Clinical Service Line/Specialty peers

The below states the aims, methodology and outcomes behind how Clinical Service Line/Specialty peers are created. 

Aim:

Identify which NHS Trusts/Organisation are ‘alike’ at a chosen service line level based on various characteristics

 

Method:

3-stage approach was used:

1. Analysis was based on HRG codes, creating a unique case provided at the selected service level by concatenating: HRG code, Admission method (and flagged for Day Case)

A % rate of each Concatenation is calculated at Providers’ service line level to mark for importance and calculate local HRG weight. The model was then additionally adjusted by size of the department (based on the total number of activity) and number of unique characteristics

 

2. The model was adjusted by additional characteristics of the service (% of service characteristics at Trust level): Elective/Non Elective, Over/Under 60 , Ethnicity, Urban/Rural, Day Cases

Association analysis

A Linear regression between the variables and Cost per WAU as the independent variable was performed to understand how these variables associate with the outcome (Service Level Cost per WAU). The estimates were used as weights to adjust selected variables (i.e. marking for importance in the model)

 

3. Euclidean Distances calculations:

List of standardised values for each provider - to allow comparison, e.g.:

                                    (% HRG – [Avg % HRG])/StdDev

 

Distances between variables for each pair of providers :

                    ( [Trust A Standardised Value] - [Trust B Standardised Value] )2

 

Distance between Trusts (adjusted by the count distance and square):

                        SQRT([Sum_Distance]/[Count distances])

 

Variables included in model Development:

 • Elective/Non Elective
• Over/Under 60
• Ethnicity
• Urban/Rural
•Day Cases

 

Data sources:

• Hospital Episode Statistics (HES):
• Admissions, A&E attendances and outpatient appointments
• Service characteristics
• Casemix - Healthcare Resource Groups (HRGs)
• DQ: dependent on providers reporting completeness
• Model Health System- Cost per WAU

 

Primary Outcome:

Lists of 10 peers for each service line for each Trust.

 

Secondary output:

Output imported onto Model Health System dashboard to support benchmarking and opportunity estimations.

 

Feedback and Knowledge Base