Multidisciplinary Diagnostic and Rechallenge Workflow for Suspected Escitalopram-Associated Neutropenia in Adolescent Obsessive-Compulsive Disorder

Summary

This case report presents a multidisciplinary workflow for evaluating suspected escitalopram-associated neutropenia in an adolescent with obsessive-compulsive disorder, integrating diagnostic assessment, structured complete blood count monitoring, cognitive behavioral therapy with exposure and response prevention, and monitored escitalopram rechallenge to support clinical decision-making under hematologic uncertainty.

Abstract

Selective serotonin reuptake inhibitors are evidence-based pharmacologic treatments for obsessive-compulsive disorder (OCD) in adolescents; however, hematologic adverse effects such as leukopenia and neutropenia are rare and incompletely characterized. Management becomes particularly challenging when an effective medication is suspected to contribute to clinically significant hematologic abnormalities. This case report describes a multidisciplinary diagnostic and monitoring workflow for evaluating suspected escitalopram-associated neutropenia in a 14-year-old female with moderate-to-severe OCD and substantial functional impairment. The patient was initially treated with escitalopram, titrated from 10 mg/day to 20 mg/day, resulting in a 39–55% reduction in symptom severity as assessed by retrospectively reconstructed Children’s Yale-Brown Obsessive Compulsive Scale (CY-BOCS) scores. Persistent leukopenia and neutropenia identified during routine complete blood count (CBC) monitoring prompted escitalopram discontinuation and a structured multidisciplinary evaluation involving hematology, rheumatology, infectious diseases, allergy and immunology, endocrinology, neurology, and cardiology. Extensive investigations, including laboratory testing, peripheral blood smear examination, brain magnetic resonance imaging, and pelvic ultrasonography, did not identify a definitive alternative etiology. Following medication discontinuation, white blood cell counts did not normalize and OCD symptoms relapsed despite cognitive behavioral therapy with exposure and response prevention, delivered in nine weekly sessions by a clinical psychologist. After multidisciplinary risk-benefit assessment and shared decision-making with the patient and legal guardian, escitalopram rechallenge was initiated using gradual dose escalation from 5 mg/day to 20 mg/day over three weeks, with weekly CBC monitoring, predefined stopping criteria, and infection surveillance. The final CY-BOCS score improved to 11/40 (65% reduction from baseline), while hematologic parameters remained stable. The procedural objective of this report is to present a reproducible multidisciplinary framework for diagnostic evaluation, risk stratification, monitoring, and cautious pharmacologic rechallenge in adolescents with suspected SSRI-associated hematologic abnormalities.

Introduction

Obsessive-compulsive disorder (OCD) is defined by the presence of obsessions (recurrent and intrusive thoughts, images, or impulses) and/or compulsions (repeated behavioral or mental acts) that disrupt daily life and generate significant distress¹. OCD ranks among the 10 most disabling medical and psychiatric conditions, with a lifetime prevalence of 2%–3% in the general population^2,3,4,5. For several decades, research has focused on identifying effective therapies for OCD, including psychotherapy, pharmacological interventions, or combined approaches. Evidence from clinical trials supports cognitive behavioral therapy (CBT) and selective serotonin reuptake inhibitors (SSRIs) as standard treatments for OCD because of their efficacy and safety profiles⁶.

Leukopenia is defined as a white blood cell count below 4,000 cells/µL, impairing host defense against infections⁷. Neutropenia, a common subtype, refers to an absolute neutrophil count (ANC) of <1,500 cells/µL and may be transient or chronic, with etiologies ranging from genetic and autoimmune disorders to infections and drug exposure⁸. Despite being rare, neutropenia, among other hematological disorders, has been reported as a potential adverse effect of certain SSRIs, including sertraline and paroxetine^9,10. Such adverse effects pose a dual risk: the clinical consequences of the hematologic abnormality itself and compromised psychiatric disease control resulting from treatment interruption.

We report a challenging case of suspected escitalopram-associated leukopenia and neutropenia in an adolescent with OCD, resulting in substantial diagnostic and therapeutic uncertainty. The objective of this report is to present a reproducible multidisciplinary management framework for suspected SSRI-associated hematologic abnormalities in pediatric OCD, integrating structured diagnostic evaluation, risk stratification, shared decision-making, and cautious pharmacologic rechallenge under intensive hematologic monitoring. Such an approach may provide a clinically viable alternative to permanent medication discontinuation in carefully selected patients when psychiatric deterioration occurs despite unresolved diagnostic uncertainty.

Case Presentation
This case report was prepared in accordance with the CARE (CAse REport) reporting guidelines, encompassing patient demographic information, clinical timeline, diagnostic reasoning, therapeutic interventions, follow-up outcomes, patient and family perspectives, and informed consent documentation.

A 14-year-old girl, followed in our department for OCD diagnosed at 13 years of age (June 2024), developed incidental leukopenia during treatment with escitalopram. She is the eldest of four siblings, lives with her family in a rural village approximately 200 km south of Taif City, Saudi Arabia, and is in her second year of intermediate school. Her developmental history was unremarkable, with no prior psychiatric or significant medical diagnoses.

The patient underwent comprehensive psychiatric evaluation within a specialized child and adolescent psychiatry setting. Baseline assessment included clinical interviews with both the patient and her mother, longitudinal symptom and developmental history, psychosocial and family psychiatric history, functional impairment assessment, and mental status examination. Diagnosis was established and supervised by a consultant child and adolescent psychiatrist in accordance with DSM-5 criteria for OCD, with particular attention to the nature, frequency, distress level, insight, and functional impact of obsessive-compulsive symptoms. Differential diagnoses were systematically evaluated and included psychotic disorders, autism spectrum disorder, tic-related OCD, mood disorders, and primary anxiety disorders.

The OCD onset was insidious. Symptoms were predominantly contamination- and religious-themed, centering on intrusive, ego-dystonic fears of impurity, incomplete cleansing, and persistent doubt regarding the correctness and adequacy of her religious practices. These obsessive cognitions were experienced as repetitive, unwanted, and distressing, generating marked anxiety and a pervasive sense of incompleteness that the patient felt compelled to resolve. In direct response to these obsessions, she developed a pattern of compulsive behaviors aimed at temporarily neutralizing anxiety and achieving a subjective sense of ritual completeness. These included repetitive handwashing, prolonged and repeated ablution (wudu) rituals, repeated performance of prayer, excessive reassurance seeking, compulsive checking of ritual completion, and avoidance of perceived contamination triggers. Ablution rituals were performed multiple times in succession, driven by persistent intrusive doubts about cleanliness or ritual correctness despite objective completion. Prayer was similarly prolonged or repeated because of obsessive uncertainty that it had been performed inadequately. Handwashing became excessive in both frequency and duration, ultimately resulting in visible skin irritation and ulcerative lesions of the hands. The compulsive behaviors afforded only transient relief, after which obsessive doubts rapidly re-emerged, perpetuating a self-reinforcing obsession-compulsion cycle. Symptom burden progressively escalated, interfering with academic functioning, daily activities, and family interactions. The patient concealed her symptoms for several months before her mother observed the behaviors and sought psychiatric evaluation.

Retrospective reconstruction of symptom severity was performed using the Children's Yale-Brown Obsessive Compulsive Scale (CY-BOCS)¹¹, as prospective scoring had not been systematically recorded across all clinical visits. Scores were estimated from detailed psychiatric documentation, serial clinical assessments, functional impairment records, and multidisciplinary follow-up notes. At the time of initial psychiatric assessment, the clinical picture was consistent with severe OCD, with an estimated CY-BOCS total score of 31/40 (Obsessions: 16/20; Compulsions: 15/20), reflecting the marked distress, functional impairment, and physical sequelae described above.

Management was supervised by a psychiatrist at Eradah Mental Health Hospital. Pharmacological treatment was initiated with escitalopram 10 mg/day and titrated to 20 mg/day after five days. Serial psychiatric follow-up assessments were scheduled weekly during the acute treatment period and monthly thereafter. Initial adverse effects occurred during the first two weeks of treatment and included nausea, headache, vomiting, and flu-like symptoms, all of which resolved with supportive management.

At the first follow-up visit (month 1), the patient reported marked regression of compulsive washing duration, improved ability to complete daily activities, improved family interaction, and decreased distress during prayer. The estimated CY-BOCS total score decreased from 31/40 at baseline to 19/40 (Obsessions: 10/20; Compulsions: 9/20), representing approximately 39% reduction in symptom severity. Escitalopram 20 mg/day was continued.

At the second follow-up visit (month 2), the patient reported that escitalopram had been discontinued by another psychiatrist approximately three weeks after the first follow-up under unspecified circumstances and that fluoxetine 20 mg/day had been substituted, resulting in symptom relapse within one week. Escitalopram 20 mg/day was accordingly reintroduced at our clinic.

Over the subsequent month (month 3), further improvement was observed, with shorter ritual duration, reduced contamination distress, improved school participation, and partial restoration of daily functioning. The estimated CY-BOCS total score decreased to 14/40 (Obsessions: 8/20; Compulsions: 6/20), corresponding to approximately 55% reduction from baseline severity.

During the month-3 visit, the patient presented with laboratory results from a private facility revealing leukopenia (white blood cell [WBC] count 2.73 × 10³/µL), neutropenia (neutrophil fraction 27.4%), and relative lymphocytosis (61.2%). A confirmatory complete blood count (CBC) performed at our clinic demonstrated persistent leukopenia (WBC 2.78 × 10³/µL, neutrophils 23.4%, lymphocytes 65.1%). No signs of infection were identified on clinical assessment or reported by the patient at either evaluation.

Diagnosis, Assessment, and Plan
In the absence of a baseline CBC and given the persistence of leukopenia across two consecutive measurements, drug-induced leukopenia secondary to escitalopram was suspected. The treating psychiatrist discontinued escitalopram as a precautionary measure and initiated CBT with exposure and response prevention (CBT/ERP).

Subsequent laboratory monitoring revealed fluctuating leukocyte counts, with WBC transiently rising to 4.3 × 10³/µL (laboratory reference range: 3.5–10.0 × 10³/µL) before declining again to 3.2 × 10³/µL with a neutrophil fraction of 27.5%. Over the following month, the patient remained off pharmacotherapy and experienced a gradual relapse of OCD symptoms, accompanied by emerging depressive features and increasing concern regarding her perceived psychological vulnerability.

Physical examination revealed a weight of 35.9 kg and height of 146 cm, with no dysmorphic features or clinical signs of chronic illness. There were no findings suggestive of eczema, asthma, or autoimmune disease. Notably, the patient had not yet attained menarche and showed no secondary sexual characteristics, prompting referral to pediatric endocrinology. Family history was non-contributory. Psychologically, the patient displayed interpersonal sensitivity, low frustration tolerance, and somatic preoccupations while maintaining adequate academic performance.

Given the diagnostic complexity, multidisciplinary evaluation was undertaken to investigate secondary causes of leukopenia. Consultations included hematology, rheumatology, infectious diseases, allergy and immunology, endocrinology, neurology, and cardiology.

Differential diagnoses included escitalopram-associated neutropenia, benign ethnic neutropenia (BEN)/Duffy-null associated neutrophil count, chronic idiopathic neutropenia, autoimmune neutropenia, infectious and inflammatory etiologies, nutritional deficiencies, endocrine abnormalities, and constitutional hematologic variants. Psychiatric management and overall care coordination were supervised by the treating consultant child and adolescent psychiatrist.

Laboratory investigations included CBC with differential, inflammatory markers, autoimmune panel (RF, C3, and C4 within normal limits), endocrine profile (TSH, prolactin, estradiol, progesterone, PTH, and vitamin D), ferritin, folate, hereditary hemophagocytic lymphohistiocytosis (HLH) panel, protein C and S (protein S: 47%; reference range: 54–82%), and infectious screening. ESR was markedly elevated at 112 mm/hour. Anti-streptolysin O (ASO) titer was elevated at 566.9 IU/mL, and Epstein–Barr virus nuclear antigen (EBNA) IgG was positive, indicating previous exposure. Peripheral blood smear demonstrated mild leukopenia and moderate neutropenia. Brain magnetic resonance imaging (MRI) and pelvic ultrasonography were unremarkable. No definitive alternative etiology for the hematologic findings was identified. Detailed investigations are presented in the supplementary file.

In light of these findings, additional considerations included post-infectious and neuroimmune conditions, particularly Pediatric Acute-onset Neuropsychiatric Syndrome (PANS) and Pediatric Autoimmune Neuropsychiatric Disorders Associated with Streptococcal Infections (PANDAS); however, the patient's insidious symptom onset and absence of a documented temporal association with acute streptococcal infection made these diagnoses less likely.

Although no definitive alternative etiology for the hematologic findings was identified following the documented multidisciplinary evaluation, direct attribution to escitalopram remained uncertain because of the absence of a baseline CBC, persistence of leukopenia after discontinuation, and fluctuating leukocyte counts independent of medication exposure. Escitalopram discontinuation was therefore maintained as a precautionary measure pending etiologic clarification. The patient was discharged with regular hematologic monitoring and continuation of CBT.

Protocol

The case report was prepared in accordance with institutional policies and standard ethical guidelines. Under institutional policy, anonymized case studies, including case reports and case series that do not involve experimental interventions, are exempt from ethics committee approval. The principles outlined in the Declaration of Helsinki were followed throughout the clinical assessment and documentation process. Written informed consent was obtained from the patient’s legal guardian for publication of this case report and accompanying clinical details. Efforts were made to ensure complete anonymization of the patient's identity.

1. Escitalopram initiation and initial treatment phase

1. Escitalopram was initiated following a comprehensive psychiatric evaluation confirming moderate-to-severe OCD with significant functional impairment. The medication was administered orally once daily in the evening after food to optimize gastrointestinal tolerability and adherence.
2. Eligibility for treatment was established through a comprehensive psychiatric evaluation conducted by a consultant child and adolescent psychiatrist. Assessment included clinical interviews with the patient and caregiver, developmental and psychosocial history, functional impairment assessment, mental status examination, and Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-5)-based diagnostic confirmation of OCD. Differential diagnoses were systematically excluded. Symptom severity was documented using the Children's Yale-Brown Obsessive Compulsive Scale (CY-BOCS), with retrospective reconstruction from detailed clinical records when prospective scores were unavailable. Pharmacologic treatment was considered in the context of severe OCD associated with substantial distress, functional impairment, and inadequate symptom control through supportive measures alone.
3. Escitalopram was initiated at a dose of 10 mg/day.
4. The medication was administered orally once daily in the evening after food to optimize gastrointestinal tolerability and treatment adherence.
5. The dose was increased to 20 mg/day after five days because clinically significant symptoms persisted and no severe adverse effects were reported. Detailed information regarding the specific tolerability assessments and decision-making process used during this initial dose escalation was not available in the clinical records; however, no serious adverse events were documented during this period.
6. Dose-escalation decisions were based on routine clinical assessment rather than a formalized protocol. Tolerability and early treatment response were reviewed during follow-up visits, including evaluation of adverse effects and overall clinical status. Dose escalation proceeded when treatment was considered clinically acceptable and no significant safety concerns were identified.
7. Clinical follow-up assessments were performed during the early treatment period to evaluate symptom response, tolerability, and adverse effects.
8. During the initial treatment phase, psychiatric follow-up assessments were conducted weekly during the acute phase and monthly thereafter. Evaluations focused on OCD symptom severity, functional status, treatment response, and adverse effects. Clinical progress was documented using serial CY-BOCS assessments and patient-reported changes in daily functioning and distress.
9. Transient non-severe adverse effects, including nausea, headache, vomiting, and flu-like symptoms, occurred during the initial treatment period and resolved without dose modification or additional pharmacologic intervention.
10. Adverse effects were assessed during routine follow-up visits through clinical interviews with the patient and caregiver and review of treatment tolerability. Reported symptoms were documented in the medical record and followed clinically over time. No formal adverse-effect rating scale or structured assessment protocol was used.

2. Escitalopram discontinuation and CBT initiation

1. An abnormal CBC obtained at a private facility was confirmed by a subsequent CBC at our clinic, prompting precautionary discontinuation of escitalopram, multidisciplinary evaluation, and serial laboratory monitoring pending etiologic clarification.
2. Confirmation was based on two consecutive abnormal CBC measurements obtained approximately one month apart. The initial CBC showed a WBC count of 2.73 × 10³/µL with neutrophils at 27.4%, whereas the repeat CBC demonstrated a WBC count of 2.78 × 10³/µL with neutrophils at 23.4%.
3. Escitalopram was temporarily discontinued after repeat CBCs confirmed persistent leukopenia and neutropenia. Given the absence of a baseline CBC and uncertainty regarding causality, discontinuation was undertaken as a precautionary measure while multidisciplinary evaluation was pursued.
4. No predefined hematologic thresholds were used. Discontinuation was based on persistent leukopenia and neutropenia on sequential CBCs, inability to exclude a drug-related cause, and clinical judgment favoring diagnostic clarification before further pharmacologic treatment.
5. CBT with ERP was initiated following medication discontinuation.
6. Psychotherapy was delivered by a licensed clinical psychologist (PhD) with specialized training in CBT and ERP for children and adolescents under the supervision of the treating consultant child and adolescent psychiatrist.
7. The intervention was initiated in January 2025 and consisted of nine consecutive weekly outpatient sessions lasting 50–60 min each.
8. Family involvement was incorporated throughout treatment and included psychoeducation, participation in treatment planning, review of clinical progress, and support for implementing exposure and response-prevention strategies at home. Family participation was integrated periodically throughout the nine weekly CBT/ERP sessions rather than delivered as separate dedicated sessions. The frequency and duration of participation varied according to clinical needs and treatment progress.
9. Psychoeducation regarding OCD mechanisms and anxiety cycles was provided during therapy sessions.
10. Obsessive thoughts and compulsive behaviors were identified and reviewed during treatment.
11. Cognitive restructuring techniques targeting maladaptive beliefs related to contamination fears and religious rituals were implemented.
12. Exposure exercises were introduced in a graded manner, beginning with less distressing contamination-related situations and progressing to more challenging exposures over time. Advancement through the hierarchy was guided by the patient's response to previous exercises, reduction in distress, and overall readiness for further exposure work.
13. No formal exposure hierarchy, readiness scale, or predefined progression criteria were used. Exposure exercises were individualized according to symptom targets, clinical response, and therapist judgment throughout treatment.
14. Response prevention targeted compulsive handwashing, repeated ablution, reassurance seeking, and ritual repetition. Progress was monitored through patient feedback and review of assigned practice exercises.
15. No formal response-prevention adherence measure was used. Progress was evaluated through clinical discussion, patient self-report, therapist observation, and review of assigned practice exercises during follow-up sessions.
16. Homework assignments were provided after each weekly session and included exposure exercises, response-prevention practice, and symptom-monitoring tasks. Completion and difficulties encountered were reviewed at subsequent sessions through patient report and therapist feedback, and assignments were adjusted according to progress and treatment goals.
17. Engagement and adherence were assessed through session attendance, completion of homework assignments, active participation in therapy sessions, and therapist evaluation of treatment involvement and cooperation.
18. No formal assessment procedures or validated instruments were used to measure engagement or adherence.
19. Clinically significant OCD symptoms persisted throughout the CBT-only phase, including contamination obsessions, compulsive rituals, emotional distress, and moderate-to-severe symptom severity. CY-BOCS scores were reassessed at approximately monthly intervals using retrospective reconstruction from clinical records, psychotherapy documentation, and psychiatric follow-up assessments.
20. CBT improved coping skills and symptom awareness; however, psychotherapy alone was insufficient to achieve adequate symptom control or functional recovery. Additional pharmacologic treatment was considered because moderate-to-severe OCD symptoms, functional impairment, emotional distress, and clinically significant symptom burden persisted despite completion of structured CBT/ERP.
21. No predefined response threshold was used. Additional pharmacologic treatment was considered on the basis of persistent clinical symptoms, ongoing functional impairment, and inadequate response to psychotherapy.

3. Reassessment following escitalopram discontinuation

1. Clinical follow-up was continued during the period of CBT without concurrent pharmacotherapy.
2. Progressive worsening of OCD symptoms was observed during the weeks following escitalopram discontinuation.
3. A comprehensive psychiatric reassessment was conducted three months after discontinuation (January 2025) by the treating consultant child and adolescent psychiatrist during routine follow-up. The evaluation included a clinical interview with the patient and her mother, mental status examination, assessment of functional impairment, review of symptom progression, treatment response, available laboratory findings, and reassessment of OCD symptom severity using estimated CY-BOCS scores derived from contemporaneous clinical records.
4. The patient presented with depressed mood, moderate OCD symptomatology, and intermittent anxiety.
5. Symptom severity was reassessed using the CY-BOCS.
6. The estimated CY-BOCS total score increased to 26/40 (Obsessions: 14/20; Compulsions: 12/20), consistent with moderate-to-severe symptom recurrence.
7. CY-BOCS scores were retrospectively reconstructed from contemporaneous clinical documentation and follow-up assessments.
8. The patient reported frustration regarding symptom relapse and concerns about the long-term implications of persistent leukopenia.
9. Additional symptoms included headache, diffuse musculoskeletal pain, lethargy, and fatigue.
10. No sleep disturbances or appetite disturbances were reported.
11. Subsequent investigations comprised hematologic, autoimmune, infectious, endocrine, neurologic, and immunologic evaluations, together with targeted imaging studies, to identify potential causes of the persistent hematologic abnormalities and associated clinical symptoms.

4. Escitalopram rechallenge

1. A multidisciplinary reassessment was completed following progressive psychiatric deterioration and inadequate response to CBT alone.
2. By late April 2025, OCD symptom severity had increased, with an estimated CY-BOCS score of 28/40 (Obsessions: 15/20; Compulsions: 13/20).
3. Concurrent hematologic reassessment demonstrated a WBC count of 2.7 × 10³/µL and a neutrophil fraction of 27%.
4. The corresponding ANC was approximately 729 cells/µL (0.73 × 10³/µL), consistent with moderate neutropenia and considered during multidisciplinary risk-benefit discussions regarding escitalopram rechallenge.
5. Inadequate symptom control during the CBT-only phase was documented, and the patient and her family formally requested resumption of pharmacotherapy.
6. Shared decision-making discussions were conducted with the patient and legal guardian.
7. The discussions addressed diagnostic uncertainty, potential risks and benefits, treatment alternatives, monitoring requirements, and predefined stopping criteria. Patient assent and guardian informed consent were obtained before rechallenge initiation.
8. The decision to rechallenge escitalopram was made collaboratively by the treating consultant child and adolescent psychiatrist and the hematology team, with input from rheumatology, infectious diseases, allergy and immunology, endocrinology, neurology, and cardiology. Clinical findings, serial laboratory results, potential risks and benefits, and worsening psychiatric symptoms were reviewed before proceeding.
9. Escitalopram was reintroduced at a dose of 5 mg/day.
10. The dose was increased gradually to 20 mg/day over three weeks.
11. Dose titration followed a structured schedule of 5, 10, 15, and 20 mg/day.
12. Dose increases occurred at one-week intervals during rechallenge.
13. Dose escalation proceeded only when no decline in WBC count or ANC was observed.
14. Dose escalation was undertaken only in the absence of febrile illness, recurrent infections, oral ulceration, or other signs of clinical deterioration. Escalation also required stable serial CBC findings following multidisciplinary review.
15. No predefined quantitative WBC or ANC thresholds were required for dose escalation. Escalation decisions were based on hematologic stability across serial CBC measurements, with no evidence of progressive leukopenia, worsening neutropenia, or emerging cytopenias.
16. Systemic deterioration referred to clinically significant adverse events, including febrile illness, recurrent or serious infections, oral or mucosal ulceration, pancytopenia, abnormal peripheral smear findings, or other clinical features prompting urgent reassessment.

5. Hematologic and clinical monitoring

1. A structured hematologic monitoring plan was implemented throughout the escitalopram rechallenge period.
2. A CBC with differential was obtained before rechallenge initiation.
3. CBC monitoring was performed weekly during dose escalation and early stabilization.
4. Following stabilization, CBC monitoring continued at individualized intervals based on sustained hematologic stability across serial assessments, absence of infectious complications, lack of progressive cytopenia, and overall clinical stability following completion of dose escalation. No fixed stabilization period was predefined, and monitoring frequency was gradually reduced according to multidisciplinary recommendations.
5. Monitoring focused on WBC count trends, ANC trends, emergence of additional cytopenias, abnormal peripheral smear findings, and other laboratory indicators of hematologic deterioration.
6. Hematologic deterioration was defined by declining WBC or ANC trends, emergence of additional cytopenias, abnormal peripheral smear findings, progression toward severe neutropenia, or other laboratory evidence suggesting worsening hematologic status.
7. Psychiatric monitoring was conducted by the treating consultant child and adolescent psychiatrist during routine follow-up visits and included assessment of OCD symptom severity, emotional status, functional impairment, treatment response, and retrospective CY-BOCS-based symptom tracking.
8. Infection surveillance was incorporated into routine follow-up and included screening for fever, oral or mucosal ulceration, sore throat, recurrent infections, and other symptoms suggestive of systemic infection. Findings prompted further evaluation when clinically indicated.
9. Medication tolerability was assessed through clinical interviews with the patient and caregiver and review of adverse effects during follow-up visits. Findings were documented in the medical record, and no formal tolerability scale or grading system was used.
10. Functional outcomes were evaluated clinically through assessment of school participation and performance, daily activities, family interactions, and OCD-related interference. Improvement was determined by reduced functional impairment and restoration of routine activities over the course of treatment.

6. Stopping thresholds and emergency reassessment criteria

1. Predefined safety criteria were established prior to escitalopram rechallenge and were reviewed throughout follow-up.
2. Urgent clinical reassessment was undertaken when a notable decline in ANC was observed. No predefined quantitative ANC decline threshold was used, and decisions were based on serial ANC trends, overall hematologic stability, accompanying clinical findings, and multidisciplinary assessment of potential risk.
3. Urgent reassessment was undertaken if laboratory findings suggested progression to severe neutropenia (ANC < 500 cells/µL) or agranulocytosis (ANC < 100 cells/µL), consistent with standard hematologic definitions used during monitoring.
4. Reassessment was triggered by fever (≥38.0°C) or other signs of possible infection and included clinical evaluation, repeat CBC testing, infection surveillance, and multidisciplinary review when clinically indicated.
5. Reassessment was performed in the presence of recurrent or clinically significant infections, defined as repeated infectious episodes, infections requiring medical treatment, or infections accompanied by worsening hematologic parameters.
6. Reassessment was performed when oral or other mucosal ulcerations were identified during follow-up.
7. Reassessment was performed when serial CBC testing demonstrated a notable decline in WBC count compared with previous measurements. No predefined quantitative WBC threshold was used, and deterioration was assessed through serial trends, particularly when accompanied by worsening neutropenia, additional cytopenias, or other concerning clinical findings.
8. Reassessment was performed if additional cytopenias emerged during monitoring. Pancytopenia was defined as concurrent reductions in white blood cell count, hemoglobin level, and platelet count below their respective laboratory reference ranges.
9. Reassessment was performed when peripheral smear review identified clinically significant abnormalities, including atypical cells, abnormal cell morphology, or findings suggestive of an underlying hematologic disorder requiring further investigation.
10. Reassessment was performed in the presence of significant systemic instability, defined as fever, hemodynamic compromise, hospitalization, serious infection, or other clinically significant deterioration warranting urgent reassessment and multidisciplinary review.

7. Patient and family counseling

1. Structured counseling was provided to the patient and family before and during escitalopram rechallenge.
2. Counseling included guidance on hand hygiene, prompt reporting of fever or signs of infection, adherence to follow-up appointments, and the importance of ongoing clinical and laboratory monitoring. Counseling also focused on recognition of warning signs requiring urgent medical reassessment, including fever, oral ulceration, mucosal symptoms, recurrent infections, and systemic deterioration.
3. Counseling discussions were documented in the medical record, and understanding was confirmed through patient and caregiver feedback during follow-up visits.
4. Infection surveillance was incorporated into all follow-up visits.

8. Treatment continuation and timeline documentation

1. Hematologic parameters were reviewed throughout follow-up after escitalopram rechallenge.
2. No progressive cytopenia or infectious complications were identified during the monitoring period.
3. Following achievement of the target escitalopram dose, the patient remained under multidisciplinary follow-up for approximately 2–3 months, during which serial psychiatric assessments, CBC monitoring, infection surveillance, and evaluation of treatment response were continued.
4. Clinical and laboratory events were documented chronologically throughout treatment and follow-up.
5. The detailed clinical timeline was summarized in Table 1.

Timepoint	Clinical Event	Pharmacotherapy	Reconstituted CY-BOCS Score	Hematology (WBC × 103/µL)	Notes
Baseline (June 2024)	Initial psychiatric assessment; DSM-5 OCD diagnosis confirmed	Escitalopram 10 mg/day initiated	31/40 (severe)	No baseline CBC available	Referral from pediatric hospital
Day 5	Tolerability review	Escitalopram titrated to 20 mg/day	Not assessed	Not available	Transient nausea, headache, vomiting, and flu-like symptoms resolved spontaneously
Month 1 (July 2024)	First follow-up visit	Escitalopram 20 mg/day continued	19/40 (−39%)	Not available	Functional improvement documented
Approximately Week 7 (August 2024)	Escitalopram discontinued by another psychiatrist; fluoxetine substituted	Fluoxetine 20 mg/day	Not assessed	Not available	Symptom relapse within one week
Month 2 (August 2024)	Second follow-up at our clinic; fluoxetine failure noted	Escitalopram 20 mg/day reintroduced	Not assessed	Not available	Shared decision to reintroduce escitalopram
Month 3 (September 2024)	Third follow-up; further improvement	Escitalopram 20 mg/day	14/40 (−55%)	2.73 (neutrophils 27.4%)	Leukopenia detected on private laboratory CBC; confirmatory CBC requested
Month 4 (October 2024)	Confirmatory CBC; escitalopram discontinued	Escitalopram discontinued; CBT/ERP initiated	Not assessed	2.78 (neutrophils 23.4%)	No signs of infection; suspected escitalopram-associated leukopenia
Months 4–5 (October–November 2024)	Serial CBC monitoring	Off pharmacotherapy	Not assessed	4.3 → 3.2 (neutrophils 27.5%)	Fluctuating WBC counts; OCD relapse with depressive features
Month 6 (January 2025)	Reassessment; multidisciplinary evaluation initiated	Off pharmacotherapy; CBT/ERP ongoing	26/40 (moderate-to-severe)	Not available	Nine weekly CBT/ERP sessions completed; extensive diagnostic workup performed
January–March 2025	Multidisciplinary consultations (hematology, rheumatology, infectious diseases, allergy and immunology, endocrinology, neurology, and cardiology)	Off pharmacotherapy; CBT/ERP ongoing	Not assessed	Not available	No definitive alternative etiology identified; benign ethnic neutropenia considered; bone marrow biopsy deferred
April 2025 (pre-rechallenge)	Pre-rechallenge assessment and shared decision-making	None	28/40 (severe)	2.7 (neutrophils 27%)	Formal rechallenge request by patient and family
Late April 2025 (Week 1)	Escitalopram rechallenge initiated	Escitalopram 5 mg/day	Not assessed	Weekly CBC monitoring	Structured hematologic monitoring plan implemented
Week 2	Dose escalation	Escitalopram 10 mg/day	Not assessed	Weekly CBC monitoring	No ANC decline and no infectious complications
Weeks 3–4	Further dose escalation	Escitalopram 15–20 mg/day	Not assessed	Weekly CBC monitoring	Tolerability confirmed; escalation criteria met
May–July 2025	Final follow-up	Escitalopram 20 mg/day continued	11/40 (−65%)	2.79 (stable)	No agranulocytosis, febrile neutropenia, or infectious complications; CBT ongoing

Table 1: Chronological timeline of clinical events, investigations, and management decisions. Chronological summary of psychiatric assessments, pharmacologic interventions, hematologic findings, multidisciplinary investigations, cognitive behavioral therapy with exposure and response prevention (CBT/ERP), escitalopram discontinuation, rechallenge procedures, and follow-up outcomes in an adolescent with obsessive-compulsive disorder (OCD) and suspected escitalopram-associated leukopenia/neutropenia. Symptom severity was assessed using the Children's Yale-Brown Obsessive Compulsive Scale (CY-BOCS). Hematologic monitoring included complete blood count (CBC) measurements and white blood cell (WBC) trends throughout treatment and follow-up. Abbreviations: ANC, absolute neutrophil count; CBC, complete blood count; CBT, cognitive behavioral therapy; CY-BOCS, Children’s Yale-Brown Obsessive Compulsive Scale; ERP, exposure and response prevention; OCD, obsessive-compulsive disorder; WBC, white blood cell.

Results

Obsessive-compulsive symptom severity, assessed via retrospectively reconstructed CY-BOCS scores, followed a trajectory consistent with the clinical course across all treatment phases. At baseline, the estimated CY-BOCS total score was 31/40, consistent with severe OCD. Following escitalopram titration to 20 mg/day, scores decreased to 19/40 at month 1 (−39%) and to 14/40 at month 3 (−55%), accompanied by functional improvement in daily activities, academic participation, and family interaction. Following escitalopram discontinuation, progressive symptom relapse was documented, with CY-BOCS scores increasing to 26/40 at reassessment (January 2025) and 28/40 immediately prior to rechallenge (April 2025), despite structured CBT/ERP.

Hematologically, leukopenia was first detected at month 3 (WBC 2.73 × 10³/µL, neutrophil fraction 27.4%; ANC approximately 748 cells/µL) and confirmed on repeat testing (WBC 2.78 × 10³/µL, neutrophils 23.4%; ANC approximately 651 cells/µL). The absence of a pre-treatment baseline CBC precluded determination of whether leukopenia predated escitalopram exposure. Following discontinuation, WBC transiently rose to 4.3 × 10³/µL before declining again to 3.2 × 10³/µL (neutrophils 27.5%; ANC approximately 880 cells/µL), with leukopenia persisting beyond two months after withdrawal. At the pre-rechallenge assessment, WBC was 2.7 × 10³/µL with a neutrophil fraction of 27% (ANC approximately 729 cells/µL). Concurrently, extensive multidisciplinary evaluation failed to identify a definitive alternative etiology.

These findings resulted in continued diagnostic uncertainty regarding the relationship between escitalopram exposure and the observed hematologic abnormalities. Given progressive psychiatric deterioration despite CBT/ERP, the patient and family requested resumption of pharmacotherapy. Following multidisciplinary review and shared decision-making, escitalopram rechallenge was initiated under structured hematologic monitoring.

Throughout rechallenge and follow-up, hematologic parameters remained stable, with a final documented WBC count of 2.79 × 10³/µL. No progression to severe neutropenia, agranulocytosis, pancytopenia, febrile neutropenia, recurrent infections, emergency reassessment events, or hospitalization occurred during follow-up, and no predefined stopping criteria were reached. Simultaneously, the final CY-BOCS score improved to 11/40 (−65% from baseline), accompanied by improved family functioning, reduced compulsive rituals, and better participation in CBT/ERP. The overall clinical course, laboratory findings, multidisciplinary evaluation, treatment interventions, and follow-up outcomes are summarized in Figure 1 and Table 1.

Figure 1: Multidisciplinary diagnostic evaluation, monitoring workflow, and escitalopram rechallenge strategy in an adolescent with suspected escitalopram-associated neutropenia and obsessive-compulsive disorder. Timeline of the patient's clinical course, including initial escitalopram treatment, detection of leukopenia and neutropenia on complete blood count (CBC) testing, medication discontinuation, cognitive behavioral therapy (CBT) with exposure and response prevention (ERP), multidisciplinary diagnostic evaluation, and escitalopram rechallenge under hematologic monitoring. The figure summarizes changes in Children's Yale-Brown Obsessive Compulsive Scale (CY-BOCS) scores, white blood cell (WBC) counts, and clinical outcomes throughout follow-up. Abbreviations: ADR, adverse drug reaction; ANC, absolute neutrophil count; ASO, anti-streptolysin O; BEN, benign ethnic neutropenia; EBNA, Epstein–Barr virus nuclear antigen; ESR, erythrocyte sedimentation rate; HLH, hemophagocytic lymphohistiocytosis; MRI, magnetic resonance imaging. Please click here to view a larger version of this figure.

Discussion

This case presents a multidisciplinary framework for managing a high-stakes therapeutic decision under irreducible diagnostic uncertainty: whether to maintain precautionary pharmacotherapy discontinuation at the cost of progressive psychiatric deterioration or to attempt cautious rechallenge in the absence of a confirmed etiological diagnosis. Thus, beyond presenting as suspected drug-induced neutropenia, this case illustrates the diagnostic complexity, ethical reasoning, and procedural rigor required when a hematologic association cannot be confirmed or refuted, yet the psychiatric burden of inaction is clinically unacceptable. The workflow presented, encompassing structured multidisciplinary evaluation, differential diagnosis, shared decision-making, and monitored rechallenge, is intended to be reproducible across similar clinical scenarios and to complement existing pharmacovigilance frameworks that default to permanent discontinuation.

Hematologic toxicity with SSRIs is exceptionally rare¹². In our patient, leukopenia was detected temporally during escitalopram treatment; however, the evidence for direct drug causality is limited and inconsistent. The absence of a pre-treatment baseline CBC represents a fundamental limitation, as it precludes determination of whether leukopenia predated escitalopram exposure. Furthermore, leukopenia persisted for more than two months following discontinuation, a pattern inconsistent with classic drug-induced neutropenia, which typically resolves within days to weeks of withdrawal¹³. Fluctuating leukocyte counts independent of medication exposure further weaken a direct causal link. Accordingly, the hematologic findings in this case are best characterized as a suspected association rather than a confirmed adverse drug reaction.

The differential diagnosis of neutropenia in this clinical context was broad, warranting systematic evaluation. Given the patient’s Saudi Arabian background and the pattern of stable, longitudinal neutropenia without infectious complications, benign ethnic neutropenia (BEN), also described as Duffy-null associated neutrophil count, represents the most clinically relevant alternative diagnosis¹⁴. BEN is a well-recognized constitutional variant prevalent in individuals of African, Middle Eastern, and Yemeni descent, characterized by persistently low neutrophil counts without increased susceptibility to infection or progressive hematologic deterioration¹⁵. In the present case, definitive confirmation of BEN was limited by the absence of historical hematologic records and Duffy antigen genotyping, which was not available at our institution. Additional differential diagnoses systematically considered included chronic idiopathic neutropenia, autoimmune neutropenia, nutritional deficiencies (folate, vitamin B12, and copper), viral etiologies (Epstein–Barr virus [EBV], cytomegalovirus [CMV], and parvovirus B19), inflammatory disorders, laboratory variability across testing platforms, and constitutional low neutrophil count variants¹⁶. Extensive multidisciplinary investigations did not identify an alternative etiology. A few abnormal findings were observed, although they were not conclusive. For instance, the markedly elevated erythrocyte sedimentation rate (ESR; 112 mm/h) suggested a nonspecific inflammatory process; however, ESR lacks diagnostic specificity and, in the absence of corroborating infectious, autoimmune, or malignant findings, was not considered sufficient to explain the persistent neutropenia. Similarly, the elevated anti-streptolysin O (ASO) titer was interpreted cautiously as evidence of previous streptococcal exposure rather than active disease. Although PANS and PANDAS were considered, the patient’s obsessive-compulsive symptoms evolved gradually over time rather than demonstrating the abrupt neuropsychiatric onset characteristic of these syndromes¹⁷, and no temporal relationship with an acute streptococcal infection was documented. Positive Epstein–Barr nuclear antigen (EBNA) immunoglobulin G (IgG) serology was likewise consistent with previous EBV exposure rather than active infection. Additional findings, including low protein S levels and delayed pubertal development, prompted further specialist evaluation but did not reveal a mechanistic explanation for the hematologic abnormalities. Collectively, these findings were considered clinically relevant and were carefully reviewed through multidisciplinary consultation; however, none demonstrated a convincing temporal, pathophysiologic, or causal relationship sufficient to explain the persistent neutropenia. They were therefore interpreted as associated findings rather than definitive etiologies, reinforcing the diagnostic uncertainty surrounding this case. Therefore, the hematologic findings were ultimately interpreted as representing either a possible escitalopram-associated phenomenon, an underlying benign constitutional neutropenic variant, or a multifactorial process involving individual hematologic susceptibility. Importantly, the absence of infectious complications, hematologic stability during rechallenge, and lack of progressive cytopenia collectively favor a benign, non-progressive variant over true agranulocytosis or bone marrow suppression.

The literature indicates that drug-induced neutropenia is most commonly associated with agents such as clozapine and antithyroid drugs, whereas SSRIs are implicated only in sporadic case reports^9,12. For example, Ay⁹ described a 45-year-old man on long-term sertraline (150 mg/day) who developed severe leukopenia (WBC count = 1.26 × 10³/µL, ANC = 150/µL). In that case, tapering sertraline to 50 mg/day led to rapid normalization of neutrophils within days. Another report concerned a 16-year-old adolescent with Guillain–Barré syndrome who developed agranulocytosis on sertraline; ANC fell to 80/µL after 24 days, requiring filgrastim, with recovery after discontinuation¹³. Paroxetine and fluoxetine have also been implicated: one patient on fluoxetine developed neutropenia at six months that promptly resolved after drug cessation¹⁸, and a 38-year-old patient receiving paroxetine (40 mg) developed neutropenia at six months, with neutrophil normalization six weeks after discontinuation.10 In an elderly patient, low-dose sertraline (50 mg) induced agranulocytosis after four weeks of treatment, requiring granulocyte colony-stimulating factor support¹⁹. Another case of pancytopenia after intravenous escitalopram abuse has been reported²⁰. These cases underscore that neutropenia has been reported idiosyncratically with various SSRIs; however, mechanistic explanations for SSRI-associated neutropenia remain elusive. Notably, no prior cases of neutropenia have been reported in pediatric patients receiving escitalopram at standard therapeutic doses. Accordingly, this case may represent the first detailed report of suspected SSRI-associated neutropenia in an adolescent with OCD, successfully managed with drug rechallenge. Importantly, neutropenia did not worsen after escitalopram rechallenge, suggesting a stable, non-progressive process rather than ongoing immune-mediated or toxic marrow injury.

The decision to rechallenge escitalopram was reached through a structured ethical and clinical deliberation process, the components of which are detailed here to support reproducibility. Following progressive psychiatric deterioration and inadequate response to nine weeks of structured CBT/ERP, the multidisciplinary team conducted repeated shared decision-making discussions with the patient and her legal guardian. These discussions explicitly addressed the diagnostic uncertainty regarding neutropenia etiology; the psychiatric risks of continued pharmacotherapy withdrawal, including functional decline and emerging affective symptoms; the available alternative pharmacologic options and their evidence base in pediatric OCD; the expected psychiatric benefits based on the prior documented response; the structured rechallenge protocol, including gradual dose titration; the predefined stopping criteria and emergency escalation thresholds; and the infection surveillance and family education plan. Patient assent and guardian informed consent were obtained before rechallenge initiation. This framework departs from the conventional pharmacovigilance principle of permanent avoidance following suspected drug-induced cytopenia but was considered ethically justifiable given the high psychiatric morbidity of undertreated OCD in this adolescent, the absence of a confirmed alternative etiology, the non-progressive hematologic course, and the absence of infectious complications. By analogy, individualized risk-benefit reassessment has been applied in other psychiatric contexts where pharmacotherapy carries hematologic risk and therapeutic alternatives are limited. However, this reasoning cannot be directly transposed from clozapine protocols, which operate within a specific, institutionally mandated monitoring framework with defined ANC thresholds that does not exist for escitalopram²¹. The rechallenge decision in this case was therefore based on clinical individualization rather than established protocol, and this distinction should be preserved when considering reproducibility. SSRIs represent evidence-based pharmacologic options and remain first-line medications for pediatric OCD when pharmacotherapy is indicated²². In this case, escitalopram was selected after individualized risk-benefit assessment and in accordance with patient and family preference, despite the need for careful monitoring and possible off-label consideration. The rechallenge monitoring strategy was designed to be structured, graduated, and reproducible. Escitalopram was titrated from 5 mg/day to 20 mg/day over three weeks, with dose escalation contingent on hematologic stability at each step. Weekly CBC with differential was performed throughout dose escalation and early stabilization, focusing on total WBC count, ANC trends, and evidence of progressive cytopenia. Serial measurements were obtained within the same laboratory system where feasible to minimize inter-laboratory variability. Infection surveillance was integrated into every clinical contact, and the family received structured counseling regarding fever, oral ulceration, mucosal symptoms, and indications for urgent reassessment. Predefined stopping thresholds included significant ANC decline, progression toward severe neutropenia or agranulocytosis, febrile neutropenia, recurrent infections, pancytopenia, or systemic deterioration.

Several implementation considerations deserve attention. First, the absence of a baseline CBC before treatment initiation represented the most consequential procedural limitation; baseline hematologic assessment should be considered in patients with potential hematologic vulnerability. Second, the unavailability of Duffy antigen genotyping precluded definitive confirmation of BEN, and centers with access to this test should incorporate it early in the differential workup. Third, the multidisciplinary framework may not be universally available. In lower-resource settings, a simplified approach incorporating serial CBC monitoring, hematology consultation, and predefined stopping criteria may represent a pragmatic alternative. Finally, patient and family engagement, facilitated through structured education and shared decision-making, was an important component of successful implementation. In particular, symptom improvement and the absence of hematologic progression supported a favorable benefit-risk balance in this case. This case highlights several clinical teachings that can be considered for broader applicability. Prompt discontinuation of the suspected agent remains appropriate when neutropenia is first detected, with peripheral smear confirmation and infection assessment. Careful clinical judgment is needed when psychiatric benefit is substantial and hematologic risk remains uncertain. Supportive measures, including empiric antibiotics or granulocyte colony-stimulating factor (G-CSF), may be required in severe or febrile cases, but they were not indicated in the present patient. Persistence of neutropenia beyond the expected recovery window should prompt broad differential diagnosis and multidisciplinary evaluation rather than premature drug attribution, particularly for agents such as SSRIs, for which hematologic adverse effects are exceptionally rare. Rechallenge may be cautiously considered in severe, refractory psychiatric illness when alternatives have proven insufficient, alternative etiologies have been systematically evaluated, the hematologic course is non-progressive, and the decision is reached through structured shared decision-making with predefined monitoring and stopping criteria. Routine CBC surveillance is not universally recommended for SSRIs; however, baseline testing and context-driven follow-up are prudent when hematologic risk is clinically suspected. The procedural and ethical framework described in this case is intended to complement, rather than replace, existing pharmacovigilance guidance and to provide clinicians with a reproducible reference when confronted with analogous diagnostic and therapeutic dilemmas.

Materials

List of materials used in this article

Name	Company	Catalog Number	Comments
Anti-Streptolysin O (ASO) Titer Assay	Not available retrospectively	Not available retrospectively	Streptococcal exposure assessment
Autoimmune Panel (RF, C3, C4)	Not available retrospectively	Not available retrospectively	Autoimmune evaluation
Brain Magnetic Resonance Imaging (MRI) System	Not available retrospectively	Not available retrospectively	Exclusion of neurologic pathology
Children's Yale-Brown Obsessive Compulsive Scale (CY-BOCS)	Yale University OCD Program	N/A	OCD severity assessment
Clinical Medical Record System	Oasis Plus Electronic Medical Record (EMR) System	N/A	Retrospective clinical data retrieval
Cognitive Behavioral Therapy (CBT) Protocol	Clinical Psychology Service	N/A	Psychotherapy intervention
Complete Blood Count (CBC) Analyzer	Not available retrospectively	Not available retrospectively	Hematologic monitoring
Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-5) Diagnostic Criteria	American Psychiatric Association	N/A	OCD diagnosis
Endocrine Laboratory Panel (TSH, prolactin, estradiol, progesterone, parathyroid hormone [PTH], vitamin D)	Not available retrospectively	Not available retrospectively	Endocrine evaluation
Epstein–Barr Virus (EBV) Serology Assay	Not available retrospectively	Not available retrospectively	Infectious disease evaluation
Erythrocyte Sedimentation Rate (ESR) Assay	Not available retrospectively	Not available retrospectively	Inflammatory evaluation
Escitalopram	Lundbeck	Not documented	SSRI treatment and rechallenge
Exposure and Response Prevention (ERP) Materials	Clinical Psychology Service	N/A	ERP intervention
Ferritin Assay	Not available retrospectively	Not available retrospectively	Nutritional and hematologic assessment
Folate Assay	Not available retrospectively	Not available retrospectively	Nutritional assessment
Hemophagocytic Lymphohistiocytosis (HLH) Screening Panel	Not available retrospectively	Not available retrospectively	Exclusion of hematologic disorders
Informed Consent Documentation	Institutional Clinical Records System	N/A	Guardian consent for publication
Multidisciplinary Consultation Workflow	Psychiatry, Hematology, Rheumatology, Infectious Diseases, Allergy and Immunology, Endocrinology, Neurology, and Cardiology Services	N/A	Diagnostic assessment and treatment decision-making
Pelvic Ultrasound System	Not available retrospectively	Not available retrospectively	Endocrine and developmental evaluation
Peripheral Blood Smear Analysis	Institutional Clinical Laboratory	N/A	Morphologic hematology evaluation
Protein C and Protein S Assays	Not available retrospectively	Not available retrospectively	Hematologic evaluation
Structured Clinical Psychiatric Interview	Child and Adolescent Psychiatry Clinic	N/A	Baseline and follow-up psychiatric assessment