Primary Care Triage Research Articles

Current Capacity for Diagnosing Alzheimer's Disease in Germany and Implications for Wait Times.

Amyloid-targeting therapies for Alzheimer's disease (AD) might become available in Germany soon. The combination of a large pool of prevalent cases and a complex diagnostic process to determine eligibility for these treatments is likely to challenge health systems' capacity. To analyze Germany's healthcare system capacity to identify treatment-eligible patients in a timely and equitable manner. We modeled patients' diagnostic journey and projects wait times due to capacity constraints for AD specialist visits and PET scans from 2024 to 2043. Model parameters were derived from published data and expert input. Wait times would be ∼50 months over the model horizon, if patients were referred to specialists based on a brief cognitive assessment in primary care. Wait times for patients with social health insurance are projected to be 1.9 times those of patients with private insurance, with peak wait times of around 76 and 40 months, respectively. Adding a blood test for the AD pathology as additional triage step would reduce wait times to below 24 months. In spite of having a well-resourced health system, Germany is projected to be unable to cope with the demand for biomarker-based AD diagnosis, if a disease-modifying AD treatment were introduced. As these treatments might become available by the end of 2024, decisive action, in particular dissemination of high-performing AD blood tests for triage in primary care, will be needed to prevent delays in access and potentially avoidable and inequitable disease progression.

Early economic evaluation of magnetic resonance imaging for prostate cancer detection in primary care.

To explore the potential impacts of incorporating prebiopsy magnetic resonance imaging into primary care as a triage test within the prostate cancer diagnostic pathway. Decision analytic modelling with decision trees was utilised for this early economic evaluation. A conceptual model was developed reflecting the common primary care routes to diagnosis for prostate cancer: opportunistic, asymptomatic prostate-specific antigen (PSA) screening or symptomatic presentation. The use of multiparametric MRI (mpMRI) or biparametric MRI (bpMRI) as a primary care triage test following an elevated PSA result was evaluated. A health system perspective was adopted with a time horizon of 12 months. Health effects were expressed in terms of utilities drawn from the literature. The primary outcome was prostate cancer diagnosis. Evidence used to inform the model was drawn from published primary studies, systematic reviews, and secondary analyses of primary and secondary care datasets. Base case analysis showed that the PSA pathway was dominated by both mpMRI- and bpMRI-based pathways for patients undergoing opportunistic screening and symptomatic assessment. bpMRI pathways had greater improvement in cost and utility than mpMRI pathways in both clinical scenarios. Significantly more MRI scans would be performed using the modelled approach (66 626 scans vs. 37 456 scans per 100 000 patients per annum), with fewer subsequent urgent suspected cancer referrals for both mpMRI (38% reduction for screening and symptomatic patients) and bpMRI (72% reduction for screening; 71% for symptomatic) pathways, and a small increase in number of missed cancer diagnoses. Deterministic sensitivity analyses, varying each parameter to its upper and lower 95% confidence intervals, showed no significant change in the dominance of the MRI-based prostate cancer diagnostic pathways. Using prostate MRI as a second-level triage test for suspected prostate cancer in primary care could reduce health service costs without a detrimental effect on patient utility.

Healthcare leaders’ experiences of implementing artificial intelligence for medical history-taking and triage in Swedish primary care: an interview study

BackgroundArtificial intelligence (AI) holds significant promise for enhancing the efficiency and safety of medical history-taking and triage within primary care. However, there remains a dearth of knowledge concerning the practical implementation of AI systems for these purposes, particularly in the context of healthcare leadership. This study explores the experiences of healthcare leaders regarding the barriers to implementing an AI application for automating medical history-taking and triage in Swedish primary care, as well as the actions they took to overcome these barriers. Furthermore, the study seeks to provide insights that can inform the development of AI implementation strategies for healthcare.MethodsWe adopted an inductive qualitative approach, conducting semi-structured interviews with 13 healthcare leaders representing seven primary care units across three regions in Sweden. The collected data were subsequently analysed utilizing thematic analysis. Our study adhered to the Consolidated Criteria for Reporting Qualitative Research to ensure transparent and comprehensive reporting.ResultsThe study identified implementation barriers encountered by healthcare leaders across three domains: (1) healthcare professionals, (2) organization, and (3) technology. The first domain involved professional scepticism and resistance, the second involved adapting traditional units for digital care, and the third inadequacies in AI application functionality and system integration. To navigate around these barriers, the leaders took steps to (1) address inexperience and fear and reduce professional scepticism, (2) align implementation with digital maturity and guide patients towards digital care, and (3) refine and improve the AI application and adapt to the current state of AI application development.ConclusionThe study provides valuable empirical insights into the implementation of AI for automating medical history-taking and triage in primary care as experienced by healthcare leaders. It identifies the barriers to this implementation and how healthcare leaders aligned their actions to overcome them. While progress was evident in overcoming professional-related and organizational-related barriers, unresolved technical complexities highlight the importance of AI implementation strategies that consider how leaders handle AI implementation in situ based on practical wisdom and tacit understanding. This underscores the necessity of a holistic approach for the successful implementation of AI in healthcare.

Innovating Healthcare Delivery: Enhancing Patient Engagement with Rule-Based Machine Learning Solutions

In contemporary healthcare, chatbots, driven by rule-based machine learning and sophisticated algorithms, are revolutionizing traditional models. This study explores their use in primary care triage, mental health support, chronic disease management, and telemedicine, highlighting their potential to enhance patient engagement, optimize clinical workflows, and improve care quality. Despite their promise, chatbots face challenges such as safeguarding patient privacy, mitigating biases, ensuring regulatory compliance, and maintaining ethical standards. The paper emphasizes the need for interoperability, scalability, and usability for seamless integration into healthcare systems. Future advancements include integrating advanced machine learning, evolving voice recognition, and personalized medicine approaches. By leveraging data analytics and predictive modelling, chatbots could proactively identify health risks, leading to a new era of preventive and personalized healthcare. Overall, this research highlights chatbots' profound impact on promoting patient-centred care, initiative-taking health management, and data-driven decision-making, fostering equitable access to high-quality healthcare services. Keywords— Rule-based chatbots, Machine Learning, Healthcare, Natural Language Processing, Patient Engagement, Clinical Support, Case Studies, Prospects.

Impact of primary care triage using the Head and Neck Cancer Risk Calculator version 2 on tertiary head and neck services in the post-coronavirus disease 2019 period.

This study investigates the impact of primary care utilisation of a symptom-based head and neck cancer risk calculator (Head and Neck Cancer Risk Calculator version 2) in the post-coronavirus disease 2019 period on the number of primary care referrals and cancer diagnoses. The number of referrals from April 2019 to August 2019 and from April 2020 to July 2020 (pre-calculator) was compared with the number from the period January 2021 to August 2022 (post-calculator) using the chi-square test. The patients' characteristics, referral urgency, triage outcome, Head and Neck Cancer Risk Calculator version 2 score and cancer diagnosis were recorded. In total, 1110 referrals from the pre-calculator period were compared with 1559 from the post-calculator period. Patient characteristics were comparable for both cohorts. More patients were referred on the cancer pathway in the post-calculator cohort (pre-calculator patients 51.1 per cent vs post-calculator 64.0 per cent). The cancer diagnosis rate increased from 2.7 per cent in the pre-calculator cohort to 3.3 per cent in the post-calculator cohort. A lower rate of cancer diagnosis in the non-cancer pathway occurred in the cohort managed using the Head and Neck Cancer Risk Calculator version 2 (10 per cent vs 23 per cent, p = 0.10). Head and Neck Cancer Risk Calculator version 2 demonstrated high sensitivity in cancer diagnosis. Further studies are required to improve the predictive strength of the calculator.

Age‐related relative comorbidity burden of mild cognitive impairment: A US database study

AbstractBackgroundAs novel therapies emerge, timely detection and diagnosis of early Alzheimer’s disease (AD) become increasingly relevant. While primary care would be the most suitable setting for early detection, diagnosis rates of mild cognitive impairment (MCI) are low and systematic screening of asymptomatic individuals is currently not recommended. We investigated whether a predictive algorithm based on established AD risk factors can help to identify individuals at elevated risk of having undiagnosed MCI, in particular in younger age cohorts.MethodOur retrospective study used MarketScan insurance claims data to identify individuals aged ≥50 years with MCI (but without dementia) and assemble controls without MCI/dementia diagnoses matched on age, sex, and geographic region. Medical conditions previously identified as AD risk factors in the literature were considered MCI risk factors if their prevalence in MCI individuals was statistically significantly higher than in controls. Logistic regression and extreme gradient boosting (XGBoost; a machine learning method) were applied to discriminate MCI from control individuals based on area under the curve (AUC). Analyses were conducted for the overall cohort and by age group (50‐64, 65‐79, and ≥80 years).ResultTwenty‐five medical conditions had statistically higher frequency in the MCI cohort (n = 5,185) compared to controls (n = 15,555). The top MCI risk factors (with odds ratio for the 50‐64, 65‐79, and ≥80 years subgroups, respectively) were: depression (4.4, 3.1, 1.9); stroke/transient ischemic attack, TIA (6.4, 3.0, 2.1); obstructive sleep apnea, OSA (3.6, 2.5, 1.7); hyperlipidemia (1.7, 1.5, 1.4); and hypertension (1.6, 1.5, 1.5) (all p‐values <0.0001). The logistic regression models for identifying MCI individuals reached AUCs of 0.75, 0.69, and 0.66, for the 3 age groups, respectively. The XGBoost approach reached AUC = 0.94 (training set) and AUC = 0.78 (test set). Each approach was most accurate when discriminating MCI for the youngest age group.ConclusionOur two modeling approaches were able to predict MCI with acceptable accuracy. The predictions were more accurate in the 50‐64 age group than in the older groups, suggesting that the use of prediction models for triage in primary care will play a greater role in younger age groups.

How well is China’s healthcare system prepared to provide access to a disease‐modifying Alzheimer’s treatment?

AbstractBackgroundWhile the majority of patients with Alzheimer’s disease (AD) reside in low and middle‐income countries, little is known of their preparedness for disease‐modifying treatments. We project wait times for access to treatment in China, the world’s most populous and still a middle‐income country.MethodMarkov model to estimate wait times based on population, prevalence and capacity data for 2024 to 2044, by desk research and interviews with 45 experts. Base case assumes that 20% of individuals aged 50 and older with no prior evaluation for cognitive impairment would visit a hospital‐based specialist clinic for evaluation and biomarker testing, and if positive, to confirm presence of amyloid and determine treatment eligibility would be determined with PET scanning (37%) or CSF analysis (63%). In two alternative scenarios, individuals would first see a primary care clinician and be referred to a clinic if (1) a brief test suggested mild cognitive impairment, or if (2) in addition a blood test with sensitivity of 0.89 and specificity of 0.73 suggested presence of AD pathology.ResultUnder current practice in China, where gatekeeping by primary care is uncommon and referrals to AD specialty care not required, specialist capacity would be overwhelmed, resulting in lengthy wait‐times throughout the simulation. (Figure 1). Using a model of specialist referral based on cognitive testing in primary care, wait times would peak at 90 months in 2028 and later remain at 50‐60 months. Confirmatory biomarker testing in specialist clinics would become the rate‐limiting constraint. (Figure 2) Adding an AD blood test would lower peak wait times to 55 months and to about two years later (Figure 3).ConclusionWe project extremely long wait times in China to gain access to AD treatment under current practice. Triage in primary care, including a blood test before PET/CSF testing would reduce wait times, albeit only to about two years. With dementia specialists per capita comparable to Japan and growing, the main constraint is capacity for confirmatory biomarker testing. Higher performing blood tests, greater reliance on CSF and expanding PET capacity would rein in wait times, provided that growth in specialist capacity is maintained.

Diagnosis of pelvic endometriosis: A systematic review and accuracy meta-analysis of non-invasive tests available in primary care

Background: Endometriosis is a chronic, often debilitating condition with a current significant delay from symptom onset to diagnosis with much of this in primary care. Methods: A systematic review and meta-analysis of the primary literature was conducted to investigate the accuracy of symptoms, clinical history and first-line non-invasive tests to predict pelvic endometriosis (PROSPERO: CRD42020187543). We searched Medline, Embase, Web of Science and Scopus from conception (1966; 1972; 1997; 2004 respectively) to September 2022 for primary test accuracy studies assessing non-invasive tests against reference standard diagnosis for endometriosis. Two authors independently conducted data extraction and quality assessment. Grading of evidence was performed using a novel visual pentagon model. Meta-analyses of test accuracy was estimated using bivariate random effects models. Results: The 125 included studies (250,574 participants) showed mixed quality. Studies applying non-surgical (database/self-reporting) reference standard had a greater risk of bias. In 98 studies applying surgical reference standard, summary diagnostic odds ratios for endometriosis were: dysmenorrhoea 2.56 (95% confidence interval 1.99-3.29); pelvic pain 2.56 (1.73-3.74); dyschezia 2.05 (1.36-3.10); dyspareunia 2.45 (1.71-3.52); family history of endometriosis 6.79 (4.08-11.3); nulligravidity of 2.01 (1.62-2.50); body mass index (BMI) ≥30kg/m2 0.37 (0.19-0.68); trans-vaginal ultrasound scan (TVUSS) endometrioma 91.2 (44.0-189); TVUSS invasive endometriosis 26.1 (9.28-73.5); and cancer antigen-125 (CA-125) >35U/mL 16.0 (8.09-31.7). Sensitivity analysis excluding all high-risk studies found concordant results. Conclusions: This meta-analysis collated the performance of non-invasive tests for endometriosis across a comprehensive and geographically varied population. Study quality was mixed, however results were consistent with high-risk studies excluded. These findings will inform future prediction models for triage in primary care.

Impact of faecal haemoglobin based triage of bowel symptoms presenting to primary care on colorectal cancer diagnosis.

Faecal immunochemical testing (FIT) for faecal haemoglobin was introduced into primary care in National Health Service Tayside in 2015 as an adjunct to clinical assessment of new bowel symptoms. We aimed to assess the impact of FIT-based triage in primary care on colorectal cancer (CRC) diagnosis. Cancer audit data between January 2016 and December 2019 were reviewed to identify all patients diagnosed locally with CRC. The mode of presentation and stage at diagnosis were noted and patient records were interrogated to identify whether FIT and full blood count (FBC) had been performed prior to referral. Results were compared between the FIT and non-FIT groups. In all, 1245 patients were diagnosed with CRC of whom 581 (46.7%) presented through primary care. FIT was performed prior to referral in 440/581 (75.7%), with the proportion increasing from 62.3% in 2016 to 85.8% in 2019. At faecal haemoglobin ≥10μg Hb/g faeces, sensitivity for CRC was 94.1%. Over the study period the annual proportion of non-emergency presentations increased significantly; presentations from primary care increased from 43.1% to 53.5% (P=0.0096). After excluding non-FIT patients who had an overt CRC at referral, there was no difference in stage at diagnosis between FIT and non-FIT cancers. Safety-netting with FBC was widely used in our cohort with 97.3% of FIT patients having also had FBC. FIT-based triage of new bowel symptoms in primary care is associated with increased non-emergency presentation of CRC but this did not influence stage at diagnosis.

Triage Errors in Primary and Pre-Primary Care.

Triage errors are a major concern in health care due to resulting harmful delays in treatments or inappropriate allocation of resources. With the increasing popularity of digital symptom checkers in pre–primary care settings, and amid claims that artificial intelligence outperforms doctors, the accuracy of triage by digital symptom checkers is ever more scrutinized. This paper examines the context and challenges of triage in primary care, pre–primary care, and emergency care, as well as reviews existing evidence on the prevalence of triage errors in all three settings. Implications for development, research, and practice are highlighted, and recommendations are made on how digital symptom checkers should be best positioned.

Digital communication in primary care: A qualitative descriptive study of patient interviews considering person-centred care

With the growing demand for primary care, provision needs to be efficient, yet retain person-centred and integrated care. Digital communication is suggested as a way to settle these aspects, although there is insufficient knowledge regarding the end-user's perspective. The aim of this study was to describe patients’ experiences of digital communication in their primary care contact, while considering aspects of person-centred care. A purposeful sample of 16 patients (aged 28–85 years) was interviewed via telephone in a qualitative design, and verbatim transcripts were analysed using inductive thematic analysis and a subsequent deductive approach in relation to a person-centred care framework. The study was reported following the COREQ guidelines. The findings indicate that, from a patient perspective, digital communication enables independent contact where one crafts one's medical history at a time and place convenient for oneself, providing opportunities for the primary care triage to consider one's individual input. These aspects of digital communication serve person-centredness, although they were less suitable for certain health issues, but a human interaction was most appropriate when health concerns were ambiguous or complex. Thus, digital communication may serve patients in primary care, but the development and implementation of such services must move beyond technical and professional perspectives. In order to facilitate integrated care, further emphasis is needed on both patients’ experiences, and the principles of person-centred care.

IDF21-0299 An automatic deep learning-based system for screening and management of DME

Background: Diabetic macular oedema (DME) is the most common complication form of sight-threatening retinopathy in people with a diabetic condition. Previous studies demonstrate specifically that poor glycaemic and blood pressure control are associated with the presence and development of DME condition. The aim of this study is to validate the implementation of deep learning systems as a tool into routinely practiced diabetic retinopathy screening to improve diagnosis and management of DME. Method: 62,383 color images of the eye fundus were used. These images were acquired between 2016 and 2020 with a non mydriatic Centervue DRS camera. Images were graded as DME and no DME by retinal specialists according to the classification of American Association of Ophthalmology. The dataset was randomly divided into train-validation and test sets in proportions of 89.5% and 10.5%, respectively. We trained 4 CNNs based on the ResNeXt architecture with standardised convolutions as in Kolesnikov et al (ECCV 2020). Model parameters were initialized from a solution to the ImageNet image classification task and further optimized by minimizing the cross-entropy loss in its Sharpness-Aware extension, which attempts to find a parameter configuration lying in neighborhoods with uniformly low loss values, leading to improved generalization ability. The training was carried out by error back-propagation with weight updates given by the Stochastic Gradient Descent algorithm, and a learning rate of 0.01 that was cyclically annealed towards 0. The optimization ran for 20 cycles but was stopped as soon as the AUC computed on the validation set did not improve for two cycles. Final predictions were computed from the average ensemble of the four models. Results: The resulting model obtained a high performance on the test dataset. The area under the curve (AUC) was 0.9591(0.9536-0.9645). In addition, at an optimal cut-off given by the Youden index, the accuracy, the sensitivity and the specificity achieved were 0.9033(0.8959-0.9104), 0.8942 and 0.9054 respectively. An F1 score of 0.7759 and a Kappa score of 0.7156 indicate a high agreement of the model’s predictions with ophthalmologists’ opinions. Conclusion: The proposed system achieved high sensitivity and specificity in DME screening from retinal images of the eye fundus. The performance of the proposed approach is comparable to that of human experts and could be a valuable tool to help primary care triage.

Artificial Intelligence Support for Skin Lesion Triage in Primary Care and Dermatology

Background Primary care providers, dermatology specialists, and health care access are key components of primary prevention, early diagnosis, and treatment of skin cancer. Artificial intelligence (AI) offers the promise of diagnostic support for nonspecialists, but real-world clinical validation of AI in primary care is lacking. Objective We aimed to (1) assess the reliability of an AI-based clinical triage algorithm in classifying benign and malignant skin lesions and (2) evaluate the quality of images obtained in primary care using the study camera (3Gen DermLite Cam v4 or similar). Methods This was a single-center, prospective, double-blinded observational study with a predetermined study design. We recruited participants with suspected skin cancer in 20 primary care practices who were referred for assessment via teledermatology. A second set of photographs taken using a standardized camera was processed by the AI algorithm. We evaluated the image quality and compared two teledermatologists’ diagnoses by consensus (the “gold standard”) with AI and histology where applicable. Results Our primary outcome assessment stratified 391 skin lesions by management as benign, uncertain, or malignant. Uncertain lesions were not included in the sensitivity and specificity analyses. Uncertain lesions included lesions that had either diagnostic or management uncertainties. For the remaining 242 lesions, the sensitivity was 97.26% (95% CI 93.13%-99.25%) and the specificity was 97.92% (95% CI 92.68%-99.75%). The AI algorithm was compared with the histological diagnoses for 123 lesions. The sensitivity was 100% (95% CI 95.85%-100%) and the specificity was 72.22% (95% CI 54.81%-85.80%). Conclusions The AI algorithm demonstrates encouraging results, with high sensitivity and specificity, concordant with previous AI studies. It shows potential as a triage tool in conjunction with teledermatology to augment health care and improve access to dermatology. Further real-life studies need to be conducted on a larger scale to assess the reliability, usability, and cost-effectiveness of the algorithm in primary care. Acknowledgments MoleMap NZ, who developed the AI algorithm, provided some funding for this study. HT's salary was partially sponsored by MoleMap NZ, who developed the AI algorithm. AB is a shareholder and consultant to Molemap Ltd provider of the AI algorithm. Conflicts of Interest None declared.

Development of an Electrochemical CCL17/TARC Biosensor toward Rapid Triage and Monitoring of Classic Hodgkin Lymphoma.

A point-of-care blood test for the detection of an emerging biomarker, CCL17/TARC, could prove transformative for the clinical management of classic Hodgkin lymphoma (cHL). Primary care diagnosis is challenging due to nonspecific clinical presentation and lack of a diagnostic test, leading to significant diagnostic delays. Treatment monitoring encounters false-positive and negative results, leading to avoidable chemotherapy toxicity, or undertreatment, impacting patient morbidity and mortality. Here, we present an amperometric CCL17/TARC immunosensor, based on the utilization of a thiolated heterobifunctional cross-linker and sandwich antibody assay, to facilitate novel primary care triage and chemotherapy monitoring strategies for cHL. The immunosensor shows excellent analytical performance for clinical testing; linearity (R2 = 0.986), detection limit (194 pg/mL), and lower and upper limits of quantitation (387-50 000 pg/mL). The biosensor differentiated all 42 newly diagnosed cHL patients from healthy volunteers, based on serum CCL17/TARC concentration, using blood samples collected prior to treatment intervention. The immunosensor also discriminated between paired blood samples of all seven cHL patients, respectively, collected prior to treatment and during chemotherapy, attributed to the decrease in serum CCL17/TARC concentration following chemotherapy response. Overall, we have shown, for the first time, the potential of an electrochemical CCL17/TARC biosensor for primary care triage and chemotherapy monitoring for cHL, which would have positive clinical and psychosocial implications for patients, while streamlining current healthcare pathways.

Efficacy of Chest Radiography as a Primary Care Triage Tool in Severe Coronavirus Disease.

Objective Severe acute respiratory syndrome coronavirus 2 has spread globally, and it is important to utilize medical resources properly, especially in critically ill patients. We investigated the validity of chest radiography as a tool for predicting aggravation in coronavirus disease (COVID-19) cases. Methods A total of 104 laboratory-confirmed COVID-19 cases were referred from the cruise ship “Diamond Princess” to the Self-Defense Forces Central Hospital in Japan from February 11 to 25, 2020. Fifty-nine symptomatic patients were selected. Chest radiography was performed upon hospitalization; subsequently, patients were categorized into the positive radiograph (Group A) and negative radiograph (Group B) groups. Radiographic findings were analyzed with a six-point semiquantitative score. Group A was further classified into two additional subgroups: patients who required oxygen therapy during their clinical courses (Group C) and patients who did not (Group D). Clinical records, laboratory data, and radiological findings were collected for an analysis. Results Among 59 patients, 34 were men with a median age of 60 years old. Groups A, B, C, and D consisted of 33, 26, 12, and 21 patients, respectively. The number of patients requiring oxygen administration was significantly larger in Group A than in Group B. The consolidation score on chest radiographs was significantly higher in Group C than in Group D. When chest radiographs showed consolidation in more than two lung fields, the positive likelihood ratio of deterioration was 10.6. Conclusions Chest radiography is a simple and easy-to-use clinic-level triage tool for predicting the severity of COVID-19 and may contribute to the allocation of medical resources.

The Delta Study - Prevalence and characteristics of mood disorders in 924 individuals with low mood: Results of the of the World Health Organization Composite International Diagnostic Interview (CIDI).

ObjectivesThe Delta Study was undertaken to improve the diagnosis of mood disorders in individuals presenting with low mood. The current study aimed to estimate the prevalence and explore the characteristics of mood disorders in participants of the Delta Study, and discuss their implications for clinical practice.MethodsIndividuals with low mood (Patients Health Questionnaire‐9 score ≥5) and either no previous mood disorder diagnosis (baseline low mood group, n = 429), a recent (≤5 years) clinical diagnosis of MDD (baseline MDD group, n = 441) or a previous clinical diagnosis of BD (established BD group, n = 54), were recruited online. Self‐reported demographic and clinical data were collected through an extensive online mental health questionnaire and mood disorder diagnoses were determined with the World Health Organization Composite International Diagnostic Interview (CIDI).ResultsThe prevalence of BD and MDD in the baseline low mood group was 24% and 36%, respectively. The prevalence of BD among individuals with a recent diagnosis of MDD was 31%. Participants with BD in both baseline low mood and baseline MDD groups were characterized by a younger age at onset of the first low mood episode, more severe depressive symptoms and lower wellbeing, relative to the MDD or low mood groups. Approximately half the individuals with BD diagnosed as MDD (49%) had experienced (hypo)manic symptoms prior to being diagnosed with MDD.ConclusionsThe current results confirm high under‐ and misdiagnosis rates of mood disorders in individuals presenting with low mood, potentially leading to worsening of symptoms and decreased well‐being, and indicate the need for improved mental health triage in primary care.

Health Care Professionals' Experience of a Digital Tool for Patient Exchange, Anamnesis, and Triage in Primary Care: Qualitative Study.

BackgroundDespite a growing body of knowledge about eHealth innovations, there is still limited understanding of the implementation of such tools in everyday primary care.ObjectiveThe objective of our study was to describe health care staff’s experience with a digital communication system intended for patient-staff encounters via a digital route in primary care.MethodsIn this qualitative study we conducted 21 individual interviews with staff at 5 primary care centers in Sweden that had used a digital communication system for 6 months. The interviews were guided by narrative queries, transcribed verbatim, and subjected to content analysis.ResultsWhile the digital communication system was easy to grasp, it was nevertheless complex to use, affecting both staffing and routines for communicating with patients, and documenting contacts. Templates strengthened equivalent procedures for patients but dictated a certain level of health and digital literacy for accuracy. Although patients expected a chat to be synchronous, asynchronous communication was extended over time. The system for digital communication benefited assessments and enabled more efficient use of resources, such as staff. On the other hand, telephone contact was faster and better for certain purposes, especially when the patient’s voice itself provided data. However, many primary care patients, particularly younger ones, expected digital routes for contact. To match preferences for communicating to a place and time that suited patients was significant; staff were willing to accept some nuisance from a suboptimal service—at least for a while—if it procured patient satisfaction. A team effort, including engaged managers, scaffolded the implementation process, whereas being subjected to a trial without likely success erected barriers.ConclusionsA digital communication system introduced in regular primary care involved complexity beyond merely learning how to manage the tool. Rather, it affected routines and required that both the team and the context were addressed. Further knowledge is needed about what factors facilitate implementation, and how. This study suggested including ethical perspectives on eHealth tools, providing an important but novel aspect of implementation.

Teledermatology reduces dermatology referrals and improves access to specialists.

BackgroundTeledermatology may be used for triage in primary care to address skin conditions, improving access and reducing time to treat the most severe or surgical cases. We aimed to evaluate the proportion of individuals who could be assessed in primary care using teledermatology, and how this affected the waiting time for an in-person dermatologist appointment.MethodsA cross-sectional retrospective study, involving 30,976 individuals and 55,624 skin lesions, was performed from July 2017 to July 2018 We assessed the frequency of diagnoses and referrals to biopsy, to in-person dermatologists, or to primary care, and compared the waiting time for an in-person dermatologist appointment before and after the teledermatology implementation.Findings53% of the patients were managed with the primary care physician, 43% were referred to in-person dermatologists and 4% directly to biopsy, leading to a reduction of 78% in the waiting time for in-person appointments when compared to the previous period. The most frequent diseases were: melanocytic nevus, seborrheic keratosis, acne, benign neoplasms, onychomycosis, atopic dermatitis, solar lentigo, melasma, xerosis, and epidermoid cyst, with significant differences according to sex, age and referrals. The most frequent treatment prescribed was emollient.InterpretationThe use of teledermatology as a triage tool significantly reduced the waiting time for in-person visits, improving health care access and utilizing public resources wisely. Knowledge of sex, age, diagnoses and treatment of common skin conditions can enable public policies for the prevention and orientation of the population, as it can be used to train general physicians to address such cases.FundingNone.

Human- Versus Machine Learning-Based Triage Using Digitalized Patient Histories in Primary Care: Comparative Study.

BackgroundSmartphones have made it possible for patients to digitally report symptoms before physical primary care visits. Using machine learning (ML), these data offer an opportunity to support decisions about the appropriate level of care (triage).ObjectiveThe purpose of this study was to explore the interrater reliability between human physicians and an automated ML-based triage method.MethodsAfter testing several models, a naïve Bayes triage model was created using data from digital medical histories, capable of classifying digital medical history reports as either in need of urgent physical examination or not in need of urgent physical examination. The model was tested on 300 digital medical history reports and classification was compared with the majority vote of an expert panel of 5 primary care physicians (PCPs). Reliability between raters was measured using both Cohen κ (adjusted for chance agreement) and percentage agreement (not adjusted for chance agreement).ResultsInterrater reliability as measured by Cohen κ was 0.17 when comparing the majority vote of the reference group with the model. Agreement was 74% (138/186) for cases judged not in need of urgent physical examination and 42% (38/90) for cases judged to be in need of urgent physical examination. No specific features linked to the model’s triage decision could be identified. Between physicians within the panel, Cohen κ was 0.2. Intrarater reliability when 1 physician retriaged 50 reports resulted in Cohen κ of 0.55.ConclusionsLow interrater and intrarater agreement in triage decisions among PCPs limits the possibility to use human decisions as a reference for ML to automate triage in primary care.

Rationale and design of a cohort study evaluating triage of acute chest pain in out-of-hours primary care in the Netherlands (TRACE).

The aims of this study are (1) to evaluate the performance of current triage for chest pain; (2) to describe the case mix of patients undergoing triage for chest pain; and (3) to identify opportunities to improve performance of current Dutch triage system for chest pain. Chest pain is a common symptom, and identifying patients with chest pain that require urgent care can be quite challenging. Making the correct assessment is even harder during telephone triage. Temporal trends show that the referral threshold has lowered over time, resulting in overcrowding of first responders and emergency services. While various stakeholders advocate for a more efficient triage system, careful evaluation of the performance of the current triage in primary care is lacking. TRiage of Acute Chest pain Evaluation in primary care (TRACE) is a large cohort study designed to describe the current Dutch triage system for chest pain and subsequently evaluate triage performance in regard to clinical outcomes. The study consists of consecutive patients who contacted the out-of-hours primary care facility with chest pain in the region of Alkmaar, the Netherlands, in 2017, with follow-up for clinical outcomes out to August 2019. The primary outcome of interest is 'major event', which is defined as the occurrence of death from any cause, acute coronary syndrome, urgent coronary revascularization, or other high-risk diagnoses in which delay is inadmissible and hospitalization is necessary. We will evaluate the performance of the triage system by assessing the ability of the triage system to correctly classify patients regarding urgency (accuracy), the proportion of safe actions following triage (safety) as well as rightfully deployed ambulances (efficacy). TRACE is designed to describe the current Dutch triage system for chest pain in primary care and to subsequently evaluate triage performance in regard to clinical outcomes.