Mistaken Identity: Many Diagnoses are Frequently Misattributed to Lyme Disease

Mapping the distribution of Lyme disease at a mid-Atlantic site in the United States using electronic health data

Lyme disease is a spatially heterogeneous tick-borne infection, with approximately 85% of US cases concentrated in the mid-Atlantic and northeastern states. Surveillance for Lyme disease and its causative agent, including public health case reporting and entomologic surveillance, is necessary to understand its endemic range, but currently used case detection methods have limitations. To evaluate an alternative approach to Lyme disease surveillance, we have performed a geospatial analysis of Lyme disease cases from the Johns Hopkins Health System in Maryland. We used two sources of cases: a) individuals with both a positive test for Lyme disease and a contemporaneous diagnostic code consistent with a Lyme disease-related syndrome; and b) individuals referred for a Lyme disease evaluation who were adjudicated to have Lyme disease. Controls were individuals from the referral cohort judged not to have Lyme disease. Residential address data were available for all cases and controls. We used a hierarchical Bayesian model with a smoothing function for a coordinate location to evaluate the probability of Lyme disease within 100 km of Johns Hopkins Hospital. We found that the probability of Lyme disease was greatest in the north and west of Baltimore, and the local probability that a subject would have Lyme disease varied by as much as 30-fold. Adjustment for demographic and ecological variables partially attenuated the spatial gradient. Our study supports the suitability of electronic medical record data for the retrospective surveillance of Lyme disease.

Evaluation and 1-year follow-up of patients presenting at a Lyme borreliosis expertise centre: a prospective cohort study with validated questionnaires

OBJECTIVES

To describe the course of symptoms reported by patients with symptoms attributed to Lyme borreliosis (LB) without being subsequently diagnosed with LB.

METHODS

We performed a prospective cohort study with patients presenting at the outpatient clinic of two clinical LB centres. The primary outcome was the prevalence of persistent symptoms, which were defined as clinically relevant fatigue (CIS, subscale fatigue), pain (SF-36, subscale bodily pain), and cognitive impairment (CFQ) for ≥ 6 months and onset < 6 months over the first year of follow-up. Outcomes were compared with a longitudinal cohort of confirmed LB patients and a general population cohort. Prevalences were standardised to the distribution of pre-defined confounders in the confirmed LB cohort.

RESULTS

Participants (n = 123) reported mostly fatigue, arthralgia, myalgia, and paraesthesia as symptoms. The primary outcome could be determined for 74.8% (92/123) of participants. The standardised prevalence of persistent symptoms in our participants was 58.6%, which was higher than in patients with confirmed LB at baseline (27.2%, p < 0.0001) and the population cohort (21.2%, p < 0.0001). Participants reported overall improvement of fatigue (p < 0.0001) and pain (p < 0.0001) but not for cognitive impairment (p = 0.062) during the follow-up, though symptom severity at the end of follow-up remained greater compared to confirmed LB patients (various comparisons p < 0.05).

CONCLUSION

Patients with symptoms attributed to LB who present at clinical LB centres without physician-confirmed LB more often report persistent symptoms and report more severe symptoms compared to confirmed LB patients and a population cohort.

Coinfection of Babesia and Borrelia in the Tick Ixodes ricinus-A Neglected Public Health Issue in Europe?

Ixodes ricinus nymphs and adults removed from humans, and larvae and nymphs from birds, have been analysed for infection with Babesia species and Borrelia species previously in separately published studies. Here, we use the same data set to explore the coinfection pattern of Babesia and Borrelia species in the ticks. We also provide an overview of the ecology and potential public health importance in Sweden of I. ricinus infected both with zoonotic Babesia and Borrelia species. Among 1952 nymphs and adult ticks removed from humans, 3.1% were PCR-positive for Babesia spp. Of these Babesia-positive ticks, 43% were simultaneously Borrelia-positive. Among 1046 immatures of I. ricinus removed from birds, 2.5% were Babesia-positive, of which 38% were coinfected with Borrelia species. This study shows that in I. ricinus infesting humans or birds in Sweden, potentially zoonotic Babesia protozoa sometimes co-occur with human-pathogenic Borrelia spp. Diagnostic tests for Babesia spp. infection are rarely performed in Europe, and the medical significance of this pathogen in Europe could be underestimated.

Leveraging machine learning approaches for predicting potential Lyme disease cases and incidence rates in the United States using Twitter

BACKGROUND

Lyme disease is one of the most commonly reported infectious diseases in the United States (US), accounting for more than [Formula: see text] of all vector-borne diseases in North America.

OBJECTIVE

In this paper, self-reported tweets on Twitter were analyzed in order to predict potential Lyme disease cases and accurately assess incidence rates in the US.

METHODS

The study was done in three stages: (1) Approximately 1.3 million tweets were collected and pre-processed to extract the most relevant Lyme disease tweets with geolocations. A subset of tweets were semi-automatically labelled as relevant or irrelevant to Lyme disease using a set of precise keywords, and the remaining portion were manually labelled, yielding a curated labelled dataset of 77, 500 tweets. (2) This labelled data set was used to train, validate, and test various combinations of NLP word embedding methods and prominent ML classification models, such as TF-IDF and logistic regression, Word2vec and XGboost, and BERTweet, among others, to identify potential Lyme disease tweets. (3) Lastly, the presence of spatio-temporal patterns in the US over a 10-year period were studied.

RESULTS

Preliminary results showed that BERTweet outperformed all tested NLP classifiers for identifying Lyme disease tweets, achieving the highest classification accuracy and F1-score of [Formula: see text]. There was also a consistent pattern indicating that the West and Northeast regions of the US had a higher tweet rate over time.

CONCLUSIONS

We focused on the less-studied problem of using Twitter data as a surveillance tool for Lyme disease in the US. Several crucial findings have emerged from the study. First, there is a fairly strong correlation between classified tweet counts and Lyme disease counts, with both following similar trends. Second, in 2015 and early 2016, the social media network like Twitter was essential in raising popular awareness of Lyme disease. Third, counties with a high incidence rate were not necessarily related with a high tweet rate, and vice versa. Fourth, BERTweet can be used as a reliable NLP classifier for detecting relevant Lyme disease tweets.

Citizen Science on Lyme Borreliosis in Spain Reveals Disease-Associated Risk Factors and Control Interventions

Background: Lyme borreliosis (LB) caused by Borrelia burgdorferi sensu lato complex spirochetes is one of the tick-borne diseases with high prevalence and social/economic burden in the United States, Spain, and other European countries. The objective is to address limited information available about the incidence, prevalence, and symptoms of LB, current prevention, and treatment interventions that are not adequately focused and thus not very effective against this disease. Methods: To address these limitations, in this study, we used a citizen science approach to evaluate the LB-associated risks and implementation of control interventions in Spain. A total of 405 participants in the survey were included in the analysis. Responses to the questionnaire were received during January-July 2022. The questionnaire contained qualitative and quantitative questions. Homogeneity among binary variables was analyzed using a Fisher's exact test. Results: Despite limitations of the study associated with response to the questionnaire and information on tick species, the results evidenced the effect of factors such as age, gender, tick bites, disease clinical signs, comorbidities such as alpha-gal syndrome, health care services, and treatment effectiveness affecting LB. Conclusions: The main conclusions of the study highlight the need for better surveillance of tick infestations, pathogen infection, and diagnosis of LB and related comorbidities. To advance in disease prevention, diagnosis, and treatment, new interventions need to be developed and implemented in both public and private health care services. Providing access to these results to the society, health care system, and scientists is important to further advance in disease surveillance, diagnosis, control, and prevention.

Diagnostic Testing for Lyme Disease

Standard 2-tier testing (STTT), incorporating a screening enzyme immunoassay (EIA) or an immunofluorescence assay (IFA) that reflexes to IgM and IgG immunoblots, has been the primary diagnostic test for Lyme disease since 1995. In 2019, the Food and Drug Administration approved a modified 2-tier test strategy using 2 EIAs: offering a faster, less expensive, and more sensitive assay compared with STTT. New technologies examine early immune responses to Borrelia burgdorferi have the potential to diagnose Lyme disease in the first weeks of infection when existing serologic testing is not recommended due to low sensitivity.

An Efficient Stacked Deep Transfer Learning Model for Automated Diagnosis of Lyme Disease

Lyme disease is one of the most common vector-borne infections. It typically causes cardiac illnesses, neurologic illnesses, musculoskeletal disorders, and dermatologic conditions. However, most of the time, it is poorly diagnosed due to many similarities with other diseases such as drug rash. Given the potentially serious consequences of unnecessary antimicrobial treatments, it is essential to understand frequent and uncommon diagnoses that explain symptoms in this population. Recently, deep learning models have been used for the diagnosis of various rash-related diseases. However, these models suffer from overfitting and color variation problems. To overcome these problems, an efficient stacked deep transfer learning model is proposed that can efficiently distinguish between patients infected with Lyme (+) or infected with other infections. 2^nd order edge-based color constancy is used as a preprocessing approach to reduce the impact of multisource light from images acquired under different setups. The AlexNet pretrained learning model is used for building the Lyme disease diagnosis model. To prevent overfitting, data augmentation techniques are also used to augment the dataset. In addition, 5-fold cross-validation is also used. Comparative analysis indicates that the proposed model outperforms the existing models in terms of accuracy, f-measure, sensitivity, specificity, and area under the curve.