Association between probable postictal psychosis and subsequent diagnosis of primary psychotic disorders: a retrospective cohort study Authors: Varsha Nandwana, Geetika Bajpai, Ching-Fang Sun, Anita S Kablinger, Aashit K Shah Published: 2026 ABSTRACT AbstractRationalePostictal psychosis (PIP) is a transient psychiatric phenomenon characterised by psychotic symptoms following a seizure. While its immediate manifestations have been well documented, there is limited understanding of its long-term implications, particularly its relationship with subsequent primary psychotic disorders.MethodsWe conducted a retrospective cohort study using de-identified patient data from the TriNetX database (2006–2026). Individuals aged 18–65 years with a diagnosis of probable PIP (International Classification of Diseases, Tenth Revision, Clinical Modification (ICD-10-CM): F06.0, F06.2) or brief psychotic disorder within 7 days of the epileptic disorder diagnosis (ICD-10-CM: F23) were propensity-score matched 1:1 with controls without PIP based on age, sex, race, ethnicity and relevant psychiatric and chronic physical comorbidities. We excluded individuals with prior psychotic disorders or delirium. The primary outcome was any psychotic disorder diagnosis within a 5-year follow-up period. Kaplan-Meier survival analyses and HRs were calculated to assess the differences in survival probabilities.ResultsA total of 1275 individuals were identified and matched for analysis. Patients with probable PIP were at significantly higher risk for subsequent psychotic disorders (HR=10.78, 95% CI 6.21 to 18.70, p=0.002). The 5-year survival probability for developing any psychotic disorder was 85.79% in the probable PIP group compared with 98.21% in controls (log-rank test, χ2=112.25; p<0.0001).ConclusionsProbable PIP is associated with a heightened risk of subsequent primary psychotic disorders, including interictal forms, within 5 years.Keywords: NEUROPSYCHIATRY, EPILEPSY WHAT IS ALREADY KNOWN ON THIS TOPIC Postictal psychosis is an established acute psychiatric complication of epilepsy, but its long-term outcomes are not well defined. Postictal psychosis is an established acute psychiatric complication of epilepsy, but its long-term outcomes are not well defined. WHAT THIS STUDY ADDS This study shows that individuals with probable postictal psychosis have a substantially higher risk of later psychotic disorders compared with patients with seizures alone. This study shows that individuals with probable postictal psychosis have a substantially higher risk of later psychotic disorders compared with patients with seizures alone. HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY These findings support closer psychiatric follow-up after postictal psychosis and highlight the need for future studies on early identification and prevention of chronic psychosis in this population. These findings support closer psychiatric follow-up after postictal psychosis and highlight the need for future studies on early identification and prevention of chronic psychosis in this population. Introduction Epilepsy is frequently accompanied by psychiatric conditions, which may precede, coexist or result from seizures. Postictal psychosis (PIP) is a transient yet severe neuropsychiatric phenomenon characterised by the emergence of psychotic symptoms following the resolution of epileptic seizures.1–3 These symptoms typically include delusions, hallucinations and disorganised thinking, with an onset occurring hours to days postictally. While the acute clinical presentation of PIP has been extensively studied, less is understood about its long-term consequences, particularly its association with chronic psychotic disorders.4 Epilepsy itself has been associated with an elevated risk of psychosis, potentially due to shared neuropathological and genetic underpinnings involving abnormal cortical excitability, altered neurotransmitter systems and structural brain abnormalities.5 Recent population-based studies have further highlighted the substantial burden of psychiatric comorbidity in epilepsy, including psychotic disorders.6 Epilepsy and psychosis can significantly impact a person’s quality of life, with substantial economic implications. PIP, situated at the nexus of epilepsy and psychosis, may represent a unique clinical phenotype that predisposes individuals to later-onset primary psychotic disorders (including interictal or primary forms), such as schizophrenia spectrum disorders.7 8 However, evidence supporting this association remains limited, and the mechanistic pathways remain poorly understood. This study aims to address this gap by evaluating the risk of subsequent psychotic disorders in individuals with PIP. Methods Study design This retrospective cohort study was conducted on 13 June 2026, using the TriNetX Research Network. This real-time electronic medical record network provides access to de-identified electronic medical information from 111 large healthcare institutions. The study aimed to evaluate the long-term risk of primary psychotic disorders following a diagnosis of probable PIP in individuals with epilepsy. Setting Data were extracted from the TriNetX electronic health record (EHR) network, which provides access to longitudinal, real-world clinical data including demographics, diagnoses, procedures, laboratory results and medication histories. The study period spanned from 2006 to 2026. Participants We included patients aged 18–65 years who had a diagnosis of seizure disorder or epileptic activity (International Classification of Diseases, Tenth Revision, Clinical Modification (ICD-10: G40.0–G40.4, G40.41, R56.9) and a subsequent diagnosis of psychotic disorder (ICD-10: F06.0, F06.2 or F23) occurring within 7 days of the seizure diagnosis. Because seizure diagnoses in EHRs may not correspond precisely to the timing of an individual seizure event, the 7-day window used in this study represents an approximation based on recorded diagnosis dates rather than confirmed seizure occurrence. A key limitation of this operational definition is that EHR diagnosis dates may not accurately reflect the timing of individual seizure events or psychotic symptom onset. Consequently, an unknown number of patients may have experienced psychotic symptoms outside the accepted 7-day post-seizure window and therefore may not have met the clinical definition of PIP. As a result, some cases included in this cohort may not represent true clinically defined PIP. This operational definition identified cases of probable PIP, with the first qualifying psychotic episode serving as the index event. Due to the retrospective nature of this database-based study and reliance on diagnostic coding, it was not possible to confirm the presence of a lucid interval or to characterise specific psychopathological features. Accordingly, cases identified in this study represent an operational definition of probable PIP rather than a definitive clinical diagnosis. The study excluded individuals with any prior diagnosis of psychotic disorders (ICD-10 codes F20–F29) or delirium (F05). Patients who developed delirium following the index event were also excluded to avoid diagnostic overlap with acute confusional states. The control cohort comprised individuals with seizure disorders but no recorded diagnosis of PIP. To minimise baseline differences between groups, propensity-score matching was performed in a 1:1 ratio based on age, sex, race, ethnicity and relevant psychiatric and chronic physical comorbidities. Baseline demographic characteristics are shown in table 1, and matched comorbidities are summarised in table 2. Table 1Baseline demographic characteristics of patients with seizure and PIP and control patients after propensity score matchingStudy cohort, n (%)Control cohort, n (%)Total population (18–65 years)1275 (100)1275 (100)Age at Index (mean±SD)1275 (100) (36.5±12.9)1275 (100) (36.9±13.8)Gender Male689 (54.03)683 (53.56) Female585 (45.8)592 (46.4) Unknown––Ethnicity Not Hispanic or Latino754 (59.1)761 (59.6) Unknown ethnicity419 (32.8)420 (32.9) Hispanic or Latino102 (8.0)94 (7.3)Race White656 (51.4)666 (52.2) Black or African American222 (17.4)211 (16.5) Unknown race268 (21.0)259 (20.3)PIP, postictal psychosis. Baseline demographic characteristics of patients with seizure and PIP and control patients after propensity score matching PIP, postictal psychosis. Table 2Comparison between medical and psychiatric comorbidities between the study cohort (seizure with PIP) and control cohort (seizure only)Study cohort, n (%)Control cohort, n (%)P valueMood/affective disorders (F30–F39)368 (28.8)364 (28.5)0.861Anxiety dissociative, stress related, somatoform and other non-psychotic mental disorders (F40–F48)382 (29.9)389 (30.5)0.762Mental and behavioural disorders due to psychoactive substance use (F10–F19)335 (26.2)346 (27.1)0.622Disorders of adult personality and behaviour (F60–F69)67 (5.2)58 (4.5)0.409Disorders of the thyroid gland (E00–E07)92 (7.2)93 (7.3)0.939Diabetes mellitus (E08–E13)87 (6.8)72 (5.6)0.219Disorders of lipoprotein metabolism and other lipidaemia (E78)119 (9.3)131 (10.2)0.424Other disorders of fluid, electrolyte and acid-base balance (E87)190 (14.9)187 (14.6)0.867Poisoning by adverse effects of and underdosing of drugs, medicaments and biological substances (T36–T50)135 (10.5)126 (9.8)0.556Injuries to the head (S00–S09)240 (18.8)232 (18.1)0.683Sleep disorders (G47)179 (14.0)180 (14.1)0.954PIP, postictal psychosis. Comparison between medical and psychiatric comorbidities between the study cohort (seizure with PIP) and control cohort (seizure only) PIP, postictal psychosis. Variables The primary exposure variable was probable PIP, as defined above. The primary outcome was the diagnosis of any psychotic disorder within 5 years following the index PIP event. Covariates used for propensity-score matching included demographic factors (age, sex, race and ethnicity) and clinical variables such as mood or affective disorders, anxiety disorders, substance use disorders, personality disorders and chronic medical conditions. Data sources and measurement All data were obtained from TriNetX, which compiles standardised ICD-10 coded data from EHRs across multiple institutions. A complete list of ICD-10 codes used to define seizure and psychotic disorders is provided in online supplemental appendix A. Because there is no specific ICD-10 code for PIP, our operational definition may also capture related entities such as antiseizure-medication–induced or acute interictal psychoses. In this database, ‘diagnosis of seizure disorder or epileptic activity’ refers to an ICD-10 code entry indicating an active seizure diagnosis recorded during a clinical encounter, which may represent a habitual seizure or seizure disorder rather than a single event. Therefore, the recorded date of seizure diagnosis may not precisely correspond to the actual timing of an individual seizure. Misclassification of the interval between seizure occurrence and subsequent psychotic symptoms is thus possible, which may affect the accuracy of the 7-day post-seizure window used to define probable PIP. Because TriNetX aggregates data from real-world EHRs across multiple institutions, individual episode documentation and diagnostic consistency cannot be independently verified. Psychiatric diagnoses were also based solely on ICD-10 coding by treating clinicians rather than standardised structured assessments, which may introduce diagnostic variability. Bias Potential sources of bias include selection bias inherent to retrospective database studies and residual confounding from unmeasured clinical variables such as seizure frequency, epilepsy subtype and treatment details. Propensity-score matching was applied to reduce baseline differences between groups, and the multicentre design of TriNetX helps mitigate institutional selection bias. Study size TriNetX automatically determines cohort size based on inclusion/exclusion criteria. After matching, 1275 patients were included in each group (probable PIP and control). The system automatically rounds small cell counts (<10) to maintain patient privacy, consistent with TriNetX data governance standards. Quantitative variables Continuous variables (eg, age) were summarised using means and SD, and categorical variables (eg, sex, comorbidities) were expressed as frequencies and percentages. Group comparisons were conducted using t-tests for continuous variables and χ2 tests for categorical variables. Statistical methods Propensity-score matching was performed in a 1:1 ratio using demographic and clinical covariates. Risk analyses included calculation of risk ratios, risk differences and ORs comparing PIP and control cohorts. Kaplan-Meier survival analyses and log-rank tests were used to evaluate time-to-event outcomes. HRs and 95% CIs were derived using the TriNetX Analytics platform, which applies validated survival modelling algorithms for observational datasets. Statistical significance was defined as p<0.05 (two-tailed). Results Participants After propensity-score matching, 1275 patients were included in each cohort (probable PIP and seizure-only controls). Baseline demographic and clinical characteristics were well balanced across groups (table 1). There were no significant differences in age, sex, race, ethnicity or comorbid medical and psychiatric conditions (table 2). Descriptive data Within the matched cohort, the mean age at index event was approximately 37 years, and both groups demonstrated similar distributions of sex (male≈54%) and race (predominantly White and African American). The follow-up period extended to 5 years after the index event for assessment of psychotic outcomes. Outcome data During the 5-year follow-up, patients with probable PIP demonstrated a substantially higher prevalence and risk of developing psychotic disorders compared with controls. The relative risk for any psychotic disorder was 9.42 (95% CI 5.46 to 16.26; p<0.0001), with an associated HR of 10.78 (95% CI 6.21 to 18.70; p=0.002). Main results The 5-year survival probability for remaining free of any psychotic disorder was 85.79% in the probable PIP group versus 98.21% in controls (log-rank χ²=112.25; p<0.0001) (figure 1). Kaplan-Meier survival analysis comparing the 5-year probability of remaining free of any psychotic disorder between individuals with probable postictal psychosis (PIP) (violet) and seizure-only controls (green). Other analyses Sensitivity analyses based on matched demographic and psychiatric variables demonstrated consistent findings, reinforcing the robustness of the association between probable PIP and subsequent psychotic disorders. No significant differences were observed in baseline comorbidity rates that could account for the elevated psychosis risk. Discussion This retrospective cohort study highlights the significant association between probable PIP and an elevated risk of subsequent chronic psychotic disorders that, in the epilepsy spectrum, would generally be categorised as interictal psychoses. These findings emphasise the need to view PIP not only as an acute phenomenon but as a potential indicator of long-term psychiatric morbidity. The elevated risk of schizophrenia and other psychotic disorders among individuals with seizure disorders is well documented, with contemporary reviews highlighting the broad spectrum of psychiatric comorbidities observed in epilepsy,9 and our findings extend this evidence by showing significantly increased HRs in patients with PIP.10 11 Although the underlying mechanisms remain unclear, Kanner et al proposed shared neurobiological factors linking epilepsy and psychosis, including neurotransmitter dysregulation, genetic susceptibility and structural brain abnormalities. This reinforces the concept that PIP may represent a bridge condition connecting seizure activity to chronic psychiatric disorders.12 Notably, a small but clinically relevant proportion of patients with PIP later develop interictal psychosis, suggesting a continuum from transient postictal states to chronic psychoses.4 13 14 Prior cohort studies have estimated a 7%–10% conversion rate from PIP to chronic or interictal psychotic disorders, supporting the concept that probable PIP may represent a clinically meaningful risk factor for long-term psychotic outcomes.13–15 Although previous studies have suggested that PIP is a transient phenomenon without long-term psychiatric consequences, these findings must be interpreted with caution due to notable methodological limitations. Many of the studies, such as those by Torta and Keller, Kanemoto et al and Clarke et al, have relatively small sample sizes, cross-sectional designs or minimal follow-up, limiting statistical power and generalisability.16 Additionally, inconsistent definitions of PIP across studies may contribute to diagnostic misclassification, particularly in distinguishing PIP from interictal psychosis or primary psychiatric disorders. Several longitudinal studies have also indicated that a minority of PIP cases later progress to interictal or chronic psychoses, reinforcing the need for long-term psychiatric follow-up.13 14 These factors highlight the necessity for more thorough and prospective research to better understand the unique psychiatric risks associated with PIP.15–17 The elevated HRs observed in this study emphasise the need for proactive screening and follow-up of individuals diagnosed with PIP. Early psychiatric intervention may help mitigate the risk of progression to chronic psychotic disorders, consistent with contemporary frameworks that emphasise the high prevalence and clinical impact of psychiatric comorbidities in epilepsy, which contribute substantially to disability and reduced quality of life.18 Furthermore, these findings highlight the need for further investigation into the shared neuropathophysiological mechanisms linking epilepsy and psychosis. Identifying predictive biomarkers or neuroimaging features in this high-risk population could enhance risk stratification and support the development of personalised treatment strategies. In epilepsy, psychoses occurring independently of seizures are broadly termed interictal psychoses, encompassing epileptic psychosis, alternative psychosis due to forced normalisation and primary psychotic disorder. Because diagnostic coding in TriNetX does not distinguish among these entities, our use of the term ‘psychotic disorder’ reflects this aggregate group rather than exclusively primary psychoses. Limitations The retrospective nature of our study introduces several limitations that warrant consideration when interpreting our findings. First, because probable PIP was identified using an ICD-based operational definition within a large EHR database, misclassification is possible. The most important risk is that some included cases may have had psychosis outside the accepted 7-day post-seizure window and therefore may not represent true clinically defined PIP. These findings should therefore be interpreted as hypothesis-generating and require confirmation in prospective studies using standardised PIP criteria. Second, although mood disturbances often accompany PIP and may evolve into affective disorders, our outcomes were restricted to psychotic diagnoses. Third, because the timing of seizure diagnoses in EHRs may not reflect the precise occurrence of seizure events, we could not reliably determine the exact seizure-to-psychosis interval or the duration of subsequent psychotic episodes, which may influence psychiatric outcomes. Fourth, psychiatric diagnoses within TriNetX rely on clinician-entered ICD-10 codes without structured assessments, introducing potential diagnostic variability. Finally, other relevant factors such as epilepsy severity, familial psychiatric history, cognitive function and psychosocial context were unavailable in the registry and may confound observed associations. Conclusion In conclusion, this study contributes to the growing body of evidence suggesting that probable PIP is associated with an increased vulnerability to primary psychotic disorders. Early recognition of PIP prompts vigilant follow-up and intervention strategies for individuals at risk of developing primary psychotic disorders. Future research is warranted to incorporate data on epilepsy subtypes, seizure frequency and age-specific analyses to a refined understanding of the complex interplay between epilepsy and psychiatric outcomes. Supplementary material Footnotes Funding: The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors. Provenance and peer review: Not commissioned; externally peer reviewed. Patient consent for publication: Not applicable. Ethics approval: This study involves human participants, but the Carilion Clinic Institutional Review Board (IRB) determined that this project does not meet the definition of human subjects research; therefore, IRB approval was not required (reference number: IRB-23-2052). Data availability statement All data relevant to the study are included in the article or uploaded as supplementary information. References Associated Data Supplementary Materials Data Availability Statement All data relevant to the study are included in the article or uploaded as supplementary information.