Pen CS will be opening up the full capability of PAT CAT though a purpose built API.

Taking Your Population Health Reporting to the Next Level

The PAT BI database provides unprecedented access to the data derived and submitted from CAT4 to your PAT CAT. A purpose built granular database designed for use with Business Intelligence (BI) Tools such as Qlik, Power BI etc. empowers your PHN/ACCHO to explore the depth of data currently at your disposal. Practice data is stored at patient level with an encrypted unique patient identifier allowing for comprehensive longitudinal reporting.
KEY FEATURES

  • Uninterrupted access to your data
  • Granular patient level database structure
  • Fully documented database schema
  • Includes PAT CAT database cleaning tools and services
  • Capacity for patient and practice longitudinal reporting
  • If you are using the Pen CS Geocoding Service you will have the added advantage of bringing together your practices CAT4 clinical data with additional data sets using the exposed Statistical Area 2 (SA2) codes. Each patient, in CAT4, has their address mapped to a SA2 code. Once an SA2 code is assigned to a patient, it can be linked to data sets available outside of the practice clinical information system such as immunisation rates and ABS data.

PAT BI Database Server Specifications


 

Minimum Requirements

Recommended Requirements

Operating System

Windows Server 2012 (Standard/R2) X64 or higher

 

Processor

8 Logical Cores

32 Logical Cores or more

RAM

16 GB

64 GB or more

Hard Disk

500 GB*

1 TB* or more

Microsoft .NET
Framework

.NET 3.5, 4.5 and 4.6.1

 

Database

Microsoft SQL Server 2014 Standard Edition

Microsoft SQL Server 2016 Standard Edition

Other Requirements

  • SQL Server Management Studio
  • Mixed Mode Authentication
  • SQL Server Agent
  • Firewall exception for connectivity

    - Standard SQL server ports are
  • TCP 1433
  • UDP 1434
  • Remote access facilities for installation and maintenance by Pen CS
  • with local admin privileges and database owner privileges

 

Operational Requirements

  • We recommend having an experienced database administrator to

maintain the database and the server

  • The PHN holds the sole responsibility to protect and/or manage the

database (including related artifacts) and the server

 


* Refer to the Estimated Database Size Growth table.


Estimated Database Size Growth


Patients per Extract

Total Practices

Total Duration of Submission (months)

Total Patients for the Duration

Estimated Size (GB)

12,000

250

3

9,000,000

140

12,000

250

6

18,000,000

279

12,000

250

9

27,000,000

419

12,000

250

12

36,000,000

558

12,000

250

15

45,000,000

698

12,000

250

18

54,000,000

837

12,000

250

21

63,000,000

977

12,000

250

24

72,000,000

1,116

12,000

250

27

81,000,000

1,256

12,000

250

30

90,000,000

1,395

12,000

250

33

99,000,000

1,535

12,000

250

36

108,000,000

1,674

12,000

250

39

117,000,000

1,814

12,000

250

42

126,000,000

1,953

12,000

250

45

135,000,000

2,093

12,000

250

48

144,000,000

2,232

12,000

250

51

153,000,000

2,372

12,000

250

54

162,000,000

2,511

12,000

250

57

171,000,000

2,651

12,000

250

60

180,000,000

2,790

12,000

250

63

189,000,000

2,930

12,000

250

66

198,000,000

3,069

12,000

250

69

207,000,000

3,209

12,000

250

72

216,000,000

3,348

12,000

250

75

225,000,000

3,488

12,000

250

78

234,000,000

3,627

12,000

250

81

243,000,000

3,767

12,000

250

84

252,000,000

3,906

12,000

250

87

261,000,000

4,046

12,000

250

90

270,000,000

4,185



"Estimated Size (GB) =((Patients Per Extract ×Total Practices ×Total Duration)/1,000,000)×15.50


Complex Database Storage

Data growth will eventually hit a limit where increasing disk space is no longer possible. In order to solve the limit issue, we can split the data into multiple drives. Splitting data over multiple disk increases performance of read and writes operations as this spreads the I/O loads, allowing for parallel queries and specific backup/restore of data.


When a drive reach its storage limit, it’s possible to split the data into multiple drives. An example (below) shows the Diagnosis split into 4 disks (other secondary tables has been omitted).



Important: Splitting data into multiple disk incurs a maintenance overhead and complexity in backups, they should be reviewed when splitting data into multiple disks. We recommend having an experienced database administrator to maintain the database and the server. The PHN holds the sole responsibility to protect and/or manage the database (including related artifacts) and the server
Large databases are the domain of a DBA/System engineer for a large database. The below items form the basis of the decision that a DBA/System engineer will need to make as the database grows. These include but are not limited to:



Data Disk o Quality SSD etc

  • Size (one large, many small)
  • Disk Controllers (number of, disk per controller, quality of, speed, cache size) o RAID
  • Type (SAN, NAS, DAS)
  • CPU
    • Type
    • Speed
    • Number of
    • Number configured for SQL Server
  • Memory
    • Type
    • Speed
    • Total Size
  • High Availability
    • fallover clustering
    • Database mirroring
    • Log Shipping
    • Replication
    • Scalable shared data
  • Network
    • Type
    • Speed

Backup/
Restoration Plan

  • Recovery Model
  • Full backup cycle
  • Differential backup cycle
  • Log backup
  • Restoration procedures

Maintenance Plan

  • Data and log file
    management (size of files)
  • Index fragmentation (when to defrag)
  • Statistics (when to recalculate)
  • Corruption detection

Monitoring

  • All of the above
  • Long running quires
  • File growth
  • Number of users/ applications




In addition to the PAT BI Database you also get a detailed Database Schema document that will explain exactly how the database tables are connected to each other. Below is s sample template of the PAT BI Database Schema,



Medication_Hormone_Replacement_Therapy_Oestrogen_Only

 

 

 

Active

bit

Yes

The status of 'Hormone Replacement Therapy Oestrogen Only'.
This will either be True, False or NULL.
For example:

  • Patient has an active (True) 'Hormone

    Replacement Therapy Oestrogen Only'
  • Patient has an inactive (False) 'Hormone Replacement Therapy Oestrogen Only'
  • Patient does not have (NULL) 'Hormone
    Replacement Therapy Oestrogen Only'

End_Date

datetime

Yes

The end date and time of 'Hormone Replacement Therapy Oestrogen Only'

ExtractID

bigint

No

The reference identifier of an extract. Use this column to query data for an extract or collection of extracts

ID

bigint

No

The primary key and seed

PatientID

bigint

No

The reference identifier of the patient. Use this column to query data for a patient or collection of patients

Prescription_Date

datetime

Yes

The date and time of the prescription