Abstract

Background

Coroners' death certificates form the basis of suicide statistics in England and Wales. Recent increases in coroners' use of narrative verdicts may affect the reliability of local and national suicide rates.

Method

We used Ministry of Justice data on inquests held between 2008 and 2009 and Local Authority suicide data (2001–02 and 2008–09) to investigate variations between coroners in their use of narrative verdicts and the impact of these on suicide rates, using ‘other’ verdicts (79% of which are narratives) as a proxy for narrative verdicts.

Results

There was wide geographic variation in Coroners' use of ‘other’ (mainly narrative) verdicts—they comprised between 0 and 50% (median = 9%) of verdicts given by individual coroners in 2008–09. Coroners who gave more ‘other’ verdicts gave fewer suicide verdicts (r = − 0.41; P< 0.001). In the 10 English Coroners' jurisdictions where the highest proportion of ‘other’ verdicts were given, the incidence of suicide decreased by 16% between 2001–02 and 2008–09, whereas it did not change in areas served by the 10 coroners who used narratives the least.

Conclusions

Variation in Coroners' use of narrative verdicts influences the validity of reported regional suicide rates. Small-area suicide rates, and changes in these rates over time in the last decade, should be interpreted with caution.

Background

Accurate suicide statistics are an important tool for Public Health surveillance and for research into the causes and geographic distribution of suicide.1–5 In England and Wales, the assessment of whether a death is a suicide is determined by a coroner's inquest. Coroners investigate deaths of uncertain or unnatural cause occurring within defined geographic areas and their findings are summarized at a public inquest. There are just over 110 coroners in England and Wales and they conduct ∼30 000 inquests every year (http://www.justice.gov.uk/publications/statistics-and-data/coroners-and-burials/deaths.htm). After the inquest, the coroner gives a verdict, most often taking the form of one of a number of ‘short-form’ categories. These include natural causes, suicide, accident/misadventure, ‘open’ (undetermined cause) or homicide.

In recent years there has been a growing trend for coroners to record ‘narrative’ verdicts. In 2001, 111 inquests (0.3% of all inquests) resulted in narrative verdicts; this had risen to 3012 (10% of inquests) by 2009.4 A narrative verdict gives the circumstances of the death, often in several sentences, without attributing the death to a single short-form category such as ‘suicide’.4 Coroners verdicts of suicide are based on legal guidance surrounding the weight of evidence that a person intended to take their life beyond reasonable doubt.6 Many cases judged to be probable suicides by clinicians receive ‘open’ verdicts as the available evidence is not definitive and for this reason official suicide statistics combine ‘open’ and ‘suicide’ verdicts.7 National suicide statistics are produced by the Office for National Statistics (ONS) based on death certificates they receive from coroners. Recent research indicates that narrative verdicts may be difficult to code4; likely suicides are often coded as accidental deaths when ONS follow the World Health Organization's International Classification of Disease (ICD) procedures in assigning underlying cause of death (http://www.who.int/classifications/icd/ICD-10_2nd_ed_volume2.pdf).

Based on a recent analysis, ONS estimated that if narrative verdicts categorized as accidental poisoning or accidental hangings were in fact suicides (as many are likely to be), then suicide rates in 2009 could be underestimated by 7 deaths per million per year, a 6% underestimate.4 The growing use of narrative verdicts may therefore compromise the analysis of secular trends in national suicide rates.8 Furthermore, as coroners investigate deaths within defined geographic areas, variations in Coroners' practice may distort area differences in suicide incidence, as differences may arise from variations in Coroners' use of narrative verdicts rather than true differences in suicide.

The temporal and spatial epidemiology of suicide is dynamic, with incidence changing in different age bands and different areas over time.5 It is therefore important that suicides are consistently recorded regionally and nationally to enable accurate documentation of any changes in incidence and facilitate effective planning and evaluation of health-care services and policy. In the current study, we have investigated small-area variations in the use of narrative verdicts in Coroners' inquests in England and Wales and have assessed their impact on apparent local trends in suicide.

Methods

Data on the number and type of Coroners' verdicts for the 115 jurisdictions in England and Wales for 2008–09 were obtained from annual statistics published by the Ministry of Justice (www.justice.gov.uk/publications/coronersannual.htm). Data for the Isles of Scilly and the Queens Household jurisdictions were excluded due to the very small numbers of deaths dealt with by the coroners for these jurisdictions.

To investigate geographic variation in the use of narrative verdicts, we used a proxy measure: the number of verdicts classified as ‘other’ verdicts by the Ministry of Justice (MoJ). ONS data suggest that of all ‘other’ verdicts in 2008 and 2009, 78 and 79%, respectively, were narrative verdicts.4 We were unable to obtain data on the specific number of narrative verdicts by coroner jurisdiction since the MoJ only receive data on the breakdown of ‘other’ verdicts from around three quarters of coroners and these data are of varying quality, making the estimation of narrative verdicts at this level difficult (personal communication, 2 June 2011: MoJ).

We used simple descriptive statistics to summarize variation in the use of ‘other verdicts’ from coroner to coroner. Characteristics of the top and bottom 10 coroner jurisdictions ranked on their proportionate use of ‘other verdicts’ were described. We investigated whether coroners who gave a high proportion of narrative verdicts tended to give a low proportion of suicide and open verdicts and vice versa. Previous research has shown that a high proportion of open verdict deaths are probable suicides7 and for this reason national suicide statistics combine deaths given ‘suicide’ and ‘open’ short-form verdicts by coroners.1 We used Spearman's rank correlation coefficients to investigate associations between use of ‘other’ verdicts and deaths certified as due to suicide, ‘open’, natural causes, accidental or due to industrial disease—the five most frequently used verdicts.

To investigate the possible impact of the use of narrative verdicts on apparent trends in suicides we compared changes in suicide rates between 2001–02 and 2008–09 in the English local authorities (LA) served by the 10 coroners who gave the highest proportion of ‘other’ verdicts with the 10 LAs whose coroners gave the lowest proportion of such verdicts. We obtained the suicide data (ICD-10 codes X60–X84 [suicide—‘Intentional self harm’) and Y10–Y34 (open verdicts—‘Event of undetermined intent’)] for the LAs served by each coroner from the National Compendium of Clinical and Health Indicators (http://www.nchod.nhs.uk), as this source only includes data for England, this element of our analysis was restricted to English LAs.

We combined suicide data for 2 year periods to reduce the impact of small-area variations in suicide rates based on small numbers of events. As many coroners served more than one LA, we present rates for all LAs within relevant Coroners' jurisdiction and pooled LA rates for the top and bottom 10 Coroners' jurisdictions. Weighted mean rates of suicide for the 10 coroners making the most frequent use of narrative verdicts and those using them least often were calculated by summing the total number of suicides in each LA divided by the sum of the populations of these LAs.

Results

In the 2 years (2008–09), there were 58 777 verdicts delivered in coroners inquests, 29 781 in 2009 and 28 996 in 2008. The mean annual numbers of verdicts given per coroner was 262 (range: 17–979). Of these, the most common verdict given (figures for 2009) was ‘accident or misadventure’ (n= 8673, 29%), followed by ‘natural causes’ (n= 8281, 28%); there were 3797 (13%) ‘other’ (mainly narrative) verdicts, 3330 (12%) suicide verdicts and 2240 (8%) ‘open’ verdicts. The remaining main verdict classes were industrial disease (n= 2623), dependence on drugs (n= 316), non-dependent use of drugs (n= 250), killed unlawfully (n= 222) and neglect/self neglect (36).

Table 1 shows the number (%) of verdicts given by the coroners with the 10 highest and 10 lowest proportionate use of ‘other’ verdicts. There was considerable variation between jurisdictions in the proportion of inquest outcomes recorded as ‘other’ (mainly narrative) verdicts, these ranged from 0 to 50% (median = 9%) of all inquests. In Birmingham and Solihull an ‘other’ verdict was recorded in 985 (50.3%) of the 1958 inquests in 2008 and 2009; all jurisdictions ranked in the top 10 on the basis of their use of ‘other’ verdicts recorded such verdicts in >20% of inquests. In contrast in the jurisdictions with the lowest proportionate use of ‘other’ verdicts, they accounted for <2% of all verdicts. The proportions of suicide verdicts were higher in the bottom 10 jurisdictions, with a mean proportion of 15.2% compared with 9.3% in the top 10. Of note, the jurisdictions with the greatest proportion of ‘other’ verdicts seem to also deal with a larger number of inquests overall, with an average of 669 inquests in 2008 and 2009 compared with an overall mean of 450 across England and Wales. Conversely, jurisdictions with a lower proportion of ‘other’ verdicts had a smaller average number of inquests over the same period (mean = 218 per year).

Table 1

Top and bottom 10 coroners jurisdictions in England and Wales ranked on their percentage of ‘other’ verdicts based on total 2008 and 2009 data

Coroner jurisdiction Populationa Total number of inquests 2008–2009 Number (%) of other verdicts Number (%) of suicide verdicts Number (%) of open verdicts 
Top 10 coroners ranked according to percentage of ‘other’ verdicts 
 Birmingham and Solihull 1 233 900 1958 985 (50.3) 36 (1.8) 60 (3.1) 
 Cardiff and Vale of Glamorgan 460 800 830 258 (31.1) 34 (4.1) 69 (8.3) 
 Stoke-On-Trent and North Staffordshire 458 900 820 237 (28.9) 50 (6.1) 36 (4.4) 
 North Lincolnshire and Grimsby 318 100 290 82 (28.3) 42 (14.5) 13 (4.5) 
 Telford and Wrekin 162 300 147 38 (25.9) 20 (13.6) 10 (6.8) 
 South and East Cumbria 226 500 329 85 (25.8) 23 (7.0) 22 (6.7) 
 Blackburn, Hyndburn and Ribble Valley 278 700 641 145 (22.6) 51 (8.0) 14 (2.2) 
 Wolverhampton 238 500 232 50 (21.6) 17 (7.3) 29 (12.5) 
 Suffolk 714 100 555 116 (20.9) 94 (16.9) 78 (14.1) 
 Preston and West Lancashire 708 700 883 184 (20.8) 117 (13.3) 39 (4.4) 
Bottom 10 coroners ranked according to percentage of ‘other’ verdicts 
 Central Hampshire 347 500 388 5 (1.3) 59 (15.2) 36 (9.3) 
 York City 198 800 197 3 (1.5) 34 (17.3) 12 (6.1) 
 North and West Cumbria 268 700 222 3 (1.4) 38 (17.1) 25 (11.3) 
 Blackpool/Fylde 216 300 235 3 (1.3) 43 (18.3) 15 (6.4) 
 North West Kent 338 200 416 3 (0.7) 59 (14.2) 39 (9.4) 
 Isle of Wight 140 200 156 2 (1.3) 15 (9.6) 27 (17.3) 
 Ceredigion 76 400 76 1 (1.3) 8 (10.5) 4 (5.3) 
 Western Dorset 160 800 156 1 (0.6) 21 (13.5) 39 (25.0) 
 Carmarthenshire 180 800 175 0 (0.0) 33 (18.9) 15 (8.6) 
 Pembrokeshire 117 400 158 0 (0.0) 27 (17.1) 4 (2.5) 
Coroner jurisdiction Populationa Total number of inquests 2008–2009 Number (%) of other verdicts Number (%) of suicide verdicts Number (%) of open verdicts 
Top 10 coroners ranked according to percentage of ‘other’ verdicts 
 Birmingham and Solihull 1 233 900 1958 985 (50.3) 36 (1.8) 60 (3.1) 
 Cardiff and Vale of Glamorgan 460 800 830 258 (31.1) 34 (4.1) 69 (8.3) 
 Stoke-On-Trent and North Staffordshire 458 900 820 237 (28.9) 50 (6.1) 36 (4.4) 
 North Lincolnshire and Grimsby 318 100 290 82 (28.3) 42 (14.5) 13 (4.5) 
 Telford and Wrekin 162 300 147 38 (25.9) 20 (13.6) 10 (6.8) 
 South and East Cumbria 226 500 329 85 (25.8) 23 (7.0) 22 (6.7) 
 Blackburn, Hyndburn and Ribble Valley 278 700 641 145 (22.6) 51 (8.0) 14 (2.2) 
 Wolverhampton 238 500 232 50 (21.6) 17 (7.3) 29 (12.5) 
 Suffolk 714 100 555 116 (20.9) 94 (16.9) 78 (14.1) 
 Preston and West Lancashire 708 700 883 184 (20.8) 117 (13.3) 39 (4.4) 
Bottom 10 coroners ranked according to percentage of ‘other’ verdicts 
 Central Hampshire 347 500 388 5 (1.3) 59 (15.2) 36 (9.3) 
 York City 198 800 197 3 (1.5) 34 (17.3) 12 (6.1) 
 North and West Cumbria 268 700 222 3 (1.4) 38 (17.1) 25 (11.3) 
 Blackpool/Fylde 216 300 235 3 (1.3) 43 (18.3) 15 (6.4) 
 North West Kent 338 200 416 3 (0.7) 59 (14.2) 39 (9.4) 
 Isle of Wight 140 200 156 2 (1.3) 15 (9.6) 27 (17.3) 
 Ceredigion 76 400 76 1 (1.3) 8 (10.5) 4 (5.3) 
 Western Dorset 160 800 156 1 (0.6) 21 (13.5) 39 (25.0) 
 Carmarthenshire 180 800 175 0 (0.0) 33 (18.9) 15 (8.6) 
 Pembrokeshire 117 400 158 0 (0.0) 27 (17.1) 4 (2.5) 

aPopulation data are based on mid-year estimates in 2009 for LAs/districts included in coroner jurisdictions published by ONS.

Analysis based on 113 jurisdictions confirmed that the use of other (mainly narrative) verdicts was strongly inversely related to the recording of suicide verdicts (r = − 0.41; P< 0.001), although there was only a weak association with the proportion of open verdicts (r = − 0.16; P= 0.09; Table 2). There was no association between the proportion of natural death verdicts given by a coroner and their use of ‘other’ verdict categories (r = 0.01), although the proportion of accidental death verdicts was also inversely associated with ‘other’ verdicts (r = − 0.50; P< 0.001). Surprisingly, there was no association between the proportion of suicide and open verdicts given by a coroner (r = 0.02; P= 0.804).

Table 2

Spearman's ranked correlation coefficients for the associations between the percentage of ‘other’ (mainly narrative) verdicts and the percentage of suicide, open, natural death, accidental and industrial disease verdicts in England and Wales, 2008–09 (n= 113)

Verdicts 2008–09
 
 R P 
Other and suicide −0.41 <0.001 
Other and open −0.16 0.090 
Other and natural death 0.01 0.904 
Other and accidental death −0.50 <0.001 
Other and industrial disease −0.03 0.715 
Verdicts 2008–09
 
 R P 
Other and suicide −0.41 <0.001 
Other and open −0.16 0.090 
Other and natural death 0.01 0.904 
Other and accidental death −0.50 <0.001 
Other and industrial disease −0.03 0.715 

The weighted average change in the rate of suicide between 2001–02 and 2008–09 was a reduction of 16% (95% CI, −27 to −5%) in the 30 LA areas served by the 10 English Coroners' jurisdictions with the highest proportion of ‘other’ (mainly narrative) verdicts (Table 3), whereas in 30 LA areas served by the 10 English coroners giving the lowest proportion of ‘other’ verdicts, the rate did not change (0%; 95% CI, −15 to 16) (Table 4). There was weak statistical evidence that these two changes in rates differed: difference in rates 1.51 (95% CI −0.14 to +3.36) per 100 000 (P = 0.09). These findings were based on English data and therefore we were unable to include the four Welsh LAs, highlighted in Table 1, in this analysis.

Table 3

Coroner jurisdictions with the highest proportion of ‘other’ verdicts in England in 2008–09 and their corresponding LA directly standardized rate of suicidea

Jurisdiction LA Population Average rate: 2001–02 Average rate: 2008–09 Difference (%) 
Birmingham and Solihull Birmingham MCD 1 028 700 10.11 7.08 −3.03 (−30) 
Solihull MCD 205 200 6.14 4.48 −1.66 (−27) 
Stoke-on-Trent and North Staffordshire Stoke-on-Trent UA 239 300 13.26 6.83 −6.44 (−49) 
Staffordshire Moorlands CD 95 400 9.53 8.61 −0.93 (−10) 
Newcastle-under-Lyme CD 124 200 10.01 5.59 −4.42 (−44) 
North Lincolnshire And Grimsby North East Lincolnshire UA 157 100 10.64 8.48 −2.16 (−20) 
North Lincolnshire UA 161 000 8.33 8.73 0.40 (5) 
Telford and Wrekin Telford and Wrekin UA 162 300 10.38 7.82 −2.57 (−25) 
South and East Cumbria Barrow-in-Furness CD 70 900 11.97 7.25 −4.73 (−39) 
Eden CD 51 800 2.26 6.31 4.05 (180) 
South Lakeland CD 103 800 7.56 7.64 0.08 (1) 
Blackburn, Hyndburn and Ribble Valley Blackburn with Darwen 139 900 14.98 9.96 −5.02 (−33) 
Hyndburn 81 100 10.76 8.74 −2.02 (−19) 
Ribble Valley 57 700 7.27 11.87 4.60 (63) 
Wolverhampton Wolverhampton MCD 238 500 10.30 9.85 −0.45 (−4) 
Suffolk Waveney CD 117 700 12.04 10.50 −1.54 (−13) 
Suffolk Coastal CD 124 100 6.58 7.45 0.87 (13) 
Ipswich CD 126 600 9.71 10.06 0.35 (4) 
Babergh CD 85 800 9.75 7.68 −2.07 (−21) 
Mid Suffolk CD 94 200 8.33 10.23 1.90 (23) 
St Edmundsbury CD 103 500 7.60 9.18 1.58 (21) 
Forest Heath CD 62 200 8.56 13.41 4.85 (57) 
Preston and West Lancashire Chorley 104 800 8.57 14.24 5.67 (66) 
Lancaster 139 800 11.33 6.18 −5.15 (−45) 
Preston 134 600 11.85 13.52 1.67 (14) 
South Ribble 108 200 8.38 6.65 −1.73 (−21) 
West Lancashire 110 200 3.25 9.15 5.90 (182) 
Wyre 111 100 9.20 11.43 2.23 (24) 
East Riding and Hull Kingston upon Hull, City of UA 261 100 11.19 10.82 −0.38 (−3) 
East Riding of Yorkshire UA 336 100 10.48 5.62 −4.86 (−46) 
  Weighted mean Weighted mean Difference (%) 
Total   9.82 8.27 −1.55 (−16) 
Jurisdiction LA Population Average rate: 2001–02 Average rate: 2008–09 Difference (%) 
Birmingham and Solihull Birmingham MCD 1 028 700 10.11 7.08 −3.03 (−30) 
Solihull MCD 205 200 6.14 4.48 −1.66 (−27) 
Stoke-on-Trent and North Staffordshire Stoke-on-Trent UA 239 300 13.26 6.83 −6.44 (−49) 
Staffordshire Moorlands CD 95 400 9.53 8.61 −0.93 (−10) 
Newcastle-under-Lyme CD 124 200 10.01 5.59 −4.42 (−44) 
North Lincolnshire And Grimsby North East Lincolnshire UA 157 100 10.64 8.48 −2.16 (−20) 
North Lincolnshire UA 161 000 8.33 8.73 0.40 (5) 
Telford and Wrekin Telford and Wrekin UA 162 300 10.38 7.82 −2.57 (−25) 
South and East Cumbria Barrow-in-Furness CD 70 900 11.97 7.25 −4.73 (−39) 
Eden CD 51 800 2.26 6.31 4.05 (180) 
South Lakeland CD 103 800 7.56 7.64 0.08 (1) 
Blackburn, Hyndburn and Ribble Valley Blackburn with Darwen 139 900 14.98 9.96 −5.02 (−33) 
Hyndburn 81 100 10.76 8.74 −2.02 (−19) 
Ribble Valley 57 700 7.27 11.87 4.60 (63) 
Wolverhampton Wolverhampton MCD 238 500 10.30 9.85 −0.45 (−4) 
Suffolk Waveney CD 117 700 12.04 10.50 −1.54 (−13) 
Suffolk Coastal CD 124 100 6.58 7.45 0.87 (13) 
Ipswich CD 126 600 9.71 10.06 0.35 (4) 
Babergh CD 85 800 9.75 7.68 −2.07 (−21) 
Mid Suffolk CD 94 200 8.33 10.23 1.90 (23) 
St Edmundsbury CD 103 500 7.60 9.18 1.58 (21) 
Forest Heath CD 62 200 8.56 13.41 4.85 (57) 
Preston and West Lancashire Chorley 104 800 8.57 14.24 5.67 (66) 
Lancaster 139 800 11.33 6.18 −5.15 (−45) 
Preston 134 600 11.85 13.52 1.67 (14) 
South Ribble 108 200 8.38 6.65 −1.73 (−21) 
West Lancashire 110 200 3.25 9.15 5.90 (182) 
Wyre 111 100 9.20 11.43 2.23 (24) 
East Riding and Hull Kingston upon Hull, City of UA 261 100 11.19 10.82 −0.38 (−3) 
East Riding of Yorkshire UA 336 100 10.48 5.62 −4.86 (−46) 
  Weighted mean Weighted mean Difference (%) 
Total   9.82 8.27 −1.55 (−16) 

aThese data are based on the National Compendium of Clinical and Health Indicators and do not include data on Welsh LAs.

Table 4

Coroner jurisdictions with the lowest proportion of ‘other’ verdicts in England in 2008–09 and their corresponding LA directly standardized rate of suicidea

Jurisdiction LA Population Average rate: 2001–02 Average rate 2008–09 Difference (%) 
Western Dorset West Dorset CD 96 500 9.18 11.77 2.60 (28) 
North Dorset CD 64 300 4.97 11.96 7.00 (141) 
North West Kent Sevenoaks CD 113 200 5.53 4.30 −1.24 (−22) 
Tunbridge Wells CD 107 600 9.69 10.04 0.35 (4) 
Tonbridge and Malling CD 117 400 4.44 8.52 4.08 (92) 
Blackpool/Fylde Blackpool UA 140 000 20.99 12.34 −8.65 (−41) 
Fylde CD 76 300 6.67 6.25 −0.42 (−6) 
Isle of Wight Isle of Wight UA 140 200 13.16 14.68 1.52 (12) 
Central Hampshire Winchester CD 113 300 7.78 11.29 3.51 (45) 
Test Valley CD 113 400 8.28 9.36 1.09 (13) 
Eastleigh CD 120 800 9.41 7.20 −2.21 (−23) 
North and West Cumbria Allerdale CD 94 300 8.90 13.40 4.50 (51) 
Carlisle CD 104 700 12.38 8.79 −3.59 (−29) 
Copeland CD 69 700 12.81 9.01 −3.80 (−30) 
York City York UA 198 800 5.82 8.55 2.73 (47) 
Teesside Redcar and Cleveland UA 137 800 10.65 10.00 −0.65 (−6) 
Middlesbrough UA 140 100 12.51 7.96 −4.55 (−36) 
Stockton-on-Tees UA 189 800 7.98 8.76 0.78 (10) 
Spilsby and Louthb East Lindsey CD 140 800 7.54 15.16 7.62 (101) 
Essex and Thurrock Thurrock UA 157 200 7.96 4.11 −3.85 (−48) 
Brentwood CD 73 800 5.55 5.96 0.41 (7) 
Basildon CD 174 100 7.38 7.29 −0.09 (−1) 
Epping Forest CD 124 000 9.53 4.75 −4.78 (−50) 
Chelmsford CD 167 800 4.20 7.75 3.55 (84) 
Maldon CD 62 900 5.74 6.85 1.11 (19) 
Uttlesford CD 75 600 6.11 8.56 2.45 (40) 
Braintree CD 142 700 7.32 5.49 −1.84 (−25) 
Colchester CD 177 100 6.39 7.84 1.45 (23) 
Tendring CD 148 000 8.31 4.49 −3.82 (−46) 
Harlow CD 80 600 8.36 3.90 −4.46 (−53) 
   Weighted mean Weighted mean Difference (%) 
Total   8.56 8.52 −0.04 (0) 
Jurisdiction LA Population Average rate: 2001–02 Average rate 2008–09 Difference (%) 
Western Dorset West Dorset CD 96 500 9.18 11.77 2.60 (28) 
North Dorset CD 64 300 4.97 11.96 7.00 (141) 
North West Kent Sevenoaks CD 113 200 5.53 4.30 −1.24 (−22) 
Tunbridge Wells CD 107 600 9.69 10.04 0.35 (4) 
Tonbridge and Malling CD 117 400 4.44 8.52 4.08 (92) 
Blackpool/Fylde Blackpool UA 140 000 20.99 12.34 −8.65 (−41) 
Fylde CD 76 300 6.67 6.25 −0.42 (−6) 
Isle of Wight Isle of Wight UA 140 200 13.16 14.68 1.52 (12) 
Central Hampshire Winchester CD 113 300 7.78 11.29 3.51 (45) 
Test Valley CD 113 400 8.28 9.36 1.09 (13) 
Eastleigh CD 120 800 9.41 7.20 −2.21 (−23) 
North and West Cumbria Allerdale CD 94 300 8.90 13.40 4.50 (51) 
Carlisle CD 104 700 12.38 8.79 −3.59 (−29) 
Copeland CD 69 700 12.81 9.01 −3.80 (−30) 
York City York UA 198 800 5.82 8.55 2.73 (47) 
Teesside Redcar and Cleveland UA 137 800 10.65 10.00 −0.65 (−6) 
Middlesbrough UA 140 100 12.51 7.96 −4.55 (−36) 
Stockton-on-Tees UA 189 800 7.98 8.76 0.78 (10) 
Spilsby and Louthb East Lindsey CD 140 800 7.54 15.16 7.62 (101) 
Essex and Thurrock Thurrock UA 157 200 7.96 4.11 −3.85 (−48) 
Brentwood CD 73 800 5.55 5.96 0.41 (7) 
Basildon CD 174 100 7.38 7.29 −0.09 (−1) 
Epping Forest CD 124 000 9.53 4.75 −4.78 (−50) 
Chelmsford CD 167 800 4.20 7.75 3.55 (84) 
Maldon CD 62 900 5.74 6.85 1.11 (19) 
Uttlesford CD 75 600 6.11 8.56 2.45 (40) 
Braintree CD 142 700 7.32 5.49 −1.84 (−25) 
Colchester CD 177 100 6.39 7.84 1.45 (23) 
Tendring CD 148 000 8.31 4.49 −3.82 (−46) 
Harlow CD 80 600 8.36 3.90 −4.46 (−53) 
   Weighted mean Weighted mean Difference (%) 
Total   8.56 8.52 −0.04 (0) 

aThese data are based on the National Compendium of Clinical and Health indicators and do not include data on Welsh LAs.

bSuicide data were unavailable for the parishes from West Lindsey that are included in this jurisdiction (population ∼16 000).

The differences seen in the weighted average rate of suicide between the two sets of coroners were mirrored when looking at the incident rate ratios (IRRs). The overall IRR was 0.93 (0.87–1.00) for 2008–09 vs. 2001–02 in the LAs served by coroners with the highest proportion of ‘other’ (mainly narrative) verdicts, while in the jurisdictions with low proportionate use of ‘other’ verdicts, the IRR was 1.01 (0.93–1.09).

Discussion

Main findings of this study

The use of ‘other’ (mainly narrative) verdicts varies markedly across different parts of England and Wales. Coroners who record a high proportion of narrative verdicts record fewer suicide verdicts. This variation has affected trends in the incidence of officially recorded suicide between 2001–02 and 2008–09 in the most affected areas of England. Areas served by coroners favouring the use of narrative verdicts experienced apparent declines in suicide, whereas those served by coroners who rarely used other verdicts experienced little change in suicide rates. These results indicate that, in some areas, frequent use of narrative verdicts may lead to the underestimation of suicide and recent increases in use of this verdict may have resulted in artefactual reductions in the incidence of suicide.

Our findings also suggest that the proportion of narrative verdicts given by a coroner is inversely associated with their use of ‘accidental death’ verdicts, indicating that narratives are used as an alternative to accident/misadventure verdicts. However, this is unlikely to lead to an underestimate of accidental deaths as ONS-coding difficulties mainly relate to the identification of suicide and proof of intent. By default, where cause/intent of death is unclear in a narrative verdict, ONS will code it as an accident.

What is already known on this topic

Studies in other countries have documented inaccuracies in suicide statistics9; and so recording practices require regular review. A recent report indicated that Norwegian suicide rates may be underestimated through increased categorization of suicides as accidental poisoning.10 Likewise an analysis of Taiwanese mortality data reported likely misreporting of suicides by hanging and pesticide self-poisoning as accidents.11 Studies from Australia12 and Malaysia13 suggest that apparent declines in national suicide rates can be caused by coding changes, although the impact of miscoding on Australian trends is contested.14 In the UK previous studies have shown how a coroner's practice (in terms of the balance of suicide and open verdicts they give) can lead to misinterpretation of suicide rates if these are based on deaths given suicide verdicts alone.15 Furthermore, a recent cross-national analysis indicates that official suicide rates may be influenced by declining autopsy rates.16

What this study adds

The National Suicide Prevention Strategy for England outlined a minimum target of reducing the incidence of suicide by 20% by 2010. To ensure progress towards this goal is accurately monitored, especially at a local level, national suicide statistics need to be robust. The findings of this study indicate that differences between LAs in the incidence of suicide, and trends in its incidence, may be caused by variation in Coroners' use of narrative verdicts as well as real differences in the incidence of suicide in an area. This study highlights the need for caution when interpreting local area data on the incidence of suicide, for instance when evaluating the success of public health interventions or planning the commissioning of services.

The situation is continuing to deteriorate. The most recent MoJ data (http://www.justice.gov.uk/publications/statistics-and-data/coroners-and-burials/deaths.htm) indicate that there was a further rise in ‘other’ (mainly narrative) verdicts in 2010 (n= 4180; 14% of all verdicts in 2010 and a 10% rise compared with figures for 2009).

Limitations

The primary limitation of the current analysis is that we used ‘other’ verdicts as a proxy for narrative verdicts. The recent ONS analysis4 indicates that this is a reasonable assumption, with over three quarters of ‘other’ verdicts in 2008–09 being narratives. However, it remains possible that there are regional variations in the proportion of narratives amongst ‘other’ verdicts. The MoJ does not receive consistent nor reliable data from individual coroners on their other verdicts which prevents the accurate reporting of narrative verdict use in each jurisdiction.

Conclusions

In addition to the growth in narrative verdicts leading to possible under-estimation of overall national suicide rates,4,8 geographic variation in their use is contributing to inaccurate reporting of the incidence of suicide and trends in suicide rates in different localities, depending on the practice of the local coroner. These local variations will not only result in misleading evaluations of the success or failure of local suicide prevention activities but also mislead local health policy-makers concerning the relative importance of suicide in their locality. Action is warranted to ensure coroners who record a narrative verdict also include a likely cause of death in a short-form verdict, in the meantime we recommend that the Office of National Statistics issues a note of caution concerning the validity of small-area suicide rates, perhaps carrying out correction, for example based on assumptions that all deaths by hanging and overdose given narrative verdicts are suicides. A speedy resolution to the current problems would have been helped by national oversight to ensure consistency in Coroners' practice. However, the current Government's proposed abolition of the post of Chief Coroner (created by the 2009 Coroners and Justice Act) is likely to delay the development and dissemination of consistent practice across the country.

Funding

We acknowledge financial support from the National Institute for Health Research (NIHR) Programme Grants for Applied Research scheme as part of an NIHR programme of research related to suicide prevention (RP-PG-0606-1247). K.H. and D.G. are NIHR Senior Investigators. K.H. is also supported by Oxford Health NHS Foundation Trust and N.K. by the Manchester Mental Health and Social Care Trust.

Acknowledgements

The views and opinions expressed in this paper are those of the authors and do not necessarily reflect those of the NIHR, NHS or the Department of Health. The NIHR and Department of Health had no role in study design, the collection, analysis and interpretation of data, the writing of the report and the decision to submit the paper for publication.

References

1
Brock
A
Griffiths
C
Trends in suicide by method in England and Wales
Health Stat Q
 , 
2003
, vol. 
20
 (pg. 
7
-
18
)
2
Charlton
J
Kelly
S
Dunnell
K
, et al.  . 
Trends in suicide deaths in England and Wales
Popul Trends
 , 
1992
, vol. 
69
 (pg. 
10
-
6
)
3
Department of Health
National Suicide Prevention Strategy for England
 , 
2002
 
London: Department of Health
4
Hill
C
Cook
L
Narrative verdicts and their impact on mortality statistics in England and Wales
Health Stat Q
 , 
2011
, vol. 
49
 (pg. 
81
-
100
)
5
Gunnell
D
Wheeler
B
Chang
SS
, et al.  . 
Changes in the geography of suicide in young men: England and Wales 1981–2005
J Epidemiol Community Health
 , 
2010
6
Matthews
P
Jervis on the Office and Duties of Coroners
 , 
2002
12th edn
London
Sweet & Maxwell Ltd
7
Linsley
KR
Schapira
K
Kelly
TP
Open verdict v. suicide—importance to research
. Br J Psychiatry
 , 
2001
, vol. 
178
 (pg. 
465
-
8
)
8
Gunnell
D
Hawton
K
Kapur
N
Coroners' verdicts and suicide statistics in England and Wales
BMJ
 , 
2011
 
doi=10.1136/bmj.d6030
9
Goldney
RD
A note on the reliability and validity of suicide statistics
Psychiat Psychol Law
 , 
2010
, vol. 
17
 (pg. 
52
-
6
)
10
Gjertsen
F
Johansson
LA
Changes in statistical methods affected the validity of official suicide rates
J Clin Epidemiol
 , 
2011
11
Chang
SS
Sterne
JAC
Lu
TH
, et al.  . 
‘Hidden’ suicides amongst deaths certified as undetermined intent, accident by pesticide poisoning and accident by suffocation in Taiwan
Soc Psychiatry Psychiatr Epidemiol
 , 
2010
, vol. 
45
 (pg. 
143
-
52
)
12
De Leo
D
Dudley
M
Aebersold
C
, et al.  . 
Achieving standardised reporting of suicide in Australia: rationale and program for change
MJA
 , 
2010
, vol. 
192
 (pg. 
452
-
6
)
13
Maniam
T
Suicide and undetermined violent deaths in Malaysia, 1966–1990: evidence for missclassification of suicide statistics
Asia-Pacific J Public Health
 , 
1995
, vol. 
8
 (pg. 
181
-
5
)
14
Page
A
Taylor
R
Martin
G
Recent declines in Australian male suicide are real, not artefactual
Aust Nz J Psychiat
 , 
2010
, vol. 
44
 (pg. 
358
-
63
)
15
Nicholson
S
Suicide in North Devon: epidemic or problem of classification
Health Trends
 , 
1992
, vol. 
24
 (pg. 
95
-
6
)
16
Kapusta
ND
Tran
US
Rockett
IR
, et al.  . 
Declining autopsy rates and suicide misclassification: a cross-national analysis of 35 countries
Arch Gen Psychiatry
 , 
2011