Mental Disorders, Substance Use Disorders, and Psychotropic Medication Use Among Sudden-Death Victims
Abstract
Objective:
The authors sought to estimate the prevalence of mental and substance use disorders and psychotropic medication prescriptions among working-age sudden-death victims.
Methods:
Using a written protocol, the authors screened for sudden deaths attended by emergency medical services (EMS) in a large metropolitan county in North Carolina from March 1, 2013, to February 28, 2015. Sudden-death cases were adjudicated by three cardiologists. Mental health and chronic disease diagnoses and treatments were abstracted from EMS, medical examiner, toxicology, and autopsy reports and from clinical records for the past 5 years before death.
Results:
Sudden death was identified for 399 adults ages 18–64 years, 270 of whom had available medical records. Most sudden-death victims were White (63%) and male (65%), had a comorbid condition such as hypertension or respiratory disease, and had a mean±SD age of death of 53.6±8.8 years. Most victims (59%) had at least one mental health or substance use disorder documented in a recent medical record; 76%–78% of victims with a mental disorder had a documented psychotropic medication prescription. However, fewer than one-half (41%) had a documented referral to a mental health professional. The most common diagnostic categories were depressive, anxiety, and alcohol-related disorders. Almost one-half (46%) of the victims had a recent psychotropic prescription, most commonly antidepressants (29%) and benzodiazepines (19%).
Conclusions:
Mental illness, substance use disorders, and psychotropic medication prescriptions were prevalent among sudden-death victims. The health care needs of these individuals may be better addressed by collaborative care for general medical and mental disorders.
HIGHLIGHTS
Most sudden-death victims had a mental disorder or substance-related diagnosis within 5 years of their death.
Psychotropic medications were documented in the medical record of three-quarters of victims with mental disorder diagnoses.
Coordinated primary care and mental health referrals may play an important role in future efforts to reduce sudden-death incidence.
Sudden death is an event that leaves many families overwhelmed and burdened by the abrupt loss of their loved ones. In the United States, the estimated incidence of sudden death is 35.1 per 100,000 persons or approximately 10% of all deaths among working-age adults (1). Review of medical and death records of sudden-death victims may improve our understanding of sudden-death risk factors and guide interventions to help prevent these premature deaths.
Mental illness and substance-related disorders are highly prevalent worldwide and are associated with reduced life expectancy (2). In the United States, the life expectancy of people with serious mental illness is approximately 30%, or 25 years, shorter than that of people in the general population (3). Approximately 50%–90% of people with serious mental illness have at least one chronic medical illness (4), often with an earlier age at onset of medical illness compared with the general population (5). Cardiovascular disease likely contributes to the reduction in life expectancy for those with mental disorders (6). Indeed, young adults with serious mental illness carry a threefold elevated mortality risk associated with coronary heart disease (7). The risk for developing cardiovascular disease, arrhythmia, and cardiac arrest is particularly elevated among those with depression (8, 9), anxiety disorders (10), substance-related disorders (11, 12), bipolar disorder, and schizophrenia (13, 14). Limited information is available on mental illness, substance use, and treatment received prior to death of sudden-death victims. By characterizing this population, we might better understand the role of psychiatric care in sudden death prevention.
The Sudden Unexpected Death in North Carolina (SUDDEN) registry of out-of-hospital deaths in Wake County, North Carolina, provides a unique resource to estimate the prevalence of mental illness, substance use disorders, and mental illness treatments among victims of sudden death (15). Our aims were to estimate the prevalence of depressive, anxiety, bipolar, psychotic, and substance use disorders as well as treatment history in a population-based sample of sudden-death victims and to compare the characteristics of victims with mental illness with those without mental illness.
Methods
Human Protections
This study was reviewed and approved by the institutional review board of the University of North Carolina at Chapel Hill (IRB no. 14–2036).
Study Sample and Inclusion Criteria
SUDDEN is a population-based registry of sudden-death victims ages 18–64 years in Wake County, North Carolina. The current analyses used data collected from March 1, 2013, to February 28, 2015. The study design and methods have been described previously (15, 16). Briefly, reporting software of the Wake County emergency medical services (EMS) patient care was queried to identify EMS-attended out-of-hospital deaths. Deaths in the hospital or emergency department were not included because individuals who died in these facilities might not have shared risk factors and prevention strategies with individuals who had out-of-hospital deaths. Incarcerated individuals and residents of hospice or skilled nursing facilities were also excluded. We sought to study sudden deaths among working-age adults because of the large impact of these types of deaths on years of productivity loss, the heterogeneity of insurance coverage among working-age adults, and the relative simplicity of adjudicating unexpected deaths from medical examiner’s reports and clinical records in younger adults.
For the remainder, postmortem medical records (medical examiner’s external examination, toxicology, and autopsy reports) were obtained from the North Carolina Office of the Chief Medical Examiner. Those with an evidently nonnatural mode of death were excluded. Data on specific natural causes of death were not collected or used in case selection or the analysis. Area hospitals and providers were contacted, with follow-up at specific intervals, to obtain clinical records up to 5 years before death. Three cardiologists unaffiliated with this study adjudicated sudden and unexpected deaths. Deaths were considered sudden if the circumstances before death suggested an abrupt pulseless condition in the absence of any terminal disease (e.g., end-stage cancer or liver disease), drug overdoses, death from trauma, or other nonnatural death. Cases of overdose deaths were excluded after reviews of toxicology reports, medical examiner’s reports, and adjudication. No criteria based on timing since last seen alive or resuscitation efforts were employed, and in the event of disagreement on the cause of death, cases were adjudicated on the basis of the majority opinion among the cardiologists. Overall, of 1,592 EMS-attended, out-of-hospital deaths over a 2-year period in Wake County, 399 cases of sudden death were identified, 270 of which had available medical records.
Data Collection and Management
Demographic data (age, sex, education, race, and marital status) were imported from cross-matched electronic death certificates. Data on smoking status and chronic medical conditions were collected from medical records and medical examiner’s reports as described elsewhere (15). Quality assurance of abstractions included error checking for concordance and accuracy with retaining and reabstraction as outlined previously (16).
Mental health diagnoses.
Two trained research assistants followed a written protocol to abstract mental health data from each case’s medical record and medical examiner’s report. Depressive, anxiety, bipolar, psychotic, substance-related, and alcohol-related disorders were abstracted. For each condition, all available medical records, for example, from emergency department visits, outpatient visits, and hospitalizations, were reviewed for the presence of any psychiatric diagnosis, past or current. If a diagnosis was present in at least one medical record, the case was identified as having received the diagnosis for that mental health condition. No precedence was given to one source of record, and all diagnoses were included in the primary analysis. Details on the specific mental disorder (including ICD-9-CM codes, if available), whether the patient had seen a mental health care provider, whether a referral record had been placed to a specialty service, and the recorded substance (in cases of substance use) were abstracted and entered into the study database. From the medical examiner’s report, we reviewed the medical history and the text summary for all sudden-death victims for any mention of a mental illness.
Psychotropic medications.
We abstracted information on psychotropic and opioid medications from medical examiners’ and medical records for the 2 years prior to death. We reported use of antidepressants, antipsychotics, benzodiazepines, nonbenzodiazepine anxiolytics and hypnotics, nonbenzodiazepine antiepileptics, psychostimulants, anticholinergics, and opioids. Given its use for pain, gabapentin was classified separately and not included in the antiepileptic category.
Statistical Analyses
Descriptive statistics (percentages and 95% confidence intervals or means and standard deviations) were calculated for demographic variables, mental and substance use disorders, and psychotropic utilization with IBM SPSS Statistics version 24 (17) and SAS version 9.4 (18).
Results
Demographic and lifestyle variables for the sudden-death victims with medical records (N=270) are presented in Table 1. Most victims were male, unmarried, and White and had low educational attainment. Their mean±SD age was 53.6±8.8 years. Chronic comorbid conditions, including hypertension (N=194, 72%), diabetes (N=93, 34%), coronary disease (N=77, 29%), and respiratory disease (N=108, 40%), were prevalent among the victims, both among those with and without mental disorders.
Victim category | ||||||||||
---|---|---|---|---|---|---|---|---|---|---|
All (N=270)a | With no mental or substance use disorder (N=110)b | With depressive or anxiety disorder (N=123)c | With psychotic or bipolar disorder (N=37)d | With substance use disorder (N=93)e | ||||||
Characteristic | N | % | N | % | N | % | N | % | N | % |
Age (mean±SD) | 53.6±8.8 | 53.8±8.0 | 53.5±9.2 | 53.7±7.6 | 53.8±8.7 | |||||
Male | 175 | 65 | 83 | 76 | 64 | 52 | 22 | 60 | 61 | 66 |
White | 171 | 63 | 56 | 51 | 93 | 76 | 31 | 84 | 66 | 71 |
Married | 102 | 38 | 50 | 46 | 41 | 33 | 10 | 28 | 25 | 27 |
Education (high school graduate or higher) | 140 | 52 | 63 | 58 | 59 | 48 | 18 | 55 | 44 | 47 |
Body mass index (mean±SD)f | 30.6±8.9 | 31.6±10.2 | 30.4±8.3 | 30.5±8.5 | 28.6±6.7 | |||||
Smoking status (in medical record) | 158 | 59 | 57 | 52 | 81 | 66 | 28 | 76 | 63 | 68 |
Demographic and lifestyle characteristics of sudden-death victims, by psychiatric diagnosis type
More than one-half (59%) of the sudden-death victims had at least one documented mental disorder or substance-related disorder in their medical record (Table 2). Among substance use disorders, alcohol-related disorders were especially common, including alcohol abuse (N=37), alcohol dependence (N=18), alcohol use disorder (N=4), and alcohol withdrawal (N=9). Cannabis-related disorders (N=13), opioid-related disorders (N=14), and stimulant-related disorders (N=16) were also present among victims.
Diagnosis or treatment variable | N | % | 95% CI |
---|---|---|---|
Mental or substance use disorder (ever) | 160 | 59 | 53–65 |
Active diagnosis in the past 2 years | 110 | 41 | 35–47 |
Diagnostic category of disorder | |||
Depressive | 98 | 36 | 31–42 |
Anxiety | 90 | 33 | 28–39 |
Psychotic | 14 | 5 | 3–9 |
Bipolar | 31 | 12 | 8–16 |
Substance use | 93 | 34 | 29–40 |
Documented treatment among cases with mental or substance use disorder diagnosis (ever) | |||
Referred for mental health treatment | 66 | 41 | 34–49 |
Known visit with a mental health care provider | 52 | 33 | 27–40 |
Mental or substance use disorder diagnoses and treatment among 270 sudden-death victims, by psychiatric diagnosis type
Overall, 39% (N=106) of the sudden-death victims had more than one mental disorder or substance use disorder diagnosis. Fewer than one-half (41%) of victims with a mental disorder or substance use disorder diagnosis had a documented referral to a mental health professional, and about one-third (33%) had a documented visit with a mental health provider (Table 2).
Almost one-half (46%) of the sudden-death victims had received prescriptions for psychotropic medications (Table 3). Psychotropic prescriptions were more common among victims with mental disorder diagnoses (76%–78%) than among those without mental disorder diagnoses (19%). In cases with a diagnosis of bipolar disorder or schizophrenia, antipsychotics were the most commonly recorded psychotropic medication class (46%). Among those with depressive or anxiety disorders, antidepressants (51%) were most common. Second-generation antidepressants and second-generation antipsychotics were used more frequently than older antidepressants and first-generation antipsychotics. Among victims with a substance-related disorder, 60% had a prescription for a psychotropic medication. Use of opioids and gabapentin were reported more often among victims with a mental disorder diagnosis than among those without a diagnosis of mental disorder or substance use disorder.
Victim category | |||||||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
All (N=270) | With no known mental or substance use disorder (N=110) | With depressive or anxiety disorder (N=123) | With psychotic or bipolar disorder (N=37) | With substance use disorder (N=93)a | |||||||||||
Medicationb | N | % | 95% CI | N | % | 95% CI | N | % | 95% CI | N | % | 95% CI | N | % | 95% CI |
Psychotropic | 123 | 46 | 40–52 | 21 | 19 | 12–28 | 94 | 76 | 68–84 | 29 | 78 | 62–90 | 56 | 60 | 50–70 |
Antidepressantc | 79 | 29 | 24–35 | 11 | 10 | 5–17 | 63 | 51 | 42–60 | 16 | 43 | 27–61 | 32 | 34 | 25–45 |
SSRI | 41 | 15 | 11–20 | 8 | 7 | 3–14 | 29 | 24 | 16–32 | 8 | 22 | 10–38 | 14 | 15 | 8–24 |
SNRI | 21 | 8 | 5–12 | 1 | 1 | 0–5 | 18 | 15 | 9–22 | 5 | 14 | 5–29 | 11 | 12 | 6–21 |
Tricyclic antidepressant | 13 | 5 | 3–8 | 2 | 2 | 0–6 | 11 | 9 | 5–15 | 2 | 5 | 1–18 | 5 | 5 | 2–12 |
Otherd | 24 | 9 | 6–13 | 2 | 2 | 0–6 | 21 | 17 | 11–25 | 4 | 11 | 3–25 | 11 | 12 | 6–20 |
Antipsychotic | 35 | 13 | 9–18 | 2 | 2 | 0–6 | 30 | 24 | 17–33 | 17 | 46 | 29–63 | 20 | 22 | 14–31 |
Second generation | 29 | 11 | 7–15 | 1 | 1 | 0–5 | 25 | 20 | 14–29 | 15 | 41 | 25–58 | 18 | 19 | 12–29 |
First generation, othere | 7 | 3 | 1–5 | 1 | 1 | 0–5 | 6 | 5 | 2–10 | 3 | 8 | 2–22 | 3 | 3 | 1–9 |
Benzodiazepinef | 52 | 19 | 15–24 | 6 | 5 | 2–11 | 42 | 34 | 26–43 | 12 | 32 | 18–50 | 28 | 30 | 21–41 |
Nonbenzodiazepine anxiolytic or hypnoticg | 24 | 9 | 6–13 | 5 | 5 | 1–10 | 19 | 15 | 10–23 | 5 | 14 | 5–29 | 10 | 11 | 5–19 |
Antiepileptich | 33 | 12 | 9–17 | 7 | 6 | 3–13 | 25 | 20 | 14–29 | 15 | 41 | 25–58 | 14 | 15 | 8–24 |
Other psychotropici | 9 | 3 | 2–6 | 2 | 2 | 0–6 | 7 | 6 | 2–11 | 3 | 8 | 2–22 | 5 | 5 | 2–12 |
Gabapentin | 31 | 11 | 8–16 | 7 | 6 | 3–13 | 22 | 18 | 12–26 | 8 | 22 | 10–38 | 15 | 16 | 9–25 |
Opioid | 55 | 20 | 16–26 | 12 | 11 | 6–18 | 39 | 32 | 24–41 | 13 | 35 | 20–53 | 28 | 30 | 21–41 |
Psychotropic medication prescriptions among sudden-death victims, by psychiatric diagnosis type
Discussion
The findings of our study highlight the association between mental disorders and sudden death. More than one-half of the sudden-death victims in the present study had at least one mental disorder or substance use disorder documented in their medical records within 5 years of their death, and more than one-third of the victims had more than one mental or substance use disorder. Among those with mental and substance use disorders, cardiovascular, respiratory, and metabolic conditions were common. Our findings extend the results of previous research on the relationships of depression with sudden cardiac death among women (8) and of schizophrenia with sudden death (14). Our results also expand on the diversity of mental health and substance use disorders in working-age adults with sudden death.
Our study was conducted in a generally healthy population. Wake County has a heterogeneous population and is ranked as one of the healthiest counties in the United States. The prevalence of mental illness and substance use disorder in this county is low. In the 2014 Behavioral Risk Factor Surveillance System survey for Wake County, only 5.9% of the population reported heavy drinking, and 17.2% were “ever told” that they had a depressive disorder (19). During this period, Wake County was ranked first among the 100 counties in North Carolina for health outcomes and healthy behaviors and had a very low drug overdose rate (20). Our findings are important and unique because we identified the prevalence of chronic mental disorders among all sudden-death victims in a diverse, healthy population of an entire U.S. county. We included only medically reported diagnoses, an approach that may have caused underestimation of the prevalence of mental illness and substance use disorder among sudden-death victims because some mental and substance use disorders may have been unreported or undiagnosed. Further, we did not know the prevalence of mental illness and substance use among individuals without medical records, which warrants future investigation.
Our results are clinically important because they suggest that multiple mental and substance use disorders are strongly associated with sudden death in working-age adults. In combination with the high rates of chronic hypertension, diabetes, and respiratory disease, the presence of mental health concerns suggests a potent, but treatable, potential contributor to premature, sudden death. This interaction of mental illness, substance abuse, and chronic illness is consistent with the causes of life-threatening arrhythmias and underscores the need to focus on mental illness in intervention programs, especially in integrated primary care settings, to prevent sudden death.
Disparity in health care access contributes to poor health and increased mortality rates among people with mental illness (21). Although referrals to and treatment by mental health professionals were uncommon, most victims with mental illness had prescriptions for psychotropic medications, as recommended by guidelines (22); yet, many antipsychotics and antidepressants increase the risk for sudden death, mainly because they can increase the QT interval and cause life-threatening arrhythmias (13, 14, 23, 24). We also observed potentially concerning benzodiazepine prescriptions among persons with substance use disorders (25) and concurrent prescriptions of benzodiazepines and opioids in cases with a previous mental disorder diagnosis, which increases the risk for overdose (26). We do not know whether psychotropic medications contributed to the sudden death of victims in our cohort; however, the high use of psychotropic medications highlights the importance of treatment coordination between medical and behavioral specialists.
Our study had some limitations. Although case ascertainment was highly inclusive, with rigorous quality control and adjudication processes, we could study only victims with medical records. However, a comparison of death certificates between victims with and without medical records indicated no important differences in demographic and social characteristics. We did not study sudden-death cases involving children, older adults, incarcerated individuals, residents of nursing homes, or sudden-death victims transported to an emergency department, and therefore our findings cannot be extended to these groups.
We could not confirm the reliability or severity of recorded diagnoses and whether these diagnoses were present when the patient died. Also, treatment and referral information may have been inflated because our cohort represented only victims who were engaged with the health care system. We examined psychotropic medications under the nervous system Anatomical Therapeutic Chemical Classification, and we do not know whether the medications were taken or whether some medications were taken concurrently. Finally, Wake County has a diverse population and was ranked as one of the healthiest counties in the United States, with low rates of mental illness and substance abuse; our findings therefore may not be generalizable to all U.S. regions.
Conclusions
Most working-age adult victims of sudden death had mental disorders, substance use disorders, and chronic medical conditions, and many of these victims had psychotropic medication prescriptions. The observed combination of mental disorders and medical conditions in sudden-death victims implies the need for collaborative care involving mental health and clinical professionals to help prevent sudden death among working-age adults.
1 : Years of life and productivity loss because of adult sudden unexpected death in the United States. Med Care 2019; 57:498–502Crossref, Medline, Google Scholar
2 : Mortality in mental disorders and global disease burden implications: a systematic review and meta-analysis. JAMA Psychiatry 2015; 72:334–341Crossref, Medline, Google Scholar
3 : Morbidity and Mortality in People With Serious Mental Illness. Alexandria, VA, National Association of State Mental Health Program Directors Medical Directors Council, 2006. https://nasmhpd.org/sites/default/files/Mortality%20and%20Morbidity%20Final%20Report%208.18.08_0.pdfGoogle Scholar
4 : Primary care of patients with serious mental illness: your chance to make a difference. J Fam Pract 2008; 57:515–525Medline, Google Scholar
5 : The impact of serious mental illness on health and healthcare. Psychosomatics 2010; 51:458–465Crossref, Medline, Google Scholar
6 : Higher mortality after myocardial infarction in patients with severe mental illness: a nationwide cohort study. J Intern Med 2015; 277:727–736Crossref, Medline, Google Scholar
7 All-cause mortality among people with serious mental illness (SMI), substance use disorders, and depressive disorders in southeast London: a cohort study. BMC Psychiatry 2010; 10:77Crossref, Medline, Google Scholar
8 : Depression and risk of sudden cardiac death and coronary heart disease in women: results from the Nurses’ Health Study. J Am Coll Cardiol 2009; 53:950–958Crossref, Medline, Google Scholar
9 : Depression and risk of sudden cardiac death and arrhythmias: a meta-analysis. Psychosom Med 2017; 79:153–161Crossref, Medline, Google Scholar
10 : An anxious heart: anxiety and the onset of cardiovascular diseases. Prog Cardiovasc Dis 2013; 55:524–537Crossref, Medline, Google Scholar
11 : Global burden of disease attributable to illicit drug use and dependence: findings from the Global Burden of Disease Study 2010. Lancet 2013; 382:1564–1574Crossref, Medline, Google Scholar
12 : Factors affecting the occurrence of out-of-hospital sudden cardiac arrest. Emerg Med Int 2015; 2015:
13 : Cardiac sudden death in psychiatric patients. Can J Psychiatry 2015; 60:203–205Crossref, Medline, Google Scholar
14 : Etiology of cardiovascular disease in patients with schizophrenia: current perspectives. Neuropsychiatr Dis Treat 2015; 11:2493–2503Crossref, Medline, Google Scholar
15 : Sudden unexpected death in North Carolina (SUDDEN): methodology review and screening results. Open Heart 2014; 1:
16 : Estimated incidence and risk factors of sudden unexpected death. Open Heart 2016; 3:
17 IBM SPSS Statistics for Windows: Version 22.0. Armonk, NY, IBM Corp, 2013Google Scholar
18 SAS 9.4 Product Documentation. Cary, NC, SAS Institute, https://support.sas.com/documentation/94. Accessed Oct 1, 2018Google Scholar
19 Calendar Year 2018 Results: 2018 BRFSS Topics for NC Local Health Regions. Raleigh, NC, North Carolina State Center for Health Statistics, 2019. https://schs.dph.ncdhhs.gov/data/brfss/2018/nc/nccr/topics.htm#hsGoogle Scholar
20 County Health Ranking & Roadmaps. Madison, WI, University of Wisconsin Population Health Institute, 2019. https://www.countyhealthrankings.org/app/north-carolina/2020/overviewGoogle Scholar
21 : Inequalities in healthcare provision for people with severe mental illness. J Psychopharmacol 2010; 24(suppl):61–68Crossref, Medline, Google Scholar
22 : Practice guideline for major depressive disorder in adults. Am J Psychiatry 1993; 150(suppl):1–26Link, Google Scholar
23 : Tricyclic antidepressants, QT interval prolongation, and torsade de pointes. Psychosomatics 2004; 45:371–377Crossref, Medline, Google Scholar
24 : Impact of psychotropic drugs on QT interval dispersion in adult patients. Arq Bras Cardiol 2014; 102:465–472Medline, Google Scholar
25 : Benzodiazepines I: upping the care on downers: the evidence of risks, benefits and alternatives. J Clin Med 2018; 7:17Crossref, Medline, Google Scholar
26 : FDA Warns About Serious Risks and Death When Combining Opioid Pain or Cough Medicines With Benzodiazepines; Requires Its Strongest Warning. Drug Safety Communications 2016. www.fda.gov/drugs/drug-safety-and-availability/fda-drug-safety-communication-fda-warns-about-serious-risks-and-death-when-combining-opioid-pain-orGoogle Scholar