Variables & COSY

Back to Prediction Scores
Click here

This calculation tool combines two distinct yet complementary approaches derived from the MRC Multicentre Trial for Early Epilepsy and Single Seizures (MESS) to assess the likelihood of seizure recurrence and inform decision-making in clinical and practical scenarios, including driving eligibility and treatment planning.

Part 1: Risk of Recurrence After a First Seizure and Implications for Driving

The Risk of Recurrence After a First Seizure and Implications for Driving tool calculates the likelihood of seizure recurrence after an individual’s first unprovoked seizure.

Variables Used:

  1. Aetiology:
    • Cryptogenic/Idiopathic: No identifiable cause or underlying condition.
    • Symptomatic: Seizure linked to a past condition (e.g., head injury, infection, stroke, or other neurological issues).
  2. Epilepsy in First-Degree Relative:
    • Yes: A history of epilepsy in immediate family members (parents or siblings).
    • No: No family history of epilepsy.
  3. Seizures Only While Asleep:
    • Yes: All seizures occurred exclusively during sleep.
    • No: Seizures occurred during wakefulness or a combination of sleep and wakefulness.
  4. EEG Results:
    • Normal: No epileptiform activity detected.
    • Abnormal: Presence of epileptiform activity (e.g., focal or generalized spikes, spike-and-wave patterns).
    • Not Done: EEG test was not conducted.
  5. CT/MRI Scan Results:
    • Normal: No structural abnormalities detected.
    • Abnormal: Structural abnormalities found (e.g., lesions, scarring, or other brain abnormalities).
    • Not Done: Imaging was not conducted.

Outputs:

  1. Risk of Seizure Recurrence:
    • This calculated risk is based on the variables entered and indicates the probability of a seizure recurrence.
  2. Chance of Seizure in the Next Year (COSY):
    • This tool informs whether an individual may be fit to drive, based on their specific seizure recurrence risk. 

Part 2: MESS Prognostic Index Tool

The MESS Prognostic Index Tool is designed to estimate the likelihood of seizure recurrence and guide treatment decisions after a single seizure or early epilepsy diagnosis. Derived from the MRC Multicentre Trial for Early Epilepsy and Single Seizures (MESS), the tool categorizes patients into six distinct groups based on a combination of risk level (low, medium, high) and treatment timing (immediate or delayed).

Variables Used:

    1. Number of Seizures Before Diagnosis:
      • Single Seizure: Indicates a low baseline risk of recurrence.
      • Two to Three Seizures: Suggests a moderate increase in risk.
      • Four or More Seizures: Associated with a high risk of recurrence.
    2. EEG Results:
      • Normal EEG: Indicates no abnormal epileptiform activity, contributing to a lower risk profile.
      • Abnormal EEG: Includes epileptiform or slow-wave abnormalities, significantly increasing the risk.
    3. Presence of Neurological Disorders:
      • No Neurological Disorder: Suggests a lower risk category.
      • Neurological Disorders (e.g., impairments, learning disabilities): Associated with higher seizure recurrence risks.
    4. Impact of Treatment Timing:
      • Immediate Treatment: Reduces recurrence risk in medium and high-risk groups during the initial years.
      • Delayed Treatment: Allows natural progression, recommended for low-risk patients where benefits of treatment are minimal.

Risk Classifications:

Patients are categorized into three primary risk groups based on the cumulative score from these variables:

  • Low Risk (0 points): Single seizure, normal EEG, and no neurological disorders.
  • Medium Risk (1 point): Two to three seizures, or a single seizure with either an abnormal EEG or neurological disorder.
  • High Risk (2–4 points): Four or more seizures, or a combination of abnormal EEG and neurological disorders.

Outputs:

  1.  
    1. Risk of Seizure Recurrence:
      • This calculated risk is based on the variables entered and indicates the probability of a seizure recurrence.
    2. Chance of Seizure in the Next Year (COSY):
      • This tool informs whether an individual may be fit to drive, based on their specific seizure recurrence risk.
When is it safe to return to driving following first-ever seizure (J W L Brown et al., 2015) DOI: 10.1136/jnnp-2013-307529
Click here

The “When is it Safe to Return to Driving Following First-Ever Seizure” Tool is designed to assess the safety and timing for resuming driving after a first seizure. It uses clinical and imaging data to calculate individualized risk and provides outputs for both seizure recurrence risk and COSY.

Variables Used:

  1. Aetiological Subgroup of First Seizure:
    • Options include:
      • Provoked Systemic: Acute systemic issues, e.g., metabolic disturbances.
      • Provoked CNS Lesion: Structural brain lesions identified via imaging.
      • Unprovoked Idiopathic: No identifiable cause; possibly genetic.
      • Unprovoked Remote Symptomatic: Linked to prior CNS insults like tumors or old strokes.
    • This variable is key to determining seizure recurrence probability.
  2. Slider for Time Points in Months:
    • Indicates the time interval in months since the first seizure.
    • This is essential to analyze the longitudinal risk changes over time.

Outputs:

  • Risk of Seizure Recurrence: Reflects the likelihood of another seizure occurring within the specified timeframe, adjusted for aetiology.
  • COSY (Chance of Seizure in the Next Year): Predicts the percentage risk of a seizure in the next 12 months, based on elapsed time since the first seizure.
Risk of Seizure Recurrence Due to Autoimmune Encephalitis With NMDAR, LGI1, CASPR2, and GABABR Antibodies: Implications for Return to Driving (Anna Rada et al., 2024) DOI: 10.1212/NXI.0000000000200225.
Click here

This tool provides the cumulative risk of seizure recurrence over time and the likelihood of experiencing a seizure in the upcoming months, based on the type of autoimmune encephalitis and the patient’s age group.

The study analyzed data from 981 patients across 14 international centers. Key inclusion criteria were:

  • Patients aged 15 years or older at disease onset.
  • At least one epileptic seizure during the course of the disease.
  • A seizure-free period of at least 3 months at any point.

Variables Used:

Type of Autoimmune Encephalitis:

  • NMDAR Encephalitis: Diagnosed by the presence of antibodies against the N-methyl-D-aspartate receptor. This subtype predominantly affects younger patients, with a mean age of 28 years, and includes 383 patients, 81.7% of whom are female.
  • LGI1 Encephalitis: Identified through antibodies targeting leucine-rich glioma-inactivated 1, primarily observed in older patients with a mean age of 63 years. This group comprises 440 patients, 34.3% of whom are female.
  • CASPR2 Encephalitis: Confirmed via antibodies against contactin-associated protein-like 2. This subtype includes 114 patients with a mean age of 64 years, predominantly male (95.6%).
  • GABABR Encephalitis: Detected by gamma-aminobutyric acid type B receptor antibodies. This smaller cohort comprises 44 patients with a mean age of 63 years, 43.2% of whom are female.

Age Category:

  • ≤70 Years: Includes patients with lower age-related risks of seizure recurrence.
  • >70 Years: Accounts for patients with increased seizure recurrence risks and associated age-related factors.

Time (Months):

  • Tracks seizure recurrence and probability dynamics over time, spanning up to 24 months for CoSy and 276 months for cumulative risk.

 

Outputs:

Cumulative Risk of Seizure Recurrence:

  • Provides the total percentage risk of seizure recurrence over time.
  • Based on the chosen age group and encephalitis type, this metric reflects the gradual increase in recurrence risk, starting at 0% and progressing dynamically across the selected time period.

 

Change of Seizure in the Next Year (COSY):

  • Estimates the likelihood of experiencing a seizure in the next year based on the selected time point .
Individualised prediction model of seizure recurrence and long-term outcomes after withdrawal of antiepileptic drugs in seizure-free patients: a systematic review and individual participant data meta-analysis (Herm J Lamberink et al., 2017) DOI: 10.1016/S1474-4422(17)30114-X.
Click here

The ASM Withdrawal Risk Retrieval Prediction Tool calculates the cumulative risk of seizure recurrence and long-term outcomes (e.g., seizure freedom) following the withdrawal of antiaeizure medications (ASMs) in seizure-free individuals. This tool is informed by meta-analytical data to guide individualized decision-making.

Variables Used:

  1. EEG Findings:
    • Options include “Normal” or “Pathological.”
    • Indicates whether EEG findings showed epileptiform activity before withdrawal.
    • Normal EEG correlates with a lower risk of recurrence compared to Pathological EEG.
  2. Time Since ASM Withdrawal:
    • A slider to input the time (in months) since ASM withdrawal began.
    • Calculates evolving risks based on longitudinal data.

Outputs:

  1. Risk of Recurrence:
    • Predicts the cumulative likelihood of seizure recurrence over time, helping physicians identify critical milestones for patients with varying EEG findings.
    • The tool highlights significant differences in risk between Normal and Pathological EEG groups.
  2. COSY (Change of Seizure in Next Year):
    • Indicates the likelihood of seizure occurrence in the next year for individuals, considering all patients or EEG-specific subgroups.
Prospective study of epilepsy with generalized tonic-clonic seizures alone: Clinical features, response to treatment, and likelihood of medication withdrawal (Jaafar et al., 2024) DOI: 10.1002/epi4.12981
Click here

This tool estimates the risk of seizure recurrence in patients with generalized tonic-clonic seizures (GTCs) after antiseizure medication (ASM) tapering. The predictions differentiate between patient-initiated and physician-initiated tapering and assess how treatment status (treated vs. untreated) impacts outcomes. The COSY (Change of Seizure in the Next Year) field calculates the likelihood of seizure frequency changes within the following year.

Variables Used:

  1. Patient-Initiated vs. Physician-Initiated ASM Taper:
    • Determines if ASM discontinuation was guided by a medical professional or undertaken independently by the patient. Physician-guided tapering is associated with a significantly lower risk of recurrence.
  2. Treatment Status (Untreated vs. Treated):
    • Indicates whether the patient received treatment with ASMs. Patients who remained untreated experienced a much higher recurrence rate (73%) compared to treated individuals (14%).
  3. Slider for Time Points in Months:
    • Allows users to input a specific time duration to calculate the cumulative risk of seizure recurrence at different points.

Outputs:

  1. Risk of Seizure Recurrence in % After x Months:
    • Provides the cumulative percentage risk of seizure recurrence based on whether the ASM discontinuation was physician-initiated or patient-initiated.
    • Example: Untreated patients show significantly higher risks, while treated individuals benefit from the structured tapering process.
  2. Change in Seizure in the Next Year (COSY) in % After x Months:
    • Displays the predicted percentage change in seizure occurrence frequency over the next year, accounting for the type of ASM discontinuation.
    • Predictions are based on data collected up to 72 months.
Epileptology of the first tonic-clonic seizure in adults and prediction of seizure recurrence (Koutroumanidis et al., 2018) DOI: 10.1684/epd.2018.1014.
Click here

This calculator estimates the risk of seizure recurrence and the predicted change in the likelihood of seizures (COSY) over time for adults experiencing a first tonic-clonic seizure. The tool leverages clinical and imaging data to provide personalized predictions based on epilepsy type and other factors.

Variables Used:

  1. Genetic vs. Non-Genetic Aetiology: Indicates whether the seizure has a genetic (e.g., hereditary conditions) or non-genetic origin (e.g., acquired conditions).
  2. Absences/Myoclonic Seizures vs. GTC Only: Presence of other seizure types like absences or myoclonic seizures versus generalized tonic-clonic (GTC) seizures alone.
  3. Temporal Lobe Epilepsy (TLE) vs. Frontal Lobe Epilepsy (FLE) vs. Undetermined Focus: The brain region where the seizure originates, assessed through imaging and EEG findings.
  4. Slider for Time Points in Months: Allows selection of a specific time interval for risk predictions.

Outputs:

  • Risk of Seizure Recurrence in % after X Months: Displays the cumulative likelihood of experiencing another seizure based on the variables selected, using longitudinal clinical data.
  • Change in Seizure in the Next Year (COSY) in % after X Months: Predicts the change in seizure occurrence over the next year, accounting for the type of epilepsy and other clinical factors.
Hippocampal sclerosis and temporal lobe epilepsy following febrile status epilepticus: The FEBSTAT study (Lewis et al., 2024) DOI: 10.1111/epi.17979
Click here

This tool predicts the risk of seizure recurrence and changes in seizure probability (COSY) after febrile status epilepticus (FSE), particularly focusing on hippocampal abnormalities detected through MRI. It provides an estimate of seizure recurrence risk and the likelihood of hippocampal sclerosis (HS) or mesial temporal lobe epilepsy (MTLE) based on detailed diagnostic imaging and patient characteristics.

Variables Used:

  1. MRI Findings:
    • T2 Hyperintensity: Indicates acute hippocampal inflammation or injury, strongly associated with increased risk of HS and MTLE.
    • Extra-hippocampal Abnormalities: Includes structural issues such as mass lesions or lobar atrophy.
    • Hippocampal Malrotation (HIMAL) or Dysmorphic Hippocampi: Suggests moderate risk due to structural irregularities.
    • T2 Normal: Indicates normal hippocampal appearance with the lowest risk.
  2. Time in Months (Slider): Specifies the interval since the FSE to assess longitudinal outcomes, providing seizure risks over different time periods.
  3. Risk Fields:
    • Risk of Seizure Recurrence: Displays cumulative risk (%) based on MRI findings and time interval.
    • Change in Seizure in the Next Year (COSY): Predicts seizure probability changes (%) over a one-year period depending on hippocampal and other MRI abnormalities.

Outputs:

  • Risk of Seizure Recurrence: Indicates the likelihood of experiencing another seizure at specific time points after FSE, factoring in MRI-defined hippocampal conditions. For patients with T2 Hyperintensity, the risk of MTLE and recurrent seizures is substantially higher compared to those with normal MRI findings.
  • COSY: Reflects the probability of seizure occurrence changes in the next year. For patients with abnormal hippocampal features (e.g., HIMAL or T2 Hyperintensity), COSY predicts worsening risks compared to those with normal imaging.
Risk of Epilepsy Diagnosis After a First Unprovoked Seizure in Dementia (Z Mahamud et al., 2020) DOI: 10.1016/j.seizure.2020.09.001.
Click here

This tool estimates the risk of seizure recurrence and changes in seizure probability (COSY) following a first unprovoked seizure in individuals diagnosed with dementia. The predictions focus on understanding the varying risks associated with different dementia subtypes, as explored in the study “Risk of epilepsy diagnosis after a first unprovoked seizure in dementia” (Mahamud et al., 2020). 

Variables Used:

Dementia Types:

  • Early-Onset Alzheimer’s Disease: Higher epilepsy risk compared to other subtypes.
  • Late-Onset Alzheimer’s Disease: Moderate risk with age-related influences.
  • Mixed Type Dementia: Combines Alzheimer’s and vascular dementia features.
  • Vascular Dementia: Increased epilepsy risk linked to cerebrovascular abnormalities.
  • Lewy Body Dementia: Lower epilepsy risk.
  • Frontotemporal Dementia: Notable epilepsy risk despite lower prevalence.
  • Parkinson’s Disease with Dementia: Minimal risk due to subcortical degeneration.
  • Unclassified Dementia and Other Types: Capture atypical or ambiguous dementia diagnoses.

Time Interval in Months :

Specifies the time since the first unprovoked seizure, allowing longitudinal analysis of seizure recurrence risks.

Outputs:

  • Risk of Seizure Recurrence: Provides individualized seizure recurrence risk estimates for different dementia subtypes and time points.
    • Example: Patients with early-onset Alzheimer’s show an elevated 5-year epilepsy risk compared to those with late-onset Alzheimer’s or vascular dementia.
  • COSY: Highlights dynamic changes in seizure probability over time.
    • Example: Risk tends to decrease for patients with low initial recurrence likelihood but remains high for early-onset Alzheimer’s cases.
Risk of Seizure Relapse After Antiepileptic Drug Withdrawal in Adult Patients with Focal Epilepsy
(Ru-Qian He et al., 2016) DOI: 10.1016/j.yebeh.2016.08.006. Epub 2016 Oct 17.
Click here

This tool estimates the risk of seizure recurrence and changes in seizure probability (COSY) following antiseizure medication (ASM) withdrawal in adults with focal epilepsy.  

Variables Used:

  • ASM Withdrawal:
    Withdrawal is defined as the planned tapering or cessation of ASMs after at least two years of seizure remission. Cave: Other risk factors should be kept in mind: patients with a shorter seizure-free period (<4 years), a longer history of active epilepsy, or those who underwent rapid tapering faced significantly higher risks of seizure recurrence.

  • Time Interval in Months:
    Specifies the months since ASM withdrawal (0–96 months for recurrence risk, 0–60 months for COSY), enabling longitudinal analysis of seizure recurrence probability and dynamic changes in seizure risk.

Outputs:

  • Risk of Seizure Recurrence:
    Provides individualized seizure recurrence risk estimates over time.

    • Example: 48.5% of recurrences occur in the first year post-withdrawal, and 79.8% occur within two years. After five years, the recurrence risk drops to below 1%.

  • COSY:
    Highlights the dynamic changes in seizure probability in the year following each time point.

    • Example: Patients with longer seizure-free durations before withdrawal show a reduced COSY compared to those with shorter remission periods.
Long-term seizure, psychiatric and socioeconomic outcomes after frontal
lobe epilepsy surgery (Anthony Khoo et al., 2022) DOI: 10.1016/j.eplepsyres.2022.106998.
Click here

This calculator provides data-driven predictions for individuals undergoing epilepsy surgery, incorporating key clinical, imaging, and pathological variables. The tool is designed to estimate outcomes over time using established longitudinal data from a cohort of 274 patients who underwent frontal lobe epilepsy surgery. These patients were followed up for a median period of 7.5 years, enabling the analysis of long-term surgical outcomes.

Variables

1. Age at the Time of Surgery (<30 vs. ≥30 years):

  • The tool categorizes individuals based on age at the time of surgery, grouping them into younger patients (under 30 years) and older patients (30 years and above).

2. Anti-Seizure Medications (ASMs) at the Time of Surgery (<4 vs. ≥4 ASMs):

  • Patients are grouped by the number of anti-seizure medications they are prescribed at the time of surgery.
    • Fewer than four ASMs typically indicates lower medication dependency.
    • Four or more ASMs often reflects drug-resistant epilepsy and higher disease burden.

3. MRI Findings (Focal Abnormality vs. Diffuse Abnormality/Normal MRI):

  • Imaging findings are classified into two categories:
    • Focal Abnormality: Defined as localized lesions that can be precisely identified on MRI scans, suggesting a specific pathological source of epilepsy.
    • Diffuse Abnormality/Normal MRI: Includes cases with widespread or poorly defined abnormalities, or normal imaging findings.

4. Pathology Type:

The pathological findings from resected brain tissue are categorized into the following groups based on histopathological examination:

  • Focal Cortical Dysplasia (FCD):
    • A developmental cortical malformation frequently associated with drug-resistant epilepsy. Diagnosis is confirmed through the presence of disrupted cortical layers, balloon cells, or other architectural anomalies.
  • Dysembryoplastic Neuroepithelial Tumor (DNET):
    • A benign glioneuronal tumor that commonly presents with longstanding epilepsy. Histological features include specific patterns of neuronal and glial cells within a myxoid matrix.
  • Cavernoma:
    • A vascular malformation composed of dilated, thin-walled blood vessels. Pathological examination often reveals hemosiderin deposits and surrounding gliosis, consistent with repeated micro-hemorrhages.
  • Gliosis:
    • A reactive condition characterized by astrocyte proliferation in response to injury, inflammation, or disease. The absence of neoplastic changes distinguishes gliosis from other pathologies.
  • Glioma:
    • Encompasses low- and high-grade tumors derived from glial cells, with classification based on the World Health Organization (WHO) grading system. Histological features include atypical cellular growth and mitotic activity, depending on the tumor grade.

Outcomes

Risk of Seizure Recurrence:

This model estimates the cumulative probability of seizure recurrence after surgery.

Change in Likelihood of Seizures Over Time:

The calculator also predicts the expected change in seizure likelihood in the year following a selected time point. 

 

Frontal lobe low-grade tumors seizure outcome: a pooled analysis of clinical predictors
(Martín A Merenzon et al., 2023) DOI: 10.1016/j.clineuro.2023.107600.
Click here

This tool estimates the risk of seizure recurrence and changes in seizure probability (COSY) following frontal lobectomy surgery for tumor-related epilepsy. It is intended to support clinicians in assessing individualized recurrence risks and guiding postoperative care decisions for patients undergoing this procedure.

Variables Used:

Frontal Lobectomy Surgery:

This variable defines whether a patient underwent a frontal lobectomy (Yes/No). Surgery outcomes were analyzed in the context of tumor-related epilepsy in low-grade glial or glioneuronal tumors.

Further Variables not included in the prediction model:

Frontal lobectomy outcomes are influenced by several critical factors. Tumor type plays a significant role, with included cases involving low-grade glial or glioneuronal tumors such as oligodendrogliomas (50.4%), astrocytomas, gangliogliomas, and dysembryoplastic neuroepithelial tumors (DNETs). The extent of resection (EOR) is a pivotal variable, as gross total resection (GTR) is associated with significantly higher odds of seizure freedom (OR = 8.77, 95% CI: 1.99–47.91, p = 0.006). Additionally, awake surgery, often performed for tumors near eloquent brain areas, is another positive predictive factor for seizure freedom (OR = 9.94, 95% CI: 1.93–87.81, p = 0.015). The time interval post-surgery, measured in months, allows tracking of the probability of seizure freedom over periods of up to 120 months, providing a longitudinal perspective on surgical outcomes.

Outputs:

Risk of Seizure Recurrence: Provides individualized seizure recurrence risk estimates over time after a frontal lobectomy.

  • Example: After surgery, 92.6% of patients remained seizure-free at three months, and 85.5% remained seizure-free at 27.3 months.

 

COSY: Highlights the dynamic changes in seizure probability following surgery.

 

Prognostic Factors and Longitudinal Change in Long-Term Outcome of Frontal Lobe Epilepsy Surgery
(Cuiping Xu et al., 2018) DOI: 10.1016/j.wneu.2018.08.192.
Click here

This tool estimates the risk of seizure recurrence and changes in seizure probability (COSY) following frontal lobectomy surgery for refractory frontal lobe epilepsy.

Variables Used:

Frontal Lobectomy Surgery:

  • This variable indicates whether a patient underwent a frontal lobectomy (Yes/No). Outcomes were analyzed for patients with refractory frontal lobe epilepsy, focusing on long-term postoperative seizure freedom and recurrence dynamics.

 

Further Variables Not Included in the Prediction Model:

Several critical factors influence surgical outcomes, though they are not directly integrated into this tool:

  • Imaging Predictors: Favorable outcomes were strongly associated with identifiable frontal lesions on preoperative magnetic resonance imaging (MRI). Patients with abnormal frontal lobe findings had significantly better outcomes compared to those with normal MRIs (P = 0.027).

  • Pathological Subgroups: Histopathological findings such as tumors (e.g., oligodendrogliomas, gangliogliomas) and malformations of cortical development (MCDs) were significant factors. Tumor-related etiologies were associated with better outcomes (62% seizure freedom at follow-up) than MCDs (52%).

  • Extent of Surgery and Intraoperative Monitoring: Complete resection of the epileptogenic zone (e.g., gross total resection or lobectomy) and intraoperative electrocorticography positively impacted outcomes, improving seizure freedom rates.

  • Electrophysiologic Predictors: Favorable outcomes correlated with localized ictal patterns and frontal rhythms on electroencephalography (EEG), while generalized or non-localized discharges were linked to poorer outcomes.

  • Acute Postoperative Seizures (APOS): APOS, defined as seizures occurring within the first postoperative week, were predictive of poorer long-term seizure control (P < 0.0001).

Outputs:

Risk of Seizure Recurrence:

  • This output provides an individualized, dynamic estimate of seizure recurrence over time following surgery. For instance, the estimated probability of seizure freedom was 57.3% at 1 year, 51.2% at 2 years, and 50.0% at 5 years postoperatively.

 

COSY (Change of Seizure in the Next Year):

  • COSY predicts the likelihood of a seizure occurring in the next year following surgery.

 

Seizure outcome and its predictors after frontal lobe epilepsy surgery
(Joseph Samuel P et al., 2019) DOI: 10.1111/ane.13139.
Click here

This tool estimates the risk of seizure recurrence and changes in seizure probability (COSY) following frontal lobe epilepsy surgery for drug-resistant frontal lobe epilepsy (FLE). 

 

Variables Used:

Frontal Lobe Surgery:
This variable captures whether a patient underwent frontal lobe epilepsy surgery (Yes/No). It evaluates outcomes in patients with refractory FLE, helping to assess the likelihood of seizure freedom over time.

 

Further Variables Not Included in the Prediction Model:

The following factors were identified in the study as predictors of long-term seizure freedom through multivariate analysis. These variables, while not integrated into the tool, provide crucial insights into surgical outcomes:

  1. Age at Surgery:
    Younger patients at the time of surgery were more likely to achieve seizure freedom (P = 0.032).

  2. Duration of Epilepsy:
    A shorter duration of epilepsy before surgery was a significant predictor of better outcomes (P = 0.021), emphasizing the importance of early intervention.

  3. Postoperative EEG Findings:
    The absence of interictal epileptiform discharges (IEDs) on postoperative EEGs predicted higher seizure freedom rates:

    • 7th–10th day (P = 0.005)
    • 3 months (P = 0.005)
    • 1 year (P = 0.017)

 

Outputs:

Risk of Seizure Recurrence:

  • This output predicts the probability of seizure recurrence over time following surgery. Based on study data, the probability of seizure freedom was 45% at 1 year, 34% at 2 years, and 14% at 5 years postoperatively.

 

COSY (Change of Seizure in the Next Year):

  • COSY estimates the likelihood of a seizure occurring in the next year following surgery.
Risk of Epilepsy Following a First Posttraumatic Seizure: A Register-Based Study (Markus Karlander et al., 2025) DOI: 10.1212/CPJ.0000000000200409.
Click here

This tool analysis epilepsy risk following a first posttraumatic seizure (PTS). The study utilized a register-based cohort design and included 4,239 individuals with a first seizure, of whom 2,286 had a prior traumatic brain injury (TBI) and 1,953 served as non-TBI controls. Data were sourced from Swedish national healthcare and population registers, covering hospitalizations for TBI between 2000 and 2010, with follow-up extending up to 10 years (median: 2.0 years). The median age was 51 years in the TBI group and 67 years in the non-TBI group. High diagnostic accuracy (over 90%) was achieved through the use of validated registry data.

Variables Used:

TBI Severity and Classification:
Participants were categorized into six levels of TBI severity based on ICD-10 diagnostic codes:

  1. Non-TBI (Control Group): No history of traumatic brain injury.
  2. Mild TBI: Minimal or transient symptoms, coded as S060.
  3. Fracture: Skull fractures without direct evidence of brain injury, coded as S020, S021, S027, S029.
  4. Extracerebral Injury: External injuries such as scalp or cranial damage, coded as S064, S065, S066.
  5. Diffuse Cerebral Injury: Widespread brain injuries, including concussions or cerebral edema, coded as S061, S062, S067-S069.
  6. Focal Cerebral Injury: Localized brain injuries, such as contusions or hematomas, coded as S063, associated with the highest risk of epilepsy.

 

Time Interval in Months:
The time from TBI to the first PTS was categorized into less than two years and more than two years. Earlier seizures (<2 years) were associated with a significantly higher risk of subsequent epilepsy.

Comorbidities:
The analysis also accounted for key co-variates, including stroke, CNS tumors, and CNS infections, which can influence the risk of epilepsy following a first PTS.

Outputs:

Risk of Seizure Recurrence:
This tool provides individualized risk estimates for epilepsy following a first PTS, stratified by TBI severity and the time interval since the injury.
Example: Patients with focal cerebral injuries and seizures occurring within two years post-TBI have a substantially elevated risk of seizure recurrence compared to those with mild TBI or seizures occurring more than two years after the injury.

COSY (Chance of a Seizure in the Next Year):
COSY quantifies the probability of experiencing another seizure within the next year, reflecting dynamic changes in seizure risk over time.
Example: For patients with mild TBI, the probability of another seizure decreases over time, whereas it remains consistently high for individuals with severe injuries, particularly focal cerebral injuries, or seizures occurring shortly after the injury.

Clinical and neuroimaging predictors of seizure recurrence in solitary calcified neurocysticercosis: A prospective observational study (Alok Kumar Singh et al., 2017) DOI: 10.1016/j.eplepsyres.2017.09.010.
Click here

This tool providesthe cumulative risk of seizure recurrence over time and the likelihood of experiencing a seizure in the upcoming months, based on the presence or absence of perilesional edema in solitary calcified neurocysticercosis.

 

Variables Used:

  1. Presence of Edema:

    • “Yes” indicates perilesional edema, significantly increasing the risk of seizure recurrence due to persistent neuroinflammation.
    • “No” indicates the absence of edema, associated with a lower recurrence risk and a more favorable prognosis.

  2. Time (Months):

    • The model tracks seizure recurrence and probability dynamics across 12 months.

 

Outputs:

  1. Cumulative Risk of Seizure Recurrence:

    • Provides the total percentage risk of seizure recurrence over time. For patients without edema, the risk reaches 18% by 12 months. For those with edema, the risk rises steeply to 69% by 12 months.

  2. Change of Seizure Probability (CoSy):

    • Estimates the change of a seizure in the next month. 
  • Guidelines for Use: This resource is intended for use by medical professionals. It is important to note that the prediction tool provided on this platform is not currently included in official clinical guidelines. Users should consider this limitation when utilizing the tool. The interpretation and application of the content and calculations presented are the sole responsibility of the user. This tool is designed to supplement, not replace, standard medical care and the comprehensive information patients should receive from their healthcare providers. It serves as an adjunct to clinical decision-making and should not be used as a standalone diagnostic or treatment tool. Users are encouraged to emphasize to patients that the results generated by the tool are based on scientific research, with links to the relevant publications available alongside each score. These publications provide additional context and detail about the underlying research methodology and findings. Consequently, the outcomes presented may not accurately reflect the unique circumstances of an individual patient.
  • 2024, University Hospital Zurich