RESUMO
BACKGROUND: Ovarian cancer is the most lethal and endometrial cancer the most common gynaecological cancer in the UK, yet neither have a screening program in place to facilitate early disease detection. The aim is to evaluate whether online search data can be used to differentiate between individuals with malignant and benign gynaecological diagnoses. METHODS: This is a prospective cohort study evaluating online search data in symptomatic individuals (Google user) referred from primary care (GP) with a suspected cancer to a London Hospital (UK) between December 2020 and June 2022. Informed written consent was obtained and online search data was extracted via Google takeout and anonymised. A health filter was applied to extract health-related terms for 24 months prior to GP referral. A predictive model (outcome: malignancy) was developed using (1) search queries (terms model) and (2) categorised search queries (categories model). Area under the ROC curve (AUC) was used to evaluate model performance. 844 women were approached, 652 were eligible to participate and 392 were recruited. Of those recruited, 108 did not complete enrollment, 12 withdrew and 37 were excluded as they did not track Google searches or had an empty search history, leaving a cohort of 235. RESULTS: The cohort had a median age of 53 years old (range 20-81) and a malignancy rate of 26.0%. There was a difference in online search data between those with a benign and malignant diagnosis, noted as early as 360 days in advance of GP referral, when search queries were used directly, but only 60 days in advance, when queries were divided into health categories. A model using online search data from patients (n = 153) who performed health-related search and corrected for sample size, achieved its highest sample-corrected AUC of 0.82, 60 days prior to GP referral. CONCLUSIONS: Online search data appears to be different between individuals with malignant and benign gynaecological conditions, with a signal observed in advance of GP referral date. Online search data needs to be evaluated in a larger dataset to determine its value as an early disease detection tool and whether its use leads to improved clinical outcomes.
Assuntos
Neoplasias dos Genitais Femininos , Neoplasias Ovarianas , Humanos , Feminino , Adulto Jovem , Adulto , Pessoa de Meia-Idade , Idoso , Idoso de 80 Anos ou mais , Neoplasias dos Genitais Femininos/diagnóstico , Estudos Prospectivos , Detecção Precoce de Câncer , Londres/epidemiologiaRESUMO
BACKGROUND: Selective Serotonin Reuptake Inhibitors (SSRIs) represent a diverse class of medications widely prescribed for depression and anxiety. Despite their common use, there is an absence of large-scale, real-world evidence capturing the heterogeneity in their effects on individuals. This study addresses this gap by utilizing naturalistic search data to explore the varied impact of six different SSRIs on user behavior. METHODS: The study sample included â¼508 thousand Bing users with searches for one of six SSRIs (citalopram, escitalopram, fluoxetine, fluvoxamine, paroxetine, sertraline) from April-December 2022, comprising 510 million queries. Cox proportional hazard models were employed to examine 30 topics (e.g., shopping, tourism, health) and 195 health symptoms (e.g., anxiety, weight gain, impotence), using each SSRI as a reference. We assessed the relative hazard ratios between drugs and, where feasible, ranked the SSRIs based on their observed effects. We used Cox proportional hazard models in order to account for both the likelihood of users searching for a particular topic or symptom and the associated time to that search. The temporal aspect aided in distinguishing between potential symptoms of the disorder, short-term medication side effects, and later appearing side effects. RESULTS: Differences were found in search behaviors associated with each SSRI. E.g., fluvoxamine was associated with a significantly higher likelihood of searching weight gain compared to all other SSRIs (HRs 1.85-2.93). Searches following citalopram were associated with significantly higher rates of later impotence queries compared to all other SSRIs (HRs 5.11-7.76), except fluvoxamine. Fluvoxamine was associated with a significantly higher rate of health related searches than all other SSRIs (HRs 2.11-2.36). CONCLUSIONS: Our study reveals new insights into the varying SSRI impacts, suggesting distinct symptom profiles. This novel use of large-scale, naturalistic search data contributes to pharmacovigilance efforts, enhancing our understanding of intra-class variation among SSRIs, potentially uncovering previously unidentified drug effects.
RESUMO
There is an appreciable mental health treatment gap in the United States. Efforts to bridge this gap and improve resource accessibility have led to the provision of online, clinically-validated tools for mental health self-assessment. In theory, these screens serve as an invaluable component of information-seeking, representing the preparative and action-oriented stages of this process while altering or reinforcing the search content and language of individuals as they engage with information online. Accordingly, this work investigated the association of screen completion with mental health-related search behaviors. Three-year internet search histories from N=7,572 Microsoft Bing users were paired with their respective depression, anxiety, bipolar disorder, or psychosis online screen completion and sociodemographic data available through Mental Health America. Data was transformed into network representations to model queries as discrete steps with probabilities and times-to-transition from one search type to another. Search data subsequent to screen completion was also modeled using Markov chains to simulate likelihood trajectories of different search types through time. Differences in querying dynamics relative to screen completion were observed, with searches involving treatment, diagnosis, suicidal ideation, and suicidal intent commonly emerging as the highest probability behavioral information seeking endpoints. Moreover, results pointed to the association of low risk states of psychopathology with transitions to extreme clinical outcomes (i.e., active suicidal intent). Future research is required to draw definitive conclusions regarding causal relationships between screens and search behavior.
RESUMO
We propose a method to estimate the household secondary attack rate (hSAR) of COVID-19 in the United Kingdom based on activity on the social media platform X, formerly known as Twitter. Conventional methods of hSAR estimation are resource intensive, requiring regular contact tracing of COVID-19 cases. Our proposed framework provides a complementary method that does not rely on conventional contact tracing or laboratory involvement, including the collection, processing, and analysis of biological samples. We use a text classifier to identify reports of people tweeting about themselves and/or members of their household having COVID-19 infections. A probabilistic analysis is then performed to estimate the hSAR based on the number of self or household, and self and household tweets of COVID-19 infection. The analysis includes adjustments for a reluctance of Twitter users to tweet about household members, and the possibility that the secondary infection was not acquired within the household. Experimental results for the UK, both monthly and weekly, are reported for the period from January 2020 to February 2022. Our results agree with previously reported hSAR estimates, varying with the primary variants of concern, e.g. delta and omicron. The serial interval (SI) is based on the time between the two tweets that indicate a primary and secondary infection. Experimental results, though larger than the consensus, are qualitatively similar. The estimation of hSAR and SI using social media data constitutes a new tool that may help in characterizing, forecasting and managing outbreaks and pandemics in a faster, affordable, and more efficient manner.
RESUMO
Text messaging can promote healthy behaviors, like adherence to medication, yet its effectiveness remains modest, in part because message content is rarely personalized. Reinforcement learning has been used in consumer technology to personalize content but with limited application in healthcare. We tested a reinforcement learning program that identifies individual responsiveness ("adherence") to text message content and personalizes messaging accordingly. We randomized 60 individuals with diabetes and glycated hemoglobin A1c [HbA1c] ≥ 7.5% to reinforcement learning intervention or control (no messages). Both arms received electronic pill bottles to measure adherence. The intervention improved absolute adjusted adherence by 13.6% (95%CI: 1.7%-27.1%) versus control and was more effective in patients with HbA1c 7.5- < 9.0% (36.6%, 95%CI: 25.1%-48.2%, interaction p < 0.001). We also explored whether individual patient characteristics were associated with differential response to tested behavioral factors and unique clusters of responsiveness. Reinforcement learning may be a promising approach to improve adherence and personalize communication at scale.