Transcript Pharmaceutical Business Intelligence- Going beyond

Pharmaceutical Business
IntelligenceGoing Beyond Numbers
By Prakash SM
The Pharma Reality
Like every other industry…
 Large amount of data at every stage of the products/ company lifecycle
– Till 21 CRF R11, a typical US FDA NDA filing was about 21 container trucks large
 All decision and movement from one stage to the next is essentially data driven
 Data accuracy and verifiability critical.
But different
 Errors have huge costs– Viaox
 Data can be presented in multiple format- qualitative (end points, inclusion criteria etc.)
and quantitative (LCMS, QSARs, dose dependency, sales etc)
 While the basic intelligence process is the same, the degree of industry information
required is deep and varies with each stage of the lifecycle ( highly specialized industry
with a vastly varied end-user group)
BI Requirements Change with the Lifecycle
Discovery
Preclinical
Clinical
Registration
Manufacture
Commercial
Focus
Fail early
Develop
Efficacy
justification and
to go further SAFETY
Consolidati
on
Reliability,
Consistenc
y
Marketing
efficiency,
SAFETY
Sample
Require
ments
HTS,
Lead
optimization
QSARs
Lab data
processing,
Animal data
Multiple
Compliance
across
multiple
regions
Record
keeping
and
consistency
across
batches
Performanc
e,
Sales,
Ranking,
PV
Sample
Software
vendors
Genologics
Waters
Oracle,
Documentu SAP,
m and other Oracle
ECM
vendors
MSTR,
Oracle,
ARIS G
Sample
Service
providers
Multiple
Multiple
CROs,
Accenture,
Indegene
IBM,
Accenture
Cognizant,
Indegene,
Genpact
Usually
internal
3
Clinical Trial Data Mining and Analytics
4
Clinical Trial Information Data Mining
Example of Indegene’s implementation of BI tools and domain skills
Background
 Clinical trials- Study on human subjects to determine the safety and efficacy
of the drug/procedure
 No of subjects can vary from 8 to >15000, duration from few weeks to many
years
 Very closely monitored for
– Safety and Efficacy
– Competitive monitoring
• Predicting the success of the drug- both approval and marketing
• Determining the future trends in treatment
 Reporting in mandatory in most countries, with specified schedules
 Significant amount of data available to public consumption
5
Clinical Trial Information Data Mining
The Challenge
– No common repository- over 275 trial sites
– Large number of clinical trials – Big sites have over 100000 clinical trials each, average
of 2000+ trials
– Multiple formats of reporting
– Multiple reporting of same trial data
– Text and quantitative information mixed, both critical for evaluation
– Some degree of search and categorization possible
 Needs a lot of reading from multiple sources and further classification to
determine any trend
 Needs significant understanding of the disease area and expertise- Usually by
mid level resources
 What is right for one disease need not be right for another
6
Clinical Trial Information Data Mining
The Solution
 Comprehensive cloud-based platform for searching and analyzing clinical trials
in a given indication/disease area
 Clean redundancy from multiple sources, develop standardized parameters
 Standardized parameters for each disease area
 Allow analysis by parameters specific to each indication
Analytical platform of global clinical trials, aimed at aiding clients in their effort to
understand specific clinical trial landscapes and their competitive environment.
Aimed at a wide variety of functions such as clinical operations, brand
management, CI professionals as well as strategic marketing.
7
Clinical Trial Information Data Mining
Sources:
Clinicaltrials.gov
ANZCTR
CTRI
UKCRN
NTR
IRCT
JapicCTI
ChiCTR
Parsers
Internal
output
engine
Human intelligence
Manual
Process
UMIN
Data warehouse
ISRCTN
SANCTR
Output
Trial Analytics Platform
ICTRP
Publications
Conference
presentation
Manual
tagging of
parameters
8
As you get it on one of the best source site
9
Possible refinement
10
Indegene soluion:
Eg: Breast cancer: Endpoints Distribution Analytics*(1/2)
Completed trials have mainly focused on Overall Response Rate and Progression Free Survival
as endpoints
* For selected drugs
11
Breast cancer: Endpoints Distribution Analytics*(2/2)
Ongoing trials are focused on Overall Response Rate, Progression Free Survival and Disease Free
Survival as endpoints
* For selected drugs
12
Breast cancer: Patient Segment Distribution Analytics*(1/2)
Molecules of interest were evaluated in first line metastatic breast cancer and in treatment refractory
breast cancer
* For selected drugs
13
Breast cancer: Patient Segment Distribution Analytics*(2/2)
Majority of molecules are being evaluated in patients with metastatic breast cancer and in primary
breast cancer.
* For selected drugs
14
Breast cancer: Site distribution
Majority of the clinical trial site are present in North America and Europe.
* For selected drugs
15
Comparison View
16
Trial Timelines
17
Implementation of BI solution for a big Pharma
18
Custom Performance BI and Reporting
Initial client state
Data Sources
IMS Midas, NPA,
NDTI, NDC
AMR, Synovate,
Cegedim
CAM, CD Promo
Impact Rx
Company
Financial data
Data types
Sales $ and units
Patient level
Promotional
Financials
• Patient share
by indication
• Continuing
patient share
• New, switch
patient share
• Reach &
frequency by
reps
• Samples
• Journal ad
Analysis
Reporting by
product or
franchise (e.g.
Oncology,
Immunology)
•
•
•
•
•
$ share
Unit share
NRx share
TRx share
Sales by
indication
Global dashboards for
Senior Management
Regional dashboards for
Regional Teams
• Net Sales
Country level dashboards
for country managers
19
Solution Options
• Indegene provides multiple solutions:
1. An XML feed delivered from data warehouse servers and read by Excel front end
2. Custom built analysis and web delivery platform
3. Business Objects or Cognos based BI platform working along with client’s internal solution
20
Analysis and delivery platform
Stage 2
Stage 3
Stage 4
Data source
OLTP
Analysis and Reporting
Subscribed
data sources
Analysis
Services
Data
Warehouse
Country/
Regional
reports
Data
Transformation
(DTS) / Extract
Reporting
Services
Operational Data
Store(ODS)
Transform
Legacy
Load
(ETL)
Extract
Clean
Conform
Stage 5
Delivery platform- SharePoint, Web Portal, Direct delivery
Stage 1
End users
Reports
Dashboard
reports
Deliver
• Delivery Platform include sophisticated Oracle and Cognos toolset for global data connectivity and faster processing
of large amount of data from country ops, global etc.
Benefit to Client:
1. Reduced cycle time for ongoing reporting
2. Reduced total cost of ownership
3. Standardization, ease of use, compliance
21
Report Delivery Options/Formats
Delivery Options
• Push Method: This type delivery option is preferred for reports with long running times with
complex calculations, extensive run-time roll-ups, aggregations etc. All the reports mentioned
below can be generated by this approach:
 Standardized brand reports
 Weekly Performance Reports
 Monthly Performance Reports
 Quarterly Reports
• Pull Method: Suitable for ad-hoc reports that aren’t standardized.
Publishing capabilities
• Web portals
• Email reports
• Intranet postings
• SharePoint
Delivery Formats

MS XLS

MS PPT

PDF

HTML

XML formats

Flash
22
Progressing of the engagement
Dashboard Continuum
Year 3
Year 2
Year 1
Recommendations
Analysis
Simple Dashboards
Analysis
• Brand X grew its share in this period
by y%, which is -% more than the
market growth
Analysis
• Brand X grew its share in this period
by y%, which is -% more than the
market growth
Recommendations
• Sales calls to segment Y physicians
need to go up to realize higher share
of Brand X
Value
23
Samples
SFE Dashboard
Std Performance Dashboard
Tool to collate and track KPIs across
multiple countries
Standardized performance report
collation and reporting
IMS query service
Market Performance Tracking
Quick search and reporting for IMS
queries
IMS driven sales tracking across
competitor and regions
Therapy Area Dashboard
Competitor tracking reporting by TA
Product Reports and Market
Analysis
Thank you