Those with drug-related problems tend to be difficult to find, and addiction is difficult to measure. Experts consistently fail to agree on what constitutes an addict, problematic use or problematic user. Estimates as to how many people are experiencing drug problems have to be drawn from different sources, using different ways of measuring.
National Drug Treatment Monitoring System
In the UK, figures for drug users presenting for treatment related to problematic drug misuse are drawn directly from the National Drug Treatment Monitoring System (NDTMS). The NDTMS relates to the process of collecting, collating and analysing information from and for those involved in the drug treatment sector. The NDTMS is a development of the regional drug misuse databases (RDMDs), which have been in place since the late 1980s.
The NDTMS collects data on both those drug misusers presenting for treatment and those in treatment. It does not count the number of people addicted to a drug, but those seeking help for their drug use or associated problems (and subsequently have their information sent to the database) and is therefore not reflective of the community at large.
While the data on new notifications helps track how many people are newly seeking help for their drug use, it does not indicate how many people are addicted or having problems with drugs in total. To find this we have to find out firstly how many people are in treatment as a whole and then how many people are having drug-related problems but not seeking help.
Estimating the number of addicts in and not in treatment
According to the report 'Provisional Statistics from the National Drug Treatment Monitoring System in England, 2001/02 and 2002/03', the number of users reported as being in treatment with drug misuse agencies and GPs was around 140,900. However, this only provides an estimate of the number of addicts in treatment rather than the total experiencing problems but not receiving treatment. Although it is almost impossible to count the number of people experiencing problems at any one time, there are ways of guessing. 
A number of local studies have used what is called the capture-recapture method. Originally used to estimate population sizes, such as the number of fish in a pool, this method simply calculates the number of recaptured or missed fish every time the net is cast. This ratio is then applied to the whole population. In essence, two consecutive samples of the same population are taken. The first sample tags and releases all that are captured. The second then reveals the number of fish recaptured and those that are not (by way of the tags). This ratio of tagged to untagged can then be applied to the rest of the population. Therefore, if, on the second catch, five out of 100 fish are tagged, then we can estimate that every time we cast the net, we catch only 5 per cent of the total fish population (1 in 20 fish are caught every time we fish). From this we can estimate that in that pool, there are roughly 2000 fish.
Using this method, several studies have estimated the number of problematic drug users in their area. In 1984, Hartnoll collected data on the number of opiate users in North London who had attended a drug clinic and those admitted to hospital for infectious diseases. Comparing the sources, they found that a fifth of the hospital sample had also attended the drug clinic. The researchers used this ratio to estimate that the total number of opiate users was five times the number who attended the drug clinic. 
The figure was revised down to 1:3 when the advent of HIV/AIDS meant that more methadone was being prescribed for longer periods. Taking that figure would mean that in 1995, there were around 120,000 problem drug users in the UK. The lowest current estimate in the updated drug strategy is 250,000, but recent Home Office research conducted by York University indicated a range of 280,000–500,000. 
Whatever the figure, the number of drug users in treatment greatly underestimates the number of people dependent on one substance or another. Using the figure mentioned above for 2001/02, we can estimate the total number of addicts in Great Britain during that period by applying this multiplier of three.
Using this formula, we can predict that for 2001/02, there were roughly 422,700 (140,900 x 3) people experiencing drug-related problems.
This figure represents the number of people counted as addicted across a one year period and not at any one point in time. This figure will be much less, due to many people stopping or starting after or before any point in time.
Factors affecting these estimates are availability and accessibility of treatment services, and people's willingness to seek treatment.
 Provisional Statistics from the National Drug Treatment Monitoring System in England, 2001/02 and 2002/03 (PDF 36KB) UK. Department of Health. London: DOH, 2003. 9p.
 Hartnoll R, Daviaud E, Lewis R & Mitcheson, M (1985) Drug Problems: Assessing local needs. A Practical manual for assessing the nature and extent of problematic drug use in a community, London: Drug Indicators Project.
 Godfrey C., Eaton G., McDougall C., Culyer A. University of York.The economic and social costs of Class A drug use in England and Wales, 2000. London: Home Office Research; Development and Statistics Directorate, 2002. 62p Home Office Research Study 249 Dup.
ISBN 1840828749. http://www.homeoffice.gov.uk/rds/pdfs2/hors249.pdf
Updated January 2004