A machine learning tool for identifying metastatic colorectal cancer in primary care.
Background: Detection of colorectal cancer (CRC) is mainly achieved by clinical assessment. As new treatments become available for metastatic CRC (MCRC), it is important to accurately identify these patients. Aim: To develop a predictive model for identifying MCRC in primary health care patients usi...
| Publicado en: | Scandinavian Journal of Primary Health Care Vol. 43; no. 3; pp. 585 - 594 |
|---|---|
| Autores principales: | , , , , , , |
| Formato: | research tables/charts Journal Article |
| Publicado: |
Taylor & Francis Ltd
Sep2025
|
| Acceso en línea: | Ver este registro en EBSCOhost |
| fields | @attributes: recordID: 1 pdfLink: plink: https://search.ebscohost.com/login.aspx?direct=true&db=ccm&AN=187499140&site=ehost-live header: @attributes: shortDbName: ccm uiTerm: 187499140 longDbName: CINAHL Complete uiTag: AN controlInfo: bkinfo: dissinfo: jinfo: jid: 02813432 BDB jtl: Scandinavian Journal of Primary Health Care issn: 02813432 maglogo: Y pubinfo: dt: Sep2025 vid: 43 iid: 3 pid: 377 pub: Taylor & Francis Ltd place: Philadelphia, Pennsylvania artinfo: ui: 187499140 183629795 187499140 187499140 10.1080/02813432.2025.2477155 187499140 ppf: 585 ppct: 9 formats: fmt: @attributes: type: P tig: atl: A machine learning tool for identifying metastatic colorectal cancer in primary care. aug: au: Abedi, Eliya Ewing, Marcela Nemlander, Elinor Hasselström, Jan Sjövall, Annika Carlsson, Axel C. Rosenblad, Andreas affil: Division of Family Medicine and Primary Care, Department of Neurobiology, Care Sciences and Society, Karolinska Institutet, Solna, Sweden sug: subj: Colorectal Neoplasms Diagnosis Neoplasm Metastasis Diagnosis Data Analysis, Computer Assisted Machine Learning Primary Health Care Prediction Models Predictive Value of Tests Evaluation Human Algorithms Case Control Studies Sweden Odds Ratio Descriptive Statistics Sensitivity and Specificity Decision Making, Clinical Colorectal Neoplasms Risk Factors Neoplasm Metastasis Risk Factors Risk Assessment Funding Source ab: Background: Detection of colorectal cancer (CRC) is mainly achieved by clinical assessment. As new treatments become available for metastatic CRC (MCRC), it is important to accurately identify these patients. Aim: To develop a predictive model for identifying MCRC in primary health care patients using diagnostic data analysed with machine learning. Design and setting: A case-control study utilising data on primary health care visits for 146 patients >18 years old diagnosed with MCRC in the Västra Götaland Region, Sweden during 2011, and 577 sex-, age, and primary health care centre-matched controls. Method: Stochastic gradient boosting was used to construct a model for predicting the presence of MCRC based on diagnostic codes from primary health care consultations during the year before index (diagnosis) date and number of consultations. Variable importance was estimated using the normalised relative influence (NRI) score. Risks of having MCRC were calculated using odds ratios of marginal effects (ORME). Results: The optimal model included 76 variables with non-zero influence, had an area under the curve of 76.5%, a sensitivity of 77.8%, and a specificity of 69.2%. The 10 most important variables had a combined NRI of 61.0%. Number of consultations during the year before index date had the highest NRI at 19.2%, with an ORME of 3.3. Conclusion: A machine learning method based on primary health care consultation frequency and diagnoses may be used to identify important variables for predicting presence of MCRC. Both primary health care consultations and associated diagnostic codes need to be taken into consideration. A study in Sweden developed a machine learning model to predict metastatic colorectal cancer (MCRC) in primary health care patients. The model used diagnostic codes and consultation data, achieving a sensitivity of 77.8% and specificity of 69.2%. The number of consultations with a general practitioner emerged as the most important variable. Consideration of both consultation frequency and diagnostic codes is crucial for MCRC detection. CLINICAL PRACTICE POINTS: Colorectal cancer (CRC) is a significant global health concern, with higher mortality rates for patients with metastatic CRC (MCRC) compared to non-metastatic CRC (NMCRC). While screening programs help detect CRC at earlier stages, clinical assessment remains the primary method for diagnosis. This study aimed to develop a predictive model using machine learning to identify MCRC in primary healthcare patients. The model analysed diagnostic data and consultation frequency from electronic medical records. The machine learning model achieved a sensitivity of 77.8% and specificity of 69.2% in predicting MCRC. Key variables influencing MCRC prediction included the number of consultations, abdominal and pelvic pain, other anaemias, and senile cataract. Interestingly, hypertension, dorsalgia, and other joint disorders indicated a decreased risk of MCRC. The study highlights the potential of machine learning-based predictive models to aid in early detection and risk assessment of MCRC in primary health care. Successful implementation of such models could lead to improved clinical decision-making and contribute to better patient outcomes in the foreseeable future, but first further development, external validation, and testing to confirm model accuracy is needed. POINT OF INTEREST: We used artificial intelligence to predict the presence of metastatic colorectal cancer, in a case-control study based on all diagnostic codes in primary health care. Both number of consultations and associated diagnostic codes need to be considered. The prediction tool had a sensitivity of 77.8% and a specificity of 69.2%. pubtype: Academic Journal doctype: research tables/charts Journal Article ougenre: Article language: English refInfo: holdings: @attributes: islocal: N |
|---|