Beyond Feedback: A Rhetorical Analysis of Not-Upheld Complaints in Adult Neurodevelopmental Assessment

Abstract

Objective: To analyse the rhetorical strategies employed in formal complaints regarding adult Attention Deficit Hyperactivity Disorder and Autism assessments, specifically where no service failure was identified. Methods: A rhetorical analysis was conducted on 48 complaints determined to be not upheld overall, submitted to a UK NHS Trust between 2024 and 2025. Results: Complainants demonstrated high rhetorical sophistication, frequently deploying a “Triple Core” strategy: Causal Attribution (blaming the service for life failures), Emotional Impact Description (framing dissatisfaction as medical trauma), and Procedural Challenge (alleging administrative breach). A minority (18.8%) employed economic arguments, whilst 81.2% included specific outcome demands, such as a prescription or a specific diagnosis. Conclusions: Complaints in this area of clinical practice demonstrate sophisticated rhetorical construction, functioning as instruments of organisational pressure. The “Triple Core” strategy creates an epistemic conflict where clinical judgement is contested by the patient’s lived experience narrative. These findings suggest that high complaint volumes may reflect a systemic gap between public expectation and clinical criteria rather than safety failures.

1. Introduction

The healthcare literature consistently frames patient complaints as a critical resource for patient safety, service quality, and organisational learning, rather than as expressions of blame, hostility, or compensation seeking. Empirical reviews demonstrate that complainants most often aim to prevent recurrence of harm, improve care for future patients, and prompt system-level learning, particularly within primary care and hospital services [1,2,3]. Across qualitative and mixed-methods studies, patients commonly seek clear explanations, acknowledgement of error or distress, apologies, and visible accountability, with financial redress typically secondary or absent [4,5]. Analyses using structured coding frameworks, especially the Healthcare Complaints Analysis Tool, show that complaint content consistently clusters around three domains: clinical care and safety, organisational and management failures such as access, continuity, and coordination, and relational or communication problems including feeling unheard, disrespected, or insufficiently informed [6,7,8,9]. When complaints are systematically aggregated rather than handled as isolated events, they can reliably identify safety “hot spots” and “blind spots”, particularly diagnostic concerns, omissions of care, and failures during transitions, which are often poorly captured by conventional incident reporting systems [2,10]. Even despite this clear valuable contribution to quality, multiple reviews conclude that healthcare organisations frequently underuse complaint ‘signals’ focusing on individual case resolution rather than synthesis, learning, and feedback into quality improvement processes, thereby limiting the preventive and corrective potential of patient voice [11,12,13].

Historically, whilst complaints were viewed primarily as a mechanism for reporting safety failures there is evidence that more recently increasingly, they can be used as an instrument of negotiating service outputs [14]. In this contemporary landscape, the patient is positioned as an informed consumer, transforming complaints into interactions between consumerist demands and organisational resource constraints [6,15]. Parallel developments in patient rights discourse and online health activism have shifted the patient-provider dynamic mirroring the broader systemic transition from paternalistic medicine to consumerist healthcare model [16,17]. Patients increasingly present as informed consumers with clear expectations regarding diagnosis and treatment, rather than as passive recipients of professional opinion [18]. While the majority of complaints continue to serve as tools for quality improvement, accountability, and early warning of systemic issues [19,20], a significant minority present distinctive features. Segal [21,22] identifies these as ‘contestable complaints’, wherein patients challenge professional judgement in areas of expertise asymmetry. Furthermore, Bismark et al. [15] note that a small cohort of repeat complainants generates a disproportionate volume of these grievances with lower upheld rates, suggesting the mechanism is sometimes employed to pursue desired outcomes rather than provide constructive feedback.

Within this context, adult Attention Deficit Hyperactivity Disorder and Autism Spectrum Disorder services in the UK, particularly within the NHS, operate under substantial pressure as demand has risen exponentially whilst capacity has remained unable to match this growth [23]. NHS England reports unprecedented increases in adult Attention Deficit Hyperactivity Disorder referrals, with many areas facing multi-year waits and some trusts unable to maintain open waiting lists. In parallel, autism assessment waits have increased dramatically, with national datasets showing steep year-on-year rises since 2019. Taken together, these trends demonstrate that neurodevelopmental pathways are experiencing sustained demand far beyond capacity, resulting in prolonged waits, unmet need, and increasing clinical and operational risk [24,25,26,27]. Despite operating under this degree of pressure, these services face levels of public scrutiny unprecedented when compared with other mental health services, such as dementia services. In cases of dementia, diagnosis is generally clear and not contestable, whilst diagnoses of ADHD and Autism Spectrum Disorder rely heavily on subjective report and observation rather than objective biomarkers [28]. Diagnosis often confers access to medication, workplace adjustments, benefits, and identity validation, raising the stakes of non-diagnosis considerably. Online communities provide detailed guidance on diagnostic criteria and assessment questions [29,30], enabling patients to arrive highly prepared and often expecting to receive a diagnosis.

This combination of high demand, diagnostic complexity, and elevated public engagement raises questions about whether external pressures might influence clinical practice in neurodevelopmental services. A recent healthcare service review in the UK has documented a parallel dynamic when examining concerns about external pressures influencing clinical judgement in contested diagnostic areas. The Cass Review [31] whilst focused on gender identity services, noted that the “exceptional” toxicity of the public debate, characterised by polarisation and vilification on social media, has left clinicians “afraid to openly discuss their views” and “fearful of working” with gender-questioning young people, contributing to an environment where some practitioners abandoned normal clinical standards ([31], pp. 13, 20–22). Similar challenges have been described in chronic fatigue syndrome services and Lyme disease diagnosis [32,33]. Should comparable external pressures operate within neurodevelopmental services, similar challenges to clinical practice might emerge. Understanding whether such pressures exist, and how they manifest in patient–clinician interactions, requires methods capable of examining not only what patients communicate but how they construct their arguments when challenging clinical decisions.

These accounts raise a broader question about how healthcare systems should balance responsiveness to patient voice with protection of evidence-based clinical decision-making. Traditional complaint analysis methods, such as thematic analysis, provide a valid tool for healthcare providers by systematically categorising and summarising patient feedback [34]. This approach primarily addresses the “what” of patient communication, identifying specific issues, patterns, and recurring themes within complaints. By focusing on content, thematic analysis highlights areas where care delivery may fall short, such as communication breakdowns, delays, or perceived neglect. These insights are important for quality improvement initiatives and help prioritise interventions based on the frequency and severity of reported problems [19]. However, thematic analysis alone may not capture the complexity of patient expression, including the emotional undertones, intentions, and persuasive elements embedded in complaints.

A rhetorical approach offers a new lens by examining the “how” of patient communication. This method examines the construction of arguments, the use of language, and the deployment of persuasive strategies patients employ to convey their concerns effectively. Understanding these rhetorical techniques reveals deeper layers of meaning, such as emotional appeals, credibility assertions, and narrative framing [21], which influence how complaints are received and interpreted by healthcare professionals. This approach recognises that patient feedback is not merely a list of grievances, but a form of social interaction shaped by power dynamics, cultural contexts, and individual motivations. By analysing rhetorical strategies, healthcare systems can better appreciate the patient’s perspective, tailor responses empathetically, and design policies that acknowledge the complexity of patient experiences.

Rhetoric, understood as the strategic use of language to persuade, is ethically neutral [35,36]. Sophisticated rhetorical strategies are legitimate when people navigate resistant institutions [37]. However, challenges arise when rhetoric is deployed instrumentally to secure predetermined outcomes regardless of clinical evidence [22]. Distinguishing constructive advocacy from instrumental pressure therefore requires systematic attention to specific rhetorical patterns, explicit outcome demands, and comparison with independent investigation findings.

To examine these dynamics empirically, this study applies systematic rhetorical analysis to formal complaints in adult neurodevelopmental assessment services. The dataset comprises 48 complaints concerning ADHD and autism assessments that proceeded to full independent investigation.

The study addresses four research questions:

• Which rhetorical patterns characterise complaints that are subsequently not upheld?
• How do complainants construct argumentative force through emotional, procedural, and economic appeals?
• What do the rhetorical patterns identified in not-upheld complaints suggest about the relationship between argumentative construction and the contestation of clinical decisions?
• What are the implications for healthcare governance and complaint-handling procedures?

As the Clinical Lead of the service concerned, the author conducted this research within an action research framework [38]. To mitigate potential bias arising from insider knowledge, analysis was restricted to complaint texts alone, without reference to clinical records or prior knowledge of complainants. The study employed systematic application of established rhetorical frameworks, with all analytical decisions reported transparently. The independent investigation outcomes provided external validation of analytical findings.

Whilst the literature on patient complaints includes rhetorical analyses in general healthcare contexts [39], no published study has systematically applied rhetorical analysis to complaints in neurodevelopmental assessment services. This study addresses that gap by examining not-upheld complaints, offering insight into rhetorical strategies and their implications for governance.

2. Methodology

2.1. Research Design and Epistemological Position

This study employed rhetorical analysis to examine how complainants constructed arguments regarding a specialist neurodevelopmental assessment service. This methodology was selected because it treats language as a strategic communicative act and examines how discourse is deployed to persuade audiences and influence decision-making in organisational settings [40,41,42]. In contrast to thematic content analysis, which primarily identifies what topics are present, rhetorical analysis enables examination of how texts are organised to produce particular accounts of causality, responsibility, harm, and entitlement. It further examines how these accounts are oriented toward compelling organisational action.

The study was conducted within a critical realist epistemology. This position holds that underlying reality exists in clinical processes and procedural compliance, but that access to that reality is mediated through interpretive practices, including complainant narratives and organisational documentation [43]. Within this framework, rhetorical patterns are treated as observable textual features with social functions, rather than as direct indicators of truth or falsity. Where interpretive context was required to provide balanced reporting, investigation outcomes were consulted. However, this consultation occurred only after the coding framework had been developed and applied to complaint texts, ensuring that code identification remained grounded in textual features rather than being influenced by investigation outcomes.

2.2. Setting, Data Source, and Sampling Strategy

The study was conducted within the adult Attention Deficit Hyperactivity Disorder and Autism assessment service at South West Yorkshire Partnership NHS Trust. The dataset comprised all formal written complaints submitted to the service between 1 January 2024 and 31 December 2025 that met inclusion criteria. No sampling was undertaken; the dataset represents the complete eligible corpus within the defined period.

Complaints were included if they: (i) constituted formal written complaints submitted through statutory NHS complaints processes; (ii) concerned the clinical assessment process or diagnostic outcome; and (iii) had completed investigation with documented outcome recorded as “not upheld” overall was selected to characterise the rhetorical architecture of formally contested grievances where no procedural or clinical fault was established. Complaints concerning purely administrative matters without clinical assessment content were excluded. The final dataset comprised 48 complaints.

The focus on not-upheld complaints reflects this study’s aim to characterise the rhetorical architecture of grievances that function as instruments of organisational pressure in the absence of verified service failure. This is a distinctive analytical category that has not previously been examined systematically in neurodevelopmental contexts. Restricting the sample to not-upheld complaints does not imply comparison with upheld ones but it defines a coherent population of interest in its own right, enabling characterisation of how formal complaints are constructed when independent investigation finds no procedural or clinical breach.

Formal NHS complaint investigations assess multiple elements within each complaint and determine outcomes for each element separately before reaching an overall determination. The 48 complaints in this study were classified as “not upheld” overall, meaning that the preponderance of complained-about elements were found not to constitute service failures upon investigation. Whilst some complaints contained individual elements that were partially upheld (for example, administrative delays or communication issues), these represented a minority of the total elements raised, and the overall complaint determination was recorded as “not upheld.”

The analytic focus was directed specifically toward those elements of complaints where investigation did not identify procedural or clinical failure. This approach was not premised on an assumption that the service was error-free but rather served to isolate rhetorical strategies used when complainants’ experiential accounts diverged from organisational determinations of procedural compliance. By examining how grievances are linguistically constructed in contexts where independent investigation found no substantive breach, the study addresses a specific question: how do complainants build persuasive force when contesting professional judgement in the absence of identifiable service failure?

2.3. Data Management, Anonymisation, and Variables Extracted

All complaint texts were de-identified prior to analysis, with removal of names, addresses, and other direct identifiers. Each complaint was assigned a study identifier. Three classes of variables were extracted.

First, complainant characteristics were extracted from service records for descriptive reporting and subgroup analyses: age at complaint, gender, and the index assessment pathway referenced (ADHD or Autism). Second, complaint length was operationalised as word count. Word count was computed consistently across documents using a standardised counting method applied to the body of the complaint text, excluding duplicated headers, email footers, and appended administrative boilerplate where present, to minimise artefactual inflation. Third, inductively derived rhetorical codes were produced through the coding process described below. The complaint corpus was taken as written by the individual.

2.4. Inductive Rhetorical Coding Framework Derivation

The primary analytical procedure involved deriving the rhetorical codes directly from the text. Initial familiarisation involved reading each complaint in full repeatedly to understand overall structure, tone, and recurrent argumentative moves. During this phase, analytic notes were recorded on repeated linguistic forms (for example, causal connectors) and repeated structural forms (for example, serial interrogatives and enumerated allegations).

Following familiarisation, the corpus underwent line-by-line open coding. Open codes captured recurring rhetorical moves, including how complainants established causation, described harm, contested procedures, invoked financial considerations, challenged professional authority, and specified desired outcomes.

Provisional categories were developed by constant comparison across complaints, with attention to whether a candidate category was (a) recurring across multiple documents, (b) sufficiently distinct from adjacent categories, and (c) operationalisable through identifiable textual markers rather than inferred intent. Through this iterative clustering and refinement, six codes were derived that captured the dominant rhetorical architecture present across the corpus: Causal Attribution, Emotional Impact Description, Procedural Challenge, Cost Argument, Clinician Competence Questioning, and Specific Outcome Demand. These codes were retained because they appeared repeatedly across the dataset, were distinguishable by observable textual criteria, and together provided broad coverage of the principal argumentative structures.

To ensure that the inductively derived coding framework remained grounded within an established rhetorical tradition, the six codes were mapped onto core classical rhetorical concepts, particularly the Aristotelian appeals of ethos, pathos, and logos. Emotional Impact Description corresponds primarily to pathos, in which the complainant seeks to elicit moral urgency through the evocation of harm, injustice, or psychological distress. Procedural Challenge and Cost Argument align most closely with logos, constructing organisational, administrative, or economic reasoning to support the complaint’s position. Clinician Competence Questioning functions as an ethos-destabilising strategy, undermining the credibility of the clinician so that their professional judgement appears unreliable or morally compromised.

The remaining two codes, Causal Attribution and Specific Outcome Demand, operate as bridging devices across appeals. Causal Attribution links perceived harm to organisational culpability, establishing the logical foundation upon which other appeals rest. Specific Outcome Demand frames the complaint as a necessary remedial action rather than a discretionary request, combining rational justification with moral urgency. Explicit theoretical grounding enables the analysis to treat complaints as purposive rhetorical artefacts structured around recognisable persuasive logics, rather than as spontaneous narrative expressions of dissatisfaction.

Each code was defined with explicit inclusion and exclusion criteria to permit consistent application.

Causal Attribution was defined as explicit language linking service actions or inactions to negative consequences, signalled by causal connectors such as “because”, “as a result”, “led to”, “caused”, or equivalent constructions that attribute downstream harms to the service decision or process. This code did not include implicit causation or temporal sequencing without explicit causal language.

Emotional Impact Description was defined as language describing emotional or psychological harm attributed to the service interaction or outcome, including general distress as well as clinically framed terms (for example, anxiety, depression, trauma). The emphasis was on psychological impact rather than general dissatisfaction. This code did not include expressions of disappointment or frustration without framing as psychological harm.

Procedural Challenge was defined as challenging whether proper procedures had been followed. It included claims or questions suggesting that policies, guidelines, diagnostic criteria, legal duties, documentation standards, or communication processes had been breached or not properly applied. It also covered rights-based arguments (such as references to equality or data protection) when these were used to frame service decisions as breaches of statutory duties. This code did not include requests for procedural information without allegation of breach.

Cost Argument was defined as rhetorical use of financial framing or economic consequence to support the complaint’s demands or to pressure organisational action, including itemised costs, claims about cost to the NHS, cost-effectiveness assertions, or compensation framing where deployed as argument rather than incidental mention. This code did not include passing references to personal expense without argumentative function.

Clinician Competence Questioning was defined as explicit challenge to a clinician’s professional competence, conduct, ethics, qualifications, or fitness to practise, going beyond disagreement with outcome to impugn capability or professionalism. This code did not include disagreement with clinical judgement without challenge to competence.

Specific Outcome Demand was defined as explicit requests for predetermined outcomes, including specific clinical actions (for example, reassessment by a different clinician, acceptance of private diagnoses, prescribing of named medications such as Elvanse), procedural changes (for example, clinician removal from the process), or compensation, such that the complaint was oriented toward a specified remedy rather than open investigation. This code did not include general requests for review or reconsideration without specification of desired outcome.

Coding was binary at the complaint level for quantitative analyses: each code was recorded as present or absent in each complaint. Binary coding was selected because it permitted reliable cross-coder agreement and supported the prevalence, co-occurrence, and subgroup analyses reported in the Results.

2.5. Pilot Application and Refinement Prior to Full Coding

Before full-corpus coding, the provisional codebook was piloted on a subset of complaints selected to reflect variation in length, tone, and apparent argumentative density. This pilot phase served to resolve boundary issues, particularly the distinction between general dissatisfaction and Emotional Impact Description, and between procedural clarification requests and Procedural Challenge. Definitions were tightened and examples of prototypical textual markers were documented in a codebook to support consistent application.

2.6. Full Systematic Coding and Audit Trail

All 48 complaints were then coded using the refined codebook. Coding was conducted at the level of text segments (sentence and clause) to allow transparent linkage between coded decisions and source text, whilst the quantitative dataset used complaint-level presence or absence. Coding was managed using NVivo14 which facilitated systematic application of codes to text segments, maintained linkage between coded segments and their source locations, and supported extraction of coded excerpts for analysis. For each code marked present, at least one supporting excerpt location was recorded to permit auditability. Coding was performed exclusively on complaint texts. Investigation reports and clinical records were not consulted during the coding stage, but were reviewed after coding completion to provide interpretive context for reporting, as described in Section 2.1.

2.7. Inter-Rater Reliability Assessment

To validate that the six rhetorical codes were not dependent on the single analyst’s interpretive framework, inter-rater reliability was assessed through independent coding by a second analyst who held postgraduate qualifications in qualitative research methods, had no involvement in service provision or clinical training, and was naive to complaint outcomes. A subset of 10 complaints was selected from the 48-complaint corpus through purposive maximum variation sampling to ensure representation across the range of complaint length (521 to 4892 words) and argumentative density observed in the full dataset.

The second coder independently applied the six operational code definitions (Causal Attribution, Emotional Impact Description, Procedural Challenge, Cost Argument, Clinician Competence Questioning, and Specific Outcome Demand) to the 10-complaint subset, producing binary presence/absence decisions for each code in each complaint. Agreement was computed as percentage agreement across the 60 total coding decisions (10 complaints × 6 codes). Percentage agreement was selected over Cohen’s kappa because the binary coding structure and relatively balanced code prevalence meant that chance agreement was unlikely to artificially inflate concordance rates. Following reliability assessment, disagreements were reviewed to identify any systematic interpretive differences, and final coding of the full corpus was completed by the primary analyst using the refined operational definitions. Inter-rater reliability results are reported in Section 3.3.

2.8. Quantitative Integration and Statistical Procedures

After qualitative coding, the binary presence matrix (48 complaints × 6 codes) was used to produce quantitative outputs. Code prevalence was calculated as the proportion of complaints containing each code. Confidence intervals for prevalence were computed using the Wilson score interval, which provides more accurate coverage for small to moderate sample sizes than the normal approximation. Co-occurrence was calculated by counting, for each pair of codes, the number of complaints in which both codes were present, producing the 6 × 6 co-occurrence matrix. Rhetorical sophistication was operationalised as the number of distinct codes present per complaint, enabling calculation of mean code count and distribution across 0–6.

Descriptive statistics for age were computed (mean, standard deviation, median, range), and the mean age confidence interval was reported. Group comparisons were performed to examine whether code prevalence or complainant characteristics varied by condition pathway (ADHD versus autism) or by gender. For categorical comparisons, Chi-square tests of independence were used where all expected cell frequencies exceeded 5; Fisher’s exact test was applied where any expected cell frequency was less than 5. For continuous variables, data distribution was checked for normality using visual inspection of histograms and Q-Q plots. Independent-samples t-tests were used to compare mean ages between groups where normality assumptions were met; Mann–Whitney U tests were employed where assumptions were violated. All statistical analyses were conducted using SPSS version 29.

2.9. Use of Investigation Findings for Contextual Interpretation

Investigation outcomes were used only after completion of code derivation and full-corpus coding. Their role was interpretive rather than constitutive; they provided contextual balance when presenting exemplar quotations in the Results, enabling the reader to see how rhetorical claims related to the documented outcome of formal investigation. Investigation findings were not used to determine whether a rhetorical code was present.

2.10. Quality, Rigour, and Reflexivity

Credibility was supported by full-corpus coding with an explicit audit trail, a defined codebook with inclusion and exclusion criteria, and an independent reliability check. Dependability was supported through documentation of the analytic sequence from open coding through refinement and full application. Confirmability was supported by restricting code assignment to observable textual features and by separating coding from investigation outcome review until after coding was complete. Transferability was supported by providing operational definitions that can be applied to other complaint datasets.

Given the author’s dual role as Clinical Lead and researcher, specific reflexive attention was paid to the tension between organisational accountability and data interpretation. Because the primary analyst held a service leadership role, reflexive safeguards were necessary to minimise confirmation bias. The principal safeguard was procedural separation between (i) inductive code derivation from the texts and (ii) interpretive consideration of investigation outcomes. A further safeguard was independent reliability coding by a second coder who was independent of the service hierarchy and had no prior involvement in complaint handling. This prevented organisational power dynamics from influencing the resolution of coding discrepancies. A full discussion of positionality and its epistemological implications is provided in the Discussion.

2.11. Ethics and Governance

This study constituted service evaluation and quality improvement activity conducted within the author’s clinical governance responsibilities. As such, it did not require Health Research Authority ethical approval under UK policy framework for health and social care research. All data were handled in accordance with NHS confidentiality and information governance principles. Complaints were analysed as de-identified texts for service-level learning purposes. Where direct quotations were used in reporting, identifying details were removed or masked. Where verbatim quotations risked jigsaw identification, they were paraphrased to preserve rhetorical structure whilst preventing re-identification.

Data are stored securely in accordance with NHS information governance policies and are not available for sharing due to confidentiality restrictions. Statistical procedures followed standard methods as implemented in SPSS.

The use of numerical counts in this study reflects a structured form of rhetorical analysis, not a move toward thematic content analysis. The numbers are used to show how often different rhetorical strategies appear and how they occur together, rather than to assess sentiment or psychological state. This approach is consistent with empirical work in rhetorical studies that uses patterns across large datasets to examine how persuasive forms remain stable, spread, and shape organisational practice [44].

3. Results

3.1. Sample Characteristics

Across the review period, the Service made in total 7447 decisions relating to diagnostic outcomes and/or screening and triage in both ADHD and Autism pathways. The complaint rate for the ADHD pathway was 0.98% (n = 39), which was significantly higher than the 0.26% (n = 9) observed for the Autism pathway (χ² = 13.68, p < 0.001). Although ADHD decision volume increased by 93% in 2025, the complaint rate per 100 decisions reduced by approximately half, decreasing from 1.47% to 0.72%. By contrast, the Autism pathway showed a stable and consistently lower complaint rate across both years.

The sample consisted of 48 complaints received in 2024 (n = 25) and 2025 (n = 23). The mean age of complainants was 35.25 years (SD = 10.61), with a median of 34.5 years and a range of 18 to 63 years. The 95% confidence interval for the mean age was 32.17 to 38.33 years. Regarding gender, 26 complainants (54%) were female and 22 (46%) were male. The reported condition was ADHD in 39 cases (81%) and Autism in 9 cases (19% as detailed in Table 1).

Table 1. Cross-tabulation of gender and condition.

Gender	ADHD	Autism	Total
Female	21	5	26
Male	18	4	22
Total	39	9	48

A chi-square test of independence indicated no association between gender and condition (χ² = 0.00, p = 1.00). An independent-samples t-test showed no evidence of an age difference between male complainants (M = 32.64 years, SD = 9.27) and female complainants (M = 37.46 years, SD = 11.36; t = −1.60, p = 0.117). Similarly, age did not differ by reported condition, with comparable ages in the ADHD group (M = 35.95 years, SD = 10.96) and the Autism group (M = 32.22 years, SD = 8.71; t = 0.95, p = 0.348).

Complaint narratives varied substantially in length (M = 2847 words; range: 523 to 8942 words). The distribution was positively skewed: 19 complaints (40%) exceeded 3000 words and 8 (17%) exceeded 5000 words. This unusually high word count, relative to typical service feedback, is consistent with substantial cognitive and emotional investment in complaint construction.

3.2. Prevalence of Rhetorical Codes

The rhetorical analysis identified six distinct codes deployed across the complaint narratives. Table 2 presents the prevalence of each code across the 48-complaint sample.

Table 2. Prevalence of Rhetorical Codes (n = 48).

Rhetorical Code	n	%	95% CI
Procedural Challenge	46	95.8%	85.7–99.5%
Causal Attribution	44	91.7%	80.0–97.7%
Emotional Impact Description	44	91.7%	80.0–97.7%
Clinician Competence Questioning	39	81.2%	67.4–91.1%
Specific Outcome Demand	39	81.2%	67.4–91.1%
Cost Argument	9	18.8%	8.9–32.6%

The most prevalent code was Procedural Challenge, present in 46 complaints (95.8%, 95% CI 85.7–99.5%). This was followed by two codes with identical prevalence: Causal Attribution and Emotional Impact Description, each present in 44 complaints (91.7%, 95% CI 80.0–97.7%). Clinician Competence Questioning and Specific Outcome Demand demonstrated identical prevalence at 39 complaints each (81.2%, 95% CI 67.4–91.1%). Cost Argument was the least prevalent code, appearing in only 9 complaints (18.8%, 95% CI 8.9–32.6%).

The mean number of codes deployed per complaint was 4.60 (SD 1.15), indicating that complainants typically employed multiple rhetorical strategies simultaneously rather than relying on single argumentative approaches. Only two complaints (4.2%) deployed two or fewer codes, whilst 37 complaints (77.1%) deployed five or six codes, demonstrating high rhetorical sophistication across the majority of the sample.

3.3. Inter-Rater Reliability Outcomes

Following the procedure outlined in Section 2.7, the inter-rater reliability analysis produced an overall agreement of 92% (55 of 60) across all coding decisions, indicating substantial consistency in code application. The specific agreement rates for each rhetorical code are presented in Table 3.

Table 3. Inter-rater Reliability by Rhetorical Code (n = 10 complaints).

Rhetorical Code	Primary Coder Present	Second Coder Present	Agreement	Disagreements
Causal Attribution	10	10	10/10 (100%)	0
Emotional Impact Description	8	8	9/10 (90%)	1
Procedural Challenge	7	9	8/10 (80%)	2
Clinician Competence Questioning	5	6	9/10 (90%)	1
Specific Outcome Demand	9	9	10/10 (100%)	0
Cost Argument	3	3	10/10 (100%)	0
Overall Agreement			55/60 (92%)	5

The second coder independently applied the six operational code definitions (Causal Attribution, Emotional Impact Description, Procedural Challenge, Cost Argument, Clinician Competence Questioning, and Specific Outcome Demand) to the 10-complaint subset, producing binary presence/absence decisions for each code in each complaint. Agreement was computed as the percentage of concordant coding decisions across the 60 total coding decisions (10 complaints × 6 codes).

Overall, agreement across all 60 coding decisions was 92% (55/60), indicating substantial consistency in code application. Three codes demonstrated perfect agreement (100%) across the 10-complaint subset: Causal Attribution, Specific Outcome Demand, and Cost Argument. These codes showed perfect concordance because they depend on explicit linguistic markers (causal connectors for Causal Attribution, itemised financial information for Cost Argument) or unambiguous textual features (explicit requests for predetermined clinical actions for Specific Outcome Demand).

The remaining three codes each showed one or two disagreements, reflecting interpretive boundaries in their application. Emotional Impact Description achieved 90% agreement (9/10), with one disagreement concerning threshold interpretation. Clinician Competence Questioning achieved 90% agreement (9/10), with one disagreement about the boundary between clinical disagreement and competence questioning. Procedural Challenge showed 80% agreement (8/10), with two disagreements concerning the distinction between outcome-focused questions and systematic procedural interrogation.

3.4. Co-Occurrence Patterns

Analysis of code co-occurrence revealed consistent patterns of multi-strategy deployment. Table 4 presents the co-occurrence matrix for all six codes.

Table 4. Co-occurrence Matrix for Rhetorical Codes (n = 48).

	CA	EI	PC	Cost	CC	SOD
Causal Attribution (CA)	44	43	43	9	38	37
Emotional Impact (EI)	-	44	43	9	38	37
Procedural Challenge (PC)	-	-	46	9	38	38
Cost Argument (Cost)	-	-	-	9	8	9
Clinician Competence (CC)	-	-	-	-	39	33
Specific Outcome Demand (SOD)	-	-	-	-	-	39

Values represent number of complaints where both codes co-occur.

Highest co-occurrence patterns:

• CA + EI + PC (triple core): 43/48 (89.6%);
• PC + CC: 38/48 (79.2%);
• PC + SOD: 38/48 (79.2%).

The highest co-occurrence rate was observed for the triple combination of Causal Attribution, Emotional Impact Description, and Procedural Challenge, present together in 43 complaints (89.6%). This finding indicates that the vast majority of complainants simultaneously attributed negative outcomes to service failures, described emotional or psychological harm, and challenged the procedures or processes employed during assessment.

Strong dyadic co-occurrence was observed between Procedural Challenge and Clinician Competence Questioning (38/48, 79.2%), Procedural Challenge and Specific Outcome Demand (38/48, 79.2%), Causal Attribution and Clinician Competence Questioning (38/48, 79.2%), and Emotional Impact Description and Clinician Competence Questioning (38/48, 79.2%). These patterns suggest that procedural challenges, competence questioning, and outcome demands form an integrated argumentative structure in the majority of complaints.

Cost Argument demonstrated minimal co-occurrence with other codes, appearing alongside other rhetorical strategies in only 16.7–18.8% of complaints. This finding indicates that economic framing, when deployed, tends to be an additional rather than core rhetorical strategy.

Examination of the coded rhetorical patterns also revealed consistent movement through classical stasis categories within the majority of complaints. Many texts progressed from conjectural stasis (assertions concerning what occurred) into definitional stasis (contesting the nature or legitimacy of the event), and ultimately into qualitative stasis (asserting moral or procedural wrongdoing). This movement was particularly visible within the “Triple Core” configuration, in which complainants first established a narrative of causation, then framed this causation as organisational failure, and finally asserted that the resulting outcome constituted a harm requiring organisational redress. Identifying this stasis progression demonstrates that these complaints do not simply express dissatisfaction but strategically reposition the dispute at higher levels of contestation. In practical terms, the rhetorical shift from factual disagreement to moral indictment narrows the range of acceptable organisational responses and contributes to the pressure dynamics described later in the Discussion.

3.5. Rhetorical Code 1: Causal Attribution

Causal Attribution (91.7%) constructs a direct causal chain between the diagnostic outcome and comprehensive personal collapse, using “domino effect” sequencing to frame non-diagnosis as the sole cause of failures across employment, relationships, and health.

Example: A female complainant with Attention Deficit Hyperactivity Disorder described business closure, relationship breakdown, and disordered eating following her assessment, summarising the impact as “This broke me.” The grievance attributed all documented life difficulties directly to the diagnostic outcome. Investigation found that these difficulties predated the assessment by several months and that the assessment was conducted in accordance with NICE guidelines.

3.6. Rhetorical Code 2: Emotional Impact Description

Emotional Impact Description (91.7%) medicalises dissatisfaction by substituting clinical terminology such as “trauma”, “PTSD”, “psychological damage” for lay expressions of displeasure, framing the assessment interaction itself as a pathogenic event.

Example: A male Autism complainant rejected the clinician’s formulation using absolute assertions (“I have not depression”, “I am not sad”) and extended his grievance to future patients: “I despair at anyone going through your service with him.” Investigation found that depression screening indicated moderate symptoms and that the presentation did not meet Autism diagnostic criteria.

3.7. Rhetorical Code 3: Procedural Challenge

Procedural Challenge (95.8%) repurposes administrative protocols, rights-based legislation, and service specifications to contest the legitimacy of the clinical outcome, creating a bureaucratic burden that demands institutional response regardless of clinical merit.

Example: A male Autism complainant deployed five consecutive “Why?” questions escalating from clinical minutiae to legal accusations including a claimed GDPR breach. Investigation found no breach: the staff member held appropriate access, no discrimination occurred, and verbal communication was routinely accepted by the service. A second complainant invoked the Equality Act, framing a preference for verbal communication as a statutory reasonable adjustment. Investigation found that multiple communication formats had been offered and that the complaint itself of 2400 words of complex prose was inconsistent with the claimed barrier.

3.8. Rhetorical Code 4: Cost Argument

Cost Argument (18.8%) is consistent with the complainant’s individual demands with institutional fiscal responsibility, either arguing that granting the diagnosis prevents greater future NHS cost (“invest to save”) or constructing private expenditure as a compensable debt arising from service failure.

Example: A female complainant with Attention Deficit Hyperactivity Disorder characterised the non-diagnosis as generating “enormous” NHS costs and invoked her son’s existing diagnosis as a heritability argument for fiscal efficiency. Investigation found that no medication costs had been incurred and that the cited additional services predated the assessment. A second complainant provided itemised private costs to the penny (£1185 assessment; £895.58 medication), framing reimbursement as the only logical form of justice. Investigation found the diagnosis was not supported by assessment evidence and that NHS monitoring had been offered.

3.9. Rhetorical Code 5: Clinician Competence Questioning

Clinician Competence Questioning (81.2%) shifts challenge from the clinical decision to the decision-maker, deploying delegitimisation strategies that target professional standing and ethical conduct to render the diagnostic conclusion void.

Example: A male complainant with Attention Deficit Hyperactivity Disorder characterised standard triangulation of clinical records as “gaslighting” and “unethical”, arguing the clinician lacked epistemic authority to interpret notes from appointments he did not attend. Investigation found that the records reviewed documented bipolar concerns and that differential diagnosis had been appropriately applied. A second complainant declared the clinician “unfit to practise”, raising the stakes to a professional conduct matter requiring formal investigation. Investigation found that assessment followed standard procedures and that the presentation was inconsistent with Autism criteria.

3.10. Rhetorical Code 6: Specific Outcome Demand

Specific Outcome Demand (81.2%) reveals the transactional function of these complaints: unlike grievances seeking explanation or apology, these texts function as negotiation instruments, using the complaint as leverage to obtain a specific clinical resource denied through the assessment pathway. Demand types included referral to a named clinic (37.5%), medication prescription (25.0%), acceptance of a private diagnosis (16.7%), reassessment by a different clinician (12.5%), clinician dismissal (4.2%), and financial compensation (2.1%).

Example: A female complainant with Attention Deficit Hyperactivity Disorder demanded referral to a named clinic and prescription of a specific medication (Elvanse), structuring the complaint as an ultimatum: resolution conditional on supply of the named drug. A second complainant demanded organisational support for external reassessment, framing the service’s quality assurance processes as “interference” with her right to a second opinion. Both complaints were not upheld; in neither case did investigation identify a clinical or procedural basis for the demanded outcome.

3.11. Patterns of Multi-Code Deployment

Analysis of rhetorical sophistication, operationalised as the number of distinct codes deployed within a single complaint, revealed substantial variation. Table 5 presents the distribution of rhetorical sophistication across the sample.

Table 5. Distribution of Rhetorical Sophistication (n = 48).

Number of Codes Deployed	n	%	Interpretation
6 codes	8	16.7%	Maximal deployment
5 codes	29	60.4%	High sophistication
4 codes	7	14.6%	Moderate sophistication
3 codes	2	4.2%	Low-moderate
0–2 codes	2	4.2%	Minimal deployment

The majority of complaints (n = 37, 77.1%) showed high rhetorical sophistication, deploying five or six codes simultaneously. Only two complaints (4.2%) deployed three codes, and two complaints (4.2%) deployed two or fewer codes. This distribution indicates that multi-strategy deployment is normative rather than exceptional within this sample.

The most common pattern was deployment of Causal Attribution, Emotional Impact Description, Procedural Challenge, Clinician Competence Questioning, and Specific Outcome Demand without Cost Argument (n = 27, 56.2% of sample). This five-code combination appeared to constitute a standard rhetorical template employed by over half of complainants. Only eight complaints (16.7%) deployed all six codes, suggesting that Cost Argument functions as an optional addition to the core five-code structure rather than a fundamental element of the rhetorical architecture.

3.12. Variation by Assessment Type and Demographics

Analysis examined whether rhetorical code deployment varied by assessment type (ADHD versus Autism) or complainant demographics. Table 6 presents code prevalence stratified by assessment type.

Table 6. Code Prevalence by Assessment Type.

Code	ADHD (n = 39)	Autism (n = 9)	Difference
Causal Attribution	36 (92.3%)	8 (88.9%)	+3.4%
Emotional Impact	36 (92.3%)	8 (88.9%)	+3.4%
Procedural Challenge	37 (94.9%)	9 (100%)	−5.1%
Cost Argument	8 (20.5%)	1 (11.1%)	+9.4%
Clinician Competence	32 (82.1%)	7 (77.8%)	+4.3%
Specific Outcome Demand	32 (82.1%)	7 (77.8%)	+4.3%

No substantial differences in code prevalence were observed between ADHD and Autism complaints. Autism complaints demonstrated slightly higher Procedural Challenge prevalence (100% versus 94.9%), but this difference is not statistically meaningful given the small Autism subsample (n = 9). All other codes showed prevalence differences of less than 5 percentage points between assessment types.

Similarly, analysis by gender revealed no meaningful differences in rhetorical code deployment. Female complainants demonstrated marginally higher Cost Argument deployment (23.1% versus 13.6%) and Specific Outcome Demand deployment (84.6% versus 77.3%), but these differences were small and not statistically significant given sample size limitations.

Age showed minimal association with code deployment. Mean age of complainants deploying each code versus those not deploying each code differed by fewer than four years across all codes, suggesting that rhetorical strategies are employed consistently across the adult age range represented in this sample.

4. Discussion

4.1. Interpretive Framing

Before interpreting the rhetorical patterns identified in this analysis, it is essential to establish an interpretive framework. The identification of recurring rhetorical strategies does not imply insincerity, instrumentality, or manipulation on the part of complainants. Strategic rhetorical construction may coexist with genuine distress, frustration, or perceived injustice. Sophisticated use of language does not preclude authentic experience of harm or legitimate concern about service quality. The present analysis interprets rhetoric in terms of its textual form and organisational effects rather than as a judgement about complainant motive or authenticity. This distinction is maintained throughout the discussion that follows. This interpretive position is consistent with established approaches in the rhetorical analysis of patient complaints, which treat persuasive construction as analytically significant without implying negative judgements about complainant motive or sincerity [21,39].

4.2. The Rhetorical Architecture of Pressure

The analysis of 48 complaints reveals a consistent pattern in the construction of grievances against neurodevelopmental assessment services. Rather than appearing as unstructured expressions of dissatisfaction, these texts demonstrate sophisticated rhetorical architecture that functions to exert pressure across multiple dimensions. The prevalence of high rhetorical sophistication, with 77.1% of complaints deploying five or six distinct codes simultaneously, suggests that this represents a coherent genre of organisational challenge rather than isolated individual disputes.

The central structural element of this pattern is what might be termed the “Triple Core”: the simultaneous deployment of Causal Attribution, Emotional Impact Description, and Procedural Challenge, present in 89.6% of complaints. This combination creates a multi-dimensional pressure mechanism. By linking administrative procedural concerns to descriptions of emotional harm, complainants construct scenarios wherein organisations face constrained response options. When a complaint links a specific clinical decision to serious life consequences, frames this connection in terms of psychological distress, and alleges procedural irregularities, it creates a rhetorical situation wherein organisations must respond simultaneously to clinical, emotional, and administrative dimensions.

A response focused solely on clinical reasoning may be perceived as dismissing the described suffering. A response expressing empathy for distress may be interpreted as acknowledging causal responsibility. A response defending procedural compliance may be viewed as bureaucratic and unresponsive to human impact. This convergence of appeals creates a complex rhetorical environment for organisational response, positioning institutions in defensive postures regardless of whether procedural or clinical breaches occurred. The effect may be to shift institutional focus from clinical decision defence to management of multiple simultaneous concerns.

This rhetorical architecture creates particular forms of organisational pressure. Whilst Cost Argument was the least prevalent strategy (18.8%), its deployment alongside core rhetorical codes is noteworthy. When present, economic framing positions diagnosis as having calculable financial value, either through avoided future NHS costs or through framing diagnostic refusal as resource misallocation. This strategy exemplifies the consumerist healthcare model described in the literature, wherein patients position themselves as informed consumers presenting business cases rather than as recipients of clinical determinations. The effect is to introduce economic reasoning alongside clinical criteria, creating additional dimensions of contestation.

The most frequent pattern observed involves deployment of procedural challenges to contest clinical decisions. By framing clinical disagreements as procedural failures, rights violations, or breaches of statutory duties (such as those under equality or data protection legislation), complaints position institutions in defensive administrative stances. This strategy has the effect of ensuring that even where clinical judgement is sound, the operational response requirements (measured in investigation time, documentation burden, and organisational attention) are substantial. The rhetorical patterns suggest that procedural contention may be more effective than direct challenge to clinical reasoning, as procedural allegations require comprehensive administrative response regardless of clinical merit.

4.3. The Epistemic Dimension

Underlying these rhetorical patterns is a tension regarding epistemic authority in neurodevelopmental assessment. Traditional clinical models position clinicians as arbiters of diagnosis, relying on structured assessment, differential diagnosis, and application of standardised criteria. The rhetorical patterns identified here suggest an alternative framework wherein personal experience and self-identification hold primary authority, with clinical assessment serving to validate pre-existing self-understanding rather than to investigate diagnostic possibilities.

An important feature of complaint rhetoric is the simultaneous construction of complainant credibility and challenge to clinician credibility. Whilst Clinician Competence Questioning (present in 81.2% of complaints) challenges the legitimacy, neutrality, or professionalism of assessors, complainants often position themselves through appeals to moral authority, altruistic concern, or authenticity of lived experience. References to acting in others’ interests (“so that others do not suffer similarly”) or to lifelong personal experience function as credibility-affirming strategies. The resulting rhetorical structure is asymmetrical: clinician authority is contested whilst complainant authority is elevated. This dual movement strengthens persuasive force by simultaneously weakening organisational voice and amplifying complainant perspective.

The high prevalence of Specific Outcome Demands (81.2%) indicates that most complainants approach the complaints process with clear expectations regarding desired remedies, including reassessment, acceptance of alternative diagnostic opinions, prescribing of specific medications, or removal of particular clinicians from their care. Whether such specificity reflects instrumental use of complaints mechanisms, necessary assertiveness when navigating gatekeeping systems, advocacy informed by online communities, or combinations thereof cannot be determined from textual analysis alone. However, the pattern suggests that for many complainants, the complaints mechanism functions not only as an accountability process but as an avenue for revisiting concluded clinical decisions.

The frequent questioning of clinician competence in contexts where outcomes diverge from patient expectations warrants particular attention. When clinical conclusions differ from patient self-understanding, the authority to make such determinations may be contested. The application of differential diagnosis (standard clinical practice involving consideration of alternative explanations) may be experienced by some patients as dismissal or invalidation, particularly where online communities or prior self-diagnosis has established strong conviction about diagnostic identity. This creates tension wherein rigorous clinical assessment may be perceived as hostile or prejudiced when it fails to align with patient self-identification.

This dynamic shows some parallels with tensions documented in the Cass Review [31] regarding gender identity services, wherein clinicians reported concerns that differential diagnosis might be interpreted as evidence of prejudice rather than as standard clinical practice. Similarly, the rhetorical patterns observed here suggest that neurodevelopmental assessment occurs in an environment wherein clinical formulations contrary to patient expectations risk triggering comprehensive challenges to professional competence. The deployment of harm and trauma language in complaints raises the moral stakes, positioning non-diagnosis as potential injury rather than as neutral clinical conclusion. Whether this creates pressure that affects clinical decision-making cannot be determined from this study, but the rhetorical environment may create tension for clinicians between thoroughness of assessment and potential for complaint escalation.

4.4. The Divergence Between Procedural Compliance and Experiential Satisfaction

The finding that none of the analysed aspects of these 48 complaints were upheld following investigation, despite extensive documentation of alleged failures, requires interpretation within a critical realist framework. This pattern points to a divergence in what is being measured. The complaint investigation process assesses procedural adherence: whether clinicians followed guidelines, whether services followed policies, whether documentation met standards. By these metrics, the service was found to be functioning in accordance with expected standards.

However, complainants may be seeking validation of a different kind which can be an acknowledgement that their self-understanding should have determined clinical conclusions, or that their distress warrants the diagnosis they believe would explain it. Research concerning shared decision-making suggests that patients expect clinical pathways to be co-created, directly integrating their personal priorities and self-understanding. Furthermore, frameworks addressing epistemic injustice demonstrate that bypassing a patient’s self-knowledge during a clinical assessment invalidates their lived experience and generates dissatisfaction [45]. Because investigation processes are designed to audit procedure and clinical reasoning rather than to validate identity or affirm self-diagnosis, they may systematically fail to address what complainants experience as the core issue. The rhetorical strategies identified (particularly detailed Causal Attributions linking absence of diagnosis to life difficulties) may represent attempts to bridge this gap and to compel institutional acknowledgement of distress even where diagnostic criteria were not met.

When investigation reports conclude that guidelines were followed and clinical reasoning was sound, they are factually accurate regarding procedure but may not address the experiential dimension of the complaint. For complainants, this may reinforce perception of institutional unresponsiveness, creating cycles wherein procedural correctness and patient satisfaction never intersect. This divergence suggests that complaint volume in this context may not primarily indicate clinical error but rather reflects structural features of services wherein the definition and purpose of assessment are contested.

If a substantial proportion of patients approach assessment expecting validation of self-diagnosis or identity confirmation rather than open clinical investigation, then services maintaining rigorous diagnostic gatekeeping will likely generate sustained complaint volumes regardless of quality of care. The pattern of comprehensive investigation finding no upheld elements, whilst institutionally validating, presents a paradox wherein services may be procedurally compliant yet failing to address user distress, as evidenced by complaint intensity and sophistication.

The recurrence of five- and six-code rhetorical configurations suggests these complaints may be understood as an emerging rhetorical genre rather than as isolated individual expressions. Consistent with Miller’s account [42] of genre as recurrent social action, these texts appear to reflect stabilising communicative templates for challenging contested diagnostic outcomes. The patterned deployment of causation, harm language, procedural challenge, competence questioning, and explicit outcome demands indicates that complaints mechanisms may be used in regularised ways to re-enter or renegotiate previously concluded clinical decisions. Conceptualising these patterns as genre characteristics of the neurodevelopmental assessment environment rather than as individual narrative choices reframes them as structured modes of organisational contestation.

4.5. Implications for Clinical Governance and Practice

These findings have significant implications for governance and management of neurodevelopmental assessment services. Leaders should recognise that the presence of sophisticated, multi-dimensional complaints may not primarily indicate service failure but rather may represent characteristic features of domains wherein clinical authority and patient expectations diverge substantially. Traditional metrics of complaint volume may be unreliable indicators of quality in this specific pathway, functioning instead as indicators of gaps between public expectation and clinical criteria [11,12,13].

Governance frameworks would benefit from distinguishing between complaints primarily seeking investigation of potential procedural failures and those primarily seeking to revisit clinical decisions through alternative pathways [14,19]. Whilst most complaints likely contain elements of both, the distinction has implications for organisational response. Services must maintain rigorous investigation of all substantive allegations whilst also supporting clinicians in maintaining evidence-based practice under pressure.

If practitioners perceive that robust and credible diagnostic decision-making leaves them vulnerable to sustained challenges to their professional competence, there is risk of defensive practice patterns, including potential lowering of diagnostic thresholds to avoid conflict [31]. Organisations should provide explicit assurance that adherence to evidence-based guidelines will be supported even when challenged through sophisticated complaint rhetoric. This requires balancing accountability (genuine procedural errors must be identified and addressed) with protection of clinical decision-making conducted in accordance with appropriate standards.

Services should implement support mechanisms for staff who are subjects of complaints [15]. This includes supervision structures that normalise the experience of being challenged through formal complaints, contextualising such complaints as often reflecting patient distress about outcomes rather than as verdicts on clinical competence. Investigation teams may benefit from training that acknowledges rhetorical dimensions of complaints whilst maintaining focus on substantive allegations. This includes capacity to distinguish between emotional intensity or structural sophistication of complaints and validity of factual claims, whilst recognising that sophisticated rhetoric does not invalidate underlying concerns.

Services must also attend to communication quality throughout assessment processes. If patients frequently report feeling unheard or dismissed [4,5], this may indicate opportunities for improvement in how diagnostic decisions are explained, how differential diagnoses are communicated, and how individuals who do not meet diagnostic criteria are supported in addressing their difficulties through alternative means. Clear communication about the purpose of assessment, the distinction between clinical criteria and subjective symptom experience, and the reasoning behind diagnostic decisions may reduce some complaints rooted in misunderstanding of process.

Finally, whilst protecting clinical integrity is essential, services must remain sensitive to historical context. Individuals with neurodevelopmental conditions and other marginalised groups have historically faced dismissal in healthcare settings [46,47,48]. The challenge for contemporary services is to maintain boundaries of evidence-based assessment whilst ensuring that non-diagnosis is communicated with empathy and clarity that minimises distress. Achieving this balance requires understanding complaints not solely as risk management concerns but as complex signals of tensions between clinical protocols, patient expectations, and genuine unmet need in contexts of constrained resources and diagnostic complexity.

5. Limitations

The author’s dual role as both Clinical Lead for the examined service and as researcher conducting this analysis creates inherent epistemological tensions that require explicit acknowledgement and methodological response. This section addresses how organisational position has potentially shaped data interpretation, what safeguards were implemented to mitigate bias, and what alternative interpretations might emerge from different positional standpoints.

As Clinical Lead, the first author bears direct organisational accountability for service quality, complaint management, and clinician support. This creates multiple potential sources of interpretive bias. First, there exists an organisational incentive to interpret complaints as instrumentally motivated rather than as reflecting genuine service failures, particularly given that all complaints in this dataset were determined to be not upheld overall. Whilst independent investigation provided external validation of these determinations, the preponderance of not-upheld outcomes simultaneously creates circumstances wherein the insider-researcher may be predisposed to view complainant rhetoric as strategic pressure rather than legitimate advocacy when navigating organisational gatekeeping.

Second, the insider role provides intimate familiarity with service processes, clinical decision-making frameworks, and the operational pressures facing clinicians. This contextual knowledge enables nuanced interpretation of how complaints relate to actual clinical practice but simultaneously risks dismissing patient perspectives that challenge organisational norms. A clinician reading a complaint alleging that “the assessor had already decided before I arrived” may interpret this as a misunderstanding of how clinical formulation operates, whereas a patient advocate might interpret identical language as evidence of inadequate clinician openness to divergent presentations.

Third, organisational loyalty and collegial relationships with clinicians whose competence is questioned within complaints may unconsciously influence interpretation. Whilst systematic coding was applied before consulting investigation outcomes, the selection of which quotations to present as exemplars, how to characterise overall patterns, and what implications to draw for governance inevitably reflects the organisational perspective from which the analysis is conducted.

The insider-researcher position also shapes what questions are asked and how findings are framed. This analysis focuses on how complaints create organisational pressure and what implications emerge for protecting clinical decision-making. A researcher positioned outside the institution, or one embedded within patient advocacy networks, might instead focus on (a) what complaint rhetoric reveals about patient experiences of dismissal, (b) what barriers prevent patients from having their concerns heard through standard channels, or (c) how services might reduce the need for formal complaints through enhanced communication during assessment processes.

Finally, the interpretation of complaint volumes as potentially reflecting “a gap between public expectation and clinical criteria” rather than service quality problems itself reflects an organisational perspective. An alternative interpretation might hold that sustained complaint volumes could indicate service-level problems rather than mismatched expectations. If a service generates complaints at rates significantly above comparable services (though such comparison data are not available in this dataset), or if patients consistently report feeling unheard and dismissed, these patterns may indicate failures in communication, overly restrictive application of diagnostic criteria, or inadequate responsiveness to the legitimate needs of a population seeking support for genuine difficulties.

5.1. Methodological Boundaries

The study utilises a single-centre design within a specific NHS Trust. Consequently, the findings may reflect local demographic factors, waiting list durations, or specific community dynamics which limit transferability to other geographical regions or private healthcare settings.

The decision to employ binary coding (present/absent) for quantitative analysis inherently flattens the data. A complaint containing a single reference to emotional harm is coded identically to one containing numerous such references. Whilst this allows for prevalence reporting and facilitates reliable inter-rater agreement, it does not capture the intensity or density of rhetorical deployment. Future research employing frequency counts or rhetorical density measures could examine whether intensity of rhetorical deployment relates to investigation complexity or service impact.

5.2. The “Non-Upheld” Constraint

The focus on not-upheld complaint elements creates a specific boundary condition. This study analyses the rhetoric of grievances where investigation found no procedural or clinical breach. It remains unknown whether the same rhetorical strategies are present in complaints where genuine service failure occurred. A limitation of this dataset is that it does not permit testing the hypothesis that rhetorical sophistication and substantive merit are inversely related. Future research comparing the rhetorical architecture of upheld versus not-upheld complaints is required to test this relationship empirically.

5.3. The Intent-Text Gap

There is a fundamental epistemic asymmetry in analysing de-identified texts without patient involvement. This research analyses performance, not cognition. The study observes the rhetorical text but cannot access the intent of the author. The high prevalence of Emotional Impact Description (91.7%) acts as a textual marker in complaints, but the study cannot distinguish between distress arising from the assessment process itself versus distress arising from non-diagnosis or unmet expectations. Similarly, the study identifies sophisticated rhetorical construction but cannot determine whether this arises from conscious strategic calculation, unconscious modelling of online templates, access to advocacy support, or genuine desperation in navigating unresponsive systems. The absence of complainant interviews means the study captures the organisational reception of the text, not the patient experience of constructing it.

5.4. Ethical Risks and Misapplication

There is an ethical risk that these findings could be used to dismiss valid concerns. The identification of recurring rhetorical patterns carries the danger that complaint handlers might dismiss highly structured or emotionally charged complaints as “merely rhetorical” rather than investigating the substantive allegations thoroughly. Rhetorical sophistication does not preclude genuine grievance. A patient may deploy multiple rhetorical strategies and have experienced a genuine service failure. These findings are intended to assist in understanding the complexity of complaint interactions and the pressures they may create for clinical decision-making, not to function as a screening tool to invalidate patient concerns or justify reduced investigative rigour. Services must continue to investigate all complaints thoroughly and impartially, regardless of their rhetorical style or sophistication.

6. Conclusions

6.1. Summary of Findings

This study provides the first empirical rhetorical analysis of formal complaints within a neurodevelopmental assessment service. The examination of 48 complaints determined to be not upheld overall reveals a noteworthy pattern: whilst independent investigations confirmed that the service adhered to clinical and administrative standards, the rhetorical construction of complaints demonstrates that complainants experienced these interactions as failures of care.

The analysis establishes that multi-dimensional rhetorical construction is common in this domain. The frequent co-occurrence of Causal Attribution, Emotional Impact Description, and Procedural Challenge (present together in 89.6% of complaints) indicates that complaints employ multiple persuasive strategies simultaneously. Complainants commonly frame assessment outcomes in terms of medical harm (91.7% included Emotional Impact Description), procedural breach (87.5% included Procedural Challenge), and challenges to clinician competence (81.2% included Clinician Competence Questioning). A substantial majority (81.2%) included explicit demands for specific outcomes such as reassessment, acceptance of private diagnoses, or prescribing of particular medications.

6.2. Contribution to Understanding Healthcare Complaints

This research contributes empirical data on the rhetorical construction of complaints in a domain characterised by diagnostic complexity, capacity constraints, and elevated public engagement. The findings demonstrate that complaints in neurodevelopmental services employ recognisable rhetorical strategies consistent with classical persuasive frameworks, combining emotional appeals (pathos), rational or procedural arguments (logos), and challenges to professional credibility (ethos).

The high prevalence of Specific Outcome Demands (81.2%) suggests that many complainants approach the complaints mechanism with clear expectations regarding desired remedies, rather than seeking only investigation or explanation. Whether this reflects instrumental use of complaints processes, necessary assertiveness when navigating gatekeeping systems, or both, cannot be determined from textual analysis alone. The findings are consistent with the broader literature documenting a shift toward consumerist healthcare models in which patients position themselves as informed consumers with clear service expectations.

The pattern of Clinician Competence Questioning appearing in over 80% of not-upheld complaints warrants particular attention. This may indicate genuine concerns about assessment quality, frustration with outcomes perceived as predetermined, or strategic attempts to undermine professional authority. Alternatively, it may reflect defensive responses from services dismissing legitimate patient concerns as attacks on competence. The study design does not permit adjudication between these interpretations, but the prevalence of this pattern suggests tensions in the patient–clinician relationship that merit further investigation.

6.3. Implications for Healthcare Governance and Practice

For healthcare organisations managing neurodevelopmental services, these findings highlight the complex pressures operating within complaint processes. The sophisticated rhetorical construction observed across the majority of complaints, combined with the preponderance of not-upheld outcomes, suggests a disconnect between patient expectations and clinical decision-making frameworks that cannot be resolved through investigation alone.

Governance frameworks should recognise that complaint volume and rhetorical intensity in diagnostically complex areas may reflect systemic tensions rather than straightforward quality failures. However, this recognition must be balanced against the possibility that sustained complaint volumes indicate genuine problems with communication, accessibility, or responsiveness. Services should avoid using rhetorical sophistication as grounds for dismissing complaints without thorough investigation, whilst simultaneously supporting clinicians in maintaining evidence-based practice when faced with pressure for specific outcomes.

Investigation teams may benefit from training that acknowledges the rhetorical dimensions of complaints whilst maintaining focus on substantive allegations. This includes distinguishing between the emotional intensity or structural sophistication of a complaint and the validity of its factual claims. Critically, such training must emphasise that sophisticated rhetoric does not invalidate underlying distress or concern, and that multiple rhetorical strategies may co-exist with legitimate grievances.

Services should consider whether current communication practices adequately explain diagnostic decisions, thresholds, and the distinction between clinical criteria and subjective experience of symptoms. If patients consistently feel unheard or dismissed, this may indicate opportunities for improving shared decision-making processes, transparency about diagnostic reasoning, and support for individuals who do not meet diagnostic thresholds but experience genuine difficulties.

6.4. Future Research Directions

Several avenues for future research emerge from this study. First, comparative analysis of rhetorical patterns in upheld versus not-upheld complaints would clarify whether the strategies identified here are specific to unsubstantiated grievances or represent general features of complaint construction. Second, longitudinal research examining complaint patterns over time could determine whether rhetorical sophistication is increasing, as suggested by the consumerist healthcare literature, or whether current patterns are stable. Third, research incorporating complainant perspectives through interviews or surveys would provide essential balance to this text-based analysis, enabling understanding of how complainants construct and experience the complaint process.

Fourth, comparative research across different neurodevelopmental services, including private providers and services in different healthcare systems, would clarify whether the patterns observed here reflect local service characteristics, broader systemic pressures, or features inherent to contested diagnostic contexts.

Fifth, comparative cross-service rhetorical analysis of examining complaint patterns in neurodevelopmental assessment alongside other contested diagnostic contexts such as gender identity services, chronic fatigue syndrome, and Lyme disease would clarify whether the rhetorical patterns identified here reflect features specific to neurodevelopmental assessment or constitute a broader genre of organisational challenge in domains characterised by diagnostic contestability.

Finally, intervention studies examining the effects of enhanced communication, clearer explanation of diagnostic reasoning, or alternative support pathways for individuals not meeting diagnostic criteria could identify approaches that reduce the need for formal complaints whilst maintaining clinical rigour.

6.5. Concluding Reflections

This study has documented recurring rhetorical patterns in complaints about neurodevelopmental assessments where investigation found no procedural or clinical breach. These patterns of combining emotional impact description, procedural challenge, competence questioning, and specific outcome demands create complex pressures for services navigating the intersection of clinical standards, capacity constraints, and patient expectations.

The analysis necessarily captures only one dimension of a multifaceted phenomenon. Rhetorical construction reflects communicative strategy, but the study cannot access the experiences, frustrations, or genuine distress that may motivate such construction. The sophisticated rhetoric observed may represent instrumental pressure, necessary advocacy in unresponsive systems, desperation from individuals experiencing genuine difficulties, or combinations of all three.

Healthcare systems face a fundamental tension in this domain: maintaining rigorous, evidence-based diagnostic standards whilst remaining responsive to patient voice and genuinely addressing unmet need. High complaint volumes combined with low upheld rates create a challenging environment for both patients and clinicians. Patients may experience the assessment process as dismissive or gatekeeping, whilst clinicians face pressure to meet expectations that diverge from clinical criteria.

The complaints mechanism serves essential accountability and learning functions, but it cannot resolve tensions rooted in resource constraints, diagnostic complexity, and divergent understandings of what assessment services can and should provide. Addressing these tensions requires systemic responses beyond complaint handling: investment in capacity, clarity about service scope and diagnostic thresholds, enhanced communication throughout assessment pathways, and alternative forms of support for individuals experiencing difficulties regardless of whether they meet diagnostic criteria.

Ultimately, the rhetorical patterns documented here should prompt reflection on both sides of the patient–service relationship. Services must ensure they are genuinely hearing patient concerns, communicating decisions clearly, and not dismissing legitimate grievances because of their rhetorical construction. Equally, healthcare governance must support clinicians in maintaining evidence-based practice under pressure, ensuring that diagnostic decisions remain grounded in clinical criteria rather than determined by rhetorical force. Finding this balance requires ongoing dialogue, transparency, and mutual recognition of the complexities inherent in neurodevelopmental assessment in contemporary healthcare contexts.