Stephen Clinton — Power BI Portfolio

Calls Waiting Wallboard

Requirement

For cases requiring a call back, provide a near-real-time view of call volumes and wait times split across groups to enable teams to see waiting workload, and supervisors to dynamically intervene in queue management and prioritisation. Audience: Clinical contact centre staff and team leaders, senior operational command staff.

Power BI Solution

Direct Query Power BI dashboard polling every 60 seconds, displaying counts of cases by response priority, using both numerical and coloured bar visualisations for the dashboard to be viewed across the room as well as on a desktop. Measures include CALCULATE with COUNT, SUM and FILTER for metrics and MAX / MAXX for durations.

Data Modelling / Engineering

Initiated use of a known flag in the patient management system to show queue status, with support from the vendor, used multi-condition CASE logic across separate event tables within an optimised query against the live tables to prevent interference with critical response operations and ranking window functions to prevent duplicate values.

Key Performance Indicator – Outcomes Based

Requirement

Report a significant internal and external KPI with high accuracy across volumes and percentage of all cases. Add the outcome category and the clinician team resolving the case. Audience: Operations senior leadership team, Clinical Contact Centre leadership and staff, metric is reported up to board and external stakeholder level.

Power BI Solution

New Power BI Report showing the percentage and volume over time with headline metrics and group / outcome detail. Report uses bookmark functionality for percentage / volume transition and a floating information window. DAX measures include extensive CALCULATE, SUM and DIVIDE for calculating the metrics.

Data Modelling / Engineering

Redesigned KPI measurement to use case outcomes via a new reference table, replacing keyword-based calculation methodology, leading to more accurate metric. Socialised the calculation of the metric to stakeholders and data teams, and created a new SQL View of all cases with outcomes, for reference and downstream reporting.

Key Performance Indicator – Time Based

Requirement

Provide a reliable methodology to measure duration from the case being created to a specific action taking place and a Power BI Report to display the metrics over time, by clinician group assessing and with high level call details. Audience: Clinical leadership, senior operations team, performance stakeholders, with summaries to executive team and board.

Power BI Solution

Power BI report built used the newly modelled dataset with visuals for key metrics and over-time with multiple axis options. Filtering by case priorities and clinical level. Included a responsive drill-through page for incident detail and a separate help / information / version page. DAX measures include PERCENTILEX, MEDIANX and AVERAGEX with FILTERs, CALCULATE and COUNTROWS. Created documentation for stakeholders and data team to understand calculation steps and the meaning of the new metric.

Data Modelling / Engineering

Defined accurate start and end timestamps for the metric, supported by complex SQL to correctly attribute the times between different queue tables. Query contained multiple CTEs to gather data, ranked window functions to capture specific case events where multiple events existed, and STRING_AGG to show queue levels on the case journey. Created a view for further downstream reporting. All processes documented with code for future reference.

Response and Utilisation Information – Staff Group

Requirement

A new report for an existing staff group showing availability, activity and utilisation filterable by region and responder type. Audience: Group management team, operations leadership and performance management groups.

Power BI Solution

Power BI report to present headline metrics, time and event-based charts and tables and feature utilisation percentage. Highly filterable for areas, types and episode length, using field parameters and an applied-filter display. DAX measures include CALCULATE, COUNTROWS and FILTERs, CONCATENATE/X for active filters, refresh times and version display, COUNT, SUM, DIVIDE and AVERAGE for metrics and ALLEXCEPT for broader measures. Report also contains detail pages for review of individual responses and a separate help, information and version page. Socialised the methodologies to stakeholders and included in help file.

Data Modelling / Engineering

Required new raw data capture of availability from the response management system as this team reports in a non-standard way. Worked with vendor to clarify feed and wrote procedure to capture and populate table for downstream reporting with extensive data quality testing to ensure accuracy. Captured event times and calculated availability durations, using temporal modelling to link responses to availability to determine utilisation. Provided thorough documentation of data capture methodology for data teams and to stakeholders for information.

Clinical Call Flow Comparison

Requirement

Report the flow of calls through clinical teams, showing call status on entry to the queue and the outcome type, to identify patterns in how calls are processed across different clinical levels. Audience: Clinical leadership groups to enable identification of repeatable call pathways and bottlenecks, directly influencing operational decision-making. Report would replace intensive sessions using extracted patient management system data in spreadsheets.

Power BI Solution

New dashboard with visualisations that aid the viewer in comparing the input and outcome for maximum usability of complex scenarios. Also includes pages for different perspectives (outcome, category and priority) and drill through detail pages with a separate page for help, information and version updates. DAX measures include CALCULATE, with REMOVEFILTERS for overall totals, SWITCH and SELECTEDVALUE for parameterised charts. Created documentation and shared with stakeholders to break down the complexity of the call queuing process and allow understanding of the measures being shown in the report.

Data Modelling / Engineering

Designed a new state-transition model from clinical queue event data using indicators with ordered event streams using LEAD/LAG logic to capture the time a case spends within each clinician group. Enforced temporal accuracy preventing overlaps between levels to ensure the strictest accuracy. Using CTEs for data capture and structure, ranked window functions and LEAD/LAG for event timings and complex CASE statements to identify the correct events, created persistent tables to materialise the data for downstream reporting, simplifying performance reporting across all clinical queues and avoiding re-processing large datasets in the future. Prepared extensive technical documentation to share with the data teams to clarify the logic and process.

Patient Assessments Report (Questionnaire Data)

Requirement

Provide visibility of clinician decision-making in remote patient assessments using responses from questionnaire-style patient assessments built into the patient management system. Show aggregated response counts and relationships between responses, and detailed responses by Clinician for feedback. Audience: Clinicians, team leaders, operational leadership.

Power BI Solution

Multi-page Power BI report showing assessment volumes, assessment durations and outcomes with correlations between responses for all activity. Additional pages for each assessment type (7), incident level detail and help, information and version detail. Created Power Query transitions to pivot the data into episode level responses, allowing the comparisons of per-episode responses. Created documentation on the methodology and shared with stakeholders to facilitate understanding especially around the individual clinician feedback.

Data Modelling / Engineering

Designed and built a new episode-based model of questionnaire responses defining the start and end of each episode to enable volumetric analysis and sequencing. Data required significant cleaning including multiple response handling, question and response text cleaning to standardise entries and branch logic to show separate pathways within the same questionnaire. Resulting clean data tables provided a reusable dataset for all downstream questionnaire reporting across all clinician streams. Provided extensive technical documentation to share with the data teams to clarify the logic and process.

Telephony Report

Requirement

Provide information on telephone activity; call distribution, volume, handling and wait times including detail to identify specific calls for review. Audience: Clinical contact centre leadership, clinical group supervisors, operations and performance groups.

Power BI Solution

Multi-page report for volumes, with pages for detail on call durations, handling time analysis, wait time analysis and inbound and outbound calls including call level drill through and a data glossary page. Filters for call types and groups handling as well as answer status and date. DAX measures include CALCULATE with COUNT, SUM and AVERAGE with REMOVEFILTERS and KEEPFILTERS for some broader metrics. Power Query table creation for call handling and wait time intervals. Additional calculated columns referencing call fields, provide context for calls in the report and classification into groups.

Data Modelling / Engineering

Created new data warehouse procedures and data tables to report telephone activity following changes to switchboard which rendered earlier third party tables redundant. Extensive mapping and joining of CISCO native call tables was required to provide accurate reliable call tables covering all call, transfer, dial and hold activity on the switchboard. Wrote additional procedures to create tables for agent activity, logged in sessions and calls which combine to create reports used for individual agents. Each form the backbone of all telephony reporting in the organisation replacing all others previously used.

Operations Report

Requirement

Provide visibility of the call handling and clinical assessment activity from the patient management system. Include case volumes, activity frequency and durations as well as headline metrics, patient demographics, performance for internal and external indicators, and case outcome information. Replace multiple fragmented reports from other reporting systems. Audience: KPI pages for assessment times used in all external performance reports to stakeholders and all pages used daily by operations teams as the main source of operational information.

Power BI Solution

Multi-page consolidated operational dashboard with high-level KPI tracking and detailed drill-down analysis. DAX measures for time-based KPIs and aggregation across operational dimensions. Filter page with indicators on each page to show filtering impact. DAX measures include active filters using table creation with ISFILTERED, as part of 40+ measures covering various calculated metrics. 168 hour weekly matrix is used for demand analysis and staffing predictions.

Data Modelling / Engineering

Most of the base data came from a minimum data set provided by data engineering, with additional CTEs for other more complex data from other native and warehouse tables, requiring familiarity with operational processes and patient management system data. CTEs handled ranked window functions where multiple entries existed and CASE statements to categorise aspects of the data.

Electronic Patient Care Record Report

Requirement

Summarise all clinical observations in patient care records captured by the front-line clinical responders and show relationships between records. Includes volume analysis, patient demographics, questionnaire and outcome analysis. Audience: Clinical leadership, response team supervisors, clinical audit teams.

Power BI Solution

Large multi-page Power BI report to show all ePCR information including common placement of report-wide filters with metrics to ease filtering to see impact and relationships between patient findings and interventions. Additional specialist reports for Maternity calls and Medicines Management. DAX measures include CALCULATE counts throughout, REMOVE/KEEPFILTERS/ALL for broader metrics, ISFILTERED for active filters, USERELATIONSHIP for drug associations.

Data Modelling / Engineering

Vendor servers synchronised multiple times daily to organisation data warehouse. Supplemental tables queried for additional information including interaction logic to present relationships between medications administered. Incremental loads into Power BI Service given the high volume of records.