Cost of Improving Access to Psychological Therapies (IAPT)
programme: an analysis of cost of session, treatment and recovery
in selected PCTs in East of England region
Muralikrishnan Radhakrishnan, PhD
Health Economist
Cambridge and Peterborough CLAHRC
Institute of Public Health,
University of Cambridge,
Forvie Site, Robinson Way,
Cambridge CB2 0SR
E-Mail: mrk38@medschl.cam.ac.uk
Geoffrey Hammond, PhD
IAPT Evaluation Research Associate
Cambridge and Peterborough CLAHRC
18 Trumpington Road
Cambridge CB2 8AH
E-Mail: Geoffrey.C.Hammond@gmail.com
Louise Lafortune,PhD
Senior Research Associate
Cambridge and Peterborough CLAHRC
Institute of Public Health,
University of Cambridge,
Forvie Site, Robinson Way,
Cambridge CB2 0SR
E-Mail: ll394@medschl.cam.ac.uk
Public Heath Theme
NIHR CLAHRC for Cambridgeshire and Peterborough
Cambridge
JUNE 2011
i
Acknowledgements
Thanks are due to Prof.Carol Brayne, Caroline Yvonne Lee and Dr.Christine Hill, Cambridge and
Peterborough CLAHRC, Institute of Public Health, University of Cambridge for their critical comments
on the study. Muralikrishnan Radhakrishnan’s and Louise Lafortune’s research was supported by the
Cambridge and Peterborough CLAHRC. Geoffrey Hammond’s research was supported by the
Strategic Health Authority of the East of England. The views expressed in this report are those of the
authors and may not reflect those of the funder.
Competing Interest
None declared
Ethical approval
The Cambridge and Peterborough CLAHRC waived the need for consent based on independent
evaluation which stated that the data were free of patient identifiable information and that the
research project proposed represented no deviation from normal clinical care and should be
deemed service evaluation.
Disclaimer
Papers published from the Cambridge and Peterborough CLAHRC are intended as a contribution to
current research. Work and ideas reported may not always represent the final position and as such
may sometimes need to be treated as work in progress. The material and views expressed in this
report are solely those of the authors and should not be interpreted as representing the collective
views of CLAHRC staff and their research funders.
ii
CONTENTS
Abstract………………………………………………………………………………………………………………………
iv
Introduction……………………………………………………………………………………………………………….
1
Methods…………………………………………………………………………………………………………………….
Cost per Session……………………………………………………………………………………………………………..
Cost of a Treatment Course……………………………………………………………………………………………
Cost per Recovered Patient…………………………………………………………………………………………….
Sensitivity Analysis………………………………………………………………………………………………………….
3
5
6
7
7
Results………………………………………………………………………………………………………………………..
Cost per Session………………………………………………………………………………………………………….....
Cost of a Treatment Course……………………………………………………………………………………………
Cost per Recovered Patient…………………………………………………………………………………………….
Sensitivity Analysis………………………………………………………………………………………………………….
8
10
11
12
13
Discussion……………………………………………………………………………………………………………………
Key Findings…….………………………………………………………………………………………………………….....
Implications……..………………………………………………………………………………………………………….....
Limitations……….………………………………………………………………………………………………………….....
Conclusion……….………………………………………………………………………………………………………….....
15
15
15
17
17
References………………………………………………………………………………………………………………….
18
Appendix…………………………………………………………………………………………………………………….
20
iii
ABSTRACT
Background
The Improving Access to Psychological Therapies (IAPT) programme is an important innovation in the
provision of mental health services in the UK. It was originally designed to increase access to
psychological therapies, thereby alleviating the burden and distress associated with anxiety and
depression. Recent literature has reported on improvements in clinical outcomes, changes in
employment status as a result of the programme, but not on the costs of the programme.
Aims
To estimate costs associated with a single session, completed course of treatment and recovery for
four treatment types in IAPT services in selected Primary Care Trusts in the East of England region.
Methods
The analysis used clinical outcomes for patients from 5 PCTs who had completed/ended treatment
between 2009 and 2010. The main measure of clinical outcome used was recovery at the end point
of treatment. The outcome of treatment were assessed in terms of changes in symptomatic
measures (severity of depression and anxiety symptoms) which were recorded using brief
questionnaires(PHQ-9 and GAD -7) required by the IAPT routine clinical dataset. Data regarding
financial & treatment activity were combined with data assessing clinical outcomes. Generalising a
few costing assumptions, the cost of session, treatment course and recovery were estimated.
Results
Of the 10789 patients who ended treatment between 2009 and 2010, 8464 patients attended 2 or
more sessions. Out of them, 4854 patients (57.3%) received only low intensity treatment, 2230
patients (26.2%) received only high intensity treatment. In addition, 252 patients (3.0%) had stepped
down (from high to low intensity) treatment and 1128(13.3%) had stepped up (from low to high)
intensity treatment. Overall, 3371 patients (39.8%) achieved recovery according to changes in
symptomatic measures as recorded by the PHQ-9 and GAD-7. Across all PCTs, the average cost of a
high intensity session was £176.97 and the average cost for a low intensity session was £98.59. The
average cost of treatment was £493(low intensity), £1,416(high intensity), £699 (stepped down) and
£1514 (stepped up). The cost per recovered patient was £1,043(low intensity), £ 2,895(high
intensity), £1,653(stepped down) and £2,914(stepped up). Sensitivity analysis reveals that the cost
per session, treatment and recovery is sensitive to the major assumption on cost ratio, indicating
that inaccurate ratios are likely to influence the overall estimates.
Conclusions
Results indicate that costs currently exceed previous estimates. Replication and additional analyses
along with evidence based discussion surrounding alternative, cost-effective methods of
intervention is recommended. It is likely that improvements in current IAPT practice cannot occur
until current practice is scrutinised and treatment approaches that are both effective and financially
viable are identified, studied, and highlighted.
iv
Introduction
Large, multinational epidemiological studies indicate that approximately 16% of the population
experience depression and anxiety over a lifetime (Kessler et al., 2003; Singleton et al., 2001). In the
UK, a 2006 report from Centre for Economic Performance (CEP) stated that “crippling depression
and chronic anxiety are the biggest causes of misery in Britain today” (CEP,2006). The financial cost
of depression in the UK was recently estimated at approximately 150 billion pounds in 2009/2010, of
which 30 billion is thought to be work related (Sainsbury Centre,2010). For the UK, it has been
estimated that a cross subsidy of £7-10 billion on social security benefits payments are made to
cover the unemployment costs of people with high prevalence mental health problems (CEP,2006).
In order to alleviate the distress and costs associated with anxiety and depression, the UK
Department of Health announced an unprecedented increase in funding for the provision of
psychological therapies in the National Health Services (DoH, 2007). With the possible exception of
the closure of asylums and an accompanying increase in community based mental health treatment
in the 1960s, the Improving Access to Psychological Therapies (IAPT) programme represents a major
policy directive in UK mental health services in the past 50 years. Developed on economic arguments
and clinical evidence base, IAPT was developed to promote access to National Institute for Health
and Clinical Excellence (NICE) - approved, Cognitive Behavioural Therapy (CBT) based talking
therapies treatments. IAPT represents a public health approach to the treatment of mild to
moderate depression and anxiety, common conditions judged to cause a preventable enormous
burden of morbidity and disability in the UK and worldwide (Layard,2006; Richards & Suckling R,
2009). IAPT services have been commissioned throughout England, with more than 300 new
therapists recruited for training in the East of England (EoE) alone between 2008 and 2011. By
substantially increasing the number of therapists, IAPT is intended to facilitate increased referrals
and reduced waiting times, with the potential to increase patient reported satisfaction and reduce
self-reported depression and anxiety (CEP,2006; Layard,2006). The funding for the IAPT programme
was contingent on the successful pilot implementation of a new treatment programme in two
demonstration sites: Doncaster in Northern England and Newham in East London. The outcomes of
these implementations would be used to argue that the increased funding from the Government
could deliver increased clinical outcomes in terms of magnitude and volume to justify the
investment (Richards & Suckling, 2009). A recent evaluation of both sites indicated that at least 55%
of patients who attended at least two sessions (including an assessment interview) recovered and
5% transitioned from unemployment into part or full-time employment (Clark et al, 2009). This study
generally agreed to demonstrate that the talking therapies implementation model can be effective
in the treatment of depression.
CBT is recommended by the NICE as an appropriate evidence-based psychological intervention for
depression and anxiety disorders (Clark et al, 2009). NICE also recommends a stepped-care approach
to the delivery of psychological therapies in mild or moderate depression and anxiety. The basic fact
of the stepped approach dictates that patients should receive treatment that is simultaneously the
least restrictive treatment option and the most likely to provide significant health gains. In the
talking therapies model, a two tier implementation is used: a high and a low intensity intervention is
available. Clark et al (2009) states that in the light of evidence, some individual patients respond
well to ‘low-intensity’ interventions (such as guided self-help and computerised CBT). In the case of
severe depression and some anxiety disorders such as post-traumatic stress, NICE guidelines
recommend that patients be provided ‘high-intensity’ face to face psychological therapy. Broadly,
these two types of therapies are provided by two types of therapists: ‘low-intensity’ and ‘highintensity’.
The rationale for widespread implementation of the IAPT programme, besides greatly decreasing the
problem of mental illness in UK, was that economic gains associated with increased productivity and
1
reemployment of those individuals treated would dwarf all costs associated with the programme
(Layard et al 2007). Layard et al (2007) state in their paper on economic costs that providing
psychological therapy to people not now in treatment would result in “the cost to the government
to be fully covered by the savings in incapacity benefits and extra taxes that result from more people
being able to work”. Layard et al (2007) also argued that these stepped care, expanded psychological
therapies programmes would cover the cost to the government by the extra output in GDP
produced by the treated person, savings to the exchequer in incapacity benefits and extra taxes as a
result of more people being able to work. The extra GDP produced by treating one person was
estimated at £1,100, the estimated extra earnings from 1.1 additional months of work over a two
year period. The savings to the exchequer in terms of incapacity benefits and extra taxes generated
is estimated to be around £ 9,000 per year or £18,000 for a two year period. Further, they also
estimated a conservative amount of £300 as a savings to the NHS per treated person over a two year
period. This gives a total benefit of £19,400 to society over a two year period for a person treated.
Adjusting for inflation rates from 2007 figures to 2010 year figures, the total estimated benefit to the
society per treated person is £20,989 or approximately £21,000 for two years (£10,500 per year).
The cost benefit calculation described in the paper assumed that the cost of providing a standard
course of roughly ten meetings of CBT is £750 or £75 per session. These unit costs were obtained
from a recognised national source (Curtis & Netten, 2006). Curtin & Netten (2006) provides a session
cost of £66, and NICE Guidelines for Post-traumatic stress disorder (PTSD) provide a session cost of
£82. Layard et al (2007) averaged these estimates to arrive at £75 per session (Layard et al 2007).
However, this estimate has yet to be tested empirically. Given the large disparity between the
estimates, there is a pressing need to estimate the costs using data from psychological therapies
programmes to assess what the cost really is.
The recent evaluation of the two pilot sites by Clark et al (2009) reported on clinical outcomes and
improvements in employment status but did not report on the associated costs of the programme.
The cost of the programme is an important consideration for psychological therapy implementation.
Given that these sessions are provided by high and low intensity therapists who differ significantly in
their training and salary costs, there could be a significant differences in the costs of the sessions
delivered by each group.
Unpublished data from the Bedfordshire Primary Care Trust (PCT) in the East of England region
reveals that ‘high intensity activity’ costs 1.8 times more than ‘low intensity activity’. Salaries
constituted at least 3/4 of the overall expenditure for high intensity and low intensity services,
indicating that IAPT sessions are human resource intensive. Given differences in the salaries
between high and low intensity employees, it becomes important to estimate specifically the cost
per single high intensity and low intensity session. The cost of a typical course of completed
treatment is also an important facet of cost to consider. Given that Clark et al (2009) reported the
percentage of patients who recovered, it is key to estimate the cost of a recovered patient through
the IAPT programme and its activities.
This study is an attempt to estimate the cost of session, completed treatment and recovered patient
using financial data from 5 PCTs in the East of England region. The study uses clinical outcomes,
financial information, and a few explicit assumptions to arrive at the costs. The methods, results,
limitations and implications for policy are discussed in the following sections.
2
Methods
This analysis is an attempt to cost low intensity and high intensity session, treatment course and
recovery provided by the IAPT programme in the East of England region. Broadly, two types of
therapies are provided (low and high intensity) by the IAPT programme. However, preliminary data
analysis revealed that a number of patients also ‘stepped down’ (i.e. transitioned from high to low
intensity) and ‘stepped up’ (, i.e. transitioned from low to high intensity) between the low and high
intensity sessions. Costs associated with these four types of treatment course and transitions were
estimated.
Costing was done from the IAPT programme perspective. The costs calculated were based solely on
costs associated with service budgets of IAPT programmes in the EoE; they did not include costs
incurred from a wider perspective (patient and other societal costs). The population of interest was
adult patients who were above 18 years and received treatment through the IAPT programme
provided by five PCTs in the region. For the analysis, financial data was combined with clinical
outcomes for 8,464 patients who had attended more than 2 treatment sessions in the programme
between 1 April 2009 to 31 March 2010. These patients include those who completed treatment,
deceased during the treatment, declined to continue with treatment, dropped out of treatment, or
were found unsuitable for treatment by IAPT therapists. Data on patient’s clinical progress and
outcome during the financial year specified above were collected on a sessional basis by high and
low intensity clinicians as part of routine clinical assessment. This information was stored on the
clinical record keeping software, PC-MIS, for all of the areas assessed. The primary clinical outcome
assessed was recovery at the end point of treatment. Secondary clinical outcomes were assessed
based on changes in symptom severity (severity of depression and anxiety symptoms) as recorded
using two brief questionnaires required for completion of the IAPT routine clinical dataset
(DoH,2008). These were the Patient Health Questionnaire Depression scale (PHQ-9) and the
Generalised Anxiety Disorder scale (GAD-7) (Appendix 1).
The Patient Health Questionnaire Depression scale (PHQ-9) has undergone validation in both the UK
and the US (Kroenke et al., 2001; Cameron et al., 2008). Nine questions enquire about a symptom
(or set of symptoms) of depression and are scored from 0 (“Not at all bothered by the problem”) to
3 (“Bothered nearly every day”). Sum scores range from 0 to 27. A score of 10 or above has been
identified as a potential clinical cut point for diagnosis of depression (Lowe et al., 2004); severity
bands are defined as follows: 0-4 not depressed, 5-9 mild depression, 10-14 moderate depression,
15-19 moderate/severe depression,and 20-27 severe.
The Generalised Anxiety Disorder scale (GAD-7) is a seven item measure of the severity of anxiety
symptoms (Spitzer et al., 2006). The measure uses the same response options and item scores as the
PHQ-9. Sum scores range from 0 to 21. A score of 8 or higher on the GAD-7 has been suggested as a
threshold for determining those with a probable clinical diagnosis of anxiety (Spitzer et al., 2006).
Severity bands are defined as follows: 0-4 not anxious, 5-9 mild anxiety, 10-14 moderate anxiety,
and 15-21 severe anxiety.
For the current analysis, those patients who fulfilled criteria for “caseness” at their initial session and
transitioned into “non caseness” at their final session were considered recovered, irrespective of
whether they completed treatment or had any other treatment outcome (e.g. declined treatment,
found unsuitable for further treatment) (Figure 1). To be classified as cases at baseline, these
individuals required to score at an initial assessment session, a clinical cut-off for the PHQ–9 of 10
points or more and the GAD-7 of 8 points or more. Individuals were classified as being cases if their
PHQ-9 or GAD-7 scores were above the stated thresholds and were required to be below this clinical
cut-off on both measures at the final treatment session to be classified as recovered. Recovery rates
3
were computed for those individuals who have ended IAPT treatment by dividing the number of
recovered cases by the number of cases treated.
Figure 1 - Categorization of patients as recovered using PHQ9 and GAD7 Scores
Patients referred
to the IAPT
Programme
Initial Assessment Session
Is the PHQ9
Score > 10 and
GAD7 score > 8 ?
NO
Classify as 'Non
Cases'
YES
Treatment
Final/Last Treatment Session
Is the PHQ9
Score < 10 and
GAD7 score < 8?
NO
Classify as
‘Not
Recovered'
YES
Classify as
Recovered'
Financial information on the total spend on IAPT programmes from five PCTs in the East of England
region were sourced for the 2009-2010 financial year (1 April 2009 to 31 March 2010) from reports
prepared by Mental Health Strategies for the Department of Health (Mental Health Strategies
4
2010)1. The PCTs included in the study were Bedfordshire, Cambridgeshire, North & East
Hertfordshire, Suffolk and West Hertfordshire. PCTs selected for inclusion had complete data on
IAPT sessions and clinical outcomes for the entire period assessed along with available information
on their service budget over the same period. Except for North & East Hertfordshire PCT (which
started IAPT operations in April 2009), all PCTs started their IAPT operations in September 2008.
Initially, the cost per session (high and low) was established, after which the cost of a completed
treatment course and cost per recovered patient for the four types of treatment course/transition
was estimated. The following subsections explain the methods and assumptions used to estimate
the different costs.
Cost Per Session
Financial information for each of the PCTs provided the overall total spend for the financial year: the
proportion of the total amount spent on low and high intensity sessions was not specified.
Therefore, a framework was developed in order to allocate the expenditure to either high or low
intensity activity (Figure 2).
The first step involved apportioning the total IAPT expenditure in a given year to the volume of high
or low intensity interventions delivered over the same period. Unpublished data from Bedfordshire
PCT revealed that ‘high intensity activity’ costs 1.8 times more than ‘low intensity activity’. A local
tariff developed for IAPT services in the EoE region also confirmed that the cost ratio was closer to
this estimate (Nolan, 2009). This cost ratio was stable even if the cost items, salaries and overheads,
were considered individually. Hence, it was assumed that high intensity sessions would cost 1.8
times more than the low intensity sessions in all of the reviewed PCTs. In order to arrive at the total
cost of high and low intensity sessions, the total session ratio was multiplied by the assumed cost
ratios to arrive at the percentage allocation of IAPT spend. The total spend was apportioned
according to the percentage allocation to arrive at the total costs for ‘high intensity’ and ‘low
intensity’ sessions. In order to estimate the per session costs, the total costs were then divided by
the total number of high and low intensity sessions delivered during the financial year specified
above.
Average duration for high intensity and low intensity sessions varied across PCTs. The cost per
session was also estimated using mean duration of sessions, to confirm that it did not vary from the
original estimates of cost per session. The mean duration was multiplied with the number of
sessions to arrive at the total duration for high and low intensity sessions. The respective total cost
was then divided by this total duration to arrive at the cost per minute for high and low intensity
sessions. The cost per minute was then multiplied with the respective mean duration to arrive at the
cost per session.
1
According to Mental Health Strategies (2010) report, the IAPT spend for the year 2009 -2010 includes both
the SHA contribution and matching funds by the PCT, except in the case of Cambridgeshire PCT, where only
the SHA contribution is provided. Personal communications with IAPT managers at Cambridgeshire PCT
revealed that an additional £810,000 has been spent as matched funds for IAPT.
5
Figure 2 – Framework for estimating cost of session, treatment and recovery
PCT wise Total Sessions
(2009 - 2010)
Multiply High Intensity Session
proportion by Cost Ratio *(1.8 )
Mulitply Low Intensity Session
proportion by Cost Ratio* (1 )
Estimate total High Intensity Session
cost (%)
Estimate total Low Intensity Session
cost (%)
Apportion total IAPT spend (2009 2010) to estimate Total Cost for High
Intensity Activity
Apportion total IAPT spend (2009 2010) to estimate Total Cost for Low
Intensity Activity
Cost Per Session
Divide the estimated Total Cost by
the number of High Intensity Session
(2009 -2010)
Cost Per Session
Divide the estimated Total Cost by
the number of Low Intensity Session
(2009 -2010)
Cost of Treatment
Multiply respective Cost per Session with the Median number of Sessions for
Low Intensity, High Intensity, Stepped Down & Stepped Up treatment types.
Cost of Recovery
Estimate Total Cost : Multiply Cost of Treatment with Number of Patients for
each treatment type ( includes those who completed, deceased, declined,
dropped out and not suitable).
Estimate Cost Per Recovered Patient : Divide the Total Cost by Number of
Patients who recovered for each treatment type ( includes those who completed,
deceased, declined, dropped out and not suitable).
* Cost Ratio based on Bedforshire PCT costing exercise
Cost of a Treatment Course
Using the per session cost estimate, the cost for a single course of treatment delivered by different
IAPT providers was estimated (Figure 2). For this, all patients from the 5 PCTs who finished
treatment regardless of the reason for ending treatment (i.e. those who completed treatment,
deceased, declined, dropped out or were found not suitable for IAPT ) between 1 April 2009 to 31
March 2010 were included in the analysis. Each patient’s entire treatment course was traced back to
their initial session to ascertain the number of sessions they received, including the initial
assessment session. The main group of interest were those who completed treatment, since their
cost represents the cost of successfully completed course of prescribed treatment.
The median number of sessions received during the course of treatment was assessed separately for
four groups of patients: those who remained in high intensity, those who remained in low intensity,
those who transitioned from low to high intensity (stepped up) and those who transitioned from
high to low intensity (stepped down). The median rather than the mean was used in further
calculations due to the fact that sessional data was not normally distributed. The median number of
sessions was then multiplied by the session cost associated with the corresponding PCT to produce
the cost of a single course of treatment.
6
The cost of a single course of treatment was calculated primarily for those patients who successfully
completed the prescribed treatment sessions. The median number of sessions along with their
interquartile range (25th and 75th percentile) was also estimated to provide a range for the costs.
The cost of treating patients in the other categories like deceased, declined, dropped out and not
suitable were calculated using the method outlined above. In addition, the cost of providing the
single session (usually assessment) to those who attended only one session was also estimated by
multiplying the respective session cost with the number of patients. These patients were assessed
either by a low intensity or high intensity therapist. In order to apportion these costs between the
four treatment types, they were allocated on the basis of total treated patient proportions. These
costs were calculated in order to estimate the total costs of treatment and to arrive at the cost per
recovered patient, as described in the following section.
Cost per Recovered Patient
To estimate the cost per recovered patient, two estimates were required: the total cost of treating
all the patients and the number of patients who recovered, in all categories (completed deceased,
declined, dropped out and not suitable). The total cost of treatment was calculated by multiplying
the cost of treatment course with the number of patients who underwent treatment. The total cost
was then divided by the number of patients who recovered to arrive at the cost per recovered
patient for each PCT for the four different treatment courses (Figure 2).
Sensitivity Analysis
For a base case estimate, it is assumed that that high intensity sessions would cost 1.8 times more
than the low intensity sessions. To check for robustness of results, the cost ratio was varied, using
low to high ratio estimates of 1.6 and 2 respectively. The main intention of the sensitivity analysis is
to report the extent to which the cost per session, treatment and recovery changes if the cost ratios
are varied.
7
Results
In total, 10,789 patients completed or ended IAPT treatment for various reasons (Figure 3) between
1 April 2009 - 31 March 2010, of whom 21.2% attended only 1 session. Among those who attended 2
or more sessions, 4844 (44.9%) successfully completed the allotted treatment. It is to be noted that
a significant number of patients dropped out of treatment (n=1961; 18.2%) and 861 (8%) were found
unsuitable for IAPT treatment after attending 2 or more sessions.
Figure 3 – Distribution of Patients who ended IAPT treatment (2009-2010)
Patients Ending
Treatment(2009-10)
10789(100%)
Attended 1 or
more Sessions
10746(99.6%)
Attended 2 or more
Sessions
8464(78.5%)
Completed Treatment
4844(44.9%)
Deceased
12 (0.1%)
Attended No
Sessions
43(0.4%)
Attended only 1
Session
2282(21.2%)
Declined
786(7.3%)
Dropped Out
1961(18.2%)
Not Suitable
861(8.0%)
The baseline demographic and clinical characteristics of the 8464 patients who attended 2 or more
sessions are presented in Table 1. 4325 patients (51.1%) belonged to the age group 26 -45 years; the
majority of patients (5576, 65.9%) were women. In full time or part time employment at baseline
assessment were 5155 patients (60.9%). Referrals to the IAPT programme were through GPs for
7511 patients (88.7%). Patients from Suffolk PCT constituted a majority of the sample (31.9%),
followed by Cambridgeshire PCT (20.4%), Bedfordshire PCT (19.3%), West Hertfordshire PCT (17.6%)
and North & East Hertfordshire PCT (10.8%). Based on baseline PHQ-9 scores, 2146 patients (25.4%)
were classified as having moderate depressive symptoms, 1987 patients (23.5%) with moderatesevere depressive symptoms and 1787 patients (21.1%) with severe depressive symptoms. On the
GAD 7 scores at baseline, 2365 patients (27.9%) were classified with moderate anxietal symptoms
and 3348 patients (39.6%) with severe anxietal symptoms.
8
Table 1 - Baseline Demographics and Clinical Characteristics of Patients
Sample Size
Age Group
<18-25
26-35
36-45
46-55
56-65
Older than 65
Gender
Men
Women
Unknown, Not Specified
Missing
Employment Status
Employment- Full-Time
Employment- Part-Time
Employment- Unemployed
Employment- Inactive
Missing
Referral Source
GP referral
Self-referral
Any specialty
Other
Missing
PCT
Bedfordshire
Cambridgeshire
North & East Hertfordshire
Suffolk
West Hertfordshire
PHQ-9
No Impairment(0-5)
Mild Impairment(5-9)
Moderate Impairment(10-14)
Moderate-Severe Impairment(15-19)
Severe Impairment(20-27)
Missing
GAD - 7
No Impairment(0-5)
Mild Impairment(5-9)
Moderate Impairment(10-14)
Severe Impairment(15-21)
Missing
Table 1 - Baseline Demographics and clinical characteristics of patients
Completed Deceased Declined Dropped Out Not Suitable
Total
n
%
n
%
n
%
n
%
n
%
n
%
4844 57.2% 12 0.1% 786 9%
1961 23% 861 10% 8464 100
635
1125
1320
957
589
218
13.1%
23.2%
27.3%
19.8%
12.2%
4.5%
0 0.0%
2 16.7%
1 8.3%
2 16.7%
6 50.0%
1 8.3%
144
189
190
134
98
31
18.3%
24.0%
24.2%
17.0%
12.5%
3.9%
1572 32.5%
3211 66.3%
9 0.2%
52 1.1%
2 16.7%
8 66.7%
1 8.3%
1 8.3%
255 32.4%
518 65.9%
5 0.6%
8 1.0%
2197
920
694
1009
24
45.4%
19.0%
14.3%
20.8%
0.5%
3 25.0%
1 8.3%
5 41.7%
3 25.0%
0 0.0%
288
143
159
177
19
4314 89.1%
125 2.6%
228 4.7%
156 3.2%
21 0.4%
11 91.7%
0 0.0%
0 0.0%
1 8.3%
0 0.0%
36.6%
18.2%
20.2%
22.5%
2.4%
707 89.9%
24 3.1%
22 2.8%
28 3.6%
5 0.6%
437
560
523
307
109
25
22.3%
28.6%
26.7%
15.7%
5.6%
1.3%
652 33.2%
1287 65.6%
3 0.2%
19 1.0%
836
361
411
339
14
42.6%
18.4%
21.0%
17.3%
0.7%
1743 88.9%
51 2.6%
83 4.2%
77 3.9%
7 0.4%
185
183
232
155
75
31
21.5%
21.3%
26.9%
18.0%
8.7%
3.6%
1401
2059
2266
1555
877
306
16.6%
24.3%
26.8%
18.4%
10.4%
3.6%
295 34.3% 2776 32.8%
552 64.1% 5576 65.9%
2 0.2% 20
0.2%
12 1.4% 92
1.1%
273
133
267
174
14
31.7%
15.4%
31.0%
20.2%
1.6%
3597
1558
1536
1702
71
42.5%
18.4%
18.1%
20.1%
0.8%
736 85.5% 7511 88.7%
37 4.3% 237 2.8%
48 5.6% 381 4.5%
36 4.2% 298 3.5%
4 0.5% 37
0.4%
896
1025
525
1476
922
18.5%
21.2%
10.8%
30.5%
19.0%
2
1
4
2
3
16.7%
8.3%
33.3%
16.7%
25.0%
96
172
68
334
116
12.2%
21.9%
8.7%
42.5%
14.8%
421
351
232
638
319
21.5%
17.9%
11.8%
32.5%
16.3%
217
177
88
253
126
25.2%
20.6%
10.2%
29.4%
14.6%
1632
1726
917
2703
1486
19.3%
20.4%
10.8%
31.9%
17.6%
629
1057
1286
1047
820
5
13.0%
21.8%
26.5%
21.6%
16.9%
0.1%
1
3
4
2
2
0
8.3%
25.0%
33.3%
16.7%
16.7%
0.0%
66
137
191
195
188
9
8.4%
17.4%
24.3%
24.8%
23.9%
1.1%
148
316
505
528
459
5
7.5%
16.1%
25.8%
26.9%
23.4%
0.3%
47
114
160
215
318
7
5.5%
13.2%
18.6%
25.0%
36.9%
0.8%
891
1627
2146
1987
1787
26
10.5%
19.2%
25.4%
23.5%
21.1%
0.3%
540
1206
1379
1714
5
11.1%
24.9%
28.5%
35.4%
0.1%
1 8.3%
3 25.0%
4 33.3%
4 33.3%
0 0.0%
59 7.5%
157 20.0%
234 29.8%
327 41.6%
9 1.1%
124 6.3%
412 21.0%
539 27.5%
880 44.9%
6 0.3%
51 5.9% 775 9.2%
170 19.7% 1948 23.0%
209 24.3% 2365 27.9%
423 49.1% 3348 39.6%
8 0.9% 28
0.3%
Table 2 presents the number of patients who were treated and who recovered using the clinical
definition of recovery in each PCT (PHQ9 Score < 9 and GAD7 score < 8 at the final/last session). The
corresponding recovery rate is also computed. Of the 8464 patients treated, 4854 patients (57.3%)
received only low intensity treatment, 2230 patients (26.2%) received only high intensity treatment,
252 patients (3.0%) were stepped down in their treatment and 1128 (13.3%) were stepped up. Of
the 3371 patients (39.8%) who achieved recovery, 1992 patients (59.1%) received low intensity
treatment exclusively, 884 patients (26.2%) received high intensity treatment exclusively, 101
9
patients (3.0%) were stepped down and 394 patients (11.7%) were stepped up. Across all PCTs, a
recovery rate of 55.7% was observed by individuals completing treatment after receiving exclusively
low intensity therapy. The recovery rate was 54.4% among those who completed high intensity
treatment, 54.9% among those who completed treatment but stepped down and 52.9% among
those who completed treatment and stepped up. Recovery rate were in excess of around 20%
among those who dropped treatment among all the therapy types. For those who declined
treatment, the recovery rate was in the range of 14% to 25% among the four treatment types. For
those found not suitable for treatment, the recovery rate was less than 10% for the low intensity,
high intensity and stepped up treatments and 17.2% for the stepped down treatment.
Table 2 – Distribution of Patients Treated, Recovered and Recovery Rates
Bedfordshire
Cambridgeshire
North & East Hertfordshire
Treated Recovered Recovery Treated Recovered Recovery Treated Recovered Recovery Treated
(n)
(n)
Rate (%)
(n)
(n)
Rate (%)
(n)
(n)
Rate (%)
(n)
Low Intensity
Completed Treatment
Deceased
Declined
Dropped Out
Not Suitable
Total
High Intensity
Completed Treatment
Deceased
Declined
Dropped Out
Not Suitable
Total
Stepped Down(High to Low)
Completed Treatment
Deceased
Declined
Dropped Out
Not Suitable
Total
Stepped Up(Low to High)
Completed Treatment
Deceased
Declined
Dropped Out
Not Suitable
Total
Suffolk
West Hertfordshire
Total
Recovered Recovery Treated Recovered Recovery Treated Recovered Recovery
(n)
Rate (%)
(n)
(n)
Rate (%)
(n)
(n)
Rate (%)
542
1
53
251
139
986
320
0
16
71
8
415
59.0%
0.0%
30.2%
28.3%
5.8%
42.1%
774
1
121
258
96
1250
390
1
20
67
11
489
50.4%
100.0%
16.5%
26.0%
11.5%
39.1%
232
0
31
89
29
381
122
0
7
17
0
146
52.6%
0.0%
22.6%
19.1%
0.0%
38.3%
728
1
198
376
130
1433
442
0
30
113
19
604
60.7%
0.0%
15.2%
30.1%
14.6%
42.1%
507
0
53
181
63
804
276
0
10
48
4
338
54.4%
0.0%
18.9%
26.5%
6.3%
42.0%
2783
3
456
1155
457
4854
1550
1
83
316
42
1992
55.7%
33.3%
18.2%
27.4%
9.2%
41.0%
277
0
27
125
50
479
165
0
7
25
4
201
59.6%
0.0%
25.9%
20.0%
8.0%
42.0%
128
0
17
51
27
223
60
0
3
9
3
75
46.9%
0.0%
17.6%
17.6%
11.1%
33.6%
182
1
26
99
41
349
93
0
5
20
5
123
51.1%
0.0%
19.2%
20.2%
12.2%
35.2%
438
0
61
155
59
713
261
0
13
42
3
319
59.6%
0.0%
21.3%
27.1%
5.1%
44.7%
299
3
40
88
36
466
141
1
1
17
6
166
47.2%
33.3%
2.5%
19.3%
16.7%
35.6%
1324
4
171
518
213
2230
720
1
29
113
21
884
54.4%
25.0%
17.0%
21.8%
9.9%
39.6%
13
0
2
7
4
26
6
0
1
3
0
10
46.2%
0.0%
50.0%
42.9%
0.0%
38.5%
38
0
8
10
11
67
19
0
1
1
2
23
50.0%
0.0%
12.5%
10.0%
18.2%
34.3%
41
0
2
19
6
68
22
0
0
6
2
30
53.7%
0.0%
0.0%
31.6%
33.3%
44.1%
28
0
5
7
5
45
19
0
2
1
1
23
67.9%
0.0%
40.0%
14.3%
20.0%
51.1%
22
0
7
14
3
46
12
0
2
1
0
15
54.5%
0.0%
28.6%
7.1%
0.0%
32.6%
142
0
24
57
29
252
78
0
6
12
5
101
54.9%
0.0%
25.0%
21.1%
17.2%
40.1%
64
1
14
38
24
141
39
0
4
8
2
53
60.9%
0.0%
28.6%
21.1%
8.3%
37.6%
85
0
26
32
43
186
39
0
4
3
2
48
45.9%
0.0%
15.4%
9.4%
4.7%
25.8%
70
3
9
25
12
119
33
0
2
4
1
40
47.1%
0.0%
22.2%
16.0%
8.3%
33.6%
282
1
70
100
59
512
161
0
7
21
6
195
57.1%
0.0%
10.0%
21.0%
10.2%
38.1%
94
0
16
36
24
170
43
0
2
9
4
58
45.7%
0.0%
12.5%
25.0%
16.7%
34.1%
595
5
135
231
162
1128
315
0
19
45
15
394
52.9%
0.0%
14.1%
19.5%
9.3%
34.9%
Cost per Session
Table 3 provides the total IAPT spend in each of the five PCTs assessed, the total session ratio
(proportion of low and high intensity sessions), and the proportion of the total cost allocated to high
and low intensity activity for the 2009-2010 financial year. All the PCTs spent in excess of £ 2 million
on the IAPT programme, with Suffolk PCT spending the most (approximately £ 4 million).
In terms of the session ratio, Cambridgeshire PCT provided the highest proportion of low intensity
sessions (59.3%) whereas North & East Hertfordshire PCT provided the lowest proportion (46.9%).
For the high intensity sessions ratio, Suffolk PCT had the highest (54.9%) and Cambridgeshire PCT
had the lowest (40.7%). In terms of the actual number of sessions, Suffolk PCT delivered the most,
with a total of 35,936 sessions and North & East Hertfordshire PCT delivered the fewest (11,780
sessions). The low number of sessions at North & East Hertfordshire PCT is explained by the fact that
the provider started providing services in April 2009, whereas all the other PCTs started their
operations in September 2008.
High intensity sessions account for more than 60% of total costs for all of the PCTs studies except
Cambridgeshire, where it accounts for 55%. As expected, this translates into significantly higher total
spent on high intensity activity. The significantly higher total spent for Suffolk is because they
performed the highest number of individual sessions.
Across all PCTs, the average cost for a low intensity session was £98.59 and it ranged from £78.31
(Suffolk PCT) to £150.17 (North & East Hertfordshire PCT). The average cost of a high intensity
session was £176.97, and ranged from £140 (Suffolk PCT) to £270.41 (North & East Hertfordshire
PCT) across PCTs. The overall cost per session was £ 137.73 for all PCTs, and ranged from
£112.70(Suffolk PCT) to £ 214.01 (North and East Hertfordshire PCT).
10
Table 3 – Total Spend, Sessions, Cost and Cost per Session of High and Low Intensity Activity (2009 -2010)
PCT
Total Spend in £'s
Bedfordshire
Cambridgeshire
North & East Hertfordshire
Suffolk
West Hertfordshire
All PCTs
2,029,000
2,188,000
2,521,000
4,050,000
2,647,000
13,435,000
Total Sessions(Ratio)
Total Cost in £'s (%)
High Intensity Low Intensity
7875(0.485)
8357(0.515)
7436(0.407) 10838(0.593)
6254(0.531)
5526(0.469)
19729(0.549) 16207(0.451)
7415(0.484)
7912(0.516)
48711(0.499) 48838(0.501)
High Intensity
Low Intensity
1,276,165(62.9%) 752,835(37.1%)
1,209,210(55.3%) 978,790(44.7%)
1,691,165(67.1%) 829,835(32.9%)
2,780,857(68.7%) 1,269,143(31.3%)
1,662,389(62.8%) 984,611(37.2%)
8,619,787(64.2%) 4,815,213(35.8%)
Cost per Session(£)
High
Low
Intensity Intensity All
162.05
90.09 125.00
162.61
90.31 119.73
270.41 150.17 214.01
140.95
78.31 112.70
224.18 124.45 172.70
176.97
98.59 137.73
The cost per session for high intensity and low intensity was also estimated using the mean
durations (Table 4) and the results show that the estimated cost per session is similar to previous
estimates.
Table 4 – Estimated Cost per Session using Mean Session Duration
PCT
Mean Duration(Minutes)
Total Duration(Minutes)
Cost per Minute(£)
Cost per Session(£)
High Intensity Low Intensity High Intensity Low Intensity High Intensity Low Intensity High Intensity Low Intensity
Bedfordshire
47.10
36.50
370913
305031
3.44
2.47
162.05
90.08
Cambridgeshire
57.20
47.30
425339
512637
2.84
1.91
162.62
90.31
North & East Hertfordshire
56.60
52.90
353976
292325
4.78
2.84
270.41
150.17
Suffolk
56.90
38.80
1122580
628832
2.48
2.02
140.95
78.31
West Hertfordshire
57.20
47.80
424138
378194
3.92
2.60
224.19
124.45
Total Duration = No.of Sessions (Table 3) X Mean Duration
Cost per Minute = Total Costs (Table 3) / Total Duration
Cost per Session = Cost per Minute X Mean Duration
Cost of a Course of Treatment
The median number of sessions in a given course of treatment for each treatment type (high or low,
stepped up or down) is given in Table 4. For all the PCTs combined, the median number of sessions
received was 5 for individuals remaining on low intensity treatment and 8 for individuals remaining
on high intensity treatment. People who stepped down from high to low intensity sessions received
a median of 3.5 sessions of low intensity interventions and 2 sessions high intensity sessions. For the
groups who initiated treatment with low intensity intervention and were stepped up, they received
approximately 1 low intensity session and 8 high intensity sessions, suggesting that most individuals
were stepped up after initial assessment. However, substantial variation in the median number of
sessions across PCTs was observed. For instance, Bedfordshire PCT and Cambridgeshire PCT
provided 10 sessions for high intensity treatment compared to other PCTs who provided in the range
of 6-8 sessions.
Table 4 presents the cost of treatment for the four different treatment types across PCTs. For all
PCTs, the average estimated cost for a course of exclusively low intensity treatment was £493 and
fluctuated between £313 (Suffolk PCT) and £901 (North & East Hertfordshire PCT). The estimated
cost of a single course of high intensity treatment was £1,416 and ranged from £987 (Suffolk PCT) to
£1,793 (West Hertfordshire PCT). The average cost of a treatment course for individuals who were
stepped down was £699 and ranged from £579 (Suffolk PCT) to £946 (West Hertfordshire PCT). The
average cost of a treatment course for individuals who were stepped up was £1514 and ranged from
£1206 (Suffolk PCT) to £1891 (Bedfordshire PCT). Stepping up and down treatments cost more than
low intensity and high intensity treatments.
11
Table 4 – Median Number of sessions and Cost of Treatment
Activity
Median
Low Intensity 25th Percentile
75th Percentile
Median
High Intensity 25th Percentile
75th Percentile
Median
Stepped Down
25th Percentile
(High to Low)
75th Percentile
Median
Stepped Up
25th Percentile
(Low to High)
75th Percentile
Bedfordshire
Cambridgeshire
North & East Hertfordshire
Suffolk
West Hertfordshire
ALL
Low
High
Cost of
Low
High
Cost of
Low
High
Cost of
Low
High
Cost of
Low
High
Cost of
Low
High
Cost of
Intensity Intensity Treatment Intensity Intensity Treatment Intensity Intensity Treatment Intensity Intensity Treatment Intensity Intensity Treatment Intensity Intensity Treatment
Sessions Sessions
(£)
Sessions Sessions
(£)
Sessions Sessions
(£)
Sessions Sessions
(£)
Sessions Sessions
(£)
Sessions Sessions
(£)
5
0
450
5
0
452
6
0
901
4
0
313
6
0
747
5
0
493
4
0
360
3
0
271
3
0
451
3
0
235
4
0
498
3
0
296
6
0
541
6
0
542
8
0
1,201
6
0
470
8
0
996
6
0
592
0
10
1,620
0
10
1,626
0
6
1,622
0
7
987
0
8
1,793
0
8
1,416
0
7
1,134
0
6
976
0
4
1,082
0
5
705
0
5
1,121
0
5
885
0
14
2,269
0
14.5
2,358
0
9
2,434
0
13
1,832
0
12
2,690
0
13
2,301
5
2
775 3.5
2.5
723
2
2
841
2
3
579
4
2
946 3.5
2
699
3
1
432
1
1
253
1
1
421
1
1.5
290
2
1
473
1
1
276
6
6
1,513
5
7
1,590
4
5
1,953
5
8
1,519
7
3
1,544
5
6
1,555
3
10
1,891
2
8
1,481
1
5
1,502
1
8
1,206
1
6.5
1,582
1
8
1,514
2
7
1,315
1
5
903
1
4
1,232
1
5
783
1
3
797
1
5
983
5
13
2,557
5
12
2,403
1
9
2,584
3
12
1,926
2
10
2,491
3
12
2,419
The cost of treatment for those in the deceased, declined, dropped out and not suitable categories
were calculated using similar methods and are presented in Appendix 1.
Cost per Recovered Patient
The total cost of treating those who completed treatment are calculated by multiplying the number
of patients treated (Table 2) with the respective cost of a single course of treatment (Table 4). The
cost of treating patients from the deceased, declined, dropped, not suitable categories and
assessment (those who had only one session) was also estimated using a similar method. The overall
total cost of treating all categories of patients under different therapy types are given in Table 5.
Irrespective of therapy received, those who completed treatment constituted the majority of the
total IAPT expenditure (over 70%). However, fluctuation was observed across PCTs. At £2,559,111,
high intensity therapy is the highest single total cost for across all treatment types for all PCTs. This
pattern of high intensity treatment being associated with the highest cost is observed for all
individual PCTs except Cambridgeshire PCT, where low intensity therapy ranks highest among the
total costs. This is expected since Cambridgeshire PCT provided the highest proportion of low
intensity sessions (59.3%) In terms of individual PCT cost, West Hertfordshire PCT has the highest
cost of £504,533 for low intensity therapy, and £674,970 for high intensity therapy. North & East
Hertfordshire PCT has the highest cost for stepped down treatment (£55,115) while Suffolk PCT has
the highest cost for stepped up treatment (£456,425).
From the number of patients who recovered (Table 2) and the total costs (Table 5), the cost per
recovered patient is calculated and is presented in Table 6. For the low intensity therapy, the
estimated cost per recovered patient is £1,043 for all PCTs (ranging from £687 (Suffolk PCT) to
£1,952 (North & East Hertfordshire PCT)). For the high intensity therapy, the cost per recovered
patient is £ 2,895 and varied significantly across PCTs (from £1,849 (Suffolk PCT) to £4,066 (West
Hertfordshire PCT)). For the stepped down therapy, the estimated cost per recovered patient is
£1,653 and ranges from £1,012 (Suffolk PCT) to £2,111 (West Hertfordshire PCT). For the stepped up
therapy, the cost per recovered patient is £2,914 and ranged from £2,341 (Suffolk PCT) to £ 3,629
(Cambridgeshire PCT). Stepping up therapy yields the highest cost per recovered patient, followed
by remaining in high intensity, stepping down and remaining in low intensity treatment. Overall,
including all therapy types, the cost per recovered patient is £1,766 and varies from £1,301 (Suffolk
PCT) to £2,773 (North & East Hertfordshire PCT).
12
Table 5 – PCT wise Total Cost of Treatment
Bedfordshire
Low Intensity
Completed Treatment
Deceased
Declined
Dropped Out
Not Suitable
Assessment
Total
High Intensity
Completed Treatment
Deceased
Declined
Dropped Out
Not Suitable
Assessment
Total
Stepped Down(High to Low)
Completed Treatment
Deceased
Declined
Dropped Out
Not Suitable
Assessment
Total
Stepped Up(Low to High)
Completed Treatment
Deceased
Declined
Dropped Out
Not Suitable
Assessment
Total
%
Cambridgeshire
North & East
Hertfordshire
%
%
Suffolk
%
West
Hertfordshire
%
Total
%
244,137
541
14,324
67,836
25,044
18,601
370,482
66%
0%
4%
18%
7%
5%
100%
349,519
361
21,856
69,904
26,011
35,771
503,421
69%
0%
4%
14%
5%
7%
100%
209,043
9,311
40,097
13,065
13,503
285,018
73%
0%
3%
14%
5%
5%
100%
228,038
157
31,011
88,333
25,451
41,887
414,876
55%
0%
7%
21%
6%
10%
100%
378,579
19,788
67,577
15,681
22,909
504,533
75% 1,409,315
0%
1,058
4%
96,289
13%
333,746
3%
105,252
5%
132,670
100% 2,078,331
68%
0%
5%
16%
5%
6%
100%
448,870
13,126
81,023
16,205
13,987
573,211
78%
0%
2%
14%
3%
2%
100%
208,135
11,057
24,879
8,781
10,753
263,604
79%
0%
4%
9%
3%
4%
100%
295,283
541
21,092
80,310
33,260
26,931
457,416
65%
0%
5%
18%
7%
6%
100%
432,155
25,794
87,389
24,948
19,622
589,909
73%
0%
4%
15%
4%
3%
100%
536,243
1,345
17,935
78,912
24,212
16,323
674,970
79% 1,920,686
0%
1,886
3%
89,003
12%
352,514
4%
107,405
2%
87,617
100% 2,559,111
75%
0%
3%
14%
4%
3%
100%
10,069
666
3,026
2,090
759
16,610
61%
0%
4%
18%
13%
5%
100%
27,459
2,385
3,432
3,776
3,231
40,283
68%
0%
6%
9%
9%
8%
100%
34,488
2,013
10,844
2,523
5,247
55,115
63%
0%
4%
20%
5%
10%
100%
16,225
2,193
2,521
1,096
1,238
23,274
70%
0%
9%
11%
5%
5%
100%
20,816
2,440
5,752
1,046
1,611
31,665
66%
0%
8%
18%
3%
5%
100%
109,057
9,697
25,576
10,531
12,087
166,947
65%
0%
6%
15%
6%
7%
100%
121,007
414
7,314
37,980
12,102
2,660
181,477
67%
0%
4%
21%
7%
1%
100%
125,925
6,576
18,500
17,868
5,323
174,191
72%
0%
4%
11%
10%
3%
100%
105,154
5,318
3,785
17,275
6,669
4,217
142,419
74%
4%
3%
12%
5%
3%
100%
340,068
579
15,348
64,211
21,252
14,966
456,425
75%
0%
3%
14%
5%
3%
100%
148,674
5,578
20,621
13,748
4,844
193,464
77%
840,828
0%
6,311
3%
38,601
11%
158,587
7%
71,639
3%
32,010
100% 1,147,977
73%
1%
3%
14%
6%
3%
100%
Table 6 – Cost per Recovered Patient
Low Intensity
High Intensity
Stepped Down(High to Low)
Stepped Up(Low to High)
Overall
North & East
West
Bedfordshire Cambridgeshire Hertfordshire Suffolk Hertfordshire All PCTs
893
1,029
1,952
687
1,493
1,043
2,852
3,515
3,719
1,849
4,066
2,895
1,661
1,751
1,837
1,012
2,111
1,653
3,424
3,629
3,560
2,341
3,336
2,914
1,682
1,546
2,773
1,301
2,434
1,766
Sensitivity Analysis
An important assumption used to estimate the costs was that high intensity sessions cost 1.8 times
more than the low intensity sessions. To check for the sensitivity of the cost ratio, it was varied
between a lower costing ratio of 1.6 and higher one 2.0. The results are presented in Table 7. As the
cost ratio is reduced to 1.6 from the base case of 1.8 the cost per session, treatment and recovery
for low intensity session increases between 7.1 - 7.6% and decreases for the high intensity session by
4.1 - 4.3%. As the cost ratio is increased to 2.0 times from the base case, the reverse is observed the cost per session, treatment and recovery for a high intensity session increases by 3.7% and
decreases for a low intensity session by around 6.5%. A similar pattern is observed with stepping
patterns, although the fluctuation in costing as a result of changes in the ratio represents 3.5% or
less of the original estimated cost in any scenario. The sensitivity analysis reveals that the cost per
session, treatment and recovered patients is sensitive to this assumption, indicating that inaccurate
ratios are likely to influence the overall estimates of session, and treatment course costs.
13
Table 7 - Sensitivity Analysis of Cost Ratio (All PCTs)
Cost Ratio
Cost per Session
Low Intensity
High Intensity
BaseCase
Low
1.8
1.6
%
Change
High
%
Change
2
99
177
106
169
7.6
-4.3
92
183
-6.6
3.7
Cost of Treatment (Completed)
Low Intensity
High Intensity
Stepped Down
Stepped Up
493
1,416
699
1,514
530
1,356
710
1,462
7.6
-4.3
1.6
-3.5
460
1,468
689
1,560
-6.6
3.7
-1.4
3.0
Cost per Recovered Patient
Low Intensity
High Intensity
Stepped Down
Stepped Up
1,043
2,895
1,653
2,914
1,117
2,775
1,665
2,836
7.1
-4.1
0.7
-2.7
980
2,999
1,642
2,981
-6.5
3.7
-0.7
2.4
14
Discussion
Key Findings
This analysis is, to the knowledge of the authors, one of the first costing estimates associated with
implementation of psychologically based talking therapies in the UK. Reported information
concerning public sector spending on psychologically based talk therapies were combined with
clinician reported estimates of session level activity to arrive at an estimate of session costs. These
session costs were then expanded to assess costs associated with a treatment course under a
number of different treatment conditions (remaining in low or high intensity treatment, stepping up
or down) and treatment outcomes. Although the IAPT service has been found to be highly effective
in reducing symptoms of anxiety and depression, the extent to which it is cost-effective in roll out of
a whole programme has not been demonstrated or discussed in the current evidence base. Costing
up treatment under these conditions is important step to assessing the cost-effectiveness of the
implementations along with an important consideration in future service design and planning.
Layard et al (2007) estimated that the cost of providing a standard course of roughly ten meetings of
CBT is £750 or £75 per session. However, these costing were based on an average session cost of
treatment in community mental health teams and treatment in a specialist PTSD clinic, neither of
which implement the talking therapies organisation model (i.e. stepped care and two tier
workforce). Additional government estimates of cost based on a capacity model of activity (assuming
a maximum number of treatment interventions based on number of hours, days in a working year)
have estimated the high and low intensity session costs to be approximately £87 and £27,
respectively, with an average cost of 58 pounds.
Given the results above (low intensity cost of approximately £100; high intensity session cost
approximately £175), 75 or 58 pounds per session would seem to be a significant underestimation of
the cost of treatment as currently provided. Although the costing estimates provided here likely
incorporate significant start-up costs and are also influenced by the fact that the majority of low
intensity trainees are relatively inexperienced, results would seem to suggest that a 2-3 fold
reduction in cost/increase in activity is likely required in order to bring psychologically based talking
therapies into the financial envelope originally proposed.
Among those who completed treatment, there was a recovery rate of more than 50% in all types of
therapy, which is in line with what Clark et al (2009) report. This rate is also similar to what was
expected when the IAPT programme rolled out (Layard et al 2007). However the natural rate of
recovery is 30%, making the recovery rate that can be directly attributable to treatment only 20%.
Although the effectiveness of the interventions are likely to increase as workers become more
experienced, even with the cost figures estimated in this study, the IAPT programme will be cost
beneficial based on an estimate of a benefit to the society per treated person of £10500 per year (as
estimated by Layard et al (2007)).
Implications
Additional analyses performed by the authors would indicate that there are a number of ways to
reduce current costs with psychologically based talk therapies clinics. Preliminary analysis of IAPT
activity on June 2010 would suggest that the average number of patients seen by a high intensity
therapist was approximately three while for a low intensity therapist, it was approximately five
individuals. Significant fluctuation in the level of daily activity, and median number of sessions
provided was also observed across PCTs. Interestingly, longer median sessions were not necessarily
associated with better treatment outcomes. As such, an increase in daily activity would be one of
the most readily observable methods to decrease sessional costs, although this would have to be
done without a corresponding drop in treatment effectiveness. In addition, standardised treatment
15
packages of a set duration agreed upon by the clinician and patient at the outset of treatment may
be an effective method to standardise treatment duration and outcomes.
Increases in treatment volume may also be realised by changes in the way treatments are delivered.
The use of technology mediated treatment, in particular telephone mediated interventions, might
be instrumental to bring down the per session cost and thereby the cost of treatment and recovery.
Psychological interventions delivered by telephone are more convenient and, in a growing number
of situations, have been found to be as or more effective at reducing symptoms in patients treated
with CBT based interventions. In a small scale study of telephone treatment for patients with
obsessive compulsive disorder, a 40% reduction in clinician treatment time for telephone therapy is
being reported (Lovell et al 2006, Robinson et al 1990). Recent analysis of current IAPT data in this
cohort has also found equivalent reduction in anxiety, depression resulting from face-to-face and
telephone mediated treatment in propensity score matched patient cohorts (in press). If suitable
and accepted by patients and clinicians, telephone mediated therapy may allow for an increase in
productivity while simultaneously reducing the need for additional treatment spaces and travel, two
of the largest non-salaried costs reported in IAPT. With telephone therapy, service organisations and
clinicians may be no longer constrained by conventional working hours and treatment facilities,
conferring some flexibility in working hours. Literature also suggests that telephone-based delivery
may also have the potential to remove patient perceived barriers to initiating treatment (Mohr et al
2006).
Other modalities of therapy also have the potential to increase the volume of patients treated,
although additional research is required as to their efficacy. Online CBT delivered by a therapist in
real time was found to be cost-effective compared with usual care in the treatment of depression in
primary care in UK if society is willing to pay at least £20 000 per Quality Adjusted Life Year and
could be a useful alternative to face-to-face CBT (Hollinghurst et al 2010). Group psycho-educational
workshop interventions have been mentioned in NICE guidelines within the context of low intensity
treatments and may represent a valid way to reduce cost for low intensity therapists. However,
more research is required to investigate if group based interventions are effective and cost effective
compared to face to face therapy.
Stepping up and stepping down was associated with increased costs relative to remaining in high or
low intensity therapies, indicating a substantial penalty in patient misallocation to the wrong
treatment. Further investigation into predictors of response in high and low intensity and into
predictors of stepping is required in order to inform clinicians concerning likely patient response to
treatment. Similarly, identification of distinct patient typologies (known as a patient casemix) and
their observed response to treatment should provide clinicians with more information with which to
correctly allocate individuals into high or low intensity treatments.
Among the total costs for treatment, a significant amount of at least 22% of the costs goes for the
treatment of those who decline, drop out or are found not suitable for treatment. Some of these
costs could be reduced through reviewing the current system of working. Following up of those who
decline and drop out could be facilitated to convert them to complete treatment. Similarly, it takes a
median of more than 2 sessions to identify those patients who are found not suitable for treatment.
If they could be identified during the initial assessment session, this has the potential to reduce
overall operating costs.
The cost per recovered patients in this study should not be used as a basis for an economic
evaluation, since the type, distribution, and baseline severity of patients is likely to vary greatly
across regions and cannot be directly compared. If a proper economic evaluation is to be performed,
one of the first steps will be to derive a nationally representative casemix sample which
16
comprehensively characterises the types of patients presenting to talking therapies services.
Reduction in anxiety and depression (PHQ-9 and GAD-7) in each casemix can then be directly
compared along with the resulting clinician time, and provide the ability to identify standardised,
cost-effective examples of effective, good clinical practice.
Limitations
One limitation of this study was that a pivotal assumption in the costing approach adopted for all of
the data was based on observed data in a single PCT. Specifically, the observed cost ratios ratio of
1.8 to 1 for high to low intensity costs was observed in itemised costing for the Bedfordshire budget
and applied to the costing data for all other PCTs. The sensitivity analysis reveals that the cost per
session, treatment and recovered patients is sensitive to this assumption, indicating that inaccurate
ratios are likely to influence the overall estimates of session, treatment course costs. However,
additional information provided from other PCTs would indicate that this ratio is representative of
the difference in staffing costs. Average salary information submitted from four PCTs (unpublished
data) showed that the ratio of low to high intensity average salary was between 1.6 to 1.8, similar to
the costing ratio used in the analysis. Given that the IAPT therapies are highly human resources
intensive and that the vast majority of the workers are salaried (and that these salaries form the
majority of the total spend on high and low intensity services), it is expected that the ratio used is a
reliable approximation of the true difference in overall costs. A similar cost ratio was observed in the
overhead costs of high intensity and low intensity activity in the data from Bedfordshire PCT,
indicating that overhead cost ratios were unlikely to influence the overall ratios.
The development of a more accurate micro costing approach is advocated in order to confirm or
improve upon the estimates used in this analysis. This remains an area of important future research.
In addition, the costs assessed here are based on the overheads of treatment. A broader perspective
incorporating the overall societal costs may also be prudent considering potentially significant
further costs for patients in terms of medication, travel for therapy and income loss due to the
treatment course.
Conclusion
This is one of the first assessments of costs of talking based psychological therapies in the UK.
Results indicate that costs currently exceed previous estimates, although a number of caveats exist.
We invite replication and additional analyses along with evidence based discussion surrounding
alternative, cost-effective methods of intervention. It is likely that improvements in current IAPT
practice cannot occur until current practice is scrutinised and treatment approaches that are both
effective and financially viable are identified, studied, and highlighted.
17
References:
Cameron IM, Crawford JR, Lawton K, Reid IC,(2008), Psychometric comparison of PHQ-9 and HADS
for measuring depression severity in primary care. British Journal of General Practise; 58: 32-6.
Centre for Economic Performance (2006), The depression report: A new deal for depression and
anxiety disorders. London: London School of Economics and Political Science Centre for Economic
Performance
Clark D.M., Layard R, Smithies R, Richards D.A, Suckling R, Wright B, (2009), Improving access to
psychological therapy: initial evaluation of two UK demonstration sites, Behaviour Research and
Therapy; 47: 910–920.
Curtis L, & Netten A, (2006), Unit Costs of Health and Social Care. University of Kent, Canterbury:
Personal Social Services Research Unit.
Department of Health, (2007),Press release: Johnson announces £170 million boost to mental health
therapies. Wednesday 10 October 2007 12:51. GNN ref 152603P.
Department of Health (2008), Improving Access to Psychological Therapies (IAPT) Commissioning
Toolkit, Department of Health : Crown
Hollinghurst S, Peters TJ, Kaur S, Wiles N, Glyn Lewis, Kessler D, (2010), Cost-effectiveness of
therapist-delivered online cognitive–behavioural therapy for depression: randomised controlled
trial. The British Journal of Psychiatry; 197 : 297-304
Kessler RC, Berglund P, Demler O, Jin R, Koretz D, Merikangas KR, (2003), The epidemiology of major
depressive disorder: Results from the National comorbidity survey replication (NCS-R), Journal of the
American Medical Association; 289:3095–3105.
Kroenke K, Spitzer RL, and Williams JB,(2001), The PHQ-9: validity of a brief depression severity
measure. Journal of general Internal Medicine; 16:606-13.
Layard R, (2006), The case for psychological treatment centres. British Medical Journal; 7548: 10302.
Layard R, Clark D, Knapp M, Mayraz G, (2007), Cost benefit analysis of psychological therapy.
National Institute Economic Review; 202: 90 -8
Lovell K, Cox D, Haddock G, Jones C, Raines D, Garvey R, Roberts C, Hadley S,(2006), Telephone
administered cognitive behaviour therapy for treatment of obsessive compulsive disorder:
randomised controlled non-inferiority trial. British Medical Journal; 333(7574): 883-5
Lowe B, Kroenke K, Herzog W, Grafe K,(2004), Measuring depression outcome with a brief selfreport instrument: sensitivity to change of the Patient Health Questionnaire (PHQ-9). Journal of
Affective Disorders; 81: 61-66.
Mental Health Strategies (2010) 2009/10 national survey of investment in adult mental health
services.
Report
prepared
for
the
Department
of
Health,
available
at
http://www.dh.gov.uk/en/Publicationsandstatistics/Publications/PublicationsPolicyAndGuidance/D
H_117488
18
Mohr DC, Hart SL, Howard I, Julian L, Vella L, Catledge C, Feldman MD,(2006), Barriers to
psychotherapy among depressed and nondepressed primary care patients. Annals of Behavioural
Medicine; 32(3): 254-58.
Nolan G, (2009), Developing a Local Tariff for IAPT Services in NHS East of England. Cambridge: NHS
East of England
Richards D, & Suckling R,(2009), Improving access to psychological therapies: Phase IV prospective
cohort study. British Journal of Clinical Psychology; 48: 377-396.
Robinson LA, Berman JS, Neimeyer RA,(1990), Psychotherapy for the treatment of depression: a
comprehensive review of controlled outcome research. Psychological Bulletin; 108: 30-49.
Sainsbury Centre, (2010), The Economic and Social Costs of Mental Health Problems in 2009/10.
London: Sainsbury Centre for Mental Health.
Singleton N, Bumpstead R, O'Brien M, Lee A, Meltzer H,(2001), Psychiatric morbidity among adults
living in private households, 2000. London: The Stationary Office.
Spitzer RL, Kroenke K, Williams JB, Lowe B, (2006), A brief measure for assessing generalized anxiety
disorder: the GAD-7. Archives of Internal Medicine; 166: 1092-7.
APPENDIX 1 - Patient Health Questionnaire Depression scale (PHQ-9) and the Generalised Anxiety
Disorder scale (GAD-7)
19
PHQ-9
Over the last 2 weeks (or other agreed time period) how often have you been bothered by any of the
following problems?
1 Little interest or pleasure in doing things
2 Feeling down, depressed, or hopeless
3 Trouble falling or staying asleep, or sleeping too much
4 Feeling tired or having little energy
5 Poor appetite or overeating
6 Feeling bad about yourself — or that you are a failure or have let yourself or your family down
7 Trouble concentrating on things, such as reading the newspaper or watching television
Moving or speaking so slowly that other people could have noticed? Or the opposite — being so
8 fidgety or restless that you have been moving around a lot more than usual
9 Thoughts that you would be better off dead or of hurting yourself in some way
PHQ-9 total score =
Not at Several More than
Nearly
all
days half the days every day
0
1
2
3
0
1
2
3
0
1
2
3
0
1
2
3
0
1
2
3
0
1
2
3
0
1
2
3
Over the last 2 weeks (or other agreed time period) how often have you been bothered by any of the
following problems?
1 Feeling nervous, anxious or on edge
2 Not being able to stop or control worrying
3 Worrying too much about different things
4 Trouble relaxing
5 Being so restless that it is hard to sit still
6 Becoming easily annoyed or irritable
7 Feeling afraid as if something awful might happen
GAD-7 total score =
Not at Several More than
Nearly
all
days half the days every day
0
1
2
3
0
1
2
3
0
1
2
3
0
1
2
3
0
1
2
3
0
1
2
3
0
1
2
3
GAD-7
0
0
1
1
2
2
APPENDIX 2 - Median number of sessions and cost of treatment of deceased, declined, dropped and
not suitable.
20
3
3
Deceased
Activity
Median
25th Percentile
75th Percentile
Median
High Intensity 25th Percentile
75th Percentile
Median
Stepped Down
25th Percentile
(High to Low)
75th Percentile
Median
Stepped Up
25th Percentile
(Low to High)
75th Percentile
Low Intensity
Bedfordshire
Cambridgeshire
North & East Hertfordshire
Suffolk
West Hertfordshire
Low
High
Cost of
Low
High
Cost of
Low
High
Cost of
Low
High
Cost of
Low
High
Cost of
Intensity Intensity Treatment Intensity Intensity Treatment Intensity Intensity Treatment Intensity Intensity Treatment Intensity Intensity Treatment
Sessions Sessions
(£)
Sessions Sessions
(£)
Sessions Sessions
(£)
Sessions Sessions
(£)
Sessions Sessions
(£)
6
0
541
4
0
361
0
0
2
0
157
0
0
6
0
541
4
0
361
0
0
2
0
157
0
0
6
0
541
4
0
361
0
0
2
0
157
0
0
0
0
0
0
0
2
541
0
0
0
2
448
0
0
0
0
0
2
541
0
0
0
2
448
0
0
0
0
0
2
541
0
0
0
7
1,569
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
2
414
0
0
1
6
1,773
2
3
579
0
0
1
2
414
0
0
1
2
691
2
3
579
0
0
1
2
414
0
0
1
7
2,043
2
3
579
0
0
-
Declined
Activity
Median
25th Percentile
75th Percentile
Median
High Intensity 25th Percentile
75th Percentile
Median
Stepped Down
25th Percentile
(High to Low)
75th Percentile
Median
Stepped Up
25th Percentile
(Low to High)
75th Percentile
Low Intensity
Bedfordshire
Cambridgeshire
North & East Hertfordshire
Suffolk
West Hertfordshire
Low
High
Cost of
Low
High
Cost of
Low
High
Cost of
Low
High
Cost of
Low
High
Cost of
Intensity Intensity Treatment Intensity Intensity Treatment Intensity Intensity Treatment Intensity Intensity Treatment Intensity Intensity Treatment
Sessions Sessions
(£)
Sessions Sessions
(£)
Sessions Sessions
(£)
Sessions Sessions
(£)
Sessions Sessions
(£)
3
0
270
2
0
181
2
0
300
2
0
157
3
0
373
2
0
180
2
0
181
2
0
300
2
0
157
2
0
249
3
0
270
4
0
361
4
0
601
4
0
313
4
0
498
0
3
486
0
4
650
0
3
811
0
3
423
0
2
448
0
2
324
0
2
325
0
2
541
0
2
282
0
2
448
0
5
810
0
5
813
0
5
1,352
0
6
846
0
3
673
1
1.5
333
1.5
1
298
4
1.5
1,006
2
2
439
1
1
349
1
1
252
1
1
253
3
1
721
2
1
298
1
1
349
1
2
414
3.5
2.5
723
5
2
1,292
3
5
940
2
1
473
4
1
522
1
1
253
1
1
421
1
1
219
1
1
349
2
1
342
1
1
253
1
1
421
1
1
219
1
1
349
5
1
612
3
3
759
1
2
691
2
2
439
1.5
3
859
Dropped Out
Activity
Median
25th Percentile
75th Percentile
Median
High Intensity 25th Percentile
75th Percentile
Median
Stepped Down
25th Percentile
(High to Low)
75th Percentile
Median
Stepped Up
25th Percentile
(Low to High)
75th Percentile
Low Intensity
Bedfordshire
Cambridgeshire
North & East Hertfordshire
Suffolk
West Hertfordshire
Low
High
Cost of
Low
High
Cost of
Low
High
Cost of
Low
High
Cost of
Low
High
Cost of
Intensity Intensity Treatment Intensity Intensity Treatment Intensity Intensity Treatment Intensity Intensity Treatment Intensity Intensity Treatment
Sessions Sessions
(£)
Sessions Sessions
(£)
Sessions Sessions
(£)
Sessions Sessions
(£)
Sessions Sessions
(£)
3
0
270
3
0
271
3
0
451
3
0
235
3
0
373
2
0
180
2
0
181
2
0
300
2
0
157
2
0
249
4
0
360
4
0
361
4
0
601
4
0
313
4
0
498
0
4
648
0
3
488
0
3
811
0
4
564
0
4
897
0
3
486
0
3
488
0
2
541
0
2
282
0
2
448
0
7
1,134
0
6
976
0
5
1,352
0
5
705
0
7
1,569
3
1
432
2
1
343
2
1
571
1
2
360
1.5
1
411
1
1
252
1
1
253
1
1
421
0
1
141
1
1
349
5
6
1,423
4
1
524
4
1
871
2
5
861
3
1
598
3
4.5
999
1
3
578
1
2
691
1
4
642
1
2
573
2
2
504
1
2
416
1
1
421
1
2
360
1
2
573
5
8
1,747
3.5
5
1,129
1
3
961
3
5
940
2
4
1,146
Not Suitable
Activity
Median
25th Percentile
75th Percentile
Median
High Intensity 25th Percentile
75th Percentile
Median
Stepped Down
25th Percentile
(High to Low)
75th Percentile
Median
Stepped Up
25th Percentile
(Low to High)
75th Percentile
Low Intensity
Bedfordshire
Cambridgeshire
North & East Hertfordshire
Suffolk
West Hertfordshire
Low
High
Cost of
Low
High
Cost of
Low
High
Cost of
Low
High
Cost of
Low
High
Cost of
Intensity Intensity Treatment Intensity Intensity Treatment Intensity Intensity Treatment Intensity Intensity Treatment Intensity Intensity Treatment
Sessions Sessions
(£)
Sessions Sessions
(£)
Sessions Sessions
(£)
Sessions Sessions
(£)
Sessions Sessions
(£)
2
0
180
3
0
271
3
0
451
2.5
0
196
2
0
249
2
0
180
2
0
181
2
0
300
2
0
157
2
0
249
3
0
270
4
0
361
4
0
601
4
0
313
4
0
498
0
2
324
0
2
325
0
3
811
0
3
423
0
3
673
0
2
324
0
2
325
0
2
541
0
2
282
0
2
448
0
4
648
0
4
650
0
4
1,082
0
6
846
0
5
1,121
4
1
522
2
1
343
1
1
421
1
1
219
1
1
349
2.5
1
387
1
1
253
1
1
421
1
1
219
1
1
349
5.5
3
982
2
2
506
3
1
721
1
1
219
2
1
473
2
2
504
1
2
416
1
1.5
556
1
2
360
1
2
573
1
1
252
1
1
253
1
1
421
1
1
219
1
1
349
3
3.5
837
4
3
849
1
3.5
1,097
4
4
877
2
3
921
21