Research and analysis

Gambling treatment need and support in England: main findings and methodology

Updated 23 January 2024

Applies to England

Main points

Adults who might benefit from gambling treatment and support

In England, there are almost 1.6 million adults who gamble who may benefit from some type of treatment or support for harmful gambling.

Of these, the majority (970,000) might benefit from a ‘level 2 intensity’ treatment, which typically involves 2 or 3 sessions of motivational interviewing delivered by gambling-specialist practitioners. Around 243,000 adults might benefit from a ‘level 4 intensity’ treatment, which typically involves 8 to 14 sessions of psychologist-led cognitive behavioural therapy (CBT) for gambling disorder.

Almost 40,000 adults might benefit from the most intensive type of treatment, which typically involves a 12-week residential treatment programme including one-to-one therapy and group sessions. Clinical practitioners should consider the appropriateness of this type of support on a case-by-case basis, taking into account the adult’s personal circumstances.

Table 1 shows regional estimates and rates. London is estimated to have the highest rate per 100,000 of population, while the South East has the lowest rate. The difference in rates is largely due to different demographics in the regions. The table also shows 95% confidence intervals with upper and lower limits. This means that there is a 95% probability that the true value lies between these upper and lower limits.

Table 1: estimates of the number of adults who gamble who might require treatment or support by region

Region	Rate per 100,000 population	Point estimate	95% confidence interval lower limit	95% confidence interval upper limit
London	3,835	266,707	214,324	336,081
North West	3,782	219,210	181,932	266,535
West Midlands	3,751	174,641	138,291	224,863
Yorkshire and the Humber	3,721	161,956	135,285	195,521
North East	3,656	78,492	65,223	95,327
East Midlands	3,514	135, 576	112,285	165,154
South West	3,304	150,213	124,969	182,050
East of England	3,261	160,194	133,834	193,366
South East	3,223	233,185	195,065	281,096
England	3,554	1,580,175	1,301,206	1,939,991

Children living in households with adults who gamble who might need support

Table 2 shows the estimates of the numbers of children living in a household with an adult who gambles who might require treatment or support. The overall estimate is that there are 912,805 children living in these kinds of households in England.

Table 2: estimates of the number of children living with an adult who gambles who might require treatment or support by region

Region	Point estimate	95% confidence interval lower limit	95% confidence interval upper limit
London	165,748	133,776	207,942
South East	134,016	112,624	160,826
North West	125,076	104,419	151,224
West Midlands	104,588	82,994	134,310
Yorkshire and the Humber	95,296	80,097	114,355
East of England	93,106	78143	111,865
South West	75,357	63041	90,833
East Midlands	76,317	63525	92507
North East	43,301	36,245	52,211
England	912,805	754,864	1,116,074

Methodology

The Office for Health Improvement and Disparities (OHID) commissioned researchers at the University of Sheffield and the University of Glasgow to produce estimates of the number of adults who gamble in England who might benefit from treatment or support. The estimates are by national, regional and local authority level.

Producing these estimates required defining suitable treatment or support type categories as well as applying small area estimation techniques to nationally representative data. The estimates are only for the support needs of people who gamble and do not consider the support needed for other people affected by their gambling.

Delphi consensus survey

The researchers used adapted Delphi consensus survey methods in an online survey. Delphi is a research survey technique used to collect data from respondents within their area of expertise. It deals with divergent opinions to achieve consensus.

The panel of survey respondents was comprised of 22 people who have current professional knowledge of commissioning and delivering treatment and support services for harmful gambling. Many different stakeholders work in this area, including NHS clinics, charities and private providers. Potential respondents were identified through purposive sampling (sampling to deliberately include a range of relevant stakeholders) based on recommendations from:

• OHID
• the professional networks of the research team
• the project’s advisory group
• other participants

Data collection took place between September and December 2022.

Before providing informed consent, respondents were provided with high-level information about the project, and about the aims and objectives of this online Delphi study and their role in it.

The researchers have published an overview of the methods and results of the online Delphi survey on the Open Science Framework (OSF) research platform.

Types of support and treatment for harmful gambling

For the prevalence estimates, OHID and the research team grouped the support and treatment available for people who gamble into 6 types.

1. Brief advice. This typically involves a brief conversation that could be delivered by non-specialists. It also includes referral to self-help groups, such as online forums.
2. Extended brief interventions. This typically involves 2 or 3 sessions of motivational interviewing delivered by gambling-specialist practitioners.
3. Psychosocial interventions delivered by voluntary sector services. This typically involves around 6 sessions of psychosocial treatment delivered one-to-one or in a group format by gambling treatment practitioners and vary in their techniques and structure. The minimum qualifications needed to deliver these interventions depends on the service provider.
4. Psychologist-led CBT. This typically involves 8 to 14 sessions of CBT for gambling disorder, delivered by clinical psychologists or CBT-accredited psychotherapists. It may also include psychological therapy for co-existing mental health conditions.
5. Intensive residential treatment. This typically involves a 12-week residential treatment programme that would include one-to-one therapy and group sessions.
6. Peer support. For example, peer support groups like Gamblers Anonymous, GamLearn, PeerAid, as well as GamFam.

The sixth type (peer support) was added following recommendations from the project’s advisory group, because it is an important type of support for gambling problems in the UK. However, the results from the Delphi survey indicated that peer support is seen as suitable for anyone who gambles or used to gamble who feel they might benefit from it. So, the researchers did not estimate the number of adults in each local authority who might benefit from peer support.

Delphi participants were asked which of the types of treatment or support different types of people who gamble might benefit from. This was based on information such as their problem gambling severity index (PGSI) score (a scoring system used to measure at-risk behaviour in people experiencing problems with gambling) and other considerations such as their alcohol use or mental health. This process resulted in the decision that people who score 1 or more on the PGSI could benefit from some type of treatment or support.

Prevalence rate estimates

The researchers estimated the prevalence rates using NHS England’s Health Survey for England. They pooled the surveys for 2015, 2016 and 2018 to obtain a large enough sample size to base the estimates on. They then used a multinomial logistic regression (a statistical technique that relates an outcome to one or more variables) to estimate the probability of a survey respondent being classified as belonging to one of the 5 treatment types or a ‘no treatment’ option, based on:

• age group (18 to 24, 25 to 34, 35 to 54, 55 and over)
• sex (male, female)
• quintile of index of multiple deprivation (a measure of how deprived a local area is and splits the local authorities in England into 5 equal size groups)
• regional unemployment rate

They then applied probabilities to the mid-year population estimates for 2020 (the most recent available at the time) provided by the Office for National Statistics for each upper tier local authority, using their unemployment rates. This gives the estimated population for each type of treatment.

The researchers derived the estimates of children living in a household with an adult who gambles who might require treatment or support by multiplying the adult prevalence rate by the mean number of children in the household using the adults’:

• age group (as before)
• sex
• quintile of index of multiple deprivation
• region