Diversity data in our fitness to practise process: a report on our updated research and analysis
Diversity data in our fitness to practise process: a report on our updated research and analysis
Introduction
Equality, diversity and inclusion (EDI) is central to our work as an effective regulator. It is at the core of everything we do and is a key part of our values and behaviours. EDI shapes how we:
make decisions
develop policy
engage with those interested in our work
deliver our regulatory functions
EDI is embedded across all our business objectives to ensure fairness, inclusion and public protection remain deeply connected in everything we do.
We have therefore continued to conduct further analysis into our diversity data to give us great insight into our fitness to practise process to allow us to:
identify disparities
seek to understand them
drive improvements
On this page we share a summary of our findings from our most recent research and analysis.
Further information about our EDI commitments can be read in our EDI and co-production action plan 2026 to 2027.
Summary
Context
In September 2023, we examined diversity data in our fitness to practise processes and published our findings on our website: initial analysis of the diversity of social workers in relation to our fitness to practise process. We found that some groups were overrepresented in the referrals we received and at key stages of our fitness to practise process, including from triage to investigation, and from case examination to hearing.
Overrepresented groups included:
Age: those aged 40 and over
Gender identity: males
Ethnicity: people of Black/African/Caribbean/Black British ethnicity
Throughout 2024 we took steps to expand our analysis. We examined an additional year of data to see whether the initial overrepresentations were still present. In October 2024 we published an update to our analysis of diversity data in our fitness to practise process, which showed overrepresentations for the above groups were still present in referrals and progression through the fitness to practise process.
Aims of our research and analysis
Our analysis builds on our previous work. We have used more sophisticated statistical analysis (logistic regression) in this analysis. This examines a range of factors at the same time and identifies which factors are associated with progression through the fitness to practise process once other relevant factors have been considered. The analysis included characteristics of:
Social workers: age, gender identity, ethnicity, nationality, religion, sexual orientation, sex same as at birth, and disability
Employers of social workers: sector type, practice area, and geographical region
Referrers: referral type
Cases: concern type
This has allowed us to consider other factors in addition to EDI related factors, to explore how some factors drive progression compared to others.
Our analysis identifies factors associated with progression through the fitness to practise process. It does not explain why these associations occur. In our analysis we try to explain some of our findings alongside other evidence however more research is needed to understand the implications for EDI.
See our technical report for more detail on our approach, methods, and results.
Key findings
We received just under 7,000 fitness to practise referrals between the dates of 2 December 2019 and 31 March 2024. We analysed this data set to identify factors associated with case progression through the fitness to practise process. In particular, we examined whether previously observed patterns relating to ethnicity, gender identity, and age remained once other relevant factors were taken into account.
The analysis found that social workers’ personal characteristics were still associated with progression through the fitness to practise process but with some variation across the different stages and different characteristics. Personal characteristics were not associated with progression at the pre-triage stage. However, at the triage and case examiner stages, progression was associated with ethnicity. Case progression was associated with gender identity only at the triage stage, and age only at the case examiner stage.
Across the fitness to practise process, the factors most strongly and consistently associated with case progression were referral source and the nature of the concerns raised. At the same time, differences associated with ethnicity remained evident at the triage and case examiner stages even after taking account of these and other factors. Together, these findings provide a more complete understanding of the factors associated with progression and can inform our next steps understanding EDI within the fitness to practise process.
Pre-triage stage
Progression at the pre-triage stage was most strongly associated with the referral source and the nature of the concerns raised, and then by geographical region. Social worker characteristics were not associated with progression at this stage.
Pre-triage is the first decision point when cases are referred to Social Work England. Progression requires a decision-maker to confirm only that the case is about a social worker and relates to at least one statutory ground of impairment.
Social worker characteristics
Once other factors were taken into account, there were no differences in progression by age, gender identity or ethnicity at this stage of the fitness to practise process.
Other factors
Cases involving social workers employed in the regions of East of England, South East, and Yorkshire and Humber were between one-and-a-half times and twice as likely to progress than in the East Midlands region.
Cases referred by employers were more than 3-and-a-half-times as likely to progress than referrals from members if the public. Referrals from children and family services were also half more likely to progress than referrals from adult services.
Cases involving concerns about convictions were the most likely to progress, almost 5-and-a-half times as likely as cases involving poor or inappropriate communication. Cases involving record keeping and safeguarding concerns were also more likely to progress.
Triage stage
At the triage stage, progression was still most strongly associated with referral source and nature of concerns. Social worker ethnicity, gender, and religion were also associated with likelihood of progression at this stage.
Progressing through triage initiates a full investigation if a senior decision-maker determines there are 'reasonable grounds' to investigate whether a social worker’s fitness to practise is impaired.
Social worker characteristics
Differences by ethnicity, gender and religion emerged at the triage stage, but not by age. Cases involving male compared to female social workers, and cases involving non-Christian compared to Christian social workers were about a third more likely to progress. Cases involving Black social workers were almost one-and-a-half times as likely to progress than cases involving White social workers.
Other factors
Variation in social workers’ employment region observed at pre-triage disappeared at this stage. The source of referral became an even stronger predictor of progression at triage. Employer referrals were 19-and-a-half times as likely to progress than member of the public referrals. Referrals from other agencies and self-referrals were also more likely to progress.
The nature of concerns raised was strongly associated with progression. Cases involving convictions, sexual misconduct, safeguarding concerns, professional boundaries, record-keeping concerns and court proceedings were more likely to progress than cases involving poor or inappropriate communication.
Case examiner stage
At the case examiner stage, referral source continued to be the strongest predictor of progression, followed by region and concern type. Social worker ethnicity also continued to be associated with progression at this stage, but not religion. Age became associated with progression but only for some age groups.
At this stage, case examiners review the evidence gathered during the investigation and decide whether there is a realistic prospect that a social worker’s fitness to practise is impaired. They may refer the case to a hearing, propose an outcome without a hearing, or close the case.
Social worker characteristics
The difference by ethnicity became more pronounced at the case examiner stage. Cases involving Black social workers were around twice as likely to progress compared to cases involving White social workers, even after taking account of other case and participant characteristics.
Gender and religious differences observed at the triage stage were no longer evident. Age was not consistently associated with progression. However, cases involving social workers aged 40 to 49 were more likely to progress than cases involving social workers aged 30 to 39. No other significant age differences were identified.
Other factors
Variation in social workers’ employment region re-emerged at this stage. Cases from the North West, London, and the South East were between twice and 3 times as likely to progress than cases from the East Midlands.
Referrals from employers and other professional sources continued to be more likely to progress than referrals from members of the public. Employer referrals remained particularly likely to progress, 13-and-a-half-times as likely.
The pattern of concern types changed at this stage. Cases involving dishonesty were the most likely to progress, twice as likely compared to cases involving poor or inappropriate communication. In contrast, safeguarding concerns were less likely to progress, despite being associated with increased progression at earlier stages.
Possible explanations for these findings
The analysis identifies factors associated with progression through the fitness to practise process but does not explain why these associations occur. Some of the findings can, however, be considered alongside other evidence and research.
Social worker characteristics
The analysis found that cases involving Black social workers were more likely to progress at the triage and case examiner stages, even after taking account of a wide range of other factors. The reasons for this are not clear from the available data. Though discrimination, including unconscious bias, are potential explanations for these differences, there may be other reasons.
The analysis found weaker and less consistent evidence for differences by age and gender identity once other factors were taken into account. Only the 40 to 49 age group was more likely to progress than the 30 to 39 age group at the case examiner stage, while males were only more likely to progress than female social workers at the triage stage. This suggests that some of the differences identified in our earlier analysis may be linked to the types of cases being referred, rather than to age or gender identity alone.
Other factors
Cases referred by employers and other professional sources were consistently more likely to progress than referrals from members of the public. One possible explanation is that employers and other professionals are more familiar with the legal and regulatory framework governing fitness to practise. Employers therefore may be more likely to make referrals that meet the relevant thresholds and to provide evidence that enables decision-makers to assess the concerns being raised. This interpretation is consistent with the finding that referrals from professional sources become increasingly overrepresented as cases move through the fitness to practise process.
Cases involving convictions, safeguarding concerns, professional boundaries, sexual misconduct and dishonesty were more likely to progress than cases involving poor or inappropriate communication. Our earlier research into the role of seriousness in fitness to practise proceedings found broad agreement that these types of concerns are generally viewed as serious. Although the seriousness research also highlights that judgements about seriousness can be subjective and highly dependent on context.
The higher progression rates for record-keeping concerns may also reflect the seriousness of some of these cases, which may involve confidentiality breaches or inappropriate access to records. The recent research found broad agreement that such concerns are generally viewed as serious within fitness to practise proceedings. Further analysis would be needed to determine whether this explains the pattern observed here.
What have we done so far?
Our findings show that referrals are more likely to progress through our fitness to practise processes when they involve social workers with specific personal characteristics. While these findings identify patterns of disproportionality, they do not explain why those patterns exist. At this stage, we cannot determine whether the differences identified are the result of inequality, bias, broader workforce factors, or other drivers within the social work system. However, where the evidence highlights areas of potential concern, we believe it is important to take proportionate action now rather than wait for complete certainty. This includes reviewing our own processes, sharing data and insights, supporting more consistent referral decision making, and working with partners across the sector to promote fairness and reduce the risk of inequitable outcomes.
At the same time, we recognise that meaningful and sustainable change must be informed by a robust understanding of the underlying causes of disproportionality. We will therefore continue to invest in research, data analysis and stakeholder engagement. This will build our understanding of why some groups of social workers are over-represented in referrals and throughout the fitness to practise process. Our response will be iterative and evolving as new evidence emerges. We will continue to work collaboratively with other regulators, employers, representative bodies and the wider social work sector to identify and implement the actions most likely to have a positive impact.
Below are some of the activities that we have already undertaken in this area.
Professional Standards Authority - Standard 3
The Professional Standards Authority for Health and Social Care’s (PSA) annual monitoring review of our performance in 2024 to 2025 had a focus on our EDI work. The PSA acknowledged the following improvements and positive practice:
our efforts to improve board oversight of our work in relation to equality, diversity and inclusion
the development of our fair referral principles
workshops we have held with employers to encourage consideration of diversity data
the ongoing support and training relation to equality, diversity and inclusion that we have provided to staff within our fitness to practise teams
Advancing Workforce Race Equity in Health and Social Care: Shared Principles Across Regulators
We have pledged our commitment to the ‘Advancing Workforce Race Equity in Health and Social Care: Shared Principles Across Regulators’. These 9 principles have been developed collaboratively by health and social care regulators to create aligned regulatory momentum in tackling racism experienced by the workforce. They align with our public sector duty obligations under the Equality Act 2010 and our broader duty to ensure fair, transparent, and inclusive regulation. It also reflects our published outcome 4 of the PSA’s Standard 3 evidence matrix, which states that "the regulator engages with and influences others to advance EDI issues and reduce unfair differential outcomes."
Strengthening workforce race equity supports improved workforce wellbeing and performance, which in turn benefits the public we serve. We are already meeting the expectations set out in the 9 shared principles through our existing EDI work, regulatory strategy, and organisational culture initiatives. This endorsement reflects our continued commitment to driving meaningful, lasting change in this area.
Social Care Workforce Race Equality Standard (SC-WRES)
The SC-WRES is intended to support organisations to make progress towards race equality by collecting and analysing workforce data across 9 key metrics, one of which includes fitness to practise. Organisations are supported to develop action plans that drive meaningful change and measurable improvement. We continue to collaborate and share our data and insights with Skills for Care to support the SC-WRES and to assist participating organisations to review their own data and support fair referrals.
Fair referral principles
As acknowledged by the PSA, we have published a set of fair referral principles, which are designed to help employers of social workers decide when to raise a fitness to practise concern with us. By encouraging employers to apply consistent, equitable principles when making referrals, the principles aim to reduce disproportionality in the system and ensure that referrals made to Social Work England are applied fairly across the social work workforce.
To help embed the principles, we have developed a programme of engagement with employers to focus on the consistent application of these principles. Further information about our engagement or application of the principles with employers can be sought by contacting our regional engagement leads.
Single Point of Contact network
Social Work England has a Single Point of Contact at every local authority in England and other organisations employing social workers. These contacts act as a central link between employers and Social Work England. We have developed and embedded the fair referral principles through ongoing work with employers and the Single Point of Contact network. They demonstrate how local decision makers can ensure the right referrals are made at the right time, in a fair and proportionate way. They also provide further support on how employers can address concerns locally where appropriate and maintain a clear and accessible route for concerns requiring regulatory intervention.
East Region pilot
We have delivered a joint pilot with the East Region to understand the lived experience of Black, male, social workers and leaders in relation to fitness to practise. The aim of this engagement is to support a reflection on our current regulatory processes and contribute towards early resolution.
During Social Work Week 2026, a session exploring the Eastern Region’s collective journey to embedding anti-racist practice was delivered which reflected on the impact of the pilot and key lessons learned. As this is a pilot, we have the ambition that these findings could be expanded further into other regions through the Principal Social Worker Networks.
Anti-racism training for decision makers
We are developing a dedicated training programme for all Social Work England staff and partners involved in fitness to practise processes. While anti-racism is a central focus, the programme recognises that people’s experiences are shaped by multiple, overlapping characteristics such as race, gender, and disability.
Independent Review of Social Work Professional Regulation
The Independent Review of Social Work Professional Regulation acknowledged our achievements and identified priority recommendations for Social Work England.
It welcomed our commitment to analysing and better understanding diversity data within fitness to practise referrals. It also supported our efforts to establish a data oversight group with other health and social care regulators which provides independent challenge and assurance on the analytical methods and conclusions being drawn.
While the review acknowledged key areas which require our improvement we continue to be confident in our decision making and there were no concerns about fitness to practise decision making identified in the review.
Next steps
Understanding patterns in fitness to practise case progression
Our analysis shows that there is overrepresentation of some groups in our fitness to practise processes. This latest research controlled the relationships between factors to confirm that some overrepresented groups remain more likely to progress through our fitness to practise processes, even when other factors are considered. A potential reason is that there is bias in our processes, but we need to carry out further research to explore and test alternative hypotheses. For example, we know that risk is a key driver in case progression, so we should explore whether case seriousness is part of the explanation for any of the overrepresented groups.
We will, therefore, carry out a qualitative piece of work to help us to learn more about the patterns of progression through our process, and the reasons. This would involve testing if any of the patterns can be explained by discrimination and understanding and testing alternative hypotheses.
Development of an EDI focused fitness to practise data report
We will create a new data report to regularly report on demographic trends across the register and fitness to practise process. This will include hearings data which we haven’t included to date due to there being too few outcomes to provide viable data. The report will improve visibility of our EDI data, support future analysis, track changes over time, and help identify emerging issues.
An internal review of fitness to practise processes and policies
While an initial, small-scale audit did not identify any evidence of bias within our processes, we acknowledge that the sample size was not large enough to allow us to make reliable assumptions. To be confident that our own processes are not contributing towards overrepresentation, we will undertake a comprehensive review of internal decision-making guidance and associated equality impact assessments. This review may help shape whether changes should be made in key areas.
Decision making guidance
We will review and strengthen our guidance on equality, diversity and inclusion for staff and decision makers involved in fitness to practise work, ensuring that relevant EDI considerations are appropriately reflected throughout the decision-making process.
Lord Mann review
We’re working with the Department for Health and Social Care and other regulators, including the Professional Standards Authority, to consider the findings from the Lord Mann review and what steps we could or should take to address them.
Decision review group
Social Work England holds a monthly decision review group, which is an internal quality-assurance body that:
evaluates a sample of concluded fitness to practise decisions
identifies broader operational trends
drives organisational learning
Taking learning from our operation of the group to date, we have reviewed our approach and taken the decision to split the group to create:
A quality assurance group, which will review decisions made at different points in the fitness to practise process.
A thematic review group will explore themes identified by the quality assurance group and agree any actions to address them.
Conclusion
The analysis described in this report provides a more detailed understanding of the factors associated with progression through the fitness to practise process. It shows that referral source and the nature of the concerns raised are most strongly and consistently associated with progression through the process. Ethnicity, however, is still significantly associated with case progression, as are gender identity and age, although with less consistency.
It is important to remain cautious in drawing conclusions from this analysis. While it identifies factors associated with progression through the fitness to practise process, it does not explain why those associations occur. The findings should therefore not be interpreted as evidence of causation. However, the persistence of differences by ethnicity at later stages of the process warrants further exploration. We remain committed to improving our understanding of these findings and taking action where appropriate to support a fair, proportionate and inclusive fitness to practise process.
This work forms part of our wider programme of equality, diversity and inclusion activity in relation to fitness to practise detailed in the “what we have done so far” section of this report. This updated analysis strengthens the evidence base for that work by providing a more detailed understanding of where disparities may emerge and which factors appear to be most strongly associated with progression.
By improving our understanding of the factors associated with progression, we are better placed to identify where further work is needed. This insight will empower us to ensure the fitness to practise process is fair, proportionate and consistent with our commitment to EDI.
Diversity data in our fitness to practise process: technical report
This technical report describes the approach we took to analyse the impact of social worker, referrer, and case characteristics on progression through our fitness to practise process. This report is designed to be read in conjunction with our summary report which provides context and interpretation of our findings.
This report will provide the information about the data we used and the analysis we conducted, and is split into the following sections:
Scope – The purpose of the research and the research questions we were seeking to address.
Data – The data we used for the analysis, including the source, the size, and the time period.
Data preparation – The steps we took to clean and transform our data, and limitations of the prepared data set.
Analysis – Details about the multivariate analysis we conducted, including the approach and key results.
Quality assurance – The processes we followed to check and quality assure our methods and results.
1. Scope
1.1. Purpose of the research
The research was intended to build on our previously published analysis which found that particular groups of participants were:
overrepresented in the referrals we received
had higher rates of progression through our fitness to practise process – between triage and investigation and between case examination and hearings
To build on this previous analysis, we set out to use more sophisticated statistical analysis (logistic regression) to examine a range of factors at the same time and identify which factors are most strongly associated with progression through the fitness to practise process. In particular, this approach would help us explore if the social worker characteristics that appear to be overrepresented in the process continue to be significant when other factors, like referrer and case characteristics, are taken into account.
This research contributes to our wider research on equality, diversity, and inclusion (EDI). While this analysis cannot evidence causation or provide explanations for patterns in the data, identifying factors associated with case progression can help towards understanding whether there may be discrimination in the fitness to practise process, and inform the next phase of our research.
1.2. Our aims
Our aims were:
To examine a wider range of factors that could reasonably be associated with case progression through the fitness to practise process. This included the social worker characteristics previously seen to be overrepresented in our previous analysis (age, gender identity, and ethnicity), as well as employer, referrer, and case characteristics.
To specifically look at factors associated with case progression at 3 points in the fitness to practise process. This was so we could see, if and how, factors associated with case progression vary between the pre-triage, triage, and case examiner stages.
To use a statistical method (logistic regression) that can take account of multiple factors at once to identify which remain associated with case progression once other relevant factors are taken into account.
2. Data
2.1. Sample details
The data set used for this analysis was case data for all the referrals that we received between 2 December 2019 and 31 March 2024. We used the same time period as our previously published data and analysis. The data was extracted from Social Work England’s case management system and included 6,789 referrals.
This fitness to practise data set reduced in size across the stages of the analysis as cases dropped out of the process because they did not meet the threshold for progression. Additionally, there were cases at each stage that remained under investigation, without a decision at the time of analysis, and these are excluded too. Taking this into account, below are the number of fitness to practise cases included at the 3 stages of this analysis:
Pre-triage (n = 6,585 cases)
Triage (n = 5,797 cases)
Case examination (n = 1,423 cases)
In this analysis, we did not look at factors associated with case progression at the hearing stage due to there being too few case left in the data set to conduct our regression analysis – subsample sizes for the different variables would be too small for statistically significant results.
2.2. Variables
For our multivariate analysis, using logistic regression, the dependent variable (DV) for models was case progression. The DV was a binary variable – cases either passed or did not pass at each of the fitness to practise stages.
We theoretically selected a range of variables from the fitness to practise data set to test as the independent variables (IV). These IVs represented a range of social worker characteristics, as well as employer, referrer, employer, and characteristics. A full breakdown of the IVs, including frequencies, can be seen in Section 4.2 below.
Social workers characteristics
The variables selected to represent the personal characteristics of social workers are as follows:
Age
Gender identity [1]
Sex registered at birth where different to current gender identity
Disability
Sexual orientation
Ethnic group
Religion or belief
Nationality
[Note 1] Historical diversity monitoring data collected by Social Work England asked registrants about their gender identity. This report analyses those data as collected and therefore refers to gender identity throughout.
Employer characteristics
The social worker’s employer characteristics include the following:
Employer type – This is the most recently recorded employer on their social worker record.
Sector type – The sector type of the social worker, based on their current or most recent recorded role.
Area of practice – An alternative grouping used to classify the sector type of the social worker’s current or most recent recorded role.
Region - The region in which the social worker is employed, based on the social worker’s current or most recent role.
Referrer characteristics
Variables represented in each of the models for referrer characteristics include:
Source of referral – This is a grouped category of the person or organisation that made a referral to us.
Case characteristics
The only relevant case characteristics represented in each of our models as an IV was concern type. We first checked for multicollinearity with the DV and confirmed there was sufficient variation for this to be an IV.
Concern type – This is a categorical variable using high level descriptive categories for the concern type associated with a fitness to practise case. The concern type is determined by Social Work England case workers upon referral.
The concern type is not the same as seriousness and there was no objective measure for case seriousness in the fitness to practise data. While cases are allocated a risk score, this was not considered an IV in our models because the risk score determines whether case progress or not and is, unsurprisingly, multicollinear with the DV. The risk score can change across the fitness to practise process based on the available evidence and this how cases can progress or be closed between the three stages examined in our models.
3. Data protection
3.1. Data set creation
We complied our data set from 2 MySQL databases and linked them according to unique identifiers using Power Automate scripting in the Power BI business analytics platform. This provided a single data set containing all fitness to practise cases referred within our specified time period. This data set contained all the variables we needed for our analysis, including the range of social worker, referrer, and case characteristics list above.
From this linked data set, we extracted data needed for the 3 fitness to practise stages that we planned to examine using logistic regression. These 3 data sets were extracted as .csv files for loading into the statistical software packages – we used both SPSS and R.
3.2. Data cleansing and variable transformation
We removed duplicate entries for individual social workers, which is necessary to meet the assumption of independence of observations within the data set. Where social workers had multiple fitness to practise referrals, we retained the chronologically earliest referral, an approach used in other organisational data presentation.
We cleansed the master data set of all the fitness to practise cases, removing records created in error, and other invalid elements of the data set.
We assessed the extent of missing data for each of the IVs and re-categorised this missing data as “Unknown” across variables. This approach maintained as many cases as possible within the data set.
We transformed IVs so that all would be categorical variables and collapsed categories for some existing categorical variables for use in the models:
Age was transformed from a continuous into a categorical variable, “binning” ages into 10-year groupings, with the lower range being “29 and under”. This matched the approach to reporting age ranges taken in other organisational data reporting.
Ethnicity was grouped into 4 categories – White, Black, Other, and Unknown. This helped us test whether the previous finding about Black social workers being overrepresented in fitness to practise process can be explained by ethnicity alone or may be due to other relevant factors.
For variables with data categories comprising small numbers of cases, we aggregated these categories into a single category named “Other” for some variables, such as nationality. For concern type, we used the category name “Small numbers category” to differentiate it from “Other” which was an existing category name for this variable.
It is possible for case workers to select up to 3 concern types. For our final regression models, we included only the primary concern type. Before making this decision, we first checked that that retaining multiple concern types in the model did not significantly change the results
3.3. Limitations of the data set
There were some limitations with the data set used for our analysis:
The completeness of diversity data decreases at later stages of the fitness to practise process. Diversity data is available for around 92% of referred cases, but only around a third of cases at the hearing stage. This is largely because hearings often involve older cases, and the quality and completeness of diversity data have improved over time. As a result, the analysis may give greater weight to the characteristics and outcomes of more recent cases than older cases.
Some variables contained substantial levels of missing information. While these cases were retained within the analysis using an "unknown" category (as described above), a large proportion of missing data can make it more difficult to identify true differences between groups and may influence the results if the missing information is not random.
4. Analysis
4.1. Approach
We used binary logistic regression to examine which factors were most strongly associated with case progression at each stage of our fitness to practise process.
We chose logistic regression because it was the best fit for our largely categorical data set with binary outcomes for the dependent variable. Either cases passed a decision-making stage or they did not.
We chose variables based on our prior analyses in which we found some overrepresentations in our data. Using regression allowed us to examine the relationship between social worker’s characteristics, case information, and what drives progression through our processes.
We can use these findings to develop actions that focus on the factors that are most important in the cases that are referred to us, and the experience of those in our fitness to practise process.
As stated in our aims section, it is essential to interpret these results carefully. The relationships we observe may be influenced by other factors not measured or accounted for in our analyses, or by wider systemic factors.
4.2. Model design
We treated each stage of the fitness to practise process under examination as a separate model, with one model for each of the following 3 stages:
Pre-triage – The first decision point when cases are referred to Social Work England. Progression requires a decision-maker to confirm only that the case is about a social worker and relates to at least one statutory ground of impairment.
Triage – Progressing through triage initiates a full investigation if a senior decision-maker determines there are 'reasonable grounds' to investigate whether a social worker’s fitness to practise is impaired.
Case examiner – At this stage, case examiners review the evidence gathered during the investigation and decide whether there is a realistic prospect that a social worker’s fitness to practise is impaired. They may refer the case to a hearing, propose an outcome without a hearing, or close the case.
IV categories and frequencies for each stage model can be seen in Annexe 1 below.
4.3. Key results
The key findings from the logistic regressions conducted at each stage are below.
Pre-triage stage
Progression at the pre-triage stage was most strongly associated with the referral source and the nature of the concerns raised, and then by geographical region. Social worker characteristics were not associated with progression at this stage.
Triage stage
At the triage stage, progression was still most strongly associated with referral source and nature of concerns. Social worker ethnicity, gender identity, and religion also became associated with likelihood of progression at this stage.
Case examiner stage
At the case examiner stage, referral source continued to be the strongest predictor of progression, followed by region and concern type. Social worker ethnicity also continued to be associated with progression at this stage, but not religion. Age became associated with progression but only for some age groups.
Full regression outputs for each stage see Annexe 2 to 4 below.
5. Quality assurance
We conducted quality assurance across the research, including when extracting and creating the data set, and when conducting the logistic regression analysis:
When compiling our data set from 2 MySQL databases for diversity and fitness to practise case data, we checked the validity of each half of the data set independently, and then again when the extracts were combined into the master data set of all cases at all stages of fitness to practise.
When we extracted the data for each separate stage of fitness to practise for the 3 models, creating 3 separate data sets, we checked the validity of each against the master data set.
We conducted the regression analyses in both SPSS and R, completed by 2 analysts working together. Results were cross validated between each software package, ensuring findings were consistent.
We also obtained independent quality assurance and advice from a data oversight group: a cross-regulator support and discussion group formed after the publication of our earlier analysis of fitness to practise referrals and case progression. Members of this group included representatives of the General Medical Council, the Nursing and Midwifery Council and support from an independent academic advisor. The group provided support in quarterly meetings, and direct feedback on our methodological design choices, and interpretation of the regression outputs.
Annexe 1: Independent variables
Referrer type
Member of the public (reference category): 3,767 (57%)
Employer: 1,151 (17%)
Other agency: 164 (2%)
Other social worker: 149 (2%)
Self-referral: 505 (8%)
Unknown: 849 (13%)
Concern type
Poor or inappropriate communication (ref): 1,178 (18%)
Conviction: 88 (1%)
Court proceedings: 295 (4%)
Criminal proceedings/involvement: 161 (2%)
Dishonesty: 1,107 (17%)
GDPR/data protection: 92 (1%)
Health: 176 (3%)
Issue with social worker's place of work or service as a whole: 79 (1%)
Performance issues: 295 (4%)
Bias and/or prejudicial treatment: 103 (2%)
Professional boundaries: 339 (5%)
Record keeping: 218 (3%)
Safeguarding concerns: 803 (12%)
Sexual misconduct: 63 (1%)
Violent and abusive conduct: 161 (2%)
Unknown: 788 (12%)
Other: 550 (8%)
Small number category [2]: 89 (1%)
[note 2] Contains categories with fewer than 35 observations, aggregated for calculation in regression.
Employer
Local authority (ref): 4,354 (66%)
Children and Family Court Advisory and Support Service (Cafcass): 136 (2%)
Agency: 384 (6%)
NHS: 176 (3%)
Other: 398 (6%)
Unknown: 1,137 (17%)
Age group
29 and under: 1,682 (26%)
30 to 39: 649 (10%)
40 to 49 (ref): 1,856 (28%)
50 to 59: 1,826 (28%)
60 or over: 572 (9%)
Gender identity
Female (ref): 5,138 (78%)
Male: 1,438 (22%)
Unknown: 9 (0%)
Ethnicity
White ethnicities (ref): 4,095 (62%)
Black ethnicities: 1,119 (17%)
Other ethnicities (not White or Black): 763 (12%)
Unknown: 608 (9%)
Disability
No disability: 4,960 (75%)
Disability: 712 (11%)
Prefer not to say: 354 (5%)
Unknown: 559 (8%)
Sex same as at birth
Yes: 5,299 (80%)
Prefer not to say: 288 (4%)
Unknown: 573 (9%)
No: 425 (6%)
Sexual orientation
Heterosexual: 5,051 (77%)
Gay man: 109 (2%)
Gay woman: 129 (2%)
Bisexual: 147 (2%)
Prefer not to say: 536 (8%)
Unknown and Other: 613 (9%)
Religion
Christianity (ref): 2,694 (41%)
Other religion (not Christianity): 1,015 (15%)
No religion: 2,298 (35%)
Unknown: 578 (9%)
Nationality
British (ref): 6,015 (91%)
Other (not British) or Unknown: 570 (9%)
Region:
East Midlands (ref): 424 (6%)
East of England: 505 (8%)
London: 1,001 (15%)
North East: 313 (5%)
North West: 904 (14%)
South East: 840 (13%)
South West: 508 (8%)
West Midlands: 694 (11%)
Yorkshire and the Humber: 607 (9%)
Other or Unknown: 789 (12%)
Sector type
Child (ref): 4,308 (65%)
Adult: 1,475 (22%)
Other: 339 (5%)
Unknown: 463 (7%)
Area of practice
Specialist - children and families (ref): 3,866 (59%)
Specialist - adults: 896 (14%)
Disabilities: 236 (4%)
Hospital: 110 (2%)
Mental health: 503 (8%)
Crime and courts: 228 (3%)
Other: 283 (4%)
To note: Totals may not sum to 100% due to rounding.
Referrer type
Member of the public (reference category): 3,153 (61%)
Employer: 996 (19%)
Other agency: 148 (3%)
Other social worker: 131 (3%)
Self-referral: 379 (7%)
Unknown: 373 (7%)
Concern type
Poor or inappropriate communication (ref): 1,033 (20%)
Conviction: 85 (2%)
Court proceedings: 259 (5%)
Criminal proceedings/involvement: 99 (2%)
Dishonesty: 936 (18%)
GDPR/data protection: 45 (1%)
Health: 127 (2%)
Issue with social worker's place of work or service as a whole: 47 (1%)
Performance issues: 198 (4%)
Bias and/or prejudicial treatment: 79 (2%)
Professional boundaries: 288 (6%)
Record keeping: 214 (4%)
Safeguarding concerns: 689 (13%)
Sexual misconduct: 46 (1%)
Violent and abusive conduct: 152 (3%)
Unknown and other: 883 (17%)
Employer
Local authority (ref): 3,663 (71%)
Children and Family Court Advisory and Support Service (Cafcass): 106 (2%)
Agency: 322 (6%)
NHS: 145 (3%)
Other: 318 (6%)
Unknown: 626 (12%)
Age group when referred
29 and under: 516 (10%)
30 to 39: 1,296 (25%)
40 to 49 (ref): 1,449 (28%)
50 to 59: 1,469 (28%)
60 or over: 450 (9%)
Gender identity
Female (ref): 4,020 (78%)
Male: 1,151 (22%)
Unknown: 9 (0%)
Ethnicity
White ethnicities (ref): 3,155 (61%)
Black ethnicities: 889 (17%)
Other ethnicities (not White or Black): 610 (12%)
Unknown: 526 (10%)
Disability
No disability: 3,883 (75%)
Disability: 531 (10%)
Prefer not to say: 282 (5%)
Unknown: 484 (9%)
Sex same as at birth
Yes (ref): 4,129 (80%)
No: 321 (6%)
Prefer not to say: 234 (5%)
Unknown: 496 (10%)
Sexual orientation
Heterosexual: 3,937 (76%)
Gay man: 83 (2%)
Gay woman: 102 (2%)
Bisexual: 114 (2%)
Prefer not to say: 416 (8%)
Prefer to self-describe: 28 (1%)
Unknown: 500 (10%)
Religion
Christianity (ref): 2,124 (41%)
Other religion (not Christianity): 804 (16%)
No religion: 1,753 (34%)
Unknown: 499 (10%)
Nationality
British (ref): 4,731 (91%)
Other (not British) or unknown: 449 (9%)
Region:
East Midlands (ref): 313 (6%)
East of England: 407 (8%)
London: 799 (15%)
North East: 248 (5%)
North West: 695 (13%)
South East: 663 (13%)
South West: 375 (7%)
West Midlands: 547 (11%)
Yorkshire and the Humber: 491 (9%)
Other or unknown: 642 (12%)
Sector type
Child (ref): 3,389 (65%)
Adult: 1,120 (22%)
Other: 274 (5%)
Unknown: 397 (8%)
Area of practice
Specialist - children and families (ref): 3,064 (59%)
Specialist - adults: 677 (13%)
Disabilities: 183 (4%)
Hospital: 85 (2%)
Mental health: 378 (7%)
Crime and courts: 164 (0%)
Other: 232 (0%)
To note: Totals may not sum to 100% due to rounding.
Referrer type
Member of the public (reference category): 530 (37%)
Employer: 535 (38%)
Other agency: 86 (6%)
Other social worker: 32 (2%)
Self-referral: 164 (12%)
Unknown: 76 (5%)
Concern type
Poor or inappropriate communication (ref): 198 (14%)
Court proceedings: 50 (4%)
Dishonesty: 239 (17%)
Health: 49 (3%)
Conviction: 66 (5%)
Professional boundaries: 111 (8%)
Record keeping: 93 (7%)
Safeguarding concerns: 265 (19%)
Violent and abusive conduct: 63 (4%)
Unknown and Other: 289 (20%)
Employer
Local authority (ref): 967 (68%)
Children and Family Court Advisory and Support Service (Cafcass): 16 (1%)
Agency: 102 (7%)
NHS: 40 (3%)
Other: 98 (7%)
Unknown: 200 (14%)
Age group
29 and under: 111 (8%)
30 to 39: 314 (22%)
40 to 49 (ref): 425 (30%)
50 to 59: 421 (30%)
60 or over: 152 (11%)
Gender identity
Female (ref): 1,018 (72%)
Male: 401 (28%)
Unknown: 4 (0%)
Ethnicity
White ethnicities (ref): 730 (51%)
Black ethnicities: 281 (20%)
Other ethnicities (not White or Black): 162 (11%)
Unknown: 250 (18%)
Disability
No disability: 964 (68%)
Disability: 142 (10%)
Prefer not to say: 84 (6%)
Unknown: 233 (16%)
Sex same as at birth
Yes: 1,019 (72%)
No: 89 (6%)
Prefer not to say: 76 (5%)
Unknown: 239 (17%)
Sexual orientation
Heterosexual: 973 (68%)
Gay man: 26 (2%)
Gay woman: 23 (2%)
Bisexual: 26 (2%)
Prefer not to say: 129 (9%)
Prefer to self-describe: 5 (0%)
Unknown: 241 (17%)
Religion
Christianity (ref): 554 (39%)
Other religion (not Christianity): 254 (18%)
No religion: 374 (26%)
Unknown: 241 (17%)
Nationality
British (ref): 1,267 (89%)
Other (not British) or Unknown: 156 (11%)
Region:
East Midlands (ref): 74 (5%)
East of England: 115 (8%)
London: 199 (14%)
North East: 61 (4%)
North West: 176 (12%)
South East: 165 (12%)
South West: 106 (7%)
West Midlands: 152 (11%)
Yorkshire and the Humber: 112 (8%)
Unknown or Other: 263 (18%)
Sector type
Child (ref): 793 (56%)
Adult: 334 (23%)
Other: 90 (6%)
Unknown: 206 (14%)
Area of practice
Specialist - children and families (ref): 733 (52%)
Specialist - adults: 199 (14%)
Disabilities: 43 (3%)
Mental health: 120 (8%)
Other: 328 (23%)
To note: Totals may not sum to 100% due to rounding.
Annexe 2: Pre-triage stage logistic regression results
Referrer type
Member of the public (reference category)
Odds ration: 1.00
95% Cl Lower: N/A
95% Cl Upper: N/A
p-value: N/A
Employer
Odds ratio: 3.54
95% CL Lower: 2.28
95% Cl Upper: 5.47
p-value: 0.00 [3]
Other agency
Odds ratio: 1.73
95% Cl Lower: 0.78
95% Cl Upper: 3.83
p-value: 0.18
Other social worker
Odds ratio: 0.94
95% Cl Lower: 0.51
95% Cl Upper: 1.73
p-value: 0.84
Self-referral
Odds ratio: 0.62
95% Cl Lower: 0.42
95% Cl Upper: 0.91
p-value: 0.01 [3]
Unknown referrer
Odds ratio: 0.56
95% Cl Lower: 0.369
95% Cl Upper: 0.85
p-value: 0.01 [3]
[note 3] Statistically significant (p ≤ 0.05) results.
Concern type
Poor of inappropriate communication (ref)
Odds ratio: 1.00
95% Cl Lower: N/A
95% Cl Upper: N/A
p-value: N/A
Conviction
Odds ratio: 5.38
95% Cl Lower: 1.24
95% Cl Upper: 23.37
p-value: 0.02 [3]
Court proceedings
Odds ratio: 0.64
95% Cl Lower: 0.42
95% Cl Upper: 0.97
p-value: 0.04 [3]
Criminal proceedings/involvement
Odds ratio: 1.25
95% Cl Lower: 0.62
95% Cl Upper: 2.52
p-value: 0.53
Dishonesty
Odds ratio: 1.39
95% Cl Lower: 0.99
95% Cl Upper: 1.96
p-value: 0.06
GDPR/data protection
Odds ratio: 2.35
95% Cl Lower: 0.72
95% Cl Upper: 7.64
p-value: 0.16
Health [
Odds ratio: 0.27
95% Cl Lower: 0.16
95% Cl Upper: 0.44
p-value: 0.00 [3]
Issue with social worker's place of work or service as a whole
Odds ratio: 0.44
95% Cl Lower: 0.23
95% Cl Upper: 0.85
p-value: 0.01 [3]
Performance issues
Odds ratio: 0.79
95% Cl Lower: 0.49
95% Cl Upper: 1.27
p-value: 0.33
Bias and/or prejudicial treatment
Odds ratio: 1.51
95% Cl Lower: 0.64
95% Cl Upper: 3.56
p-value: 0.35
Professional boundaries
Odds ratio: 1.59
95% Cl Lower: 0.87
95% Cl Upper: 2.93
p-value: 0.13
Record keeping
Odds ratio: 4.02
95% Cl Lower: 1.45
95% Cl Upper: 11.12
p-value: 0.01 [3]
Safeguarding concerns
Odds ratio: 1.77
95% Cl Lower: 1.16
95% Cl Upper: 2.70
p-value: 0.01 [3]
Sexual misconduct
Odds ratio: 1.25
95% Cl Lower: 0.37
95% Cl Upper: 4.23
p-value: 0.72
Unknown
Odds ratio: 0.15
95% Cl Lower: 0.10
95% Cl Upper: 0.24
p-value: 0.00 [3]
Violent or abusive conduct
Odds ratio: 1.38
95% Cl Lower: 0.67
95% Cl Upper: 2.85
p-value: 0.38
Other
Odds ratio: 1.24
95% Cl Lower: 0.83
95% Cl Upper: 1.87
p-value: 0.30
Small numbers category [2]
Odds ratio: 0.34
95% Cl Lower: 0.19
95% Cl Upper: 0.61
p-value: 0.00
[note 2] Contains categories with fewer than 35 observations, aggregated for calculation in regression.
[note 3] Statistically significant (p ≤ 0.05) results.
Employer
Local authority (ref)
Odds ratio: 1.00
95% Cl Lower: N/A
95% Cl Upper: N/A
p-value: N/A
Children and Family Court Advisory and Support Service (Cafcass)
Odds ratio: 0.89
95% Cl Lower: 0.44
95% Cl Upper: 1.78
p-value: 0.74
Agency
Odds ratio: 1.03
95% Cl Lower: 0.65
95% Cl Upper: 1.65
p-value: 0.89
NHS
Odds ratio: 1.31
95% Cl Lower: 0.65
95% Cl Upper: 2.65
p-value: 0.45
Other
Odds ratio: 0.88
95% Cl Lower: 0.59
95% Cl Upper: 1.32
p-value: 0.54
Unknown
Odds ratio: 1.02
95% Cl Lower: 0.72
95% Cl Upper: 1.45
p-value: 0.91
Age group
29 and under
Odds ratio: 1.06
95% Cl Lower: 0.78
95% Cl Upper: 1.45
p-value: 0.70
30 to 39
Odds ratio: 1.03
95% Cl Lower: 0.82
95% Cl Upper: 1.30
p-value: 0.78
40 to 49
Odds ratio: 1.00
95% Cl Lower: N/A
95% Cl Upper: N/A
p-value: N/A
50 to 59
Odds ratio: 1.08
95% Cl Lower: 0.86
95% Cl Upper: 1.36
p-value: 0.51
60 or over
Odds ratio: 1.17
95% Cl Lower: 0.84
95% Cl Upper: 1.63
p-value: 0.36
Gender identity
Female (ref)
Odds ratio: 1.00
95% Cl Lower: N/A
95% Cl Upper: N/A
p-value: N/A
Male
Odds ratio: 1.08
95% Cl Lower: 0.86
95% Cl Upper: 1.36
p-value: 0.49
Unknown
Odds ratio: 40.62
95% Cl Lower: 0.00
95% Cl Upper: N/A
p-value: 1.00
Ethnicity
White ethnicities (ref)
Odds ratio: 1.00
95% Cl Lower: N/A
95% Cl Upper: N/A
p-value: N/A
Black ethnicities
Odds ratio: 1.11
95% Cl Lower: 0.84
95% Cl Upper: 1.46
p-value: 0.45
Other ethnicities
Odds ratio: 1.24
95% Cl Lower: 0.91
95% Cl Upper: 1.69
p-value: 0.18
Unknown
Odds ratio: 1.02
95% Cl Lower: 0.39
95% Cl Upper: 2.66
p-value: 0.97
Disability
No disability
Odds ratio: 1.00
95% Cl Lower: N/A
95% Cl Upper: N/A
p-value: N/A
Disability
Odds ratio: 0.82
95% Cl Lower: 0.63
95% Cl Upper: 1.07
p-value: 0.15
Prefer not to say
Odds ratio: 1.00
95% Cl Lower: 0.64
95% Cl Upper: 1.55
p-value: 0.99
Unknown
Odds ratio: 0.34
95% Cl Lower: 0.04
95% Cl Upper: 3.32
p-value: 0.35
Sex same as at birth
Yes
Odds ratio: 1.00
95% Cl Lower: N/A
95% Cl Upper: N/A
p-value: N/A
Prefer not to say
Odds ratio: 1.19
95% Cl Lower: 0.67
95% Cl Upper: 2.10
p-value: 0.55
Unknown
Odds ratio: 3.00
95% Cl Lower: 0.37
95% Cl Upper: 24.64
p-value: 0.31
No
Odds ratio: 1.05
95% Cl Lower: 0.75
95% Cl Upper: 1.47
p-value: 0.79
Sexual orientation
Heterosexual
Odds ratio: 1.00
95% Cl Lower: N/A
95% Cl Upper: N/A
p-value: N/A
Gay man
Odds ratio: 0.76
95% Cl Lower: 0.40
95% Cl Upper: 1.42
p-value: 0.39
Gay woman
Odds ratio: 1.12
95% Cl Lower: 0.61
95% Cl Upper: 2.06
p-value: 0.72
Bisexual
Odds ratio: 0.76
95% Cl Lower: 0.45
95% Cl Upper: 1.28
p-value: 0.30
Prefer not to say
Odds ratio: 0.82
95% Cl Lower: 0.55
95% Cl Upper: 1.22
p-value: 0.34
Unknown and Other
Odds ratio: 1.41
95% Cl Lower: 0.43
95% Cl Upper: 4.61
p-value: 0.57
Religion
Christianity (ref)
Odds ratio: 1.00
95% Cl Lower: N/A
95% Cl Upper: N/A
p-value: N/A
Other religion (not Christianity)
Odds ratio: 1.01
95% Cl Lower: 0.75
95% Cl Upper: 1.34
p-value: 0.97
No religion
Odds ratio: 0.95
95% Cl Lower: 0.77
95% Cl Upper: 1.18
p-value: 0.65
Unknown
Odds ratio: 0.67
95% Cl Lower: 0.10
95% Cl Upper: 4.40
p-value: 0.68
Nationality
British
Odds ratio: 1.00
95% Cl Lower: N/A
95% Cl Upper: N/A
p-value: N/A
Other (not British) or unknown
Odds ratio: 0.77
95% Cl Lower: 0.57
95% Cl Upper: 1.06
p-value: 0.11
Region
East Midlands (ref)
Odds ratio: 1.00
95% Cl Lower: N/A
95% Cl Upper: N/A
p-value: N/A
East of England
Odds ratio: 1.92
95% Cl Lower: 1.22
95% Cl Upper: 3.02
p-value: 0.00 [3]
London
Odds ratio: 1.34
95% Cl Lower: 0.92
95% Cl Upper: 1.96
p-value: 0.12
North East
Odds ratio: 1.34
95% Cl Lower: 0.81
95% Cl Upper: 2.20
p-value: 0.26
North West
Odds ratio: 1.29
95% Cl Lower: 0.89
95% Cl Upper: 1.88
p-value: 0.18
South East
Odds ratio: 1.47
95% Cl Lower: 1.01
95% Cl Upper: 2.15
p-value: 0.05
South West
Odds ratio: 1.22
95% Cl Lower: 0.81
95% Cl Upper: 1.84
p-value: 0.34
West Midlands
Odds ratio: 1.19
95% Cl Lower: 0.80
95% Cl Upper: 1.76
p-value: 0.39
Yorkshire and the Humber
Odds ratio: 1.62
95% Cl Lower: 1.07
95% Cl Upper: 2.47
p-value: 0.02 [3]
Other or Unknown
Odds ratio: 1.50
95% Cl Lower: 0.98
95% Cl Upper: 2.29
p-value: 0.06
[note 3] Statistically significant (p ≤ 0.05) results.
Sector type
Child (ref)
Odds ratio: 1.00
95% Cl Lower: N/A
95% Cl Upper: N/A
p-value: N/A
Adult
Odds ratio: 1.57
95% Cl Lower: 0.78
95% Cl Upper: 3.18
p-value: 0.21
Other
Odds ratio: 0.68
95% Cl Lower: 0.28
95% Cl Upper: 1.64
p-value: 0.39
Unknown
Odds ratio: 0.84
95% Cl Lower: 0.53
95% Cl Upper: 1.34
p-value: 0.46
Area of practice
Specialist - children and families (ref)
Odds ratio: 1.00
95% Cl Lower: N/A
95% Cl Upper: N/A
p-value: N/A
Specialist - adults [3]
Odds ratio: 0.47
95% Cl Lower: 0.22
95% Cl Upper: 0.99
p-value: 0.05
Disabilities
Odds ratio: 0.85
95% Cl Lower: 0.52
95% Cl Upper: 1.36
p-value: 0.49
Hospital
Odds ratio: 0.89
95% Cl Lower: 0.33
95% Cl Upper: 2.39
p-value: 0.49
Mental health
Odds ratio: 0.79
95% Cl Lower: 0.40
95% Cl Upper: 1.56
p-value: 0.49
Crime and courts
Odds ratio: 0.74
95% Cl Lower: 0.45
95% Cl Upper: 1.21
p-value: 0.23
Other
Odds ratio: 1.59
95% Cl Lower: 0.60
95% Cl Upper: 4.23
p-value: 0.36
[note 3] Statistically significant (p ≤ 0.05) results.
Annexe 3: Triage stage logistic regression results
Referrer type
Member of the public (reference category)
Odds ration: 1.00
95% Cl Lower: N/A
95% Cl Upper: N/A
p-value: N/A
Employer
Odds ratio: 19.42
95% CL Lower: 15.68
95% Cl Upper: 24.05
p-value: 0.00 [3]
Other agency
Odds ratio: 9.21
95% Cl Lower: 6.11
95% Cl Upper: 13.90
p-value: 0.00 [3]
Other social worker
Odds ratio: 1.34
95% Cl Lower: 0.88
95% Cl Upper: 2.03
p-value: 0.17
Self-referral
Odds ratio: 4.41
95% Cl Lower: 3.33
95% Cl Upper: 5.83
p-value: 0.00 [3]
Unknown referral
Odds ratio: 1.56
95% Cl Lower: 1.14
95% Cl Upper: 2.12
p-value: 0.00 [3]
[note 3] Statistically significant (p ≤ 0.05) results.
Concern type
Poor of inappropriate communication (ref)
Odds ratio: 1.00
95% Cl Lower: N/A
95% Cl Upper: N/A
p-value: N/A
Conviction
Odds ratio: 3.83
95% Cl Lower: 2.01
95% Cl Upper: 7.30
p-value: 0.00 [3]
Court proceedings
Odds ratio: 1.50
95% Cl Lower: 1.06
95% Cl Upper: 2.12
p-value: 0.02 [3]
Criminal proceedings/involvement
Odds ratio: 1.40
95% Cl Lower: 0.84
95% Cl Upper: 2.34
p-value: 0.19
Dishonesty
Odds ratio: 1.25
95% Cl Lower: 0.99
95% Cl Upper: 1.57
p-value: 0.06
GDPR/data protection
Odds ratio: 0.64
95% Cl Lower: 0.29
95% Cl Upper: 1.43
p-value: 0.28
Health
Odds ratio: 1.10
95% Cl Lower: 0.69
95% Cl Upper: 1.76
p-value: 0.68
Issue with social worker's place of work or service as a whole
Odds ratio: 0.57
95% Cl Lower: 0.22
95% Cl Upper: 1.44
p-value: 0.23
Performance issues
Odds ratio: 1.12
95% Cl Lower: 0.75
95% Cl Upper: 1.68
p-value: 0.59
Bias and/or prejudicial treatment
Odds ratio: 0.44
95% Cl Lower: 0.19
95% Cl Upper: 1.03
p-value: 0.06
Professional boundaries
Odds ratio: 1.69
95% Cl Lower: 1.20
95% Cl Upper: 2.37
p-value: 0.00 [3]
Record keeping
Odds ratio: 2.24
95% Cl Lower: 1.55
95% Cl Upper: 3.23
p-value: 0.00 [3]
Safeguarding concerns
Odds ratio: 1.96
95% Cl Lower: 1.52
95% Cl Upper: 2.51
p-value: 0.00 [3]
Sexual misconduct
Odds ratio: 2.28
95% Cl Lower: 1.01
95% Cl Upper: 5.12
p-value: 0.05 [3]
Violent or abusive conduct
Odds ratio: 1.53
95% Cl Lower: 1.00
95% Cl Upper: 2.34
p-value: 0.05 [3]
Unknown and Other
Odds ratio: 1.25
95% Cl Lower: 0.97
95% Cl Upper: 1.60
p-value: 0.08
[note 3] Statistically significant (p ≤ 0.05) results.
Employer
Local authority
Odds ratio: 1.00
95% Cl Lower: N/A
95% Cl Upper: N/A
p-value: N/A
Cafcass
Odds ratio: 0.73
95% Cl Lower: 0.40
95% Cl Upper: 1.36
p-value: 0.33
Agency
Odds ratio: 1.34
95% Cl Lower: 0.99
95% Cl Upper: 1.81
p-value: 0.06
NHS
Odds ratio: 0.98
95% Cl Lower: 0.63
95% Cl Upper: 1.53
p-value: 0.93
Other
Odds ratio: 1.03
95% Cl Lower: 0.77
95% Cl Upper: 1.38
p-value: 0.84
Unknown
Odds ratio: 1.48
95% Cl Lower: 1.15
95% Cl Upper: 1.89
p-value: 0.00 [3]
[note 3] Statistically significant (p ≤ 0.05) results.
Age group
29 and under
Odds ratio: 0.84
95% Cl Lower: 0.64
95% Cl Upper: 1.10
p-value: 0.20
30 to 39
Odds ratio: 0.95
95% Cl Lower: 0.78
95% Cl Upper: 1.15
p-value: 0.57
40 to 49
Odds ratio: 1.00
95% Cl Lower: N/A
95% Cl Upper: N/A
p-value: N/A
50 to 59
Odds ratio: 0.98
95% Cl Lower: 0.81
95% Cl Upper: 1.17
p-value: 0.80
60 or over
Odds ratio: 1.09
95% Cl Lower: 0.84
95% Cl Upper: 1.43
p-value: 0.51
Gender identity
Female (ref)
Odds ratio: 1.00
95% Cl Lower: N/A
95% Cl Upper: N/A
p-value: N/A
Male
Odds ratio: 1.32
95% Cl Lower: 1.10
95% Cl Upper: 1.57
p-value: 0.00 [3]
Unknown
Odds ratio: 2.57
95% Cl Lower: 0.48
95% Cl Upper: 13.88
p-value: 0.27
[note 3] Statistically significant (p ≤ 0.05) results.
Ethnicity
White ethnicities (ref)
Odds ratio: 1.00
95% Cl Lower: N/A
95% Cl Upper: N/A
p-value: N/A
Black ethnicities
Odds ratio: 1.44
95% Cl Lower: 1.16
95% Cl Upper: 1.78
p-value: 0.00 [3]
Other ethnicities
Odds ratio: 0.90
95% Cl Lower: 0.70
95% Cl Upper: 1.15
p-value: 0.39
Unknown
Odds ratio: 1.61
95% Cl Lower: 0.76
95% Cl Upper: 3.41
p-value: 0.22
[note 3] Statistically significant (p ≤ 0.05) results.
Disability
No disability (ref)
Odds ratio: 1.00
95% Cl Lower: N/A
95% Cl Upper: N/A
p-value: N/A
Disability
Odds ratio: 0.99
95% Cl Lower: 0.78
95% Cl Upper: 1.25
p-value: 0.91
Prefer not to say
Odds ratio: 1.17
95% Cl Lower: 0.84
95% Cl Upper: 1.63
p-value: 0.37
Unknown
Odds ratio: 0.13
95% Cl Lower: 0.02
95% Cl Upper: 0.71
p-value: 0.02 [3]
[note 3] Statistically significant (p ≤ 0.05) results.
Sex same as at birth
Yes (ref)
Odds ratio: 1.00
95% Cl Lower: N/A
95% Cl Upper: N/A
p-value: N/A
No
Odds ratio: 1.03
95% Cl Lower: 0.77
95% Cl Upper: 1.37
p-value: 0.84
Prefer not to say
Odds ratio: 0.87
95% Cl Lower: 0.56
95% Cl Upper: 1.35
p-value: 0.53
Unknown
Odds ratio: 2.19
95% Cl Lower: 0.54
95% Cl Upper: 8.93
p-value: 0.28
Sexual orientation
Heterosexual
Odds ratio: 1.00
95% Cl Lower: N/A
95% Cl Upper: N/A
p-value: N/A
Gay man
Odds ratio: 0.96
95% Cl Lower: 0.55
95% Cl Upper: 1.68
p-value: 0.90
Gay woman
Odds ratio: 0.94
95% Cl Lower: 0.56
95% Cl Upper: 1.58
p-value: 0.82
Bisexual
Odds ratio: 1.52
95% Cl Lower: 0.97
95% Cl Upper: 2.38
p-value: 0.07
Prefer not to say
Odds ratio: 1.35
95% Cl Lower: 0.97
95% Cl Upper: 1.87
p-value: 0.07
Prefer to self describe
Odds ratio: 0.71
95% Cl Lower: 0.26
95% Cl Upper: 1.90
p-value: 0.49
Unknown
Odds ratio: 3.35
95% Cl Lower: 0.70
95% Cl Upper: 16.14
p-value: 0.13
Religion
Christianity (ref)
Odds ratio: 1.00
95% Cl Lower: N/A
95% Cl Upper: N/A
p-value: N/A
Other religion (not Christianity)
Odds ratio: 1.28
95% Cl Lower: 1.01
95% Cl Upper: 1.61
p-value: 0.04 [3]
No religion
Odds ratio: 0.85
95% Cl Lower: 0.71
95% Cl Upper: 1.02
p-value: 0.09
Unknown
Odds ratio: 0.80
95% Cl Lower: 0.15
95% Cl Upper: 4.24
p-value: 0.79
[note 3] Statistically significant (p ≤ 0.05) results.
Nationality
British
Odds ratio: 1.00
95% Cl Lower: N/A
95% Cl Upper: N/A
p-value: N/A
Other (not British) or unknown
Odds ratio: 0.97
95% Cl Lower: 0.75
95% Cl Upper: 1.25
p-value: 0.81
Region
East Midlands (ref)
Odds ratio: 1.00
95% Cl Lower: N/A
95% Cl Upper: N/A
p-value: N/A
East of England
Odds ratio: 0.95
95% Cl Lower: 0.65
95% Cl Upper: 1.37
p-value: 0.77
London
Odds ratio: 0.80
95% Cl Lower: 0.57
95% Cl Upper: 1.12
p-value: 0.19
North East
Odds ratio: 1.12
95% Cl Lower: 0.73
95% Cl Upper: 1.71
p-value: 0.61
North West
Odds ratio: 1.03
95% Cl Lower: 0.73
95% Cl Upper: 1.43
p-value: 0.88
South East
Odds ratio: 0.93
95% Cl Lower: 0.66
95% Cl Upper: 1.30
p-value: 0.67
South West
Odds ratio: 1.07
95% Cl Lower: 0.73
95% Cl Upper: 1.56
p-value: 0.74
West Midlands
Odds ratio: 0.93
95% Cl Lower: 0.66
95% Cl Upper: 1.32
p-value: 0.69
Yorkshire and the Humber
Odds ratio: 0.91
95% Cl Lower: 0.63
95% Cl Upper: 1.30
p-value: 0.59
Unknown or Other
Odds ratio: 1.14
95% Cl Lower: 0.80
95% Cl Upper: 1.6 0.474
p-value: 0.47
Sector type
Child (ref)
Odds ratio: 1.00
95% Cl Lower: N/A
95% Cl Upper: N/A
p-value: N/A
Adult
Odds ratio: 1.48
95% Cl Lower: 0.83
95% Cl Upper: 2.64
p-value: 0.19
Other
Odds ratio: 3.15
95% Cl Lower: 1.51
95% Cl Upper: 6.58
p-value: 0.00 [3]
Unknown
Odds ratio: 1.27
95% Cl Lower: 0.89
95% Cl Upper: 1.81
p-value: 0.18
[note 3] Statistically significant (p ≤ 0.05) results.
Area of practice
Specialist - children and families (ref)
Odds ratio: 1.00
95% Cl Lower: N/A
95% Cl Upper: N/A
p-value: N/A
Specialist - adults
Odds ratio: 0.64
95% Cl Lower: 0.35
95% Cl Upper: 1.19
p-value: 0.16
Disabilities
Odds ratio: 1.05
95% Cl Lower: 0.72
95% Cl Upper: 1.55
p-value: 0.79
Hospital
Odds ratio: 0.79
95% Cl Lower: 0.36
95% Cl Upper: 1.73
p-value: 0.56
Mental health
Odds ratio: 0.80
95% Cl Lower: 0.45
95% Cl Upper: 1.44
p-value: 0.46
Crime and courts
Odds ratio: 0.96
95% Cl Lower: 0.59
95% Cl Upper: 1.56
p-value: 0.88
Other
Odds ratio: 0.42
95% Cl Lower: 0.19
95% Cl Upper: 0.94
p-value: 0.03 [3]
[note 3] Statistically significant (p ≤ 0.05) results.
Annexe 4: Case examiner stage logistic regression stage results
Referrer type
Member of the public (reference category)
Odds ration: 1.00
95% Cl Lower: N/A
95% Cl Upper: N/A
p-value: N/A
Employer
Odds ratio: 13.51
95% CL Lower: 8.88
95% Cl Upper: 20.56
p-value: 0.00 [3]
Other agency
Odds ratio: 6.89
95% Cl Lower: 3.70
95% Cl Upper: 12.82
p-value: 0.00 [3]
Other social worker
Odds ratio: 4.85
95% Cl Lower: 1.86
95% Cl Upper: 12.61
p-value: 0.00 [3]
Self-referral
Odds ratio: 5.23
95% Cl Lower: 2.97
95% Cl Upper: 9.24
p-value: 0.00 [3]
Unknown
Odds ratio: 4.00
95% Cl Lower: 1.99
95% Cl Upper: 8.03
p-value: 0.00 [3]
[note 3] Statistically significant (p ≤ 0.05) results.
Concern type
Poor of inappropriate communication (ref)
Odds ratio: 1.00
95% Cl Lower: N/A
95% Cl Upper: N/A
p-value: N/A
Court proceedings
Odds ratio: 1.04
95% Cl Lower: 0.36
95% Cl Upper: 2.98
p-value: 0.94
Dishonesty
Odds ratio: 2.08
95% Cl Lower: 1.23
95% Cl Upper: 3.50
p-value: 0.01 [3]
Health
Odds ratio: 0.63
95% Cl Lower: 0.28
95% Cl Upper: 1.43
p-value: 0.27
Conviction
Odds ratio: 0.58
95% Cl Lower: 0.26
95% Cl Upper: 1.29
p-value: 0.18
Professional boundaries
Odds ratio: 1.58
95% Cl Lower: 0.84
95% Cl Upper: 2.95
p-value: 0.15
Record keeping
Odds ratio: 0.78
95% Cl Lower: 0.39
95% Cl Upper: 1.57
p-value: 0.49
Safeguarding concerns
Odds ratio: 0.49
95% Cl Lower: 0.29
95% Cl Upper: 0.84
p-value: 0.01 [3]
Violent or abusive conduct:
Odds ratio: 1.00
95% Cl Lower: 0.47
95% Cl Upper: 2.13
p-value: 0.99
Unknown or Other:
Odds ratio: 0.85
95% Cl Lower: 0.50
95% Cl Upper: 1.44
p-value: 0.54
[note 3] Statistically significant (p ≤ 0.05) results.
Employer
Local authority (ref)
Odds ratio: 1.00
95% Cl Lower: N/A
95% Cl Upper: N/A
p-value: N/A
Cafcass
Odds ratio: 0.00
95% Cl Lower: 0.00
95% Cl Upper: N/A
p-value: 1.00
Agency
Odds ratio: 0.76
95% Cl Lower: 0.46
95% Cl Upper: 1.28
p-value: 0.31
NHS
Odds ratio: 0.65
95% Cl Lower: 0.26
95% Cl Upper: 1.62
p-value: 0.35
Other
Odds ratio: 1.50
95% Cl Lower: 0.87
95% Cl Upper: 2.56
p-value: 0.14
Unknown
Odds ratio: 1.59
95% Cl Lower: 1.00
95% Cl Upper: 2.54
p-value: 0.05 [3]
[note 3] Statistically significant (p ≤ 0.05) results.
Age group
29 and under
Odds ratio: 0.76
95% Cl Lower: 0.42
95% Cl Upper: 1.38
p-value: 0.36
30 to 39
Odds ratio: 0.59
95% Cl Lower: 0.38
95% Cl Upper: 0.89
p-value: 0.01 [3]
40 to 49 (ref)
Odds ratio: 1.00
95% Cl Lower: N/A
95% Cl Upper: N/A
p-value: N/A
50 to 59
Odds ratio: 0.94
95% Cl Lower: 0.66
95% Cl Upper: 1.35
p-value: 0.75
60 or over
Odds ratio: 0.80
95% Cl Lower: 0.49
95% Cl Upper: 1.29
p-value: 0.36
[note 3] Statistically significant (p ≤ 0.05) results.
Gender identity
Female (ref)
Odds ratio: 1.00
95% Cl Lower: N/A
95% Cl Upper: N/A
p-value: N/A
Male
Odds ratio: 1.04
95% Cl Lower: 0.76
95% Cl Upper: 1.43
p-value: 0.82
Unknown
Odds ratio: 5.61
95% Cl Lower: 0.33
95% Cl Upper: 96.92
p-value: 0.24
Ethnicity
White ethnicities (ref)
Odds ratio: 1.00
95% Cl Lower: N/A
95% Cl Upper: N/A
p-value: N/A
Black ethnicities
Odds ratio: 2.08
95% Cl Lower: 1.35
95% Cl Upper: 3.20
p-value: 0.00 [3]
Other ethnicities
Odds ratio: 1.53
95% Cl Lower: 0.93
95% Cl Upper: 2.53
p-value: 0.10
Unknown
Odds ratio: 1.44
95% Cl Lower: 0.36
95% Cl Upper: 5.81
p-value: 0.61
[note 3] Statistically significant (p ≤ 0.05) results.
Disability
No disability (ref)
Odds ratio: 1.00
95% Cl Lower: N/A
95% Cl Upper: N/A
p-value: N/A
Disability
Odds ratio: 0.70
95% Cl Lower: 0.43
95% Cl Upper: 1.15
p-value: 0.16
Prefer not to say
Odds ratio: 1.48
95% Cl Lower: 0.77
95% Cl Upper: 2.85
p-value: 0.24
Unknown
Odds ratio: 0.59
95% Cl Lower: 0.05
95% Cl Upper: 6.52
p-value: 0.67
Sex same as at birth
Yes (ref)
Odds ratio: 1.00
95% Cl Lower: N/A
95% Cl Upper: N/A
p-value: N/A
No
Odds ratio: 0.82
95% Cl Lower: 0.44
95% Cl Upper: 1.52
p-value: 0.53
Prefer not to say
Odds ratio: 0.42
95% Cl Lower: 0.18
95% Cl Upper: 0.99
p-value: 0.05 [3]
Unknown
Odds ratio: 2.00
95% Cl Lower: 0.20
95% Cl Upper: 19.70
p-value: 0.55
[note 3] Statistically significant (p ≤ 0.05) results.
Sexual orientation
Heterosexual
Odds ratio: 1.00
95% Cl Lower: N/A
95% Cl Upper: N/A
p-value: N/A
Gay man
Odds ratio: 1.51
95% Cl Lower: 0.47
95% Cl Upper: 4.85
p-value: 0.48
Gay woman
Odds ratio: 0.19
95% Cl Lower: 0.02
95% Cl Upper: 1.64
p-value: 0.13
Bisexual
Odds ratio: 1.25
95% Cl Lower: 0.39
95% Cl Upper: 4.02
p-value: 0.71
Prefer not to say
Odds ratio: 1.70
95% Cl Lower: 0.88
95% Cl Upper: 3.28
p-value: 0.11
Prefer to self-describe
Odds ratio: 5.29
95% Cl Lower: 0.53
95% Cl Upper: 52.66
p-value: 0.16
Unknown
Odds ratio: 7.08
95% Cl Lower: 0.60
95% Cl Upper: 83.72
p-value: 0.12
Religion
Christianity (ref)
Odds ratio: 1.00
95% Cl Lower: N/A
95% Cl Upper: N/A
p-value: N/A
Other religion (not Christianity)
Odds ratio: 0.95
95% Cl Lower: 0.60
95% Cl Upper: 1.50
p-value: 0.82
No religion
Odds ratio: 0.86
95% Cl Lower: 0.56
95% Cl Upper: 1.34
p-value: 0.51
Unknown
Odds ratio: 0.23
95% Cl Lower: 0.02
95% Cl Upper: 2.48
p-value: 0.22
Nationality
British
Odds ratio: 1.00
95% Cl Lower: N/A
95% Cl Upper: N/A
p-value: N/A
Other (not British) or unknown
Odds ratio: 1.26
95% Cl Lower: 0.82
95% Cl Upper: 1.93
p-value: 0.29
Region
East Midlands (ref)
Odds ratio: 1.00
95% Cl Lower: N/A
95% Cl Upper: N/A
p-value: N/A
East of England
Odds ratio: 1.25
95% Cl Lower: 0.50
95% Cl Upper: 3.13
p-value: 0.63
London
Odds ratio: 2.36
95% Cl Lower: 1.01
95% Cl Upper: 5.50
p-value: 0.05 [3]
North East
Odds ratio: 1.80
95% Cl Lower: 0.61
95% Cl Upper: 5.31
p-value: 0.28
North West
Odds ratio: 2.90
95% Cl Lower: 1.25
95% Cl Upper: 6.74
p-value: 0.01 [3]
South East
Odds ratio: 2.50
95% Cl Lower: 1.06
95% Cl Upper: 5.87
p-value: 0.04 [3]
South West
Odds ratio: 1.97
95% Cl Lower: 0.79
95% Cl Upper: 4.91
p-value: 0.15
West Midlands
Odds ratio: 2.09
95% Cl Lower: 0.89
95% Cl Upper: 4.94
p-value: 0.09
Yorkshire and the Humber
Odds ratio: 2.03
95% Cl Lower: 0.82
95% Cl Upper: 5.06
p-value: 0.13
Unknown or Other
Odds ratio: 2.93
95% Cl Lower: 1.27
95% Cl Upper: 6.73
p-value: 0.01 [3]
[note 3] Statistically significant (p ≤ 0.05) results.
Sector type
Child (ref)
Odds ratio: 1.00
95% Cl Lower: N/A
95% Cl Upper: N/A
p-value: N/A
Adult
Odds ratio: 0.72
95% Cl Lower: 0.27
95% Cl Upper: 1.93
p-value: 0.52
Other
Odds ratio: 2.12
95% Cl Lower: 0.85
95% Cl Upper: 5.26
p-value: 0.11
Unknown
Odds ratio: 4.39
95% Cl Lower: 1.58
95% Cl Upper: 12.22
p-value: 0.00
Area of practice
Specialist - children and families (ref)
Odds ratio: 1.00
95% Cl Lower: N/A
95% Cl Upper: N/A
p-value: N/A
Specialist - adults
Odds ratio: 1.40
95% Cl Lower: 0.49
95% Cl Upper: 4.05
p-value: 0.53
Disabilities
Odds ratio: 0.47
95% Cl Lower: 0.18
95% Cl Upper: 1.25
p-value: 0.13
Mental health
Odds ratio: 0.93
95% Cl Lower: 0.34
95% Cl Upper: 2.51
p-value: 0.88
Unknown or Other
Odds ratio: 0.39
95% Cl Lower: 0.15
95% Cl Upper: 0.99
p-value: 0.05 [3]
[note 3] Statistically significant (p ≤ 0.05) results.