No abstract

EP-206Introduction: State sector hospitals in Sri Lanka lack electronic database management systems. The database at the HPB unit at Colombo South Teaching Hospital was based on a rudimentary Google-Sheet that wasn't maintenance-friendly, prone to inconsistencies, and lacked data retrievability for analysis purposes. Methods: Using cloud-based Google services and AppSheet, a multiaccess mobile app was developed to store HPB data. The author spent 2 months studying the web platform to create a password-protected app. Consent was obtained from patients to maintain clinical data through the app using mobile devices of the HPB team members. After 25 months of use of the app, this abstract analyses the overall data. Results: The app can record 254 data variables per patient, of which 222 are analyzable. The database has so far 1,561 patients referred to and managed at the HPB unit since November 2018 in which 566 liver (M:F 2:1), 578 pancreatic (M:F 1.5:1), and 417 biliary pathologies (M:F 1:1.2) have been diagnosed. 857 have malignant pathologies, 523 have benign pathologies. 455 had a conclusive surgical management decision, 477 had a nonsurgical management decision. 74% (n = 420) of liver, 37% (n = 214) of pancreatic, and 53% (n = 222) of biliary patients had cancers. Data on 239 HCC, 64 CRLM, 137 pancreatic adenocarcinomas, 16 pancreatic cystic neoplasms, 141 cholangiocarcinoma, and 43 gallbladder cancers are available in the database. In whom a management decision was reached, 31% of liver, 39% of pancreatic, and 49% of biliary patients were operated on. Conclusions: This cost-free user-friendly solution can revolutionize database management for a low-income country.

Due to the rarity of the disease, there are no guidelines or consensus for the management of solid pseudopapillary neoplasm of the pancreas with liver metastasis. We describe a patient with solid pseudopapillary neoplasm of the pancreas with multiple liver metastases who was successfully managed with staged resection. A previously healthy 50-year-old female presented to the clinic with dull epigastric pain. The contrast-enhanced computed tomography showed a well-defined cystic lesion at the distal body and tail of the pancreas with multiple liver deposits. A core needle liver biopsy was suspicious of a hepatic epithelioid hemangioendothelioma. A distal pancreatectomy with splenectomy and a left hemi-hepatectomy were performed with a subsequent staged right posterior sectionectomy to address the inadequate future liver remnant. The final histology with immunohistochemistry revealed a pancreatic solid pseudopapillary neoplasm. The patient had satisfactory surgical and oncological outcomes with no recurrence. As these are slow growing tumours, appropriately spaced staged liver resection was feasible to allow hypertrophy of the liver remnant and physiological recovery of the patient. The discrepancy in the histology and cross-sectional imaging led to a pre-operative diagnostic dilemma, and the final diagnosis was reached after surgical resection. Thus, resectable metastatic solid pseudopapillary neoplasm of the pancreas may be offered a primary curative radical surgery before considering other treatment modalities.

Introduction: Maintaining robust clinical data in a low-income country is a challenge due to the lack of electronic medical record systems in state sector hospitals. To mitigate this, the author created a cloud-based cost-free app for the HPB unit of Colombo South Teaching Hospital. Methods: Analyzable data variables pertaining to HPB patients were prospectively gathered over a period of 25 months using the app. The app's on-the-go analysis section was used to extract liver-related data. Results: 36% (n = 566) of the HPB data were liver-related. 66% were males. Mean-age was 65. 75% of liver lesions were malignancies; 57% (n = 242) HCC, 16% (n = 66) colorectal liver metastasis (CRLM), 5% (n = 21) non-colorectal liver metastasis (NCRLM). 45% of HCC belonged to 61 to 70 age group. Male:female of HCC 4.8:1, CRLM 1.2:1, and NCRLM 0.9:1. 51% of HCC were solitary-lesions, 30% multiple-lesions. 25% of CRLM were solitary-lesions, 43% multiple-lesions. 53% of HCC were confined to right-lobe, 31% to leftlobe. CRLM 58% in right-lobe, 37% in both lobes. 64% of HCC were discrete-HCC, 30% were diffuse-HCC. Of those with conclusive decisions, 38% (n = 130) of liver-lesions surgically, and 62% (n = 214) non-surgically were managed. Lesions were resectable in 31% of HCC, 74% of CRLM, and 80% of NCRLM. 43% (n = 12) of resectable-HCC underwent hemihepatectomy, 36% (n = 10) NALR, 11% trisectionectomy. For CRLM, this was 60% (n = 15) NALR, 36% (n = 9) hemihepatectomy. 54% (n = 58) of inoperable-HCC had locoregional-therapy; 19%-ablation (72%-thermal, 28%-chemical) and 81%-TACE. 46% of inoperable-HCC were palliated. Conclusions: This form of mobile-based clinical data maintenance has improved the data completion rates and accuracy of HPB data.