Admission hematocrit and in-hospital progressive stroke in single small subcortical infarction: a CNSR-III cohort study

Single small subcortical infarction (SSSI) can worsen during hospitalization despite an initially mild presentation. Hematocrit (HCT) is routinely available at admission and reflects red-cell volume and oxygen-carrying capacity, but its association with neurological progression in SSSI remains uncertain. We examined the association between admission HCT and in-hospital progressive stroke in patients with SSSI. This retrospective cohort study included 2457 patients with MRI-confirmed SSSI from the Third China National Stroke Registry (CNSR-III). The primary outcome was progressive stroke, defined as an increase of ≥2 points in the National Institutes of Health Stroke Scale (NIHSS) score from admission to discharge. Logistic regression was used to estimate odds ratios (ORs) for HCT and the primary outcome. Functional dependency and mortality at 3, 6, and 12 months, alternative NIHSS thresholds, and subgroup analyses were considered secondary or exploratory. Post hoc Holm adjustment was applied to the four originally reported neurological thresholds, and archived time-to-death data were additionally evaluated using exploratory Cox regression. Progressive stroke occurred in 135 patients (5.50%). Higher admission HCT was associated with lower odds of progressive stroke (fully adjusted OR per 1% increment, 0.983; 95% CI 0.973–0.994; P = 0.002; post hoc Holm-adjusted P = 0.008). In an exploratory quartile comparison, the highest HCT quartile had lower odds than the lowest quartile (adjusted OR, 0.46; 95% CI 0.24–0.85; P = 0.013), although the category pattern did not establish a graded dose–response relationship. HCT was not associated with functional dependency at 3, 6, or 12 months. Higher HCT was associated with lower odds of 3-month mortality (adjusted OR, 0.937; 95% CI 0.892–0.984; P < 0.001), but not mortality at 6 or 12 months. In additional exploratory Cox models, adjusted mortality hazard ratios (HRs) were 0.963 at 3 months (95% CI 0.768–1.207), 0.877 at 6 months (0.771–0.997), and 0.897 at 12 months (0.821–0.979); event counts were small. Higher admission HCT was associated with lower odds of in-hospital progressive stroke in patients with SSSI. The mortality findings differed between landmark logistic and time-to-event Cox models; these and all exploratory analyses should be interpreted cautiously. These observational findings do not establish causality or support interventions intended to modify HCT.