Restricting promotions of food and drink high in fat, sugar or salt: business and regulatory impact assessment - partial

ANNEX A: Department of Health and Social Care Calorie Model

Analysis for Scotland Government Health and Social Care – Promotion Modelling 05/04/2022

Background

The Scottish Government has requested from the UK Department of Health and Social Care (DHSC) modelling on the long-term economic benefits produced by a set of calorie reductions. These calorie reductions represent the total potential health benefits of a set of promotions restrictions and interventions currently considered by the Scottish Government, the scenarios are linked to price and multibuy promotions.

DHSC maintains a “calorie model” which estimates the benefits in net present value (NPV) cash terms of reduced calorie consumption. The model does not estimate the size, scope or duration of the calorie reduction itself. Instead, it takes the reduction as an input parameter and converts it to an NPV value.

The DHSC calorie model gives total health benefits, this represents the total modelled benefits of the scenario discounted to estimate the present value of these future benefits, as consistent with the HMT green book¹. The total health benefits are made up of the four benefits modelled:

• Health benefits, the monetised increase in QALYs^[65]
• NHS costs, the reduction in NHS treatment costs
• Economic output, the increase in economic output
• Social Care costs, the reduction in Social Care costs

Modelling

Modelling was done on 1 – 5 April 2022 using the DHSC calorie model v3.1.2

The calorie model is a cohort-based Markov model that evaluates the economic impact of a weight reduction policy/intervention applied to that cohort.

Inputs

• Policy Lifetime – all scenarios and options modelled have a 25-year policy lifetime as detailed in consultation with the Scottish Population health economics team.
• Evaluation Period– all scenarios and options modelled have a 25-year evaluation period as detailed in the consultation meeting with the Scottish Population health economics team.
• Model Cohort – the model simulates a policy that targets the whole population.
• NHS opportunity cost multiplier -– all scenarios and options modelled without the NHS opportunity cost multiplier as advised by RPC following HMT guidance.
• Calorie reductions – calorie reductions are as detailed in table 1. Calorie reductions are implemented as detailed in the consultation with the Scottish Population health economics team
• Cohort: This policy is applied to all ages and affects the whole population.

Adjustments

The results given in table A2 come from the DHSC calorie model, which models health benefits for England, adjusted for Scotland.

A simple pro-rata change has been made to adjust the results for Scotland. This adjustment involves reducing the model output by the ratio of the English and Scottish populations, this ratio has been set at 10%. This is reasonable provided we make four assumptions:

• The demographic make-up of the English and Scottish populations is similar.
• The levels of obesity prevalence in England and Scotland are similar.
• The prevalence of the six diseases in the model (CHD, Stroke, Diabetes, Liver disease, Breast cancer and Colorectal cancer) in England and Scotland are similar; and
• The costs associated with these conditions are similar in both countries.

Analysis by the Scottish government suggests these are sensible assumptions

Output

The benefits of option A are 0, all other option benefits are given in table A2. The 25-year discounted total benefits, the four component benefits and the undiscounted QALY gain are given in table A2 in 2013 price and are adjusted to the Scottish population.

*Monetised benefits are discounted to reflect present values, as detailed in the HMT green book

*A QALY is valued at £60,000, as detailed in the HMT green book

*QALY count is undiscounted

*Monetised benefits are rounded to the nearest million and QALY counts to the nearest 1,000

* NHS, social care and economic benefits are based on 2013 pricing. This can be adjusted to 2019 by uprating results for these three benefits types by the GDP deflator². The QALY benefits should not be uprated.

Overview of Central results:

All price promotion

The policy for all price promotions modelled (based on a central calorie reduction of -88 calories per person per day) gives a total economic health benefit, in discounted GBP, of £9,523 million. This total NPV is made up of the 4 economic effects modelled: QALYs, economic output, social care costs and NHS costs, shown in table 2. Each of these components has been adjusted by the appropriate discount rate to give the approximate present cash value of future benefits, to be consistent with HMT green book guidance. QALYs are valued at £60,000 per QALY and all others use 2013 prices.

The policy increases the undiscounted total QALY count by 158,300 (124,900 discounted at the health discount rate of 1.5%) and results in approximately 89,100 fewer premature deaths (deaths of individuals under the age of 75) in the model cohort of 6.7 million during the evaluation period of 25 years. A QALY refers to a measurement of disease burden, which includes the quantity and quality of life lived. One QALY equates to one year in perfect health. QALY scores range from 1 (perfect health) to 0 (dead). Increased QALY counts refer to increase in the number of years spent in higher quality health.

Multibuy promotion

The policy for all multi-buy promotions modelled (based on a central calorie reduction of -22 calories per person per day) gives a total economic health benefit, in discounted GBP, of £2,418 million. This total NPV is made up of the 4 economic effects modelled: QALYs, economic output, social care costs and NHS costs, shown in table 2. Each of these components has been adjusted by the appropriate discount rate to give the approximate present cash value of future benefits, to be consistent with HMT green book guidance.

This policy increases the undiscounted total QALY count by 4,010 (31,600 discounted at the health discount rate of 1.5%) and results in approximately 22,600 fewer premature deaths (deaths of individuals under the age of 75) in the model cohort of 6.7 million during the evaluation period of 25 years. A QALY refers to a measurement of disease burden, which includes the quantity and quality of life lived. One QALY equates to one year in perfect health. QALY scores range from 1 (perfect health) to 0 (dead). Increased QALY counts refer to increase in the number of years spent in higher quality health.

Additional caveats and advice on using these results

1. The analysis is illustrative, carries significant uncertainty and should be treated with caution. The calorie model itself is a deterministic model and so gives no formal confidence intervals, therefore levels of uncertainty in the input parameters of the model (including those relating to the expected change in calorie intake) has been reflected through sensitivity analysis.

2. The model says nothing about what level of calorie reduction might be expected from any intervention. That must be assessed separately before modelling, using whatever evidence is available/appropriate.

3. The experience in England suggests that many interventions may deliver relatively small (but still worthwhile) calorie reductions, with policies in England relating to promotions delivering calorie reductions of around 10-70 kcal per day. The central estimates of 22 kcal and 88 kcal would appear reasonable on this basis assuming it includes a wide range of Scottish restrictions and interventions.

4. Although the model estimates a wide range of benefits from the calorie reduction and the subsequent reduction in obesity, it is not comprehensive. Some medical conditions linked to BMI (such as musculo-skeletal conditions) are not currently included.

5. The model takes no account of local circumstances or policy targeting. Excess intake, associated costs and/or potential reductions, and therefore results, may differ markedly across the country and population. Results are a national average and make no comment on regional or demographic variation

6. The model also says nothing about health inequalities.

7. Documentation for version 3 of the calorie model has been published, version 3.1.2 used in this analysis has the same fundamental structure as version 3 but contains many small improvements, all improvements have been subject to appropriate levels of QA within DHSC.