Agreement of abdominal palpation with point-of-care ultrasound for hepatosplenic organomegaly: a cross-sectional, population-based study in schistosomiasis-endemic settings of rural Uganda

Background: Ultrasound is commonly used to assess hepatosplenic morbidity in schistosomiasis, but access may be limited in resource-constrained settings. Abdominal palpation is simple and inexpensive, yet its clinical value and diagnostic performance for detecting ultrasound-defined organ enlargement remains uncertain. We evaluated standardised abdominal palpation against point-of-care ultrasound in a schistosomiasis-endemic setting in rural Uganda. Methods: We conducted a cross-sectional study in January-February 2024 of 3177 participants aged 5-92 years from Pakwach, Buliisa and Mayuge districts in the SchistoTrack cohort. Trained nurses who also had midwifery skills performed standardised liver and spleen palpation, assessing palpability, extension, consistency and tenderness. Palpation agreement was assessed against point-of-care B-mode ultrasound performed by sonographers with organ dimensions classified using height-standardised internal reference values. We estimated sensitivity, specificity, predictive values and Cohen's κ for palpation against ultrasound-defined liver and spleen enlargement and assessed associations between liver consistency and tenderness, and schistosomal periportal fibrosis examined from ultrasound. Results: Spleen palpation had high specificity (98.4%), with sensitivity ranging from 10.9% for moderate splenomegaly to 48.6% for severe splenomegaly. There was strong agreement with ultrasound for severe splenomegaly (κ=0.57) though weak agreement for moderate splenomegaly (κ=0.14). Liver palpation had high specificity (97.7%) but very low sensitivity (3.4-7.3%) and minimal agreement with any hepatomegaly found from ultrasound (κ=0.02-0.04). Among participants with a palpable liver, schistosomal periportal fibrosis was more frequent in those with firm/hard than soft liver consistency (53.3%, n=24/45 vs 8.6%, n=3/35; χ 2 =17.64, p<0.001); while tenderness was not associated with schistosomal periportal fibrosis. Nurses additionally identified clinically relevant findings, including umbilical collaterals, abdominal swelling, and abdominal or hypochondrial pain and tenderness. Conclusions: Standardised abdominal palpation cannot replace ultrasound for comprehensive detection or exclusion of hepatosplenomegaly. Nevertheless, severe ultrasound-defined splenomegaly was more readily detected by palpation, while liver consistency and other clinical findings provided information beyond organomegaly. In schistosomiasis-endemic settings with limited access to ultrasound, further research is needed to assess building capacity for abdominal palpation and how it can be combined with clinical information to increase clinical suspicion while possibly informing low-cost strategies for triage and referral.