Outcome Measurement in Functional Neurological Symptom Disorder

Outcome measurement in Functional Neurological Symptom Disorders (FNSDs) is particularly complex. Pressing questions include what kind of measure is more accurate or meaningful, or how to achieve standardisation in a clinically heterogenous group where subjective and objective observations of the same construct may deviate. This project aimed to build on the limited knowledge of measuring outcomes in FNSDs and attempts to address one of its inherent complexities; where clinical aspects of the disorder confound the usual prioritisation of "objective" over "subjective" (or patient-rated) measures. This PhD comprised a literature review and three research studies, each using different measures to assess the current status and (potential) outcomes in FNSD patients.

A narrative description of systematically identified literature on stress, distress, and arousal measures in FNSD presents an overarching profile of the relationships between subjective and objective study measures. Eighteen studies (12 functional seizures, six other FNSD) capturing 396 FNSD patients were included. Eleven reported no correlation between subjective and objective measures. Only four studies reported significant correlations (r's=-0.74-0.59, p's <0.05). The small number of studies and diverse methodologies limit the conclusions of this review. However, the review's findings underscore the importance of validating outcome measures in patients with FNSD, carefully selecting the most appropriate measures for the research objectives, and possibly combining different measures optimally to triangulate a patient's current state, level of functioning or disability.
Study One used factor analysis and Rasch modelling to investigate the psychometric properties of a novel FNSD-specific resource-based measure developed as an outcome measure for psychological therapies (The sElf-efficacy, assertiveness, Social support, self-awareness and helpful thinking (EASE) questionnaire). A 4-factor model identified self-efficacy (SE), self-awareness/assertiveness (SA), social support (SS) and interpersonal illness burden (IIB) as relevant domains. Each latent scale fits the Rasch model after refinement of the category responses and removing two items. With further improvement, the EASE-F has the potential to reliably measure self-reported SE, SA, SS, and IIB constructs which were found to be meaningful to patients with FNSD. This can identify patients with strengths and deficits in these constructs, allowing therapists to individualise interventions. Recommendations for refinement of future instrument versions, using the measure in clinical practice, and research in FNSDs are discussed.
Study Two sought to understand the urgent and emergency care (UEC) service usage patterns among FNSD patients. Retrospective FNSD patient data from 2013 to 2016 UEC records (including NHS 111 calls, ambulance services, A&E visits, and acute admissions) were used to compare FNSD UEC usage rates with those of the general population and to model rates before and after psychotherapy. FNSD patients displayed 23 to 60 times higher UEC usage than the general population. Emergency service usage rates showed a significant reduction in level (rate level change = -0.90--0.70, p's <0.05) immediately after psychotherapy. While this study was uncontrolled, and a causal relationship between psychotherapy and reduced UEC service use cannot be proven by its design, the decrease in pre-treatment service usage among FNSD patients mirrors treatment-related improvements in health status and functioning previously documented using self-reported outcome measures. Further research is warranted to elucidate features of emergency care service use by patients with FNSD, assess interventions' cost-effectiveness, and help to optimise limited health care resource allocation.
Study Three utilised a delay discounting and emotional bias task to assess if these measures could indicate the health state of FNSD patients and to compare findings in patients with those in healthy controls. This online-based study collected data on cognitive-affective functioning, decision-making and, indirectly, emotion regulation, alongside self-reported health data and indicators of mood while completing the tasks. Delay discounting (DD) was steeper in patients with FNSD, indicating a preference for less subjectively valuable immediate rewards. Patients displayed priming and interference effects for angry and happy facial expressions, which differed from the interference effects observed in healthy controls [F(1,76) = 3.5, p = 0.037, η2p = 0.084]. Modest associations (r's =0.26-0.33, p's <0.05) were found between the DD estimates and self-reported generalised anxiety, but not current feelings of anxiety in FNSD. There were no correlations with indices for negative affective priming or interference. These measures did not show predictive ability for self-reported difficulty regulating emotions, anxiety, depression or coping in FNSD. However, the fact that the DD task and self-reported constructs failed to correlate does not invalidate this objective test. The findings underscore the importance of using a combined approach to outcome measurement.
This project highlights the importance of a more comprehensive understanding of outcomes and measures that capture clinically valid and meaningful health information. Given that subjective and objective measures capture different aspects of health state or function, a combination of measurement approaches will likely produce the most comprehensive understanding of patients' current state or treatment outcome. Because of the attentional, emotional, and perceptual alterations implicated in FNSD and the variable external representations of these, the difference between objective and subjective measures represents an interesting observation in its own right. The size of the discrepancy between subjective and objective measures may provide additional valuable insights into the underlying pathology. Nonetheless, there is still a need for standardisation and consistency in FNSD outcome measurement and reporting. Several important factors, such as the timeframe of measures, the influence of confounding factors, and the variety of presentation of any aspect of the disorder (e.g., physiological, cognitive, social, or behavioural presentations of arousal/stress), will need to be considered when designing and interpreting measurements for research or clinical analysis of the patient group.