A cost-benefit analysis of the Improving Access to Psychological Therapies programme using its key defining outcomes

Abstract

The economic argument underpinning the Improving Access to Psychological Therapy programme has been a central component in its initial and continued investment. Using open-access data, this article undertook a cost-benefit analysis using the programme’s key defining outcomes to determine its return-on-investment. It was found that in terms of investment and efficiency gains, the programme was in the higher ends of the cost spectrum for delivering psychological therapies. Although cost-estimates appear promising at first, when set in the context of a lower number of treatment contacts and a high proportion of early disengagement, estimates increased sharply.

Keywords

anxiety disorders cost-benefit depression health economics Improving Access to Psychological Therapies

Introduction

This journal recently published a special section on the Improving Access to Psychological Therapy (IAPT) programme introducing new evidence and perspectives which challenged the often-championed successes of the initiative, particularly around its clinical outcomes (Vol. 23, No. 9, August 2018, Special Section: IAPT Under the Microscope). In the editorial, Marks (2018) underlined several issues concerning the limited research into the programme’s value-for-money, economic viability and promised return-on-investment. Indeed, most studies directly evaluating the programme’s impact have focused on clinical effectiveness and not necessarily cost-effectiveness. While there have been recent trials reporting cost-benefit analyses (Lovell et al., 2017; NHS England, 2018; Richards et al., 2018), these tend to be restricted to individual providers, settings and intervention studies. What’s more, these trials are often complex and multifaceted and involving several teams, meaning the associated cost-benefits specific to IAPT services is difficult to ascertain. Direct cost-benefit analyses have reported mixed findings and uncertainty (De Lusignan et al., 2012; Griffiths and Steen, 2013; McCrone, 2013; Mukuria et al., 2013; Radhakrishnan et al., 2013; Sreeharan et al., 2013) with any effects found to disappear at follow-up (Mukuria et al., 2013). This is surprising as the cost-savings first proposed by Layard et al. (2007) were pivotal in its initial investment. The economic case which continues to be referenced (Layard and Clark, 2014) argues that investment in treatment would effectively ‘pay for itself’ (Layard et al., 2007: 8) because any money invested is retrieved via increased work productivity, tax receipts and an overall reduction in healthcare utilisation. Yet, despite this, the current evidence supporting its economic justification appears minimal.

Optimising fair and timely access to evidence-based psychological therapies was and continues to be the focus of the IAPT service delivery model. A core feature of delivery is the stepped-care model which involves offering referrals the least intrusive intervention first time along low- to high-intensity levels (Clark et al., 2017; Van Straten et al., 2015). Those with less severe and complex conditions are offered low-intensity treatments initially with the option to step-up or ‘correct’ as needed (Van Straten et al., 2015). Low-intensity interventions are useful as they reduce the amount of practitioner contact time required, leading to cost-savings and improved accessibility (Bennett-Levy et al., 2010; Hammond et al., 2012; Radhakrishnan et al., 2013). However, evidence has shown wide variability in stepped-care modelling across services (Gyani et al., 2013; RCP, 2013; Richards et al., 2012), poor rates of stepping-up during treatment (Bower et al., 2013; Chan and Adams, 2014; Delgadillo et al., 2013; Richards et al., 2012) and less than optimal intake score discrimination between intensity levels (Bower et al., 2013; Chan and Adams, 2014). A greater emphasis on low-intensity techniques is potentially inefficient overall as studies have reported high levels of relapse (Ali et al., 2017; Cairns, 2013; Hepgul et al., 2016) and poor long-term recovery profiles for those engaging with IAPT services (Scott, 2018). What’s more, there is a notable lack of longitudinal studies reporting on the sustainability of low-intensity interventions (Zhou et al., 2016), as well as little to no evidence demonstrating the superiority of stepped-care over more stratified-care models (Van Straten et al., 2015).

It should be noted that one of the IAPT programme’s greatest achievements is in its procurement of highly representative outcomes data. Because of the decision to mandate sessional measurement, 98 per cent of referrals attending at least two treatment contacts have valid pre- and post-outcomes data (Clark et al., 2017). Indeed, this high level of data quality enhances a researcher’s ability to closely scrutinise its delivery. While this approach is rational, its scope is still limited. The first limitation concerns the high proportion of early disengagement which, according to the latest annual report, around 40 per cent of those entering treatment attend one session only (IAPT, 2018). Within the same annual report, approximately 43 per cent of assessed-only referrals were deemed suitable but declined treatment, while 23 per cent were deemed not suitable, and only 9 per cent were discharged by mutual agreement following advice and support (IAPT, 2018). The second limitation concerns the heavy reliance on brief, self-report measures and lack of long-term outcomes which, when using more in-depth and longitudinal techniques, have found intervention effects to be diminished or even temporary (Ali et al., 2017; Cairns, 2013; Hepgul et al., 2016; Marks, 2018; Scott, 2018).

In consideration of the points outlined above, an analysis was undertaken to assess to what extent the money spent on IAPT services leads to its immediate and targeted benefits. Extracting publicly available data, it is possible to generate an overall cost-per-IAPT outcome. Although the programme’s evaluation framework is likely skewed and overlooking several cost implications, it is a useful indicator of how efficient its delivery is in relation to its own metrics and total money invested.

Methodology

IAPT annual report figures

The primary source of data for analysis is the ‘Psychological Therapies: Annual report on the use of IAPT services’ data tables, which includes figures on the number of incoming referrals, entering treatment, completing treatment, as well as rates of recovery and reliable improvement, between 1 April 2016 and 31 March 2017 (IAPT, 2018). Data are submitted regularly to NHS Digital with the latest estimates reporting a 98 per cent completion rate for pre- and post-outcomes for those with at least two treatment contacts (Clark et al., 2017). To allow for the most up-to-date information gathering, a single instance for each data point across each referral is captured for the whole year, therefore some activity will naturally represent beginning in the preceding year.

NHS reference costs

A secondary source of data includes the NHS Reference Costs (2016–2017) which reports an average unit cost of providing specifically defined services to the NHS in a given financial year. Reference costs are submitted annually by providers and are used to inform national tariffs, with the system stipulating prices be linked with outcomes. Each is valued according to mental health severity clusters and episodic care, the number of assessments, and performance against 10 national quality and outcome metrics, including access, engagement, clinical outcomes and satisfaction ratings. A unit cost is the figure incurred by providing a single unit of care and the initial assessment is per patient assessed and can cover multiple attendances. The reference costs include net education and training income, as well as other income streams, including direct and indirect running costs and overheads (NHS Improvement, 2017). Each submission is mandated and supported by detailed costing and collection guidance to reduce the variation in practices. It is the responsibility of providers to improve processes to ensure there is accurate reference costing data for delivering services.

Defining outcome terms

To access an IAPT service, an individual requires a ‘referral’ (e.g. from a GP, self and other; IAPT, 2018). Any single patient can have multiple referrals at any one time, and across multiple providers, therefore IAPT has opted to count referrals instead of patients. To enter treatment, a referral must have a first treatment contact recorded during the reporting period. Completing a course of treatment is defined as having at least two treatment contacts during the referral. A referral is deemed to have reliably improved if there has been a significant improvement (exceeding measurement error) on their pre- and post-treatment scores. The shift differs according to the measure used and is tailored to a referral’s specific condition. Recovery is assessed as the proportion of eligible referrals, defined as being above caseness at intake (a specific clinical cut-off score), completing a course of treatment (at least two treatment contacts) and dropping below caseness at the end of treatment. A referral has reliably recovered if they meet the criteria for both reliable improvement and recovery.

Cost-per-outcome analysis

A cost-per-outcome analysis reflects what it costs to offer a specific service and achieve its relevant outcomes. Overall costs, calculated using the total NHS Reference Costs (2016–2017) for Adult IAPT Mental Health Care Clusters and Initial Assessments, were divided by the number of referrals within each category of completed treatment, reliably improved, recovered and reliably recovered (IAPT, 2018). Given that initial assessments can involve advice, signposting and assessing for non-suitability, the cost-per-outcome is provided with and without initial assessment costs. Considering the large proportion of referrals entering treatment do not attend at least two treatment contacts, the inclusion of assessment costs was deemed necessary.

Results

During 2016–2017, the national IAPT programme reported the number of referrals completing treatment as 567,106, with 369,254 reliably improving, 258,884 recovering and 246,456 reliably recovering (IAPT, 2018). During the same period, the costs for adult IAPT mental health care totalled £367,219,192, with £274,664,758 allocated to care clusters and £92,554,434 initial assessments. Based on total activity, there were 886,645 care cluster units and 726,002 initial assessment units, generating an average unit cost of £309.78 and £127.49, respectively (NHS Reference Costs, 2016–2017).

Based on these figures, it was possible to determine an overall cost-per-outcome analysis (Table 1). This calculation reported large differences in costs depending on whether initial assessments were included or excluded. The overall cost-per-completing treatment was £648: reliably improving £994, recovering £1419 and reliably recovering £1490. The national recovery rate target of 50 per cent was not achieved nationally during the 2016–2017 period (49.3%), although it was for the first time in early 2017 (IAPT, 2017). If a 50 per cent recovery rate of all those completing treatment was assumed during this period, then the cost-per-recovered would be £1295 or £969 excluding overall assessment costs.

Table 1.

Overall cost-per-outcome relative to IAPT NHS Reference Costs (2016–2017; including and excluding initial assessment costs).

	No. of referrals (IAPT, 2016–2017)	Relative to IAPT total overall NHS Reference Costs (2016–2017)	Relative to IAPT NHS Reference Costs (2016–2017) (excluding overall assessment costs)
Completed	567,106	£647.53	£484.33
Reliably improved	369,254	£994.49	£743.84
Recovered	258,884	£1418.47	£1060.96
Reliably recovered	246,456	£1490.00	£1114.46

IAPT: Improving Access to Psychological Therapy.

Cost per low- and high-intensity treatment

Given the cost differentials between both treatment intensity types, the cost-per-outcome is more meaningful when considering what proportion of low- to high-intensity treatment services provide. The figures provided by Radhakrishnan et al. (2013) reported high-intensity treatment accounted for more than 55 per cent of total costs, while the IAPT (2012) programme recommends a capacity ratio of 60 per cent high-intensity to 40 per cent low-intensity. According to the NHS Reference Costs (2016–2017), the number of high-intensity to low-intensity contacts was around 51 to 49 per cent, respectively. Assuming the 51:49 per cent contact ratio (NHS Reference Costs, 2016–2017), a series of analyses were run to assess how the costs differed per-outcome based on the intensity level. This ratio was applied to the number of referrals within each category as reported by the IAPT national figures (IAPT, 2018).

Replicating the sensitivity analyses undertaken by Radhakrishnan et al. (2013), ratios were varied between low- and high-intensity treatment costs of between 1.6 and 2.8 times. This was done to enhance the range of estimated figures and the finding’s robustness. Although initial assessment costs are typically one-off and, depending on the service delivery model, not necessarily a low- or high-intensity contact per se, to capture the true costing and efficiency per-outcome, these were included for reference. All contacts reported the full range of clusters treated, from 01 to 21, including, it should be noted, those in clusters 00 (Variance (unable to assign mental health care cluster code)) and 99 (Patients not assessed or clustered).¹ Table 2 provides the estimated costs based on cost differences and sensitivity analyses between low- and high-intensity treatments.

Table 2.

Overall cost-per-outcome relative to IAPT NHS Reference Costs (2016–2017; including and excluding overall assessment costs) between low- and high-intensity treatments, including varying cost ratios for sensitivity analyses.

	IAPT annual report (2016–2017) figures
Completed	567,106
Reliably improved	369,254
Recovered	258,884
Reliably recovered	246,456
	Total activity (no. of contacts) NHS Reference Costs (2016–2017) Adult IAPT Mental Health Clusters
Assessment	726,002
Low-Intensity	1,508,753
High-Intensity	1,572,974
Intensity	NHS Reference Costs (2016–2017)		NHS Reference Costs (2016–2017)(excluding overall assessment costs)
	Cost-per-completed		Cost-per-completed
	Low	High	Low	High
Ratio of contacts (49%:51%)	277,644	289,462	277,644	289,462
1.6	£508.70	£780.69	£380.49	£583.93
1.8	£472.37	£815.55	£353.31	£609.99
2	£440.88	£845.75	£329.76	£632.59
2.8	£348.06	£934.78	£260.33	£699.17
	Cost-per-reliably improved		Cost-per-reliably improved
Ratio of contacts (49%:51%)	180,780	188,474	180,780	188,474
1.6	£781.27	£1199.00	£584.36	£896.80
1.8	£725.47	£1252.53	£542.62	£936.84
2	£677.10	£1298.92	£506.45	£971.54
2.8	£534.56	£1435.65	£399.83	£1073.80
	Cost-per-recovered		Cost-per-recovered
Ratio of contacts (49%:51%)	126,745	132,139	126,745	132,139
1.6	£1114.35	£1710.17	£833.49	£1279.14
1.8	£1034.76	£1786.52	£773.96	£1336.24
2	£965.77	£1852.68	£722.36	£1385.73
2.8	£762.45	£2047.70	£570.28	£1531.60
	Cost-per-reliably recovered		Cost-per-reliably recovered
Ratio of contacts (49%:51%)	120,660	125,796	120,660	125,796
1.6	£1170.55	£1796.41	£875.52	£1343.64
1.8	£1086.94	£1876.61	£812.98	£1403.62
2	£1014.47	£1946.11	£758.78	£1455.61
2.8	£800.90	£2150.96	£599.04	£1608.83

IAPT: Improving Access to Psychological Therapy.

Absorbing overall assessment costs greatly increases the cost-per-outcome across both low- and high-intensity levels. Where these costs are included, the cost-per-recovered figure does not fall below £750, regardless of the intensity level and varying cost ratios. Based on these assumptions, the range of costs is notable. Adopting the standard cost ratio rate of 1.8, as is used by Radhakrishnan et al. (2013), for low-intensity treatment, the cost-per-completed is £472, reliably improved £725, recovered £1035 and reliably recovered £1087. For high-intensity treatment, the cost-per-completed is £816, reliably improved £1253, recovered £1787 and reliably recovered £1877. The range of costs-per-recovered outcome relative to the varying cost ratios ranged from £762 to £1114 for low-intensity and £1710 to £2048 for high-intensity. If overall assessment costs are excluded, at a cost ratio of 1.8, for low-intensity treatment, the cost-per-completed is £353, reliably improved £543, recovered £774 and reliably recovered £813. For high-intensity treatment, the cost-per-completed is £610, reliably improved £937, recovered £1336 and reliably recovered £1404. The range of costs-per-recovered outcome with overall assessment costs excluded, relative to the cost ratio assumptions, ranged from £570 to £833 for low-intensity and £1279 to £1532 for high-intensity.

Cost-per low- and high-intensity contact

In consideration of the intensity levels, it is worth considering how this translates to a cost-per-contact. The term ‘contact’ is used to match the NHS Reference Costs (2016–2017) terminology. A lower number of contacts per treatment course would expectedly increase the overall cost-per-contact estimates. Following a similar approach to the previous section, costs-per-contact were made possible by dividing the overall costs, again including and excluding initial assessment figures, by the number of contacts within each category. This provided a proportion of total costs allocation per contact, allowing for an estimated figure calculation. As before, the range of differential cost ratios (1.6–2.8 times) was also considered (Table 3). Based on this analysis, assuming a cost ratio of 1.8 and overall assessment costs are included, a low-intensity contact was estimated to be £87, while a high-intensity contact was £150. Including initial assessment costs, the sensitivity analyses reported a value range of £64–£94 for low-intensity and £144–£172 for high-intensity. Excluding assessment costs reduced the range from £48 to £70 for low-intensity and £107 to £129 for high-intensity.

Table 3.

Overall cost-per-contact between low- and high-intensity treatments, including varying cost ratios for sensitivity analyses, relative to IAPT NHS Reference Costs (2016–2017) (including and excluding overall assessment costs).

Intensity	Cost-per-contactNHS Reference Costs (2016–2017)		Cost-per-contactNHS Reference Costs (2016–2017) (excluding overall assessment costs)
	Low	High	Low	High
No. of contacts	1,508,753	1,572,974	1,508,753	1,572,974
1.6	£93.61	£143.66	£70.02	£107.46
1.8	£86.93	£150.08	£65.02	£112.25
2	£81.13	£155.64	£60.68	£116.41
2.8	£64.05	£172.02	£47.91	£128.66

IAPT: Improving Access to Psychological Therapy.

Discussion

To all appearances, the estimated cost-per-IAPT outcome compares favourably with the original economic argument and accompanying research. The simple economic case presented by Layard et al. (2007) was that treating someone would cost around £750, higher than what was calculated here, and higher still than the estimates of Radhakrishnan et al. (2013) of £877 and Mukuria et al. (2013) of £1042. Equally, the cost-per-recovered estimate compares favourably with the figure from Radhakrishnan et al. (2013) of £1766, and cost-per-reliably recovered from Mukuria et al. (2013) of £3800. However, these costs are based on 2007–2009 and 2009–2010 datasets and must be set in the context of significant start-up costs and employment of inexperienced trainees. Although these costs appear promising, it is only on closer inspection, when the high rate of early disengagement and a low number of contacts per course of treatment are considered, that these cost-estimates increase sharply.

The results clearly demonstrate the worth of including all referrals entering treatment as overall associated costs greatly increase. Failing to consider all referrals effectively skews the interpretation of effectiveness at IAPT services and given a substantial proportion of those entering treatment attend one session only (IAPT, 2018), an intention-to-treat analysis appears more appropriately suited for evaluating service delivery. While admirable in its ambition to reach an access rate of 25 per cent (NHS England and Mental Health Taskforce, 2016), the programme must recognise the limitations of increasing access at a potential cost to a robust and comprehensive assessment and treatment allocation session.

Regarding the number of treatment contacts, there appears to be a lower-than-expected rate overall. During 2016–2017, the IAPT (2018) programme reported an average session attendance of 6.6, ranging from 2.7 to 3.6 for low-intensity and 5.4 to 7.2 for high-intensity interventions. The original estimated figure of £750 by Layard et al. (2007) was based on 10 sessions at £75 each. If this figure was adjusted to a 6.6 session average, or even the highest of 7.2 sessions, the estimated cost-per-completed would be £495 or £540, respectively, each lower than the estimate figures calculated here. Elsewhere, Radhakrishnan et al. (2013) reported a median session attendance of five for low-intensity and eight for high-intensity treatments, while Mukuria et al. (2013) reported an average of three sessions overall, with a majority (over 90%) assigned to low-intensity interventions initially. In the wider literature, the Personal Social Services Research Unit (PSSRU, 2017) estimates around six sessions of face-to-face Cognitive Behavioural Therapy (CBT) to be £280, or £450 per person within the first year of treatment, around £200 cheaper than the estimated figures reported here. This suggests that the economic viability of IAPT delivery is at risk of providing little return-on-investment relative to other costs reported in the literature.

Previous analyses of IAPT services have reported a cost-per-session of low-intensity treatment to be £99 to £102 (Griffiths and Steen, 2013; Radhakrishnan et al., 2013) and of high-intensity to be £138 to £174 (Griffiths and Steen, 2013; Radhakrishnan et al., 2013). These are broadly in line with the overall range of estimates reported here. Across 21 studies, Barrett and Petkova (2013) reported a range of £31–£133 per hour of individual CBT, based on 2012 pricing. The overall cost-per-completing treatment presented in this article, if delivered over an average of 6.6 sessions, would result in a cost of £98 per session overall. Based on a 6.0 session average for individual CBT as reported by the national figures (IAPT, 2018), this would increase to £108. All things considered, this demonstrates how the cost-estimates, although encouraging at first, must be assessed with reference to a lower session attendance overall. Consequently, the cost-efficiency of IAPT interventions are found to be near the higher end of cost-estimates (Barrett and Petkova, 2013), thus challenging the claims of being value-for-money.

It is perhaps worth reflecting on the average number of sessions being recorded by the IAPT programme. These figures fall short of guideline recommendations for evidence-based psychotherapeutic interventions, which the programme claims to provide (IAPT, 2018; The National Institute for Health and Care Excellence (NICE), 2011). According to the IAPT national figures (IAPT, 2018), 39 per cent of referrals received both low- and high-intensity treatment, 35 per cent received low-intensity only and 26 per cent received high-intensity only, suggesting a greater leaning towards low-intensity delivery overall. The low number of attended sessions for low-intensity interventions is perhaps more concerning given sessions can sometimes last only 20–30 minutes. Equally, the overall average for high-intensity interventions is characteristically more in line with the numbers recommended for low-intensity interventions. What this illustrates is an emphasis on greater low-intensity styled provision. It is worth noting that high-intensity practitioners still constitute most of the workforce (62%) (NHS England and Health Education England, 2016), yet research has shown wide variability in the proportion of low- to high-intensity delivery between services (Gyani et al., 2013; RCP, 2013; Richards et al., 2012). On the whole, low-intensity interventions clearly have their value, however, given the sizeable rate of re-referrals (Cairns, 2013; Di Bona et al., 2014; Hepgul et al., 2016), lack of treatment score discrimination and stepping up practices between intensity levels (Bower et al., 2013; Chan and Adams, 2014; Delgadillo et al., 2013; Richards et al., 2012), highly complex cases being treated (Bower et al., 2013; Goddard et al., 2015; Hepgul et al., 2016) and high rates of relapse for those receiving low-intensity treatments (Ali et al., 2017; Scott, 2018); there is a potential risk of creating care inefficiencies that favour short over long-term outcomes.

It is important to note this article considers only one outcome, this being the programme’s own metrics and evaluation framework. Even with these considered, the cost-estimates appear to be high. It should be acknowledged that, had these outcomes been evaluated using more in-depth and longitudinal techniques by an independent assessor, these figures may have been very different (Scott, 2018). As other outcomes could be considered, the cost implications may be even greater. Moreover, as these estimates have not considered the wider implications of re-referrals or the local healthcare economy, the continued emphasis on judging service delivery via brief, self-report measures could generate false economies. Set against a backdrop of reduced funding in other sectors, including across adult social care (Amin-Smith et al., 2017), a distorted view of effectiveness in one sector could potentially come at the cost of another. If costs truly are absorbed elsewhere for those who do not receive any benefit from IAPT service delivery, this risks reducing effective care and increasing overall costs.

Currently, the available evidence looking into the cost-effectiveness of IAPT services, particularly beyond the service setting, are limited (De Lusignan et al., 2012; Delgadillo et al., 2017; Griffiths and Steen, 2013; Lovell et al., 2017; Mukuria et al., 2013; Parsonage et al., 2014; Radhakrishnan et al., 2013). What’s more, research used to justify the economic benefits of the IAPT programme has little relevance for how it delivers and evaluates interventions. For instance, Layard and Clark (2014) cite a study conducted by Fournier et al. (2015) to justify the potential rate at which individuals move from incapacity benefits into employment. However, this specific study focuses only on patients who had recovered from severe depression, were assessed using structured clinical interviews and diagnostic criteria, and were treated by highly trained practitioners, the majority of whom had PhDs. Similarly, research into the long-term effects of interventions appears to have been selectively chosen, omitting the generally limited to mixed findings in this area (Marks, 2018). Consequently, the current attempts to justify the cost-effectiveness of IAPT services seem to be severely lacking and should be the focus of future research. Encouragingly, IAPT reports are expected to include funding information which can be used to complement existing controlled studies and better inform statistical modelling and cost-benefit analyses (Barkham et al., 2017; Clark et al., 2017). Future work ought to focus on redesigning existing care pathways, resolving service inefficiencies and considering broader and long-term targets. There are promising avenues of research emerging regarding how to improve service delivery and cost-efficiencies including focusing on organisational factors (Clark et al., 2017), the deployment of practice research networks (Lucock et al., 2017), collaborative care models (Green et al., 2014), outcome feedback technology (Delgadillo et al., 2017), outcome adjusted benchmarking (Delgadillo et al., 2015), improved record linkage (De Lusignan et al., 2011, 2012; Parsonage et al., 2014) and integration of more diverse psychotherapeutic modalities (Barkham et al., 2017; Fonagy and Clark, 2015).

Limitations

It should be recognised that given the total number of contacts available via NHS Reference Costs (2016–2017) for low- and high-intensity treatments, attempts to differentiate between cases being stepped-up or down were not undertaken. As it was not possible to determine the number stepping up or down or indeed the overall average number of sessions attended, it was deemed acceptable to divide the proportion by a 51:49 per cent ratio to reflect actual high- to low-intensity work, rather than incorporating more assumptions. Research has reported rates of stepped-care to be less than adequate, failing to reach rates of 10 per cent in some cases (Delgadillo et al., 2013; Richards et al., 2012) although this has been shown to vary widely (Delgadillo et al., 2013; Gyani et al., 2013; IAPT, 2018; Richards et al., 2012) with certain providers leaning towards greater low- or high-intensity provision or assessment models only (IAPT, 2018). Accordingly, cost-estimates across all categories should be contextualised and interpreted with this in mind.

The analysis presented here is based on several assumptions across multiple data sources, many of which use single indicators for complex processes. Consequently, it was not possible to determine the sustainability of post-treatment outcomes or indeed tease out any costs saved or incurred within other sectors. Equally, indirect benefits such as improving the perceived value of psychological therapies and acceptance of people with mental health problems more generally cannot be accounted for. Bottom-up compared with top-down approaches have been found to vary costs (Barrett and Petkova, 2013) and there are a range of methodologies for conducting economic evaluations (Cohen and Reynolds, 2008). The cost-benefit analyses conducted in this area are often scaled at an individual or small number of services which, as of writing, are based on an almost decade-old dataset (De Lusignan et al., 2012; Griffiths and Steen, 2013; Mukuria et al., 2013; Radhakrishnan et al., 2013). As a result, comparisons with the wider literature may naturally produce discrepancies purely due to the chosen methodology.

All things considered, the range of estimated costs reported in this article are intended to provide an overview for the cost-efficiency of the IAPT programme at a range of different levels (outcomes, treatment course and session contacts). To truly measure the cost-utility of IAPT service delivery, it would require a sophisticated and longitudinal research design, incorporating measures of quality-adjusted life years and auditing the entire local healthcare economy.

Conclusion

Taken as a whole, the IAPT programme seems to be delivering treatment at an inefficient cost. Although outcome targets are being reached, this appears to be due to an increased emphasis on low-intensity styled provision which not only drives up costs-per-IAPT outcome but also potentially reduces the appropriateness of treatment allocation and sustainability of these outcomes. It should be noted that this interpretation centres on the estimated costs associated with service delivery and not the wider healthcare economy. Given that this article focuses solely on the programme’s evaluation framework and own metrics, it should be acknowledged that more in-depth and longitudinal techniques may have produced different and potentially costlier outcomes. It is commonly understood that there is wide variation between services and it is likely some are operating more efficiently than others. The IAPT programme is in a unique position given its robust data recording practices which can be used to support services in producing practice-based evidence and share learning nationally. Indeed, this should also complement the much-needed reformulation of determining service delivery and effectiveness as specified in this journal (Marks, 2018; Scott, 2018).

Footnotes

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Ethical approval

This article does not contain any studies with human participants or animals performed by any of the authors.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Notes

ORCID iD

Scott Steen

References

Ali

Rhodes

Moreea

, et al. (2017) How durable is the effect of low intensity CBT for depression and anxiety? Remission and relapse in a longitudinal cohort study. Behaviour Research and Therapy 94: 1–8.

Amin-Smith

Phillips

Simpson

(2017) Adult social care funding: A local or national responsibility? IFS briefing note BN227. London: Institute for Fiscal Studies.

Barkham

Moller

Pybis

(2017) How should we evaluate research on counselling and the treatment of depression? A case study on how the National Institute for Health and Care Excellence’s draft 2018 guideline for depression considered what counts as best evidence. Counselling and Psychotherapy Research 17(4): 253–268.

Barrett

Petkova

(2013) Cognitive behaviour therapy: A comparison of costs. In: Curtis

Burns

(eds) Unit Costs of Health and Social Care. Canterbury: Personal Social Services Research Unit, University of Kent, pp. 16–20.

Bennett-Levy

Richards

Farrand

(2010) Oxford guide to low intensity CBT interventions. In: Bennett-Levy

Richards

Farrand

, et al. (eds) Oxford Guide to Low Intensity CBT Interventions. Oxford: Oxford University Press, pp. 3–19.

Bower

Sutton

Richards

, et al. (2013) Influence of initial severity of depression on effectiveness of low intensity interventions: Meta-analysis of individual patient data. British Medical Journal 346(2): f540.

Cairns

(2013) Patients who come back: Clinical characteristics and service outcome for patients re-referred to an IAPT service. Counselling and Psychotherapy Research 14(1): 1–8.

Chan

Adams

(2014) Service use, drop-out rate and clinical outcomes: A comparison between high and low intensity treatments in an IAPT service. Behavioural and Cognitive Psychotherapy 42(6): 747–759.

Clark

Canvin

Green

, et al. (2017) Transparency about the outcomes of mental health services (IAPT approach): An analysis of public data. The Lancet 6736(17): 1–8.

10.

Cohen

Reynolds

(2008) Interpreting the results of cost-effectiveness studies. Journal of the American College of Cardiology 52(25): 2119–2126.

11.

De Lusignan

Navarro

Chan

, et al. (2011) Detecting referral and selection bias by the anonymous linkage of practice, hospital and clinic data using Secure and Private Record Linkage (SAPREL): Case study from the evaluation of the Improved Access to Psychological Therapy (IAPT) service. BMC Medical Informatics and Decision Making 11(1): 61.

12.

De Lusignan

Chan

Parry

, et al. (2012.) Referral to a new psychological therapy service is associated with reduced utilisation of healthcare and sickness absence by people with common mental health problems: A before and after comparison. Journal of Epidemiology & Community Health 66(6): e10.

13.

Delgadillo

Asaria

Ali

, et al. (2015) On poverty, politics and psychology: The socioeconomic gradient of mental healthcare utilisation and outcomes. The British Journal of Psychiatry 209(5): 429–430.

14.

Delgadillo

Overend

Lucock

, et al. (2017.) Improving the efficiency of psychological treatment using outcome feedback technology. Behaviour Research and Therapy 99: 89–97.

15.

Delgadillo

Gellatly

Stephenson-Bellwood

(2013) Decision making in stepped care: How do therapists decide whether to prolong treatment or not? Behavioural and Cognitive Psychotherapy 43: 1–14.

16.

Di Bona

Saxon

Barkham

, et al. (2014) Predictors of patient non-attendance at Improving Access to Psychological Therapy services demonstration sites. Journal of Affective Disorders 169: 157–164.

17.

Fonagy

Clark

(2015) Update on the improving access to psychological therapies programme in England: Commentary on … children and young people’s improving access to psychological therapies. BJPsych Bulletin 39(5): 248–251.

18.

Fournier

DeRubeis

Amsterdam

, et al. (2015) Gains in employment status following antidepressant medication or cognitive therapy for depression. The British Journal of Psychiatry 206(4): 332–338.

19.

Goddard

Wingrove

Moran

(2015) The impact of comorbid personality difficulties on response to IAPT treatment for depression and anxiety. Behaviour Research and Therapy 73: 1–7.

20.

Green

Richards

Hill

, et al. (2014) Cost-effectiveness of collaborative care for depression in UK primary care: Economic evaluation of a randomised controlled trial (CADET). PLoS ONE 9(8): e104225.

21.

Griffiths

Steen

(2013) Improving access to psychological therapies (IAPT) programme: Scrutinising IAPT cost estimates to support effective commissioning. Journal of Psychological Therapies in Primary Care 2(2): 142–156.

22.

Gyani

Shafran

Layard

, et al. (2013.) Enhancing recovery rates: Lessons from year one of IAPT. Behaviour Research and Therapy 51(9): 597–606.

23.

Hammond

Croudace

Radhakrishnan

, et al. (2012) Comparative effectiveness of cognitive therapies delivered face-to-face or over the telephone: An observational study using propensity methods. PLoS ONE 7(9): e42916.

24.

Hepgul

King

Amarasinghe

, et al. (2016.) Clinical characteristics of patients assessed within an Improving Access to Psychological Therapies (IAPT) service: Results from a naturalistic cohort study (Predicting Outcome Following Psychological Therapy; PROMPT). BMC Psychiatry 16(1): 52.

25.

Improving Access to Psychological Therapies (IAPT) (2012) Improving Access to Psychological Therapies: Guidance for Commissioning IAPT Training 2012/2013. London: Department of Health, Improving Access to Psychological Therapies.

26.

Improving Access to Psychological Therapies (IAPT) (2017) The Improving Access to Psychological Therapies Manual. Wakefield: NHS England and National Collaborating Centre for Mental Health.

27.

Improving Access to Psychological Therapies (IAPT) (2018) Psychological Therapies: Annual Report on the Use of IAPT Services England, 2016-17—Data Tables. Leeds: NHS Digital, Community and Mental Health Team.

28.

Layard

Clark

(2014) Thrive: The Power of Psychological Therapy. London: Penguin Group.

29.

Layard

Clark

Knapp

, et al. (2007) Cost-Benefit Analysis of Psychological Therapy (CEP Discussion). London: National Institute Economic Review, p. 90.

30.

Lovell

Bower

Gellatly

, et al. (2017) Clinical effectiveness, cost-effectiveness and acceptability of low-intensity interventions in the management of obsessive-compulsive disorder: The Obsessive-Compulsive Treatment Efficacy randomised controlled Trial (OCTET). Health Technology Assessment 21(37): 1–132.

31.

Lucock

Barkham

Donohoe

, et al. (2017) The role of Practice Research Networks (PRN) in the development and implementation of evidence: The Northern improving access to psychological therapies PRN case study. Administration and Policy in Mental Health and Mental Health Services Research 44(6): 1–13.

32.

McCrone

(2013) IAPT is probably not cost-effective. British Journal of Psychiatry 202(5): 383.

33.

Marks

(2018) IAPT under the microscope. Journal of Health Psychology 23(9): 1131–1135.

34.

Mukuria

Brazier

Barkham

, et al. (2013) Cost-effectiveness of an improving access to psychological therapies service. The British Journal of Psychiatry 202(3): 220–227.

35.

NHS England (2018) Mental health ‘game-changer’ care leads to 75 per cent reduction in hospital admissions. NHS England News, 14 May. Available at: https://www.england.nhs.uk/2018/05/mental-health-game-changer-care-leads-to-75-reduction-in-hospital-admissions/

36.

NHS England and Health Education England (2016) 2015 adult IAPT workforce census report. Version 1, 5 September, https://www.england.nhs.uk/mentalhealth/wp-content/uploads/sites/29/2016/09/adult-iapt-workforce-census-report-15.pdf

37.

NHS England and Mental Health Taskforce (2016) Five Year Forward View for Mental Health for the NHS in England. Redditch: NHS England and Mental Health Taskforce Strategy.

38.

NHS Improvement (2017) Combined Costs Collection: Education and Training Costs Collection Guidance 2016/17. London: NHS England and NHS Improvement.

39.

NHS Reference Costs (2016–2017) NHS Reference Costs 2016/17: Highlights, Analysis and Introduction to the Data. London: NHS Improvement.

40.

Parsonage

Hard

Rock

(2014.) Managing Patients with Complex Needs: Evaluation of the City and Hackney Primary Care Psychotherapy Consultation Service. London: Centre for Mental Health.

41.

Personal Social Services Research Unit (PSSRU) (2017) Unit Costs of Health and Social Care 2017. Kent: Personal Social Services Research Unit, University of Kent.

42.

Radhakrishnan

Hammond

Jones

, et al. (2013) Cost of Improving Access to Psychological Therapies (IAPT) programme: An analysis of cost of session, treatment and recovery in selected Primary Care Trusts in the East of England region. Behaviour Research and Therapy 51(1): 37–45.

43.

Richards

Duffy

Blackburn

, et al. (2018) Digital IAPT: The effectiveness & cost-effectiveness of internet-delivered interventions for depression and anxiety disorders in the Improving Access to Psychological Therapies programme: Study protocol for a randomised control trial. BMC Psychiatry 18(1): 1–13.

44.

Richards

Bower

Pagel

, et al. (2012) Delivering stepped care: An analysis of implementation in routine practice. Implementation Science 7: 3.

45.

Royal College of Psychiatrists (RCP) (2013) Report of the Second Round of the National Audit of Psychological Therapies for Anxiety and Depression (NAPT) 2013. London: Healthcare Quality Improvement Partnership.

46.

Scott

(2018) Improving access to psychological therapies (IAPT) – The need for radical reform. Journal of Health Psychology 23(9): 1136–1147.

47.

Sreeharan

Madden

Lee

, et al. (2013) Improving access to psychological therapies and antidepressant prescribing rates in England: A longitudinal time-series analysis. The British Journal of General Practice 63(614): e649–e653.

48.

The National Institute for Health and Care Excellence (NICE) (2011) Common Mental Health Problems: Identification and Pathways to Care (Clinical Guideline: CG123). London: NICE.

49.

Van Straten

Hill

Richards

, et al. (2015) Stepped care treatment delivery for depression: A systematic review and meta-analysis. Psychological Medicine 45(2): 231–246.

50.

Zhou

Pei

, et al. (2016) Internet-based cognitive behavioural therapy for subthreshold depression: A systematic review and meta-analysis. BMC Psychiatry 16(1): 356.