Clinical prediction models for treatment outcomes in newly diagnosed epilepsy: A systematic review.

Up to 35% of individuals diagnosed with epilepsy continue to have seizures despite treatment, commonly referred to as drug-resistant epilepsy. Uncontrolled seizures can directly, or indirectly, negatively impact an individual's quality of life. To inform clinical management and life decisions, it is important to be able to predict the likelihood of seizure control. Those likely to achieve seizure control will be able to return sooner to their usual work and leisure activities and require less follow-up, whereas those with a poor prognosis will need more frequent clinical attendance and earlier consideration of epilepsy surgery. This is a systematic review aimed at identifying demographic, clinical, physiological (e.g., electroencephalographic), and imaging (e.g., magnetic resonance imaging) factors that may be predictive of treatment outcomes in patients with newly diagnosed epilepsy (NDE). MEDLINE and Embase were searched for prediction models of treatment outcomes in patients with NDE. Study characteristics were extracted and subjected to assessment of risk of bias (and applicability concerns) using the PROBAST (Prediction Model Risk of Bias Assessment Tool) tool. Baseline variables associated with treatment outcomes are reported as prognostic factors. After screening, 48 models were identified in 32 studies, which generally scored low for concerns of applicability, but universally scored high for susceptibility to bias. Outcomes reported fit broadly into four categories: drug resistance, short-term treatment response, seizure remission, and mortality. Prognostic factors were also heterogenous, but the predictors that were commonly significantly associated with outcomes were those related to seizure characteristics/types, epilepsy history, and age at onset. Antiseizure medication response was often included as a baseline variable, potentially obscuring other factor relationships at baseline. Currently, outcome prediction models for NDE demonstrate a high risk of bias. Model development could be improved with a stronger adherence to recommended TRIPOD (Transparent Reporting of a Multivariable Prediction Model for Individual Prognosis or Diagnosis) practices. Furthermore, we outline actionable changes to common practices that are intended to improve the overall quality of prediction model development in NDE.

Remote myocardial fibrosis predicts adverse outcome in patients with myocardial infarction on clinical cardiovascular magnetic resonance imaging.

BACKGROUND: Heart failure (HF) most commonly occurs in patients who have had a myocardial infarction (MI), but factors other than MI size may be deterministic. Fibrosis of myocardium remote from the MI is associated with adverse remodeling. We aimed to 1) investigate the association between remote myocardial fibrosis, measured using cardiovascular magnetic resonance (CMR) extracellular volume fraction (ECV), and HF and death following MI, 2) identify predictors of remote myocardial fibrosis in patients with evidence of MI and determine the relationship with infarct size. METHODS: Multicenter prospective cohort study of 1199 consecutive patients undergoing CMR with evidence of MI on late gadolinium enhancement. Median follow-up was 1133 (895-1442) days. Cox proportional hazards modeling was used to identify factors predictive of the primary outcome, a composite of first hospitalization for HF (HHF) or all-cause mortality, post-CMR. Linear regression modeling was used to identify determinants of remote ECV. RESULTS: Remote myocardial fibrosis was a strong predictor of primary outcome (χ2: 15.6, hazard ratio [HR]: 1.07 per 1% increase in ECV, 95% confidence interval [CI]: 1.04-1.11, p < 0.001) and was separately predictive of both HHF and death. The strongest predictors of remote ECV were diabetes, sex, natriuretic peptides, and body mass index, but, despite extensive phenotyping, the adjusted model R2 was only 0.283. The relationship between infarct size and remote fibrosis was very weak. CONCLUSION: Myocardial fibrosis, measured using CMR ECV, is a strong predictor of HHF and death in patients with evidence of MI. The mechanisms underlying remote myocardial fibrosis formation post-MI remain poorly understood, but factors other than infarct size appear to be important.

'Knowledge exchange' workshops to optimise development of a risk prediction tool to assist conveyance decisions for suspected seizures - Part of the Risk of ADverse Outcomes after a Suspected Seizure (RADOSS) project.

PURPOSE: Suspected seizures present challenges for ambulance services, with paramedics reporting uncertainty over whether or not to convey individuals to emergency departments. The Risk of ADverse Outcomes after a Suspected Seizure (RADOSS) project aims to address this by developing a risk assessment tool utilizing structured patient care record and dispatch data. It proposes a tool that would provide estimates of an individual's likelihood of death and/or recontact with emergency care within 3 days if conveyed compared to not conveyed, and the likelihood of an 'avoidable attendance' occurring if conveyed. Knowledge Exchange workshops engaged stakeholders to resolve key design uncertainties before model derivation. METHOD: Six workshops involved 26 service users and their significant others (epilepsy or nonepileptic attack disorder), and 25 urgent and emergency care clinicians from different English ambulance regions. Utilizing Nominal Group Techniques, participants shared views of the proposed tool, benefits and concerns, suggested predictors, critiqued outcome measures, and expressed functionality preferences. Data were analysed using Hamilton's Rapid Analysis. RESULTS: Stakeholders supported tool development, proposing 10 structured variables for predictive testing. Emphasis was placed on the tool supporting, not dictating, care decisions. Participants highlighted some reasons why RADOSS might struggle to derive a predictive model based on structured data alone and suggested some non-structured variables for future testing. Feedback on prediction timeframes for service recontact was received, along with advice on amending the 'avoidable attendance' definition to prevent the tool's predictions being undermined by potential overuse of certain investigations in hospital. CONCLUSION: Collaborative stakeholder engagement provided crucial insights that can guide RADOSS to develop a user-aligned, optimized tool.

Improved survival prediction and comparison of prognostic models for patients with hepatocellular carcinoma treated with sorafenib.

BACKGROUND: The 'Prediction Of Survival in Advanced Sorafenib-treated HCC' (PROSASH) model addressed the heterogeneous survival of patients with hepatocellular carcinoma (HCC) treated with sorafenib in clinical trials but requires validation in daily clinical practice. This study aimed to validate, compare and optimize this model for survival prediction. METHODS: Patients treated with sorafenib for HCC at five tertiary European centres were retrospectively staged according to the PROSASH model. In addition, the optimized PROSASH-II model was developed using the data of four centres (training set) and tested in an independent dataset. These models for overall survival (OS) were then compared with existing prognostic models. RESULTS: The PROSASH model was validated in 445 patients, showing clear differences between the four risk groups (OS 16.9-4.6 months). A total of 920 patients (n = 615 in training set, n = 305 in validation set) were available to develop PROSASH-II. This optimized model incorporated fewer and less subjective parameters: the serum albumin, bilirubin and alpha-foetoprotein, and macrovascular invasion, extrahepatic spread and largest tumour size on imaging. Both PROSASH and PROSASH-II showed improved discrimination (C-index 0.62 and 0.63, respectively) compared with existing prognostic scores (C-index ≤0.59). CONCLUSIONS: In HCC patients treated with sorafenib, individualized prediction of survival and risk group stratification using baseline prognostic and predictive parameters with the PROSASH model was validated. The refined PROSASH-II model performed at least as good with fewer and more objective parameters. PROSASH-II can be used as a tool for tailored treatment of HCC in daily practice and to define pre-planned subgroups for future studies.

Modelling seizure rates rather than time to an event within clinical trials of antiepileptic drugs.

BACKGROUND: Predictive models within epilepsy are frequently developed via Cox's proportional hazards models. These models estimate risk of a specified event such as 12-month remission. They are relatively simple to produce, have familiar output, and are useful to answer questions about short-term prognosis. However, the Cox model only considers time to first event rather than all seizures after starting treatment for example. This makes assessing change in seizure rates over time difficult. Variants to the Cox model exist enabling recurrent events to be modelled. One such variant is the Prentice, Williams and Peterson - Total Time (PWP-TT) model. An alternative is the negative binomial model for event counts. This study aims to demonstrate the differences between the three approaches, and to consider the benefits of the PWP-TT approach for assessing change in seizure rates over time. METHODS: Time to 12-month remission and time to first seizure after randomisation were modelled using the Cox model. Risk of seizure recurrence was modelled using the PWP-TT model, including all seizures across the whole follow-up period. Seizure counts were modelled using negative binomial regression. Differences between the approaches were demonstrated using participants recruited to the UK-based multi-centre Standard versus New Antiepileptic Drug (SANAD) study. RESULTS: Results from the PWP-TT model were similar to those from the conventional Cox and negative binomial models. In general, the direction of effect was consistent although the variables included in the models and the significance of the predictors varied. The confidence intervals obtained via the PWP-TT model tended to be narrower due to the increase in statistical power of the model. CONCLUSIONS: The Cox model is useful for determining the initial response to treatment and potentially informing when the next intervention may be required. The negative binomial model is useful for modelling event counts. The PWP-TT model extends the Cox model to all included events. This is useful in determining the longer-term effects of treatment policy. Such a model should be considered when designing future clinical trials in medical conditions typified by recurrent events to improve efficiency and statistical power as well as providing evidence regarding changes in event rates over time.

A systematic review of methodology used in the development of prediction models for future asthma exacerbation.

BACKGROUND: Clinical prediction models are widely used to guide medical advice and therapeutic interventions. Asthma is one of the most common chronic diseases globally and is characterised by acute deteriorations. These exacerbations are largely preventable, so there is interest in using clinical prediction models in this area. The objective of this review was to identify studies which have developed such models, determine whether consistent and appropriate methodology was used and whether statistically reliable prognostic models exist. METHODS: We searched online databases MEDLINE (1948 onwards), CINAHL Plus (1937 onwards), The Cochrane Library, Web of Science (1898 onwards) and ClinicalTrials.gov, using index terms relating to asthma and prognosis. Data was extracted and assessment of quality was based on GRADE and an early version of PROBAST (Prediction study Risk of Bias Assessment Tool). A meta-analysis of the discrimination and calibration measures was carried out to determine overall performance across models. RESULTS: Ten unique prognostic models were identified. GRADE identified moderate risk of bias in two of the studies, but more detailed quality assessment via PROBAST highlighted that most models were developed using highly selected and small datasets, incompletely recorded predictors and outcomes, and incomplete methodology. None of the identified models modelled recurrent exacerbations, instead favouring either presence/absence of an event, or time to first or specified event. Preferred methodologies were logistic regression and Cox proportional hazards regression. The overall pooled c-statistic was 0.77 (95% confidence interval 0.73 to 0.80), though individually some models performed no better than chance. The meta-analysis had an I2 value of 99.75% indicating a high amount of heterogeneity between studies. The majority of studies were small and did not include internal or external validation, therefore the individual performance measures are likely to be optimistic. CONCLUSIONS: Current prognostic models for asthma exacerbations are heterogeneous in methodology, but reported c-statistics suggest a clinically useful model could be created. Studies were consistent in lacking robust validation and in not modelling serial events. Further research is required with respect to incorporating recurrent events, and to externally validate tools in large representative populations to demonstrate the generalizability of published results.

Predictors of severe asthma attack re-attendance in Ecuadorian children: a cohort study.

Asthma is a common cause of emergency care attendance in low- and middle-income countries (LMICs). While few prospective studies of predictors for emergency care attendance have been undertaken in high-income countries, none have been performed in a LMIC.We followed a cohort of 5-15-year-old children treated for asthma attacks in emergency rooms of public health facilities in Esmeraldas City, Ecuador. We collected blood and nasal wash samples, and performed spirometry and exhaled nitric oxide fraction measurements. We explored potential predictors for recurrence of severe asthma attacks requiring emergency care over 6 months' follow-up.We recruited 283 children of whom 264 (93%) were followed-up for ≥6 months or until their next asthma attack. Almost half (46%) had a subsequent severe asthma attack requiring emergency care. Predictors of recurrence in adjusted analyses were (adjusted OR, 95% CI) younger age (0.87, 0.79-0.96 per year), previous asthma diagnosis (2.2, 1.2-3.9), number of parenteral corticosteroid courses in previous year (1.3, 1.1-1.5), food triggers (2.0, 1.1-3.6) and eczema diagnosis (4.2, 1.02-17.6). A parsimonious Cox regression model included the first three predictors plus urban residence as a protective factor (adjusted hazard ratio 0.69, 95% CI 0.50-0.95). Laboratory and lung function tests did not predict recurrence.Factors independently associated with recurrent emergency attendance for asthma attacks were identified in a low-resource LMIC setting. This study suggests that a simple risk-assessment tool could potentially be created for emergency rooms in similar settings to identify higher-risk children on whom limited resources might be better focused.

Identifying patients who will not reachieve remission after breakthrough seizures.

OBJECTIVE: We aim to identify people with epilepsy who are unlikely to reachieve a 12-month remission within 2 years after experiencing a breakthrough seizure following an initial 12-month remission. METHODS: We apply a novel longitudinal discriminant approach to data from the Standard and New Antiepileptic Drugs study to dynamically predict the risk of a patient not achieving a second remission after a breakthrough seizure by combining both baseline covariates (collected at the time of breakthrough seizure) and follow-up data. RESULTS: The model classifies 83% of patients. Of these, 73% of patients (95% confidence interval [CI] = 58%-88%) who did not achieve a second remission were correctly identified (sensitivity), and 84% of patients (95% CI = 69%-96%) who achieved a second remission were correctly identified (specificity). The area under the curve from our model was 87% (95% CI = 80%-94%). Patients who did not achieve a second remission were correctly identified on average after 10 months of observation postbreakthrough. Occurrence of seizures after breakthrough and the number of seizures experienced were the most informative longitudinal variables. These longitudinal profiles were influenced by the following baseline covariates: age at breakthrough seizure, presence of neurological insult, and number of antiepileptic drugs required to achieve first remission. SIGNIFICANCE: Using longitudinal data gathered during patient follow-up allows more accurate predictions than using baseline covariates in a standard Cox model. The model developed in this paper is a useful first step in developing a tool for identifying patients who develop drug resistance after an initial remission.

Effectiveness of Protease Inhibitor/Nucleos(t)ide Reverse Transcriptase Inhibitor-Based Second-line Antiretroviral Therapy for the Treatment of Human Immunodeficiency Virus Type 1 Infection in Sub-Saharan Africa: A Systematic Review and Meta-analysis.

Background: In sub-Saharan Africa, 25.5 million people are living with human immunodeficiency virus (HIV), representing 70% of the global total. The need for second-line antiretroviral therapy (ART) is projected to increase in the next decade in keeping with the expansion of treatment provision. Outcome data are required to inform policy. Methods: We performed a systematic review and meta-analysis of studies reporting the virological outcomes of protease inhibitor (PI)-based second-line ART in sub-Saharan Africa. The primary outcome was virological suppression (HIV-1 RNA <400 copies/mL) after 48 and 96 weeks of treatment. The secondary outcome was the proportion of patients with PI resistance. Pooled aggregate data were analyzed using a DerSimonian-Laird random effects model. Results: By intention-to-treat analysis, virological suppression occurred in 69.3% (95% confidence interval [CI], 58.2%-79.3%) of patients at week 48 (4558 participants, 14 studies), and in 61.5% (95% CI, 47.2%-74.9%) at week 96 (2145 participants, 8 studies). Preexisting resistance to nucleos(t)ide reverse transcriptase inhibitors (NRTIs) increased the likelihood of virological suppression. Major protease resistance mutations occurred in a median of 17% (interquartile range, 0-25%) of the virological failure population and increased with duration of second-line ART. Conclusions: One-third of patients receiving PI-based second-line ART with continued NRTI use in sub-Saharan Africa did not achieve virological suppression, although among viremic patients, protease resistance was infrequent. Significant challenges remain in implementation of viral load monitoring. Optimizing definitions and strategies for management of second-line ART failure is a research priority. Prospero Registration: CRD42016048985.

Comparing the Efficacy of Drug Regimens for Pulmonary Tuberculosis: Meta-analysis of Endpoints in Early-Phase Clinical Trials.

Background: A systematic review of early clinical outcomes in tuberculosis was undertaken to determine ranking of efficacy of drugs and combinations, define variability of these measures on different endpoints, and to establish the relationships between them. Methods: Studies were identified by searching PubMed, Medline, Embase, LILACS (Latin American and Caribbean Health Sciences Literature), and reference lists of included studies. Outcomes were early bactericidal activity results over 2, 7, and 14 days, and the proportion of patients with negative culture at 8 weeks. Results: One hundred thirty-three trials reporting phase 2A (early bactericidal activity) and phase 2B (culture conversion at 2 months) outcomes were identified. Only 9 drug combinations were assessed on >1 phase 2A endpoint and only 3 were assessed in both phase 2A and 2B trials. Conclusions: The existing evidence base supporting phase 2 methodology in tuberculosis is highly incomplete. In future, a broader range of drugs and combinations should be more consistently studied across a greater range of phase 2 endpoints.

Dynamic classification using credible intervals in longitudinal discriminant analysis.

Recently developed methods of longitudinal discriminant analysis allow for classification of subjects into prespecified prognostic groups using longitudinal history of both continuous and discrete biomarkers. The classification uses Bayesian estimates of the group membership probabilities for each prognostic group. These estimates are derived from a multivariate generalised linear mixed model of the biomarker's longitudinal evolution in each of the groups and can be updated each time new data is available for a patient, providing a dynamic (over time) allocation scheme. However, the precision of the estimated group probabilities differs for each patient and also over time. This precision can be assessed by looking at credible intervals for the group membership probabilities. In this paper, we propose a new allocation rule that incorporates credible intervals for use in context of a dynamic longitudinal discriminant analysis and show that this can decrease the number of false positives in a prognostic test, improving the positive predictive value. We also establish that by leaving some patients unclassified for a certain period, the classification accuracy of those patients who are classified can be improved, giving increased confidence to clinicians in their decision making. Finally, we show that determining a stopping rule dynamically can be more accurate than specifying a set time point at which to decide on a patient's status. We illustrate our methodology using data from patients with epilepsy and show how patients who fail to achieve adequate seizure control are more accurately identified using credible intervals compared to existing methods.

Ablation index, a novel marker of ablation lesion quality: prediction of pulmonary vein reconnection at repeat electrophysiology study and regional differences in target values.

AIMS: Force-Time Integral (FTI) is commonly used as a marker of ablation lesion quality during pulmonary vein isolation (PVI), but does not incorporate power. Ablation Index (AI) is a novel lesion quality marker that utilizes contact force, time, and power in a weighted formula. Furthermore, only a single FTI target value has been suggested despite regional variation in left atrial wall thickness. We aimed to study AI's and FTI's relationships with PV reconnection at repeat electrophysiology study, and regional threshold values that predicted no reconnection. METHODS AND RESULTS: Forty paroxysmal atrial fibrillation patients underwent contact force-guided PVI, and the minimum and mean AI and FTI values for each segment were identified according to a 12-segment model. All patients underwent repeat electrophysiology study at 2 months, regardless of symptoms, to identify sites of PV reconnection. Late PV reconnection was seen in 53 (11%) segments in 25 (62%) patients. Reconnected segments had significantly lower minimum AI [308 (252-336) vs. 373 (323-423), P < 0.0001] and FTI [137 (92-182) vs. 228 (157-334), P < 0.0001] compared with non-reconnected segments. Minimum AI and FTI were both independently predictive, but AI had a smaller P value. Higher minimum AI and FTI values were required to avoid reconnection in anterior/roof segments than for posterior/inferior segments (P < 0.0001). No reconnection was seen where the minimum AI value was ≥370 for posterior/inferior segments and ≥480 for anterior/roof segments. CONCLUSION: The minimum AI value in a PVI segment is independently predictive of reconnection of that segment at repeat electrophysiology study. Higher AI and FTI values are required for anterior/roof segments than for posterior/inferior segments to prevent reconnection.

A novel marker to predict early recurrence after atrial fibrillation ablation: the ablation effectiveness quotient.

INTRODUCTION: Inability to predict clinical outcome despite acutely successful pulmonary vein isolation (PVI) remains the Achilles' heel of atrial fibrillation ablation (AFA). Arrhythmia recurrence is frequently due to recovery of radiofrequency (RF) ablation lesions believed to be complete at the original procedure. OBJECTIVES: We hypothesized that a high ratio between post-AFA levels of serum high sensitivity cardiac troponin T (HScTnT), a highly specific marker of acute myocardial injury, and duration of RF application (the ablation effectiveness quotient, AEQ) would indicate effective ablation and correlate with early clinical success. METHODS: We prospectively measured HScTnT levels in 60 patients (42 [70%] male, 22 [37%] with paroxysmal AF [PAF], mean age 62.5 ± 10.6 years) 12-18 hours after AFA and calculated the AEQ for each. Patients were followed-up with ECGs and Holter monitors for recurrence of atrial tachyarrhythmia (AT). RESULTS: Early recurrence of AT within 6 months occurred in 22 (37%). AT recurrence was not significantly related to left atrial size or comorbidities, nor to RF time or HScTnT level. Mean AEQ was significantly lower in those with recurrence than those without (0.35 ± 0.14 ng/L/s vs. 0.45 ± 0.18 ng/L/s), P = 0.02. Subgroup analysis showed this finding was due to patients with PAF in whom early significance was maintained to one year, with an AEQ >0.4 ng/L/s having 75% sensitivity and 90% specificity in predicting freedom from AT. CONCLUSION: A high AEQ correlates well with freedom from AT in patients with PAF in both the short and medium term. If confirmed in further studies, AEQ may become a useful marker of risk of AT post-AFA.

Quality-of-life benefits of catheter ablation of persistent atrial fibrillation: a reanalysis of data from the SARA study.

AIMS: The recently published SARA study was a prospective, multi-centre randomized controlled trial that compared CA to antiarrhythmic drug therapy (ADT) in 146 patients with persistent atrial fibrillation (AF). The study found that recurrence of AF or atrial flutter occurred significantly less often in the CA arm compared to the ADT arm (29.6% vs. 56.3%, p = 0.002). Despite this clear superiority in terms of efficacy, the authors were not able to demonstrate a corresponding Quality of Life (QoL) improvement. We sought to investigate this apparent disparity using alternative analytical methods. METHODS AND RESULTS: We were able to show that a high coefficient of variation existed for all QoL measures at each time point which may explain the lack of statistical difference originally reported. We reanalyzed the raw QoL data from the SARA study using paired sample t-tests for the change in QOL for individual patients between baseline and 12 month (final) follow up. For patients randomized to ADT the difference in QoL after 12 months was not significant for any of the four QoL domains (global, physical, psychological and sexual) whereas for patients randomized to CA all comparisons were significant (global, p < 0.001; physical, p = 0.001; psychological, p < 0.001; sexual, p = 0.003). CONCLUSION: In the SARA study, after 12 months' follow up, CA significantly improved QoL for patients with persistent AF whereas medical therapy had no appreciable effect.

Improving safety in catheter ablation for atrial fibrillation: a prospective study of the use of ultrasound to guide vascular access.

INTRODUCTION: The most frequent complications of AF ablation (AFA) are related to vascular access, but there is little evidence as to how these can be minimized. METHODS: Consecutive patients undergoing AFA at a high-volume center received either standard care (Group S) or routine ultrasound-guided vascular access (Group U). Vascular complications were assessed before hospital discharge and by means of postal questionnaire 1 month later. Outcome measures were BARC 2+ bleeding complications, postprocedural pain, and prolonged bruising. RESULTS: Patients in Group S (n = 146) and U (n = 163) were well matched at baseline. Follow-up questionnaires were received from 92.6%. Patients in Group U were significantly less likely to have a BARC 2+ bleed, 10.4% versus 19.9% P = 0.02, were less likely to suffer groin pain after discharge (27.1% vs. 42.8%; P = 0.006) and were less likely to experience prolonged local bruising (21.5% vs. 40.4%; P = 0.001). Multivariable logistic regression analysis revealed a significant association of vascular complications with nonultrasound guided access (OR 3.12 95%CI 1.54-5.34; P = 0.003) and increasing age (OR 1.05 95%CI 1.01-1.09; P = 0.02). CONCLUSION: Routine use of ultrasound-guided vascular access for AFA is associated with a significant reduction in bleeding complications, postprocedural pain, and prolonged bruising when compared to standard care.

Time to 12-month remission and treatment failure for generalised and unclassified epilepsy.

OBJECTIVES: To develop prognostic models for time to 12-month remission and time to treatment failure after initiating antiepileptic drug monotherapy for generalised and unclassified epilepsy. METHODS: We analysed data from the Standard and New Antiepileptic Drug (arm B) study, a randomised trial that compared initiating treatment with lamotrigine, topiramate and valproate in patients diagnosed with generalised or unclassified epilepsy. Multivariable regression modelling was used to investigate how clinical factors affect the probability of achieving 12-month remission and treatment failure. RESULTS: Significant factors in the multivariable model for time to 12-month remission were having a relative with epilepsy, neurological insult, total number of tonic-clonic seizures before randomisation, seizure type and treatment. Significant factors in the multivariable model for time to treatment failure were treatment history (antiepileptic drug treatment prior to randomisation), EEG result, seizure type and treatment. CONCLUSIONS: The models described within this paper can be used to identify patients most likely to achieve 12-month remission and most likely to have treatment failure, aiding individual patient risk stratification and the design and analysis of future epilepsy trials.

Thiopurine monitoring in children with inflammatory bowel disease: a systematic review.

AIMS: The aim was to systematically review the evidence on the clinical usefulness of thiopurine metabolite and white blood count (WBC) monitoring in the assessment of clinical outcomes in children with inflammatory bowel disease (IBD). METHODS: Medline, Embase, Cochrane Central Register of controlled trials and http://www.clinicaltrials.gov were screened in adherence to the PRISMA statement by two independent reviewers for identification of eligible studies. Eligible studies were randomized controlled trials (RCTs), cohort studies and large case series of children with inflammatory bowel disease (IBD) (<18 years) who underwent monitoring of thiopurine metabolites and/or WBC. RESULTS: Fifteen papers were identified (n = 1026). None of the eligible studies were RCTs. High 6-thioguanine nucleotide (6TGN) concentrations were not consistently associated with leucopenia. Leucopenia was not associated with achievement of clinical remission. A positive but not consistent correlation between 6TGN and clinical remission was reported. Haematological toxicity could not be reliably assessed with 6TGN measurements only. A number of studies supported the use of high 6-methylmercaptopurine ribonucleotides (6MMPR) as an indicator of hepatotoxicity. Low thiopurine metabolite concentration may be indicative of non-compliance. CONCLUSION: Thiopurine metabolite testing does not safely predict clinical outcome, but may facilitate toxicity surveillance and treatment optimization in poor responders. Current evidence favours the combination of thiopurine metabolite/WBC monitoring and clinic follow-up for prompt identification of haematologic/hepatic toxicity safe dose adjustment, and treatment modification in cases of suboptimal clinical outcome or non-compliance. Well designed RCTs for the identification of robust surrogate markers of thiopurine efficacy and toxicity are required.

Systematic review of methods used in prediction models with recurrent event data.

BACKGROUND: Patients who suffer from chronic conditions or diseases are susceptible to experiencing repeated events of the same type (e.g. seizures), termed 'recurrent events'. Prediction models can be used to predict the risk of recurrence so that intervention or management can be tailored accordingly, but statistical methodology can vary. The objective of this systematic review was to identify and describe statistical approaches that have been applied for the development and validation of multivariable prediction models with recurrent event data. A secondary objective was to informally assess the characteristics and quality of analysis approaches used in the development and validation of prediction models of recurrent event data. METHODS: Searches were run in MEDLINE using a search strategy in 2019 which included index terms and phrases related to recurrent events and prediction models. For studies to be included in the review they must have developed or validated a multivariable clinical prediction model for recurrent event outcome data, specifically modelling the recurrent events and the timing between them. The statistical analysis methods used to analyse the recurrent event data in the clinical prediction model were extracted to answer the primary aim of the systematic review. In addition, items such as the event rate as well as any discrimination and calibration statistics that were used to assess the model performance were extracted for the secondary aim of the review. RESULTS: A total of 855 publications were identified using the developed search strategy and 301 of these are included in our systematic review. The Andersen-Gill method was identified as the most commonly applied method in the analysis of recurrent events, which was used in 152 (50.5%) studies. This was closely followed by frailty models which were used in 116 (38.5%) included studies. Of the 301 included studies, only 75 (24.9%) internally validated their model(s) and three (1.0%) validated their model(s) in an external dataset. CONCLUSIONS: This review identified a variety of methods which are used in practice when developing or validating prediction models for recurrent events. The variability of the approaches identified is cause for concern as it indicates possible immaturity in the field and highlights the need for more methodological research to bring greater consistency in approach of recurrent event analysis. Further work is required to ensure publications report all required information and use robust statistical methods for model development and validation. PROSPERO REGISTRATION: CRD42019116031.

Systematic review of clinical prediction models for psychosis in individuals meeting At Risk Mental State criteria.

Objectives: This study aims to review studies developing or validating a prediction model for transition to psychosis in individuals meeting At Risk Mental State (ARMS) criteria focussing on predictors that can be obtained as part of standard clinical practice. Prediction of transition is crucial to facilitating identification of patients who would benefit from cognitive behavioural therapy and, conversely, those that would benefit from less costly and less-intensive regular mental state monitoring. The review aims to determine whether prediction models rated as low risk of bias exist and, if not, what further research is needed within the field. Design: Bibliographic databases (PsycINFO, Medline, EMBASE, CINAHL) were searched using index terms relating to the clinical field and prognosis from 1994, the initial year of the first prospective study using ARMS criteria, to July 2024. Screening of titles, abstracts, and subsequently full texts was conducted by two reviewers independently using predefined criteria. Study quality was assessed using the Prediction model Risk Of Bias ASessment Tool (PROBAST). Setting: Studies in any setting were included. Primary and secondary outcome measures: The primary outcome for the review was the identification of prediction models considering transition risk and a summary of their risk of bias. Results: Forty-eight unique prediction models considering risk of transition to psychosis were identified. Variables found to be consistently important when predicting transition were age, gender, global functioning score, trait vulnerability, and unusual thought content. PROBAST criteria categorised four unique prediction models as having an overall low-risk bias. Other studies were insufficiently powered for the number of candidate predictors or lacking enough information to draw a conclusion regarding risk of bias. Conclusions: Two of the 48 identified prediction models were developed using current best practice statistical methodology, validated their model in independent data, and presented low risk of bias overall in line with the PROBAST guidelines. Any new prediction model built to evaluate the risk of transition to psychosis in people meeting ARMS criteria should be informed by the latest statistical methodology and adhere to the TRIPOD reporting guidelines to ensure that clinical practice is informed by the best possible evidence. External validation of such models should be carefully planned particularly considering generalisation across different countries. Systematic review registration: https://www.crd.york.ac.uk/PROSPEROFILES/108488_PROTOCOL_20191127.pdf, identifier CRD42018108488.

An international study to investigate and optimise the safety of discontinuing valproate in young men and women with epilepsy: Protocol.

Valproate is the most effective treatment for idiopathic generalised epilepsy. Currently, its use is restricted in women of childbearing potential owing to high teratogenicity. Recent evidence extended this risk to men's offspring, prompting recommendations to restrict use in everybody aged <55 years. This study will evaluate mortality and morbidity risks associated with valproate withdrawal by emulating a hypothetical randomised-controlled trial (called a "target trial") using retrospective observational data. The data will be drawn from ~250m mainly US patients in the TriNetX repository and ~60m UK patients in Clinical Practice Research Datalink (CPRD). These will be scanned for individuals aged 16-54 years with epilepsy and on valproate who either continued, switched to lamotrigine or levetiracetam, or discontinued valproate between 2014-2024, creating four groups. Randomisation to these groups will be emulated by baseline confounder adjustment using g-methods. Mortality and morbidity outcomes will be assessed and compared between groups over 1-10 years, employing time-to-first-event and recurrent events analyses. A causal prediction model will be developed from these data to aid in predicting the safest alternative antiseizure medications. Together, these findings will optimise informed decision-making about valproate withdrawal and alternative treatment selection, providing immediate and vital information for patients, clinicians and regulators.