When the Missouri Department of Health and Senior Services (DHSS) studied death certificates at hospitals with high inpatient death rates, it discovered that 45.8% of the certificates listed a cause of death inconsistent with CDC guidelines, according to a study published in the Morbidity and Mortality Weekly Report (MMWR).
Heart disease, renal disease, and cancer were studied as underlying causes of death, meaning the illness that began the sequence of events leading to death. Some hospitals were found to underreport cancer and overreport heart disease and renal disease.
Among the 32 hospitals in the study, heart disease was incorrectly identified on up to 85% of the death certificates examined, renal disease on up to 44%, and cancer on up to 9%.
Causes of death are an extremely important metric within the medical field. Data are used to spot and track disease outbreaks, as a basis for health policies and to direct research priorities and funding. So inaccuracies can result in real impacts. Another recent study examining causes of death in urban versus rural areas is an apt example of the kind of information that could be in question as two of the causes of death in that study were inaccurate in the MMWR study: heart disease and cancer.
Not to mention, medical errors are the third leading cause of death, The BMJ reported in May 2016.
That said, the researchers named three important limitations to the study. The researchers assumed that the cause of health in hospital medical charts was more accurate than that listed on the death certificate. They also commented that different reviewers of the same information might have reached different conclusions. Most notably, 12% of the medical charts considered for the study sample were either unavailable or lacked sufficient information, a lapse that suggests a broader problem with medical documentation.
However, all three of these limitations could potentially be more than offset by the fact that inaccuracies were identified by comparing the Missouri hospitals’ data to national benchmarks established by the CDC. If discrepancies like those found in the Missouri data are commonplace, the CDC’s benchmark data are likely to be flawed.
The Missouri DHSS is now offering web-based training to help physicians improve the accuracy of their cause-of-death reports, an idea that may need to be widely implemented.