Cost-effectiveness of a domestic violence and abuse training and support programme in primary care in the real world: updated modelling based on an MRC phase IV observational pragmatic implementation study

Objectives To evaluate the cost-effectiveness of the implementation of the Identification and Referral to Improve Safety (IRIS) programme using up-to-date real-world information on costs and effectiveness from routine clinical practice. A Markov model was constructed to estimate mean costs and quality-adjusted life-years (QALYs) of IRIS versus usual care per woman registered at a general practice from a societal and health service perspective with a 10-year time horizon. Design and setting Cost–utility analysis in UK general practices, including data from six sites which have been running IRIS for at least 2 years across England. Participants Based on the Markov model, which uses health states to represent possible outcomes of the intervention, we stipulated a hypothetical cohort of 10 000 women aged 16 years or older. Interventions The IRIS trial was a randomised controlled trial that tested the effectiveness of a primary care training and support intervention to improve the response to women experiencing domestic violence and abuse, and found it to be cost-effective. As a result, the IRIS programme has been implemented across the UK, generating data on costs and effectiveness outside a trial context. Results The IRIS programme saved £14 per woman aged 16 years or older registered in general practice (95% uncertainty interval −£151 to £37) and produced QALY gains of 0.001 per woman (95% uncertainty interval −0.005 to 0.006). The incremental net monetary benefit was positive both from a societal and National Health Service perspective (£42 and £22, respectively) and the IRIS programme was cost-effective in 61% of simulations using real-life data when the cost-effectiveness threshold was £20 000 per QALY gained as advised by National Institute for Health and Care Excellence. Conclusion The IRIS programme is likely to be cost-effective and cost-saving from a societal perspective in the UK and cost-effective from a health service perspective, although there is considerable uncertainty surrounding these results, reflected in the large uncertainty intervals.


Abstract
Objectives: To evaluate the cost-effectiveness of the implementation of the IRIS programme using up-todate real-world information on costs and effectiveness from routine clinical practice. A Markov model was constructed to estimate mean costs and quality-adjusted life-years (QALYs) of IRIS versus usual care per woman registered at a general practice from a societal and health service perspective with a ten-year time horizon.

Design and Setting:
Cost-utility analysis in UK general practices, including data from six sites which have been running IRIS for at least two years across England.

Participants:
Based on the Markov model, we stipulated a hypothetical cohort of 10,000 women aged 16 years or older.

Interventions
The Identification and Referral to Improve Safety (IRIS) was a randomised controlled trial that tested the effectiveness of a primary care training and support intervention to improve the response to women experiencing DVA, and found it to be cost-effective. As a result, the IRIS programme has been implemented across the UK, generating data on costs and effectiveness outside a trial context.

Results:
The IRIS programme saved £14 per woman aged 16 or older registered in general practice  cost-effective in 61% of simulations using real life data when the cost-effectiveness threshold was £20 000 per QALY gained as advised by NICE.

Conclusion:
The IRIS programme is likely to be cost-effective and cost-saving from a societal perspective in the UK and cost effective from a health service perspective, though there is considerable uncertainty surrounding these results, reflected in the large uncertainty intervals.

Strengths and limitations of this study
• We have used up-to-date routine data from several sites across England to evaluate the value for money of IRIS, a domestic violence training programme.
• We were unable to include any impact of the IRIS programme on children exposed to DVA, as to our knowledge, there are no available cohort studies focusing on the cost and benefits of DVA interventions for this population.
• Using up-to-date data on costs and effectiveness from routine clinical practice the national implementation of the IRIS programme is likely to be cost-effective and even cost-saving.

Introduction
The lifetime prevalence of domestic violence and abuse (DVA) against women varies internationally from 15% to 71% (1). In the United Kingdom, in the year ending March 2017, 7.5% of women (1.2 million) experienced domestic abuse (2). Women who experience DVA suffer chronic health problems including gynaecological problems, gastrointestinal disorders, neurological symptoms, chronic pain, cardiovascular conditions and mental health problems (3)(4)(5)(6). In 2012, the cost of DVA in the UK, including medical and social services, lost economic output and emotional costs, was estimated to be £11 billion (7). While such estimates highlight the importance of DVA as a public health and clinical problem, information on cost-effectiveness is needed to make an economic case for investment in DVA interventions in health care, particularly when health systems are dominated by austerity.
The Identification and Referral to Improve Safety (IRIS) trial tested the effectiveness of a training and support intervention for general practice teams in two English cities (8).
Discussions about DVA between clinicians and patients were 22 times greater in the intervention practices compared with the control practices. Primary care practices that delivered the intervention also experienced a 6 fold and 3 fold increase in referrals received by DVA agencies and DVA-related notes in the patient medical records, respectively. The IRIS programme can now be commissioned across the UK: as of December 2016, 34 UK areas had commissioned IRIS; more than 800 GP practices nationally have had IRIS training, and over 5,000 women have been referred in to DVA support services by IRIS since 2010.
The cost-effectiveness of the IRIS trial was assessed using data from the trial and the programme was estimated to be good value for money (9). Given its national implementation, IRIS became a real-life, long-term intervention, raising the need for a new economic

Overview of economic evaluation
This was a cost-utility analysis, comparing IRIS with usual care in general practices. The outcome measure was quality-adjusted life years (QALYs), as recommended for economic evaluations in the UK (12). The main analysis was from a societal perspective, as many of the costs of DVA are borne outside the health system; we also estimated cost utility from an NHS perspective. Costs were calculated in 2015/16 UK£. We calculated costs and benefits over a 10-year time horizon, with future costs and outcomes discounted at an annual rate of 3.5% (12).

Model structure
We developed a Markov model (Figure1) based on the previous analysis (9). The model has five states and the cycle length was six months; this length was chosen as it reflects the average amount of time women stay in contact with DVA advocacy services. A hypothetical cohort of 10,000 women aged 16 years or older was simulated moving between the states ( Figure 1). Other than death, which is an absorbing state, women can transition between each of the other states 'Not abused', 'Abused but not identified', 'Abused and identified, seeing advocate educator', 'Abuse and identified, not seeing advocate educator'. The IRIS programme is a multi-component intervention that has been described in detail elsewhere (8,9). In brief, it consists of two two-hour multidisciplinary training sessions, for the practice clinical team and one hour training for reception and ancillary staff. They are delivered jointly by an IRIS advocate educator from a local collaborating specialist DVA agency, alongside a clinician interested in DVA, the IRIS clinical lead. The advocate educator is central to the intervention, combining a training and support role to the practices with provision of advocacy to women referred. Other intervention components include a HARK template (13) in the electronic medical record triggered by entry of clinical problem codes (such as depression, anxiety, irritable bowel syndrome, pelvic pain and assault), an explicit referral pathway to a named IRIS advocate educator, and publicity materials about DVA visible in practices. Patients referred to the advocate educator are usually seen at the referring general practice, enhancing safety and confidentiality.

Prevalence of domestic abuse
The proportion of women aged 16 years or older experiencing abuse was estimated based on published epidemiological data. This was taken from a cross sectional study carried out by Richardson and colleagues in east London (14), which reported a prevalence of 0.17 or 17% in the population of women consulting a general practitioner or practice nurse. This is an estimate of the prevalence of DVA in general practice, generalizable for England. .

Transition probabilities
There are eight transitions between states in the model. Transition probabilities were obtained using observational data from the IRIS programme, the MOSAIC (MOthers' Advocates In the conducted a cross-sectional survey to estimate community preferences for health states resulting from intimate partner violence. Using a UK-based algorithm, they found the utility of women experiencing any abuse was 0.64. When the severity/frequency of violence was low, the mean utility was 0.65 and when the severity/frequency was moderate or severe the mean utility was 0.63. For women who were abused in our model, we assumed this was moderate to severe, giving a utility score of 0.63 (21). For women seeing an advocate educator, we used the utility value of women with low abuse (0.65), implying that seeing an advocate educator slightly increased their quality-of-life scores.

Costs
We included: intervention costs, costs of onward referral, and costs associated with DVA (including costs to the UK National Health Service (NHS), lost economic output, costs to the criminal/civil justice system, and personal costs).
One IRIS advocate educator typically provides training, support and advocacy services for 24 general practices at any one point in time. Intervention costs were calculated based on the actual budget of the IRIS programme in the six sites (including advocate educator salaries, travel, recruitment, laptop, telephone, publicity, clinician consultancy, evaluation and central management costs) at a total six month cost across all sites of £272,613. This was divided by the number of registered women aged 16+ in IRIS-trained general practices in these sites (n=595,902). Costs of onward referral from the advocate educator was based on the finding of contact time from the IRIS trial, in which an onward referral was given to 57% of women in contact with an advocate educator and 63% of these women accepted this referral. Therefore, Costs associated with intimate partner violence in the UK are described by Walby and Olive (7). In their report, costs of lost economic output, health services, criminal justice system, civil justice system, social welfare, personal costs, specialised services and physical/emotional impact were individually reported, and total costs were €13,732 million (£11 billion) in 2012. We excluded costs of physical/emotional impact (€6,614 million), as they were not financial costs, but consisted of monetary valuing of health status, which in cost-effectiveness models ought to be captured in terms of QALYs; these were also not included in the original cost-effectiveness analysis. The remaining costs were converted to UK£ and inflated to 2015/16. Total costs per six months were £2,933 million. Based on the 2015 Crime Survey for England and Wales, it was estimated that 1.3 million women experienced intimate partner violence in 2015/16 in the UK (2). Mean costs per abused woman were therefore £2,043. We assumed that the costs of intimate partner abuse are similar to the costs of abuse by other family members, and that the costs would not differ between identified or unidentified abuse. In sensitivity analyses we have allowed the costs of identified abuse to increase or decrease by 10% compared to abuse that was not identified; similarly the costs of Abused and identified, seeing advocate educator were allowed to increase or decrease by 25%. Costs and utilities were applied to each health state. Total costs and QALYs for the hypothetical cohort were generated for the IRIS programme and the control group. The main outcome was the incremental costs per QALY gained. In the UK an intervention is generally considered cost-effective when the incremental costs per QALY gained are less than £20,000 (12). We also presented the results of cost-effectiveness analysis in terms of incremental net monetary benefit (NMB). This was calculated as the mean incremental QALYs per woman registered at the general practice accruing to IRIS multiplied by the decision-makers' maximum willingness to pay for a QALY (assumed to be £20,000), minus the mean incremental cost per woman. Negative incremental NMBs indicate that usual care was preferred on cost-effectiveness grounds and positive incremental NMBs favour IRIS.

Cost-utility analysis
The cost-utility analysis was conducted using pooled national data, but we have also evaluated the cost-effectiveness at different local sites. We allowed all parameters, including costs and benefits, to vary across sites and reported them individually.

Sensitivity analysis
All parameters were varied in a one-way sensitivity analysis, using lower and upper limits based on 95% uncertainty intervals. We undertook a probabilistic sensitivity analysis, drawing random samples from the probability distributions of all parameters in 1,000 simulations. The proportion of simulations with an incremental cost per QALY gained below the cost-effectiveness threshold was calculated for different values, ranging from £0 to £50,000. The results were presented in a cost-effectiveness acceptability curve. Parameter values used in the base case analysis are shown in Table 1. Over the ten-year time horizon, mean total costs per woman were £4,416 in the intervention group, compared to £4,430 in the control group (Table 2(a)). The IRIS programme therefore saves £14 per woman aged 16 and older registered to GP practices, from a societal perspective over 10 years. Total

Base case
QALYs per woman were 0.001 higher in the intervention group (6.671) than in the control group (6.669). Because the intervention was associated with lower costs and greater effectiveness the incremental cost per QALY gained was negative (i.e. IRIS dominates current practice as it is both cost-saving and more effective than usual care) and the incremental NMB was positive (£42). The incremental NMB was also positive (£22) when using an NHS-only perspective (Table 2(b)).  Figure 2). When it was varied similarly in the intervention arm, the incremental NMB varied from -£25 to £109. Figure 2 shows the 12 parameters that when varied had the highest impact on the incremental NMB.
Incremental costs and QALYs varied widely in probabilistic sensitivity analyses. The 95% uncertainty interval for incremental costs was -£151 to £37, for incremental QALYs it was -0.005 to 0.006 and for the incremental NMB it was -£247 to £351. Figure 3(a) shows a scatter plot of the incremental costs and incremental QALYs from the 1,000 simulations. The IRIS programme is cheaper and more effective than the absence of the programme (usual care), dominating current practice in 35% of the simulations and was dominated by the absence of the programme in 18% of the simulations. The IRIS programme was cost-effective in 61% of simulations when the cost-effectiveness threshold was £20,000 ( Figure 3(b)).

Summary
We found that the IRIS GP training and service programme is likely to be cost-effective and cost-saving in the UK compared to usual care. There is considerable uncertainty surrounding these results, but the probability that IRIS is cost-effective was more than 60% at the costeffectiveness threshold commonly used in the UK. IRIS was more cost-effective when costs were measured from a societal perspective as the cost savings from reducing DVA were higher. IRIS was also cost-effective when taking an NHS-only perspective. There was some variation in value for money between sites. We contacted researchers in the field and searched the NHS Economic Evaluations Database and the HTA Database at the Centre for Reviews and Dissemination (22) for costeffectiveness analyses of DVA programmes using the search terms "domestic violence" and "cost*" (28/08/2017). We identified four economic impact studies, all using modelling methods: one based on the pilot of the IRIS trial (22), another based on the main trial (9), the third based on an evaluation of independent domestic violence Advisors (IDVA) (23), and the fourth of a trial of cognitive trauma therapy for abused women who have left the abusive relationship (23). All the studies found the interventions cost-effective, despite uncertainty.
Our findings are consistent with these previous studies. Our study is the only one that analyses the economic impact of a primary care-based programme implemented outside of trial settings.

Strengths and limitations
Our analysis has the strength of being based on a previously published cost-effectiveness model, updated with real-life data. Importantly, intervention costs and the probability of referral with IRIS were based on actual clinical practice, rather than in a research setting. We also had new data for the probability of identifying abuse and for what happened to women who were abused in current practice without the programme. However, it was not possible to update all parameter values. In particular, we were unable to update the utility value estimates, although in the sensitivity analysis, we have allowed these to vary and results were relatively stable. Costs of the intervention were calculated by dividing the total costs of the programme over all registered women in practices with the IRIS programme. Many of these women will never experience abuse and therefore cannot directly benefit from the programme. If programme costs were divided over women experiencing abuse only, mean costs per woman would be higher. However, the QALYs gained would also be higher, as Another limitation is that we have used mainly data on short-term outcomes, although modelled long-term outcomes. There is unfortunately little data on long-term outcomes of DVA and the effect of advocacy, although it is generally agreed that effects last for a long time.

Implications for research and/or practice
The IRIS programme is likely to be cost-effective and cost-saving when implemented in the real life of the in the UK National Health System. In order to decrease uncertainty around the cost-effectiveness estimates of IRIS and programmes like it, more data are needed on the utilities of women identified and women seeing an advocate and on long-term outcomes associated with DVA. Furthermore, future research should endeavour to understand the impacts and economic burden of DVA on exposed children, other family members and friends.   1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57  58  59 1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57  58  59  60   F  o  r  p  e  e  r  r  e  v  i  e  w  o  n  l  y   19 We would like to thank our IRIS partners who deliver the programme in the sites, especially those in northern England, south-west England and London who took the time and effort to provide us with data.

12
Based on the Markov model, which uses health states to represent possible outcomes of the 13 intervention, we stipulated a hypothetical cohort of 10,000 women aged 16 years or older.

15
The IRIS trial was a randomised controlled trial that tested the effectiveness of a primary care 16 training and support intervention to improve the response to women experiencing DVA, and 17 found it to be cost-effective. As a result, the IRIS programme has been implemented across 18 the UK, generating data on costs and effectiveness outside a trial context. The IRIS programme is likely to be cost-effective and cost-saving from a societal perspective 4 in the UK and cost effective from a health service perspective, though there is considerable 5 uncertainty surrounding these results, reflected in the large uncertainty intervals. • We have used up-to-date routine data from several sites across England to evaluate the 9 value for money of IRIS, a domestic violence training programme.

10
• We were unable to include any impact of the IRIS programme on children exposed to 11 DVA, as to our knowledge, there are no available cohort studies focusing on the cost 12 and benefits of DVA interventions for this population.

13
• We have used mainly data on short-term outcomes, although modelled long-term 14 outcomes, as to our knowledge, no study has tracked women subject to DVA over 15 long periods of time. The lifetime prevalence of domestic violence and abuse (DVA) against women, as defined by 2 the United Nations (1), varies internationally from 15% to 71% (2). In the United Kingdom, 3 in the year ending March 2017, 7.5% of women (1.2 million) experienced domestic abuse (3). 4 Women who experience DVA suffer chronic health problems including gynaecological  The cost-effectiveness of the IRIS trial was assessed using data from the trial and the 24 programme was estimated to be good value for money (11). Given its national   identified', 'Abused and identified, seeing advocate educator', 'Abuse and identified, not 1 seeing advocate educator'. As the hypothetical cohort of women aged 16 or older were 2 considered eligible for the intervention, all results were reported as "per woman aged 16 or 3 older registered to GP practice". The IRIS programme is a multi-component intervention that has been described in detail 7 elsewhere (10, 11). In brief, it consists of two two-hour multidisciplinary training sessions, for  This research project was given exemption from NHS Research Ethics processes, as it was 1 classified as service evaluation. When observational data were unavailable, we have chosen to 2 use peer-reviewed published data that was relevant to general practice and the UK. Each 3 relevant parameter and its source are described in detail below. 4 5 Prevalence of domestic abuse 6 The proportion of women aged 16 years or older experiencing abuse was estimated based on 7 published epidemiological data. This was taken from a cross sectional study carried out by 8 Richardson and colleagues in east London (17), which reported a prevalence of 0.17 or 17% 9 in the population of women consulting a general practitioner or practice nurse. This is an 10 estimate of the prevalence of DVA in general practice, generalizable for England.  women in the three Abused states was also determined by this process.  We included: intervention costs, costs of onward referral, and costs associated with DVA 5 (including costs to the UK National Health Service (NHS), lost economic output, costs to the 6 criminal/civil justice system, and personal costs). Costs were also reported per woman aged 7 16 or older registered to GP practice.  The cost-utility analysis was conducted using pooled national data, but we have also 5 evaluated the cost-effectiveness at different local sites. We allowed all parameters, including 6 costs and benefits, to vary across sites and reported them individually. Patient and Public Involvement (PPI) 19 We did not directly include PPI in this study, but the data collected from local IRIS 20 Programmes was developed with PPI. Parameter values used in the base case analysis are shown in Table 1, which also includes the 1 parameters used in the original trial to allow for a direct comparison. The main differences 2 between the parameters for this study and the trial parameters lie in the transition probabilities 3 relating to the health state of 'abuse but not identified' and its cost. 4 Over the ten-year time horizon, mean total costs per woman were £4,416 in the intervention 5 group, compared to £4,430 in the control group (Table 2(a)). The IRIS programme therefore 6 saves £14 per woman aged 16 and older registered to GP practices, from a societal 7 perspective over 10 years. Total QALYs per woman were 0.001 higher in the intervention 8 group (6.671) than in the control group (6.669). Because the intervention was associated with 9 lower costs and greater effectiveness the incremental cost per QALY gained was negative (i.e.

10
IRIS dominates current practice as it is both cost-saving and more effective than usual care) 11 and the incremental NMB was positive (£42). The incremental NMB was also positive (£22) 12 when using an NHS-only perspective ( Across all sites combined, results were most sensitive to varying the transition probability 2 from Abused but not identified to Not abused. When in the control arm this was varied from 3 0.049 to 0.051, the incremental NMB varied from £110 to -£26 ( Figure 2). When it was 4 varied similarly in the intervention arm, the incremental NMB varied from -£25 to £109.
5 Figure 2 shows the 12 parameters that when varied had the highest impact on the incremental 6 NMB.  We found that the IRIS GP training and service programme is likely to be cost-effective and results, but the probability that IRIS is cost-effective was more than 60% at the cost- of uncertainty, and the fact that the probability that IRIS is cost-effective is just higher than 5 50% reflects the fact that IRIS is (slightly) favoured over the alternative according to our base 6 case estimates. IRIS was more cost-effective when costs were measured from a societal 7 perspective as the cost savings from reducing DVA were higher. IRIS was also cost-effective 8 when taking an NHS-only perspective. There was some variation in value for money between 9 sites, which appears to be driven mainly by the different rates of identification and/or referral, 10 although different local costs have also contributed.  the only one that analyses the economic impact of a primary care-based programme 1 implemented outside of trial settings.  Another limitation is that we have used mainly data on short-term outcomes, although 3 modelled long-term outcomes. There is unfortunately little data on long-term outcomes of 4 DVA and the effect of advocacy, although it is generally agreed that effects last for a long 5 time. This, however, bias our estimates against the intervention, implying our results are 6 conservative.  We would like to thank our IRIS partners who deliver the programme in the sites, especially 1 those in northern England, south-west England and London who took the time and effort to 2 provide us with data.
3 Data sharing 4 The anonymised data used in this study can be obtained from the corresponding author.

Figure 1. Health states and movement between health states in Markov model.
Legend: The model starts with all women in either the 'Not abused' state or one of the states associated with abuse, based on the prevalence of DVA (see text). Women in the 'Not abused' state could stay in this state, move to 'Abused but not identified' or die from any cause. Once women were in the 'Abused but not unidentified' state, they could stay in that state, move back to 'Not abused', move to 'Abused and identified, seeing advocate' or 'Abused and identified, not seeing advocate' or die. Women in the 'Abused and identified' states could stay in these states, move back to 'Not abused' or die.  Figure 2. Univariate sensitivity analysis.
Legend: All analyses are as for the base case analysis with univariate adjustment of the parameters listed (see text). Results are point estimates of the incremental net monetary benefit (NMB) of the intervention vs. control. The incremental net monetary benefit is calculated at a maximum willingness to pay for a QALY of £20 000.    . Women in the 'Not abused' state could stay in this state, move to 'Abused but not identified' or die from any cause. Once women were in the 'Abused but not unidentified' state, they could stay in that state, move back to 'Not abused', move to 'Abused and identified, seeing advocate' or 'Abused and identified, not seeing advocate' or die. Women in the 'Abused and identified' states could stay in these states, move back to 'Not abused' or die.
160x90mm (300 x 300 DPI)  control. The incremental net monetary benefit is calculated at a maximum willingness to pay for a QALY of £20 000.
280x90mm (300 x 300 DPI)   Cost-utility analysis in UK general practices, including data from six sites which have been 10 running IRIS for at least two years across England.

12
Based on the Markov model, which uses health states to represent possible outcomes of the 13 intervention, we stipulated a hypothetical cohort of 10,000 women aged 16 years or older.

15
The IRIS trial was a randomised controlled trial that tested the effectiveness of a primary care 16 training and support intervention to improve the response to women experiencing DVA, and 17 found it to be cost-effective. As a result, the IRIS programme has been implemented across 18 the UK, generating data on costs and effectiveness outside a trial context.  1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57  58  59  60   F  o  r  p  e  e  r  r  e  v  i  e  w  o  n  l  y   6 cost-effective in 61% of simulations using real life data when the cost-effectiveness threshold 1 was £20 000 per QALY gained as advised by NICE.

Conclusion:
3 The IRIS programme is likely to be cost-effective and cost-saving from a societal perspective 4 in the UK and cost effective from a health service perspective, though there is considerable 5 uncertainty surrounding these results, reflected in the large uncertainty intervals. • We have used up-to-date routine data from several sites across England to evaluate the 9 value for money of IRIS, a domestic violence training programme.

10
• We were unable to include any impact of the IRIS programme on children exposed to 11 DVA, as to our knowledge, there are no available cohort studies focusing on the cost 12 and benefits of DVA interventions for this population.

5
Base case 6 Parameter values used in the base case analysis are shown in Table 1, which also includes the 7 parameters used in the original trial to allow for a direct comparison. The main differences 8 between the parameters for this study and the trial parameters lie in the transition probabilities 9 relating to the health state of 'abuse but not identified' and its cost.    in 61% of simulations when the cost-effectiveness threshold was £20,000 ( Figure 3(b)).  We found that the IRIS GP training and service programme is likely to be cost-effective and considerable uncertainty surrounding these results, but the probability that IRIS is cost-7 effective was more than 60% at the cost-effectiveness threshold commonly used in the UK.
160x90mm (300 x 300 DPI)  control. The incremental net monetary benefit is calculated at a maximum willingness to pay for a QALY of £20 000.

Characterising uncertainty 20a
Single study-based economic evaluation:Describe the effects of sampling uncertainty for the estimated incremental cost and incremental effectiveness parameters, together with the impact of methodological assumptions (such as discount rate, study perspective). 20b Model-based economic evaluation: Describe the effects on the results of uncertainty for all input parameters, and uncertainty related to the structure of the model and assumptions. If applicable, report differences in costs, outcomes, or cost-effectiveness that can be explained by variations between subgroups of patients with different baseline characteristics or other observed variability in effects that are not reducible by more information.  1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57  58  59  60