|
|
|
|
|
Improving Death
Certificates for Data-Driven |
|
A few weeks ago, a colleague of ours distributed an article to our leadership team. The findings reported that 60% of men who died from suicide had no known mental health conditions. This is a critical issue for our health policy non-profit, because we share current research with policymakers to support policies that provide access for all people to mental health and substance use treatment when and where they need it. Needless to say, we were surprised by the findings (and even more by the headlines it generated – “Most male suicides in US show no link to mental health issues, study reveals”). We dug into the methods of this research to see how the authors arrived at their findings. Once we located the relevant information (results were buried in appendices that were not readily available), we found that the authors used the CDC’s National Violent Deaths Reporting System (NVDRS). While this is currently the best system for analyzing death certificate data; it includes records from law enforcement, coroner/medical examiners, and toxicology reports, but it does not systemically include health care data. This is a major problem as people were identified as having mental health conditions only if the condition surfaced serendipitously in law enforcement, coroner/medical examiner, or toxicology reports. In the discussion, the authors acknowledge that other research has found that 90% of deaths involved a previously diagnosed (or diagnosable) mental health condition. These unexplained discrepancies are problematic and, if they receive traction in the media, could take away from the decades of work done to bring the ramifications of untreated or under-treated mental illness into the spotlight. To prevent this, we need to improve the quantity (and validity) of information included on death certificates. In 2022, the data available for research on deaths from suicide is generally limited to what is detailed on death certificates because of data privacy regulations and the relatively small number of suicides in any given jurisdiction. In most cases, a death certifier may have never met (or treated) the person whose death they are charged with certifying. Access to the decedent’s medical records may be restricted if the death certifier does not have access to the decedent’s primary care provider’s EHR (or if the primary care doctor is unknown). This can lead to missing and potentially incorrect conclusions. Missing data in the form of omitted diagnoses also play a role. The omission of key diagnoses from death certificates is commonly observed in studies of suicide and drug overdose deaths, where most deaths have no listed contributing diagnoses related to mental health or substance use disorders. This leads to publications of research findings that are likely invalid, given the substantial measurement error and missing data. Thus, this presents a challenge for epidemiologists charged with translating the science into something that policymakers can understand and use. The “less muddy” the science is, the easier the story is to tell. There is an opportunity to tide us over until we reach utopia – a place where data from health care providers, law enforcement investigators, and medical examiners/ coroners/ justices of the peace is freely shareable for research and death certification purposes, while patient privacy is also appropriately protected. That opportunity lies in the expanded use of Part 2 of death certificates. Death Certificate 101 The U.S. Standard Certificate of Death has two sections used to identify the situational factors and health conditions related to the death: · Part 1 includes only the chain of events leading directly to the death, from the underlying cause to the immediate cause of death. (e.g., the immediate cause of death may be an upper gastrointestinal bleed, while the underlying cause of death is chronic alcohol abuse) · Part 2 can include all “other significant conditions contributing to death but not resulting in the underlying cause from Part 1.” This is where lingering mental or physical health conditions should be listed if there is no evidence that they were in the direct pathway leading to the death. While we await utopia, the real opportunity to better track the previously unmeasured impact of chronic health and social conditions is through expanding the use of Part 2. While up to 20 contributing causes can be listed, Part 2 is frequently left blank. Studies have also confirmed that certain key factors related to death are more likely to be omitted. For example, obesity is a clear risk factor for mortality, yet it is often not listed as a contributing cause of death, even when individuals die from an obesity-related condition (such as diabetes mellitus or obstructive sleep apnea). If Part 2 was completed fully, we would speculate that most articles like this one and associated confusing headlines could be prevented. The Future We envision a software solution that integrates medical records from multiple health systems along with law enforcement and emergency call center data (for mental health crisis calls to 911; or now, 988) to enable rapid and accurate death certificate completion by the certifier. This information, potentially coupled with a peer-review verification system, would permit the death certifier to access a range of health and contextual information that would aid in a cause of death determination. The rise of the SARS-CoV-2 (COVID-19) pandemic demonstrated that such integrated data collection is possible in the United States if the issue becomes a priority. Updating data technology and systems for non-infectious diseases remains a priority of the CDC and other groups, including the National Response. In summary, we described the urgent need for more complete health and mental health information to be recorded on death certificates. This will enable researchers to obtain and publish valid information on the contextual and health-related causes of mortality – no matter how indirect those causes might be. Although we propose a future, integrated data system to increase precision in the diagnoses identified as causes of death, we can focus right now on ensuring that Part 2 of the death certificate is as complete as possible. We have our work cut out for us. ■ |
|
|
