing model
Adam Moreton1, Mawada Adam1 and Ahmed Samei Huda1
March 2018
Psychiatrists are said by critical psychiatrists to use a disease-centered prescribing model in which medication is used to
reverse an underlying disease-causing mechanism. Consultant psychiatrists were invited to complete an online
questionnaire exploring their rationale in prescribing medication for specific indications to discover their prescribing
model(s). In most clinical situations psychiatrists were more likely to choose medication based on an outcome-centered
model where medication has been shown to be effective in research but the mechanism of this effect is unclear. An
exception was the use of thiamine to prevent Wernicke-Korsakoff syndrome in alcohol withdrawal where a disease-
centered model was commonest. In contradiction to the assertion by critical psychiatrists, a disease-centered model is
rarely used except when there is clear evidence for this model. In most clinical situations prescribers acknowledged
uncertainty in the evidence base.
INTRODUCTION
Critical Psychiatry is a movement which argues that
utilising a medical framework for the treatment of mental
disorder is not appropriate [1]. Members of this group refer
to themselves as Critical Psychiatrists. Key members of this
movement have suggested two conceptual models of
prescribing: drug- and disease-centered; and stated in a
recent article [2] that a disease-centered model is the
prevailing approach used by psychiatrists.
A disease-centered model presumes psychiatric
medications correct a hypothesised underlying mechanism
which is causing the mental disorder and that medications
are beneficial. In the drug-centered model, psychiatric
medications have several psychoactive effects some of
which can be perceived as beneficial in some patients, whilst
in others they can be harmful, e.g. emotional blunting.
There are other conceptual models used by psychiatrists.
In an outcome-centered model, a medication may be chosen
as research has shown it to be effective for certain clinical
indications but that how this effect is achieved is uncertain;
in addition, the balance of benefit and harm should be
carefully considered. Psychiatrists may also use different
prescribing models across different conditions due to highly
variable evidence for the aetiology underlying different
conditions and the mechanisms of action for different drugs.
A literature review failed to identify any studies measuring
the frequency of use of the various conceptual models of
prescribing used by psychiatrists. Research tended to focus
on attitudes to prescribing specific medications such as depot
antipsychotics [3], therapeutic drug monitoring [4], and
patient attitudes to antipsychotics [5]. Therefore a survey of
psychiatrists about their prescribing in generality and in
various clinical situations would provide useful data to add
to this debate.
METHODS
Consultant psychiatrists within a North West England
NHS Mental Health Trust covering a mix of urban and rural
areas were invited to complete an online questionnaire
assessing their rational for prescribing in common clinical
scenarios.
Two questions collected data on the respondent’s
specialty and duration of experience in psychiatry. The
remaining questions assessed their general attitude to
prescribing medication within the field of mental health in
utilising different medication classes for specified
indications.
For each of the eight questions relating to prescribing
four multiple choice options were available. The available
answers were presented in a random sequence each time the
survey was accessed to prevent a tendency to select a
particular order of answer giving a false impression of
consistent choice in prescribing model. For each scenario the
four options corresponded to drug-, disease- and outcome-
centered models along with a fourth option where
medication was not felt to be indicated.
For example, in the depression scenario the survey asked
‘what is your attitude towards prescribing antidepressants in
an adult with depression of at least moderate severity?’
Respondents were asked to choose from these options
(comments in parentheses were not visible but are added
here to indicate the prescribing-model):
• It reverses a known pathological process or corrects an
underlying abnormality such as a mono-amine deficit.
This leads to the resumption of normal functioning
i.e. a cure [disease-centered: medication corrects an
underlying mechanism causing the mental disorder];
• It causes an altered mental state by a psychoactive effect
and this altered mental state is regarded as superior to
being depressed. It can cause emotional blunting which
may be regarded as helpful in depression [drug-
centered: medications have several psychoactive effects,
some of which harmful and some are beneficial];
1Pennine Care NHS Foundation Trust, Tameside General
Hospital, Greater Manchester, UK
Correspondence to Dr A. S. Huda at Pennine Care Early
Intervention Team, 133 Astley Street, Dukinfield, Cheshire,
SK16 4PU; ahmed.huda@nhs.net.
A truncated version of this study was published an eLetter
response to https://doi.org/10.1192/bjb.2017.11 in BJPsych
Bulletin March 2018

2 Psychiatrists use outcome centered prescribing
• It has been shown in research to be effective by reducing
depression symptom scores. It effects the activity of
mono-amines at receptors but how this leads
to improvement is not fully understood [outcome-
centered: research has shown medications to
be effective for certain clinical indications but
how this effect is achieved is uncertain]; and
• I would not use antidepressants in adults with depression
as I do not think they are effective [medication not
indicated].
For each question the percentage of respondents choosing
each answer was calculated. The proportion choosing the
most frequently selected answer was compared to the
number selecting any of the other three answers combined
using the Clopper-Pearson method, a conservative approach
to assessing binomial proportions, with two sided p-values
reported [6]. For each question the three least frequently
chosen options were combined due to low numbers (or zero)
choosing some options.
To assess the impact of clinical experience the frequency
of choosing the most popular answer for each question,
as indicated by the undifferentiated analysis, was compared
for those with less than 25 years and those with 25 years or
more experience using Fisher's exact test. All statistical
analysis was undertaken using StatsDirect v. 2.8 [7].
RESULTS
The survey link was sent to 88 consultants in March 2016
with a reminder sent six weeks later. Responses were
received from 56 psychiatrists from various psychiatric
subspecialties (general adult n=23; liaison n=3; child and
adolescent n=10; intellectual disability n=2; psychotherapy
n=1; rehabilitation n=5; older adults n=10; and substance
misuse n=2), resulting in a 64% response rate. Not all
respondents answered every question.
All respondents had worked in psychiatry for a minimum
of six years. Five (9%) declared between six and 10 years
experience, 37 (66%) reported between 10 and 25 years, and
14 (25%) have been working for 25 years or more.
In a general question about using psychiatric medication,
respondents reported that when they prescribe they do so as
it may be beneficial, but with the acknowledgement that
medications can also cause harmful side-effects.
In all but one prescribing situation respondents most
frequently chose answers corresponding to an outcome-
centered model (Table 1). The exception was use of thiamine
in alcohol withdrawal where disease-centered prescribing
was most prevalent.
No statistical difference was seen in the most popular
choice of model between psychiatrists of different
experience levels.
DISCUSSION
The decision to prescribe requires consideration of
various benefits and drawbacks in the context of the
individual patient - such an approach corresponds to the
outcome-centered prescribing model which was found to be
most prevalent in this survey. The exception of preference
for a disease-centered model for thiamine in Wernicke-
Korsakoff syndrome demonstrated that respondents did not
have a fixed model of prescribing for all situations but used
different models based on the available knowledge base.
The lack of variation in responses between the two levels
of clinical experience indicated that an outcome-centered
model of prescribing occurred across psychiatrists regardless
of their duration in the specialty.
The main limitations of this survey was its response rate
and that it was conducted in a single NHS trust. Although a
high degree of consensus was seen in choosing the most
popular answer in each scenario.
CONCLUSION
It was clear that the psychiatrists surveyed utilised
prescribing models that were reactive to their clinical
experience. In contradiction to the assertion that a disease-
centered model is most prevalent amongst psychiatrists, an
outcome-centered model was overwhelmingly chosen by
respondents. In most clinical scenarios respondents’
prescribing models used empirical evidence of effectiveness
but with an awareness of uncertainty around mechanism of
effect.
Further surveys amongst other mental health
professionals to ascertain their views on medication to detect
Model n= (%)
Scenario
Disease
centered
Drug
centered
Outcome
centered
Medica-
tion not
indicated
p =
General attitude
to psychiatric
medication
2 (4)
0 (0)
54 (96)
0 (0)
<0.01
Antidepressants
in depression
5 (9)
4 (8)
44 (83)
0 (0)
<0.01
Antidepressants
in anxiety
disorders
3 (6)
1 (2)
46 (88)
2 (4)
<0.01
Antipsychotics in
psychosis
3 (6)
3 (6)
46 (88)
0 (0)
<0.01
Lithium in
bipolar
disorder
1 (2)
0 (0)
50 (96)
1 (2)
<0.01
Benzodiazepines
in short term
anxiety
1 (2)
11 (22)
36 (71)
3 (6)
<0.01
Stimulants in
ADHD
0 (0)
5 (10)
43 (84)
3 (6)
<0.01
Thiamine to
prevent
Wernicke-
Korsakoff
syndrome
35 (70)
0 (0)
15 (30)
0 (0)
<0.01
Table 1: Responses to each scenario

3 Psychiatrists use outcome centered prescribing
similarities or otherwise with psychiatrists would help
develop knowledge in this area.
CONFLICT OF INTEREST
The authors have declared no conflicts of interest.
REFERENCES
[1] Moncrief J, Huws R. About Us. The Critical Psychiatry Network.
Available from: http://www.criticalpsychiatry.co.uk/index.php/36-
info/info/226-about-us [cited 19th June 2018].
[2] Yeomans D, Moncrieff J, Huws R. Drug-Centred Psychopharmacology:
a Non-Diagnostic Framework for Drug Treatment” BJPsych
Advances 2015; 21: 229–236.
[3] Waddell L, Taylor M. Attitudes of patients and mental health staff to
antipsychotic long-acting injections: systematic review. British
Journal of Psychiatry 2009; 195: (Supplement 52) S43-S50.
[4] Law S, Haddad P, Chaudhry I, Husain N, Drake R, Flanagan R, David
A, Patel M. Antipsychotic therapeutic drug monitoring:
psychiatrists' attitudes and factors predicting likely future use.
Therapeutic Advances in Psychopharmacology 2015; 5: 214-223.
[5] Day JC, Bentall RP, Roberts C, Randall F, Rogers A, Cattell D, Healy
D, Rae P, Power C. Attitudes toward antipsychotic medication: the
impact of clinical variables and relationships with health
professionals. Archives of General Psychiatry 2005; 62: 717-724.
[6] Newcombe RC. Two sided confidence intervals for the single
proportion: a comparative evaluation of seven methods. Statistics
in Medicine 1998; 17: 857-872.
[7] StatsDirect. StatsDirect Statistical Software 2013. Available from:
www.statsdirect.co.uk.
A truncated version of this study was published an eLetter response to https://doi.org/10.1192/bjb.2017.11 in BJPsych Bulletin March 2018