Diagnostic disparities in sickle cell disease: Rates and implications Free

Diagnosing sickle cell disease later in life can be challenging. Research has highlighted the importance of early diagnosis, particularly during the neonatal period, as it significantly improves morbidity and mortality rates. While multiple diagnostic methods and expert opinions can enhance accuracy, resource availability varies across settings, leading to potential questions about the reliability of recorded diagnoses. To this end, there is a significant lack of research identifying the factors that contribute to diagnostic error in identifying individuals with sickle cell disease. The ASH Research Collaborative SCD Data Hub Report from 2025 (ASH Research Collaborative, 2025) does outline key findings in relation to sickle cell disease diagnosis, noting discrepancies in diagnostic codes documented on electronic health records versus individuals with a physician-attested diagnosis. In our study, we explore this notion further via an analysis of the patients documented as having sickle cell disease within our healthcare system.

We conducted a retrospective analysis of patient electronic chart data versus ICD-10 coding of sickle cell disease. This included the examination of rates of misdiagnosis for sickle cell disease and a detailed evaluation via regression analysis of the number of encounters associated with these diagnostic codes. Statistical significance was set at p < 0.05. All analyses were performed using IBM SPSS Statistics version 30.

A total of 563 adult subjects were identified by the ICD-10 code of D57 in our system. Five hundred and one (89%) were alive at the time of analysis, and 62 (11%) were deceased. Twenty-five patients (4.4%) had a coded diagnosis of hemoglobin SS (D57.1), 91 (16%) had a diagnosis of sickle cell crisis without designation of sickle cell type, 83 (14.7%) were coded with hemoglobin SC disease, and 1 patient was coded as sickle cell trait. Twenty-seven (4.8%) were coded as sickle cell thalassemia, 4 (0.7%) as other sickle cell disorders, and 3 (0.5%) had CVA due to hemoglobin S. Actual diagnosis from electrophoresis was incorrect in 69 patients (12.3%) with sickle cell disease. Within this subset, 27 patients with normal AA variant and 36 patients with sickle cell trait were coded as sickle cell disease or crisis despite electrophoresis findings to the contrary. In 76 (15%) of subjects, the diagnosis could not be determined by chart review. An incorrect diagnosis was more likely if the initial code was from primary care (p= 0.0428). There was a statistically significant difference between the number of coded encounters between individuals with the correct diagnosis (65.6) and those with incorrect diagnosis (7.6; p = 0.0002).

Our findings indicate a significant shift from the findings of the 2025 SCD Data Hub Report, which reported an overall discrepancy of 47.8%, versus our finding of 12.3%, for active individuals with confirmed SCD diagnosis relative to active individuals reported within the system. Our data, which is also included in the Data Hub Report, demonstrates a more conservative trend for sickle cell misdiagnosis specifically amongst adults, which certainly merits further exploration. Indeed, the etiology of this difference is likely to be multifactorial, particularly when considering additional variables such as the location of diagnosis, which may play a large role when considering heterogeneity of resources available. Additional factors, such as age at diagnosis or receiving care in multiple systems, may also play a significant role, although this may be more difficult to ascertain. Moreover, the significant difference in the number of encounters between correctly versus incorrectly diagnosed individuals suggests a decreased illness burden, further highlighting the importance of establishing an accurate diagnosis, opening more avenues for research to understand these discrepancies. These results signify an important avenue for further research into the factors that may be contributory, which in turn may have significant ramifications on disease burden as well as burden on our healthcare system.