Albumin-Based Inflammatory–Nutritional Indices as Novel Biomarkers for Severity Stratification and Re-Hospitalization Risk in Hyperemesis Gravidarum: A Retrospective Case–Control Study

Background: The aim of this study was to evaluate the diagnostic and prognostic performance of albumin-based inflammatory–nutritional indices in hyperemesis gravidarum (HG) and to determine their associations with disease severity and risk of re-hospitalization. Methods: This retrospective case–control study included 246 women with HG and 246 gestational-age-matched healthy pregnant controls at 6–16 weeks of gestation. Disease severity was classified as mild, moderate, or severe using the Pregnancy-Unique Quantification of Emesis (24 h scale) (PUQE-24) score. A comprehensive panel of albumin-based inflammatory indices—including C-reactive protein-to-albumin ratio (CAR), fibrinogen-to-albumin ratio (FAR), neutrophil-to-albumin ratio (NAR), leukocyte-to-albumin ratio (LAR), neutrophil percentage-to-albumin ratio (NPAR), monocyte-to-albumin ratio (MAR), hemoglobin–albumin–lymphocyte–platelet (HALP) score, modified HALP (m-HALP) score, prognostic nutritional index (PNI) score, systemic immune-inflammation index-to-albumin (SII/Alb), and systemic inflammatory response index-to-albumin (SIRI/Alb)—was calculated from routine complete blood count and serum biochemistry results obtained at diagnosis. Receiver operating characteristic analysis, along with univariate and multivariate logistic regression models, was performed to evaluate diagnostic performance and identify predictors of severe HG and re-hospitalization. Results: Albumin-based indices exhibited severity-associated alterations, with an overall trend toward worsening immuno-nutritional status across increasing HG severity. Among these, m-HALP score demonstrated the strongest inverse correlations with PUQE-24 score, ketonuria grade, length of hospital stay, and re-hospitalization risk (r = −0.74 to −0.52; all p < 0.001) and achieved the highest discriminative accuracy for both severe HG (AUC 0.864, 95% CI 0.836–0.892, p < 0.001) and re-hospitalization (AUC 0.722, 95% CI 0.675–0.766, p < 0.001). In multivariable analysis, higher HALP, m-HALP, and PNI were independently associated with a lower likelihood of severe HG. For re-hospitalization, higher m-HALP and HALP were independently associated with a lower risk, whereas higher NPAR, higher ketonuria grade, and higher PUQE-24 score were independently associated with an increased risk of re-hospitalization. Conclusions: Albumin-based indices, particularly m-HALP, demonstrated robust diagnostic and prognostic performance in HG compared with conventional biomarkers. These readily available, cost-neutral composite biomarkers enable objective severity stratification and accurate identification of patients at elevated risk of recurrent hospitalization, offering immediate potential to guide personalized, evidence-based clinical management.