Data

A data platform to help PHSO understand access, inequality and regional variation

Name: UBDS IT Consulting Limited
Price range: $

Overview

The Parliamentary and Health Service Ombudsman (PHSO) is an independent organisation that investigates complaints about UK government departments, other public bodies, and the NHS in England when people have been unable to resolve them elsewhere. Acting as a final and impartial stage in the complaints process, the PHSO helps ensure citizens are treated fairly, identifies where services have fallen short, and highlights systemic issues that can drive wider improvement across public services.

Key Services

Data & AI

Technologies

Azure Databricks

AzureDevOps

Microsoft Fabric

Executive Summary

Turning casework data into population-level insight.

A geo-demographic insight platform that helps PHSO benchmark service usage against population data, identify under- and over-representation, and prioritise outreach, policy focus and deeper investigation.

The Parliamentary and Health Service Ombudsman needed to better understand systemic issues and regional differences across the populations and communities it serves.

While the organisation held rich internal casework data, including demographic and geographic information, it lacked a clear population benchmark and visibility of wider external trends. As a result, it could see internal volumes and patterns but could not confidently answer critical questions such as:

Where are systemic differences in how communities engage with our services?
Are some groups, including those with protected characteristics such as ethnicity or disability, over- or under-represented?
Are patterns statistically meaningful, or simply reflective of population size?
Where should outreach, policy focus, or deeper investigation be prioritised?

For example, if 30% of cases came from people aged over 65, was that high? Without knowing whether over-65s made up 15% or 35% of the population, leadership could not assess whether this reflected over-representation, under-representation, or normal variation.

More importantly, this context strengthens the organisation’s ability to identify and evidence potential systemic barriers to access. If certain ethnic groups, disabled people, or other protected communities were under-represented, this could indicate that those groups were not reaching the service, rather than not experiencing issues. Equally, over-representation could signal underlying systemic challenges requiring further investigation.

UBDS Digital partnered with the organisation as a cross-functional rainbow team to design and implement a geo-demographic insight platform that anchored service usage to population context from open datasets from ONS, MHCLG and others.

The solution was delivered iteratively using an Agile approach. The initial phase established the core product capability, including the demographics insight tool, statistical significance framework, and reusable data foundation. A second phase onboarded two new datasets for deprivation and community group context, regional exploration and benchmarking, and executive insight summaries.

In parallel, we conducted a structured research exercise, engaging with over 40 organisations to understand the wider external data landscape, including sensitive and restricted data. This informed a practical roadmap for integrating additional datasets and expanding analytical capability over time.

The Challenge

Seeing beyond case volumes to identify meaningful systemic insight.

The organisation lacked the contextual data and statistical framework needed to understand what ‘normal’ looked like compared to the wider population and to confidently identify where systemic or access issues may exist.

1. Volume Without Population Context

The organisation could see how many cases were received by age, ethnicity, disability status, or geography, but could not determine whether those numbers were higher or lower than expected given the underlying population profile.

For example, if one region generated 5,000 cases, that might appear significant. However, if that region accounts for a large share of the national population, the volume may simply reflect scale rather than disproportionate demand.

Without comparing their data to the wider population, it was difficult to determine whether services were being accessed equitably, or whether certain groups were not reaching the service at all.

2. Complex Geography

The organisation needed to analyse patterns across multiple geographic systems, including NHS geographies, counties and regions, parliamentary constituencies, and postcode-level searches.

However, these boundary systems do not align neatly.

For example, an ICB boundary may overlap multiple parliamentary constituencies. Deprivation data may be available at Lower Super Output Area level, while casework data may be recorded at postcode level.

Without a consistent approach, comparisons risked being misleading, limiting confidence in identifying geographic patterns or targeting interventions effectively.

3. Risk of Over-Interpretation

Most requests for interpreting data had to involve a small number of experts. Users did not have a simple statistical framework to determine whether differences were meaningful.

Without statistical safeguards:

Normal variation could be mistaken for meaningful insight
Small-number effects could distort interpretation
Conclusions could drift into unsupported explanations

This created a risk that perceived differences, particularly for smaller or minority groups, could either be over-interpreted or overlooked entirely.

4. Integrating External Data at Scale

While valuable open datasets existed, onboarding them in a robust and sustainable way was a challenge. Datasets varied in format, granularity, quality, and governance requirements.

The Solution

A geo-demographic insight platform.

UBDS Digital designed and implemented a scalable geo-demographic platform anchored to authoritative population data and robust statistical methodology.

Solution Components

Population-Benchmarked Analytical Framework
Our analytical methodology of service usage against population profiles supported identification of meaningful over- or under-representation, particularly across protected and minority groups, while avoiding unsupported conclusions.
Multi-layer Geospatial Visualisations
Our visualisation approach allowed for intuitive analysis across NHS geographies, regions, constituencies, and postcode-level views.
Agile Delivery of Core and Enhanced Capability
Core capabilities were delivered first in a rapid build/release followed by enhancements driven by user feedback, including deprivation context, regional benchmarking, and executive insight summaries.
Modern Data Platform Foundation
A scalable Lakehouse architecture supports automated ingestion of internal and external datasets, governed analytical models, and reproducible pipelines.
Robust External Data Integration Approach
A data model and pipeline architecture was established to onboard open and external datasets, ensuring relevance, quality, and governance alignment, and sustainable expansion over time.
Data Landscape Research and Roadmap
A focused research exercise engaged over 40 organisations and assessed potential datasets, resulting in a prioritised roadmap for future data integration and capability expansion.

The Outcome

Defensible Insight into Systemic and Access Issues.

Leadership now has a clear, evidence-based view of who is and is not reaching the service.

This enables the organisation to:

Identify potential systemic barriers affecting minority and protected groups
Detect under-representation that may indicate access issues
Highlight over-representation that may signal systemic challenges
Prioritise further investigation and engagement where it matters most

Clearer Strategic Decision-Making

The organisation can:

Target outreach and policy interventions more effectively
Focus resources on areas where access or outcomes may be unequal
Avoid unsupported assumptions and anecdotal interpretation

Improved Confidence and Transparency

Users can confidently interpret results, supported by clear statistical tiering that distinguishes signal from noise.

Better Alignment of Resources

By anchoring service data to population context, the organisation can align outreach and engagement efforts more effectively and support resourcing decisions with objective evidence.

A Scalable Foundation for Future Growth

The organisation now has:

A reusable data foundation
A robust approach to onboarding open and external data
A prioritised roadmap for expanding analytical capability

This provides a model for other public sector organisations seeking to understand who their services are reaching, who they are not, and where systemic inequalities may exist, using data in a defensible and scalable way.