ABSTRACT Aim Hypovolaemic shock (HS) is an uncommon but life‐threatening complication in paediatric nephrotic syndrome (NS), yet current guidelines lack a quantitative admission‐time instrument to flag high‐risk children. We aimed to develop and internally validate a simple model for early HS using routinely available clinical‐laboratory data. Methods We performed a single‐centre observational study including 258 children with NS admitted to a tertiary paediatric hospital. HS within 24‐h of admission was the outcome. Candidate predictors were routine admission variables. Multivariable logistic regression was used to derive a parsimonious model, which was presented as a nomogram. Internal validation and clinical utility were assessed using cross‐validation and decision‐curve analysis (DCA). Results HS occurred in 37/258 (14.3%). The final 4‐variable model retained age, haematocrit, serum albumin and serum sodium. Two markers remained strongly and independently associated with HS: serum albumin, OR = 2.17 per 1 g/L decrease (95% CI 1.68–3.07; p < 0.001); and serum sodium, OR = 5.79 per 5 mmol/L decrease (95% CI 1.86–22.75; p = 0.006). Age (OR 0.90 per year) and haematocrit (OR 1.47 per +5%) modestly improved discrimination. Apparent discrimination was excellent (AUC = 0.966; 95% CI 0.945–0.988; Brier 0.049) and remained high on cross‐validation (AUC = 0.959; 95% CI 0.934–0.983; Brier 0.055). DCA showed a higher net benefit of the model than ‘treat all’ or ‘treat none’ between 20% and 50% risk. Conclusion A four‐variable admission nomogram can stratify early HS risk in children with NS using routine data. External multicentre validation remains necessary.
Quang et al. (Sun,) studied this question.