Glass box machine learning for retrospective cohort studies using many patient records. The complex example of bleeding peptic ulcer

Glass Box Machine Learning is, in this study, a type of partially supervised data mining and prediction technique, like a neural network in which each weight or pattern of mutually relevant weights is now replaced by a meaningful "probabilistic knowledge element." We apply it to retrospective cohort studies using large numbers of structured medical records to help select candidate patients for future cohort studies and similar clinical trials. Here it is applied to aid analysis of approaches to aid Deep Learning, but the method lends itself well to direct computation of odds with "explainability" in study design that can complement "Black Box" Deep Learning. Cohort studies and clinical trials traditionally involved at least one 2 × 2 contingency table, but in the age of emerging personalized medicine and the use of machine learning to discover and incorporate further relevant factors, these tables can extend into many extra dimensions as a 2 × 2 x 2 × 2 x ….data structure by considering different conditional demographic and clinical factors of a patient or group, as well as variations in treatment. We consider this in terms of multiple 2 × 2 x 2 data substructures where each one is summarized by an appropriate measure of risk and success called DOR*. This is the diagnostic odds ratio DOR for a specified disease conditional on a favorable outcome divided by the corresponding DOR conditional on an unfavorable outcome. Bleeding peptic ulcer was chosen as a complex disease with many influencing factors, one that is still subject to controversy and that highlights the challenges of using Real World Data.

Analysing Crime Scene Data, Medical Records and Forensic Information to Determine the Causes of Suicides

OBJECTIVE

 To analyse crime scene data, medical records, and forensic information to unveil insights into the causes and traits of suicides.

STUDY DESIGN

 Descriptive study. Place and Duration of the Study: Department of Family Medicine, Kayseri City Hospital, Kayseri, Turkiye, between January 2020 to December 2021.

METHODOLOGY

 A suicide investigation team (doctor, social worker, psychologist) was created to study cases and conduct on-site psychological autopsies. Triggered by emergency calls, the team interviewed suicide victims' relatives using semi-structured questionnaires, gathering data on personal details, time, method, and potential motives. Medical records revealed psychiatric history and medication use, while national judicial systems were reviewed for legal records.

RESULTS

 A total of 158 fatal suicides were studied. Males accounted for 73.4%, females 26.6%. The leading cause was psychiatric illness (43%), chiefly depression (39%). Suicide peaked in the fall, especially in September, mainly at 23:00-23:59. Home was the common site (58.9%), and hanging was the primary method (44.3%). Prior hospitalisation for suicide attempts was 7.5%. Criminal records were held by 16.4% (26 individuals).

CONCLUSION

 The results support the idea that suicides have seasonal patterns and that there are temporal windows of increased risk for suicide.

KEY WORDS

 Suicide reasons, Suicide time, Psychological autopsy, Seasonal and temporal patterns.

Pruritus in primary biliary cholangitis is under-recorded in patient medical records

OBJECTIVE

Cholestatic pruritus in primary biliary cholangitis (PBC) reduces patients' health-related quality of life (HRQoL). Despite this, existing research suggests that pruritus is under-recorded in patients' health records. This study assessed the extent to which pruritus was recorded in medical records of patients with PBC as compared with patient-reported pruritus, and whether patients reporting mild itch were less likely to have pruritus recorded. We also evaluated clinico-demographic characteristics and HRQoL of patients with medical record-documented and patient-reported pruritus.

DESIGN

This cross-sectional study used clinical information abstracted from medical records, together with patient-reported (PBC-40) data from patients with PBC in the USA enrolled in the PicnicHealth cohort. Medical record-documented pruritus was classified as 'recent' (at, or within 12 months prior to, enrolment) or 'ever' (at, or any point prior to, enrolment). Patient-reported pruritus (4-week recall) was assessed using the first PBC-40 questionnaire completed on/after enrolment; pruritus severity was classified by itch domain score (any severity: ≥1; clinically significant itch: ≥7). Patient clinico-demographic characteristics and PBC-40 domain scores were described in patients with medical record-documented and patient-reported pruritus; overlap between groups was evaluated. Descriptive statistics were reported.

RESULTS

Pruritus of any severity was self-reported by 200/225 (88.9%) patients enrolled; however, only 88/225 (39.1%) had recent medical record-documented pruritus. Clinically significant pruritus was self-reported by 120/225 (53.3%) patients; of these, 64/120 (53.3%) had recent medical record-documented pruritus. Patients reporting clinically significant pruritus appeared to have higher mean scores across PBC-40 domains (indicating reduced HRQoL), versus patients with no/mild patient-reported pruritus or medical-record documented pruritus.

CONCLUSION

Compared with patient-reported measures, pruritus in PBC is under-recorded in medical records and is associated with lower HRQoL. Research based only on medical records underestimates the true burden of pruritus, meaning physicians may be unaware of the extent and impact of pruritus, leading to potential undertreatment.

Application of virtual diagnosis and treatment combined with medical record teaching method in standardized training of general practitioner

The aim of this study was to explore the effect of virtual diagnosis and treatment combined with the medical record teaching method in standardized training of general practitioners. Eighty students who had standardized general practice training, from March 2020 to March 2022, in the grassroots practice base of general practitioner training in the affiliated Hospital of our Medical College were retrospectively analyzed and divided into 2 groups according to the teaching method that they received. The differences in assessment scores, critical thinking, clinical thinking ability, learning autonomy ability, and classroom teaching effectiveness were compared, and the students' satisfaction with teaching was investigated. The scores of theoretical knowledge, skill operation, medical history collection, and case analysis in the study group were notably higher (P < .05). In the study group, scores in truth-seeking, openness to knowledge, analytical ability, systematic ability, self-confidence, curiosity, and cognitive maturity were significantly higher (P < .05). A notable improvement was observed in the study group's scores on systematic thinking ability and evidence-based thinking ability, as well as the scores on critical thinking ability after teaching (P < .05). The scores of learning interest, self-management, plan implementation, and cooperation ability improved notably after teaching (P < .05). Learning target, learning processes, learning effects, classroom environment construction, teaching strategy, and technology application in the study group were significantly higher than those in the control group (P < .05). The satisfaction rate in the study group was significantly higher than that in the control group (P < .05). Virtual diagnosis and treatment combined with case-based learning teaching has a very good effect in the standardized training of general practitioners. Students are generally satisfied with their learning experience, which can improve their critical thinking ability and clinical thinking skills. This teaching method is worth further popularizing.

Improving medical record completeness at Wallaga University Referral Hospital: a multidimensional quality improvement project

BACKGROUND

Appropriately documented medical records enhance coordination, patient outcomes and clinical research.

OBJECTIVE

The aim of this project was to improve Wallaga University Referral Hospital's (WURH) medical record completeness rate from 53% to 80% from 1 January 2023 to 31 August 2023.

METHODS

A hospital-based interventional study was conducted at WURH. The Plan-Do-Study-Act cycle was used to test change ideas. A fishbone diagram and a driver diagram were used to identify root causes and address them. Key interventions consisted of supportive supervision, developing and distributing standardised formats, orientation for staff, establishing a chart audit team and assigning data owners.

RESULT

On the completion of the project, the overall implementation of inpatient medical record completeness increased from 53% to 82%. This improvement varies from department-to-department. It increased from 51% to 79%, 53% to 79%, 46% to 81% and 64% to 91% in the departments of internal medicine, paediatrics, obstetrics and gynaecology and surgery, respectively. The project brought improvements in the completeness of physician notes (84% to 100%), physician order sheet (54% to 84%), nursing care plan (26% to 69%), admission sheet (76% to 98%), discharge summary (94% to 98%), progress note (38% to 91%), medication administration (80% to 100%), appropriate attachment of documents (78% to 93%) and documentation of vital signs (50% to 100%).

CONCLUSION AND RECOMMENDATION

The rate of medical record completeness was significantly improved in the study area. This was achieved through the application of multidimensional change ideas related to health professionals, supplies, health management information systems and leadership. However, in some of the parameters, the national targets were not met. Therefore, we recommend providing regular technical updates, conducting frequent chart audits and providing supportive supervision for the enhancement of medical record completeness. It is also advisable for the hospital management to work on its sustainability.

Predicting survival in locally advanced gastric cancer using prognostic factors - neoadjuvant rectal score and downstaging depth score

BACKGROUND

Clinical prediction models are needed to accurately predict the prognosis of patients with gastric cancer who have received neoadjuvant therapy and to determine the best treatment strategies. The aim of this study is to determine the role of two prognostic factors, the neoadjuvant rectal (NAR) score and the downstaging depth score (DDS), in predicting survival in patients with gastric cancer who received neoadjuvant therapy and underwent curative gastrectomy.

METHODS

We reviewed the medical records of 129 patients who had been diagnosed with primary gastric cancer and underwent radical gastrectomy after receiving neoadjuvant therapy. We calculated the NAR score and DDS values for each patient and conducted a survival analysis to assess the accuracy of these prognostic factors in predicting overall survival.

RESULTS

The median overall survival time of the patients was found to be 29 months. Patients with low NAR scores and high DDS had significantly longer overall survival. Univariate analyses based on clinical and laboratory characteristics showed that gender, surgery type, resection type, neural invasion, grade, adjuvant radiotherapy, lymphocyte level, carcinoembryonic antigen (CEA) level, NAR score, and DDS were associated with survival. Moreover, multivariate analyses showed that lymphocyte level, DDS, and NAR score were independent prognostic factors.

CONCLUSION

In summary, our research indicates that NAR score and DDS may serve as useful prognostic markers for predicting overall survival in patients with locally advanced gastric cancer who receive neoadjuvant chemotherapy followed by curative surgery. Patients with high DDS and low NAR scores were found to have better prognoses.

[Cybersecurity and use of ICT in the health sector]

Cybercrime in the health sector is a growing threat in the digital age. With computerization of medical records and telemedicine on the rise, cyberattacks can have devastating consequences. Leaking sensitive data or hijacking systems can compromise patient's privacy and jeopardize healthcare. To counter this threat, robust cybersecurity measures are required as a protective measure. This article aims to expose the main dangers and threats faced by ICT, as well as present cybersecurity with its bioethical implications and, finally, the ideal scheme for it in the health sector in order to create a safer and more efficient environment. This article aims to address these issues and provide a comprehensive view of how cybersecurity and ICT can coexist safely and effectively in the healthcare field.

Clinical signs, causes, and outcome of central cord syndrome in 22 cats

OBJECTIVE

To describe the signalment, clinical findings, presumptive or definitive diagnosis, and outcome in cats with central cord syndrome (CCS).

ANIMALS

22 cats.

CLINICAL PRESENTATION

Cats evaluated for CCS at 7 referral hospitals between 2017 and 2021 were included. Information retrieved from medical records included signalment, physical and neurological examination findings, diagnostic investigations, definitive or presumptive diagnosis, treatment, and follow-up.

RESULTS

Median age at presentation was 9 years. Two neuroanatomical localizations were associated with CCS: C1-C5 spinal cord segments in 17 (77.3%) cats and C6-T2 spinal cord segments in 5 (22.7%) cats. Neuroanatomical localization did not correlate with lesion location on MRI in 8 (36.3%) cats. The most common lesion location within the vertebral column was over the C2 and C4 vertebral bodies in 6 (27.2%) and 5 (22.7%) cats, respectively. Peracute clinical signs were observed in 11 (50%) cats, acute in 1 (4.5%), subacute in 4 (18%), and chronic and progressive signs were seen in 6 (40.9%) cats. The most common peracute condition was ischemic myelopathy in 8 (36.3%) cats, whereas neoplasia was the most frequently identified chronic etiology occurring in 5 (22.7%) cats. Outcome was poor in 13 (59%) cats, consisting of 4 of 11 (36.6%) of the peracute cases, 3 of 4 (75%) of the subacute cases, and 6 of 6 of the chronic cases.

CLINICAL RELEVANCE

Central cord syndrome can occur in cats with lesions in the C1-C5 and C6-T2 spinal cord segments. Multiple etiologies can cause CCS, most commonly, ischemic myelopathy and neoplasia. Prognosis depends on the etiology and onset of clinical signs.

Linking Trauma Registry Patients With Insurance Claims: Creating a Longitudinal Patient Record

INTRODUCTION

Trauma registries and their quality improvement programs only collect data from the acute hospital admission, and no additional information is captured once the patient is discharged. This lack of long-term data limits these programs' ability to affect change. The goal of this study was to create a longitudinal patient record by linking trauma registry data with third party payer claims data to allow the tracking of these patients after discharge.

METHODS

Trauma quality collaborative data (2018-2019) was utilized. Inclusion criteria were patients age ≥18, ISS ≥5 and a length of stay ≥1 d. In-hospital deaths were excluded. A deterministic match was performed with insurance claims records based on the hospital name, date of birth, sex, and dates of service (±1 d). The effect of payer type, ZIP code, International Classification of Diseases, Tenth Revision, Clinical Modification diagnosis specificity and exact dates of service on the match rate was analyzed.

RESULTS

The overall match rate between these two patient record sources was 27.5%. There was a significantly higher match rate (42.8% versus 6.1%, P < 0.001) for patients with a payer that was contained in the insurance collaborative. In a subanalysis, exact dates of service did not substantially affect this match rate; however, specific International Classification of Diseases, Tenth Revision, Clinical Modification codes (i.e., all 7 characters) reduced this rate by almost half.

CONCLUSIONS

We demonstrated the successful linkage of patient records in a trauma registry with their insurance claims. This will allow us to the collect longitudinal information so that we can follow these patients' long-term outcomes and subsequently improve their care.

Experience with a hybrid recruitment approach of patient-facing web portal screening and subsequent phone and medical record review for a neurosurgical intervention trial for chronic ischemic stroke disability (PISCES III)

BACKGROUND

Recruitment of participants is the greatest risk to completion of most clinical trials, with 20-40% of trials failing to reach the targeted enrollment. This is particularly true of trials of central nervous system (CNS) therapies such as intervention for chronic stroke. The PISCES III trial was an invasive trial of stereotactically guided intracerebral injection of CTX0E03, a fetal derived neural stem cell line, in patients with chronic disability due to ischemic stroke. We report on the experience using a novel hybrid recruitment approach of a patient-facing portal to self-identify and perform an initial screen for general trial eligibility (tier 1), followed by phone screening and medical records review (tier 2) prior to a final in-person visit to confirm eligibility and consent.

METHODS

Two tiers of screening were established: an initial screen of general eligibility using a patient-facing web portal (tier 1), followed by a more detailed screen that included phone survey and medical record review (tier 2). If potential participants passed the tier 2 screen, they were referred directly to visit 1 at a study site, where final in-person screening and consent were performed. Rates of screening were tracked during the period of trial recruitment and sources of referrals were noted.

RESULTS

The approach to screening and recruitment resulted in 6125 tier 1 screens, leading to 1121 referrals to tier 2. The tier 2 screening resulted in 224 medical record requests and identification of 86 qualifying participants for referral to sites. The study attained a viable recruitment rate of 6 enrolled per month prior to being disrupted by COVID 19.

CONCLUSIONS

A tiered approach to eligibility screening using a hybrid of web-based portals to self-identify and screen for general eligibility followed by a more detailed phone and medical record review allowed the study to use fewer sites and reduce cost. Despite the difficult and narrow population of patients suffering moderate chronic disability from stroke, this strategy produced a viable recruitment rate for this invasive study of intracranially injected neural stem cells.

TRIAL REGISTRATION

ClinicalTrials.gov Identifier: NCT03629275.

AI-Generated Clinical Summaries Require More Than Accuracy

This Viewpoint discusses AI-generated clinical summaries and the necessity of transparent development of standards for their safe rollout.

Validation of human immunodeficiency virus diagnosis codes among women enrollees of a U.S. health plan

BACKGROUND

Efficiently identifying patients with human immunodeficiency virus (HIV) using administrative health care data (e.g., claims) can facilitate research on their quality of care and health outcomes. No prior study has validated the use of only ICD-10-CM HIV diagnosis codes to identify patients with HIV.

METHODS

We validated HIV diagnosis codes among women enrolled in a large U.S. integrated health care system during 2010-2020. We examined HIV diagnosis code-based algorithms that varied by type, frequency, and timing of the codes in patients' claims data. We calculated the positive predictive values (PPVs) and 95% confidence intervals (CIs) of the algorithms using a medical record-confirmed diagnosis of HIV as the gold standard.

RESULTS

A total of 272 women with ≥ 1 HIV diagnosis code in the administrative claims data were identified and medical records were reviewed for all 272 women. The PPV of an algorithm classifying women as having HIV as of the first HIV diagnosis code during the observation period was 80.5% (95% CI: 75.4-84.8%), and it was 93.9% (95% CI: 90.0-96.3%) as of the second. Little additional increase in PPV was observed when a third code was required. The PPV of an algorithm based on ICD-10-CM-era codes was similar to one based on ICD-9-CM-era codes.

CONCLUSION

If the accuracy measure of greatest interest is PPV, our findings suggest that use of ≥ 2 HIV diagnosis codes to identify patients with HIV may perform well. However, health care coding practices may vary across settings, which may impact generalizability of our results.

Valproic acid monitoring: Serum prediction using a machine learning framework from multicenter real-world data

BACKGROUND

Our study employs machine learning to predict serum valproic acid (VPA) concentrations, aiming to contribute to the development of non-invasive assays for therapeutic drug monitoring.

METHODS

Medical records from 2002 to 2019 were obtained from the Taiwan Chang Gung Research Database. Using various machine learning algorithms, we developed predictive models to classify serum VPA concentrations into two categories (1-50 μg/ml or 51-100 μg/ml) and predicted the exact concentration value. The models were trained on 5142 samples and tested on 644 independent samples. Accuracy was the main metric used to evaluate model performance, with a tolerance of 20 μg/ml for continuous variables. Furthermore, we identified important features and developed simplified models with fewer features.

RESULTS

The models achieved an average accuracy of 0.80-0.86 for binary outcomes and 0.72-0.88 for continuous outcome. Ten top features associated with higher serum VPA levels included higher VPA last and daily doses, bipolar disorder or schizophrenia spectrum disorder diagnoses, elevated levels of serum albumin, calcium, and creatinine, low platelet count, low percentage of segmented white blood cells, and low red cell distribution width-coefficient of variation. The simplified models had an average accuracy of 0.82-0.86 for binary outcome and 0.70-0.86 for continuous outcome.

LIMITATIONS

The study's predictive model lacked external test data from outside the hospital for validation.

CONCLUSIONS

Machine learning models have the potential to integrate real-world data and predict VPA concentrations, providing a promising tool for reducing the need for frequent monitoring of serum levels in clinical practice.

The Environmental Impacts of Electronic Medical Records Versus Paper Records at a Large Eye Hospital in India: Life Cycle Assessment Study

BACKGROUND

Health care providers worldwide are rapidly adopting electronic medical record (EMR) systems, replacing paper record-keeping systems. Despite numerous benefits to EMRs, the environmental emissions associated with medical record-keeping are unknown. Given the need for urgent climate action, understanding the carbon footprint of EMRs will assist in decarbonizing their adoption and use.

OBJECTIVE

We aimed to estimate and compare the environmental emissions associated with paper medical record-keeping and its replacement EMR system at a high-volume eye care facility in southern India.

METHODS

We conducted the life cycle assessment methodology per the ISO (International Organization for Standardization) 14040 standard, with primary data supplied by the eye care facility. Data on the paper record-keeping system include the production, use, and disposal of paper and writing utensils in 2016. The EMR system was adopted at this location in 2018. Data on the EMR system include the allocated production and disposal of capital equipment (such as computers and routers); the production, use, and disposal of consumable goods like paper and writing utensils; and the electricity required to run the EMR system. We excluded built infrastructure and cooling loads (eg. buildings and ventilation) from both systems. We used sensitivity analyses to model the effects of practice variation and data uncertainty and Monte Carlo assessments to statistically compare the 2 systems, with and without renewable electricity sources.

RESULTS

This location's EMR system was found to emit substantially more greenhouse gases (GHGs) than their paper medical record system (195,000 kg carbon dioxide equivalents [CO2e] per year or 0.361 kg CO2e per patient visit compared with 20,800 kg CO2e per year or 0.037 kg CO2e per patient). However, sensitivity analyses show that the effect of electricity sources is a major factor in determining which record-keeping system emits fewer GHGs. If the study hospital sourced all electricity from renewable sources such as solar or wind power rather than the Indian electric grid, their EMR emissions would drop to 24,900 kg CO2e (0.046 kg CO2e per patient), a level comparable to the paper record-keeping system. Energy-efficient EMR equipment (such as computers and monitors) is the next largest factor impacting emissions, followed by equipment life spans. Multimedia Appendix 1 includes other emissions impact categories.

CONCLUSIONS

The climate-changing emissions associated with an EMR system are heavily dependent on the sources of electricity. With a decarbonized electricity source, the EMR system's GHG emissions are on par with paper medical record-keeping, and decarbonized grids would likely have a much broader benefit to society. Though we found that the EMR system produced more emissions than a paper record-keeping system, this study does not account for potential expanded environmental gains from EMRs, including expanding access to care while reducing patient travel and operational efficiencies that can reduce unnecessary or redundant care.

Comparing patient reported and medical record data of adverse drug reactions to anti-seizure drugs

BACKGROUND

Anti-seizure drugs (ASDs) can potentially cause serious adverse drug reactions (ADRs). Patient self-reporting can increase the rate of ADR detection, but studies examining patient self-reporting of ADRs caused by ASDs are lacking.

AIM

To determine the characteristics of ADRs reported by patients receiving ASDs, assess laboratory data and medical record confirmation of patient-reported ADRs, and explore factors associated with laboratory data and medical record confirmation.

METHOD

A self-reporting questionnaire was distributed to patients prescribed ASDs at outpatient clinics. Patients assessed the causality of suspected ADRs using Causality Assessment Tool. Naranjo's algorithm was used by researchers for causality assessment. Medical records were used to gather information on ADR symptoms, ASD medication, and abnormal laboratory data.

RESULTS

From 478 distributed questionnaires, 93.1% completed the questionnaire and 67.4% of respondents reported at least one ADR. The most common ADRs were drowsiness (50.7%), dizziness (9.7%), and ataxia (4.3%). For causality, suspected ADRs were classified as possible in 52.3% of cases and probable in 46.3% of cases by patients, and possible in 64.7% of cases and probable in 25.7% of cases by researchers. Only 12.7% of patients had laboratory data and/or medical record confirmation of suspected ADRs. The psychiatry clinic was less likely to confirm suspected ADRs compared to the epilepsy clinic (OR = 0.412, p = 0.022).

CONCLUSION

Confirmation of patient-reported ADRs with either laboratory data or medical records was uncommon. Recording patient-reported ADRs in patients' medical history and monitoring laboratory tests related to patient-reported symptoms should be promoted to increase the safety of ASD treatment.

Blockchain technology in Cardiovascular Medicine: a glance to the future? Results from a social media survey and future perspectives

The leverage of digital facilities in medicine for disease diagnosis, monitoring, and medical history recording has become increasingly pivotal. However, the advancement of these technologies poses a significant challenge regarding data privacy, given the highly sensitive nature of medical information. In this context, the application of Blockchain technology, a digital system where information is stored in blocks and each block is linked to the one before, has the potential to enhance existing technologies through its exceptional security and transparency. This paradigm is of particular importance in cardiovascular medicine, where the prevalence of chronic conditions leads to the need for secure remote monitoring, secure data storage and secure medical history updating. Indeed, digital support for chronic cardiovascular pathologies is getting more and more crucial. This paper lays its rationale in three primary aims: 1) to scrutinize the existing literature for tangible applications of blockchain technology in the field of cardiology; 2) to report results from a survey aimed at gauging the reception of blockchain technology within the cardiovascular community, conducted on social media; 3) to conceptualize a web application tailored specifically to cardiovascular care based on blockchain technology. We believe that Blockchain technology may contribute to a breakthrough in healthcare digitalization, especially in the field of cardiology; in this context, we hope that the present work may be inspiring for physicians and healthcare stakeholders.

Accounting for informative observation process in transition models of binary longitudinal outcome: Application to medical record data

When extracting medical record data to form a retrospective cohort, investigators typically focus on a pre-specified study window, and select subjects who had hospital visits during that study window. However, such data extraction may suffer from an informative observation process, since sicker patients may have hospital visits more frequently. For example, Consecutive Pregnancy Study is a retrospective cohort study of women with multiple pregnancies in 23 Utah hospitals from 2003 to 2010, where the interest is to understand the risk factors of recurrent pregnancy outcomes, such as preterm birth. The observation process is informative in the sense that, women with adverse pregnancy outcomes may be less likely/willing/able to endure subsequent pregnancies. We proposed a three-part joint model with shared random effects structure to address this analytic complication. Particularly, a first-order transition model is used to model the longitudinal binary outcome; a gamma regression model is assumed for the inter-pregnancy intervals; a continuation ratio model specifies the probability of continuing with more births in the future. We note that the latter two parts give rise to a parametric cure-rate survival model. The performance of the proposed method was examined in extensive simulation studies, with both correctly and mis-specified models. The analyses of Consecutive Pregnancy Study data further demonstrate the inadequacies of fitting the transition model alone ignoring the informative observation process.

Delirium is frequently underdiagnosed among older hospitalised patients despite available information in hospital medical records

BACKGROUND

In-hospital delirium is associated with adverse outcomes and is underdiagnosed, limiting research and clinical follow-up.

OBJECTIVE

To compare the incidence of in-hospital delirium determined by chart-based review of electronic medical records (D-CBR) with delirium discharge diagnoses (D-DD). Furthermore, to identify differences in symptoms, treatments and delirium triggers between D-CBR and D-DD.

METHOD

The community-based cohort included 2,115 participants in the Hordaland Health Study born between 1925 and 1927. Between 2018 and 2022, we retrospectively reviewed hospital electronic medical records from baseline (1997-99) until death prior to 2023. D-DD and D-CBR were validated using The Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition, criteria for delirium.

RESULTS

Of the 2,115 participants, 638 had in-hospital delirium. The incidence rate (IR) of D-CBR was 29.8 [95% confidence interval 28, 32] per 1,000 person-years, whereas the IR by D-DD was 3.4 [2.8, 4.2]. The IR ratio was 9.14 (P < 0.001). Patients who received pharmacological treatment for delirium (n = 121, odds ratio (OR) 3.4, [2.1, 5.4], P < 0.001), who were affected by acute memory impairment (n = 149, OR 2.8, [1.8, 4.5], P < 0.001), or change in perception (n = 137, OR 2.9, [1.8, 4.6] P < 0.001) had higher odds for D-DD. In contrast, post-operative cases (OR 0.2, [0.1, 0.4], P < 0.001) had lower odds for D-DD.

CONCLUSION

Underdiagnosis of in-hospital delirium was a major issue in our study, especially in less severe delirium cases. Our findings emphasise the need for integrating systematic delirium diagnostics and documentation into hospital admission and discharge routines.

AERMNet: Attention-enhanced relational memory network for medical image report generation

BACKGROUND AND OBJECTIVES

The automatic generation of medical image diagnostic reports can assist doctors in reducing their workload and improving the efficiency and accuracy of diagnosis. However, among the most existing report generation models, there are problems that the weak correlation between generated words and the lack of contextual information in the report generation process.

METHODS

To address the above problems, we propose an Attention-Enhanced Relational Memory Network (AERMNet) model, where the relational memory module is continuously updated by the words generated in the previous time step to strengthen the correlation between words in generated medical image report. And the double LSTM with interaction module reduces the loss of context information and makes full use of feature information. Thus, more accurate disease information can be generated by AERMNet for medical image reports.

RESULTS

Experimental results on four medical datasets Fetal heart (FH), Ultrasound, IU X-Ray and MIMIC-CXR, show that our proposed method outperforms some of the previous models with respect to language generation metrics (Cider improving by 2.4% on FH, Bleu1 improving by 2.4% on Ultrasound, Cider improving by 16.4% on IU X-Ray, Bleu2 improving by 9.7% on MIMIC-CXR).

CONCLUSIONS

This work promotes the development of medical image report generation and expands the prospects of computer-aided diagnosis applications. Our code is released at https://github.com/llttxx/AERMNET.

Outcomes of ultraminipercutaneous nephrolithotomy in infants: our experiences at a single center in an endemic region

OBJECTIVE

In this study, we aimed to contribute to the literature by sharing the perioperative and postoperative outcomes of infants (0-24 months) who underwent ultra-mini percutaneous nephrolithotomy (PNL) for kidney stones in our clinic.

METHODS

Infants under 24 months old with kidney stones of 2 cm and larger, who applied to our clinic between January 2018 and May 2023, were included in the study. The patients' demographic and clinical characteristics were obtained from the medical records. The collected data were analyzed retrospectively.

RESULTS

A total of 26 patients were included in the study. The mean age of the patients was 17.3 ± 3.90 (12-24) months. The mean operation time was 50.7 ± 6.43 min. The mean stone size was 2.66 ± 0.59 cm. Stone-free was achieved in 23 patients (88.5%). In one patient (3.8%) with residual fragments, SWL was performed, and in two patients (7.7%), RIRS was performed to achieve stone-free. Postoperatively, fever was observed in 3 patients (11.5%). There were no patients requiring blood transfusion.

CONCLUSIONS

In experienced centers, ultra-mini-PNL performed by experienced surgeons is an effective and reliable treatment option for infants under 24 months of age with kidney stones larger than 2 cm. It provides high-stone clearance rates and low complication rates.

Interoperability in the Wild: Comparison of Real-World Electronic C-CDA Documents from Two Sources

Although health information exchange (HIE) networks exist in multiple nations, providers still require access multiple sources to obtain medical records. We sought to measure and compare differences in data presence and concordance across regional HIE and EHR vendor-based networks. Using 1,054 randomly selected patients from a large health system in the US, we generated consolidated clinical document architecture (C-CDA) documents from each network. 778 (74%) patients had at least one C-CDA document present from either source. Among these patients, two-thirds had information in only one source. All documents contained demographics, but less than half of patients had data in clinical data domains. Moreover, data across HIE networks were not concordant. Results suggest that HIE networks have different, likely complementary, data available for the same patient, suggesting the need for better integration and deduplication for national HIE efforts.

Graph Representation Learning-Based Fixed-Length Clinical Feature Vector Generation from Heterogeneous Medical Records

Transformation of patient data extracted from a database into fixed-length numerical vectors requires expertise in topical medical knowledge as well as data manipulation-thus, manual feature design is labor-intensive. In this study, we propose a machine learning-based method to for this purpose applicable to electronic medical data recorded during hospitalization, which utilizes unsupervised feature extraction based on graph embedding. Unsupervised learning is performed on a heterogeneous graph using Graph2Vec, and the inclusion of clinically useful data in the obtained embedding representation is evaluated by predicting readmission within 30 days of discharge based on it. The embedded representations are observed to improve predictive performance significantly as the information contained in the graph increases, indicating the suitability of the proposed method for feature design corresponding to clinical information.

Predicting In-Hospital Death from Derived EHR Trajectory Features

Medical histories of patients can predict a patient's immediate future. While most studies propose to predict survival from vital signs and hospital tests within one episode of care, we carried out selective feature engineering from longitudinal medical records in this study to develop a dataset with derived features. We thereafter trained multiple machine learning models for the binary prediction of whether an episode of care will culminate in death among patients suspected of bloodstream infections. The machine learning classifier performance is evaluated and compared and the feature importance impacting the model output is explored. The extreme gradient boosting model achieved the best performance for predicting death in the next hospital episode with an accuracy of 92%. Age at the time of the first visit, length of history, and information related to recent episodes were the most critical features.

Quality of Person-Generated Healthy Walking Data: An Explorative Analysis

Despite the potential benefits of Person Generated Health Data (PGHD), data quality issues impede its use. This study examined the effect of different methods for filtering armband data on determining the amount of healthy walking and the consistency between healthy walking captured using armbands and health diaries. Four weeks of armband and health diary data were acquired from 103 college students. Armband data filtering was performed using heart rate measures and minimum daily step counts as a proxy for adequate daily wear time. No substantial differences in the filtered armband datasets were observed by filtering methods. Significant gaps were observed between healthy walking amounts determined from armband data and through the health diary. Future studies need to explore more diverse data filtering methods and their impact on health outcome assessments.

Rule-Based Text Classification of Dental Diagnosis

Unstructured medical records boast an abundance of information that could greatly facilitate medical decision-making and improve patient care. With the development of Natural Language Processing (NLP) methodology, the free-text medical data starts to attract more and more research attention. Most existing studies try to leverage the power of such unstructured data using Machine Learning algorithms, which would usually require a relatively large training set, and high computational capacity. However, when faced with a smaller-scale project, opting for an alternative approach may be more effective and practical. This project proposes an efficient and light-weight rule-based approach to categorize dental diagnosis data. It not only fills the void of dental records in the medical free-text processing area, but also demonstrates that with expertly designed research structure and proper implementation, simple method could achieve our study goal very competently.

Applications of artificial intelligence in the field of oral and maxillofacial pathology: a systematic review and meta-analysis

BACKGROUND

Since AI algorithms can analyze patient data, medical records, and imaging results to suggest treatment plans and predict outcomes, they have the potential to support pathologists and clinicians in the diagnosis and treatment of oral and maxillofacial pathologies, just like every other area of life in which it is being used. The goal of the current study was to examine all of the trends being investigated in the area of oral and maxillofacial pathology where AI has been possibly involved in helping practitioners.

METHODS

We started by defining the important terms in our investigation's subject matter. Following that, relevant databases like PubMed, Scopus, and Web of Science were searched using keywords and synonyms for each concept, such as "machine learning," "diagnosis," "treatment planning," "image analysis," "predictive modelling," and "patient monitoring." For more papers and sources, Google Scholar was also used.

RESULTS

The majority of the 9 studies that were chosen were on how AI can be utilized to diagnose malignant tumors of the oral cavity. AI was especially helpful in creating prediction models that aided pathologists and clinicians in foreseeing the development of oral and maxillofacial pathology in specific patients. Additionally, predictive models accurately identified patients who have a high risk of developing oral cancer as well as the likelihood of the disease returning after treatment.

CONCLUSIONS

In the field of oral and maxillofacial pathology, AI has the potential to enhance diagnostic precision, personalize care, and ultimately improve patient outcomes. The development and application of AI in healthcare, however, necessitates careful consideration of ethical, legal, and regulatory challenges. Additionally, because AI is still a relatively new technology, caution must be taken when applying it to this industry.

A four-way patient search method for the retrospective identification of poisoning patients

When studying emergency department (ED) visits, electronic health record systems of hospitals provide a good basis for retrospective studies. However, many intoxication patients presenting to the ED, may not be identified retrospectively if only a single search method is applied. In this study, a new four-way combined patient search method was used to retrospectively identify intoxication patients presenting to the ED. The search included reason for admission to the ED, laboratory results related to intoxication diagnostics, ICD-10 codes, and a novel free word search (FWS) of patient records. After the automated search, the researcher read the medical records of potential substance abuse patients to form comprehensive profiles and remove irrelevant cases. The addition of a free word search identified 36% more substance abuse patients than the combination of the other three methods mentioned above. Patients identified by the FWS search alone were generally admitted to the ED for trauma or mental health problems and were often found to be heavily under the influence of alcohol and/or drugs. The main intoxicants were ethanol and benzodiazepines. The free word search was highly complementary to traditional patient search methods, highlighting the importance of the combined patient search method in retrospective data collection.

Utilizing patient-nurse verbal communication in building risk identification models: the missing critical data stream in home healthcare

BACKGROUND

In the United States, over 12 000 home healthcare agencies annually serve 6+ million patients, mostly aged 65+ years with chronic conditions. One in three of these patients end up visiting emergency department (ED) or being hospitalized. Existing risk identification models based on electronic health record (EHR) data have suboptimal performance in detecting these high-risk patients.

OBJECTIVES

To measure the added value of integrating audio-recorded home healthcare patient-nurse verbal communication into a risk identification model built on home healthcare EHR data and clinical notes.

METHODS

This pilot study was conducted at one of the largest not-for-profit home healthcare agencies in the United States. We audio-recorded 126 patient-nurse encounters for 47 patients, out of which 8 patients experienced ED visits and hospitalization. The risk model was developed and tested iteratively using: (1) structured data from the Outcome and Assessment Information Set, (2) clinical notes, and (3) verbal communication features. We used various natural language processing methods to model the communication between patients and nurses.

RESULTS

Using a Support Vector Machine classifier, trained on the most informative features from OASIS, clinical notes, and verbal communication, we achieved an AUC-ROC = 99.68 and an F1-score = 94.12. By integrating verbal communication into the risk models, the F-1 score improved by 26%. The analysis revealed patients at high risk tended to interact more with risk-associated cues, exhibit more "sadness" and "anxiety," and have extended periods of silence during conversation.

CONCLUSION

This innovative study underscores the immense value of incorporating patient-nurse verbal communication in enhancing risk prediction models for hospitalizations and ED visits, suggesting the need for an evolved clinical workflow that integrates routine patient-nurse verbal communication recording into the medical record.

Novel linkage approach to join community-acquired and national data

BACKGROUND

Community optometrists in Scotland have performed regular free-at-point-of-care eye examinations for all, for over 15 years. Eye examinations include retinal imaging but image storage is fragmented and they are not used for research. The Scottish Collaborative Optometry-Ophthalmology Network e-research project aimed to collect these images and create a repository linked to routinely collected healthcare data, supporting the development of pre-symptomatic diagnostic tools.

METHODS

As the image record was usually separate from the patient record and contained minimal patient information, we developed an efficient matching algorithm using a combination of deterministic and probabilistic steps which minimised the risk of false positives, to facilitate national health record linkage. We visited two practices and assessed the data contained in their image device and Practice Management Systems. Practice activities were explored to understand the context of data collection processes. Iteratively, we tested a series of matching rules which captured a high proportion of true positive records compared to manual matches. The approach was validated by testing manual matching against automated steps in three further practices.

RESULTS

A sequence of deterministic rules successfully matched 95% of records in the three test practices compared to manual matching. Adding two probabilistic rules to the algorithm successfully matched 99% of records.

CONCLUSIONS

The potential value of community-acquired retinal images can be harnessed only if they are linked to centrally-held healthcare care data. Despite the lack of interoperability between systems within optometry practices and inconsistent use of unique identifiers, data linkage is possible using robust, almost entirely automated processes.

Best practice: quality assessment outcomes of the Practice Enhancement Program among family physicians in Saskatchewan, Canada

Increased family physician workloads have strained primary care. The objective of this study was to describe the frequency and types of quality concerns identified among Saskatchewan's family physicians, changes in these concerns over time, associated physician characteristics, and recommendations made for improvement. In this repeated cross-sectional study (1997-2020), we examined family physician assessment reports from the Saskatchewan Practice Enhancement Program, a mandatory practice review strategy, for quality concerns on three outcomes: care, medical record, and facility. We recorded demographic and practice characteristics, the presence or absence of quality concerns, and the type of recommendations made. Concern incidence was calculated both overall and across subperiods, and three outcome-specific multiple logistic regression models were developed. Recommendations made were quantified, and their nature was evaluated using thematic analysis. Among 824 assessments, 20.8% identified concerns, with a statistically significant increase in 2015-20 over earlier years (14.2% versus 43.4%, P < .001). Corresponding proportions also significantly increased within each quality outcome (6.0%-37.1%, P < .001 for care concerns; 12.7%-19.6%, P = .03 for medical record concerns; 3.9%-21.0%, P < .001 for facility concerns). We found statistically significant adjusted associations between care concerns and both urban location [odds ratio (OR): 2.2; 95% confidence interval (CI): 1.30, 3.8] and international medical training (OR: 2.4; 95% CI: 1.34, 4.2); facility concerns and solo practice (OR: 2.5 95% CI: 1.10, 5.7); and medical record concerns and male gender (OR: 1.88; 95% CI: 1.09, 3.3), solo practice (OR: 1.67; 95% CI: 1.01, 2.7), and increased age. Reflecting a statistically significant interaction found between age as a continuous covariate and time period, older physicians were more likely to have a medical record concern in later years (OR: 1.072; 95% CI: 1.026, 1.120) compared to earlier ones (OR: 1.021; 95% CI: 1.001, 1.043). Among physicians where a concern was identified, recommendations most frequently pertained to documentation (91.2%), chronic disease management (78.2%), cumulative patient profiles (62.9%), laboratory investigations (53.5%), medications (51.8%), and emergency preparedness (51.2%). A concerning and increasing proportion of family physicians have quality gaps, with identifiable factors and recurring recommendations. These findings provide direction for strategic support development.

The impact of hospitalization on mortality in patients with connective tissue disease-associated interstitial lung disease: a medical records review study

BACKGROUND

Interstitial lung disease (ILD) remains one of the most important causes of morbidity and mortality in patients with Connective Tissue Diseases (CTD). This study evaluated the impact of hospitalization on mortality in an ethnically and racially diverse cohort of CTD-ILD patients.

METHODS

We conducted a medical records review study at Montefiore Medical Center, Bronx, NY. We included 96 patients and collected data on demographic characteristics, reasons for hospitalization, length of stay, immunosuppressant therapy use, and mortality. We stratified our patients into two cohorts: hospitalized and non-hospitalized. The hospitalized cohort was further subdivided into cardiopulmonary and non-cardiopulmonary admissions. Two-sample tests or Wilcoxon's rank sum tests for continuous variables and Chi-square or Fisher's exact tests for categorical variables were used for analyses as deemed appropriate.

RESULTS

We identified 213 patients with CTD-ILD. Out of them, 96 patients met the study's inclusion criteria. The majority of patients were females (79%), and self-identified as Hispanic (54%) and Black (40%). The most common CTDs were rheumatoid arthritis (RA) (29%), inflammatory myositis (22%), and systemic sclerosis (15%). The majority (76%) of patients required at least one hospitalization. In the non-hospitalized group, no deaths were observed, however we noted significant increase of mortality risk in hospitalized group (p = 0.02). We also observed that prolonged hospital stay (> 7 days) as well as older age and male sex were associated with increased mortality.

CONCLUSIONS

Prolonged (> 7 days) hospital stay and hospitalization for cardiopulmonary causes, as well as older age and male sex were associated with an increased mortality risk in our cohort of CTD-ILD patients.

Reflections on the use of patient records: Privacy, ethics, and reparations in the history of psychiatry

One of the most common questions we get asked as historians of psychiatry is "do you have access to patient records?" Why are people so fascinated with the psychiatric patient record? Do people assume they are or should be available? Does access to the patient record actually tell us anything new about the history of psychiatry? And if we did have them, what can, or should we do with them? In the push to both decolonize and personalize the history of psychiatry, as well as make some kind of account or reparation for past mistakes, how can we proceed in an ethical manner that respects the privacy of people in the past who never imagined their intensely personal psychiatric encounter as subject for future historians? In this paper, we want to think through some of the issues that we deal with as white historians of psychiatry especially at the intersection of privacy, ethics, and racism. We present our thoughts as a conversation, structured around questions we have posed for ourselves, and building on discussions we have had together over the past few years. We hope that they act as a catalyst for further discussion in the field.

Validation of an ICD-10 case-finding algorithm for endometrial cancer in US insurance claims

PURPOSE

To evaluate the positive predictive value (PPV) of an endometrial cancer case finding algorithm using International Classification of Disease 10th revision Clinical Modification (ICD-10-CM) diagnosis codes from US insurance claims for implementation in a planned post-marketing safety study. Two algorithm variants were evaluated.

METHODS

Provisional incident endometrial cancer cases were identified from 2016 through 2020 among women aged ≥50 years. One algorithm variant used diagnosis codes for malignant neoplasms of uterine sites (C54.x), excluding C54.2 (malignant neoplasm of myometrium); the other used only C54.1 (malignant neoplasm of endometrium). A random sample of medical records of recent incident provisional cases (2018-2020) was requested for adjudication. Confirmed cases showed biopsy evidence of endometrial cancer, documentation of cancer staging, or hysterectomy following diagnosis. We estimated the PPV of the variants with 95% confidence intervals (CI) excluding cases that had insufficient information.

RESULTS

Of 294 provisional cases adjudicated, 85% were from outpatient settings (n = 249). Mean age at diagnosis was 69.3 years. Among the 294 adjudicated cases (identified with the broader algorithm variant), the same 223 were confirmed endometrial cancer cases by both algorithm variants. The PPV (95% CI) for the broader algorithm variant was 84.2% (79.2% and 88.3%), and for the variant using only C54.1 was 85.8% (80.9% and 89.8%).

CONCLUSION

We developed and validated an algorithm using ICD-10-CM diagnosis codes to identify endometrial cancer cases in health insurance claims with a sufficiently high PPV to use in a planned post-marketing safety study.

Leveraging Historical Medical Records as a Proxy via Multimodal Modeling and Visualization to Enrich Medical Diagnostic Learning

Simulation-based Medical Education (SBME) has been developed as a cost-effective means of enhancing the diagnostic skills of novice physicians and interns, thereby mitigating the need for resource-intensive mentor-apprentice training. However, feedback provided in most SBME is often directed towards improving the operational proficiency of learners, rather than providing summative medical diagnoses that result from experience and time. Additionally, the multimodal nature of medical data during diagnosis poses significant challenges for interns and novice physicians, including the tendency to overlook or over-rely on data from certain modalities, and difficulties in comprehending potential associations between modalities. To address these challenges, we present DiagnosisAssistant, a visual analytics system that leverages historical medical records as a proxy for multimodal modeling and visualization to enhance the learning experience of interns and novice physicians. The system employs elaborately designed visualizations to explore different modality data, offer diagnostic interpretive hints based on the constructed model, and enable comparative analyses of specific patients. Our approach is validated through two case studies and expert interviews, demonstrating its effectiveness in enhancing medical training.

Exploring the Documentation of Delirium in Patients After Cardiac Surgery: A Retrospective Patient Record Study

Delirium is a common disorder for patients after cardiac surgery. Its manifestation and care can be examined through EHRs. The aim of this retrospective, comparative, and descriptive patient record study was to describe the documentation of delirium symptoms in the EHRs of patients who have undergone cardiac surgery and to explore how the documentation evolved between two periods (2005-2009 and 2015-2020). Randomly selected care episodes were annotated with a template, including delirium symptoms, treatment methods, and adverse events. The patients were then manually classified into two groups: nondelirious (n = 257) and possibly delirious (n = 172). The data were analyzed quantitatively and descriptively. According to the data, the documentation of symptoms such as disorientation, memory problems, motoric behavior, and disorganized thinking improved between periods. Yet, the key symptoms of delirium, inattention, and awareness were seldom documented. The professionals did not systematically document the possibility of delirium. Particularly, the way nurses recorded structural information did not facilitate an overall understanding of a patient's condition with respect to delirium. Information about delirium or proposed care was seldom documented in the discharge summaries. Advanced machine learning techniques can augment instruments that facilitate early detection, care planning, and transferring information to follow-up care.

Leveraging automated approaches to categorize birth defects from abstracted birth hospitalization data

BACKGROUND

The Surveillance for Emerging Threats to Pregnant People and Infants Network (SET-NET) collects data abstracted from medical records and birth defects registries on pregnant people and their infants to understand outcomes associated with prenatal exposures. We developed an automated process to categorize possible birth defects for prenatal COVID-19, hepatitis C, and syphilis surveillance. By employing keyword searches, fuzzy matching, natural language processing (NLP), and machine learning (ML), we aimed to decrease the number of cases needing manual clinician review.

METHODS

SET-NET captures International Classification of Diseases, 10th Revision, Clinical Modification (ICD-10-CM) codes and free text describing birth defects. For unstructured data, we used keyword searches, and then conducted fuzzy matching with a cut-off match score of ≥90%. Finally, we employed NLP and ML by testing three predictive models to categorize birth defect data.

RESULTS

As of June 2023, 8326 observations containing data on possible birth defects were submitted to SET-NET. The majority (n = 6758 [81%]) were matched to an ICD-10-CM code and 1568 (19%) were unable to be matched. Through keyword searches and fuzzy matching, we categorized 1387/1568 possible birth defects. Of the remaining 181 unmatched observations, we correctly categorized 144 (80%) using a predictive model.

CONCLUSIONS

Using automated approaches allowed for categorization of 99.6% of reported possible birth defects, which helps detect possible patterns requiring further investigation. Without employing these analytic approaches, manual review would have been needed for 1568 observations. These methods can be employed to quickly and accurately sift through data to inform public health responses.

[Detection of adverse events in patients interned in medical clinic using the Global Trigger Tool]

INTRODUCTION

The Global Trigger Tool (GTT) is a tool that accurately identifies adverse events that represent a significant problem in hospitals.

METHODS

Cross-sectional study based on retrospective review of randomized medical records using the GTT tool.

RESULTS

A total of 161 adverse events (AEs) were detected: 51 events per 100 admissions, 66 per 1000 patient-days, and 30% of admissions with AEs. The most frequent triggers were from the care module, with 25% complications associated with the use of procedures, 10% pressure ulcers, and 9% care-associated infections. The presence of AEs had a statistically significant association with a stay of more than 5 days, and a moderate association with age and number of triggers. Regarding the damage, 78% of the patients presented mild events and 4% fatal events. The ROC curves analysis showed that the triggers with the greatest area under the curve were: procedural complication (0.70), pressure ulcers (0.61) and rapid response code (0.60).

DISCUSSION

The number of events per 100 admissions was higher than that reported in the literature, but there were no differences in events per 1000 patientdays. Fatal cases were caused by respiratory infectious diseases in patients with comorbidities, nasogastric tube needs and cognitive decline. The study highlights the scarce use of the tool in public hospitals and the implementation of trigger analysis with ROC curves. Knowing the frequency and the most frequent type of event will allow the implementation of measures that improve patient safety.

[The quality of dental care in children with permanent teeth trauma according to analysis of medical records in an emergency unit of a municipal dental clinic]

THE AIM OF THE STUDY

Was to assess the efficacy and timing of emergency dental care in children with permanent teeth trauma according to analysis of medical records in an emergency unit of a municipal dental clinic.

MATERIAL AND METHODS

The study involved 320 medical records of pediatric patients admitted to emergency dental care unit of a municipal dental clinic in 2021 because of maxillofacial trauma from which 221 records of children with acute dental trauma were extracted. The quality of documentation of the medical records, rationale for diagnosis and adequacy of emergency dental treatment were analyzed.

RESULTS

No records included diagnosis code according to ICD-10. Trauma history was described in the majority of records by in 67% of them no trauma time was stated with proper precision. In 67.6% of permanent teeth trauma cases emergency aid was carried out inadequately. All patients with uncomplicated crown fractures were dismissed with no treatment. In complicated crown fractures needing pulp vitality preservation the pulp was devitalized or just anesthetized. Tooth replantation in avulsion cases was not performed. In 13.5% of records the treatment was not properly described. In 67.6% of records there were no recommendations for follow-ups.

CONCLUSION

There is a strong need for the improvement of knowledge of traumatic dental injuries management among Russian pediatric dentists by elaboration and implementation of protocols for dental traumas treatment.

The Scottish Medical Imaging Archive: 57.3 Million Radiology Studies Linked to Their Medical Records

Keywords: MRI, Imaging Sequences, Ultrasound, Mammography, CT, Angiography, Conventional Radiography Published under a CC BY 4.0 license. See also the commentary by Whitman and Vining in this issue.

Real-World Evidence for Risk Factors of Bruises and Fractures from Falls in Patients with Overactive Bladder: A Medical Record Analysis

Aim

To identify the risk factors for bruises and fractures from falls in patients with overactive bladder (OAB).

Methods

We evaluated 1136 patients with OAB and aged ≥50 years who visited our hospital. Age, sex, frequency of nocturnal urination, and urinary incontinence type were investigated in the 360 eligible patients. Patients were divided into three groups: those patients without falls (no-fall group), those with fall bruises (bruise group), and those with fall fractures (fracture group). The risk factors for bruises and fractures in patients with OAB were evaluated using the logistic regression analysis. In addition, association between the bruises or fractures from falls and the behavior around urination during the night was investigated.

Results

The multivariate logistic regression analysis showed that female sex (odds ratio (OR) 2.888, p = 0.030) and nocturnal urination frequency ≥3 times/night (OR vs. ≤2 times/night, 2.940; p = 0.040) were significantly associated with bruises. Nocturnal urination frequency ≥3 times/night (OR vs. ≤2 times/night, 2.835; p = 0.026) and urge incontinence (OR 3.415, p = 0.016) were significantly associated with fractures. Behavior around urination during the night was significantly associated with fractures (p = 0.009).

Conclusion

In the real-world clinical setting, increasing nocturnal urination frequency is a common risk factor for bruises and fractures. Also, female sex and urge incontinence were the risk factors for bruises and fractures, respectively. OAB patients with urge incontinence would especially require aggressive intervention to prevent fractures during night-time voiding.

Time to normalisation of C-reactive protein and incidence of relapse in microscopic polyangiitis: A medical records review study in Japan

OBJECTIVES

Despite the identification of risk factors for relapses in antineutrophil cytoplasmic antibody-associated vasculitis, the relationship between changes in C-reactive protein (CRP) levels after the initial treatment and the incidence of relapse remains unknown. This study aimed to assess the association between the time taken for normalisation of CRP levels and the incidence of relapse in Japanese adult patients with microscopic polyangiitis.

METHODS

This study included 85 consecutive patients with newly diagnosed microscopic polyangiitis who achieved remission after 6 months of immunosuppressive treatment at the Aichi Medical University Hospital between 2009 and 2017. The relationship between the time to normalisation of CRP after the initial immunosuppressive treatment and relapse incidences was evaluated using multivariable Cox proportional hazard models.

RESULTS

During the follow-up period, 13 (30.2%), 7 (41.2%), and 16 (64.0%) patients relapsed (P = .025) within 1-14, 15-28, and ≥29 days of normalisation, respectively. The hazard ratios [95% confidence intervals (CIs)] for the time to normalisation of CRP of 1-14, 15-28, and ≥29 days were 1.00 (reference), 2.42 (95% CI: 0.92-6.39), and 3.48 (95% CI: 1.56-7.76), respectively.

CONCLUSIONS

A significant association between the time to normalisation of CRP and the relapse incidence in Japanese patients with microscopic polyangiitis was observed.

Comprehensive evaluation of disease coding quality in gastroenterology and its impact on the diagnosis-related group system: a cross-sectional study

OBJECTIVE

According to the diagnosis-related group (DRG) requirement, issues of diagnosis and procedure coding in the gastroenterology department of our hospital were analyzed and improvement plans were proposed to lay the foundation for effective implementation of DRGs.

METHODS

The title page of case-history of 1600 patients admitted to the Department of Gastroenterology of this hospital from January 1, 2021 to December 31, 2021 was sampled as a data source, and the primary and other diagnostic codes, operation or procedure codes involved in the title page of case-history were categorized and statistically analyzed.

RESULTS

Of the 531 cases treated with gastrointestinal endoscopy in our hospital in 2021, coding errors were identified in 66 cases and unsuccessful DRG enrollment in 35 cases, including 14 cases with incorrect coding of the primary diagnosis (8 cases with unsuccessful DRG enrollment), 37 cases with incorrect coding of the primary operation (23 cases with unsuccessful DRG enrollment), and 8 cases with incorrect coding of both the primary diagnosis and the primary operation (4 cases with unsuccessful DRG enrollment). Analysis of 66 inpatient cases with coding problems showed a total of 167 deficiencies, including 36 deficiencies in major diagnoses, 84 deficiencies in other diagnoses, and 47 deficiencies in surgery or operation coding.

CONCLUSION

The accuracy of coding of disease diagnosis and surgical operation is the basis for the smooth implementation of DRGs. The medical staff of this hospital has poor cognition of DRGs coding and fails to recognize the important role of the title page of case-history quality to DRGs system, and their attention to DRGs and knowledge base of disease classification coding should be improved. In addition, the high incidence of coding errors, especially the omission of disease diagnosis, requires increased training of physicians and nurses on clinical knowledge and requirements for DRGs medical records, thereby improving the quality of medical cases and ensuring the accuracy of DRGs information.

Large Language Models in Neurology Research and Future Practice

Recent advancements in generative artificial intelligence, particularly using large language models (LLMs), are gaining increased public attention. We provide a perspective on the potential of LLMs to analyze enormous amounts of data from medical records and gain insights on specific topics in neurology. In addition, we explore use cases for LLMs, such as early diagnosis, supporting patient and caregivers, and acting as an assistant for clinicians. We point to the potential ethical and technical challenges raised by LLMs, such as concerns about privacy and data security, potential biases in the data for model training, and the need for careful validation of results. Researchers must consider these challenges and take steps to address them to ensure that their work is conducted in a safe and responsible manner. Despite these challenges, LLMs offer promising opportunities for improving care and treatment of various neurologic disorders.

The association of environmental factors with neurocognitive outcomes in survivors of childhood acute lymphoblastic leukemia (ALL)

PURPOSE

To investigate the association of environmental factors, rehabilitation services during therapy and socioeconomic status (SES - insurance type), with neurocognitive outcomes at the end of therapy for survivors of childhood acute lymphoblastic leukemia (ALL).

METHODS

Survivors (n = 236) treated on the St. Jude Total Therapy Study 16 completed end of therapy testing with performance measures (IQ, attention, processing speed, fine motor skills, academics) and caregiver ratings (attention, executive function, adaptive skills). Environmental factors were abstracted from the medical record.

RESULTS

Distribution of sex (47.3% female, p = 0.399), treatment arm (45.5% low risk, 54.5% standard/high risk p = 0.929), insurance type (47.7% private, 52.3% public/none, p = 0.117), and mean age at diagnosis (7.7 vs. 6.8 years, p = 0.143) were similar for groups with (n = 110; 46.6%) and without (n = 126; 53.6%) rehabilitation services during therapy. Compared to those without rehabilitation, the rehabilitation group (n = 110; 46.4%) had more caregiver reported problems with attention (Z = -0.28 vs. 0.43, p = 0.022), executive function (Z = -0.50 vs. -0.08, p = 0.003), and adaptive skills (Z = -0.41 vs.-0.13, p = 0.031). Among the rehabilitation group, there was no difference in outcomes by insurance status. Among those without rehabilitation, those with public insurance had worse neurocognitive outcomes than those with private insurance in IQ (Z = -0.04 vs. -0.45, p = 0.0115), processing speed (Z = -0.10 vs. -0.75, p = 0.0030), reading (Z = 0.18 vs. -0.59, p < 0.0001), and math (Z = -0.04 vs. -0.50, p = 0.0021).

CONCLUSION

Participation in rehabilitation services during early intensive therapy is associated with end of therapy caregiver-reported neurocognitive outcomes in daily life.