1
|
Mendoza-Urbano DM, Garcia JF, Moreno JS, Bravo-Ocaña JC, Riascos AJ, Zambrano Harvey A, Prada SI. Automated extraction of information from free text of Spanish oncology pathology reports. Colomb Med (Cali) 2023; 54:e2035300. [PMID: 37614525 PMCID: PMC10443791 DOI: 10.25100/cm.v54i1.5300] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 06/10/2022] [Revised: 08/02/2022] [Accepted: 09/20/2022] [Indexed: 08/25/2023] Open
Abstract
Background Pathology reports are stored as unstructured, ungrammatical, fragmented, and abbreviated free text with linguistic variability among pathologists. For this reason, tumor information extraction requires a significant human effort. Recording data in an efficient and high-quality format is essential in implementing and establishing a hospital-based-cancer registry. Objective This study aimed to describe implementing a natural language processing algorithm for oncology pathology reports. Methods An algorithm was developed to process oncology pathology reports in Spanish to extract 20 medical descriptors. The approach is based on the successive coincidence of regular expressions. Results The validation was performed with 140 pathological reports. The topography identification was performed manually by humans and the algorithm in all reports. The human identified morphology in 138 reports and by the algorithm in 137. The average fuzzy matching score was 68.3 for Topography and 89.5 for Morphology. Conclusions A preliminary algorithm validation against human extraction was performed over a small set of reports with satisfactory results. This shows that a regular-expression approach can accurately and precisely extract multiple specimen attributes from free-text Spanish pathology reports. Additionally, we developed a website to facilitate collaborative validation at a larger scale which may be helpful for future research on the subject.
Collapse
Affiliation(s)
| | | | - Juan Sebastian Moreno
- Quantil SAS. Bogotá, Colombia
- Centro de Analítica para Políticas Públicas. Bogotá, Colombia
| | | | - Alvaro José Riascos
- Quantil SAS. Bogotá, Colombia
- Centro de Analítica para Políticas Públicas. Bogotá, Colombia
- Universidad de los Andes, Facultad de Economía. Bogotá, Colombia
| | | | - Sergio I Prada
- Fundación Valle del Lili, Centro de Investigaciones Clínicas, Cali, Colombia
- Universidad Icesi, Centro PROESA, Cali, Colombia
| |
Collapse
|
2
|
Parra-Lara LG, Mendoza-Urbano DM, Zambrano ÁR, Valencia-Orozco A, Bravo-Ocaña JC, Bravo-Ocaña LE, Rosso F. Methods and implementation of a Hospital-Based Cancer Registry in a major city in a low-to middle-income country: the case of Cali, Colombia. Cancer Causes Control 2022; 33:381-392. [PMID: 35013914 DOI: 10.1007/s10552-021-01532-z] [Citation(s) in RCA: 6] [Impact Index Per Article: 3.0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 03/16/2020] [Accepted: 11/30/2021] [Indexed: 11/29/2022]
Abstract
PURPOSE To describe our experience upon developing and implementing a hospital-based cancer registry (HBCR) in a quaternary-level of care private non-profit academic medical center in Cali, Colombia. METHODS HBCRs capture, in a given institution, every single patient with a confirmed malignancy. In this study, all cases evaluated between 2014 and 2018 were included in the HBCR. In compliance with the International Agency for Research on Cancer recommendations, cases were classified as analytic or non-analytic. Data derived from an exhaustive selection of patients was stored in a computing platform owned by the institution, meeting the 2016 Facility Oncology Registry Data Standards recommendations. Quality control was performed by evaluating comparability, timeliness, validity, and completeness. RESULTS A total of 24,405 new cases were registered between 2014 and 2018, from which 4253 (17.4%) died. Among all cases, based on the anatomic location, most common malignancies were breast (n = 1554), thyroid (n = 1346), hematolymphoid (n = 1251), prostatic (n = 805), and colorectal (n = 624). The behavior of the new cases was consistent with an incremental trend. CONCLUSION Upon implementing the HBCR, major challenges were identified (i.e., a precise definition of cases, the development of processes for capturing new cases, a standardized data collection strategy, and carrying-out an appropriate patient follow-up). Based on our experience, the success of an HBCR largely relies on the interest from the institution, the engagement of stakeholders and financial support, that is, it depends on the adequate access over time to funding, technological, and staffing resources.
Collapse
Affiliation(s)
- Luis G Parra-Lara
- Centro de Investigaciones Clínicas (CIC), Fundación Valle del Lili, Cra 98 No. 18-49, 760032, Cali, Colombia. .,Facultad de Ciencias de la Salud, Universidad Icesi, Calle 18 No. 122-135, 760031, Cali, Colombia.
| | - Diana M Mendoza-Urbano
- Centro de Investigaciones Clínicas (CIC), Fundación Valle del Lili, Cra 98 No. 18-49, 760032, Cali, Colombia
| | - Ángela R Zambrano
- Servicio de Hemato-Oncología, Departamento de Medicina Interna, Fundación Valle del Lili, Cra 98 No. 18-49, 760032, Cali, Colombia
| | - Andrea Valencia-Orozco
- Centro de Investigaciones Clínicas (CIC), Fundación Valle del Lili, Cra 98 No. 18-49, 760032, Cali, Colombia
| | - Juan C Bravo-Ocaña
- Departamento de Patología y Laboratorio Clínico, Fundación Valle del Lili, Cra 98 No. 18-49, 760032, Cali, Colombia
| | - Luis E Bravo-Ocaña
- Registro Poblacional de Cáncer de Cali, Departamento de Patología, Facultad de Salud, Universidad del Valle, 760032, Cali, Colombia
| | - Fernando Rosso
- Centro de Investigaciones Clínicas (CIC), Fundación Valle del Lili, Cra 98 No. 18-49, 760032, Cali, Colombia.,Facultad de Ciencias de la Salud, Universidad Icesi, Calle 18 No. 122-135, 760031, Cali, Colombia.,Servicio de Infectología, Departamento de Medicina Interna, Fundación Valle del Lili, Cra 98 No. 18-49, 760032, Cali, Colombia
| |
Collapse
|
3
|
Meel M, Choudhary N, Kumar M, Mathur K. Epidemiological Profiling and Trends of Primary Intracranial Tumors: A Hospital-Based Brain Tumor Registry from a Tertiary Care Center. J Neurosci Rural Pract 2021; 12:145-152. [PMID: 33531774 PMCID: PMC7846331 DOI: 10.1055/s-0040-1721622] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 11/29/2022] Open
Abstract
Background and Objectives Hospital-based cancer registry is an essential tool for augmentation of the standard of care, administration motive, and resource for population-based cancer registries. Here, we presented hospital-based brain tumor registry (HBBTR) to outline a comprehensive epidemiological data, both clinical and histopathological, as well as trends of central nervous system tumors. In addition, we compare this data with national brain tumor data as well as an international brain tumor registry. Materials and Methods For the generation of this 7-year HBBTR data of all primary intracranial tumors operated, diagnosed, and registered at the Department of Pathology, Sawai ManSingh, between January 1, 2013 and December 31, 2019, was collected, analyzed, and compared with Tata Memorial Hospital, National Institute of Mental Health and Neurosciences, and Central Brain Tumor Registry of the United States. Results A total of 3,526 patients were of primary intracranial tumors. Out of which, male patients were 1,982 (56.2%), while 1,544 (43.8%) were female patients. Maximum proportion of tumors was in fifth decade. Overall, pediatric and adult patients constituted of 15.5 and 84.5% of the cases, respectively. Among all primary intracranial tumors, meningiomas (20%) were most common followed by glioblastoma multiformat (18%) and least common were germ cell tumors (0.1%) followed by pineal tumors (0.3%). In pediatric cohort astrocytic tumors (30.1%) are most common followed by embryonal tumors (20.8%), while in adults meningiomas (23.1%) were most common followed by glioblastomas (20.3%). Our registry showed similar trends of tumors with national data as compared with international data in median age of presentation. Conclusion This HBBTRs provide prevalence of primary intracranial tumors at a tertiary care center and could be a part of population-based registry.
Collapse
Affiliation(s)
- Mukta Meel
- Department of Pathology, SMS Medical College, Jaipur, Rajasthan, India
| | - Nikita Choudhary
- Department of Pathology, SMS Medical College, Jaipur, Rajasthan, India
| | - Mukesh Kumar
- Department of Neurosurgery, SMS Medical College, Jaipur, Rajasthan, India
| | - Kusum Mathur
- Department of Pathology, SMS Medical College, Jaipur, Rajasthan, India
| |
Collapse
|
4
|
Gondhowiardjo S, Ekaputra E, Randi A, Jayalie VF. The challenge of the implementation and evaluation of hospital-based cancer registry in Indonesia’s national referral hospital. MEDICAL JOURNAL OF INDONESIA 2020. [DOI: 10.13181/mji.bc.203785] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 10/22/2022] Open
Abstract
BACKGROUND To lower the burden caused by cancer, the Ministry of Health of the republic of Indonesia requires valid data collection to plan and evaluate cancer programs. This study aimed to evaluate the challenge of developing a cancer registry (CanReg) in Cipto Mangunkusumo Hospital.
METHODS This was an observational study on the implementation of cancer registration from the initial licensing until the creation of valid and accurate data, also the challenges in implementing hospital-based cancer registry (HBCR) in Cipto Mangunkusumo Hospital.
RESULTS Cancer registry was developed in 2016 using the 2013 Indonesian version of CanReg5 program called SriKandI. We identified some problems in this registry implementation, such as legal and human resources, medical records, electronic health records, and the CanReg5 program. Moreover, this team processed 886,086 raw patients’ data with fairly good topography and age data completeness.
CONCLUSIONS Several obstacles were encountered upon the establishment of HBCR at Cipto Mangunkusumo Hospital from program to bureaucracy and resources. Nevertheless, CanReg data can be used as a basis for decision making by stakeholders.
Collapse
|
5
|
MacCallum C, Skandarajah A, Gibbs P, Hayes I. The Value of Clinical Colorectal Cancer Registries in Colorectal Cancer Research: A Systematic Review. JAMA Surg 2019; 153:841-849. [PMID: 29926104 DOI: 10.1001/jamasurg.2018.1635] [Citation(s) in RCA: 16] [Impact Index Per Article: 3.2] [Reference Citation Analysis] [Abstract] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 01/06/2023]
Abstract
Importance Clinical colorectal cancer registries (CCCRs) are potentially powerful tools in colorectal cancer research. They are resource intensive, but to our knowledge, no formal review of their value exists. While quality control, clinical audit, and benchmarking are important factors in assessing the value of maintaining CCCRs, they are difficult to quantify. This study focuses on registry research output as a measure of value; the study hypothesizes that CCCRs do not produce sufficient published research output of clinical significance to justify the resources required to maintain them. Objective To assess the value of maintaining CCCRs by identifying and characterizing existing CCCRs and measuring their comparative research impact. Evidence Review We searched MEDLINE (PubMed) and Google Scholar for articles published from January 1990 to July 2016 that identified multi-institutional CCCRs with peer-reviewed published outcomes. Purely population-based registries were excluded. We then searched the same databases in the same time period for articles that were published by each included CCCR. The articles must have been based on outcomes relating to individual CCCR data. We categorized published outcomes into oncological, surgical, or other outcomes. We measured the research impact of each CCCR using the number of articles, citation index, impact factor, and Altmetric score. Findings A total of 18 CCCRs were identified, with sample sizes between 104 and 1 400 000 cases. Data fields, published aims, and outcomes were similar between registries. The most frequently published outcomes related to anastomotic leak following colorectal surgery. The National Cancer Database formed the basis of the highest number of publications (66), the Northern Region Colorectal Cancer Audit Group had the highest median article citation number (28.5), the National Bowel Cancer Audit had the highest median impact factor (4.72), and the National Cancer Database had the highest median Altmetric score (4.5). Conclusions and Relevance There is a significant body of colorectal cancer outcomes research generated from the CCCRs. However, given the enormous resources required, the overall research output and impact of CCCRs is low in proportion to the size of the data sets. These registries hold key oncological and surgical outcomes data; focusing on data linkage between registries and developing automated data collection will enable international comparisons in colorectal cancer management and will increase the research impact of CCCRs, thereby increasing their value.
Collapse
Affiliation(s)
- Caroline MacCallum
- Department of Surgery, University of Melbourne, Melbourne, Victoria, Australia
| | - Anita Skandarajah
- Department of Surgery, University of Melbourne, Melbourne, Victoria, Australia
| | - Peter Gibbs
- Peter MacCallum Cancer Centre, Melbourne, Victoria, Australia
| | - Ian Hayes
- Colorectal Surgery Unit, Department of General Surgical Specialties, Royal Melbourne Hospital, Melbourne, Victoria, Australia
| |
Collapse
|