Machine learning and natural language processing (NLP) approach to predict early progression to first-line treatment in real-world hormone receptor-positive (HR+)/HER2-negative advanced breast cancer patients.

BACKGROUND: CDK4/6 inhibitors plus endocrine therapies are the current standard of care in the first-line treatment of HR+/HER2-negative metastatic breast cancer, but there are no well-established clinical or molecular predictive factors for patient response. In the era of personalised oncology, new approaches for developing predictive models of response are needed. MATERIALS AND METHODS: Data derived from the electronic health records (EHRs) of real-world patients with HR+/HER2-negative advanced breast cancer were used to develop predictive models for early and late progression to first-line treatment. Two machine learning approaches were used: a classic approach using a data set of manually extracted features from reviewed (EHR) patients, and a second approach using natural language processing (NLP) of free-text clinical notes recorded during medical visits. RESULTS: Of the 610 patients included, there were 473 (77.5%) progressions to first-line treatment, of which 126 (20.6%) occurred within the first 6 months. There were 152 patients (24.9%) who showed no disease progression before 28 months from the onset of first-line treatment. The best predictive model for early progression using the manually extracted dataset achieved an area under the curve (AUC) of 0.734 (95% CI 0.687-0.782). Using the NLP free-text processing approach, the best model obtained an AUC of 0.758 (95% CI 0.714-0.800). The best model to predict long responders using manually extracted data obtained an AUC of 0.669 (95% CI 0.608-0.730). With NLP free-text processing, the best model attained an AUC of 0.752 (95% CI 0.705-0.799). CONCLUSIONS: Using machine learning methods, we developed predictive models for early and late progression to first-line treatment of HR+/HER2-negative metastatic breast cancer, also finding that NLP-based machine learning models are slightly better than predictive models based on manually obtained data.

Factors associated with mortality among elderly people hospitalized due to femoral fractures.

OBJECTIVE: To analyze factors associated with mortality among elderly people hospitalized in a single-center regional hospital due to femoral fractures. METHODS: This was a retrospective cohort study. Patients aged 60 years or over who were hospitalized with a diagnosis of femoral fracture (ICD S72) between 2008 and 2013 were selected through the electronic medical records. RESULTS: The study evaluated 195 individuals of mean age 78.5 ± 9.6 years; females predominated (68.2%). The main mechanism for falls was low-energy (87.2%). Surgery was performed on 93.3% of the patients; the mean length of hospital stay was 13.6 ± 7.5 days and the mean waiting time for the surgery was 7.7 ± 4.2 days. The prevalence of mortality was 14.4%, and this occurred mostly among older individuals (p = 0.029); patients with leukocytosis (p < 0.001); those who needed intensive care (p < 0.001); and those who did not undergo surgery (p < 0.001). The mean survival was significantly longer among patients who underwent surgery and shorter among those who needed intensive care. CONCLUSION: Women predominated among the hospitalizations, and the degree of leukocytosis associated with advanced age presented a relationship with mortality, independent of the type of lesion or surgical procedure. More studies still need to be conducted in order to assess other factors associated with mortality.

OBJETIVO: Analisar os fatores associados à mortalidade em idosos hospitalizados por fratura de fêmur em um hospital unicêntrico regional. MÉTODOS: Estudo de coorte retrospectiva. Foram selecionados, por meio do prontuário eletrônico, pacientes internados com diagnóstico de fratura de fêmur (CID S72) com 60 anos ou mais de 2008 a 2013. RESULTADOS: Foram avaliados 195 indivíduos com idade média de 78,5 ± 9,6 e o gênero feminino foi mais prevalente (68,2%). O principal mecanismo de queda foi o de baixa energia (87,2%), a feitura de cirurgia foi de 93,3%, o tempo de internação médio foi de 13,6 ± 7,5 dias, o tempo de espera para a cirurgia médio foi de 7,7 ± 4,2 dias. A prevalência de mortalidade foi de 14,4%, ocorreu principalmente nos indivíduos mais idosos (p = 0,029), com leucocitose (p < 0,001), com necessidade de cuidados intensivos (p < 0,001) e que não foram submetidos a cirurgia (p < 0,001). A sobrevida média foi significativamente maior nos pacientes submetidos a cirurgia e inversamente nos pacientes que necessitaram da unidade de terapia intensiva. CONCLUSÃO: As mulheres predominaram nas internações e o grau de leucocitose associado a idade avançada apresentou relação com a mortalidade, independentemente do tipo de lesão e procedimento cirúrgico. Ainda devem ser feitos mais estudos para avaliar outros fatores associados à mortalidade.

Factors associated with mortality among elderly people hospitalized due to femoral fractures / Fatores associados à mortalidade em idosos hospitalizados por fraturas de fêmur

ABSTRACT OBJECTIVE: To analyze factors associated with mortality among elderly people hospitalized in a single-center regional hospital due to femoral fractures. METHODS: This was a retrospective cohort study. Patients aged 60 years or over who were hospitalized with a diagnosis of femoral fracture (ICD S72) between 2008 and 2013 were selected through the electronic medical records. RESULTS: The study evaluated 195 individuals of mean age 78.5 ± 9.6 years; females predominated (68.2%). The main mechanism for falls was low-energy (87.2%). Surgery was performed on 93.3% of the patients; the mean length of hospital stay was 13.6 ± 7.5 days and the mean waiting time for the surgery was 7.7 ± 4.2 days. The prevalence of mortality was 14.4%, and this occurred mostly among older individuals (p = 0.029); patients with leukocytosis (p < 0.001); those who needed intensive care (p < 0.001); and those who did not undergo surgery (p < 0.001). The mean survival was significantly longer among patients who underwent surgery and shorter among those who needed intensive care. CONCLUSION: Women predominated among the hospitalizations, and the degree of leukocytosis associated with advanced age presented a relationship with mortality, independent of the type of lesion or surgical procedure. More studies still need to be conducted in order to assess other factors associated with mortality.

RESUMO OBJETIVO: Analisar os fatores associados à mortalidade em idosos hospitalizados por fratura de fêmur em um hospital unicêntrico regional. MÉTODOS: Estudo de coorte retrospectiva. Foram selecionados, por meio do prontuário eletrônico, pacientes internados com diagnóstico de fratura de fêmur (CID S72) com 60 anos ou mais de 2008 a 2013. RESULTADOS: Foram avaliados 195 indivíduos com idade média de 78,5 ± 9,6 e o gênero feminino foi mais prevalente (68,2%). O principal mecanismo de queda foi o de baixa energia (87,2%), a feitura de cirurgia foi de 93,3%, o tempo de internação médio foi de 13,6 ± 7,5 dias, o tempo de espera para a cirurgia médio foi de 7,7 ± 4,2 dias. A prevalência de mortalidade foi de 14,4%, ocorreu principalmente nos indivíduos mais idosos (p = 0,029), com leucocitose (p < 0,001), com necessidade de cuidados intensivos (p < 0,001) e que não foram submetidos a cirurgia (p < 0,001). A sobrevida média foi significativamente maior nos pacientes submetidos a cirurgia e inversamente nos pacientes que necessitaram da unidade de terapia intensiva. CONCLUSÃO: As mulheres predominaram nas internações e o grau de leucocitose associado a idade avançada apresentou relação com a mortalidade, independentemente do tipo de lesão e procedimento cirúrgico. Ainda devem ser feitos mais estudos para avaliar outros fatores associados à mortalidade.

Radiographic evaluation of complete and incomplete discoid lateral meniscus.

OBJECTIVES: The aim of this retrospective study was (1) to evaluate the radiographic features to differentiate arthroscopically confirmed complete and incomplete discoid lateral meniscus (DLM) (2) to determine the cutoff values for any parameter that was found to differentiate complete from incomplete DLM. MATERIALS AND METHODS: We retrospectively analyzed plain knee radiographs of 130 arthroscopically proven DLM. Seventy-nine patients had complete DLM and 51 patients incomplete DLM. Knee radiographs from 52 patients with arthroscopically proven normal lateral meniscus acted as control group. Radiographic parameters measured included fibular height, lateral joint space, condylar cutoff sign, height of lateral tibial spine, obliquity of lateral tibial spine, squaring of the lateral femoral condyle, and cupping of the lateral tibial plateau. RESULTS: Among radiographic parameters, high fibular head, widening of the lateral joint space and femoral condylar cutoff sign showed statistically significant (p<0.0001) differences between complete and incomplete DLM. At specific threshold points of fibular height<11 mm, lateral joint space>6 mm and condylar cutoff sign<0.80, the diagnosis of complete DLM revealed 87.3% sensitivity, 81.6% specificity and 78.4% positive predictive value (PPV) for the fibular height, 81.0% sensitivity, 86.6% specificity and 83.1% PPV for the lateral joint space, and 86.1% sensitivity, 83.5% specificity and 80% PPV for the condylar cutoff sign. CONCLUSIONS: Radiographic features of fibular height, lateral joint space and condylar cut off sign can be used for screening of a complete type of DLM. However, radiographs are not a reliable screening tool for an incomplete DLM. LEVEL OF EVIDENCE: IV, Case Series.