This article is more than 5 years old. Information may no longer be current.
Swedish registry data overstate prevalence of Cushing's disease
Click Here to Manage Email Alerts
An analysis of Swedish registry data revealed that approximately half of 1,300 patients identified as having Cushing’s disease were misdiagnosed, according to findings published in Pituitary.
“More than 50% of individuals who had received a diagnostic code for Cushing’s syndrome did not — after review of the medical records — prove to have the syndrome,” Oskar Ragnarsson, MD, PhD, associate professor and a senior consultant in internal medicine and endocrinology at the Sahlgrenska Academy at the University of Gothenburg, Sweden, told Endocrine Today. “Thus, the study demonstrates the importance of validation of the diagnosis when using registries for epidemiological research in rare diseases, such as Cushing’s disease. Otherwise, the epidemiologic data may provide misleading information on both mortality and comorbidities.”
In a retrospective study, Ragnarsson and colleagues analyzed medical records from 1,317 patients who received a diagnostic code for Cushing’s disease between 1987 and 2013 in the Swedish Patient Register. Researchers validated diagnoses through a collection of clinical, biochemical, imaging and histopathological data reviewed by an endocrinologist experienced in diagnosing Cushing’s disease. Researchers stratified patients by five groups: confirmed Cushing’s disease; other forms of the disease (such as cortisol-producing adenoma); possible but not confirmed Cushing’s disease diagnosis; diagnoses related to Cushing’s disease (suspected Cushing’s disease ruled out after evaluation); and diagnoses unrelated to Cushing’s disease.
After a review of patient records, researchers verified a diagnosis of Cushing’s disease in 534 patients (41%), whereas 610 patients (46%) did not have the disease. In 41 patients (3%), the diagnosis was considered probable but could not be confirmed, according to researchers, and medical records for 10% of patients were missing.
Of the 610 patients without Cushing’s disease, 334 patients (25%) had diagnoses unrelated to Cushing’s syndrome, 156 (12%) had endogenous Cushing’s syndrome of other causes, 86 (7%) had suspected Cushing’s syndrome that was ruled out after evaluation and 32 (2%) had drug-induced Cushing’s syndrome.
Researchers observed that 390 patients were diagnosed with Cushing’s disease in Sweden between 1987 and 2013 (144 patients diagnosed before 1987 were excluded). Mean annual incidence was 1.6 cases per million (95% CI, 1.4-1.8). Mean annual incidence was 1.5 cases per million between 1987 and 1995, 1.4 cases per million between 1996 and 1995 and two cases per million between 2005 and 2013.
“A higher incidence between 2005 and 2013 compared with 1987 and 2004 was noticed,” Ragnarsson said. “We don’t know if this reflects an increasing incidence of the disease or simply increased awareness, earlier recognition and earlier diagnosis.”
The researchers noted that the findings underscore the importance of validating diagnoses in patients with presumed Cushing’s disease from large health care registries.
“A correct diagnosis of [Cushing’s disease] was verified in less than half of the patients,” the researchers wrote. “Thus, the results from previous and future epidemiological studies using ICD coding for inclusion of patients with [Cushing’s disease], as well as other rare and complex diseases, should be questioned.” – by Regina Schaffer
For more information:
Oskar Ragnarsson, MD, PhD, can be reached at Institute of Medicine at Sahlgrenska Academy, University of Gothenburg, Gröna Stråket 8, 413 45, Gothenburg, Sweden; email: oskar.ragnarsson@medic.gu.se.
Disclosures: The authors report no relevant financial disclosures.
Click Here to Manage Email Alerts