Witnessed and unwitnessed sudden cardiac death: a nationwide study of persons aged 1–35 years

Abstract

Aims

The aim of this study is to compare clinical characteristics and causes of death among witnessed and unwitnessed sudden cardiac death (SCD) cases aged 1–35 years.

Methods and results

In this retrospective nationwide study, all deaths in persons aged 1–35 years in Denmark during 2000–09 were included (23.7 million person-years). Using the in-depth descriptive Danish death certificates and Danish nationwide registries, 860 cases of sudden, unexpected death were identified. Through review of autopsy reports and register data, we identified 635 cases of SCD of which 266 (42%) were witnessed and 326 (51%) were unwitnessed. In 43 cases (7%), witnessed status was missing. Clinical characteristics were overall similar between the two groups. We found a male predominance among unwitnessed SCD compared to witnessed SCD (71% and 62%, respectively, P-value 0.012), as well as more psychiatric comorbidity (20% and 13%, respectively, P-value 0.029). Unwitnessed SCD more often occurred during sleep whereas witnessed SCD more often occurred while the individual was awake and relaxed (P-value < 0.001). The autopsy rate among all SCD cases was 70% with no significant difference in autopsy rate between the two groups. Sudden unexplained death, which was the leading autopsy conclusion in both groups, was more frequent among unwitnessed SCD (P-value 0.001).

Conclusion

Several clinical characteristics and autopsy findings were similar between witnessed and unwitnessed SCD cases. Our data support the inclusion of both witnessed and unwitnessed cases in epidemiological studies of SCD cases aged 1–35 years, although the risk of misclassification is higher among unwitnessed and non-autopsied cases of SCD.

Open in new tabDownload slide

Introduction

Sudden cardiac death (SCD) is one of the most common causes of natural death in both persons of all ages and among the young. Reliable information on the epidemiology of SCD is important to initiate preventive strategies. However, the incidence rate of SCD has been variably reported with SCD incidences ranging between 15–159 and 1–2.7 per 100 000 person years in persons of all ages and among the young, respectively. Important contributors to the variability in reported incidences are differences in applied SCD definition and case ascertainment criteria, methods of estimation and extrapolation, and demographics of the populations studied.^1–4

Most recent studies adhere to the World Health Organization’s (WHO) sudden death (SD) criteria. The specificity of these criteria varies depending on whether the event is witnessed or unwitnessed: In witnessed SD, death occurs within 1 h of an observed acute change in cardiovascular status. In unwitnessed cases, the deceased must have been seen alive and functioning normally within 24 h of being found dead.^1–3,5

However, the time aspect of ‘sudden’ continues to be a subject of debate and the inclusion time from symptom onset till death ranges from 1 to 24 h in the literature. Another crucial question is whether unwitnessed deaths can be categorized as sudden and unexpected. Unwitnessed cases of SCD potentially represent a different subset of patients and the circumstances and cause of death might be too uncertain.

In a recent study, Tseng et al.⁶ report observations from a prospectively designed community-wide autopsy study of victims of apparent SCD in the San Francisco area. The investigators found that 40% of cases attributed to stated cardiac arrest as defined by the WHO were not sudden or unexpected. Additionally, Tseng and colleagues report data suggesting that nearly half of all unwitnessed WHO-defined SCD were non-cardiac compared with 27% among witnessed cases. In a subsequent article, the same investigators suggest that the conventional SCD definition can be improved to better specify sudden arrhythmic death by restricting witnessed SCD to ventricular tachycardia/fibrillation or non-pulseless electrical activity rhythms and unwitnessed cases to <1 h since last observed in usual health condition.⁷

We have previously identified all SCD cases in persons aged 1–35 years in Denmark from 2000 to 2009.^2,3 In this study, we aimed to compare clinical factors and autopsy findings among witnessed and unwitnessed cases of SCD. We hypothesized that unwitnessed cases of SCD could represent different patient demographics with more comorbidity when compared with witnessed SCD cases.

Methods

Study design

We conducted a retrospective study using information from all death certificates, the registration of all in- and outpatient activity in Danish hospitals and emergency rooms together with access to all medical records and autopsy reports.^2,3 All deaths in persons aged 1–35 years in Denmark in 2000–09 were included in the initial analysis to identify SD and subsequently SCD.

The Danish health care system and Danish registries

All Danish residents are assigned a unique personal Civil Registration Number, which can be linked unambiguously to national registries on an individual level. Using the Civil Registration Number, information on the cause of death can be obtained from the National Causes of Death Register, in which immediate, contributory, and underlying causes of death are recorded using International Classification of Diseases (ICD) 10 codes. This register is based on information from death certificates and autopsy reports. Causes of deaths are subsequently evaluated and corrected by the Danish Health and Medicines Authorities in case of obvious mistakes.

Information on prior medical history can be retrieved from the Danish National Patient Register which contains information on all inpatient activities at Danish hospitals and emergency departments since 1977 as well as outpatient contacts since 1995, classified by ICD codes for each visit (ICD-8 for 1978–94 and ICD-10 from 1995 and onwards).

Death certificates and forensic and clinical autopsy

Whenever a person dies in Denmark, a death certificate is issued. The death certificate is always issued by a medical doctor, who based on all available information, including medical files, determines the most likely cause of death. Police involvement is mandatory whenever a person is found dead and/or death is sudden and unexpected. The police decide whether a medicolegal external examination should be performed. The medicolegal examination is carried out by a Medical Doctor of Public Health who has access to the police report comprising interviews with first responder, relatives, and the family doctor, any medical files related to the deceased and the body of the deceased, which is thoroughly externally examined for any signs of illness. Information from all of these sources is printed in a supplementary information field on the death certificate, which makes Danish death certificates highly suitable for identification of sudden and unexpected death.^2,3,8

A forensic autopsy is conducted if the manner of death is not fully elucidated after medicolegal external examination. When indicated, a toxicological examination is performed by the forensic chemical department.⁹ Furthermore, physicians and relatives of the deceased can request a hospital autopsy if it is decided not to perform a forensic autopsy.

Definitions

We defined SD as a sudden, natural, and unexpected death including all various causes of natural death; in witnessed cases as an acute change in cardiovascular status with time to death being <1 h and in unwitnessed cases as a person last seen alive and symptom free <24 h before being found dead. Unexpected death was defined as a death occurring in an individual free from severe chronic diseases or other diseases or lifestyles that can be expected to lead to death.

Sudden cardiac death was defined as SD cases with the cause of death being of presumed cardiac or unknown origin. Autopsied SCD cases were subdivided into two groups:

1. Explained SCD, in which a cardiac cause of death was established (including thoracic aortic dissections).

2. Sudden unexplained death (SUD), where the autopsy revealed no cause of death.

In non-autopsied SCD cases, the cause of death was presumed to be of cardiac origin unless circumstances related to the death or information from the registries, death certificates, or discharge summaries indicated a more likely non-cardiac cause of death.

All cases meeting the criteria of SCD were included. Individuals who were successfully resuscitated after out-of-hospital cardiac arrest but died shortly later in hospital were also included. Cases with aborted SCD were not included.

Statistical analysis

Calculations and data analyses were performed using SAS software package 9.4. Confidence intervals were calculated using Poisson distribution. Differences in categorized nominal data were tested with the Fisher’s exact test. Medians were compared using Wilcoxon rank-sum test. A two-sided P-value <0.05 was considered statistically significant.

Results

During the study period, there was a mean population in Denmark of 5.41 million inhabitants, of whom 2.37 million were in the age group 1–35 years. From 2000 to 2009, there were 8 756 deaths among persons aged 1–35 years (Figure 1).

We identified 860 (10% of all deaths) cases of SD by thoroughly reviewing all death certificates and autopsy reports. Of the 860 SD cases, 230 were not autopsied, 630 were autopsied. Thus, the autopsy rate among sudden and unexpected death was 73%.

After reviewing all autopsy reports, the Danish registries and discharge summaries of those with sudden unexpected death (n = 860), 225 (26%) were found to have died a sudden non-cardiac death. We identified a total of 635 SCD (74%); 431 autopsied and 204 non-autopsied cases. Among the 635 SCD cases, 326 (51%) were unwitnessed and 266 (42%) were witnessed. There was no information on whether death was witnessed in 43 (7%) cases (Figure 1).

Witnessed and unwitnessed sudden cardiac death

We found a male predominance among unwitnessed SCD cases compared with witnessed SCD cases (71% and 62%, respectively, P-value 0.012), as well as more psychiatric comorbidity (20% and 13%, respectively, P-value 0.029). Unwitnessed SCD more often took place at home, while more witnessed SCD died in-hospital (P-value <0.001). Unwitnessed SCD more often occurred during sleep, whereas witnessed SCD more often occurred while the individual was awake and relaxed (P-value < 0.001). The autopsy rate among all SCD cases was 70% with no significant differences in autopsy rates between unwitnessed and witnessed SCD cases (P = 0.148). A comparison of clinical characteristics between witnessed and unwitnessed SCD cases is presented in Table 1 and stratified by autopsy status in Tables 2 and 3.

Autopsy findings

Autopsy findings among cases of SCD and non-cardiovascular SD are provided in Figure 2A and B, respectively.

From the 635 SCD cases, 431 were autopsied; information regarding witnessed status and time aspect was available in 414 cases. Of these, 178 (43%) were witnessed and 236 (57%) unwitnessed. The most frequent autopsy finding in both groups was SUD. However, SUD was significantly more frequent among unwitnessed SCD cases (P-value 0.001). Among witnessed SCD cases, the cause of death was more commonly myocarditis (P-value 0.006) and congenital heart diseases (P-value 0.033) (Table 2).

Those who died suddenly and unexpectedly from a non-cardiovascular disease most often died from pulmonary (36%), infectious (24%), and cerebrovascular disease (15%). Pulmonary disease was a significantly more predominant cause of death among witnessed cases of SD (P-value 0.001), whereas a cerebrovascular cause of death was more common among unwitnessed cases of SD (P-value 0.023).

Discussion

By utilizing autopsy reports, the informative Danish death certificates, discharge summaries, and nationwide registries, we have identified all cases of SCD in persons aged 1–35 years in Denmark during 2000–09 and examined differences in clinical characteristics and autopsy findings between cases of witnessed and unwitnessed SCD.

Witnessed vs. unwitnessed sudden cardiac death

Approximately 50% of all SCD among persons <35 years of age in Denmark were unwitnessed. In contrast, Tseng et al.⁷ found that 77% of all WHO-defined SCD occurred unwitnessed; however, the mean age in the POST SCD study was 63 years (age range 18–90 years). Our study population is younger (mean age 29 years and age range 1–35 years) and probably less prone to social isolation and this could partly explain the observed lower frequency of unwitnessed SCD.

Unwitnessed SCD cases had more psychiatric comorbidity (20% vs. 13%) when compared with witnessed SCD cases. This may be explained by the fact that individuals with a mental disorder have a smaller network size and tend to live more in solitude when compared to the general population.¹⁰ Recent healthcare initiatives focusing on improving bystander resuscitation attempts are likely to benefit individuals with psychiatric disorders to a lesser degree as these patients more often have unwitnessed cardiac arrests due to solitary lifestyles. Furthermore, persons with psychiatric disorders who experience cardiac arrest more often present with non-shockable rhythms, which is potentially related to a larger comorbidity burden and the use of psychotropic drugs among these individuals.^8,11,12

Not surprisingly, we found that more unwitnessed SCD often took place at home and more while sleeping whereas witnessed SCD often occurred in-hospital and often while being awake and relaxed. Previous studies have found that a significant proportion of SCD cases occur in-hospital.^3,13,14 This finding warrants for caution in the interpretation of results from studies that rely solely on out-of-hospital deaths.

Autopsy findings

The autopsy frequency was relatively high among both witnessed and unwitnessed SCD cases and there were no significant differences in the autopsy frequencies between witnessed and unwitnessed cases. In total, 600 SD were autopsied in persons aged 1–35 years from 2000 to 2009 (70% of all SD), of which 414 (69%) were SCD. The leading cause of SCD was SUD. The relative frequency of SUD was significantly higher among unwitnessed SCD cases. An underlying channelopathy such as Brugada syndrome, long QT syndrome, and catecholaminergic polymorphic ventricular tachycardia is often identified among SUD cases and their families, subsequently increasing their risk of arrhythmia and SCD.^15–18 Brugada syndrome and particularly long QT syndrome Type 3 can debut with SD and both tend to have symptom onset at nighttime. Accordingly, this could explain the increased occurrence of SUD cases among unwitnessed SCD.

Inclusion of unwitnessed deaths?

We found similar clinical characteristics among witnessed and unwitnessed SCD cases aged 1–35 years. However, uncertainty of the circumstances of death will always remain among unwitnessed SCD and in particular among non-autopsied cases. In those cases, non-SD and non-cardiac death (e.g. occult overdose) can rarely be excluded. In the POST SCD Study, Tseng et al.⁶ found that of 896 OHCA deaths, 371 (41%) were not sudden or unexpected. The authors identified 525 WHO-defined SCD cases and after autopsy including toxicology and histology 210 (40%) were found to be caused by non-cardiac causes and 71 (14%) were caused by occult overdose. Subsequently, the same authors suggest refining the definition of SCD by restricting witnessed SCD to ventricular tachycardia/fibrillation or non-pulseless electrical activity rhythms and unwitnessed cases to <1 h since last seen normal.⁷ This would likely increase the positive predictive value of sudden arrhythmic death cases. However, narrowing the SCD definition to deaths that occur <1 h since last seen normal will undoubtful exclude many unwitnessed SCD cases at a high cost of the sensitivity. A solution could be to stratify SCD into ‘definite’ and ‘probable’ SCD depending on witnessed status as previously done by Chugh et al.¹⁹ or to use sensitivity analyses excluding unwitnessed cases.

Although the autopsy frequency in the present study was high, 178 (30%) non-autopsied SCD cases were included in the SCD population, based on information from death certificates and registries. Among these cases, the true cause of death is uncertain. However, we have previously found that 67–75% of the autopsied sudden unexpected deaths were categorized as SCD after reviewing the autopsy report. Based on that, we find it likely that most of the non-autopsied sudden unexpected deaths would have been attributed to SCD if they had been autopsied.^2,3 Also, in a recent study, systematic work-up of relatives to non-autopsied possible SCD cases revealed a definite inherited cardiac disease in 13% of the referred families, increasing to a yield of 23% when families with borderline diagnoses were included.²⁰

Accordingly, we advocate using both witnessed and unwitnessed cases to provide most accurate data of the SCD epidemiology. Advantageously, future studies may also use sensitivity analyses excluding unwitnessed cases, as some risk of misclassification remains.

Study strengths and limitations

The main strength of this study is the nationwide approach by which we identified all cases of SD using the unique Danish death certificates, autopsy reports, as well as discharge summaries and information regarding past medical histories found in the Danish registries.

The study had limitations consistent with other retrospective studies. While it was easy to extract data on whether a death was witnessed or if the person was seen alive <24 h before death, it was difficult to assess more precise time limits (i.e. 1 h) in many of the unwitnessed cases.

All individuals in Denmark aged 1–35 years were included and the results in the present study may not necessarily apply to individuals aged <1 or >35 years.

Autopsy remains the gold standard for identifying the cause of death and uncertainty regarding the true cause of death remains in non-autopsied SCD cases. In our study, 30% of SCD cases were not autopsied. Although information from all available sources, including medical files, heavily suggested that the death was caused by cardiovascular disease, some of these deaths could have had other causes, leading to an overestimation of SCD.

Conclusion

In this 10-year nationwide study on SCD-cases ages 1–35 years, we found surprisingly few clinical differences between witnessed and unwitnessed SCD cases. Autopsy findings were also similar between the two groups, although cause of death remained unexplained more often among unwitnessed SCD cases compared with witnessed SCD cases. In this study, 51% of all SCD cases were unwitnessed and omitting these would dramatically change estimates of the burden of SCD. Our data support the use of witnessed and unwitnessed SCD-cases in epidemiological studies of SCD. Precaution is, however, warranted when including unwitnessed and non-autopsied cases of SCD in epidemiologic studies since a certain risk of misclassification remains.

Funding

This project has received funding from the Novo Nordisk Foundation SADS-Young grant No. NFF18OC0031634.

Conflict of interest: none declared.

Data availability

The data underlying this article are available in the article.

References

Author notes