Objective To determine whether coding a developmental disability as the underlying cause of death obscures mortality trends of adults with developmental disability.
Design National Vital Statistics System 2012–2016 US Multiple Cause-of-Death Mortality files.
Participants Adults with a developmental disability indicated on their death certificate aged 18 through 103 at the time of death. The study population included 33 154 adults who died between 1 January 2012 and 31 December 2016.
Primary outcome and measures Decedents with a developmental disability coded as the underlying cause of death on the death certificate were identified using the International Statistical Classification of Diseases and Related Health Problems, Tenth Revision code for intellectual disability, cerebral palsy, Down syndrome or other developmental disability. Death certificates that coded a developmental disability as the underlying cause of death were revised using a sequential underlying cause of death revision process.
Results There were 33 154 decedents with developmental disability: 7901 with intellectual disability, 11 895 with cerebral palsy, 9114 with Down syndrome, 2479 with other developmental disabilities and 1765 with multiple developmental disabilities. Among all decedents, 48.5% had a developmental disability coded as the underlying cause of death, obscuring higher rates of choking deaths among all decedents and dementia and Alzheimer’s disease among decedents with Down syndrome.
Conclusion Death certificates that recorded the developmental disability in Part I of the death certificate were more likely to code disability as the underlying cause of death. While revising these death certificates provides a short-term corrective to mortality trends for this population, the severity and extent of this problem warrants a long-term change involving more precise instructions to record developmental disabilities only in Part II of the death certificate.
- developmental disability
- underlying cause of death
- mortality trends
This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.
Statistics from Altmetric.com
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.
Strengths and limitations of this study
This study revises death certificates that had developmental disability coded as underlying cause of death.
This study provides evidence that coding developmental disability as underlying cause of death obscures often preventable diseases or injuries other than disability that initiated the sequence of events leading to death.
The high percentage of death certificates coded with a developmental disability as the underlying cause of death reveals the need to revise death certificates for adults with developmental disability prior to reporting mortality trends.
Since the study population only includes decedents who had a developmental disability recorded on their death certificate, it is not inclusive of all decedents with a developmental disability in the USA.
Limitations to retrospective revising of death certificates highlight the need for a long-term solution involving formal changes to policies guiding coding of developmental disability on death certificates.
Public health and preventive care efforts aimed at reducing premature mortality rely on population mortality data from death certificates.1 2 Thus, it is imperative that death certificates accurately identify a valid and informative underlying cause of death (UCOD), the disease or injury that initiated the causal sequence of events leading to death. Beyond general concerns with the overall accuracy of death certificate data,3–5 researchers express specific concern regarding the frequency in which the UCOD identifies a cause that is not valid or informative for public health and preventive care efforts aimed at reducing premature mortality.2 6 7 One such cause of concern occurs when decedents with developmental disability have their disability coded as their UCOD.8–14
Developmental disabilities comprise a diverse array of conditions that originate at birth or during the early developmental part of life—intellectual disability, cerebral palsy, Down syndrome, autism as well as other chromosomal abnormalities.15 These disabilities are attributable to physical, learning, language or behavioural impairments, directly impact daily functioning and extend across a person’s life course.15 16 Developed by WHO to facilitate global medical communication and research, the International Statistical Classification of Diseases and Related Health Problems, Tenth Revision (ICD-10) classifies developmental disabilities variably as mental and behavioural disorders in the cases of intellectual disability, cerebral palsy and autism, or as chromosomal or congenital abnormalities in the cases of Down syndrome and spina bifida.17
Longevity for individuals with developmental disabilities improved since the early 1960s, resulting in a larger percentage of adults in this population now living into their 60s.18–22 Despite this improvement, on average, research informs that adults with developmental disability die at ages 20–25 years earlier than those in the general population.18 19 21 There is increasing consensus that developmental disabilities should not be considered a valid UCOD, as doing so prohibits identification of the preventable medical cause of death, and is not advantageous to public health or preventive care efforts.2 8–15 23 Instead, the goal in coding the UCOD on death certificates for persons with developmental disability should be to identify the specific, and often preventable, disease or injury other than disability that initiated the sequence of events leading to death.
Cerebral palsy offers insight to the limitations of coding a developmental disability as an UCOD. Cerebral palsy describes a group of various permanent non-progressive motor disorders which stem from an injury to or malfunction in the developing brain and vary in severity, cause and presentation.24 Besides the motor disorders that cause activity limitations, there can be associated conditions that accompany the diagnosis of cerebral palsy such as epilepsy or problems with sensation, cognition, communication and behaviour. People with cerebral palsy also have risk factors for general health conditions shared with the general population, such as chronic heart disease.25 Due to a higher prevalence of gastro-oesophageal reflux and/or dysphagia among adults with moderate-to-severe forms of cerebral palsy, there is increased risk among this population for aspiration on either saliva, food or gastric contents leading to pneumonia or pneumonitis, and subsequent death.26 27 In the event that an individual with cerebral palsy aspirates on food, develops pneumonia and dies, it is not useful to identify the UCOD as cerebral palsy. Doing so conceals the fact that death was due to aspiration and subsequent development of pneumonia, and fails to differentiate this causal pathway from other causal pathways such as arteriosclerotic heart disease leading to acute myocardial infarction and death. As this example illustrates, coding a developmental disability as the UCOD is uninformative,2 28 29 and, as Trollor et al 12 contend, ‘obscures’ actual mortality trends related to preventable conditions that need increased attention in public health or preventive care efforts.
Prior studies report the prevalence and obscuring effect of coding a developmental disability as the UCOD. Internationally, 16% of persons with developmental disability in general,12 and between 34% and 56% of persons with cerebral palsy in particular,8 30 had their disability coded as their UCOD. Among the US population, 20% of decedents with an intellectual disability,10 and 21% of decedents with Down syndrome,31 had their disability coded as their UCOD. Two studies, from Canada14 and Australia,12 that mention the obscuring effect of this practice on mortality trends report a substantial increase in deaths from respiratory disease after revising death certificates coded with a developmental disability as the UCOD. To date, researchers have not detailed the obscuring effect of coding a developmental disability as the UCOD on mortality trends for adults with developmental disability in the USA. To determine the severity of this problem, we compare mortality trends for adult decedents with developmental disability using the originally reported and a revised UCOD from US death certificate data.
Death certificate data for this study are from the National Vital Statistics System 2012–2016 US Multiple Cause-of-Death Mortality files. As mortality coding in the USA converted to ICD-10 in 2009,32 this study included the death certificates for adults aged 18–103 years at the time of death that recorded an ICD-10 code for a developmental disability as the UCOD, or as a multiple cause of death (other comorbidities present at time of death)—intellectual disability (F70-79), cerebral palsy (G80), Down syndrome (Q90) and other developmental disabilities (F80-89, Q91-99). This resulted in a sample of 33 154 death certificates over the 5-year period.
Sequential UCOD revision process
If the UCOD listed on the death certificate was not coded as a developmental disability, it was accepted as valid and retained for analysis. In the event that the developmental disability was identified as the UCOD, it was revised by the study team using a sequential UCOD revision process.8 12 We chose to identify a singular UCOD code to construct a straightforward methodology, recognising some recent studies contend that multiple morbidity data could be helpful, especially in understanding complex conditions with comorbidities, as in dementia and Alzheimer’s disease.33 34 In the instances when the UCOD was revised, we identified a valid UCOD by working sequentially from the last line to the first line of Part I of the death certificate, moving from the first to the last listed code per line. ICD-10 codes that the US Centers for Disease Control and Prevention (CDC) states are not to be used as the UCOD were not considered valid options for the revised UCOD.35 In addition, ICD-10 chapter XVIII R-codes were only used for the UCOD if no other valid UCOD was listed in part I of the death certificate.2 As they record comorbidities present at the time of death that were not part of the sequence of events leading to death, ICD-10 codes in Part II of the death certificate were not considered in the revision process.
The following examples demonstrate our sequential UCOD revision process. The first example is a decedent who had J96.9 (respiratory failure, unspecified) recorded on line 1, position 1; J69.0 (pneumonitis due to inhalation of food and vomit) recorded on line 2, position 1 and F79 (unspecified intellectual disability) on line 3, position 1, with F79 identified as the UCOD. In this instance, working sequentially from the last line (line 3), we would dismiss F79 as it is not a valid UCOD, and identify J69.0 on line 2 as the first listed valid UCOD. A second example, involving a more complex revision process, is for a decedent who had T17.9 (foreign body in respiratory tract, part unspecified) recorded on line 1, position 1; W80 (inhalation and ingestion of other objects causing obstruction of respiratory tract—commonly termed choking) recorded on line 1, position 2; G80.9 (cerebral palsy, unspecified) recorded on line 2, position 1 and R56.8 (other and unspecified convulsions) listed in line 3, position 1, with G80.9 identified as the UCOD. In this instance, we would initially identify R56.8 as the UCOD. However, as R-codes are unspecified causes and not useful for public health,2 we would continue looking for a more valid UCOD. We would dismiss G80.9 as it is not a valid UCOD, dismiss T17.9 as CDC rules do not allow this ICD-10 code to be an UCOD35 and identify W80 as the valid UCOD, superseding R56.8. In instances where no valid UCOD was present in Part I of the death certificate, as in the cases where the death certificate only listed an ICD-10 code for a developmental disability, we identified the UCOD as ‘unknown’.
All analyses were conducted using STATA V.15.0 (College Station, Texas, USA). Results are presented by developmental disability group due to disparate cause of death trends—intellectual disability, cerebral palsy, Down syndrome, other developmental disability and multiple developmental disabilities. For each group, we report: UCOD percentages by ICD-10 chapter code for all death certificates prior to and after revisions; changes in the cause of death rank order and percentage point differences between original and revised data. We also detail the leading specific underlying causes of death within ICD-10 chapters reflecting the most drastic changes.
Patient and public involvement
Patients and public were not involved in any aspect of this study, inclusive of development of research question and design, outcomes measures, analytic plan and interpretation of results. As all study participants were deceased and data are anonymous, results cannot and will not be disseminated to study participants.
Full comparative results of the original and revised UCOD by ICD-10 chapter code are presented for each disability group in tables 1–5. Although detailed in the text, information on specific underlying causes of death per ICD-10 chapter are not included in the tables. We compared results from aggregated data with results from each specific year prior and subsequent to revisions. The only dissimilarity was a slight increase in the percentage of decedents with Down syndrome who had Q90 coded as their UCOD beginning in 2014.
There were 7901 decedents with intellectual disability, 11 895 decedents with cerebral palsy, 9114 decedents with Down syndrome, 2479 decedents with other developmental disabilities and 1765 decedents with multiple developmental disabilities. The UCOD was coded as a developmental disability, and thus was revised, on the death certificates of 48.52% of all decedents with a developmental disability, but varied by disability type—25.88% of decedents with intellectual disability, 59.53% of decedents with cerebral palsy, 58.99% of decedents with Down syndrome, 26.42% of decedents with other developmental disabilities and 52.69% of decedents with multiple developmental disabilities.
As expected, after revising the death certificates, there was a steep decline in the percentage and rank order of decedents in each disability group with an UCOD in the ICD-10 chapter inclusive of their disability. Due to these revisions, the percentages increased for all other ICD-10 chapters for all disability groups. We detail the most remarkable changes.
Across groups, the percentage of deaths caused by diseases of the respiratory system, ICD-10 chapter X, rose dramatically after the revision from: 15.37% to 23.67% for decedents with intellectual disability; 8.95% to 29% for decedents with cerebral palsy; 8.56% to 27% for decedents with Down syndrome; 8.67% to 16.3% for decedents with other developmental disabilities and 12.01% to 29.24% for decedents with multiple developmental disabilities. As a result, respiratory deaths rose from the third to the first leading cause of death for decedents with intellectual disability, cerebral palsy, Down syndrome and multiple developmental disabilities; and from the fifth to the second leading cause of death for decedents with other developmental disabilities such as autism.
For all disability groups, the most prevalent underlying causes of death identified among diseases of the respiratory system were J69 pneumonitis due to inhalation of food/vomit and J18.9 pneumonia, unspecified organism. The percentage of deaths from pneumonitis due to inhalation of food/vomit increased from: 5.73% to 9.57% among decedents with intellectual disability; 3.14% to 10.08% among decedents with cerebral palsy; 5.43% to 9.84% among decedents with Down syndrome; 3.03% to 5.77% among decedents with other developmental disabilities and 5.72% to 12.18% among decedents with multiple developmental disabilities. The percentage of deaths from pneumonia, unspecified organism rose from: 4.38% to 7.1% among decedents with intellectual disability; 2.3% to 8.53% among decedents with cerebral palsy; 0.07% to 10.19% among decedents with Down syndrome; 2.02% to 4.4% among decedents with other developmental disabilities and 2.21% to 8.67% among decedents with multiple developmental disabilities.
To varying degrees, deaths from external causes of mortality, ICD-10 chapter XX, increased in per cent and rank order for four of the disability groups from: 2.67% to 7.53% (rank order increase from 9 to 3) for decedents with intellectual disability; 1.65% to 5.89% (rank order increase from 9 to 5) for decedents with cerebral palsy; 7.34% to 8.75% (rank order increase from 7 to 4) for decedents with other developmental disabilities and 1.76% to 6.06% (rank order increase from 10 to 6) for decedents with multiple developmental disabilities. While the percentage of deaths from external causes of mortality rose from 2.79% to 3.86% for decedents with Down syndrome, the rank order dropped one position from 5 to 6.
For each disability group, the most prevalent underlying causes of death in the external causes chapter were W78-80, accidental inhalation and ingestion of food or other objects causing obstruction of the respiratory tract. The percentage of deaths from accidental choking increased from: 0.1% to 4.80% among decedents with intellectual disability; 0.02% to 3.76% among decedents with cerebral palsy; 1.32% to 2.28% among decedents with Down syndrome; 3.19% to 4.48% among decedents with other developmental disabilities and 0.17% to 4.09% among decedents with multiple developmental disabilities.
One other remarkable finding among all disability groups regarded the percentage of decedents who had either an unknown or an imprecise UCOD. Unknown causes of death were recorded when no other ICD-10 code was listed in Part I of the death certificate other than disability. Imprecise underlying causes occurred when no other ICD-10 code was listed in Part I of the death certificate other than the developmental disability and/or a single or multiple R-code from chapter XVIII. As a result, the UCOD was either unknown or imprecise for: 4.96% of decedents with intellectual disability; 17.17% of decedents with cerebral palsy; 8.88% of decedents with Down syndrome; 8.51% of decedents with other developmental disabilities and 11.51% of decedents with multiple developmental disabilities. This means that even after revision, the death certificate provided little to no indication of the disease or injury that initiated the sequence of events leading to the death of these individuals.
Particular to decedents with Down syndrome, deaths from diseases of the nervous system, ICD-10 chapter VI, rose from 5.31% to 13.18% (rank order increase from 4 to 3). The most prevalent UCOD for decedents with Down syndrome in this chapter, G30.9 Alzheimer’s disease, unspecified, increased from 3.27% to 9.5%. In addition, contrary to trends among the other disability groups, deaths from mental, behavioural and neurodevelopmental disorders, ICD-10 chapter V, rose from 1.73% to 11.82% (rank order increase from 10 to 4). The most prevalent cause of death in this chapter, F03 unspecified dementia, increased from 1.62% to 11.4%.
In 2012–2016 US death certificate data, the coding of a developmental disability as the UCOD is widespread, but varied among specific disability groups. Results from this study expose that the practice of coding a developmental disability as the UCOD on a death certificate obscures the higher prevalence of choking-related deaths (coded as either pneumonitis due to inhalation of food/vomit, or accidental choking) among all developmental disability groups, and obscures dementia and Alzheimer’s disease-related deaths among decedents with Down syndrome.
In instances when death certificates that code a developmental disability as the UCOD are not revised, population mortality trends for persons with various types of developmental disability are likely to be inaccurate and may, therefore, misinform public health and preventive care efforts aimed at preventing premature mortality. The primary strength of this study is that results demonstrate that the prominence of choking-related deaths as a mortality risk among persons with developmental disability was severely underestimated prior to revision. As revision of the UCOD revealed, similar to the older adult population,36–38 careful attention is needed to attend to the heightened risk of preventable choking-related deaths among persons with developmental disability.15 39 40 Thus, it is of the utmost importance that public health and preventive care efforts focus on management of the swallowing function among this population. Physicians conducting routine exams should be screening for swallowing disorders, abnormalities in the cough and gag reflex as well as respiratory distress.15 40 Mortality data based on original death certificates, often coding developmental disability as UCOD, fail to reveal this important mortality trend.
Unfortunately, current CDC instructions and WHO guidelines for completing death certificates permit identifying a developmental disability as the UCOD in at least two known instances.41 42 ICD-10 codes for intellectual disability (F70-79) are permitted as UCOD in instances that the actual UCOD is unknown. In addition, Down syndrome is suggested as the UCOD in the event the death certificate indicates unspecified dementia or Alzheimer’s disease. Continued permission of the practice of coding a developmental disability as the UCOD demands both short-term and long-term redress. As this study highlights, analysis of mortality trends for adults with developmental disability must recognise the obscuring effect of coding a developmental disability as the UCOD and take the necessary steps to revise the UCOD on these death certificates prior to reporting findings. The sequential UCOD revision process used in this study reduces obfuscation of mortality trends among populations with developmental disabilities in the short-term.
While revising death certificates for adults with developmental disability is a necessary short-term solution, the primary limitations of this study indicate that retrospective revision is not a sufficient long-term solution. The first limitation regards accuracy of revised death certificates. Although sequentially revising the UCOD for US death certificates that identify a developmental disability as the UCOD increases the reliability of mortality trends for this population, it is not possible to verify the accuracy of the revised UCOD without access to the medical records of the decedent and the medical personnel that completed the death certificate. Although the only remedy present at the time, the method of retrospectively revising the UCOD should not be viewed as a sustainable long-term strategy for surveilling mortality trends among this population. Instead, focus should be on ensuring that at the time of death, the medical certifier of the death certificate, the individual completing the medical portion of the death certificate, accurately identifies the disease process or injury, other than developmental disability, that initiated the chain of events leading to death. In all US states, this would be either the attending physician present at the time of death or the decedent’s personal physician, with some states allowing the chief medical officer of medical facilities to certify.43 44 In instances when death occurs without an attending physician present, cause of death is unknown, or death occurs by accident, suicide or homicide, the individual responsible for certifying the cause of death would be the medical examiner or coroner, with responsibility varying by US state.43 45 In addition, it is unlikely that the death certificates of all decedents with a developmental disability included an ICD-10 code for developmental disability. Based on inconsistencies regarding the proper location to code developmental disability on death certificates we observed in this study, it is obvious that there is confusion regarding where to record a developmental disability on the death certificate. This confusion may result in the medical certifier not recording the developmental disability on the death certificate at all. Thus, the results from this study only describe mortality trends for decedents with developmental disability who had their disability recorded on their death certificate, and do not account for possible state level variation in the cause of death certification process.
In order to fully address these limitations, it is necessary to move beyond retrospective remedies, and formally change instructions for coding a developmental disability on the death certificate. As CDC instructions on cause of death coding are developed based on and in cooperation with WHO guidelines,46 47 change is needed at both a national and international level. Due to concerns that population representative datasets rarely include adults with developmental disability, which limits surveillance of morbidity and mortality trends among this population,48 49 it is imperative to continue recording developmental disabilities on death certificates. Per CDC and WHO guidelines, Part II of the death certificate is intended for the recording of comorbidities that were present at the time of death, but were not part of the sequence of events leading to death.41 42 To ensure that developmental disabilities are recorded on the death certificate, but not identified as the UCOD, instructions for completing death certificates should specify that developmental disabilities should not be recorded in Part I or permitted as the UCOD. Instead, developmental disabilities should be recorded in Part II of the death certificate.10 11 13 Evidence from this study supports this recommendation. Among the death certificates that recorded a developmental disability in Part I (with or without also recording a developmental disability in Part II), 82% coded developmental disability as the UCOD. In contrast, among the death certificates that recorded a developmental disability only in Part II, 19% coded developmental disability as the UCOD. This proposed change would preserve the ability to surveil mortality trends for this population by ensuring the recording of the developmental disability on the death certificate while minimising the possibility a developmental disability is coded as the UCOD. Changing the instructions for coding developmental disability would also increase the accuracy of mortality data for adults with developmental disability by deterring the coding of a developmental disability as an UCOD. As a result, the death certificates of this population would more accurately represent actual mortality trends, allowing for better-informed public health and preventive care efforts to reduce premature mortality for this population.
Contributors SL had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Concept and design: SL, JDS, MT. Acquisition of data: SL. Analysis of data: SL, JDS. Interpretation of data: SL, JDS, MT. Drafting of the manuscript: SL, JDS. Critical revision of the manuscript for important intellectual content: SL, JDS, MT.
Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.
Competing interests None declared.
Provenance and peer review Not commissioned; externally peer reviewed.
Data sharing statement All data are publicly available from the Centers for Disease Control and Prevention.
Patient consent for publication Not required.