Sage Journals: Discover world-class research

Abstract

French

Background

Compared to treatment as usual (TAU), early psychosis intervention programs (EPI) have been shown to reduce mortality, hospitalizations and days of assisted living while improving employment status.

Aims

The study aim was to conduct a cost-benefit analysis (CBA) and a cost-effectiveness analysis (CEA) to compare EPI and TAU in Canada.

Methods

A decision-analytic model was used to estimate the 5-year costs and benefits of treating patients with a first episode of psychosis with EPI or TAU. EPI benefits were derived from randomized controlled trials (RCTs) and Canadian administrative data. The cost of EPI was based on a published survey of 52 EPI centers in Canada while hospitalizations, employment and days of assisted living were valued using Canadian unit costs. The outcomes of the CBA and CEA were expressed in terms of net benefit (NB) and incremental cost per life year gained (LYG), respectively. Scenario analyses were conducted to examine the impact of key assumptions. Costs are reported in 2019 Canadian dollars.

Results

Base case results indicated that EPI had a NB of $85,441 (95% CI: $41,140; $126,386) compared to TAU while the incremental cost per LYG was $26,366 (95% CI: EPI dominates TAU (less costs, more life years); $102,269). In all sensitivity analyses the NB of EPI remained positive and the incremental cost per LYG was less than $50,000.

Conclusions

In addition to EPI demonstrated clinical benefits, our results suggest that large-scale implementation of EPI in Canada would be desirable from an economic point of view .

Keywords

cost-benefit analysis economic analysis early onset Schizophrenia psychosis

Introduction

Psychotic illnesses result in impairment in reality testing, social and vocational functioning in people affected from them. The first episode of psychosis occurs generally between the ages of 15 and 30 years¹ and the incidence of first episode of non-affective psychosis has been reported in Ontario as 33.3 per 100,000 per year.² Early psychosis intervention programs (EPI) programs consist of multidisciplinary teams providing rapid psychiatric assessment and follow-up for youth aged typically 16 to 35 (criteria can vary by program) and include an array of psychosocial interventions (e.g., case management, family interventions).^3–6 Several randomized controlled trials (RCTs) have shown that EPI was associated with better outcomes than treatment as usual (TAU) (e.g., in terms of hospitalizations, remission, involvement in work or school, symptom severity, quality of life).⁷ Canadian administrative data from the catchment area of the Prevention and Early Intervention Program for Psychoses (PEPP) in London, Ontario, have also shown that mortality rates were four times lower in EPI users compared to a matched control group.⁸ While EPI programs are now considered best practice in Europe, North America and Australia, differences in the availability of EPI programs between countries and different regions within countries remain.^9–12

EPI programs have also been found to be cost saving or cost-effective compared to TAU as shown by a recent systematic review of 19 economic evaluations of EPI versus TAU.⁹ However, only one study was conducted in Canada based on a matched cohort analysis¹³ and this study did not consider the cost of the EPI program itself. As such the value of EPI programs in Canada is unknown, which represents an important gap in knowledge as EPI programs are not universally available across Canada.^11,14 It is therefore critical to document their economic value to support the broader implementation of EPI programs in Canada. Therefore, the objective of this study was to conduct an economic evaluation of EPI compared to TAU in Canada.

Methods

Methods Overview

A decision analytic model was created to estimate the 5-year costs and benefits of EPI compared to TAU in individuals with first time psychosis in terms of hospitalizations, employment, number of assisted living days and mortality. Costs and benefits beyond one year were discounted annually at a rate of 1.5%.¹⁵ A cost-benefit analysis (CBA) and a cost-effectiveness analysis (CEA) were conducted to compare EPI and TAU from a societal perspective. Scenario analyses were conducted to test the robustness of the results when changing key assumptions. Costs were reported in 2019 dollars.

Model Structure

The structure of the model is shown in Figure 1 and is similar to two previous EPI decision-analytic models.^16,17 The model assumes that each month, a proportion of individuals will die. Alive individuals can be hospitalized or employed while the remainder will not. Additionally, an overall mean number of assisted living days are applied to all patients who are alive.

Figure 1.

Model structure.

Clinical Inputs

In the absence of RCT data, mortality data were based on linked administrative databases from Ontario comparing over 5 years a cohort of individuals with first-episode psychosis using EPI services and a propensity matched control group of non-EPI users (i.e., TAU).⁸ Based on an absolute risk reduction of 2.5% (95% CI: 1.3%, 3.7%) in favor of EPI, all cause mortality was modeled as 3.9% in TAU and 1.4% for EPI in the first two years following admission.⁸ Since there were no differences in all cause mortality between EPI (1.5%) and TAU (1.7%) in the period between 2 and 5 years after admission,⁸ TAU mortality rate was assumed for EPI for years 2–5.

Estimates of the mean number of hospitalizations and employment rates between EPI and TAU were based on a recent meta-analysis of RCTs comparing EPI versus TAU in early phase psychosis.⁷ Based on a pooled analysis⁷ of 7 RCTs (LEO,¹⁸ STEP,¹⁹ OTP,²⁰ COAST,²¹ RAISE-ETP,²² JECP,²³ PIANO²⁴), the mean (standard deviation [SD]) number of hospitalizations was 0.41 (0.30) for EPI and 0.59 (1.11) for TAU. To accommodate the monthly time periods of our model, these estimates were divided by the sample size weighted mean duration of the 7 RCT's (i.e., 17 months) included in the calculation of the number of hospitalizations. Based on Canadian data¹⁴ and the duration of the RCTs included in the meta-analysis, the model assumes that EPI was provided for two years. After 24 months of EPI treatment, the model conservatively assumes the same hospitalization rates between EPI and TAU (i.e., impact of EPI on hospitalization does not continue after treatment cessation).

The proportion of individuals who were employed or in school after being treated with EPI or TAU were derived from the same meta-analysis.⁷ The results which were based on 6 RCTs (LEO,¹⁸ STEP,¹⁹ RAISE-ETP,²² JCEP,²³OPUS,²⁵ PIANO²⁴) indicated that the proportion of patients that were either employed or in school following TAU treatment was 45.3% while the relative risk of being in school or employed with EPI compared to TAU was 1.13 (95% CI: 1.03, 1.24, p = 0.01).⁷ Therefore the proportion of patients assumed to be working or in school following EPI treatment was calculated at 51.2% (45.3% × 1.13). To evaluate the impact of EPI on employment and associated productivity gains we used data from a Canadian study²⁶ which indicated that 72% of subjects in the EPI program who were either employed or in school, were in fact employed. In the absence of data, we also assumed 72% of TAU patients who were either employed or in school, were in fact employed. Therefore, the model assumes that 36.9% (i.e., 51.2% x 72%) and 32.6% (i.e., 45.3% x 72%) of EPI and TAU patients were employed following treatment, respectively. We further applied EPI Canadian data to estimate that amongst EPI and TAU patients that were employed, the proportion working full-time was 64% while the proportion working part-time was 36%.²⁶ Since it is unlikely that the effect of EPI on employment would be immediate, a 6-month delay from entering the program until EPI led to higher employment was assumed based on clinical opinion. Beyond the duration of the EPI intervention (i.e., 24 months in the base case analysis), we conservatively assumed that the effect of EPI would not last after EPI is stopped. Therefore, after 24 months, the proportion of individuals employed or in school in the TAU arm was applied to both the EPI and TAU arms.

To model the number of days in assisted living, data from the OPUS RCT was used as it is the only RCT reporting data on the mean number of days in assisted living associated with EPI and TAU.²⁷ In OPUS,²⁷ assisted living referred to facilities that were staffed around the clock for individuals with mental health problems. In the five-year follow-up of OPUS, EPI patients were found to have a lower mean number of assisted living days compared to TAU patients for a period of 3–5 years post treatment initiation, but not during the first two years post-treatment initiation. During years three to five, the mean (SD) number of assisted living days was reported to be 57 (213) and 102 (282) when EPI was compared to TAU, for a difference of 45.1 days (95% CI: 0.31, 89.9, p = 0.05).²⁷ This is equivalent to a monthly mean number of assisted living days of 1.58 for EPI and 2.83 for TAU. These differential rates were applied during months 25 through 60 in the 5-year base case analysis. During the first 24 months the mean number of supported housing days for TAU was applied to both treatment arms.

Cost Inputs

All costs are reported in 2019 dollars. The per-patient cost of providing EPI in Canada was based on a 2012 survey of 52 EPI programs in Ontario¹¹ in which participants were asked to provide the annual EPI budget (i.e., “What is your EPI program budget per annum?”) (Mean of $462,411) and the annual number of patients using EPI services (i.e., please indicate the “Currently registered clients") (mean of 62). Using this information, the per-patient annual cost of treating with EPI was calculated at $7,458 (i.e., $462,411/62). This study also reported the results by small or large catchment areas (e.g., $6,902 and $7,985, respectively) which were used in scenario analyses. Since the information reported in the survey included overheads but not psychiatrist fees, psychiatrist fees were added to the EPI cost per patient derived from the survey. Based on expert opinion, it was assumed EPI patients would have a psychiatric visit every 2 weeks for the first 3 months of EPI treatment, every 4 weeks during months 4 through 12 and every 6 weeks during the second year of EPI services. Assuming an initial psychiatric consultation fee of $300.70 (Ontario Schedule of Benefits, billing code A190)²⁸ and following psychiatrist consultations fees of $173.70 (billing code K198; 2 units)²⁸ the per-patient total physician fees applied to the first and second year of EPI treatment were $2,733 and $1,390 respectively. To the best of our knowledge no Canadian studies have compared the cost of treating first episode psychosis with TAU. Therefore we relied on two economic evaluations based on the Lambeth Early Onset (LEO) (TAU = 60% of EPI cost¹⁶) and OPUS (TAU = 40% of EPI costs²⁹) RCTs to estimate the cost of TAU at 50% (e.g. average of the two studies) of the costs of EPI including psychiatrist fees. The EPI and TAU per-patient annual costs of treatment were converted to 1-month cost to accommodate the model's one-month time periods. Differential costs for the EPI and TAU arm were applied in the first 24 months of the model. After 24 months the cost of TAU was applied to EPI arm as our base case analysis assumes that the duration of EPI programs was 24 months.¹⁴

Based on data from the Canadian Institute for Health Information (CIHI) the model considers an average cost of $13,266 for a hospitalization for schizophrenia/schizoaffective disorder (Case Mix Group 707, length of stay of 19.3 days).³⁰ CIHI hospitalization costs do not include physician costs. Therefore, based on expert opinion, the costs of psychiatrist visits during the EPI hospital stay were added assuming three psychiatrist visits per week. The model assumes a cost per day in assisted living of $87 based on the median cost per day in assisted living among five provinces reported in a recent Canadian study and adjusted to 2019 dollars using the Canadian Consumer Price Index for shelter.³¹ Because of the variation in the cost of assisted living in this study,³¹ the daily cost of assisted living was varied in scenario analysis. To estimate productivity costs associated with employment, a monthly wage was applied to individuals who were employed during each month. Since young people with early psychosis may be more likely to be employed in lower paying jobs,¹⁷ the Ontario minimum wage of $14 per hour³² was assumed to be conservative in our estimates. Furthermore, it was assumed that full-time employees would work 40 h per week, while part-time employees would work 20 h per week.

Economic Evaluation

A CBA and a CEA were performed to determine the economic value of EPI compared to TAU. Under the cost-benefit framework, the net benefit (NB) associated with treating patients with EPI instead of TAU was calculated as the difference between the incremental monetized benefits (i.e., hospitalization, employment, assisted living days and mortality) and the incremental treatment costs of EPI versus TAU. To model monetized life years lost, a value of $100,000 per year of life was applied to each year of the expected remaining life of our population based on the value of statistical lives recommended by the Treasury Board of Canada for valuing premature mortality (i.e., $6.1 million in 2014 or $7.9 million in 2019 dollars).³³ A standardized mortality ratio for schizophrenia compared to the general population of 2.5³⁴ was applied to Canadian life table data to estimate life years lost and lifetime monetized life years lost assuming an average age of 25 years⁸ for our population. In the CEA, costs (treatment costs and costs related to hospitalization, productivity and assisted living days) and effectiveness (i.e., life-years) were compared between EPI and TAU to determine the incremental cost per life year gained (LYG) associated with EPI.

Several scenario analyses were conducted to evaluate the impact of key model assumptions on the results including: 1) using different values for the relative price of TAU compared to EPI (i.e., base case 50%, scenario analyses using 40% and 60%) and for EPI annual treatment cost (base case = $7,458) to reflect low ($6,902) and high ($7,985) catchment areas; 2) varying the number of psychiatrist visits during the EPI program by + /-20% and the daily cost of assisted living (base case: $87, scenario analysis using $60 or $100); and 3) assuming that the effect of EPI on employment would be immediate (base case: 6-month delay).

To deal with the uncertainty associated with the model parameters, probabilistic analyses was undertaken for the base case and scenario analyses. In probabilistic analysis,¹⁵ each parameter is assigned a probability distribution and the model is simulated a large number of times (e.g., N = 1,000 simulations). A 95% CI can then be evaluated for each model outcome. For the CBA, the proportion of simulations for which the NB was positive was calculated while cost-effectiveness acceptability curves (CEAC)³⁵ were used in the CEA to determine the probability of EPI being cost-effective at different levels of willingness to pay per LYG (e.g., $100,000/LYG). Table 1 provides a summary of the base case values and the associated distributions considered for the simulations.³⁵ To measure the impact of the assumptions made for the parameters’ distributions, we conducted additional probabilistic analyses by changing the value of the standard errors (SEs) (i.e., assuming the SEs are half or twice the SEs used in the reference case).

Table 1.

Summary of Values, Parameters and Sources of Clinical and Cost Variables Used in the Analyses.

Model Variable	Base case Value	Probabilistic distribution and parameters	Source
Clinical variables
Mean hospital admissions (over 17 months)
TAU	0.59	normal (mean = 0.59, SE = 0.11)	Correll⁷
Difference EPI minus TAU	−0.18	normal (mean = −0.18, SE = 0.071)	Correll⁷
Mean number of days in assisted living (over 36 months)
TAU	102	normal (mean = 102, SE = 20)	Bertleson²⁷
Difference EPI minus TAU	45	normal (mean = 45, SE = 28.8)	Bertleson²⁷
Probability of being employed or in school
TAU	0.453	Fixed	Correll⁷
Relative risk of EPI vs. TAU	1.13	Lognormal (mean = 0.12, SE = 0.057)	Correll⁷
Proportion of individuals employed or in school who are employed	0.72	beta (alpha = 70, beta = 27)	Abdel Baki²⁶
Probability that individuals employed had full time employment	0.64	beta (alpha = 62, beta = 35)	Abdel Baki²⁶
Number of Psychiatric visits first year of treatment	15	Gamma (alpha = 25, beta = 0.06)	Expert opinion
Number of Psychiatric visits second year of treatment	8	Gamma (alpha = 25, beta = 0.032)	Expert opinion
Mortality (over 24 months)
TAU	3.9%	fixed	Anderson⁸
Absolute reduction in mortality (EPI-TAU)	2.5%	normal (mean = 0.025, SE = 0.006)	Anderson⁸
Mortality (24–60 months)
TAU	1.7%	fixed	Anderson⁸
EPI	1.7%	fixed	Anderson⁸
Costing variables
Annual cost of EPI	$7,458	gamma (alpha = 25, beta = 298.32)	Durbin¹¹
Initial Psychiatric consultation Fee	$300.70	Fixed	Code A190
Outpatient psychiatrist fees	$173.70	Fixed	Code K198*2
Inpatient psychiatrist fees	$200.30	Fixed	Code K100*2
Relative cost of TAU versus EPI	0.50	varied in sensitivity analysis	McCrone¹⁶
Mean cost per assisted living day	$87	gamma (alpha = 25, beta = 3.28)	Latimer³¹
Cost per psychiatric hospitalization	$9,552	gamma (alpha = 25, beta = 382.08)	CIHI³⁰
Monthly productivity cost per employed individual	$2,315	gamma (alpha = 25, beta = 92.61)	assumption

Notations: TAU: treatment as usual; EPI: early psychosis intervention; SE: standard error.

Results

In the base case, the CBA results indicated that the additional costs associated with treating patients with EPI compared to TAU ($10,909; 95% CI: $7,779, $14,999) were lower than the associated monetized benefits ($96,351, 95% CI: $51,211; $137,500), yielding a NB of $85,441 (95% CI: $41,140; $126,386). The NB was positive in 100% of the simulations. In the CEA, the incremental cost per LYG of EPI compared to TAU was $26,366 (95% CI: dominates TAU, [i.e., less costly, more life years]; $102,269) and the probability of EPI to be cost-effective at a willingness to pay of $100,000/LYG was 0.96 (Figure 2). Table 2 presents the details.

Figure 2.

Cost-effectiveness acceptability curve.

Table 2.

Cost Benefit and Cost-Effectiveness Analyses (EPI Compared to TAU): Base Case Probabilistic Results.

		95% confidence intervals
	Mean (Average of simulations)	2.5%	97.5%
COST BENEFIT ANALYSIS
Costs (a)
Increased Intervention costs	$10,909	$7,779	$14,999
Benefits
Reduced Hospitalizations	$3,202	$440	$6,743
Reduced assisted living days	$2,887	-$625	$6,896
Increased Productivity	$2,251	$867	$4,094
Value of life year lost	$88,011	$44,083	$129,703
Total Benefits (b)	$96,351	$51,211	$137,500
Net Benefits: (b)-(a)	$85,441	$41,140	$126,386
Percentage of simulations where the NB is positive	100%

COST-EFFECTIVENESS ANALYSIS*
Incremental Costs	$2,569*	-$3,641	$8,634
Incremental LYGs	0.0974	0.04890	0.14383
Incremental cost per LYG	$26,366	dominates	$102,269
Probability of EPI being cost effective at a willingness to pay of $100,000/LYG**	0.96

Notations: TAU: treatment as usual; EPI: early psychosis intervention; LYG: life years gained.

* In the CEA, incremental costs associated with EPI include treatment costs and costs related to hospitalization, productivity and assisted living days while life years gained are used as the effectiveness measure.

**: The probability of EPI being cost effective at a willingness to pay of $100,000/LYG corresponds to the number of simulations in which the incremental cost per LYG is less than $100,000 per LYG.

Using wider distributions for the probabilistic analysis had the most impact on the results (Table 3) but still the NB was positive in 97% of the simulations in the CBA and the probability of EPI to be cost-effective at a willingness to pay of $100,000/LYG was 0.84 with incremental cost-effectiveness ratios of less than $50,000/LYG.

Table 3.

Scenario Analyses (Probabilistic): Early Psychosis Intervention (EPI) vs. Treatment as Usual (TAU).

	COST-BENEFIT ANALYSIS (EPI versus TAU)		COST-EFFECTIVENESS ANALYSIS (EPI versus TAU)
Variable/value	Net Benefits	Percentage of simulations when the net Benefits are positive	Incremental cost per LYG	Probability of EPI being cost-effective at a willingness to pay of $100,000/LYG*
Ratio of TAU costs to EPI costs (base case = 0.50)
0.4	$83,974	100%	$41,422	93%
0.6	$86,908	100%	$11,310	98%
Annual EPI program cost (base case = $7,458)
Small catchment area(EPI annual cost $6,902)	$86,084	100%	$19,773	97%
Large catchment area(EPI annual cost $7,985	$84,832	100%	$32,622	95%
Psychiatrist visits during EPI program (15 year one, 8 year two)
-20% (13 year 1, 7 year 2)	$85,730	100%	$23,402	97%
+20% (18 year 1, 10 year 2)	$84,952	100%	$31,393	95%
Cost per Assisted living day (base case = $87)
$60	$84,540	100%	$35,621	95%
$100	$85,863	100%	$22,037	96%
Impact of EPI on Employment immediate – no time lag (base case = 6-month time lag)
No lag	$85,951	100%	$21,134	97%
Impact of using tighter or wider distributions
Tighter: Assume that SE is half of the base case SE.	$85,731	100%	$25,971	100%
Wider: Assume SE is twice the base case SE	$82,400	97%	$27,562	84%

Notations: EPI: early psychosis intervention; TAU: treatment as usual; SE: standard error.

*: percentage of simulations for which the incremental cost effectiveness per LYG is less than $100,000.

Discussion

Despite the recommendations of several provincial guidelines^5,36 EPI is not yet universally offered in Canada.¹⁴ This study is a first step in understanding the economic value associated with the provision of EPI in Canada and makes an important contribution to the economic literature on EPI. Our base case results suggest that EPI provides good value for money compared to TAU as shown by the CBA (NB of $85,441) and the CEA (Incremental cost per LYG of $26,366). In all scenario analyses comparing EPI versus TAU, the NB remained positive and the incremental cost per LYG was less than $50,000/LYG. The evidence provided here does suggest that EPI should be offered to all young Canadians experiencing first episode psychosis.

Our base case results were aligned with a recent systematic review of 19 economic evaluations of EPI services across the world, which showed that EPI was a cost-effective treatment strategy. Compared to these economic evaluations,⁹ our study has several strengths. First the clinical data used to model the impact of EPI on hospitalizations and employment was based on a recent meta-analysis of several RCTs comparing EPI and TAU in different countries and settings.⁷ Data on the impact of EPI on assisted living days was based on one of the largest RCTs to date.²⁷ While the mortality data were derived from Canadian non-randomized evidence, the risk of confounding in this study was reduced due to the use of propensity scoring matching to create comparable cohorts of EPI and non-EPI users.⁸ Another strength of the analysis is that it is fully probabilistic and therefore, the uncertainty around the model parameters is considered. We also used several conservative assumptions in our estimates. For example, we assumed that the treatment duration and benefits of EPI lasted 2 years based on the duration of RCTs included in the meta-analysis while there may be value in extending EPI beyond 2 years as shown in a recent Canadian³⁷ study and as now suggested in some provincial guidelines.^5,6 In terms of limitations, although Canadian unit costs were used in the analysis, none of the RCTs included in the meta-analysis used for this evaluation⁷ were conducted in Canada although they represented many different jurisdictions across the world. However, the generalizability of the trial results to current Canadian practice is unknown. In addition, RCTs may have had higher standards of fidelity than real world practices of EPI although the recent OPUS-real-world³⁸ data confirmed the effectiveness of EPI in real-world settings. We also did not model the costs and monetized benefits of providing EPI beyond 2 years. While we were able to derive a per-patient cost of EPI treatment based on a survey of 52 EPI services in Ontario,¹¹ the costs of EPI may be different in other Canadian provinces. The cost of TAU was not available in Canada and we had to rely on international data from LEO¹⁶ and OPUS.²⁹ Furthermore, what is considered as TAU is variable and may differ between jurisdictions. However, the results of our scenario analysis varying the cost of EPI or TAU did not change the study conclusions. Due to the lack of Canadian data, community medication costs associated with EPI and TAU were not included. However, although not directly comparable, the recent Canadian economic evaluation of a 3-year RCT extension following two years of EPI did not find a statistically significant difference in medication costs between EPI and TAU over the three years of participation.³⁹ Another major limitation of our economic evaluation is that due to the lack of data, we did not model the impact of EPI and TAU on health-related quality of life (HRQoL) using an outcome such as quality adjusted life years (QALYs). This is not unique to our study as two recent published economic evaluations of EPI^39,40 have used relapse avoided or months in total symptom remission in their cost-effectiveness calculations. Due to the scope of our CBA and CEA, we did not evaluate issues related to the universal implementation of EPI services in Canada. Those issues are important as the recent Ontario data² from the catchment area of the London-based EPI program indicated that approximately 50% of first-episode psychosis patients did not contact the EPI program for screening or admission.⁴¹ Differences in EPI access were also seen across socio-economic strata, which raised important issues regarding equity in access.⁴¹ In addition, while EPI programs can be offered in specialized settings by a dedicated team (e.g., academic hospitals), recent RCT data from the US/Canada (RAISE²²) and Italy (PIANO²⁴) have shown that integrating the provision of EPI services in community mental health teams (CMHT) by training CMHT staff to deliver EPI model-based care is also effective^22,24 and cost-effective.⁴² This may help provide access to EPI services especially in smaller communities where the small number of incident cases does not allow the regrouping of first episode psychosis patients and EPI staff in a separate specialised service. Finally, to ensure quality of care, maximise access, attainment of expected superior EPI outcomes, and similar cost between EPI programs, fidelity to the model is of utmost importance. Therefore, continuous training and monitoring involving community of practice and/or learning healthcare systems are also key when implementing EPI programs.¹²

This economic evaluation also identified a few important areas for future research priorities. First and foremost, more information on the cost of providing EPI and TAU is needed in Canada and elsewhere. Second, efforts should be made to determine the economic value of providing EPI beyond 2 years although the recent Canadian economic evaluation of a 3-year RCT extension following two years of EPI did not find a statistically significant difference in hospitalizations between EPI and TAU.³⁹ Finally, there is a need to generate HRQoL data associated with EPI or TAU treatments which could be used to conduct cost-utility analyses, and to leverage administrative databases to document the epidemiology, costs, and outcomes associated with EPI compared to TAU.

Given the demonstrated clinical benefits of EPI, large scale implementation of EPI has been recommended in different Canadian provinces. Our results suggest that the large-scale implementation of EPI in Canada would also be desirable from an economic point of view. However, future research is required to provide more certainty about the costs of providing EPI and TAU to better support decision-making about large scale implementation of EPI in Canada.

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.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the McConnell Family Foundation and the Canadian Institute of Health Research, (grant number 156947).

ORCID iDs

Gord Blackhouse

Amal Abdel-Baki

References

Jones

. Adult mental health disorders and their age at onset. Br J Psychiatry Suppl. 2013;54:s5-10.

Anderson

Norman

MacDougall

, et al. Estimating the incidence of first-episode psychosis using population-based health administrative data to inform early psychosis intervention services. Psychol Med. 2019;49(12):2091-2099.

Early Intervention in Psychosis Network (EIPN). Standards for early intervention in psychosis services. London: Royal College of Psychiatrists; 2016.

Early Psychosis Guidelines Writing Group and EPPIC National Support Program. Australian clinical guidelines for early psychos: The National Centre of Excellence in Youth Mental Health:Melbourne; 2016.

Early psychosis intervention program standards: Ontario Ministry of Health and Long-Term Care; 2011.

Programmes d’interventions pour premiers épisodes psychotiques (PIPEP). Quebec: Ministère de la santé et des services sociaux. Cadre de référence; 2017.

Correll

Galling

Pawar

, et al. Comparison of early intervention services vs treatment as usual for early-phase psychosis: a systematic review, meta-analysis, and meta-regression. JAMA Psychiatry. 2018;75(6):555-565.

Anderson

Norman

MacDougall

, et al. Effectiveness of early psychosis intervention: comparison of service users and nonusers in population-based health administrative data. Am J Psychiatry. 2018;175(5):443-452.

Aceituno

Vera

Prina

, et al. Cost-effectiveness of early intervention in psychosis: systematic review. Br J Psychiatry. 2019;215(1):388-394.

10.

Bertulies-Esposito

Nolin

Iyer

, et al. Ou en sommes-nous? An overview of successes and challenges after 30 years of early intervention services for Psychosis in Quebec: Ou en sommes-nous? Un apercu des reussites et des problemes apres 30 ans de services d'intervention precoce pour la psychose au Quebec. Can J Psychiatry. 2020;65(8):536-547.

11.

Durbin

Selick

Hierlihy

, et al. A first step in system improvement: a survey of early psychosis intervention programmes in ontario. Early Interv Psychiatry. 2016;10(6):485-493.

12.

Bertulies-Esposito

Iyer

Abdel-Baki

. The impact of policy changes, dedicated funding and implementation support on early intervention programs for psychosis. Can J Psychiatry. 2022:7067437211065726.

13.

Goldberg

Norman

Hoch

, et al. Impact of a specialized early intervention service for psychotic disorders on patient characteristics, service use, and hospital costs in a defined catchment area. Can J Psychiatry. 2006;51(14):895-903.

14.

Nolin

Malla

Tibbo

, et al.

Early intervention for psychosis in Canada: what is the state of affairs?

Can J Psychiatry. 2016;61(3):186-194.

15.

Guidelines for the economic evaluation of health technologies: anada. 4th ed. Ottawa: Canadian Agency for Drugs and Technologies in Health (CADTH); 2017 March.

16.

McCrone

Knapp

Dhanasiri

. Economic impact of services for first-episode psychosis: a decision model approach. Early Interv Psychiatry. 2009;3(4):266-273.

17.

Park

McCrone

Knapp

. Early intervention for first-episode psychosis: broadening the scope of economic estimates. Early Interv Psychiatry. 2016;10(2):144-151.

18.

Craig

Garety

Power

, et al. The Lambeth Early Onset (LEO) team: randomised controlled trial of the effectiveness of specialised care for early psychosis. Br Med J. 2004;329(7474):1067.

19.

Srihari

Tek

Kucukgoncu

, et al. First-Episode services for psychotic disorders in the U.S. Public sector: A pragmatic randomized controlled trial. Psychiatr Serv. 2015;66(7):705-712.

20.

Grawe

Falloon

Widen

, et al. Two years of continued early treatment for recent-onset schizophrenia: a randomised controlled study. Acta Psychiatr Scand. 2006;114(5):328-336.

21.

Kuipers

Holloway

Rabe-Hesketh

, et al. An RCT of early intervention in psychosis: Croydon Outreach and Assertive Support Team (COAST). Soc Psychiatry Psychiatr Epidemiol. 2004;39(5):358-363.

22.

Kane

Robinson

Schooler

, et al. Comprehensive versus usual community care for first-episode psychosis: 2-year outcomes from the NIMH RAISE early treatment program. Am J Psychiatry. 2016;173(4):362-372.

23.

Hui

Chang

Chan

, et al. Early intervention and evaluation for adult-onset psychosis: the JCEP study rationale and design. Early Interv Psychiatry. 2014;8(3):261-268.

24.

Ruggeri

Bonetto

Lasalvia

, et al. A multi-element psychosocial intervention for early psychosis (GET UP PIANO TRIAL) conducted in a catchment area of 10 million inhabitants: study protocol for a pragmatic cluster randomized controlled trial. Trials. 2012;13:73.

25.

Petersen

Jeppesen

Thorup

, et al. A randomised multicentre trial of integrated versus standard treatment for patients with a first episode of psychotic illness. Br Med J. 2005;331(7517):602.

26.

Abdel-Baki

Letourneau

Morin

, et al. Resumption of work or studies after first-episode psychosis: the impact of vocational case management. Early Interv Psychiatry. 2013;7(4):391-398.

27.

Bertelsen

Jeppesen

Petersen

, et al. Five-year follow-up of a randomized multicenter trial of intensive early intervention vs standard treatment for patients with a first episode of psychotic illness: the OPUS trial. Arch Gen Psychiatry. 2008;65(7):762-771.

28.

Ontario Ministry of Health. Schedule of benefits physician services under the health insurance act. https://www.health.gov.on.ca/en/pro/programs/ohip/sob/physserv/sob_master20210314.pdf; 2021.

29.

Hastrup

Kronborg

Bertelsen

, et al. Cost-effectiveness of early intervention in first-episode psychosis: economic evaluation of a randomised controlled trial (the OPUS study). Br J Psychiatry. 2013;202(1):35-41.

30.

Patient Cost Estimator. Canadian Institute for Health Information (CIHI). https://www.cihi.ca/en/patient-cost-estimator. Updated Last Updated Date. Accessed 2019 June 19.

31.

Latimer

Rabouin

Cao

, et al. Costs of services for homeless people with mental illness in 5 Canadian cities: a large prospective follow-up study. CMAJ Open. 2017;5(3):E576-E585.

32.

Ontario Go. Your guiide to the emopoyments standards act. https://www.ontario.ca/document/your-guide-employment-standards-act-0/minimum-wage.

33.

Treasury Board of Canada: Canadian Cost-Benefit Analysis-Regulatory Proposals 2007.

34.

Saha

Chant

McGrath

. A systematic review of mortality in schizophrenia: is the differential mortality gap worsening over time? Arch Gen Psychiatry. 2007;64(10):1123-1131.

35.

Briggs

Sculpher

Claxton

. Decision modelling for health economic evaluation. United Kingdom: Oxford University Press; 2006.

36.

Plan d’action en santé mentale 2015-2020: Gouvernement du Québec; 2015. IISBN: 978-2-550-77085-5. Publication year is 2017. https://publications.msss.gouv.qc.ca/msss/fichiers/2017/17-914-18W.pdf

37.

Malla

Joober

Iyer

, et al. Comparing three-year extension of early intervention service to regular care following two years of early intervention service in first-episode psychosis: a randomized single blind clinical trial. World Psychiatry. 2017;16(3):278-286.

38.

Posselt

Albert

Nordentoft

, et al. The danish OPUS early intervention services for first-episode psychosis: a phase 4 prospective cohort study With comparison of randomized trial and real-world data. Am J Psychiatry. 2021;178(10):941-951.

39.

Groff

Latimer

Joober

, et al. Economic evaluation of extended early intervention service vs regular care following 2 years of early intervention: secondary analysis of a randomized controlled trial. Schizophr Bull. 2021;47(2):465-473.

40.

Behan

Kennelly

Roche

, et al. Early intervention in psychosis: health economic evaluation using the net benefit approach in a real-world setting. Br J Psychiatry. 2020;217(3):484-490.

41.

Anderson

Norman

MacDougall

, et al. Disparities in access to early psychosis intervention services: comparison of service users and nonusers in health administrative data. Can J Psychiatry. 2018;63(6):395-403.

42.

Rosenheck

Leslie

Sint

, et al. Cost-effectiveness of comprehensive, integrated care for first episode psychosis in the NIMH RAISE early treatment program. Schizophr Bull. 2016;42(4):896-906.

Economic Evaluation of Early Psychosis Interventions From A Canadian Perspective

Abstract

Background

Aims

Methods

Results

Conclusions

Keywords

Introduction

Methods

Methods Overview

Model Structure

Clinical Inputs

Cost Inputs

Economic Evaluation

Results

Discussion

Footnotes

Declaration of Conflicting Interests

Funding

ORCID iDs

References