Question: How
effective is sertraline compared with cognitive behavioural therapy (CBT) in
treating adults with generalised anxiety disorder (GAD) in the third step of


Hypothesis:  Participants undergoing CBT will have a
greater reduction in their GAD symptoms and will better tolerate therapy than participants
that take sertraline; measured at 6 months by the GAD-7 questionnaire.

Primary objective: to determine
which single step three treatment option is the most effective in the
management of generalized anxiety disorder.

Secondary objective: To put
forward an efficient and streamlined treatment algorithm for generalized anxiety


anxiety disorder (GAD) is a highly impactful, underdiagnosed chronic mental
health disease. It is clinically defined as “Excessive anxiety and worry about
a number of events or activities that occur more days than not for at least 6
months.” (1) It is a multifactorial disease that seems to have many triggers.
These range from genetic predisposition to general life stressors and physical
or emotional trauma. It mainly affects people between the ages of 35 to 59 and
is slightly more common in women than men (2). Patients that suffer from GAD
feel a constant sense of restlessness or worry and find it difficult to sleep
or concentrate and may suffer from dizziness or palpitations. This is
particularly important because the impact of these symptoms are wide and far
reaching. Not only is it incredibly distressing for the patients themselves,
but it also affects their family members who witness their loved ones change
and become unable to function due to such high levels of anxiety. Additionally,
the lack of sleep and concentration may cause an inability to work effectively
which leads to a lack of productivity or worse, lost days at work. This
combined with the prevalence of the disease leads to a large amount of money
lost to the economy. The lack of evidence on the most effective treatment to
offer also increases the burden on the health service as patients are not being
treated in the best possible way. There has been a great deal of research done
on the impact of GAD, it’s prevalence and attitudes towards it. In 2001 (3), a
quantitative study used questionnaires to find out how general practitioners
(GPs) viewed GAD and how likely it was to be diagnosed correctly. They found
that only “59.6% of physicians viewed GAD as a genuine mental disorder” (3) and
only “34.4%” (3) of patients were correctly diagnosed with GAD. While a
questionnaire is not the best instrument to gather information due to the many
associated biases, it does serve to give an overview of how GAD is seen. As for
the prevalence of GAD, a review and meta-analysis on the “Patterns and
correlates of generalized anxiety disorder in community samples” (4) concluded
that GAD is a “commonly occurring mental disorder that can seriously impair
functioning.” (4) This idea of GAD being highly prevalent was also corroborated
by another meta-analysis titled “A review and meta-analysis of the genetic
epidemiology of anxiety disorders.” (5) This used data from several studies to
conclude that “Panic disorder, generalized anxiety disorder, phobias, and OCD
all have significant familial aggregation.” (5) To evaluate the
impact of GAD, a 2008 study (6) reviewed 34 studies that reported the
associations between “GAD and role functioning, QOL, and/or economic costs” (6)
and determined that the “QOL impairments of GAD were at least comparable in
magnitude to those of other anxiety disorders… and physical
conditions, and greater than those of substance use disorders” (6)
which helps us realise just how much of an effect GAD has on the lives of those
affected. While it is very difficult to effectively compare poorly understood
mental health problems and substance use disorders, this does help to give us
an idea of how highly impactful GAD is. In terms of the current NICE management
for GAD, the algorithm has a 4-step process. In the first step, patients are
assessed and given education and monitoring. In the second step, they may be
referred to psychoeducational groups. In the third step, they are offered a
“high intensity psychological intervention” (7) which comprises of either
cognitive behavioural therapy(CBT) or applied relaxation or they are given
sertraline as a pharmacological treatment. In the forth step they are then given
a combination. My research protocol is aimed at the third step and seeks to
determine which one of the two options (CBT or sertraline) is more effective in
treating GAD. This study is important because in determining this, and making
the more effective treatment the definitive third step, it may be possible to
stop the progression of GAD into the forth step where the patient is then
struggling with suicidal thoughts and thoughts of self-harming.

Research Method.

This will be a
randomized control trial with two intervention groups. These will be patients
that are given sertraline and patients that are given cognitive behavioural therapy.

The population
to be studied will be patients over the age of 18 who live in Norwich and have had
a GAD diagnosis for between 6 months to 2 years. These patients will be recruited
by looking through GP records in Norwich to find patients over 18 that have had
a GAD diagnosis for over 6 months but not more than 2 years. I have chosen 2 years
because I specifically want to see patients that have already been started on step
management for GAD. This time frame means that the patients retrieved in the
search are likely to have already been started on the step management for GAD
but are unlikely to have gone past step 2. Selected individuals will then be
sent a letter detailing the study and the purpose of the study. They will also
be sent an invitation to join the study and those that respond will be assessed
via the inclusion and exclusion criteria.

Inclusion criteria

Patients must be over the age of 18 as it is a
study in adults.

Patients must have a confirmed diagnosis of GAD
and must have failed to respond to step 2 management.

Patients must not have any other coexisting
mental health problems.

Patients must have their primary residence in

Exclusion criteria

1.       Patients
cannot have any other chronic organic disease that may be the cause of their

2.       Patients
cannot have been previously on a selective serotonin reuptake inhibitor (SSRI)
or have undergone CBT for GAD or other mental health issues.

3.       Patients
must not have any contraindications to the use of sertraline.

4.       Patients
that will not be present for the full duration of the trial (6 months).

The patients
will then be assessed by an independent psychologist as well as an independent
psychiatrist to see if they are medically fit to take part in the study. After
they have been cleared to take part in the study, they will then be given a
consent form to sign. Each consent form must read and signed by the individual.
This ensures that they know what will happen in the study and they are aware of
any potential adverse occurrences i.e. from taking sertraline. When this is
done, the patients will each then complete the Generalised Anxiety Disorder Assessment (GAD-7) questionnaire. This
will serve to give a baseline for the severity of the anxiety.

Anxiety Disorder Assessment

The diagnostic
criteria will be the GAD -7 Anxiety questionnaire. This is a questionnaire that
assesses the severity of the patient’s anxiety and how much it affects their
day to day life. It has 7 sections with each section having a maximum score of 3.
This adds up to 21 overall and gives a reflection of the severity of the
patient’s anxiety. 5 is the cut-off point for mild GAD; 10 for moderate and 15
for severe. (8)



As this will be
a double-blind trial with two interventions, it is important to eliminate as
many biases as possible. To do this, a random number generator will be used to
assign each participant a number. This will then be randomly shuffled digitally
to put the participants into one of two intervention groups.

Intervention groups

The study will
go on for 6 months/24 weeks. The two intervention groups will be as follows.

Group one: This
group will be given 25mg of sertraline daily for the first week. This will then
be increased to 50mg daily and then by 50mg (only if needed, as determined by
the psychiatrist) each subsequent week up to a maximum of 200mg daily. This dose
(or the maximum tolerated dose) will then be maintained until the trial ends. Any
adverse effects due to sertraline will result in the participant being made to
withdraw from the study and the drug being stopped. To account for patient
safety, the patients will be assessed weekly by an independent psychiatrist.

Group two: The
patients will have 24 sessions of CBT. These will be once weekly sessions that
last for one hour each. This will last for the duration of the trial.


Data collection

The participants
will complete the GAD-7 questionnaire every 2 months/8 weeks. This will make
for a total of 3 times altogether in the study. This will allow us to study any
improvements in the patient’s levels of anxiety at regulated intervals. I have
chosen an interval of 8 weeks as while sertraline can show effects in the first
2 weeks, it takes roughly 8 weeks for symptoms to show significant improvement.
(10) These results will then be collected by researchers who have no knowledge
of the allocation groups (blinded).

Date analysis

My primary
outcome measure is ‘a reduction in the level of anxiety as determined by a numerical
decrease in a participant’s GAD-7 score’.

Sample size

As my study is
looking at those over the age of 18, I need to deduct the people below the age
of 18 from the total population of Norwich.  The estimated total population of Norwich is
147,921 (11). Since roughly 20% (12) are below the age of 18, my total
population is approximately 118,337. As the prevalence of GAD is approximately 3%,
my population with a diagnosis of GAD is roughly 3550 people. Thus, with a
confidence level of 95% and a confidence interval of 5% my calculations show
that my sample size should be 347. Due to a high dropout rate, particularly in
studies relating to mental health acts where patients may not stick to their
assigned therapy, I am adding 20% to my sample size. This gives 416 (rounded
down from 416.4) total participants. These participants will then be split into
the two-intervention group with 208 participants in each group. The least number
of participants needed per group is 174 (173.5 rounded down). Having this
number of people will mean that my sample size is great enough to make sure
that my results are statistically significant.

The data
analysis will then be done via the T-test/T-score. This will allow for a comparison
between the GAD-7 scores of both groups at baseline and at weeks 8, 16 and 24
(months 2,4 and 6). This will then help us measure a difference in the mean
reduction (if there is any) of GAD-7 scores between the two groups at those
time periods. Results will be considered statistically significant if P is < 0.05. Ethics For this study, I will seek ethical approval from the relevant ethics committees. As both treatments are currently accepted management options for GAD, there should be no issues in this. The only true avenue for harm to occur is via adverse effects from sertraline and if this does occur, the participant will be withdrawn. Participants can also withdraw from the study at any time and additionally, all their information will be kept confidential within the research team. Exposure of their information to any other sources will require their permission in the form of another consent form. Every participant will be cleared by a psychiatrist to ensure that they fully understand the study and all gathered information will be protected as directed by the data protection act. Timeline My study spans 6 months. I chose this time frame because for GAD diagnosis, patients must have been experiencing symptoms for 6 months and thus I reason that recovery should be given the same time frame. In terms of patient recruitment, I believe that I will need an additional 2 months. This is not only to ensure that the needed number of people are recruited but also to ensure that they were correctly diagnosed with GAD and fit both the inclusion and exclusion criteria. The study will then begin, and they will be randomised. They will then complete the GAD questionnaire at this time and then subsequently at months 2,4 and 6. Follow up will then be done for an additional 12 months after the study is completed.   Cost The numerical cost is yet to be calculated. Expenses include: 1)      The cost of drugs 2)      Cost of the questionnaires. 3)      Cost of the staff including researchers, psychiatrists and psychologists. 4)      The cost of a location to supply the relevant therapies. 5)      The cost of the therapies themselves. 6)      The cost of data storage. Limitations of the study 1.       Questionnaires have inherent biases and thus are not the best way to gather information from the patients. Patients may lie in their responses, particularly as the study goes on, to be perceived better or to give a more favourable outcome to the researchers. 2.       Patients with mental health issues may have poor adherence to the administered therapy. 3.       Patients with mental health issues will have a higher dropout rate. 4.       GAD is a disease with a high association with other mental health conditions such as depression which may make it hard for my study's results to be purely applicable to GAD. 5.       The varied doses of sertraline make it difficult to determine a minimally effective dose for the drug or what drug dose gives the best effect with the lease amount of side effects. Practical implications of the study. This study will give a greater insight to health professionals about which therapy is more effective in step 3 management of GAD and which one should be the first line treatment in step 3. This could drastically improve outcomes for patients and stop their progression to step 4 of the management ladder.