Andrew I. Kim

Patients with acute liver failure (ALF) can be listed status I for liver transplantation (LT) whereas patients with cirrhosis must follow the MELD scoring system. Liver imaging can mistakenly diagnose submassive hepatic necrosis in ALF as cirrhosis. The purpose of our study was to assess the accuracy of ultrasound (US) and computed tomography (CT) in distinguishing cirrhosis from ALF. All patients listed for ALF and transplanted during the study period were included. Controls were age- and gender-matched cirrhotic patients who underwent LT during the same period. Abdominal US or CT scans obtained on all patients were independently reviewed by three blinded abdominal radiologists. Explants from all patients were reviewed by two blinded pathologists, and histological diagnosis was correlated with radiological diagnosis. Forty-one patients with ALF and 42 patients with cirrhosis were analyzed. Univariate and multivariate analyses both revealed overall accuracy of 85% for ultrasound and 93% for CT. US and CT scans both provide high levels of accuracy in terms of discriminating ALF from cirrhosis but measures taken to determine whether a patient has ALF vs. cirrhosis needs to approach 100% accuracy. Thus, imaging studies alone should not definitively diagnosis one etiology of liver failure over the other.

Object: In low- and middle-income countries, hydrocephalus is one of the most common congenital birth defects. Traditionally, shunt placement has been the major intervention for hydrocephalus. More recently, Endoscopic Third Ventriculostomy with Choroid Plexus Cauterization (ETV/CPC) was introduced as a low-cost, effective alternative. This study seeks to characterize hydrocephalus and its neurosurgical outcomes in Tanzania and to assess the feasibility of training a pediatric general surgeon in performing ETV/CPC.\nMethods: Part 1 of this combined prospective/retrospective cohort study characterizes hydrocephalus and its treatment in Arusha, Tanzania. Measured variables include etiologies of hydrocephalus, ages and demographics at treatment, severity of disease, numbers of patients, 30-day morbidity and mortality of shunt placement, and rates of shunt failure/infection (based on Kaplan-Meier method).\nIn part 2 of this study, a pediatric general surgeon from Tanzania was provided with ETV/CPC training. We allowed a 3-month latency period for her to become familiarized with the procedure. Then, we prospectively collected data from a consecutive cohort of 9 months of patients with hydrocephalus. Our main variables of interest were failure rate, failure reason, and 30-day mortality. We used the Kaplan-Meier method to determine if there was a significant difference in neurosurgical outcomes for hydrocephalus between VPS and ETV.\nResults:\nPart 1: Of 136 patients studied, average age of hydrocephalus onset was 2.98 months, average age of first illness was 1.81 months, and average age at treatment was 11.22 months. The most common etiologies were myelomeningocele (32 percent), congenital idiopathic (30 percent), post-infectious (26 percent), and encephalocele (4 percent). Overall, the VPS failure rate was 39.2 percent with an average time to failure of 8.18 months and average number of failures as 1.84. Most common reasons for failure included shunt malfunction (78 percent), shunt infection (16 percent), and wound dehiscence (6 percent). The 30-day mortality rate was 3.1 percent. Kaplan-Meier analyses showed no significant survival or VPS failure differences based on identified patient characteristics.\nPart 2: Of 32 patients evaluated for ETV, 23 ETV surgeries were attempted and 17 (73.9 percent) were successful. 6 were converted to VPS for reasons of technical issues (33.3 percent), scarred third ventricle floor (33.3 percent), and poor visualization (33.3 percent). The etiologies of hydrocephalus were 38 percent PIH and 62 percent NPIH. After a median follow-up of 57 days, the VPS failure rate was 37.5 percent and ETV failure rate was 5.9 percent (p=0.3192). The direction of revisions were as follows: prior VPS to new VPS (n=2), prior VPS to new ETV (n=1), and prior ETV to new VPS (n=1). There were no deaths during our follow-up period. Kaplan-Meier analyses showed no significant survival or VPS failure differences based on patient characteristics or surgery. There was a significantly shorter treatment delay in part 2 compared to part 1 (0.31 vs. 8.2 months, p= 0.0276).\n3\nConclusion: Overall, these analyses in Tanzania show that (1) etiologies of hydrocephalus are evenly split between myelomeningocele, congenital idiopathic, and post-infectious, (2) neurosurgical outcomes based on shunt placement showed high shunt failure rates within one year, most commonly due to shunt malfunction, and (3) early outcomes of ETV by a pediatric general surgeon were not significantly different from VPS outcomes, with results limited by small sample size. A determination of equivalence or superiority between these surgical approaches by a pediatric general surgeon requires further research