Patients diagnosed during or shortly following emergency hospital care may have higher 1-year mortality across a range of noncancer conditions.
In parallel disease-specific cohort studies, investigators used linked primary care, hospital, and mortality records from the Clinical Practice Research Datalink, Hospital Episode Statistics, and Office for National Statistics to compare the frequency of emergency diagnosis, 1-year all-cause mortality, and total overnight hospital stays during the year postdiagnosis between 1.7 million patients diagnosed with one of 13 non-neoplastic conditions as an emergency or not as an emergency from 1999 to 2019. The conditions included axial spondyloarthritis, celiac disease, coronary or ischemic heart disease, chronic obstructive pulmonary disease (COPD), inflammatory bowel disease (IBD), Lyme disease, multiple sclerosis (MS), Parkinson's disease, polyendocrine metabolic ovarian syndrome, rheumatoid arthritis, schizophrenia and other chronic psychoses, subacute bacterial endocarditis, and tuberculosis.
Emergency diagnosis was defined as a diagnosis occurring on the day of or within 30 days following the start of an emergency hospital admission. Models adjusted for age, year of diagnosis, deprivation, comorbidity burden, and diagnosis source.
Emergency diagnoses occurred in at least 20% of patients in nine of the conditions, including more than 30% of those with Parkinson's disease and 35% of those with COPD. The frequency varied substantially by condition, ranging from 6% among female participants with Lyme disease to 83% among male participants with subacute bacterial endocarditis.
The difference in 1-year mortality between patients diagnosed as an emergency and those who were not was at least 10 percentage points for both male and female participants in nine of the conditions. Male participants with axial spondyloarthritis, (20% vs. 2%), female participants with coeliac disease (11% vs. 1%), and male participants with MS (15% vs 1%) had greater 1-year mortality following emergency diagnosis. The mortality associations generally persisted after adjustment. Some of the largest associations occurred in conditions typically associated with good prognosis, such as celiac disease and IBD.
Hospital use showed a similar pattern. Male participants with celiac disease diagnosed as an emergency spent a mean 19 vs. 2 nights in the hospital within 1 year postdiagnosis compared with those who were not diagnosed as an emergency. Among female participants with IBD and MS, the corresponding durations were 19 vs. 3 and 17 vs. 2 nights.
Emergency diagnosis varied by patient characteristics. Patients diagnosed as an emergency were generally older with greater comorbidity burdens. For most conditions, the frequency of emergency diagnosis followed a U- or J-shaped pattern with age and was higher among those living in areas with greater deprivation.
In sensitivity analyses, expanding the definition of emergency diagnosis to include emergency department attendance modestly increased its frequency, while the association with poorer outcomes persisted. Excluding patients diagnosed in 2019 to account for potential effects of the COVID-19 pandemic did not meaningfully alter the findings.
The investigators cautioned that they could not establish whether emergency diagnosis contributed to poorer outcomes. Patients diagnosed as an emergency may have differed from other patients in disease severity and other characteristics not fully captured by the analysis. The investigators relied on accurate recording of diagnoses and diagnosis dates. In addition, its definition of emergency diagnosis could have captured diagnoses made incidentally during emergency admissions prompted by unrelated conditions. Whether preventing emergency diagnoses or diagnosing patients through alternative routes would improve outcomes remains uncertain.
“Emergency diagnoses affect many patients across a wide range of conditions, and patients diagnosed as an emergency are consistently more likely to die and to spend longer in hospital in the year after diagnosis,” wrote lead study author Emma Whitfield, of Epidemiology of Cancer Healthcare & Outcomes in the Department of Behavioural Science and Health in the Institute of Epidemiology and Health Care at the University College London, and colleagues.
Full disclosures of the study authors can be found in the study
Source: PLOS Medicine
