Lyme borreliosis and medical wandering: what do patients think about multidisciplinary management? A qualitative study in the context of scientific and social controversy

INTRODUCTION

To answer to patients' medical wandering, often due to "unexplained symptoms" of "unexplained diseases" and to misinformation, multidisciplinary care centers for suspected Lyme borreliosis (LB), such as the 5 Tick-Borne Diseases (TBDs) Reference Centers (TBD-RC), were created a few years ago in France, the Netherlands and Denmark. Our study consisted of a comprehensive analysis of the satisfaction of the patients managed at a TBD-RC for suspected LB in the context of scientific and social controversy.

METHODS

We included all adults who were admitted to one of the TBD-RC from 2017 to 2020. A telephone satisfaction survey was conducted 12 months after their first consultation. It consisted of 5 domains, including 2 free-text items: "What points did you enjoy?" and "What would you like us to change or to improve?". In the current study, the 2 free-items were analyzed with a qualitative method called reflexive thematic analysis within a semantic and latent approach.

RESULTS

The answer rate was 61.3% (349/569) and 97 distinctive codes from the 2-free-text items were identified and classified into five themes: (1) multidisciplinarity makes it possible to set up quality time dedicated to patients; (2) multidisciplinarity enables seamless carepaths despite the public hospital crisis compounded by the COVID-19 pandemic; (3) multidisciplinarity is defined as trust in the team's competences; (4) an ambivalent opinion and uncertainty are barriers to acceptance of the diagnosis, reflecting the strong influence of the controversy around LB; and (5) a lack of adapted communication about TBDs, their management, and ongoing research is present.

CONCLUSION

The multidisciplinary management for suspected LB seemed an answer to medical wandering for the majority of patients and helped avoid misinformation, enabling better patient-centered shared information and satisfaction, despite the context of controversy.

Acceptance of diagnosis and management satisfaction of patients with "suspected Lyme borreliosis" after 12 months in a multidisciplinary reference center: a prospective cohort study

INTRODUCTION

Because patients with a "suspicion of Lyme borreliosis (LB)" may experience medical wandering and difficult care paths, often due to misinformation, multidisciplinary care centers were started all over Europe a few years ago. The aim of our study was to prospectively identify the factors associated with the acceptance of diagnosis and management satisfaction of patients, and to assess the concordance of the medical health assessment between physicians and patients 12 months after their management at our multidisciplinary center.

METHODS

We included all adults who were admitted to the Tick-Borne Diseases Reference Center of Paris and the Northern Region (TBD-RC) (2017-2020). A telephone satisfaction survey was conducted 12 months after their first consultation. It consisted of 5 domains and 13 items rated between 0 (lowest) and 10 (highest grade): (1)Reception; (2)Care and quality of management; (3)Information/explanations given to the patients; (4)Current medical condition and acceptance of the final diagnosis; (5)Overall appreciation. Factors associated with diagnosis acceptance and management satisfaction at 12 months were identified using logistic regression models. The concordance of the health status as assessed by doctors and patients was calculated using a Cohen's kappa test.

RESULTS

Of the 569 patients who consulted, 349 (61.3%) answered the questionnaire. Overall appreciation had a median rating of 9 [8;10] and 280/349 (80.2%) accepted their diagnoses. Patients who were "very satisfied" with their care paths at TBD-RC (OR = 4.64;CI95%[1.52-14.16]) had higher odds of diagnosis acceptance. Well-delivered information was strongly associated with better satisfaction with the management (OR = 23.39;CI95%[3.52-155.54]). The concordance between patients and physicians to assess their health status 12 months after their management at TBD-RC was almost perfect in the groups of those with confirmed and possible LB (κ = 0.99), and moderate in the group with other diagnoses (κ = 0.43).

CONCLUSION

Patients seemed to approve of this multidisciplinary care organization for suspected LB. It helped them to accept their final diagnoses and enabled a high level of satisfaction with the information given by the doctors, confirming the importance of shared medical decisions, which may help to reduce health misinformation. This type of structure may be useful for any disease with a complex and controversial diagnosis.

Positive intrathecal anti-Borrelia antibody synthesis: what are the implications for clinical practice? Clinical features and outcomes of 138 patients in a French multicenter cohort study

We aimed to describe the clinical characteristics, management, and residual symptoms (RS) in patients with definite and possible Lyme neuroborreliosis (LNB). We conducted a retrospective French multicenter cohort study (2010-2020). Cases of LNB were defined as clinical manifestations attributed to LNB and a positive Borrelia-specific intrathecal antibody index (AI) ("possible" LNB) and with pleocytosis ("definite" LNB). Risk factors of RS were determined using a logistic regression model. We included 138 adult patients with a positive AI. Mean age was 59.5 years (± 14.7). The median duration of symptoms before diagnosis was 1.0 [0.5-4.0] months. The most frequent manifestation was radicular pain (n = 79, 57%). Complete cerebrospinal fluid (CSF) leukocyte analysis was available in 131 patients, of whom 72 (55%) had pleocytosis. Patients with definite LNB had a shorter duration of symptoms (median 1.0 [0.5-2.6] vs. 3.0 [0.6-7.0] months, p < 0.01) and more radicular pain (74% vs 44%, p < 0.01) than patients with possible LNB. At the last visit (median duration of follow-up: 70 [30-175] days), 74/124 patients (59.7%) reported RS, mostly radicular pain (n = 31, 25%). In multivariate analysis, definite LNB (OR = 0.21 [0.05-0.931], p = 0.039) and duration of symptoms less than 3 months (OR = 0.04 [0.01-0.37], p = 0.005) were protective factors against RS at last follow-up. Our study highlights the challenges of LNB management, especially for patients with a positive AI without pleocytosis, questioning whether LB is still ongoing or not. Early diagnosis and treatment are important to improve outcomes and to lower potential RS.

Exploring convolutional neural networks with transfer learning for diagnosing Lyme disease from skin lesion images

BACKGROUND AND OBJECTIVE

Lyme disease which is one of the most common infectious vector-borne diseases manifests itself in most cases with erythema migrans (EM) skin lesions. Recent studies show that convolutional neural networks (CNNs) perform well to identify skin lesions from images. Lightweight CNN based pre-scanner applications for resource-constrained mobile devices can help users with early diagnosis of Lyme disease and prevent the transition to a severe late form thanks to appropriate antibiotic therapy. Also, resource-intensive CNN based robust computer applications can assist non-expert practitioners with an accurate diagnosis. The main objective of this study is to extensively analyze the effectiveness of CNNs for diagnosing Lyme disease from images and to find out the best CNN architectures considering resource constraints.

METHODS

First, we created an EM dataset with the help of expert dermatologists from Clermont-Ferrand University Hospital Center of France. Second, we benchmarked this dataset for twenty-three CNN architectures customized from VGG, ResNet, DenseNet, MobileNet, Xception, NASNet, and EfficientNet architectures in terms of predictive performance, computational complexity, and statistical significance. Third, to improve the performance of the CNNs, we used custom transfer learning from ImageNet pre-trained models as well as pre-trained the CNNs with the skin lesion dataset HAM10000. Fourth, for model explainability, we utilized Gradient-weighted Class Activation Mapping to visualize the regions of input that are significant to the CNNs for making predictions. Fifth, we provided guidelines for model selection based on predictive performance and computational complexity.

RESULTS

Customized ResNet50 architecture gave the best classification accuracy of 84.42% ±1.36, AUC of 0.9189±0.0115, precision of 83.1%±2.49, sensitivity of 87.93%±1.47, and specificity of 80.65%±3.59. A lightweight model customized from EfficientNetB0 also performed well with an accuracy of 83.13%±1.2, AUC of 0.9094±0.0129, precision of 82.83%±1.75, sensitivity of 85.21% ±3.91, and specificity of 80.89%±2.95. All the trained models are publicly available at https://dappem.limos.fr/download.html, which can be used by others for transfer learning and building pre-scanners for Lyme disease.

CONCLUSION

Our study confirmed the effectiveness of even some lightweight CNNs for building Lyme disease pre-scanner mobile applications to assist people with an initial self-assessment and referring them to expert dermatologist for further diagnosis.

Incidence, associated factors, and effect on renal function of amoxicillin crystalluria in patients receiving high doses of intravenous amoxicillin (The CRISTAMOX Study): A cohort study

BACKGROUND

Amoxicillin crystalluria (AC), potentially responsible for acute kidney injury (AKI), is reported more and more frequently in patients treated with high doses of intravenous amoxicillin (HDIVA). The main objective of this study was to evaluate AC incidence in these patients. The secondary objectives were to identify factors associated with AC and to evaluate its impact on the risk of AKI.

METHODS

This multicentre, observational, cohort study was conducted between Mar 18, 2014 and Aug 16, 2019 in Dijon, Nancy, and Reims University Hospitals as well as Châlon-sur-Saône, Charleville-Mézières, and Troyes general hospitals in France. Adult patients (≥18 years) treated with HDIVA and having been tested for AC at least once during treatment were included. Clinical, biological, and therapeutic characteristics of the patients were collected. A univariable mixed logistic regression model assessed the factors associated with AC. A multivariable Cox model with AC as a time-dependent variable assessed the prognostic factors for AKI. ClinicalTrials.gov number: NCT02853292.

FINDINGS

Of the 112 included patients, 27 (24.1%, 95% CI [16.2-32.0]) developed at least one episode of AC within a mean of 5.1 days. The factors associated with its occurrence were the concomitant use of angiotensin converting enzyme (ACE) inhibitors (OR=4.6, 95% CI [2.2-9.3], p<0.0001) and the decrease of urinary pH (OR=2.1 for one pH point decrease, 95% CI [1.2-3.7], p=0.009). 20 patients (17.9%) presented with AKI, within a mean time of 10.9 days. The main factor associated with the occurrence of AKI was the occurrence of AC (aHR=7.4, 95% CI [2.5-22.2], p=0.0003).

INTERPRETATION

AC occurred in a quarter of patients treated with HDIVA and was highly prognostic of AKI.

FUNDING

None.

Outcomes of patients admitted to intensive care units for acute manifestation of small-vessel vasculitis: a multicenter, retrospective study

Background

The outcomes of patients admitted to the intensive care unit (ICU) for acute manifestation of small-vessel vasculitis are poorly reported. The aim of the present study was to determine the mortality rate and prognostic factors of patients admitted to the ICU for acute small-vessel vasculitis.

Methods

This retrospective, multicenter study was conducted from January 2001 to December 2014 in 20 ICUs in France. Patients were identified from computerized registers of each hospital using the International Classification of Diseases, Ninth Revision (ICD-9). Inclusion criteria were (1) known or highly suspected granulomatosis with polyangiitis, eosinophilic granulomatosis with polyangiitis, microscopic polyangiitis (respectively, ICD-9 codes M31.3, M30.1, and M31.7), or anti–glomerular basement membrane antibody disease (ICD-9 codes N08.5X-005 or M31.0+); (2) admission to the ICU for the management of an acute manifestation of vasculitis; and (3) administration of a cyclophosphamide pulse in the ICU or within 48 h before admission to the ICU. The primary endpoint was assessment of mortality rate 90 days after admission to the ICU.

Results

Eighty-two patients at 20 centers were included, 94 % of whom had a recent (<6 months) diagnosis of small-vessel vasculitis. Forty-four patients (54 %) had granulomatosis with polyangiitis. The main reasons for admission were respiratory failure (34 %) and pulmonary-renal syndrome (33 %). Mechanical ventilation was required in 51 % of patients, catecholamines in 31 %, and renal replacement therapy in 71 %. Overall mortality at 90 days was 18 % and the mortality in ICU was 16 %. The main causes of death in the ICU were disease flare in 69 % and infection in 31 %. In univariable analysis, relevant factors associated with death in nonsurvivors compared with survivors were Simplified Acute Physiology Score II (median [interquartile range] 51 [38–82] vs. 36 [27–42], p = 0.005), age (67 years [62–74] vs. 58 years [40–68], p < 0.003), Sequential Organ Failure Assessment score on the day of cyclophosphamide administration (11 [6–12] vs. 6 [3–7], p = 0.0004), and delayed administration of cyclophosphamide (5 days [3–14] vs. 2 days [1–5], p = 0.0053).

Conclusions

Patients admitted to the ICU for management of acute small-vessel vasculitis benefit from early, aggressive intensive care treatment, associated with an 18 % death rate at 90 days.

Electronic supplementary material

The online version of this article (doi:10.1186/s13054-016-1189-5) contains supplementary material, which is available to authorized users.