You are in Research

UN World Drug Report 2000: Contents, Omissions and Distortions

by Carla Rossi [1]
Department of Mathematics
Università di Roma "Tor Vergata"
Via della Ricerca Scientifica
00133 Roma

 

1. Overall judgement

When analysing the new-born WDR 2000, we cannot help comparing it with its "big brother", the WDR 1997. The first worrying aspects emerge from a comparison of the number of pages, the contents, the assumption of responsibility by a clearly identified work group, the length of the introduction by the Executive Director in relation to the rest of the report, and the inclusion of specific contributions by authoritative international experts.

Table 1.

Qualitative and quantitative comparison between the

WDR 1997 and the WDR 2000.

	WDR 1997	WDR 2000
Number of pages	332	172
Chapter titles	1-Recent trends and development in cultivation, production, trafficking and consumption - an overview, 2- Theories and interpretations of illicit drug use, 3-The health and social consequences of drug abuse, 4-The illicit drug industry: production, trafficking and distribution, 5-Drugs and public policy, 6-Strategic programmes, 7-Country profiles.	1-Recent trends in production, trafficking and consumption: an overview, 2-The three pillars of demand reduction: epidemiology, prevention, treatment, 3-Alternative development.
Length of introduction: number pages	1	21
Existence of specific working group and assumption of responsibility	Yes	No
Number of specific contributions by authoritative international experts	6	None

 

The information contained in Table 1 clearly demonstrates the limited and limiting overall structure of the WDR 2000. A particular cause for concern is the systematic exclusion of quantitative information on the black market and on money laundering (Chap. 4 WDR 1997), of information on policies implemented at national and international level and on strategic programmes (Chaps. 5 and 6 WDR 1997), and of relevant quantitative information on the development of the phenomenon provided by the Country profiles (Chap.7 WDR 1997). I believe that the exclusion of these subjects, perhaps planned right from the beginning, can be explained in the light of the twisted interpretations and distortions of data and information in order to back what is evidently a form of self-defence, clearly expressed in the 21-page introduction. Such interpretation would not have been possible if the report had contained exhaustive information on the subjects excluded, as can be demonstrated using data from official national and international sources, and as is proven below, also on the basis of some of the information which is present in the WDR 2000, especially that which is reproduced in BOX 1A on pages 36-38 (graphs reproduced in Graph 1) and in Figure 17 on page 39 (reproduced in Graph 2).

From an analysis of the information contained in the BOX and in the figure, it is possible to deduce the continuing growth in the market and in the abuse of drugs at global level. The box could well be entitled: the failure of supply reduction policies at global level, as we will see.

From figures 13 and 14 of Graph 1, it is clear that the trend in seizures is directly proportional to the trend in production. In other words, the greater the amount of drugs produced, the greater the amount of drugs seized. The proportion of the amount seized is therefore constant with respect to production, which means that it can be considered a sort of fixed tax which traffickers have to pay, much lower than the taxes paid by any retail traders to be able to run their business (in the UN ODCCP publication Global Illicit drug trends 2000 we learn that this tax, called the "interception rate", does not exceed 17% for opiates and 46% for cocaine in recent years). Figures 15 and 16 in the Graph 1 show how the amount of seizures is directly proportional to consumption, measured by deaths for heroin and by survey data among American 18-year-olds for cannabis. These figures confirm what has already emerged from the two previous figures: that the more the market grows, the more seizures grow. The effectiveness of seizure action has not changed, and remains rather low. The Box therefore provides the only information contained in the Report on the global trend of the impact of current policies on the drugs market, and is not positive.

The four figures considered tell us nothing about the development over the years of the market and of consumption. However, the results obtained from the previous analyses allow us to state with reasonable certainty (given the correlation coefficients shown above, which are very close to one) that while the data show an increase in seizures, behind this increase there lies an identical increase in the market and in consumption.

Let us now consider Graph 2, which reproduces Figure 17 published on page 39 of the WDR 2000. The trend of the curve is clearly upward, and we can consequently conclude that the heroin market and heroin consumption have been growing constantly at global level since the early 1980s.

This does not mean, of course, that the trend shown is true of every country, but only that the global mean is rising. This means that while in some countries there may have been a reduction in heroin use, in other countries consumption has risen so much that it not only compensates for the reductions but also produces an upward trend at global level, as we can see from the graph. In other words, what we see is a phenomenon well known to economists: when a market contracts due to natural saturation or maybe because of action taken at local level, it moves to open another market elsewhere which will more than make up for the losses sustained in the first market. All this contradicts the presumed, much-advertised positive effects of current policies at global level.

Graph 1.
Figures published in BOX 1A on pages 36-38 of the WDR 2000.

 

Graph 2.
Figure 17 on page 39 of the WDR 2000.

 

In any case the WDR 2000 itself contains a summary of the results of the standard ARQ questionnaires, reproduced here in Table 2, which show a wide diffusion and expansion of the drugs markets at global level for all substances in general, and above all for the stimulants that form the "new fashion" (+100% of countries report an increase between 1992 and 1998). If we look in detail at the situation with regard to heroin, however, what we see is precisely the effect of the saturation of old markets and the opening of new markets; there is an increase, in fact, in both countries that report a reduction in use (saturation) and countries that report an increase (new markets). Overall, however, there is a prevailing increase at global level, as the previous graph shows.

Table 2.

Trends in the prevalence of some drugs (source: World Drug Report 2000)

% of countries reporting	Increasing use 1992	Increasing use 1998	Decreasing use 1992	Decreasing use 1998
Substance Cannabis	40	57	13	13
ATS	25	49	13	13
Heroin	42	47	8	13
Cocaine	33	33	8	3
Opium	10	16	19	11

 

 

2. Brief analysis of the contents of the 3 chapters of the WDR 2000.

Chapter 1, apart from the BOX cited above and some results of the analysis of the "ARQ" questionnaires, is just a summary (a kind of potted guide) of the publication cited above (UN ODCCP Global Illicit Drug Trends 2000), which contains a lot more information and data. It is not possible to verify the information and the quality of the data provided from UN ODCCP sources (the tables reproduced in the appendix to the WDR 2000 refer to this chapter).

Chapter 2 is a collage of analyses of data drawn from various sources, many also available on-line: it could be defined as merely careless if it were not for the evident distortion which we will return to later (section 3 and section 4).

Chapter 3, only 15 pages long, describes a number of cases and episodes relative to international interventions. It is not possible to verify the accuracy of the information provided, which consists mainly of qualitative considerations based on internal information.

3. The main distortions and "tendentious?" omissions of data from the EMCDDA[2].

Let us begin with Table 3 (Table 2 on page 93 of the WDR 2000), which refers to the national prevalence estimates for problem drug users, citing the EMCDDA Annual Report 2000, the ONDCP Annual Report 2000 and the UNDCP ARQ Data as the sources of data. With regard to the data from the EMCDDA table (the other data cannot be checked immediately), we find:

Table 3.
Estimates for ‘problem drug users’[3] per 1000 inhabitants,
age 15-64 in the late 1990s…

Country	Range of estimates	Mid-range estimate	Mean estimate
Finland	0.5 – 4.2	2.4	1.9
Sweden	2.5 – 3.5	3.0	3.0
Denmark	2.9 – 4.0	3.4	3.6
Norway	3.2 – 4.6	3.9	3.9
Germany	1.4 – 3.0	2.2	2.2
Austria	2.9 – 3.4	3.2	3.2
Ireland	1.9 – 5.7	3.8	3.4
France	3.2 – 4.6	3.9	4.1
Spain	3.1 – 6.6	4.9	4.9
Benelux	2.3 – 7.7	5.0	2.8
U.K.	2.3 – 8.9	5.6	6.2
Italy	4.4 – 8.3	6.4	6.6

 

In realty the definition of ‘problem drug use’ given by EMCDDA (Annual Report 2000 page 14) is the following:

‘intravenous or long-duration/regular use of opiates, cocaine and/or amphetamines’,

which is completely different from that which the WDR 2000 attributes to EMCDDA in the caption to the Table.

What is even more serious, however, is the distortion of the data[4].

The first distortion is of a methodological nature, and consists in the calculation of the mean estimates shown in the final column of Table 3. These mean estimates are meaningless in that they have been obtained by putting together, in an incorrect manner, all the estimates for the same country obtained through different methods and from different data - a bit like adding up pears and books, for example. This type of error can be put down to methodological incompetence on the part of those responsible for the table, and to carelessness.

The second distortion is not justifiable in anyone with a degree of expertise on the subject, unless he or she is trying deliberately to back a thesis outlined a priori by bending the data in the direction desired, given that without distortion this thesis would not be supported. To explain this in simple terms, and to make it easy for everyone to understand, the relevant part of the original EMCDDA Table (Table 4) is reproduced below (the complete table is available at www.emcdda.org).

The estimate given in the WDR 2000 as the estimate for Benelux is therefore not present in the original table: it is thus not drawn from EMCDDA data, but is an ad hoc calculation made by putting together the estimates for Belgium, Holland and Luxembourg in a worthless jumble (there has not even been an attempt to produce a weighted mean by taking account of the different populations of the three countries) The three countries actually behave in completely different ways with regard to this problem. In Luxembourg, for example, the figures are more than double those for Belgium as three times higher than those for Holland. I believe that an operation such as that performed to put together the WDR 2000 Table can only have been decided on to obscure the Dutch data, which are much more positive than all the others, by using a shameful trick. The operation is not only seriously flawed from a methodological point of view; it also suggests that the source of the data as they are presented is the EMCDDA, and there is no mention of any ad hoc synthesis of data.

Table 4.
National prevalence estimates of problem drug use in the EU and Norway
(prevalence rates of problem drug use per 1000 inhabitants aged 15-64 1996-1998).

Country	Overall range of estimates
Finland	0.5 – 4.2
Sweden	2.5 – 3.5
Denmark	2.9 – 4.0
Norway	3.2 – 4.6
Germany	1.4 – 3.0
Austria	2.9 – 3.4
Ireland	1.9 – 5.7
France	3.2 – 4.6
Spain	3.1 - 6.6
Luxembourg	6.7 – 7.7
Belgium	3.0
Netherlands	2.3 – 2.7
U.K.	2.3 – 8.9
Italy	4.4 – 8.3

 

 

On page 93 of the WDR 2000 we find the following comment:

Estimates for Western Europe indicate that "problem drug use" affects on average[5] 4 ½ persons per 1000 inhabitants age 15-64 (mean estimate of 15 countries). Estimates range from an average 2-3 in Finland and Sweden and some other countries of continental Europe, including Germany and Austria, to level around 4 in France, 5 in Spain and 6-7 in the UK, Switzerland, Italy and Luxembourg.

We can compare this with the comment on the original Table, on pages 14-15 of the EMCDDA Annual Report 2000:

…Prevalence rates seem highest in Spain, Italy, Luxembourg and the UK…and lowest in Belgium, Germany, the Netherlands, Austria, Finland and Sweden…Intermediate rates are reported in Denmark, Spain, France, Ireland and Norway.

It is odd that in the WDR 2000 it is precisely Belgium and Holland that have disappeared from the original EMCDDA list of countries with low prevalence rates, while the countries remaining high on the list include Finland (which is much less important), and above all Sweden, which in the original table has a mean prevalence similar to that of Belgium and much higher than that of Holland. What does remain is the explicit comment on Luxembourg, which is not consistent with the table in which the three countries (Belgium, Holland and Luxembourg) are presented together as Benelux. In this case, leaving the explicit reference to Luxembourg is like suggesting that the whole Benelux area behaves in the same way, which is obviously untrue.

Another distortion regards the graph presented as Figure 11 on page 99, for which the EMCDDA is the stated source. The figure is reproduced in Graph 3. The comment on the figure is reproduced in the BOX that follows it.

Graph 3.
Figure 11 on page 99 of the World Drug Report 2000.

 

 

Comment on figure 11 on page 99 of the WDR 2000.

Data provided by the European Monitoring Centre for Drugs and Drug Addiction suggest that a stabilization - following a strong intensification of demand reduction efforts – was actually achieved in the countries of European Union in the 1990s. Following strong increase in the 1980s, the number of acute drug deaths stagnated in the European Union in the 1990s. If the trends of the 1980s had continued – which might have been the case without appropriate interventions – the number of acute drug-related death cases, less than 7000 a year in the late 1990s, could well have been three times higher in the late 1990s (See figure 11).

In actual fact the source for the figure is not the EMCDDA as stated: the clear curve for the extrapolation of death trends does not appear in any EMCDDA document, as it is clearly an arbitrary interpretation, methodologically and substantially incorrect, just as the curve regarding cases observed has not been drawn from the stated source. This distortion is particularly serious in that the caption attributes the analysis to the EMCDDA, which on pages 18-19 of the 2000 Annual Report (wrongly stated as the source) actually states as follows:

EMCDDA comment

In many countries, acute drug-related deaths increased markedly from the late 1980s to the mid-1990s. This rise has since stabilized in the EU as a whole, but divergent national trends can still be identified.

- In Spain, France and to some extent Germany (although a recent increase was reported), Italy and Austria, acute drug-related deaths have stabilized or decreased. This may reflect levels of problem drug use, reduced injecting and/or increases in access to treatment,

- Following few deaths in the early 1990s, Greece, Ireland and Portugal have since reported substantial increases. These may be related to rising heroin use, but also reflect improved recording practices.

- Following significant numbers of drug-related deaths in the early 1990s, increases continue in Sweden, the UK and, to some extent, Denmark. The reasons for this tendency need further investigation.

As we can see by analysing the source, the European Monitoring Centre never uses data aggregated at a supranational level because such aggregation leads to distorted and inaccurate analysis, both due to the different definition of cases that makes it impossible to compare the data from different countries, as is clearly pointed out in the captions of the EMCDDA tables, and because the epidemic situation relative to the use of opiates in the various countries is very different. The extrapolations and the interpretations about the presumed efficacy of demand reduction efforts are clearly false. It is no coincidence that in Sweden itself, where the repressive policy is very close to the UNDCP direction, the trend of deaths in recent years is rising. The extrapolation with an exponential curve has clearly been performed to support the interpretation that the report wishes at all costs to give: anyone who works in the field, in fact, knows that no actual trend can have an exponential curve that rises indefinitely, since "problem drug use" is a saturation epidemic phenomenon (see the bibliography). For the sake of completeness, both the original EMCDDA data (Table 5) and the graphs of the trends of drug-related deaths in several countries (Graph 4), as drawn from the table provided by the EMCDDA (available on-line), are reproduced below. The curves have been normalised by dividing each by the total number of deaths over the entire period observed in order to obtain comparable graphs without altering the qualitative trends.

As is clear, in some countries the trend is oscillating but substantially constant over time (Holland) with alternating upward and downward periods, in others we see the typical saturation epidemic trend with an upward phase followed by a downward phase, more or less marked (France and Germany) while in Sweden we see a substantially linear growth and in Ireland exponential followed by linear growth. Similar divergent trends can be seen in the other countries which are not included in the graph, but whose data are reproduced in Table 5. It is therefore not possible to claim that there is a European trend - there are only very divergent national trends - let alone to make any hypothesis about possible extrapolated trends with interpretations about the presumed efficacy of the European policy, whatever that may be. In conclusion, there are evident distortions of the official EMCDDA data, and an attempt to conceal such distortions by means of captions that suggest the EMCDDA itself is the author of the material. To put it another way, the report contains propaganda and disinformation under the guise of scientific data, behind the screen of attribution to an agency which has always shown itself to be impartial in the publication of information, in order to twist the data in support of pre-established theses that the original data do not confirm. There are at least two other cases of evident distortion of data drawn from sources other than the EMCDDA, which are described in section 4.

Graph 4.
Trend of drug-related deaths in several European countries.

 

 

Table 5.
Drug-related deaths in some EU countries
(source EMCDDA Annual Report 2000, tables at www.EMCDDA.org).

Year	Austria	Denmark	Finland	France	Germany	Greece	Luxembourg	Netherlands	Belgium	Portugal	Spain	Sweden	U.K.
1985		150		172	324	10	1	40	12		143	150	1254
1986		109		185	348	28	3	42	20	18	163	138	1362
1987		140		228	442	56	5	23	17	22	234	141	1332
1988		135	11	236	670	62	4	33	37	33	337	125	1348
1989	20	123	23	318	991	72	8	30	49	52	455	113	1321
1990	36	115	41	350	1491	66	9	43	96	82	455	143	1339
1991	70	188	34	411	2125	79	16	49	90	121	579	147	1411
1992	121	208	27	499	2099	79	17	43	75	156	556	175	1533
1993	130	210	26	454	1738	78	14	38	80	115	442	181	1615
1994	140	271	35	564	1624	146	29	50	46	143	388	205	1796
1995	160	274	51	465	1565	176	20	33	48	198	394	194	1956
1996	179	266	45	393	1712	222	16	63		232	429	250	2150
1997	132	275		228	1501	232	9	70		235	360		2144
1998	108	250		143	1674	244	16	61		337	310
Totals	1096						167	618	570	1744	5245	1962	20561

 

 

4. The main distortions, omissions and "tendentious?" interpretations of data from other sources.

Let us consider Figure 1 on page 96 and Figure 12 on page 115 of the WDR 2000. Both are reproduced in Graph 5.

Graph 5.
Figure 1 on page 96 and Figure 12 on page 115 of the WDR 2000.

 

 

 

Figure 1 is included in the text to support to the following line of argument:

The largest funds for systematic research into understanding the problem of drug abuse and for implementing prevention and treatment programmes, have been made available over the last decade in the USA. Spending on demand reduction (research, prevention and treatment) increased at the federal level from US$ 0.9 billion in 1985 to US$ 5.6 billion in 1999, equivalent to US$ 20 per inhabitant (a very high figure by international standards), or a third of all drug control spending in the country. Even if inflation is taken into account, there was a more than fourfold increase in annual spending for demand reduction over the 1985-1999 period. Parallel to increase spending, drug abuse (annual prevalence as well as current use of all drugs as revealed in the annual household survey) fell by some 40% and cocaine abuse fell by as much as 70% over the 1985-1998/99 period. Though changes in human behavior are usually the result of a multitude of factors, the above example indicates, nonetheless, that a massive increase in demand reduction efforts, based on in depth research of the problem, seems to play an important role in curbing drug abuse.

While Figure 12 is included to support the other line of argument:

Modern media campaigns, such as ONDCP’s current five-year National Youth Anti-Drug Campaign (US$ 200 million), which started in 1997, use more sophisticated and targeted approaches, developed in close consultation with experts from various behavioural sciences, drug prevention, medicine as well as experts from teen marketing and advertising, and representatives from various professional, civic and community based organisations. The approach has been inter alia, based on evidence that social marketing, which involves all parties concerned, segments the audience and tailors messages, usually obtains better results….. The results achieved thus far are impressive. The strong upward trend in drug use among youths (in contrast to the US population as a whole), observed over the 1991-1996 period, was reversed, notably among the younger age groups. The Monitoring the Future studies, which independently collect data on substance abuse among US high school students, showed that annual prevalence of drug use among 8th graders (14 years-old) in the USA fell between 1996 and 1999 by 12% , and was in 1999 a third lower than could have been expected if the prior upward trend had continued (see Figure 12).

In this case, too, we can return to the sources to verify the original data and the interpretation. The complete report on the National Household Survey (SAMHSA) for 1999 can be consulted on-line at www.samhsa.gov/OAS/NHSDA/1999. Naturally the site does not include the graph of aggregate annual prevalence shown in 1, but this can easily be reconstructed from the data given in the various tables containing much more detailed information on use (prevalence) and on the first use (incidence) of the various substances in different segments of the population divided according to sex, age, ethnic origin….. Before considering a number of comments contained in the SAMHSA Report which confute entirely the optimistic and simplistic interpretation of the WDR, it is worth noting that Figure 1, at least from 1992, actually shows an upward trend in prevalence of drug abuse despite the parallel growth in expenditure. The comment in the WDR 2000 omits this aspect, hiding it behind the analysis of the trend in the long term, for which there are, however, various explanations. In actual fact, to obtain a more accurate analysis the entire period should be divided into two sub-periods: the first, from 1981 to 1991, in which the growth in investments is accompanied by a reduction in prevalence, and the second, from 1991 onwards, in which it seems that the growth in investments is accompanied by a growth in prevalence. This shows that the two trends (investments and prevalence) are actually largely independent and can only be interpreted by studying the various phenomena in a much deeper manner…

At this point we might consider some observations in the SAMHSA document which give the lie once and for all to the simplistic, distorted interpretation of the WDR 2000 with respect to Figure 1. We will take just two substances (for an analysis of the others, we can once again consult the above-mentioned site: it is worth mentioning, however, that the other trends observed are not substantially different from those for heroin and cocaine and crack). Let us consider the indicator "incidence of new use" for its efficacy as an instrument for the monitoring of trends, indicated by the texts cited in the bibliography and widely accepted at international level.

From Chapter 2 of the 1999 report (source: SAMHSA) :

Heroin

There was an estimated 149,000 new heroin users in 1998, not statistically different than the 189,000 new users in 1997 or the 132,000 new users in 1996. Estimates of heroin incidence are subject to wide variability and usually do not show any clear trend.

The rate of heroin initiation for the age group 12-17 increased from below 1.0 during the 1980s to nearly 2 during 1996 through 1998.

Cocaine and crack cocaine

The annual number of new users of any form of cocaine rose between 1994 and 1998 from 514,000 to 934,000. However, this was a lower level than during the early and mid 1980s. Recent initiation was at a lower level than it was at its peak in 1983, when the number of new initiates was estimated to be at 1.6 million.

The rates of initiation among different age groups have been increasing in recent years. In particular, the rate among youths age 12-17 increased from 5.1 in 1992 to 13.1 in 1996, remaining level since then. Historically, most initiation of cocaine use has taken place among young adults age 18-25. The rate for that age group fell from a high 30.5 in 1983 and 1984 to 9.1 in 1994. Initiation rates among this age group have increased to 20.8 in 1998.

The number of new crack cocaine users was 371,000 in 1998. While there has been little change in the overall number of new crack users per year since 1985, the age-specific rate of new use for age 12-17 years has increased from 1.4 in 1991 to 4.8 in 1997 and 3.6 in 1998.

The observations in the report cited above already give the lie to the optimistic conclusions regarding the prevention campaigns targeted at young people, mentioned in the comment on Figure 12. However, to get a better idea of where this figure comes from and of how it has been altered, we need to examine the original results of the Monitoring the Future studies, which can be found at: http://monitoringthefuture.org/data/data.html.

On the page containing the final report on these studies, the figure is different from that reproduced in the WDR 2000 as Figure 12, which has been cut and altered. The original, in fact, is the figure reproduced in Graph 6. If we compare the original with the WDR figure we immediately notice that the latter contains only one detail obtained by zooming in on the lower curve of the original and cutting out all the rest. Moreover, the presumed linear trend that is extrapolated does not exist in the original. In other words, the original figure has been taken and the two curves referring to older students have been cancelled, as they are not consistent with the pre-established thesis, while an arbitrary extrapolating curve, unrelated to the original source, has been added, naturally without admitting as much so that the authoritative character of the source corroborates the statement. Any further comment seems unnecessary. In the light of the results of the other American survey reproduced above, however, it is worth analysing briefly the data of the study separately for the various substances. Such analysis shows that the apparently downward trend of the curve, which takes all substances together, is only valid for certain substances and, as we have seen, for younger students.

Graph 6.
Figure 1 from Monitoring the Future 2000.

 

 

These are the data in Table 6: the numbers speak for themselves, and need no further comment.

 

Table 6.
Prevalence in the last year for various substances,
comparison between the 1996 and 2000 data,
approximate data (source: Monitoring the Future).

	1996 8th	2000 8th	1996 10th	2000 10th	1996 12th	2000 12th
Ecstasy	2.2%	3%	4.4%	5.6%	4.4%	8.4%
Crack	1.8%	1.8%	2%	2.2%	2%	2.2%
Cocaine	2%	1.8%	3.8%	3.9%	4%	4.3%
Heroin	1.6%	1.1%	1.2%	1.4%	1%	1.5%
Cannabis	19%	17%	34%	32%	35%	36%

 

 

To bring this brief critical analysis of the World Drug Report 2000 to an end, we can conclude that the volume cannot be considered of any value in terms of information, and even less so in terms of scientific rigour: it is principally a work of propaganda and disinformation under the guise of science; it presents distorted data, covering itself by attributing the data presented to bodies and agencies which enjoy international respect, with the sole imaginable aim of twisting the data in order to support pre-established theses that are not corroborated at all by real epidemiological observations.

Selected Bibliography

1. European Monitoring Centre for Drugs and Drug Addiction. Study on incidence of problem drug use and latency time to treatment in the European Union. Report CT.99.EP.05. Lisbon: EMCDDA, 2000.

2. Hser YI, Anglin MD, Grella C, Longshore D, Prendergast ML. Drug treatment careers. A conceptual framework and existing research findings. J Subst Abuse Treat 1997;14:543-58.

3. Hughes PH, Rieche O. Heroin epidemics revisited. Epidemiol Rev 1995;17:66-73.

4. Hunt LG, Chambers CD. The heroin epidemics. New York, NY: Spectrum Publications Inc., 1976.

5. Ravà L., Calvani M.G., Heisterkamp S., Wiessing L., Rossi C. "Incidence indicators for policy making: models, estimation and implications, UN Bulletin on Narcotics, 2001, in press.

6. Wiessing LG, Hartnoll R, Rossi C. Epidemiology of drug use at macro level: indicators, models and policy-making. UN Bulletin on Narcotics, 2001, in press.

Footnotes

[1] Member of the Board of the European Monitoring Centre on Drugs and Drug Addiction (EMCDDA) and of the Committee of Experts of the Permanent Watchdog on the development of the phenomenon of drugs and drug addiction established within the Department of Social Affairs of the Presidency of the Council of Ministers.

[2] European Monitoring Centre for Drugs and Drug Addiction.

[3] Problem drug use as defined by EMCDDA: drug addiction, notably to opiates and stimulants, injecting drug use, or drug use associated with criminal behaviour.

[4] I was part of the research group that produced the estimates, and am extremely familiar with all the methods used and with the starting data relative to the various countries considered. [5] We might also object that it is not possible, from a methodological point of view, to obtain a mean, but this is not the most serious problem.