KFKiyomitsu Fukaguchi

Kiyomitsu Fukaguchi

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article

Previous Organizations / Institutes

The organizations / institutes from the scholar's other articles in the past ten years


Kiyomitsu Fukaguchi has a total of 68 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Cytomegalovirus Pneumonia Requiring Intensive Care During Treatment for Rapidly Progressive Interstitial Lung Disease Associated With Anti‐Synthetase Syndrome: A Case Report

    by Hiroe Aramaki, Kiyomitsu Fukaguchi, Sachi Miyashita, Yuki Yao, Mizuki Sato, Shingo Ohki, Hiroshi Koyama, Satoshi Noma, Tomoya Fukui

    A 70‐year‐old man presented with fever and malaise and was diagnosed with anti‐synthetase syndrome based on chest computed tomography and laboratory findings. Steroid therapy was initiated because of progressive respiratory failure. However, despite steroid therapy, his respiratory status worsened, accompanied by expansion of ground‐glass opacities (GGOs) on imaging. On hospital day 26, the patient developed severe respiratory failure, requiring mechanical ventilation and prone positioning, at which time cytomegalovirus (CMV) antigenemia was detected. After clinically diagnosing CMV pneumonia superimposed on interstitial lung disease (ILD), ganciclovir was initiated while gradually tapering corticosteroid therapy. This approach led to significant clinical and radiological improvements. The patient was successfully weaned from ventilation on day 30 and discharged on day 70. In patients with rapidly progressive ILD who deteriorate despite high‐dose corticosteroids and develop new diffuse GGOs, superimposed CMV infection should be suspected and promptly evaluated. Early recognition and antiviral therapy are essential for improving outcomes.

  • Role of pathological autopsy in determining the cause of death in cases of unexpected clinical deterioration or medical incidents

    by Sachi Miyashita, Tadashi Kamio, Hiroshi Koyama, Shingo Ohki, Yumiko Tsunano, Mizuki Sato, Kiyomitsu Fukaguchi, Hiroki Hadano

    Background: The extent to which autopsies are conducted and their role in determining the cause of death has not been sufficiently examined. We aimed to evaluate the contribution of autopsy in determining the cause of death in cases of unexpected clinical deterioration or medical accidents, using data from the Japan Center for Quality Health Care database. Methods: We analysed data from patients with unexpected clinical deterioration who required pathological autopsy between January 2010 and November 2024. Cases were categorised based on whether autopsies were performed, and the reasons for not performing autopsies were examined. Results: In total, 682 reports were identified. Pathological autopsies were performed in 341 patients, and the cause of death was determined in 64.2% of cases (219 patients). The cause of death remained undetermined in 25.2% of cases (86 patients) despite autopsy. Among the 341 patients in whom autopsies were not performed, the primary reason was the family refusal (81.5%, 278 patients). Conclusions: Pathological autopsies are essential for determining the cause of death in cases of unexpected clinical deterioration. Family refusal was identified as the principal barrier to autopsy, whereas lack of autopsy proposal accounted for only a small proportion of cases.


  • Open Access

    Association between nafamostat mesylate infusion dose and dialysis circuit occlusion in continuous renal replacement therapy: a retrospective cohort study

    by Naoto Eriguchi, Kiyomitsu Fukaguchi, Mochida Yasuhiro, Hiroshi Koyama

    Background: Circuit occlusion during treatment is detrimental in patients undergoing continuous renal replacement therapy (CRRT), leading to treatment interruption, reduced clearance efficiency, and increased healthcare costs. Nafamostat mesylate (NM) is an anticoagulant commonly used during CRRT, particularly in patients with a high risk of bleeding, owing to its short half-life and reversible anticoagulant effects. However, its optimal dose range remains unclear. We aimed to investigate whether differences in the hourly dose of NM are associated with differences in the circuit occlusion rate of CRRT in patients who are critically ill. Methods: We conducted a retrospective cohort study in the intensive care unit of Shonan Kamakura General Hospital between April 2022 and June 2024. Adult patients requiring CRRT with NM as the sole anticoagulant were included in this study. The primary outcome was the ratio of circuit occlusions within 24 h. A multivariate logistic regression analysis was performed to account for confounding clinical factors. The secondary outcome was the occurrence of bleeding events until hospital discharge. We also performed subgroup analyses and stratified and compared the patient- and device-side factors thought to affect circuit occlusion. Results: We included 70 patients (mean age, 72 years; 44 males [62.9%]). The median NM dose was 30 mg/h (interquartile range [IQR]: 30–40 mg/h). There was no significant association between NM dose and occlusion (odds ratios [OR] 1.39; 95% confidence interval [CI] 0.71–2.89) or bleeding events (OR 1.39; 95% CI 0.55–4.35). Subgroup analysis suggested that the association between a higher NM dose and circuit occlusion events was generally consistent across strata, although some subgroups (such as patients with higher sequential organ failure assessment scores, bleeding risk, sepsis, or different filter types) showed possible heterogeneity in the effect estimates. However, no statistically significant interactions were observed. Conclusions: In this study, the continuous NM infusion dose was not significantly associated with circuit occlusion in CRRT. Future randomized trials comparing different NM doses and alternative anticoagulants are needed to define the optimal anticoagulation strategies.

  • Open Access

    Association between pet ownership and physical function at discharge in hospitalized older adults: A retrospective observational study

    by Yo Ishihara, Kiyomitsu Fukaguchi, Hiroshi Koyama

    Objectives: Hospitalization can lead to a loss of physical function among older adults. While pet ownership has been reported to have beneficial effects on physical function, its impact on hospitalized patients remains unclear. We examined the association between pet ownership upon admission and physical function during hospitalization. Methods: This single-center, retrospective, observational study was conducted between April 2013 and April 2023. Patients aged ≥65 years who were hospitalized for the first time due to pneumonia at our facility were included. Cases were identified through medical record searches using the keywords “pet”, “dog”, “cat”, “ownership”, and “keeping”. Patients with unclear pet ownership status in the medical records were excluded from the analysis. Patients were classified into two groups: Pet Owners (PO) and Non-Pet Owners (NPO) upon admission. The Barthel index (BI) gain (the difference between admission and discharge BI scores) and BI efficiency (BI gain divided by hospital stay) were compared. Logistic regression was performed for patients with a BI gain >0 and discharge BI scores >85, indicating independence without functional decline, after adjusting for covariates. Results: A total of 248 patients were initially screened, we finally included 172 patients (69.4%) in the PO group and 46 patients (18.5%) in the NPO group. The median BI gain was 10 (0–30) and 2.5 (0.0–26.3) in the PO and NPO groups (p = 0.24), respectively. The median BI efficiency was 0.48 (0.00–1.87) and 0.16 (0.00–1.27) in the PO and NPO groups (p = 0.062), respectively. Current pet ownership was not associated with independence without functional decline, with an adjusted odds ratio of 0.68 (95% confidence interval, 0.15–3.94; p = 0.63). Conclusions: This study found no association between current pet ownership and the maintenance of physical function in hospitalized older adults.

  • Open Access

    Analysis of aPTT predictors after unfractionated heparin administration in intensive care units using machine learning models

    by Tadashi Kamio, Masaru Ikegami, Megumi Mizuno, Seiichiro Ishii, Hayato Tajima, Yoshihito Machida, Kiyomitsu Fukaguchi

    Objectives: Predicting optimal coagulation control using heparin in intensive care units (ICUs) remains a significant challenge. This study aimed to develop a machine learning (ML) model to predict activated partial thromboplastin time (aPTT) in ICU patients receiving unfractionated heparin for anticoagulation and to identify key predictive factors. Methods: Data were obtained from the Tokushukai Medical Database, covering six hospitals with ICUs in Japan, collected between 2018 and 2022. The study included 945 ICU patients who received unfractionated heparin. The dataset comprised both static and dynamic features, which were used to construct and train ML models. Models were developed to predict aPTT following initial and multiple heparin doses. Model performance was evaluated using the area under the receiver operating characteristic curve (ROC AUC), area under the precision–recall curve (PR AUC), precision, recall, F1 score, and accuracy. SHAP analysis was conducted to determine key predictive factors. Results: The random forest model demonstrated the highest predictive performance, with ROC AUC values of 0.707 for the first infusion and 0.732 for multiple infusions. Corresponding PR AUC values were 0.539 and 0.551. Despite moderate overall predictive performance, the model exhibited high precision (0.585 for the first infusion and 0.589 for multiple infusions), indicating effectiveness in correctly identifying true positive cases. However, recall and F1 scores were lower, suggesting that some cases, particularly in sub-therapeutic and supra-therapeutic ranges, may have been missed. Incorporating time-series data, such as vital signs, provided only marginal improvements in performance. Conclusions: ML models demonstrated moderate performance in predicting aPTT following heparin infusion in ICU patients, with the random forest model achieving the highest classification accuracy. Although the models effectively identified true positive cases, their overall predictive performance remained limited, necessitating further refinement. The inclusion of static and dynamic features did not significantly enhance model accuracy. Future studies should explore additional factors to improve predictive models for optimizing individualized anticoagulation management in ICUs.

  • Open Access

    Machine learning‐based prediction of bleeding risk in extracorporeal membrane oxygenation patients using transfusion as a surrogate marker

    by Tadashi Kamio, Masaru Ikegami, Megumi Mizuno, Seiichiro Ishii, Hayato Tajima, Yoshihito Machida, Kiyomitsu Fukaguchi

    Background: The increasing use of extracorporeal membrane oxygenation (ECMO) has highlighted challenges in managing bleeding complications. Optimal transfusion strategies remain uncertain for this diverse patient group, necessitating accurate predictive tools. This study developed and validated a machine learning (ML) algorithm to predict bleeding complications in patients with ECMO, using red blood cell (RBC) transfusion as a surrogate marker.Methods: Data from the Tokushukai Medical Database (2018–2022), covering 71 hospitals, were used. An ML approach was employed to predict bleeding complications, using RBC transfusion events as a surrogate marker. Model performance was evaluated using precision, recall, F1 score, and accuracy. SHapley Additive exPlanations (SHAP) analysis was conducted to identify key factors influencing model predictions.Results: Out of 470 ECMO‐treated intensive care unit patients, 357 were included for model development. Forty‐seven variables were used, with the light gradient boosting machine (LightGBM) and random forest models performing better than the other models, with receiver operating characteristic (ROC) area under the curve (AUC) above 0.7 for both (accuracy: 70.5%, ROC AUC: 0.703, recall: 0.784, and ROC AUC: 0.705, respectively). Models such as extreme gradient boosting performed similarly, while support vector classification had the lowest performance. SHAP analysis identified circulating blood volume, hemoglobin, and weight as the most important predictive factors.Discussion: The LightGBM and Random Forest models effectively predict bleeding complications in patients with ECMO, using RBC transfusion as a surrogate marker. This tool can support early identification of high‐risk patients and improve overall transfusion management.


  • Open Access

    A comparison study of temporal trends of SARS-CoV2 RNAemia and biomarkers to predict success and failure of high flow oxygen therapy among patients with moderate to severe COVID-19

    by Hiroshi Koyama, Kazuya Sakai, Kiyomitsu Fukaguchi, Hiroki Hadano, Yoshihisa Aida, Tadashi Kamio, Takeru Abe, Mototsugu Nishii, Ichiro Takeuchi

    Optimal timing for intubating patients with coronavirus disease 2019 (COVID-19) has been debated throughout the pandemic. Early use of high-flow nasal cannula (HFNC) can help reduce the need for intubation, but delay can result in poorer outcomes. This study examines trends in laboratory parameters and serum severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) RNA levels of patients with COVID-19 in relation to HFNC failure. Patients requiring HFNC within three days of hospitalization between July 1 and September 30, 2021 were enrolled. The primary outcome was HFNC failure (early failure ≤Day 3; late failure ≥Day 4), defined as transfer to intensive care just before/after intubation or in-hospital death. We examined changes in laboratory markers and SARS-CoV2-RNAemia on Days 1, 4, and 7, together with demographic data, oxygenation status, and therapeutic agents. We conducted a univariate logistic regression with the explanatory variables defined as 10% change rate in each laboratory marker from Day 1 to 4. We utilized the log-rank test to assess the differences in HFNC failure rates, stratified based on the presence of SARS-CoV2 RNAemia. Among 122 patients, 17 (13.9%) experienced HFNC failure (early: n = 6, late: n = 11). Seventy-five patients (61.5%) showed an initial SpO2/FiO2 ratio ≤243, equivalent to PaO2/FiO2 ratio ≤200, and the initial SpO2/FiO2 ratio was significantly lower in the failure group (184 vs. 218, p = 0.018). Among the laboratory markers, a 10% increase from Day 1 to 4 of lactate dehydrogenase (LDH) and interleukin (IL)-6 was associated with late failure (Odds ratio [OR]: 1.42, 95% confidence interval [CI]: 1.09–1.89 and OR: 1.04, 95%CI: 1.00–1.19, respectively). Furthermore, in patients with persistent RNAemia on Day 4 or 7, the risk of late HFNC failure was significantly higher (Log-rank test, p<0.01). In conclusion, upward trends in LDH and IL-6 levels and the persistent RNAemia even after treatment were associated with HFNC failure.


  • Top Cited Papers
    Open Access

    The Accuracy and Potential Racial and Ethnic Biases of GPT-4 in the Diagnosis and Triage of Health Conditions: Evaluation Study

    by Naoki Ito, Sakina Kadomatsu, Mineto Fujisawa, Kiyomitsu Fukaguchi, Ryo Ishizawa, Naoki Kanda, Daisuke Kasugai, Mikio Nakajima, Tadahiro Goto, Yusuke Tsugawa

    Background: Whether GPT-4, the conversational artificial intelligence, can accurately diagnose and triage health conditions and whether it presents racial and ethnic biases in its decisions remain unclear. Objective: We aim to assess the accuracy of GPT-4 in the diagnosis and triage of health conditions and whether its performance varies by patient race and ethnicity. Methods: We compared the performance of GPT-4 and physicians, using 45 typical clinical vignettes, each with a correct diagnosis and triage level, in February and March 2023. For each of the 45 clinical vignettes, GPT-4 and 3 board-certified physicians provided the most likely primary diagnosis and triage level (emergency, nonemergency, or self-care). Independent reviewers evaluated the diagnoses as “correct” or “incorrect.” Physician diagnosis was defined as the consensus of the 3 physicians. We evaluated whether the performance of GPT-4 varies by patient race and ethnicity, by adding the information on patient race and ethnicity to the clinical vignettes. Results: The accuracy of diagnosis was comparable between GPT-4 and physicians (the percentage of correct diagnosis was 97.8% (44/45; 95% CI 88.2%-99.9%) for GPT-4 and 91.1% (41/45; 95% CI 78.8%-97.5%) for physicians; P=.38). GPT-4 provided appropriate reasoning for 97.8% (44/45) of the vignettes. The appropriateness of triage was comparable between GPT-4 and physicians (GPT-4: 30/45, 66.7%; 95% CI 51.0%-80.0%; physicians: 30/45, 66.7%; 95% CI 51.0%-80.0%; P=.99). The performance of GPT-4 in diagnosing health conditions did not vary among different races and ethnicities (Black, White, Asian, and Hispanic), with an accuracy of 100% (95% CI 78.2%-100%). P values, compared to the GPT-4 output without incorporating race and ethnicity information, were all .99. The accuracy of triage was not significantly different even if patients’ race and ethnicity information was added. The accuracy of triage was 62.2% (95% CI 46.5%-76.2%; P=.50) for Black patients; 66.7% (95% CI 51.0%-80.0%; P=.99) for White patients; 66.7% (95% CI 51.0%-80.0%; P=.99) for Asian patients, and 62.2% (95% CI 46.5%-76.2%; P=.69) for Hispanic patients. P values were calculated by comparing the outputs with and without conditioning on race and ethnicity. Conclusions: GPT-4’s ability to diagnose and triage typical clinical vignettes was comparable to that of board-certified physicians. The performance of GPT-4 did not vary by patient race and ethnicity. These findings should be informative for health systems looking to introduce conversational artificial intelligence to improve the efficiency of patient diagnosis and triage.


  • Open Access

    The Accuracy and Potential Racial and Ethnic Biases of GPT-4 in the Diagnosis and Triage of Health Conditions: Evaluation Study (Preprint)

    by Naoki Ito, Sakina Kadomatsu, Mineto Fujisawa, Kiyomitsu Fukaguchi, Ryo Ishizawa, Naoki Kanda, Daisuke Kasugai, Mikio Nakajima, Tadahiro Goto, Yusuke Tsugawa

    BACKGROUND Whether GPT-4, the conversational artificial intelligence, can accurately diagnose and triage health conditions and whether it presents racial and ethnic biases in its decisions remain unclear. OBJECTIVE We aim to assess the accuracy of GPT-4 in the diagnosis and triage of health conditions and whether its performance varies by patient race and ethnicity. METHODS We compared the performance of GPT-4 and physicians, using 45 typical clinical vignettes, each with a correct diagnosis and triage level, in February and March 2023. For each of the 45 clinical vignettes, GPT-4 and 3 board-certified physicians provided the most likely primary diagnosis and triage level (emergency, nonemergency, or self-care). Independent reviewers evaluated the diagnoses as “correct” or “incorrect.” Physician diagnosis was defined as the consensus of the 3 physicians. We evaluated whether the performance of GPT-4 varies by patient race and ethnicity, by adding the information on patient race and ethnicity to the clinical vignettes. RESULTS The accuracy of diagnosis was comparable between GPT-4 and physicians (the percentage of correct diagnosis was 97.8% (44/45; 95% CI 88.2%-99.9%) for GPT-4 and 91.1% (41/45; 95% CI 78.8%-97.5%) for physicians; P=.38). GPT-4 provided appropriate reasoning for 97.8% (44/45) of the vignettes. The appropriateness of triage was comparable between GPT-4 and physicians (GPT-4: 30/45, 66.7%; 95% CI 51.0%-80.0%; physicians: 30/45, 66.7%; 95% CI 51.0%-80.0%; P=.99). The performance of GPT-4 in diagnosing health conditions did not vary among different races and ethnicities (Black, White, Asian, and Hispanic), with an accuracy of 100% (95% CI 78.2%-100%). P values, compared to the GPT-4 output without incorporating race and ethnicity information, were all .99. The accuracy of triage was not significantly different even if patients’ race and ethnicity information was added. The accuracy of triage was 62.2% (95% CI 46.5%-76.2%; P=.50) for Black patients; 66.7% (95% CI 51.0%-80.0%; P=.99) for White patients; 66.7% (95% CI 51.0%-80.0%; P=.99) for Asian patients, and 62.2% (95% CI 46.5%-76.2%; P=.69) for Hispanic patients. P values were calculated by comparing the outputs with and without conditioning on race and ethnicity. CONCLUSIONS GPT-4’s ability to diagnose and triage typical clinical vignettes was comparable to that of board-certified physicians. The performance of GPT-4 did not vary by patient race and ethnicity. These findings should be informative for health systems looking to introduce conversational artificial intelligence to improve the efficiency of patient diagnosis and triage.