Nature and Extent of External-Cause Deaths of Nursing Home Residents in Victoria, Australia Joseph E. Ibrahim, PhD,* Briony J. Murphy, BA,* Lyndal Bugeja, PhD,* and David Ranson, DMJ(Path)*†

OBJECTIVES: To describe the nature and extent of external-cause deaths of residents of nursing homes in Victoria, Australia. DESIGN: A retrospective cohort study of all decedents using routinely collected data contained within the National Coronial Information System. SETTING: Accredited nursing homes in Victoria. PARTICIPANTS: Nursing home residents who had died from external causes and whose deaths were reported to the Coroners Court between July 1, 2000, and December 31, 2012. MEASUREMENTS: Basic descriptive analysis was conducted to measure frequencies and proportion of exposures within each outcome group, and rates were calculated using population data. RESULTS: One thousand two hundred ninety-six external cause deaths of nursing home residents were identified. Deaths were due to falls (n = 1,155, 89.1%), choking (n = 89, 6.9%), suicide (n = 17, 1.3%), complications of clinical care (n = 8, 0.6%) and resident-on-resident assault (n = 7, 0.5%). Deaths occurred more frequently in women (n = 814, 62.8%), in keeping with the sex distribution in nursing homes, and residents aged 85 and older (n = 923, 71.2%). The number of inquests held to investigate a death as a matter of public interest was small (n = 24, 1.9%). CONCLUSION: A significant proportion of nursing home resident deaths are from external causes and are potentially preventable. A shift in community attitudes is required toward an understanding that premature death of a resident from injury is not a natural part of life. J Am Geriatr Soc 63:954–962, 2015.

Key words: nursing home; resident; death; external cause

From the *Department of Forensic Medicine, Monash University, Melbourne, Victoria, Australia; and †Victorian Institute of Forensic Medicine, Melbourne, Victoria, Australia. Address correspondence to Dr. Joseph E. Ibrahim, 65 Kavanagh Street, Southbank, Victoria 3006, Australia. E-mail: [email protected] DOI: 10.1111/jgs.13377

JAGS 63:954–962, 2015 © 2015, Copyright the Authors Journal compilation © 2015, The American Geriatrics Society

ealthy aging is a global public health goal,1 and Western developed nations continue to raise concerns about the current state of, and need to improve, care for the older population residing in nursing homes.2–5 These concerns must be addressed as the population of older persons is growing rapidly. In 2013, there were 841 million (11.7%) people aged 60 and older globally, which is expected to increase to more than 2 billion (21.1%) by 2050.1 In Australia, more than 186,000 residents live in approximately 2,700 nursing homes.6 These older people are at risk of death from external causes7 such as trauma and complications of clinical care because of their physical frailty, comorbidities, drug regimens, and the need for care coordination.8 However, there is little information about the frequency and determinants of deaths in nursing homes that may be premature or preventable.9,10 Reasons for this include underuse of routinely collected data, a perception that these deaths are not preventable, small public health gains, and lack of a specific responsible agency.9 In Australia, coroners (lawyers with a judicial appointment) investigate a wide range of deaths, including those related directly or indirectly to trauma and any unexpected deaths. Coroners integrate the statements of witnesses and the findings of pathologists, police and other experts and may hold inquest hearings to further explore the circumstances of the death. Coroners make formal judicial findings on how each death occurred that may include detailed recommendations about how that death may have been prevented and address broader questions of public health and safety.11 Aggregating and analyzing data from coroners’ findings provides information about the nature and burden of nursing home deaths and an evidence base for public health prevention strategies.9,12


METHODS Study design This was a retrospective cohort study of all decedents using routinely collected data from deaths reported to the Coroners Court of Victoria (CCOV).



MAY 2015–VOL. 63, NO. 5





The study population was drawn from 56,855 deaths reported to the CCOV between July 1, 2000, and December 31, 2012 (Figure 1). Information on the broader nursing home population was obtained from government reports.13–17

The primary data source was the National Coronial Information System (NCIS), a national data storage and retrieval system containing information on deaths reported to coroners in Australia since July 2000. The NCIS contains variables (coded and free text) and four full-text documents: narrative of circumstances from the police report, autopsy report, forensic toxicology report, and coroner’s finding.18

A nursing home was defined as a licensed business and accredited service with the Australian Aged Care Standards and Accreditation Agency (ACSAA) for long-term care of older persons. A resident was a person who lived in a nursing home temporarily (respite care) or permanently. An external cause was defined as any “Damage to the body caused by (acute) exchanges with environmental energy that are beyond the body’s resilience. The energy may be mechanical, thermal, electrical, ionising radiation or chemical (poisoning, asphyxiation).”7 The level of care that a nursing home provided was according to Department of Health classifications, with high level of care defined as regular assistance with most personal activities of daily living (e.g., eating, toileting) and a low level of care as assistance with domestic activities (e.g., meals, laundry) and occasional assistance with personal activities.19


Case Identification

The study population comprised nursing home residents who had died from an external cause.

Cases were identified with the assistance of NCIS staff, who provided feedback on a list of search terms that the

Data Source




Population of Victoria (13, 17) =

4.7 m

5.6 m

# of nursing homes (14, 15) =

Persons >65 years old (13, 17) =

0.6 m

0.8 m

# of residents (16) =





All deaths recorded in NCIS (July 1, 2000 – Dec. 31, 2012) n = 56, 855

Deaths where incident location is coded in NCIS as aged care, palliative care or residential care service n=5,449 Incident did not occur in a nursing home n=1,350 Death confirmed in a nursing home n=4,099 Deceased not nursing home resident (i.e., visitor) n=8 Deaths among nursing home residents n=4,091

Open cases n=64

Final cohort n=4,027

Natural-cause deaths n=2,731

External-cause deaths n=1,296

Figure 1. Flowchart for case inclusion. NCIS = National Coronial Information System.




investigators generated and searched for all deaths in which the incident location was classified as home for the elderly or retirement village; nursing home; hospice, palliative, or respite care; or residential care facility. A unit record data extract of the search results was provided to the investigators.

Inclusion Criteria The following inclusion criteria were applied: the incident resulting in death occurred at a nursing home accredited by ACSAA, the person was a resident of a nursing home, and the coroner’s investigation was closed as of June 30, 2013. Nursing homes that ceased operations during the study period were identified by a retrospective search of the Department of Health archived annual lists of accredited residential aged care services for the years 2001–2011, and were included in the study.

Data Collection The following variables were extracted for each death: age and sex of the deceased, time and day of occurrence of the incident, nursing home address, place of death, and outcomes of coroner’s investigation (mechanism of injury, cause of death, whether an inquest was held, and any formal recommendations made). Information about the nursing home was sourced from the ACSAA site audit reports for geographic location (regional or metropolitan), size according to number of residential places available, level of care available (high care only, low care only, mixed), and type of ownership (public, private, charity).

Data Analysis SPSS version 15 (SPSS, Inc., Chicago, IL) was used to conduct a series of univariate and bivariate analyses on variables related to the deceased person, the coroner’s investigation, and the circumstances of the death. Data were presented in a cross-table to measure frequencies and proportion of exposures within each outcome group (cause of death). A research officer (BM) checked types of death as classified according to the NCIS assigned categories for intent and primary mechanism of death (based on International Classification of Diseases, Tenth Revision, coding). Two independent researchers, one a clinical practitioner and the other experienced in medico-legal death investigations (LB), then verified the mechanism of death and location of the incident using all available original documentation (coroner’s finding, autopsy, police reports). A qualified medical practitioner in geriatric medicine (JEI) also reviewed all deaths from choking, suicide, resident-toresident assault (RRA), and complications of clinical care for any misclassifications. Population data regarding nursing home places from the Australian Institute of Health and Welfare were used to calculate odds ratios (ORs) to quantify the magnitude of risk of death according to age and sex. The significance of the ORs was evaluated by calculating 95% confidence intervals and P-values. Data on total bed days were available only for 2002 to 2012, so comparative study data in

MAY 2015–VOL. 63, NO. 5


Table 2 excludes deaths reported to the coroner during 2000 and 2001.

Ethics The Victorian Institute of Forensic Medicine Research Advisory Committee (RAC 011/13) granted ethics approval, and the Department of Justice Human Research Ethics Committee granted approval for access to the NCIS (CF/13/ 8187).

RESULTS Overview Of the 4,027 deaths of nursing home residents, approximately one-third (n = 1,296) were due to external causes. The majority were unintentional (1,263, 97.5%), comprising falls (1,155, 89.1%); choking (89, 6.9%); and other unintentional deaths (19, 1.5%), including other threats to breathing, poisoning, and thermal mechanisms. The intentional causes were suicide (17, 1.3%), RRA (7, 0.5%), and a single death from homicide (Table 1). Complications of clinical care comprised eight deaths (0.6%). The majority of these were medication errors (n = 5) leading to inadvertent administration of an oral hypoglycemic, cardioactive agents, antipsychotics, and a laxative. The other cases were sepsis (n = 3) from skin pressure injury, a malpositioned nasogastric tube into the lung, and a joint prosthesis.

Reporting Trends There was a sustained increase in the number of deaths over the study period. This was almost entirely due to falls, which doubled from 2002 to 2003 (11 and 21 deaths, respectively) and again from 2007 to 2008 (45 and 119 deaths, respectively). Choking as a cause of death also doubled between 2007 and 2008 (6 and 15 deaths, respectively). Other external causes of death were relatively constant (range of 0–4 deaths per annum).

Inquests and Coroners’ Recommendations for Prevention Coroners rarely held an inquest (an open court hearing) (24, 1.9%), of which almost half were for deaths due to RRA (5/24, 20.8%) and complications of clinical care (5/ 24, 20.8%). This is probably because of the public interest in such deaths, as together these two causes accounted for only 15 of 1,296 (1.2%) deaths. Coroners rarely made a formal recommendation for prevention (12/1,296, 0.9%).

Demographic Characteristics Demographically, external-cause deaths occurred more frequently in women (814, 62.8%), in keeping with the gender distribution in nursing homes, and people aged 85 and older (923, 71.2%) (Table 2). The age distribution of nursing home residents is similar to that of the study


MAY 2015–VOL. 63, NO. 5



Table 1. External-Cause Deaths in Nursing Homes, Victoria, 2000–12 Unintentional


Fall Related, n = 1,155

Choking, n = 89

Sex Female 743 (64.3) 40 (44.9) Male 412 (35.7) 49 (55.1) According to sex Female 743 (91.3) 40 (4.9) Male 412 (85.5) 49 (10.2) Agec

Nature and extent of external-cause deaths of nursing home residents in Victoria, Australia.

To describe the nature and extent of external-cause deaths of residents of nursing homes in Victoria, Australia...
132KB Sizes 0 Downloads 7 Views