Rapid Responses to:

RESEARCH: Michael Bloor, Maria Gannon, Gordon Hay, Graham Jackson, Alastair H Leyland, and Neil McKeganey Contribution of problem drug users’ deaths to excess mortality in Scotland: secondary analysis of cohort study BMJ 2008; 337: a478 [Abstract] [Full text]

Rapid Responses: Submit a response to this article

Rapid Responses published:

Death due to problem drug use: Tip of the iceberg?

Sahoo Saddichha   (24 July 2008)

Stretching credulity

Harold Saxon   (24 July 2008)

Not a third of all excess mortality

David S Gordon   (18 August 2008)

Death due to problem drug use: Tip of the iceberg? 24 July 2008
Sahoo Saddichha, WHO-BGI Senior Consultant Kolkata-700091, India Send response to journal: Re: Death due to problem drug use: Tip of the iceberg?	Bloor and colleagues [1] have eloquently defined the "Scottish effect" as the differences in deaths due to problem drug use in Scotland. One pertinent question however remains: how was problem drug use defined? Is it harmful use or dependence due to drugs as defined by ICD 10 and DSM IV? Although I am not well aware of mortality rates in England and Scotland, I believe that they would indeed be different as Mackenbach et al. [2] have observed in their study. These authors have also observed that socio-economic inequality may be responsible for differing morbidity and mortality rates in European countries. In that case, how "standarized" would it be to extrapolate from Scottish to English death rates? Regarding the DORIS sample, I would also request the authors to shed light on choosing a '33 month' review period, rather than a 36 month (3 year) or 48 month (4 year) period. In fact, the DORIS sample has mainly focussed on opioid users (88% of sample), making the death rates a reflection of opiate related deaths rather than drug related ones. I do agree however that public health initiatives that can be more direct, although lacking popularity, may go a long way in reducing these mortality rates. References: 1)Bloor et al. Contribution of problem drug users’ deaths to excess mortality in Scotland: secondary analysis of cohort study. BMJ 2008; 337: a478 2)Mackenbach et al.Socioeconomic Inequalities in Health in 22 European Countries. N Engl J Med 2008;358:2468-81. Competing interests: None declared
Stretching credulity 24 July 2008
Harold Saxon, Student University of the West of Scotland, PA1 2BE Send response to journal: Re: Stretching credulity	Bloor et al infer that 30% of the difference between mortality in Scotland compared to England can be accounted for by mortality amongst drug users. However the authors stretch credulity when they state that 'readers must decide on the windows of applicability of the different dataset.' they have employed. The authors employ assumption after assumption, and estimate after estimate to arrive at their headline figure. However their methodology raises a number of important questions Perhaps the most specious assumption is that the mortality experience of the DORIS cohort (or any cohort of drug users seeking treatment) can be applied to the total drug using population (the author's prevalence data on problem drug use). I find this to be the weakest link in the author's analysis. In terms of statistical modelling, how did the authors decide which prior distributions to use? Would different priors make substantial difference to their conclusion? The authors regurgitate the loaded but empty language of a 'Scottish Effect'. Hopefully this vocabulary is limited to a few researchers who presumably fail to realise that they could equally have employed the title 'English Effect'. Even if we had confidence in the authors estimate that 30% of the difference in mortality between Scotland and England is due to the effects of drug use it hardly supports the authors' assertion that drug misuse should be 'a prime target' in reducing health inequalities. The authors have only shown that it would reduce the overall difference in mortality between Scotland and England. Given that the overall burden of health inequality results from chronic disease, reduction in drug use should therefore only be viewed as a worthwhile target, not a prime one. Competing interests: None declared
Not a third of all excess mortality 18 August 2008
David S Gordon, Head of Public Health Observatory Division NHS Health Scotland, Elphinstone House, 65 West Regent St, Glasgow G2 2AF Send response to journal: Re: Not a third of all excess mortality	It is disappointing to see both the "what this study adds" box and "This week in numbers" claim that drug use causes 32% of the excess mortality in Scotland over England. As the paper makes clear, drug use accounts for 32% of the excess in the 15-54 age group. The "Scottish effect" is not constrained to those under age 55, and the absolute numbers below 55 will be swamped by those aged 55 and over. It makes a much less dramatic headline, though that does not diminish the importance of reducing the death toll from drugs. Competing interests: None declared