Development of a Clinical Severity Score for Indian Sickle Cell Anaemia Patients
NONGMAITHEM, S. S.; BISWAS, A.; IYER, S.; VAISHNAVI, J.; WATH, A.; CHANDAK, G. R.; JAIN, D.
Show abstract
BackgroundSickle cell anemia (SCA) is a common monogenic disorder but phenotypic heterogeneity is common among patients, especially in Indians. Scores to label them as per severity have mostly included non-Indians. We investigated the utility of existing pediatric severity score (PSS) in Indian patients and attempted to develop a severity score to facilitate informed management decisions. Materials and MethodsA total of 171 SCA patients were recruited and two clinical experts categorized them into mild, moderate and severe groups based on clinical and biochemical parameters. We generated PSS and two other modified scores viz. Indian Severity Score 1 (ISS1) by including additional four clinical parameters, and Indian Severity Score 2 (ISS2) by replacing four biochemical measures with related clinical parameters. The patients were randomized and using the training set (N=86), severity scores cutoff values were decided and severity status was established in the testing set (N=85). Overall concordance between severity score-based and clinical expert categorization was calculated in three randomized sets. ResultsUsing PSS, only 2/3rd (66.7%) of Indian patients matched with the clinical assessment; the modified scores significantly improved the concordance; ISS1 (82.8%) and ISS2 (85.1%). Results were similar in all three random sets (80.0%-84.71% and 80.0%-88.2% for ISS1 and ISS2 respectively), suggesting robustness of modified scores in Indian patients. The highest concordance was observed in mild (81-100%) followed by severe (57-85%) groups in all three severity score models. The lowest concordance was observed in the moderate group (10-48%). ConclusionsWe have developed a robust population-specific score for identification of severity status in young Indian SCA patients. Inclusion of specific clinical symptoms in Indian patients underlines the importance of population-specific features to correctly categorize SCA patients. Further exploration of its utility in other populations is needed.
Matching journals
The top 4 journals account for 50% of the predicted probability mass.
Similar papers in this journal
- Association of ABO blood group with COVID-19 severity, acute phase reactants and mortality 94%
- Coagulation abnormalities in children with uncorrected congenital heart defects seen at a teaching hospital in a developing country 94%
- Use of an automated pyrosequencing technique for confirmation of Sickle Cell Disease 94%
Similar papers in this journal
- Pulmonary haemorrhage as a frequent cause of death among patients with severe complicated Leptospirosis in Southern Sri Lanka 91%
- Comparative clinical transcriptome of pir genes in severe Plasmodium vivax malaria 90%
- Assessment of the new World Health Organization’s dengue classification for predicting severity of illness and level of healthcare required 90%
Similar papers in this journal
Similar papers in this journal
- Characteristics and outcomes of cases of children and adolescents with pediatric inflammatory multisystem syndrome in a tertiary care center in Mexico City 93%
- Delineation of Single-cell Altas Provides New Insights for Development of Coronary Artery Lesions in Kawasaki Disease: Bad and Good Signaling Molecules 89%
- Potential of host serum protein biosignatures in the diagnosis of tuberculous meningitis in children 88%
Similar papers in this journal
- Combination therapy of Tocilizumab and steroid for management of COVID-19 associated cytokine release syndrome: A single center experience from Pune, Western India 90%
- The PBL teaching method in Neurology Education in the Traditional Chinese Medicine undergraduate students: An Observational Study 89%
- Demographic and Clinical Characteristics of Pediatric COVID-19 in Arkansas: March-December 2020 88%
"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.