Table HB5

Table HB5: Drug-related deaths by NHS Board area: average for 2009 to 2013, and relative to the estimated number of problem drug users
Area	2009-2013 annual average drug-deaths (all ages)	Problem drug users (aged 15-64) in 2009/10¹				Annual average drug-deaths: 2009-2013 per 1,000 problem drug users in 2009/10⁴
		95% Confidence Interval²				Likely range of values
		Estimate	Lower end	Upper end	+/-³	Estimate	from⁵	to⁵
Footnotes: Estimates of problem drug users aged 15 to 64, as published by the Information Services Division (ISD) of NHS National Services Scotland. Some of the estimates are subject to potentially large percentage margins of error, as indicated by the 95% Confidence Intervals. The 95% Confidence Intervals are the range within which it is expected that the true value will lie. On the basis of statistical theory, there is only a 5% chance that a 95% Confidence Interval will not include the (unknown) true value of the quantity which is being estimated - so, on average, one would expect that 19 out of 20 of all 95% Confidence Intervals will include the (unknown) true values. The average of the percentage differences between (a) the estimate and the lower end of the 95% Confidence Interval and (b) the estimate and the upper end of the 95% Confidence Interval. It is calculated using the rounded values of the estimate and the two ends. These death rates are broad indications only, as (e.g.) the estimated numbers of problem drug users may be subject to wide confidence intervals. The 'from' value in the range for the rate is calculated using the upper end of the 95% Confidence Interval for the estimated number of problem drug users, and the 'to' value in the range for the rate is calculated using the lower end of the 95% Confidence Interval for the estimated number of problem drug users. NB: The numbers of drug-related deaths for each area are based on the Board boundaries that apply with effect from 1st April 2014. The figures that have been used for earlier years are the numbers that would have been seen had the new boundaries applied in those years. However, the estimated numbers of problem drug users are based on the Board boundaries that applied in 2009/10. It follows (refer to paragraph 4.7) that the 'per 1,000 problem drug user' death rates are likely to be under-estimated for Greater Glasgow & Clyde, and over-estimated for Lanarkshire. It is thought that the boundary changes will have had very little effect on the figures for other areas.
Scotland	544	59,600	58,300	61,000	2%	9.1	8.9	9.3
Ayrshire & Arran	39	5,100	4,800	5,300	5%	7.7	7.4	8.2
Borders	7	580	510	670	14%	12.8	11.0	14.5
Dumfries & Galloway	8	1,300	1,200	1,500	12%	6.3	5.5	6.8
Fife	36	3,300	3,100	3,600	8%	10.8	9.9	11.5
Forth Valley	23	2,200	2,000	2,400	9%	10.3	9.4	11.3
Grampian	47	4,900	4,600	5,300	7%	9.6	8.9	10.2
Greater Glasgow & Clyde	172	20,800	20,400	21,900	4%	8.3	7.8	8.4
Highland	21	2,100	1,800	2,300	12%	9.9	9.0	11.6
Lanarkshire	64	5,900	5,300	6,500	10%	10.8	9.8	12.0
Lothian	81	8,200	7,600	8,800	7%	9.9	9.3	10.7
Orkney	1	n-a	n-a	n-a	n-a	n-a	n-a	n-a
Shetland	1	130	90	270	69%	10.8	5.2	15.6
Tayside	43	5,000	4,700	5,300	6%	8.6	8.1	9.1
Western Isles	1	130	90	240	58%	10.8	5.8	15.6