A paucity of corroborative psychological and psychiatric evidence may be inhibiting detection of drug intoxication suicides in the United States. We evaluated the relative importance of suicide notes and psychiatric history in the classification of suicide by drug intoxication versus firearm (gunshot wound) plus hanging/suffocation—the other two major, but overtly violent methods.
This observational multilevel (individual/county), multivariable study employed a generalized linear mixed model (GLMM) to analyze pooled suicides and undetermined intent deaths, as possible suicides, among the population aged 15 years and older in the 17 states participating in the National Violent Death Reporting System throughout 2011–2013. The outcome measure was relative odds of suicide versus undetermined classification, adjusted for demographics, precipitating circumstances, and investigation characteristics.
A suicide note, prior suicide attempt, or affective disorder was documented in less than one-third of suicides and one-quarter of undetermined deaths. The prevalence gaps were larger among drug intoxication cases than gunshot/hanging cases. The latter were more likely than intoxication cases to be classified as suicide versus undetermined manner of death (adjusted odds ratio [OR], 41.14; 95% CI, 34.43–49.15), as were cases documenting a suicide note (OR, 33.90; 95% CI, 26.11–44.05), prior suicide attempt (OR, 2.42; 95% CI, 2.11–2.77), or depression (OR, 1.61; 95% CI, 1.38 to 1.88), or bipolar disorder (OR, 1.41; 95% CI, 1.10–1.81). Stratification by mechanism/cause intensified the association between a note and suicide classification for intoxication cases (OR, 45.43; 95% CI, 31.06–66.58). Prior suicide attempt (OR, 2.64; 95% CI, 2.19–3.18) and depression (OR, 1.48; 95% CI, 1.17–1.87) were associated with suicide classification in intoxication but not gunshot/hanging cases.
Without psychological/psychiatric evidence contributing to manner of death classification, suicide by drug intoxication in the US is likely profoundly under-reported. Findings harbor adverse implications for surveillance, etiologic understanding, and prevention of suicides and drug deaths.
Citation: Rockett IRH, Caine ED, Connery HS, D’Onofrio G, Gunnell DJ, Miller TR, et al. (2018) Discerning suicide in drug intoxication deaths: Paucity and primacy of suicide notes and psychiatric history. PLoS ONE 13(1): e0190200. https://doi.org/10.1371/journal.pone.0190200
Editor: Keith M. Harris, University of Queensland, AUSTRALIA
Received: September 19, 2017; Accepted: November 14, 2017; Published: January 10, 2018
Copyright: © 2018 Rockett et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: The third-party data underlying this study are from the Restricted Access Database (RAD) and are available from the National Violent Death Reporting System (NVDRS). Persons meeting criteria established by the Centers for Disease Control and Prevention (CDC) can request data from the restricted access database through the following URL: https://www.cdc.gov/violenceprevention/nvdrs/rad.html. The authors did not have any special privileges in obtaining this data set.
Funding: This research was supported by Grants R49CE002109 and R49CE002093 from the National Center for Injury Prevention and Control of the Centers for Disease Control and Prevention. They provided salary support for three authors (IRHR, EDC, HJ) and consulting fees for another (SS) (https://www.cdc.gov/injury/index.html). The funder had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.
Competing interests: The authors have declared that no competing interests exist.
Following a 19% decline between 1986 and the turn of the century, the United States (US) suicide rate rose from 10.40 per 100,000 population in 2000 to 13.75 in 2015—a 32% increase . Although large, this rise may be a serious underestimate due to increasing misclassification of intoxication suicides , a trend detected in other countries [3,4]. The US drug intoxication mortality rate for persons 15 years and older has risen by 257%—from 7.81 deaths per 100,000 in 2000 to 20.07 in 2015 . This increase primarily implicates the component involving pharmaceutical opioids, heroin, and other illicit opioids [5,6]. Distinguishing manner of death showed that 67% of drug fatalities in the 15 and older population in 2000 were classified as accident, followed by suicide, undetermined intent (hereafter “undetermined”), and homicide at 19%, 14%, and 0.1% . By 2015, accidents constituted 84% of these deaths versus 10% for suicide, 6% for undetermined, and 0.1% for homicide. At 322%, the 15-year rise in the accident component of the drug intoxication death rate was 9-fold higher than the corresponding rise in the drug intoxication suicide rate. Viewed in the context of under-resourced and overburdened emergency healthcare [7,8] and medicolegal death investigation systems [9,10], these differentials and discrepancies buttress a scenario of growing suicide misclassification. Compounding the situation, medical examiners and coroners need substantial affirmative evidence to assign suicide or homicide as the manner of death, whereas information can be scant for supporting an accident option, their major default .
Suicide is a highly stigmatized and perpetually undercounted phenomenon in the US . Firearm (gunshot wound), hanging/suffocation, and drug intoxication are the leading methods, accounting for 88.3% of suicides in 2015 , with respective components of 49.8%, 26.8%, and 11.7%. However, strong corroborative psychological and psychiatric evidence, in the form of a suicide note and psychiatric history, may be much more important to ascertainment of suicide by drug intoxication than by the other two overtly violent methods [13–15].
A meta-analysis of 27 psychological autopsy studies  and a systematic review of 22 case-control studies and 54 case series  indicated that up to 90% of suicides in the US and other Western countries had a diagnosable psychiatric disorder at time of death. An American multiple-cause-of-death study showed that, whereas diagnosed psychiatric disorders distinguished suicides from deaths whose manner was classified as accident, they were documented in less than 10% of the death certificates for the suicides . A subsequent American multiple-cause-of-death study of pooled suicides and undetermined deaths found that cases without psychiatric documentation manifested a 260% excess likelihood of an undetermined classification . An English coroner study reported a large prevalence gap in suicide notes , as potentially pivotal evidence [21,22], between suicide and undetermined cases—50% versus 11%. Undetermined is universally regarded as the manner of death category most susceptible to suicide misclassification [23–27], and drug intoxication and other poisoning is the predominant injury cause/mechanism represented among undetermined cases in the US [19, 28]. In the United Kingdom, suicide rates are officially presented as a composite of rates for registered suicides and undetermined deaths , an accommodation that increases the suicide rate by approximately 20% .
In this multilevel (individual/county), multivariable study of the association between psychological/psychiatric documentation and suicide versus undetermined manner of death classification, based on microdata from the US National Violent Death Reporting System (NVDRS), we first calculated the prevalence of a documented affective disorder, suicide note, and past suicide attempt among suicide and undetermined cases. We then evaluated associations between these variables and suicide versus undetermined classification, and whether emergent associations were more paramount among drug intoxication cases than gunshot and hanging/suffocation cases. Our incorporation of undetermined deaths, as well as registered suicides, not only provided a window on the nature of suicide misclassification within the undetermined manner of death category , but within the accident category—as a much larger reservoir for obscuring drug intoxication suicides [31,32].
Materials and methods
National Violent Death Reporting System (NVDRS) Restricted Access Database
The principal data source for this study was the Restricted Access Database in the NVDRS, a state, territory, and incident-based surveillance system that employs public health informatics for making data linkages to produce detailed, individual-level information about suicide and other violent deaths . Administered by the Centers for Disease Control and Prevention (CDC), the NVDRS primarily comprises data from death certificates, law enforcement records, and medical examiner and coroner records. This system variably includes such optional supplementary data as crime laboratory reports and hospital records. Its Restricted Access Database contains de-identified information that includes geographic location, circumstances, and personal sociodemographic characteristics. In this study, the microdata pertained to the 17 states, disaggregatable to county of death, which participated in the NVDRS throughout our observation period, 2011–2013. These states were Alaska, Colorado, Georgia, Kentucky, Maryland, Massachusetts, New Jersey, New Mexico, North Carolina, Ohio, Oklahoma, Oregon, Rhode Island, South Carolina, Utah, Virginia, and Wisconsin. Our study population comprised deaths from intentional self-harm, that is, registered suicides (ICD-10: U03, X60-X84, Y87.0), and undetermined deaths (ICD-10 Y10-Y34 and Y87.2, Y89.9) whose state and county of death were known. It was further limited to decedents aged 15 years and older, since less than 1% of known suicides nationally were younger . Decedents in the study population totaled 40,581.
Table 1 provides comparative population and mortality data for the 17 NVDRS states in this study and for the entire US in 2012, the mid-year in our observation period. With a population approaching one-third of that of the US, these states closely resembled the nation in their age and sex composition, manner of death distribution, and crude and age-adjusted all-cause, suicide, and undetermined mortality rates. This demographic concordance supports the generalizability of our results, although NVDRS states overrepresented non-Hispanic Whites and Blacks and underrepresented Hispanics. Natural causes or diseases accounted for the overwhelming preponderance of deaths in both the NVDRS states and nationally. However, the disproportionate toll that suicide and other injury exact in premature mortality and truncated life expectancy [34–36] is masked in the equating of individual deaths in most conventional summary mortality measures.
Individual-level variables from NVDRS
Manner of death (suicide versus undetermined) was the outcome variable. Individual-level predictors were overtness of selected injury mechanisms/causes (gunshot and hanging/ suffocation versus drug intoxication); suicide note (yes versus no or unknown); prior suicide attempt (yes versus no or unknown); and primary mental diagnosis (depression or bipolar disorder versus none or unknown). Other individual-level covariates were autopsy status; current mental health treatment; crisis in past two weeks; number of intimate partner or legal or job or financial or school problems; blood alcohol concentration (grams per deciliter); number of other specified drug positives; physical health problem; age; sex; race/ethnicity; marital status; and education.
County-level covariates and data sources
Linked to the microdata, there were two county-level medicolegal covariates: selection mode (elected or appointed) of the chief medical examiner or coroner and accreditation status/type of medicolegal death investigation system (accredited medical examiner, accredited coroner, unaccredited medical examiner, unaccredited coroner). The data source for selection mode was a CDC website , and those for the investigation systems were the respective websites of the two relevant accrediting agencies, the National Association of Medical Examiners  and the International Association of Coroners & Medical Examiners , and email or telephone communication with state, district, or county offices to resolve outstanding questions. A third county-level covariate was urbanicity, a surrogate for external forces that may support or inhibit medicolegal death investigations . Operationalized across five categories (large metropolitan, small metropolitan, adjacent metropolitan, micropolitan or adjacent, rural) representing the 12 ordinal categories of the 2013 Urban Influence Codes, its data source was the County Area Health Resource File for 2014–2015 .
Statistical analysis and hypotheses
Pursuant to profiling suicides and undetermined cases, we used a generalized linear mixed model (GLMM) to test the respective hypotheses that: among the pooled cases, the designation of a “suicide” manner of death would be more likely when (1) the injury mechanism/cause was gunshot or hanging/suffocation (as more overtly violent methods of suicide), versus drug intoxication; and there was documentation of (2) a suicide note, (3) a prior suicide attempt, or (4) a primary diagnosis of depression or bipolar disorder. Then we tested our final hypothesis that any observed associations, pertaining to hypotheses 2 through 4, were stronger among drug intoxication than gunshot/hanging cases. GLMM is a two-level model, which was logistic at the individual level and linear at the county level. We included a state-level random effect to incorporate the data structure of counties nested in a state and individuals nested in a county. The statistical software was SAS for Windows, version 9.4 (Cary, NC: SAS Institute Inc., 2002–2010).
There were 36,190 suicides and 4,391 undetermined deaths in the 17 NVDRS states between 2011 and 2013 (Table 2). Thirty-one percent of the suicides were accompanied by a suicide note, compared with 1.5% for the undetermined cases. Prevalence of documented suicide attempts was 18% among suicides and 1.5% among the undetermineds, with corresponding prevalences of comorbid depression and bipolar disorder of 28%, 23%, 4%, and 5%. Gunshot/hangings constituted 68% of the suicides, compared with 11% for the drug intoxication cases, whereas the latter constituted 55% of the undetermined cases versus 4% of the gunshot/hangings.
The multilevel, multivariable analysis showed that gunshot/hanging deaths were 41 times more likely than drug intoxication deaths to be classified by medical examiners and coroners as suicides (Table 3)—affirming the hypothesized association regarding overtness of injury mechanism/cause. Odds also were higher for the residual specified group, whose injury mechanisms included gassing, other poisoning, cutting, jumping (falling), and drowning (immersion). Consistent with our second hypothesis, cases with a note were 34 times more likely than cases with no note or unknown note status to be classified as suicide. Also affirmed were our third and fourth hypotheses, which concerned evidence of a prior suicide attempt or a primary diagnosis of depression or bipolar disorder, respectively. Decedents who had a prior suicide attempt were 2.4 times more likely than decedents without such history to be classified as suicide, and those with documented unipolar depression or bipolar disorder were respectively 61% and 41% more likely than their referent to be so classified.
Stratification by mechanism/cause
Focus on a comparison of drug intoxication and gunshot/hanging suicides revealed marked differentials in their respective prevalences of a documented suicide note, prior suicide attempt, and a primary diagnosis of depression and bipolar disorder (Table 4). Prevalence gaps for a suicide note, prior suicide attempt, and affective disorders between suicides and undetermined deaths were larger among drug intoxication cases than gunshot/hanging cases. Multilevel, multivariable analysis affirmed our fifth and final hypothesis that observed associations between psychological and psychiatric characteristics and a suicide versus undetermined classification were stronger among the drug intoxication cases (Table 5). Relative to their referent, cases with a note were associated with 45-fold increased odds of a suicide classification for drug deaths compared with 8-fold increased odds for gunshot/hanging deaths. Documentation of a prior suicide attempt and comorbid depression showed increased odds of suicide classification of 164% and 48%, respectively, for the drug cases, but no associations for gunshot/hanging cases.
Even the enriched microdata from the NVDRS showed less than one-third of suicide cases included a record of a comorbid affective disorder, prior suicide attempt, or authenticated suicide note, whether written, typed, digital or audio. Signaling complications for suicide accounting and prevention, our multilevel, multivariable analysis leads us to infer that full and accurate accounting of suicide by opioid and other drug intoxication in the US is far more dependent upon corroborative psychological and psychiatric evidence than is that of suicide by the other two major but overtly violent methods, firearm and hanging/suffocation. We hypothesize that absence of such evidence increases the likelihood that drug intoxication suicides are misclassified in the accident manner of death category, as well as in undetermined. Against the background of highly stressed emergency healthcare [7,8] and medicolegal death investigation systems [9,10], our findings both reinforce and inform concern and skepticism about the uniformity and validity of suicide and default manner of death classifications in the US [32,42–44].
During the late 1980s, CDC convened an expert multidisciplinary panel to develop operational criteria for suicide determination . This panel recommended that injury deaths be classified as suicide if the injury was consistent with self-infliction and there was indication of intent. They also recommended medical examiners and coroners utilize multiple data sources in their investigations, including sources that would facilitate identification of the psychological and psychiatric characteristics of suicides and possible suicides. In cases where intent was equivocal, medical examiners and coroners would need to consult psychiatrists or clinical psychologists proficient in assessing self-harm proximal to death in order to optimize suicide ascertainment . However, there continues to be no such auxiliary workforce in the US to support suicide investigations, unlike police in homicide investigations, whose training and service focus on ruling homicide in or out in death scene and other medicolegal death investigations . Moreover, medical examiners and coroners rarely employ any formal psychological inquiry in helping them resolve intentionality or manner of death in ambiguous, uncertain, and equivocal cases, epitomized by most drug poisoning deaths [11,47]. Such procedures, in some settings called a “psychological autopsy,” integrate interviews with key informants—family, friends, and acquaintances of decedents—and intensive records review and examination [17,48].
Heterogeneity of forensic investigation systems, within and across states, may impede accurate classification of suicide and other manners of death in the US [31,49]. With no national oversight, administration of death investigation systems varies from state control under an appointed state medical examiner to county control under an elected coroner or a medical examiner . Contrasting with medical examiners, coroners typically lack training in medicine, pathology or forensic pathology. Training in psychiatry or clinical psychology is not a requisite for either position. Compounding threats to the quality of injury-related manner of death data, most coroner offices were not professionally accredited by either the International Association of Coroners & Medical Examiners or the National Association of Medical Examiners, as of January 2015 .
In 2015, CDC hosted an expert meeting to commence addressing the inherent challenges fatal opioid and other drug intoxication cases pose for medical examiner and coroners in accurately assigning suicide and other manners of death . Our research will fortify the empirical underpinnings of this important and ongoing initiative.
Strengths and limitations
The first multilevel investigation of the association between psychological/psychiatric documentation and suicide manner of death classification in the US, and in the world to the best of our knowledge, this study was able to factor in heterogeneity across precipitating circumstances of death, medicolegal death investigation systems and practices, and demographics. Only the microdata from the NVDRS Restricted Access Database were appropriate for evaluating our research questions. Essential for valid geographic, sociodemographic, and temporal accounting and comparisons, the NVDRS emphasizes uniform protocols for defining manner of death and achieving consistent data collection, entry, review, and coding practices . However, the affirmation of hypothesized variation between suicides and undetermined deaths is likely conservative as a consequence. Tempering this caveat on study generalizability is the countervailing demographic concordance we reported between the 17 NVDRS states and the nation. We note that since our observation period, 2011–2013, the NVDRS has expanded to 40 states, the District of Columbia, and Puerto Rico . The goal is to cover all 50 states and territories.
NVDRS data do not include staffing and budgetary information on medicolegal offices, and data collection is not yet uniform and standardized for mental health, substance use, and other circumstances of death. Toxicology is nearly complete in drug and other poisoning cases, but fiscal constraints and associated data deficits make coverage problematic for surveillance and research in non-poisoning cases . The “yes” provision means affirmative for circumstances of the suicide and undetermined deaths, but “no” entries are not distinguished from unknown or missing information—a conservative bias. The NVDRS neither documents comorbid diseases, such as cancer and amyotrophic lateral sclerosis (ALS), which are risk factors for suicide , nor differentiates prescription opioids from illicit analogs. Data sparseness in our multilevel study precluded more refined distinctions among injury mechanisms/causes and separation of centralized and decentralized medicolegal death investigation systems.
Without psychological/psychiatric evidence contributing to manner of death classification, suicide by drug intoxication in the US is likely profoundly under-reported. Findings harbor adverse implications for surveillance, etiologic understanding, and prevention of suicides and drug deaths.
- 1. Centers for Disease Control and Prevention, 2017. Web-based Injury Statistics Query and Reporting System. Available from: https://www.cdc.gov/injury/wisqars/
- 2. Rockett IRH, Hobbs G, De Leo D, Stack S, Frost JL, Ducatman AM, et al. Suicide and unintentional poisoning mortality trends in the United States, 1987–2006: two unrelated phenomena? BMC Public Health 2010; Nov 17;10:705. pmid:21083903
- 3. Gunnell D, Bennewith O, Simkin S, Cooper J, Klineberg E, Rodway C, et al. Time trends in coroners' use of different verdicts for possible suicides and their impact on officially reported incidence of suicide in England: 1990–2005. Psychol Med 2013; 43: 1415–1422. pmid:23113986
- 4. Chan CH, Caine ED, Chang SS, Lee WJ, Cha ES, Yip ES. The impact of improving suicide death classification in South Korea: a comparison with Japan and Hong Kong. PLoS One 2015; May 20;10(5):e0125730. pmid:25992879
- 5. Rudd RA, Aleshire N, Zibbell JE, Gladden RM. Increases in drug and opioid overdose deaths—United States, 2000–2014. MMWR Morb Mortal Wkly Rep 2016; 64 (50–51): 1378–1382. pmid:26720857
- 6. Hedegaard H, Warner M, Minino AM. Drug overdose deaths in the United States, 1999–2015. NCHS Data Brief 2017; 273: 1–8.
- 7. Institute of Medicine. Emergency medical services at the crossroads. Washington, DC: The National Academies Press; 2006.
- 8. Institute of Medicine. Hospital-based emergency care: at the breaking point. Washington, DC: The National Academies Press; 2006.
- 9. Institute of Medicine; Board on Health Promotion and Disease Prevention; Committee for the Workshop on the Medicolegal Death Investigation System. Medicolegal death investigation system: workshop summary. Washington, DC: The National Academies Press; 2003.
- 10. Committee on Identifying the Needs of the Forensic Science Community; Committee on Science, Technology, and Law Policy and Global Affairs; Committee on Applied and Theoretical Statistics; Division on Engineering and Physical Sciences. Strengthening forensic science in the United States: a path forward. Washington, DC: The National Academies Press; 2009.
- 11. Rockett IRH, Smith GS, Caine ED, Kapusta ND, Hanzlick RL, Larkin GL, et al. Confronting death from drug self-intoxication (DDSI): prevention through a better definition. Am J Public Health 2014; Dec;104(12):e49–55. pmid:25320874
- 12. Goldsmith SK, Pellmar TC, Kleinman AM, Bunney WEE. Reducing suicide: a national imperative. Washington, DC: The National Academies Press; 2002.
- 13. Ovenstone IMK. A psychiatric approach to the diagnosis of suicide and its effect upon the Edinburgh statistics. Br J Psychiatry 1973; 123(572): 15–21. pmid:4729869
- 14. Carr JR, Hoge CW, Gardner J, Potter R. Suicide surveillance in the U.S. Military—reporting and classification biases in rate calculations. Suicide Life Threat Behav 2004 Autumn; 34(3): 233–241. pmid:15385178
- 15. Bohnert AS, McCarthy JF, Ignacio RV, Ilgen MA, Eisenberg A, Blow FC. Misclassification of suicide deaths: examining the psychiatric history of overdose decedents. Inj Prev 2013 Oct;19(5):326–330. pmid:23322257
- 16. Arsenault-Lapierre G, Kim C, Turecki G. Psychiatric diagnoses in 3275 suicides: a meta-analysis. BMC Psychiatry 2004; 4: 37. pmid:15527502
- 17. Cavanagh JT, Carson AJ, Sharpe M, Lawrie SM. Psychological autopsy studies of suicide: a systematic review. Psychol Med 2003; 33(3):395–405. Review. Erratum in: Psychol Med 2003; 33(5): 947. pmid:12701661
- 18. Rockett IRH, Wang S, Lian Y, Stack S. Suicide-associated comorbidity among US males and females: a multiple cause-of-death. Inj Prev 2007; 13: 311–315. pmid:17916887
- 19. Rockett IRH, Wang S, Stack S, De Leo D, Frost JL, Ducatman AM, et al. Race/ethnicity and potential suicide misclassification: window on a minority suicide paradox? BMC Psychiatry 2010; May 19;10:35. pmid:20482844
- 20. Linsley KR, Schapira K, Kelly TP. Open verdict v. suicide—importance to research. Br J Psychiatry 2001; 178: 465–468. pmid:11331564
- 21. Stone DH, Chishti P, Roulston C. Final report of the European review of suicide and violence epidemiology (EUROSAVE) Project, 2002. Available from: http://ec.europa.eu/health/ph_projects/1999/injury/fp_injury_1999_frep_10_en.pdf
- 22. Parai JL, Kreiger N, Tomlinson G, Adlaf EM. The validity of the certification of manner of death by Ontario coroners. Ann Epidemiol 2006; 16(11): 805–811. pmid:16621598
- 23. Ohberg A, Lonnqvist J. Suicides hidden among undetermined deaths. Acta Psychiatr Scand 1998; 98(3): 214–218. pmid:9761408
- 24. Värnik P, Sisask M, Värnik A, Arensman E, Van Audenhove C, van der Feltz-Cornelis CM et al. Validity of suicide statistics in Europe in relation to undetermined deaths: developing the 2–20 benchmark. Inj Prev 2012; 18: 321–325. pmid:22157205
- 25. Salmerón D, Cirera L, Ballesta M, Navarro-Mateu F. Time trends and geographical variations in mortality due to suicide and causes of undetermined intent in Spain, 1991–2008. J Public Health (Oxf) 2013; 35(2): 237–245.
- 26. Kõlves K, Milner A, Värnik P. Suicide rates and socioeconomic factors in Eastern European countries after the collapse of the Soviet Union: trends between 1990 and 2008. Sociol Health Illn 2013; 35(6): 956–970. pmid:23398609
- 27. Pritchard C, Hansen L. Examining undetermined and accidental deaths as source of 'under-reported-suicide' by age and sex in twenty Western countries. Community Ment Health J 2015; 51(3): 365–376. pmid:25536938
- 28. Huguet N, Kaplan MS, McFarland BH. Rates and correlates of undetermined deaths among African Americans: results from the National Violent Death Reporting System. Suicide Life Threat Behav 2012; 42(2): 185–96. pmid:22486604
- 29. UK Department of Health, 2014. Statistical update on suicide. Available from: https://www.gov.uk/government/uploads/system/uploads/attachment_data/file/278120/Suicide_update_Jan_2014_FINAL_revised.pdf
- 30. Thomas K, Gunnell D. Suicide in England and Wales 1861–2007: a time-trends analysis. Int J Epidemiol 2010; 39(6): 1464–1475. pmid:20519333
- 31. Warner M, Paulozzi LJ, Nolte KB, Davis GG, Nelson LS. State variation in certifying manner of death and drugs involved in drug-intoxication deaths. Acad Forensic Path 2013; 3: 231–237.
- 32. Rockett IRH, Hobbs GR, Wu D, Jia H, Nolte KB, Smith GS, et al. Variable classification of drug-intoxication suicides across US states: a partial artifact of forensics? PLoS One 2015; 10, e0135296. 2015 Aug 21;10(8):e0135296. Erratum in: PLoS One 2015;10(9):e0137933 pmid:26295155
- 33. Blair JM, Fowler KA, Jack SP, Crosby AE. The National Violent Death Reporting System: overview and future directions. Inj Prev 2016 Apr; Suppl 1: i6–i11.
- 34. Rockett IRH. Injury and violence: a public health perspective. Popul Bull 1998: 53(4): 1–40. pmid:12321948
- 35. Rockett IRH. Injury elimination and survival: a cross-national study. Public Health 1998; 112(1): 27–30. pmid:9490885
- 36. Rockett IRH, Lilly CL, Jia H, Larkin GL, Miller TR, Nelson LS, et al. Self-injury mortality in the United States in the early 21st century: a comparison with proximally ranked diseases. JAMA Psychiatry 2016; 73(10): 1072–1081. pmid:27556270
- 37. Centers for Disease Control and Prevention, 2015. Public Health Law Program. Coroner/Medical Examiner Laws by State. Available from: http://www.cdc.gov/phlp/publications/topic/coroner.html.
- 38. National Association of Medical Examiners, 2017. Available from: https://netforum.avectra.com/eweb/DynamicPage.aspx?Site=NAME&WebCode=OrgSearch.
- 39. International Association of Coroners & Medical Examiners, 2017. Available from: http://www.theiacme.com/accreditation.
- 40. Waldhoer T, Berzlanovich A, Vutuc C, Haidinger G. Rates of postmortem examination in Austria: the effect of distance between location of death and site of examination. J Clin Epidemiol 2003; 56(9): 891–895. pmid:14505775
- 41. Health Resources and Services Administration, Bureau of Health Workforce, National Center for Health Workforce Analysis. The County Area Health Resources File (AHRF) 2014–2015 Release. Rockville, Maryland: US Department of Health and Human Services, Health Resources and Services Administration; 2015.
- 42. Webster LR, Dasgupta N. Obtaining adequate data to determine causes of opioid-related overdose deaths. Pain Medicine 2001; Suppl 2: S86–S92.
- 43. Klugman J, Condran G, Wray M. The role of medicolegal systems in producing geographic variation in suicide rates. Soc Sci Q 2013; 94(2): 462–489.
- 44. Gray D, Coon H, McGlade E, Callor WB, Byrd J, Viskochil J, et al. Comparative analysis of suicide, accidental, and undetermined cause of death classification. Suicide Life Threat Behav 2014; 44(3): 304–316. pmid:25057525
- 45. Rosenberg ML, Davidson LE, Smith JC, Berman AL, Buzbee H, Gantner G, et al. Operational criteria for the determination of suicide. J Forensic Sci 1988; 33(6): 1445–1456. pmid:3204347
- 46. US Public Health Service. National strategy for suicide prevention: goals and objectives for action. Washington, DC: Department of Health and Human Services; 2001.
- 47. Curphey TJ. The forensic pathologist and the multidisciplinary approach to death. In Schneidman ES, editor, Essays in self-destruction. New York: Jason Aronson; 1967. pp. 463–474.
- 48. Botello T, Noguchi T, Sathyavagiswaran L, Weinberger LE, Gross BH. Evolution of the psychological autopsy: fifty years of experience at the Los Angeles County Chief Medical Examiner-Coroner's Office. J Forensic Sci. 2013; 58(4):924–926. pmid:23551031
- 49. Hsieh ML, Neuilly MA. Within and inter-institutional differences between death certifiers on autopsy conclusions. J Interpers Violence 2016; May. pmid:27147273
- 50. Hanzlick R. A perspective on medicolegal death investigation in the United States: 2013. Acad Forensic Pathol. 2014; 4(1): 2–9.
- 51. National Commission on Forensic Science, National Institute of Standards and Technology. Accreditation of medicolegal death investigation offices. January 30, 2015. Available from: http://www.justice.gov/sites/default/files/ncfs/pages/attachments/2015/02/25/mdi_accreditation_final_as_amended_1_30_15.pdf
- 52. Stone DM, Holland KM, Bartholow B, Logan JE, LiKamWa McIntosh W, Trudeau A, et al. Deciphering suicide and other manners of death associated with drug intoxication: a Centers for Disease Control and Prevention consultation meeting summary. Am J Public Health 2017; Aug; 107(8): 1233–1239. pmid:28640689
- 53. Kaplan MS, Caetano R, Giesbrecht N, Huguet N, Kerr WC, McFarland BH, et al. The National Violent Death Reporting System: use of the restricted access database and recommendations for the system’s improvement. Am J Prev Med 2017; Jul; 53(1): 130–133. pmid:28347589
- 54. Harris EC, Barraclough BM. Suicide as an outcome for medical disorders. Medicine (Baltimore) 1994; 73(6): 281–296