Electronic Patient Data Research Articles

Telehealth-based diagnostic testing in general practice during the COVID-19 pandemic: an observational study.

BackgroundSince the World Health Organization declared COVID-19 a pandemic on 11 March 2020, health technologies have been rapidly scaled up to ensure access to care. A significant innovation has been telehealth in general practice. Now widespread, it remains unknown how this shift to virtual care has impacted on quality-of-care indicators such as pathology testing and diagnosis.AimTo undertake a comparison of telehealth and face-to-face general practice consultations to: identify if there were differences in the proportion of pathology test referrals from 2019–2020; and quantify any change in pathology test collection and follow-up patterns.Design & settingRetrospective observational study of routinely collected electronic patient data from 807 general practices across New South Wales (NSW) and Victoria, Australia.MethodMultivariate generalised estimating equation models were used to estimate the proportion of pathology test referrals for overall, face-to-face, and telehealth consultations. Pathology test follow-up was described through median (and interquartile range [IQR]) time.ResultsPathology test referrals declined during periods of high COVID-19 cases, falling from 10.8% in February 2020 to a low of 4.5% during the first peak in April. Overall, pathology test referrals were lower for telehealth than face-to-face consultations. Median time between referral and test collection was 3 days (IQR 1–14) for telehealth and 1 day (IQR 0–7) for face to face.ConclusionFor telehealth to become part of routine care, it is crucial that gaps in functionality, including difficulty in test referral processes, be addressed. Quality improvements supporting care practices will ensure clinicians’ workflows are supported and patients receive diagnostic testing.

Sex Differences in Anemia and Red Blood Cell Transfusions Among General Internal Medicine Inpatients

Background: Epidemiological data from the general population have shown that women have lower hemoglobin values compared to men. Previous research on surgical patients showed that women were significantly more likely to receive red blood cell (RBC) transfusions compared to men in the peri-operative period due to multiple factors including the application of absolute transfusion thresholds. Sex differences in RBC transfusion frequencies have not been well studied in non-surgical settings. Here we investigated sex differences in anemia and in RBC transfusion frequencies in patients admitted to General Internal Medicine (GIM) wards.Study design and methods: This was a retrospective cohort study using de-identified electronic patient data from all patients admitted to the GIM wards from the emergency department across 4 different hospitals between 2010 and 2017, who had at least 1 CBC test during their admission. Patients with multiple hospital admissions during the study period, those who were pregnant, those who underwent surgery during the admission, and those transferred to or from another service besides GIM during their admission were excluded. Severity of anemia was stratified using the WHO sex-specific quantitative thresholds for hemoglobin: mild anemia (110 to 119 g/L in women and 110 to 129 g/L in men), moderate anemia (80 to 109 g/L in both men and women) and severe anemia (lower than 80 g/L in both men and women).Results: The study cohort consisted of 51,073 inpatients (24,666 men and 26,407 women). The median hemoglobin at the time of admission was significantly higher (p < 0.001) in men (median 135 g/L, IQR 31) compared to women (median 127 g/L, IQR 25). The prevalence of moderate and severe anemia was significantly higher (p < 0.001) in female inpatients (56%) compared to males (44%). The proportion of patients who received RBC transfusions during their admission was not significantly different (p = 0.36) between men (49%) and women (51%). There was no significant sex difference (p = 0.28) observed in the median hemoglobin prior to RBC transfusions between men (median 68 g/L, IQR 10.5) and women (median 68 g/L, IQR 11). The number of patients with pre-transfusion hemoglobin less than 70 g/L, the commonly used threshold for transfusion in the in-patient setting, was not significantly different (p = 0.91) between men and women. However, among the patients with hemoglobin < 70 g/L, a significantly lower proportion (p < 0.001) of women (82%) received RBC transfusions compared to men (91%).Conclusions: In a large cohort of GIM inpatients, we observed significantly higher prevalence of moderate and severe anemia in women compared to men. While there was no sex difference in overall RBC transfusion frequencies in this inpatient population, there was a lower proportion of women receiving RBC transfusions compared to men in the subgroup of patients with hemoglobin less than 70 g/L. DisclosuresNo relevant conflicts of interest to declare.

An atomic approach to the design and implementation of a research data warehouse.

ObjectiveAs a long-standing Clinical and Translational Science Awards (CTSA) Program hub, the University of Pittsburgh and the University of Pittsburgh Medical Center (UPMC) developed and implemented a modern research data warehouse (RDW) to efficiently provision electronic patient data for clinical and translational research.Materials and MethodsWe designed and implemented an RDW named Neptune to serve the specific needs of our CTSA. Neptune uses an atomic design where data are stored at a high level of granularity as represented in source systems. Neptune contains robust patient identity management tailored for research; integrates patient data from multiple sources, including electronic health records (EHRs), health plans, and research studies; and includes knowledge for mapping to standard terminologies.ResultsNeptune contains data for more than 5 million patients longitudinally organized as Health Insurance Portability and Accountability Act (HIPAA) Limited Data with dates and includes structured EHR data, clinical documents, health insurance claims, and research data. Neptune is used as a source for patient data for hundreds of institutional review board-approved research projects by local investigators and for national projects.DiscussionThe design of Neptune was heavily influenced by the large size of UPMC, the varied data sources, and the rich partnership between the University and the healthcare system. It includes several unique aspects, including the physical warehouse straddling the University and UPMC networks and management under an HIPAA Business Associates Agreement.ConclusionWe describe the design and implementation of an RDW at a large academic healthcare system that uses a distinctive atomic design where data are stored at a high level of granularity.

Implementation of a cloud-based electronic patient-reported outcome (ePRO) platform in patients with advanced cancer

BackgroundPatient reported outcomes (PROs) have been associated with improved symptom management and quality of life in patients with cancer. However, the implementation of PROs in an academic clinical practice has not been thoroughly described. Here we report on the execution, feasibility and healthcare utilization outcomes of an electronic PRO (ePRO) application for cancer patients at an academic medical center.MethodsWe conducted a randomized trial comparing an experimental ePRO arm to standard of care in patients with advanced cancer in the thoracic, gastrointestinal, and genitourinary oncology groups at Stanford Cancer Center from March 2018 to November 2019. We describe the pre-implementation, implementation, and post-implementation phases of the ePRO arm, technological barriers, electronic health record (EHR) integration, clinician burden, and patient data privacy and security. Feasibility was pre-specified to be at least 70% completion of all questionnaires. Acceptability was based on patient and clinician feedback. Ambulatory healthcare utilization was assessed by reviewing numbers of phone messages, electronic portal messages, and referrals for supportive care.ResultsOf 617 ePRO questionnaires sent to 72 patients, 445 (72%) were completed. Most clinicians (87.5%) and patients (93%) felt neutral or positive about the ePRO tool’s ease of use. Exposure to ePRO did not cause a measurable change in ambulatory healthcare utilization, with a median of less than two phone messages and supportive care referrals, and 5–6 portal messages.ConclusionsWeb-based ePRO tools for patients with advanced cancer are feasible and acceptable without increasing clinical burden. Key lessons include the importance of pilot testing, engagement of stakeholders at all levels, and the need for customization by disease group. Future directions for this work include completion of EHR integration, expansion to other centers, and development of integrated workflows for routine clinical practice.

Regular intake of energy drinks and multivitamin supplements is associated with elevated plasma vitamin B6 levels in post-bariatric patients

The aim of the present survey was to analyze plasma vitamin B6 levels in post-bariatric patients and to elucidate the causal factors associated with elevated plasma vitamin B6 levels. This is a retrospective analysis of electronic patient data of all post-bariatric patients evaluated at the endocrine outpatient clinic of the University Hospital Basel in 2017, for which plasma vitamin B6 values were assessed during regular follow-up visits. In total, 205 patients were included in the study, whereof a minority of 43% had vitamin B6 levels in the normal range. 50% of the patients had vitamin B6 levels up to fourfold higher than the upper normal limit and 7% had levels more than fourfold above the upper normal limit. Vitamin B6 deficiency was not observed in any patient. While multivitamin supplementation in general was associated with elevated plasma vitamin B6 levels, the highest vitamin B6 levels were found after biliopancreatic diversion (BPD) and in patients who reported daily energy drink intake. Elevated plasma vitamin B6 levels up to fourfold above the upper normal limit are common in postbariatric patients and are associated with regular multivitamin supplementation, while highly elevated plasma vitamin B6 levels were seen primarily upon regular energy drink intake. Thus, a regular follow-up of vitamin B6 plasma levels and critical evaluation of vitamin B6 supplementation, either as part of the multivitamin preparation or related to regular energy drink intake, is highly warranted and should be an integral part of the routine post-bariatric follow-up.

Machine learning in medicine: It has arrived, let's embrace it.

Machine learning and artificial intelligence (AI) have arrived in medicine and the healthcare community is experiencing significant growth in their adoption across numerous patient care settings. There are countless applications for machine learning and AI in medicine ranging from patient outcome prediction, to clinical decision support, to predicting future patient therapeutic setpoints. This commentary discusses a recent application leveraging machine learning to predict one-year patient survival following orthotopic heart transplantation. This modeling approach has significant implications in terms of improving clinical decision-making, patient counseling, and ultimately organ allocation and has been shown to significantly outperform pre-existing algorithms. This commentary also discusses how adoption and advancement of this modeling approach in the future can provide increased personalization of patient care. The continued expansion of information systems and growth of electronic patient data sources in health care will continue to pave the way for increased use and adoption of data science in medicine. Personalized medicine has been a long-standing goal of the healthcare community and with machine learning and AI now being continually incorporated into clinical settings and practice, this technology is well on the pathway to make a considerable impact to greatly improve patient care in the near future.

Epidemiology and Management of Acquired Hemophilia A in a Single Hospital System

Introduction: Acquired Hemophilia A (AHA) is an uncommon autoimmune disease that occurs when autoantibodies form against intrinsic factor VIII. Cases are typically identified due to isolated prolongation of the partial thromboplastin time (PTT) or unexplained bleeding diathesis, prompting further evaluation with PTT mixing studies and the Bethesda assay. AHA is historically associated with a high mortality rate; however, when appropriately diagnosed and treated, modern mortality from acute bleeding is approximately 4.5% (Knoebl et al Journal of Thrombosis and Haemostasis 2012). This study aims to explore the epidemiology, management, and outcomes in cases of AHA found within our hospital system to guide future care.Methods: The study design involves identifying all diagnosed cases of AHA within our hospital system using diagnosis codes and use of specific laboratory tests. The study design was presented to the University of Illinois College of Medicine at Peoria Internal Review Board and granted exempt status. 1,931 patients were identified, of which 6 patients met criteria for diagnosis of AHA.Results: The initial analysis reveals that the study population had a median age of 79.5 years old (high: 90; low: 61). All the patients presented with bleeding symptoms and 5 of 6 had major bleeding requiring transfusion. 5 of the 6 patients also developed soft tissue bleeding with hematoma or hemarthrosis. Underlying diagnosis of cancer was the suspected etiology in 4 of 6 cases. The median initial Bethesda assay value was 106.5 Bethesda units (high: 180; low: 33) and the median initial factor VIII activity assay result was less than 1%. The median PTT mixing study result was 51.5 seconds (high: 66.7; low: 44). In 3 of the 6 cases, patient survival was less than 1 year from onset of bleeding symptoms. 1 patient died prior to initiation of a bypassing agent. All the remaining patients were treated with the bypassing agent recombinant factor VIIA (NovoSeven ®). 2 of those 5 cases required management with a second line bypassing agent and in both cases porcine recombinant factor VIII (Obizur ®) was used and successfully achieved hemostasis. Remission with complete eradication of the acquired factor inhibitor was achieved in 4 of 5 cases receiving a bypassing agent with subsequent immunosuppressive therapy. Prednisone was used in all cases, with two using a second immunosuppressant (cyclophosphamide and rituximab). Of the 4 patients achieving remission, all eventually had recurrence of bleeding. Only 1 patient is still alive and none of the other patients survived to 5 years from initial presentation. 4 of those 5 patients died from direct complications of bleeding.Conclusions: Despite having access to electronic patient data from one of the largest hospital systems in Illinois, only 6 diagnoses were confirmed. We suspect that AHA is significantly underdiagnosed and that many patients developing factor inhibitors die before a full workup can be performed. The EACH2 registry changed how we view AHA; however, that analysis of that database must continue to be supported by population data from centers not included in that study (Knoebl et al Journal of Thrombosis and Haemostasis 2012). Furthermore, analysis of outcomes with newer bypassing agents and more sophisticated use of these agents is ongoing. We plan to publish the full results of our study, including a cost-benefit analysis of these 6 cases. We hope that our patient data can be compounded with other recent studies to help to shape future management decisions and guidelines. DisclosuresNo relevant conflicts of interest to declare.

Structure and concept of ICU rounds: the VIS-ITS survey.

ObjectiveTo gather data about structural and procedural characteristics of patient rounds in the intensive care unit (ICU) setting.DesignA structured online survey was offered to members of two German intensive care medicine societies.Measurements and main resultsIntensivists representing 390 German ICUs participated in this study (university hospitals 25%, tertiary hospitals 23%, secondary hospitals 36%, primary hospitals 16%). In 90% of participating ICUs, rounds were reported to take place in the morning and cover an average of 12 intensive care beds and 6 intermediate care beds within 60 min. With an estimated bed occupancy of 80%, this averaged to 4.3 min spent per patient during rounds. In 96% of ICUs, rounds were stated to include a bedside visit. On weekdays, 86% of the respondents reported holding a second ICU round with the attendance of a qualified decision-maker (e.g. board-certified intensivist). On weekends, 79% of the ICUs performed at least one round with a decision-maker per day. In 18%, only one ICU round per weekend was reported, mostly on Sundays. The highest-qualified decision-maker present during rounds on most ICUs was an ICU attending (57%). Residents (96%) and intensive care nurses (87%) were stated to be always or usually present during rounds. In contrast, physiotherapists, respiratory therapists or medical specialists such as pharmacists or microbiologist were not regular members of the rounding team on most ICUs. In the majority of cases, the participants reported examining the medical chart directly before or during the bedside visit (84%). An electronic patient data management system (PDMS) was available on 31% of ICUs. Daily goals were always (55%) or usually (39%) set during rounds.ConclusionThis survey gives a broad overview of the structure and processes of ICU rounds in different sized hospitals in Germany. Compared to other mostly Anglo-American studies, German ICU rounds appear to be shorter and less interdisciplinary.

Evaluation of comprehensiveness and reliability of electronic health records concerning resuscitation efforts within academic intensive care units: a retrospective chart analysis

BackgroundAccording to the literature, the validity and reliability of medical documentation concerning episodes of cardiopulmonary resuscitation (CPR) is suboptimal. However, little is known about documentation quality of CPR efforts during intensive care unit (ICU) stays in electronic patient data management systems (PDMS). This study analyses the reliability of CPR-related medical documentation within the ICU PDMS.MethodsIn a retrospective chart analysis, PDMS records of three ICUs of a single university hospital were searched over 5 y for CPR check marks. Respective datasets were analyzed concerning data completeness and data consistency by comparing the content of three documentation forms (physicians’ log, nurses’ log, and CPR incident form), as well as physiological and therapeutic information of individual cases, for missing data and plausibility of CPR starting time and duration. To compare data reliability and completeness, a quantitative measure, the Consentaneity Index (CI), is proposed.ResultsOne hundred sixty-five datasets were included into the study. In 9% (n = 15) of cases, there was neither information on the time points of CPR initiation nor on CPR duration available in any data source. Data on CPR starting time and duration were available from at least two data sources in individual cases in 54% (n = 90) and 45% (n = 74), respectively. In these cases, the specifications of CPR starting time did differ by a median ± interquartile range of 10.0 ± 18.5 min, CPR duration by 5.0 ± 17.3 min. The CI as a marker of data reliability revealed a low consistency of CPR documentation in most cases, with more favorable results, if the time interval between the CPR episode and the time of documentation was short.ConclusionsThis study reveals relevant proportions of missing and inconsistent data in electronic CPR documentation in the ICU setting. The CI is suggested as a tool for documentation quality analysis and monitoring of improvements.

A retrospective study of gastrointestinal disorders in a predominantly Austrian leisure horse referral hospital population

SummaryThe aim of our retrospective study was to produce and analyse epidemiological data on gastrointestinal (GI) disorders of Austrian leisure horses. Case records of horses referred to the Equine Clinic of the Veterinary University of Vienna were collected from an electronic hospital information system using its search options. All cases reported to have suffered a GI disorder between March 2003 and February 2015 were selected. In this period of 4314 days, data of 1833 GI cases, mostly accompanied by colic signs, were analysed. The vast majority of our GI cases appeared in the abdominal parts of the GI tract 96% (1760/1833) while just 1.3% (24/1833) could be definitively attributed to oesophageal disorders, whereas in 2.7% (49/1833) of the cases a definitive diagnosis could not be made. The highest prevalence of GI disorders was seen in July and October/November. A breakdown of figures for cases clustered under the colic syndrome showed that colon impactions in left ventral large and/or pelvic flexure (35%) were the major problem followed by gastric disorders (18%), spasmodic colic (9%), enteritis (7%), caecal tympany (6%); caecal impaction (4%), colitis (3%), abnormal anatomical positions of large colon including complete and partial strangulation/volvulus, nephrosplenic entrapment of the left ventral and dorsal colon and right dorsal displacement of the large colon (3%) and small intestinal volvulus (2.7%). In 12% of the cases, the cause of the abdominal discomfort was likely idiopathic paralytic ileus. The searching tool of the electronic patient data also brought up cases with malabsorbtion/IBD syndrome and GI ruptures, both these contributed less than 1% of the referred GI cases. Results suggested that the causes of colic are diverse, but the major reasons for referred cases of colic in the east Austrian horse population are large colon impaction or displacements. Seasonal distribution and specific causes of gastrointestinal disorders should be considered essential components of the initial clinical assessment of horses presenting with colic.

Derivation and validation of a machine learning risk score using biomarker and electronic patient data to predict progression of diabetic kidney disease

AimPredicting progression in diabetic kidney disease (DKD) is critical to improving outcomes. We sought to develop/validate a machine-learned, prognostic risk score (KidneyIntelX™) combining electronic health records (EHR) and biomarkers.MethodsThis is an observational cohort study of patients with prevalent DKD/banked plasma from two EHR-linked biobanks. A random forest model was trained, and performance (AUC, positive and negative predictive values [PPV/NPV], and net reclassification index [NRI]) was compared with that of a clinical model and Kidney Disease: Improving Global Outcomes (KDIGO) categories for predicting a composite outcome of eGFR decline of ≥5 ml/min per year, ≥40% sustained decline, or kidney failure within 5 years.ResultsIn 1146 patients, the median age was 63 years, 51% were female, the baseline eGFR was 54 ml min−1 [1.73 m]−2, the urine albumin to creatinine ratio (uACR) was 6.9 mg/mmol, follow-up was 4.3 years and 21% had the composite endpoint. On cross-validation in derivation (n = 686), KidneyIntelX had an AUC of 0.77 (95% CI 0.74, 0.79). In validation (n = 460), the AUC was 0.77 (95% CI 0.76, 0.79). By comparison, the AUC for the clinical model was 0.62 (95% CI 0.61, 0.63) in derivation and 0.61 (95% CI 0.60, 0.63) in validation. Using derivation cut-offs, KidneyIntelX stratified 46%, 37% and 17% of the validation cohort into low-, intermediate- and high-risk groups for the composite kidney endpoint, respectively. The PPV for progressive decline in kidney function in the high-risk group was 61% for KidneyIntelX vs 40% for the highest risk strata by KDIGO categorisation (p < 0.001). Only 10% of those scored as low risk by KidneyIntelX experienced progression (i.e., NPV of 90%). The NRIevent for the high-risk group was 41% (p < 0.05).ConclusionsKidneyIntelX improved prediction of kidney outcomes over KDIGO and clinical models in individuals with early stages of DKD.Graphical abstract

Probabilistic transmission models incorporating sequencing data for healthcare-associated Clostridioides difficile outperform heuristic rules and identify strain-specific differences in transmission.

Fitting stochastic transmission models to electronic patient data can offer detailed insights into the transmission of healthcare-associated infections and improve infection control. Pathogen whole-genome sequencing may improve the precision of model inferences, but computational constraints have limited modelling applications predominantly to small datasets and specific outbreaks, whereas large-scale sequencing studies have mostly relied on simple rules for identifying/excluding plausible transmission. We present a novel approach for integrating detailed epidemiological data on patient contact networks in hospitals with large-scale pathogen sequencing data. We apply our approach to study Clostridioides difficile transmission using a dataset of 1223 infections in Oxfordshire, UK, 2007–2011. 262 (21% [95% credibility interval 20–22%]) infections were estimated to have been acquired from another known case. There was heterogeneity by sequence type (ST) in the proportion of cases acquired from another case with the highest rates in ST1 (ribotype-027), ST42 (ribotype-106) and ST3 (ribotype-001). These same STs also had higher rates of transmission mediated via environmental contamination/spores persisting after patient discharge/recovery; for ST1 these persisted longer than for most other STs except ST3 and ST42. We also identified variation in transmission between hospitals, medical specialties and over time; by 2011 nearly all transmission from known cases had ceased in our hospitals. Our findings support previous work suggesting only a minority of C. difficile infections are acquired from known cases but highlight a greater role for environmental contamination than previously thought. Our approach is applicable to other healthcare-associated infections. Our findings have important implications for effective control of C. difficile.

Effects of the elixhauser comorbidity index on the time of discharge and 30-day mortality rate in geriatric and non-geriatric patients undergoing partial hip prosthesis arthroplasty

In this study, we aimed to analyse the non-routine discharge period of the ECI and its predicted effect on the 30-day mortality rate in geriatric and non-geriatric patients admitted for partial hip replacement. 215 patients who underwent elective partial hip arthroplasty under spinal anaesthesia in our tertiary hospital were included in the study. This was a single-centre retrospective descriptive study. Data were collected by reviewing the electronic patient data system and patient files. The patients were divided into two groups as non-geriatric (18-64-year-old, n=125) and geriatric (≥65-year-old, n=90). The impact of the Elixhauser Comorbidity Index (ECI) on postoperative mortality rates and the discharge time from the hospital was analysed by comparing the age groups. The ECI percentage values (80.86 ± 3.99) of recovered patients were higher than patients discharged as exitus (71.5 ± 7.74) (p 0.05). When ECI≤77.8%, it predicts mortality with 77.8% sensitivity and 85.92% specificity. The Elixhauser Comorbidity index could not predict the time of discharge due to complications developed after transfer of patients to the postoperative intensive care unit. There was no significant correlation between ECI and discharge time in non-geriatric and geriatric patients.

Monitoring cardiovascular disease risk in individuals with severe mental illness in an inpatient mental health setting: a secondary data analysis

Background/Aims Life expectancy in people with lived experience of mental health conditions is reduced by up to 25 years; this is from preventable physical medical comorbidities and multi-morbidities such as cardiovascular disease, diabetes, cancers and smoking-related lung disease. Two-thirds of these deaths are avoidable if people with severe mental illness are offered prompt physical screening checks and monitoring. The aim of this article was to explore barriers to the management of cardiovascular disease risk on inpatient wards and make recommendations in relation to cardiovascular disease risk management in people with severe mental illness. Methods A structured MS Excel extraction data tool informed by best practice guidance was developed and used to extract electronic patient data on screening and monitoring of cardiovascular disease risk factors (blood pressure, smoking, alcohol, lipids, body mass index/weight, blood glucose level) across 10 inpatient psychiatric wards within one London mental health trust. A target sample of 245 electronic records of patients with severe mental illness discharged between 25 August 2018 and 13 February 2019 with length of inpatient stay &gt;40 days was examined. Simple random sampling (MS Excel random number generator) was used to select a final sample of 120 electronic records. All the included samples had been prescribed psychotropic medication. Results Regarding patient demographics, there was an inverse correlation with age, with a greater proportion of inpatients being of a younger age: 51% aged 18–39 years compared with 14% aged 60–79 years. The study found an average of 71% compliance of the documentation of data on all individual parameters (smoking, alcohol, body mass index, blood pressure, serum glucose, serum lipids, electrocardiogram) at baseline. Results showed an average of 79% compliance for monitoring review at least once across the parameters within 3 months of admission. Conclusions It is recommended as a minimum for individuals with severe mental illness under the care of mental health services and/or taking psychotropic medication to have regular cardiometabolic risk assessment and management of risk at the point of entry into services and a review for weight, waist circumference, blood glucose checks, lipid profile, blood pressure, lifestyle choice behaviours and personal assessment of cardiovascular disease. Although progress is being made across provider services to implement the above, gaps in practice are still evident, as demonstrated in these findings.

Health information systems in Jordan and Palestine: the need for health informatics training.

Some Arab countries have health information systems (HIS) in place but they lack well trained IT staff. Poor management and lack of appreciation of the importance of HIS are major barriers to development and adoption of HIS in Arab hospitals. This research is part of a survey carried out to determine health informatics (HI) use and to assess the training needs of health professionals in Jordan and Palestine. A survey was conducted in 2017 among employees in all health professions at 14 hospitals in Jordan and Palestine to assess their use of the HI system and to assess the HI skills needed in both countries. The majority of respondents reported that their hospital departments were employing computer systems to run services. More than half had received training in computer skills but also half said they needed specialized training in HI. Between 58.0% and 73.6% agreed that their hospitals provided the necessary support to operate HI systems. The vast majority (86.0%) of health professionals reported that they needed skills to monitor diagnosis and treatment, including access to clinical findings. Other skills needed included using shared hospital services (85.6%), using medical records (84.7%), managing electronic patient data (84.5%), using patient medical records to conduct clinical research (83.4%), and using tele-care services and technologies effectively (75.9%). Health professionals in Palestine and Jordan are in need of training in HI and therefore educational programmes in the area of HI are strongly recommended.

Use of electronic patient data storage for evaluating and setting the risk category of late effects in childhood cancer survivors.

Many of the late effects of cancer treatment in childhood may occur even decades after the treatment, and only a minority of the survivors remain as healthy as their peers. Providing appropriate long-term care for childhood cancer survivors after transition to primary health care is a challenge. Both survivors and primary care providers need information on potential late effects. The lack of a systematic late effect follow-up plan may lead to excessive use of health care services or delayed intervention. While manual compilation of individual follow-up plans is time consuming for experienced clinicians, electronic algorithms may be feasible. In Finland, international guidelines for determining the risk of late effects have been implemented. Nationally, Turku University Hospital was asked with developing an automatized system for calculating the risk of late effects, based on electronic patient records saved in the hospital data lake. An electronic algorithm that uses details from exposure-based health screening guidelines published by the Children's Oncology Group was created. The results were compared with those manually extracted by an experienced clinician. Significant concordance between the manual and algorithm-based risk classification was found. A total of 355 patients received a classification using the algorithm, and 325 of those matched with the manual categorization, producing a Cohen's coefficient of 0.91 (95% confidence interval 0.88-0.95). Automated algorithms can be used to categorize childhood cancer survivors efficiently and reliably into late effect risk groups. This further enables automatized compilation of appropriate individual late effect follow-up plan for all survivors.

Monitoring cardiovascular disease risk in individuals with severe mental illness in an inpatient mental health setting: a secondary data analysis

Background/Aims Life expectancy in people with lived experience of mental health conditions is reduced by up to 25 years; this is from preventable physical medical comorbidities and multi-morbidities such as cardiovascular disease, diabetes, cancers and smoking-related lung disease. Two-thirds of these deaths are avoidable if people with severe mental illness are offered prompt physical screening checks and monitoring. The aim of this article was to explore barriers to the management of cardiovascular disease risk on inpatient wards and make recommendations in relation to cardiovascular disease risk management in people with severe mental illness. Methods A structured MS Excel extraction data tool informed by best practice guidance was developed and used to extract electronic patient data on screening and monitoring of cardiovascular disease risk factors (blood pressure, smoking, alcohol, lipids, body mass index/weight, blood glucose level) across 10 inpatient psychiatric wards within one London mental health trust. A target sample of 245 electronic records of patients with severe mental illness discharged between 25 August 2018 and 13 February 2019 with length of inpatient stay &gt;40 days was examined. Simple random sampling (MS Excel random number generator) was used to select a final sample of 120 electronic records. All the included samples had been prescribed psychotropic medication. Results Regarding patient demographics, there was an inverse correlation with age, with a greater proportion of inpatients being of a younger age: 51% aged 18–39 years compared with 14% aged 60–79 years. The study found an average of 71% compliance of the documentation of data on all individual parameters (smoking, alcohol, body mass index, blood pressure, serum glucose, serum lipids, electrocardiogram) at baseline. Results showed an average of 79% compliance for monitoring review at least once across the parameters within 3 months of admission. Conclusions It is recommended as a minimum for individuals with severe mental illness under the care of mental health services and/or taking psychotropic medication to have regular cardiometabolic risk assessment and management of risk at the point of entry into services and a review for weight, waist circumference, blood glucose checks, lipid profile, blood pressure, lifestyle choice behaviours and personal assessment of cardiovascular disease. Although progress is being made across provider services to implement the above, gaps in practice are still evident, as demonstrated in these findings.

Effect of COVID-19 on platelet count and its indices.

Easily accessible, inexpensive, and widely used laboratory tests that demonstrate the severity of COVID-19 are important. Therefore, in this study, we aimed to investigate the relationship between mortality in COVID-19 and platelet count, Mean Platelet Volume (MPV), and platelet distribution width. In total, 215 COVID-19 patients were included in this study. The patients were divided into two groups. Patients with room air oxygen saturation < 90% were considered as severe COVID-19, and patients with ≥90% were considered moderate COVID-19. Patient medical records and the electronic patient data monitoring system were examined retrospectively. Analyses were performed using the SPSS statistical software. A p-value <0.05 was considered significant. The patients' mean age was 64,32 ± 16,07 years. According to oxygen saturation, 81 patients had moderate and 134 had severe COVID-19. Our findings revealed that oxygen saturation at admission and the MPV difference between the first and third days of hospitalization were significant parameters in COVID-19 patients for predicting mortality. While mortality was 8.4 times higher in patients who had oxygen saturation under 90 % at hospital admission, 1 unit increase in MPV increased mortality 1.76 times. In addition to the lung capacity of patients, the mean platelet volume may be used as an auxiliary test in predicting the mortality in COVID-19 patients.

Understanding enterprise data warehouses to support clinical and translational research.

Among National Institutes of Health Clinical and Translational Science Award (CTSA) hubs, adoption of electronic data warehouses for research (EDW4R) containing data from electronic health record systems is nearly ubiquitous. Although benefits of EDW4R include more effective, efficient support of scientists, little is known about how CTSA hubs have implemented EDW4R services. The goal of this qualitative study was to understand the ways in which CTSA hubs have operationalized EDW4R to support clinical and translational researchers. After conducting semistructured interviews with informatics leaders from 20 CTSA hubs, we performed a directed content analysis of interview notes informed by naturalistic inquiry. We identified 12 themes: organization and data; oversight and governance; data access request process; data access modalities; data access for users with different skill sets; engagement, communication, and literacy; service management coordinated with enterprise information technology; service management coordinated within a CTSA hub; service management coordinated between informatics and biostatistics; funding approaches; performance metrics; and future trends and current technology challenges. This study is a step in developing an improved understanding and creating a common vocabulary about EDW4R operations across institutions. Findings indicate an opportunity for establishing best practices for EDW4R operations in academic medicine. Such guidance could reduce the costs associated with developing an EDW4R by establishing a clear roadmap and maturity path for institutions to follow. CTSA hubs described varying approaches to EDW4R operations that may assist other institutions in better serving investigators with electronic patient data.

Interpretable clinical prediction via attention-based neural network

BackgroundThe interpretability of results predicted by the machine learning models is vital, especially in the critical fields like healthcare. With the increasingly adoption of electronic healthcare records (EHR) by the medical organizations in the last decade, which accumulated abundant electronic patient data, neural networks or deep learning techniques are gradually being applied to clinical tasks by utilizing the huge potential of EHR data. However, typical deep learning models are black-boxes, which are not transparent and the prediction outcomes of which are difficult to interpret.MethodsTo remedy this limitation, we propose an attention neural network model for interpretable clinical prediction. In detail, the proposed model employs an attention mechanism to capture critical/essential features with their attention signals on the prediction results, such that the predictions generated by the neural network model can be interpretable.ResultsWe evaluate our proposed model on a real-world clinical dataset consisting of 736 samples to predict readmissions for heart failure patients. The performance of the proposed model achieved 66.7 and 69.1% in terms of accuracy and AUC, respectively, and outperformed the baseline models. Besides, we displayed patient-specific attention weights, which can not only help clinicians understand the prediction outcomes, but also assist them to select individualized treatment strategies or intervention plans.ConclusionsThe experimental results demonstrate that the proposed model can improve both the prediction performance and interpretability by equipping the model with an attention mechanism.