Electronic Patient Management System Research Articles

Machine learning in clinical practice: Evaluation of an artificial intelligence tool after implementation.

Artificial intelligence (AI) has gradually found its way into healthcare, and its future integration into clinical practice is inevitable. In the present study, we evaluate the accuracy of a novel AI algorithm designed to predict admission based on a triage note after clinical implementation. This is the first of such studies to investigate real-time AI performance in the emergency setting. The novel AI algorithm that predicts admission using a triage note was translated into clinical practice and integrated within St Vincent's Hospital Melbourne's electronic emergency patient management system. The data were collected from 1 January 2021 to 17 August 2022 to evaluate the diagnostic accuracy of the AI system after implementation. A total of 77 125 ED presentations were included. The live AI algorithm has a sensitivity of 73.1% (95% confidence interval 72.5-73.8), specificity of 74.3% (73.9-74.7), positive predictive value of 50% (49.6-50.4) and negative predictive value of 88.7% (88.5-89) with a total accuracy of 74% (73.7-74.3). The accuracy of the system was at the lowest for admission to psychiatric units (34%) and at the highest for gastroenterology and medical admission (84% and 80%, respectively). Our study showed the diagnostic evaluation of a real-time AI clinical decision-support tool became less accurate than the original. Although real-time sensitivity and specificity of the AI tool was still acceptable as a decision-support tool in the ED, we propose that continuous training and evaluation of AI-enabled clinical support tools in healthcare are conducted to ensure consistent accuracy and performance to prevent inadvertent consequences.

301 Role of comprehensive geriatric assessment in determining appropriateness of admission to long term care

Abstract Background The benefit of comprehensive geriatric assessment (CGA) in supporting discharge to home from hospital for older people has been identified. When discharge to long term care (LTC) is considered by generalist teams, through determination of potential for further remediation and support, as well as vigilance for elder abuse and inappropriate third-party pressure, CGA supports continued living at home. Although European guidelines recommend, and Irish policy mandates, CGA prior to consideration of entry to LTC, it is not clear to what extent this is routinely undertaken. We aimed to determine the value of CGA in supported continued living at home when entry to LTC is proposed by referring clinical teams. Methods Review of electronic system of geriatric medicine consultations for assessment for LTC in a Level 4 teaching hospital. Information on discharge destination was retrieved from discharge summaries and electronic patient management systems. Results Of 740 consultations over 14 months between 2020 and 2021, 99 were for assessment for LTC. At review at 8 months, of 49 (49%) discharges to LTC, 42 (42%) were discharged as new residents, 4 (4%) were existing residents and 3 (3%) were discharged for end of life care. Of those for whom CGA recommended alternatives, 22 (22%) went directly home, 7 (7%) were discharged to offsite rehabilitation, 10 (10%) remained inpatients, and 8 (8%) died during admission. Conclusion This review highlights the importance and value of the role of CGA in planning discharge destination, a key if under-emphasised role for Irish geriatricians. Only 42% of the consults were discharged to LTC as a new permanent residence. This supports previous recommendations to routinely incorporate CGA when entry to LTC is considered, ensuring the most appropriate discharge destination for patients with significant multi-morbidity and complex care needs.

Usability testing of an electronic patient-reported outcome system linked to an electronic chemotherapy prescribing and patient management system for patients with cancer

BackgroundPeople affected by cancer experience a wide range of symptoms which have a major impact on their functioning and health-related quality of life (HRQoL). One way to measure the impact of cancer symptoms is through the use of patient-reported outcomes. MethodsAn electronic patient-reported outcome (ePRO) application (ChemoPRO®) was designed to be used by cancer patients to report their symptoms and communicate with their clinical team. Usability testing sessions were conducted with people with lived experience of cancer to understand how real users interact with the ChemoPRO® system. One-to-one testing sessions were conducted to assess use of the system and identify areas for further refinement. User satisfaction was assessed using a brief satisfaction questionnaire previously used by Aiyegbusi et al. (date). ResultsTen people with lived experience of cancer took part in the usability study. Symptoms and HRQoL measures, including the Euroqol EQ5D5L and the PRO-CTCAE™ were included in the ePRO system. Participantshad a mean age of 62.3 years. Three critical errors and 21 non-critical errors were reported. All participants were enthusiastic about the app. Participants liked the simplicity and responsiveness of the patient-facing app and highlighted the potential for communicating with their clinical team. The overall usability and satisfaction score was 4.5 (sd = 0.09). ConclusionThis usability study suggests that people with lived experience of cancer found the ChemoPRO® app acceptable and easy to use. One of the key features of this particular ePRO system that should be developed further is system functionality to facilitate communication between patients and clinicians. Future testing should include testing in a clinical setting and testing with people from ethnic minorities.

Dental practice to dental hospital referrals—upgraded to urgent suspected cancer pathways: A three year service review

AbstractIntroductionInappropriate General Dental Practitioner (GDP) patient referrals may delay cancer diagnoses, increasing the risk of late‐stage presentation. Urgent suspected cancer (USC) referrals from GDPs are reviewed swiftly by clinicians. Non‐USC referrals may wait up to a week before being reviewed by clinicians.Aims and objectivesTo investigate upgraded‐to‐USC GDP referrals to a University Dental Hospital (UDH) following the establishment of the pan‐Wales electronic referral management system (e‐RMS) in 2019. To examine reasons for their upgrade to the USC pathway, and to assess diagnostic outcomes, diagnostic delay and compliance with The National Institute for Health and Care Excellence (NICE) referral recommendations.MethodologyUpgraded‐to‐USC e‐Referrals were reviewed and cross‐referenced against the electronic patient management system and patient files to assess waiting times from referral date to first appointment. Each e‐Referral was reviewed to determine the reason for upgrade to a USC referral. Patient outcomes and diagnoses were also examined.Results83 patient records were identified. Upgraded e‐Referrals accounted for 18.6% of all USC e‐Referrals. The majority of upgraded e‐Referrals (66.3%) were upgraded because they included words describing key malignancy features. Malignant and oral potentially malignant disorders (OPMDs) accounted for 3.8% and 17.5% of diagnostic outcomes respectively. Overall, 18.8% of upgraded‐to‐USC e‐Referrals did not meet the NICE guideline requirement of having a first appointment within 2 weeks.ConclusionEarly diagnosis of malignancies is fundamental in improving oral cancer survival outcomes. Some GDPs' clinical judgement and referral pathway choice needs improvement. Reinforcing GDP education in regard to USC referrals will likely reduce inappropriate referrals.

Assessment of facility-level antiretroviral treatment patient status utilizing a national-level laboratory cohort: Toward an understanding of system-level tracking and clinic switching in South Africa.

Most estimates of HIV retention are derived at the clinic level through antiretroviral (ART) patient management systems, which capture ART clinic visit data, yet these cannot account for silent transfers across HIV treatment sites. Patient laboratory monitoring visits may also be observed in routinely collected laboratory data, which include ART monitoring tests such as CD4 count and HIV viral load, key to our work here. In this analysis, we utilized the NHLS National HIV Cohort (a system-wide viewpoint) to investigate the accuracy of facility-level estimates of retention in care for adult patients accessing care (defined using clinic visit data on patients under ART recorded in an electronic patient management system) at Themba Lethu Clinic (TLC). Furthermore, we describe patterns of facility switching among all patients and those patients classified as lost to follow-up (LTFU) at the facility level. Of the 43,538 unique patients in the TLC dataset, we included 20,093 of 25,514 possible patient records (78.8%) in our analysis that were linked with the NHLS National Cohort, and we restricted the analytic sample to patients initiating ART between 1 January 2007 and 31 December 2017. Most (60%) patients were female, and the median age (IQR) at ART initiation was 37 (31-45) years. We found the laboratory records augmented retention estimates by a median of 860 additional active records (about 8% of all median active records across all years) from the facility viewpoint; this augmentation was more noticeable from the system-wide viewpoint, which added evidence of activity of about one-third of total active records in 2017. In 2017, we found 7.0% misclassification at the facility-level viewpoint, a gap which is potentially solvable through data integration/triangulation. We observed 1,134/20,093 (5.6%) silent transfers; these were noticeably more female and younger than the entire dataset. We also report the most common locations for clinic switching at a provincial level. Integration of multiple data sources has the potential to reduce the misclassification of patients as being lost to care and help understand situations where clinic switching is common. This may help in prioritizing interventions that would assist patients moving between clinics and hopefully contribute to services that normalize formal transfers and fewer silent transfers.

OGBN P14 The development, introduction and outcomes of an electronic, patient record-based, Morbidity and Mortality (M&M) reporting system

Abstract Background Recording of patient complications and outcomes has suffered from the lack of a standardised reporting system. The aim of this project was to develop and introduce within our health board a generic M&M reporting tool for all surgical and medical specialities which would allow agreed outcomes and actions, following open discussion at M&M meetings, to be documented in the electronic patient record (EPR). Methods TrakCare® (Intersystems) is the electronic patient management system used within our region to record and store information for patient consultations, investigations and treatments. The criteria felt essential were for a reporting system that was suited to all disciplines, accessible to all levels of staff, interlinked between specialities and allowed analysis and reporting of outcomes. Access to all aspects of the EPR including results and imaging should be available during the M&M meeting to facilitate informed discussion. Agreement was reached on the required drop down menus which included patient details, background/event timeline, death certification, end-of-life care, clinical governance, lessons learned and actions and outcomes. ICD-10 codes were used for diagnosis and OPCS-4 coding for procedures. A complication list was developed with word recognition of either the organ system or the complication. Complications were graded according to the Clavien-Dindo classification. All entered data was analysed using Business Intelligence. Results The development process lasted 9 months with the input of the author and programmers from the TrakCare® team. The system was first introduced in September 2017 within the general surgical unit. 1855 patients have been discussed at the general surgical M&M meetings using the system. Since introduction, an additional 37 specialities have adopted the electronic M&M reporting tool. To date 4859 patients in total have been discussed. The introduction of the system has resulted in additional improvements to the M&M process including the requirement for a designated chairperson, a meeting attendance record and documentation of actions and outcomes from meetings. Automatically generated reports show all outstanding actions and recommendations identified at M&M meetings to ensure that the full process is completed. Conclusions The widespread adoption of a standardised M&M reporting tool clearly shows the willingness of care providers to engage in open discussion and learning from M&M meetings. The introduction of a standardised electronic reporting tool has led to additional improvements in the M&M process.

385 Service Evaluation of the Management of Acute Cholecystitis in the Royal Gwent Hospital Pre and During the COVID-19 Pandemic

Abstract Aim NICE guidelines set out the criteria for the treatment of patients with acute cholecystitis and the operative timescales for cholecystectomy. These targets were greatly affected during the Covid-19 pandemic. Therefore, we aimed to assess the impact that COVID-19 had on patients presenting with acute cholecystitis at a busy district general hospital. June 2020, compared with patients who presented with the same in June 2019. Method Patient cohorts were identified for matching seasons pre- and post-covid-19 (June 2019 and June 2020). Data of all patients who presented with acute cholecystitis was obtained using an electronic patient management system. Statistical analyses were performed using a Wilcoxon test. Results The results of the study indicate that waiting times post-covid are going down (p<0.05). Thus, days until cholecystectomy have decreased but the number of patients being operated on too has decreased thus further worsening waiting times for elective patients. The median and IQR's of days to surgery post-covid are 198 (121.5–278) and pre-covid are 251 (89.5–586.5). Presentations of gallstone complications almost doubled post-covid and the percentage of patients operated on decreased by over 20%. Conclusions It is clear from the data that the NICE guidance on the management of acute cholecystitis has been difficult to adhere to during the pandemic. While the time from diagnosis to operation has reduced post-covid the total number of operations has decreased drastically, putting further strain on elective waiting lists. This, inevitably, will result in further presentations of complications from gallstones and adverse patient outcomes.

MO923: Electronic Medication Record Accuracy and Clinical Pharmacist Intervention in Haemodialysis Outpatient Settings

Abstract BACKGROUND AND AIMS Haemodialysis (HD) units in Ireland operate a national electronic Kidney Disease Clinical Patient Management System (KDCPMS). KDCMPS is not always used as the primary electronic patient record (EPR) but in conjunction with other electronic and paper record systems across the healthcare setting. HD patients on average have 6 comorbidities and have the largest pill burden for any chronic disease; consuming 19 oral doses/day, comprising of 12 different medicines [1, 2]. Frequent medication changes, polypharmacy, comorbidities and non-adherence, increase the risk of drug-related problems (DRPs). In the HD population, DRPs are prevalent at a rate of 1 for every three medication exposures and can elicit negative outcomes, including worsening morbidity, mortality and increased healthcare expenditure [3]. In this study setting, KDCPMS information accuracy has not been examined to date. This study aims to describe medication discrepancies within KDCPMS records of HD outpatients. METHOD This prospective, observational study was conducted in the HD unit of Tallaght University Hospital, Dublin. Medicine reconciliation was conducted to identify KDCPMS discrepancies, followed by medication review to document DRPs. Recommendations were issued by the clinical pharmacist to resolve DRPs. RESULTS All KDCPMS records examined contained intentional and unintentional discrepancies (n = 36). Unintentional discrepancies corresponding to 8.8 discrepancies per patient (5.13SD) were observed. A total of 143 DRPs were identified in 34 patients (94.4%) (Table 1). Out of these, 65% of pharmacist recommendations were accepted (n = 93), 22.4% were rejected (n = 32), 8.4% (n = 12) were referred to the renal multidisciplinary team (MDT) and 4.2% were not actioned (n = 6). Discrepancies and DRPs by therapeutic area are shown in Fig. 1. Parenteral anticoagulants and thrombolytics were the most common undocumented intentional discrepancies (48.6%, n = 50/103) followed by iron (32%, n = 33/103). A total of 16 (44.4%) patients had at least one medicine de-prescribed. New prescriptions were issued for 26 patients (72.2%) for 81 medicines. CONCLUSION KDCPMS contains inaccuracies that could lead to systemic error. Robust clinical governance supported by the national policy is required to improve the accuracy of information contained in KDCPMS and support its use as the primary platform for renal patients. Specialist clinical pharmacists working collaboratively within the renal MDT reduce discrepancies, improve KDCPMS accuracy and resolve DRPs. Enhanced pharmaceutical care by specialist pharmacists should be supported within national models of care for chronic disease management to enhance patient outcomes.

Assessing the scalability of an integrated falls prevention service for community-dwelling older people: a mixed methods study

BackgroundThere is growing acknowledgement of the need for a phased approach to scaling up health interventions, beginning with an assessment of ‘scalability’, that is, the capacity of an individual intervention to be scaled up. This study aims to assess the scalability of a multi-component integrated falls prevention service for community-dwelling older people and to examine the applicability of the Intervention Scalability Assessment Tool (ISAT). The ISAT consists of 10 domains for consideration when determining the scalability of an intervention, and each domain comprises a series of questions aimed at examining readiness for scale-up.MethodsMultiple methods were used sequentially as recommended by the ISAT: a review of policy documents, results from a service evaluation and falls-related literature; one-to-one interviews (n = 11) with key stakeholders involved in management and oversight of the service; and a follow-up online questionnaire (n = 10) with stakeholders to rate scalability and provide further feedback on reasons for their scores.ResultsThree of the ISAT domains were rated highly by the participants. Analysis of the qualitative feedback and documents indicated that the issue of falls prevention among older people was of sufficient priority to warrant scale-up of the service and that the service aligned with national health policy priorities. Some participants also noted that benefits of the service could potentially outweigh costs through reduced hospital admissions and serious injuries such as hip fracture. The remaining domains received a moderate score from participants, however, indicating considerable barriers to scale-up. In the qualitative feedback, barriers identified included the perceived need for more healthcare staff to deliver components of the service, for additional infrastructure such as adequate room space, and for an integrated electronic patient management system linking primary and secondary care and to prevent duplication of services.ConclusionsPlans to scale up the service are currently under review given the practical barriers that need to be addressed. The ISAT provides a systematic and structured framework for examining the scalability of this multi-component falls prevention intervention, although the iterative nature of the process and detailed and technical nature of its questions require considerable time and knowledge of the service to complete.

Impact of Covid 19 preventive measures on hospital admission due to respiratory illness in children at Professorial Paediatric unit Teaching Hospital Jaffna

Respiratory diseases are common causes of hospital admission in children. Droplet transmission is the primary route of transmition in respiratory infections. The objective was to assess the burden of respiratory diseases on hospital admission of children during this Covid 19 pandemic and to assess the impact of Covid 19 preventive measures on it. An institutional based retrospective descriptive study based on secondary data was carried out at Professorial Paediatric unit Teaching hospital Jaffna. The data was extracted from the electronic patient management system. All the records of children who were primarily diagnosed with respiratory disease during the period of June 2020 to May 2021 were extracted. This data was compared with that of 2017. The total hospital admission has reduced from 4127 to 1049 during the pandemic. There was a significant decrease in total hospital admission during the covid 19 pandemic compared to before the pandemic {t (11.488) = 8.015 p<0.001}. Similarly, there was a significance decrease in the admissions due to respiratory diseases during the covid 19 pandemic {t(11.355) =5.982 p<0.001}. The percentage of respiratory admission has significantly reduced during the pandemic {t(21.997) =5.173 p<0.001}. Hospital admission due to respiratory illness has significantly reduced when compared to total admission during this pandemic. This may be attributed to social distancing, wearing mask and good hand hygiene.

Behaviour and Socioeconomic Factors of Undocumented Migrants Missing Antiretroviral Therapy Clinic Appointments in Limpopo Province

Background: Global migration is increasingly driven by socioeconomic and political unrest, raising a potentially challenging burden for healthcare systems. A significant group of non-South African citizens present in the country are undocumented migrants and have not restricted access to healthcare services. Objectives: This study explored the behaviour and socioeconomic factors of undocumented migrants missing Antiretroviral Therapy clinic appointments. Methods: A qualitative cross-sectional approach was used. Semi-structured interviews, observations, and document analysis of 17 undocumented migrants, who missed more than four appointments in the first 12 months since antiretroviral therapy initiation, were conducted. Thematic analysis and coding categorised the findings into themes. Reliability and validity were ensured through intercoder agreement, audio recording, triangulation, bracketing, and member checking. Results: The finding revealed that registering to more than one healthcare facility, self-transfer to another facility, and job hunting resulted in clients missing their scheduled appointments. Conclusion: The conclusion drawn from this study is that the Electronic Patient Management System must improve to control the number of clients missing clinic appointments.

Increasing patient flow through neurosurgical critical care: the Leeds Improvement Method

At Leeds General Infirmary, a busy tertiary centre for neurosurgery, there has been little visibility of the step-down status of the patients from intensive care to high dependency or from...

Competing risk of mortality on loss to follow-up outcome among patients with HIV on ART: a retrospective cohort study from the Zimbabwe national ART programme

ObjectiveTo determine the loss to follow-up (LTFU) rates at different healthcare levels after antiretroviral therapy (ART) services decentralisation among ART patients who initiated ART between 2004 and 2017 using the...

P32 An audit to assess the suitability of patients at a tertiary/quaternary paediatric hospital to switch from intravenous (IV) to oral (PO) antimicrobial therapy

Introduction, Aims and ObjectivesIn 2011 the Start Smart then Focus campaign was launched by Public Health England (PHE) to combat antimicrobial resistance.1 The ‘focus’ element refers to the antimicrobial review at 48–72 hours, when a decision and documentation regarding infection management should be made. [OM1] At this tertiary/quaternary paediatric hospital we treat, immunocompromised, high risk patients. In a recent audit it was identified that 80% of antimicrobial use is IV, this may be due to several factors including good central access, centrally prepared IV therapy and oral agents being challenging to administer to children. The aim of the audit was to assess if patient have a blood culture prior to starting therapy, have a senior review at 48–72 hours, and thirdly if our high proportion of intravenous antimicrobial use is justified.MethodElectronic prescribing data from JAC was collected retrospectively over an 8 day period. IV antimicrobials for which there is a suitable oral alternative, this was defined as >80% bioavailability, were included. Patients were excluded in the ICU, cancer and transplant setting, those with absorption issues and with a high risk infection, such as endocarditis or bacteraemia. Patient were assessed against a set criteria to determine if they were eligible to switch from IV to PO therapy; afebrile, stable blood pressure, heart rate <90/min, respiratory rate < 20/min for 24 hours. Reducing CRP, reducing white cell count, blood cultures negative or sensitive to an antibiotic that can be given orally.Results100% of patients (11) had a blood cultures taken within 72 hours of starting therapy55% of patients had a positive blood culture82% of patients had a senior review at 48–72 hours46% of patients were eligible to switch from IV to PO therapy at 72 hours33% of eligible patients were switched from IV to PO therapy at 72 hoursConclusion and RecommendationsThis audit had a low sample size due to the complexity of the inclusion and exclusion criteria, and the difficulty in reviewing patient parameters on many different hospital interfaces. It is known that each patient is reviewed at least 24 hourly on most wards and therefore there is a need for improved documentation of prescribing decisions. Implementation of an IV to oral switch guideline is recommended to support prescribing decisions and educate and reassure clinicians on the bioavailability and benefits of PO antimicrobial therapy where appropriate. Having recently changed electronic patient management systems strategies to explore include hard stops on IV antimicrobial therapies, however this will require much consideration. Education of pharmacist and nurses is required to raise awareness about antimicrobial resistance and the benefits of IV to PO switches, despite the ease of this therapy at out Trust. This will promote a culture in which all healthcare professionals are active antimicrobial guardians, leading to better patient outcomes, less service pressures, and long term financial benefit.ReferenceGOV.UK. 2019. Antimicrobial stewardship: Start smart - then focus. [ONLINE]Available at: https://www.gov.uk/government/publications/antimicrobial-stewardship-start-smart-then-focus [Accessed 3 July 2019]

Users of rehabilitation services in 14 countries and territories affected by conflict, 1988-2018.

ObjectiveTo analyse the demographic and clinical characteristics of people attending physical rehabilitation centres run or supported by the International Committee of the Red Cross in countries and territories affected by conflict.MethodsOf 150 such rehabilitation centres worldwide, 38 use an electronic patient management system. We invited all 38 centres to participate. We extracted de-identified data from 1988 to 2018 and categorized them by sex, age, country or territory and reason for using rehabilitation services.FindingsThirty-one of the 38 rehabilitation centres in 14 countries and territories participated. We included data for 287 274 individuals. Of people using rehabilitation services, 61.6% (176 949/287 274) were in Afghanistan, followed by 15.7% (44 959/287 274) in Cambodia. Seven places had over 9000 service users each (Afghanistan, Cambodia, Gaza Strip, Iraq, Myanmar, Somalia and Sudan). Overall, 72.6% (208 515/287 274) of service users were male. In eight countries, more than half of the users were of working age (18–59 years). Amputation was the most common reason for using rehabilitation services; 33.3% (95 574/287 274) of users were people with amputations, followed by 13.7% (39 446/287 274) with cerebral palsy. The male predominance was greater in the population aged 18–34 years (83.1%; 71 441/85 997) and in people with amputations (88.6%; 84 717/95 574) but was evident across all places, age groups and health conditions.ConclusionThe considerably lower attendance of females at the rehabilitation centres highlights the need to understand the factors that affect the accessibility and acceptability of rehabilitation for women and girls in conflict settings.

Automatic Extraction of Risk Factors for Dialysis Patients from Clinical Notes Using Natural Language Processing Techniques.

Studies have shown that mental health and comorbidities such as dementia, diabetes and cardiovascular diseases are risk factors for dialysis patients. Extracting accurate and timely information associated with these risk factors in the patient health records is not only important for dialysis patient management, but also for real-world evidence generation. We presented HERALD, an natural language processing (NLP) system for extracting information related to risk factors of dialysis patients from free-text progress notes in an electronic dialysis patient management system. By converting semi-structured notes into complete sentences before feeding them into the NLP module, the HERALD system was able achieved 99%, 83% and 80% accuracy in identifying dementia, diabetes and infarction, respectively.

Tipping point: When patients stop eating and drinking in the last phase of their life.

Eating and drinking are essential also in social life. Nutrition and hydration (N&H) at end of life are often a source of discussion and distress. Stopping eating and drinking is a defining element of the dying phase, however, this time point is not well defined. The aim of this retrospective analysis was to investigate whether such a time point can be detected, whether there are specific characteristics associated. The time point when patients stopped oral intake was analyzed in relation to time until death on a specialist palliative care ward (sPCW) of a tertiary comprehensive cancer center. This "tipping point" (TP) was defined as the time point when total food intake fell below 25% of normal solid intake (TP-S) and "tipping point fluid" (TP-F) when fluid intake fell below 500ml/day (oral/intravenous/subcutaneous). Demographic and medical data, the N&H-module in the electronic patient management system and the daily multiprofessional treatment notes at TP (±3 days) were analyzed in all patients cared on the sPCW between 1/15 and 9/17. In these 32 months, of 1194 treated patients 683 (57%) died on the ward. A TP-S was identified in 291 patients prior to dying on the ward (43%) with a median time of six days from to TP-S death. In 75% of these patients, TP-S occurred within two weeks prior to death (range: 0-5 weeks). A TP-F was detected in 202 patients (30%) with a median TP-F-time of two days prior to death. In 75% of these patients, the TP-F was within three days prior to death (range: 0-14 days). The cancer entities in patients in whom TPs could be detected were heterogeneous. No specific disease-related or sociodemographic characteristics for patients with TPs could be determined. In the daily treatment notes, oral and swallowing problems, taste alterations and discussions about stopping artificial nutrition were mentioned. In less than half of dying patients a definitive TP could be detected. In these patients, TP-S occurred within two weeks and TP-F within few days before death. No specific characteristics of patients with TPs could be observed. This indicates the individual nature of the trajectory at end of life.

The utility of a shortened palliative care screening tool to predict death within 12\u2009months \u2013 a prospective observational study in two south African hospitals with a high HIV burden

BackgroundTimely identification of people who are at risk of dying is an important first component of end-of-life care. Clinicians often fail to identify such patients, thus trigger tools have been developed to assist in this process. We aimed to evaluate the performance of a identification tool (based on the Gold Standards Framework Prognostic Indicator Guidance) to predict death at 12 months in a population of hospitalised patients in South Africa.MethodsPatients admitted to the acute medical services in two public hospitals in Cape Town, South Africa were enrolled in a prospective observational study. Demographic data were collected from patients and patient notes. Patients were assessed within two days of admission by two trained clinicians who were not the primary care givers, using the identification tool. Outcome mortality data were obtained from patient folders, the hospital electronic patient management system and the Western Cape Provincial death registry which links a unique patient identification number with national death certificate records and system wide electronic records.Results822 patients (median age of 52 years), admitted with a variety of medical conditions were assessed during their admission. 22% of the cohort were HIV-infected. 218 patients were identified using the screening tool as being in the last year of their lives. Mortality in this group was 56% at 12 months, compared with 7% for those not meeting any criteria. The specific indicator component of the tool performed best in predicting death in both HIV-infected and HIV-uninfected patients, with a sensitivity of 74% (68–81%), specificity of 85% (83–88%), a positive predictive value of 56% (49–63%) and a negative predictive value of 93% (91–95%). The hazard ratio of 12-month mortality for those identified vs not was 11.52 (7.87–16.9, p < 0.001).ConclusionsThe identification tool is suitable for use in hospitals in low-middle income country setting that have both a high communicable and non-communicable disease burden amongst young patients, the majority under age 60.

260 Recording of Syncope Presentations in Model 4 Hospital - Large Variations in Data Collection Systems

Abstract Background Patients with unexplained falls, syncope and dizziness frequently present acutely. There are multiple IT systems in operation across the acute hospital network with varying functionality relating to activity capture. MAXIMS is an electronic healthcare record used to capture attendances in our ED. Within our department a digital patient management system Electronic Patient Management System (EPMS) records clinical information from admission through to discharge. The Hospital Inpatient Enquiry System (HIPE) is the principal national source of discharge information from acute hospitals. We aimed to look at this cohort of patients and compare the clinical data available on each of these information sources used in a Model 4 teaching hospital. Methods MAXIMS database was interrogated to examine patients who presented to the ED with the search terms “falls”, “syncope” and “dizziness”. We conducted a review of all patients who met our search terms over two non-consecutive weeks in 2018. For the same period EPMS was interrogated and HIPE data were requested and analysed. Results MAXIMS, EPMS and HIPE records over the two-week period captured 190, 104 and 24 patients respectively. Mean age was 62, 75 and 72 years. Of 190 records on MAXIMS, 107 (56%) were not coded on HIPE following discharge. Of these 107 patients, 93 (87%) did not require a hospital admission and 14 (13%) patients self-discharged. Conclusion Current available digital systems provide inconsistent information relating to patients presenting acutely with falls, dizziness and syncope. How these conditions are captured electronically can vary between systems due to which parameters are recorded. HIPE stats were lower than Maxims/EPMS which may reflect the fact that HIPE only captures discharge diagnosis as opposed to presenting symptoms. In order to get a clearer picture of these conditions, and for planning purposes, an approach amalgamating the data from all systems into one electronic source would be required.

Electronic primary dental care records in research: A case study of validation and quality assurance strategies

BackgroundIn dentistry, the use of electronic patient records for research is underexplored. The aim of this paper is to describe a case study process of obtaining research data (sociodemographic, clinical and workforce) from electronic primary care dental records, and outlining data cleaning and validation strategies. This study was undertaken at the University of Portsmouth Dental Academy (UPDA), which is a centre of education, training and provision of state funded services (National Health Services). UPDA’s electronic patient management system is R4/Clinical +. This is a widely used system in general dental practices in the UK. MethodA two-phase process, involving first Pilot and second Main data extraction were undertaken. Using System Query Language (SQL), data extracts containing variables related to patients’ demography, socio-economic status and dental care received were generated. A data cleaning and validation exercise followed, using a combination of techniques including Maletic and Marcus’s (2000) general framework for data cleaning and Rahm and Haido’s (2010) principles of data cleaning. ResultsThe findings of the case study support the use of a two-phase data extraction process. The data validation processes highlighted the need for both manual and analytical strategies when cleaning these data. Finally, the process demonstrated that electronic dental records can be validated and used for epidemiological and heath service research. The potential to generalise findings is great due to the large number of records. There are, however, limitations to the data which need to be considered, relating to quality (data input), database structure and interpretation of data codes. ConclusionElectronic dental records are useful in health service research, epidemiological studies and skill mix research. Researchers should work closely with clinicians, managers and software developers to ensure that the data generated are accurate, valid and generalisable. Following data extraction the researchers need to adapt stringent validation and data cleaning strategies to guarantee that the extracted electronic data are accurate.