Modeling results from Health Sciences data
1 like405 views
This document discusses strategies for using modeling and predictive analytics to reduce costs and improve efficiency in clinical trials. It provides examples of how modeling site activation and enrollment can help minimize delays and budget overruns. Risk-based monitoring is presented as another way to transition to more efficient monitoring. Savings projections are estimated for different therapeutic areas based on reducing clinical site monitoring costs from 30% to 3-15% of total costs. Overall, the document argues that predictive modeling approaches can help lower total costs of clinical trials through better planning, execution and oversight.
![Strategic Business Transformation
Visibility into the Unknown
• Site activation and enrolment modeling & simulation: Predict site
activation and enrolment for all trials. Quickly identify regions and sites
with greatest risk for activation and enrollment “misses”; allowing for
early defensive action (i.e., adding more sites, modifying recruiting
efforts) and minimizing margin burn-down.
• Transition to a blended site monitoring model through risk-based
monitoring: Target field and central monitoring according to risk;
leveraging predictive analytics and business process management
changes into a unique-to-industry site monitoring paradigm.
• Confident bid defense: Assess prior contract profitability , along with trial
and site metadata; and incorporate priors for site activation and
enrolment in bid preparation and defense.
Higher Business Agility & Scale
Fastest to Actionable Insight
Economies of Scale
Universal Adaptability
• Minimize duplication of effort through a central library of templates and
analytics used by all departments, all therapy areas and all phases of
clinical trials (Use cases: data review, adaptive monitoring, data cleaning,
trial operations, safety assessments). [unquantified]
• Productivity savings for clinical trial informatics: Reduce time and effort
for data source combination, data preparation, report and dashboard
creation & operations for setting up a new trial by at least 40%. Assume
one person is working to set up each new trial. Assuming an estimated
cost of $150K per FTE (fully loaded) leads to a cost saving of
approximately $60K per year per FTE. (Use case: ability to scale up trials
with limited headcount). Savings based on 50 new trials
Lower Total Cost of Ownership
Self-service Discovery
• Accelerate CRA productivity and efficiency – Focus CRA’s on process-
driven specific site tasks requiring action.
• Streamline processes - involving data management, analysis/report
preparation, dashboard creation and distribution of trial operations
metrics.
• Assuming 200 internal users and 15% productivity increase; this is the
equivalent of up to 10.5 FTE‘s. Assuming average total loaded cost for FTEs
across these roles is $150k, then potential cost saving is approximately
$4,450,000 per annum. Potential of deploying these FTE‘s to new work will
increase the value.
• Bids: Improve awareness of margin indications (history/metadata on
trial, regions, sites) when bidding for Sponsor trials. Savings = Reduced
bid effort direct + indirect costs * number of bids.
• Trial / Site Planning: Improve site/PI selection and site activation /
enrolment modeling to drive better and predictable activation and
enrolment. Savings = Reduced margin loss due to under-enrolling sites.
• Execution: Transition from calendar to risk based monitoring. Savings =
Reduced monitoring effort % * cost(s) of Phase III trials
• Monitoring: Provide better visibility into site performance and
milestones to enable faster response to activation and enrolment misses.
Savings = Reduced frequency of SMV * costs (direct + indirect) per SMV](https://image.slidesharecdn.com/modelingresultsfromhealthsciencesdata-150117132710-conversion-gate01/85/Modeling-results-from-Health-Sciences-data-2-320.jpg)
Modeling results from Health Sciences data
- 1. Modeling Results from Health Sciences Data Judson R. Chase www.linkedin.com/in/judsonchase
- 2. Strategic Business Transformation Visibility into the Unknown • Site activation and enrolment modeling & simulation: Predict site activation and enrolment for all trials. Quickly identify regions and sites with greatest risk for activation and enrollment “misses”; allowing for early defensive action (i.e., adding more sites, modifying recruiting efforts) and minimizing margin burn-down. • Transition to a blended site monitoring model through risk-based monitoring: Target field and central monitoring according to risk; leveraging predictive analytics and business process management changes into a unique-to-industry site monitoring paradigm. • Confident bid defense: Assess prior contract profitability , along with trial and site metadata; and incorporate priors for site activation and enrolment in bid preparation and defense. Higher Business Agility & Scale Fastest to Actionable Insight Economies of Scale Universal Adaptability • Minimize duplication of effort through a central library of templates and analytics used by all departments, all therapy areas and all phases of clinical trials (Use cases: data review, adaptive monitoring, data cleaning, trial operations, safety assessments). [unquantified] • Productivity savings for clinical trial informatics: Reduce time and effort for data source combination, data preparation, report and dashboard creation & operations for setting up a new trial by at least 40%. Assume one person is working to set up each new trial. Assuming an estimated cost of $150K per FTE (fully loaded) leads to a cost saving of approximately $60K per year per FTE. (Use case: ability to scale up trials with limited headcount). Savings based on 50 new trials Lower Total Cost of Ownership Self-service Discovery • Accelerate CRA productivity and efficiency – Focus CRA’s on process- driven specific site tasks requiring action. • Streamline processes - involving data management, analysis/report preparation, dashboard creation and distribution of trial operations metrics. • Assuming 200 internal users and 15% productivity increase; this is the equivalent of up to 10.5 FTE‘s. Assuming average total loaded cost for FTEs across these roles is $150k, then potential cost saving is approximately $4,450,000 per annum. Potential of deploying these FTE‘s to new work will increase the value. • Bids: Improve awareness of margin indications (history/metadata on trial, regions, sites) when bidding for Sponsor trials. Savings = Reduced bid effort direct + indirect costs * number of bids. • Trial / Site Planning: Improve site/PI selection and site activation / enrolment modeling to drive better and predictable activation and enrolment. Savings = Reduced margin loss due to under-enrolling sites. • Execution: Transition from calendar to risk based monitoring. Savings = Reduced monitoring effort % * cost(s) of Phase III trials • Monitoring: Provide better visibility into site performance and milestones to enable faster response to activation and enrolment misses. Savings = Reduced frequency of SMV * costs (direct + indirect) per SMV
- 3. Clinical site monitoring can consume up to 30% of overall trial costs (Institute of Medicine. Assuring Data Quality and Validity in Clinical Trials for Regulatory Decision Making. Workshop Report. Washington, DC: National Academies Press; 1999. Obesity Trials 3% 5% 10% 15% CSM % 0f Overall Trial Costs $1.59 MM $2.66 MM $5.32 MM $7.99 MM Projected Headroom 1,2,3 1 Based on 30% of Phase III cost estimate of lowest estimated drug (Qnexa) overall trial costs 2 CSM % presented in four savings tiers; erring on the side of being conservative in projections 3 Projected headroom from risk based monitoring are displayed as $MM April 2012 - http://www.manhattan-institute.org/html/fda_05.htm
- 4. Clinical site monitoring can consume up to 30% of overall trial costs (Institute of Medicine. Assuring Data Quality and Validity in Clinical Trials for Regulatory Decision Making. Workshop Report. Washington, DC: National Academies Press; 1999. Diabetes Trials 3% 5% 10% 15% CSM % 0f Overall Trial Costs $.62 MM $1.03 MM $2.06 MM $3.08 MM Projected Headroom 1,2,3 1 Based on 30% of Phase III cost estimate of lowest estimated drug (Byetta) overall trial costs 3 Projected headroom from risk based monitoring are displayed as $MM April 2012 - http://www.manhattan- institute.org/html/fda_05.htm 2 CSM % presented in four savings tiers; erring on the side of being conservative in projections
- 5. Clinical site monitoring can consume up to 30% of overall trial costs (Institute of Medicine. Assuring Data Quality and Validity in Clinical Trials for Regulatory Decision Making. Workshop Report. Washington, DC: National Academies Press; 1999. Cardiovascular Trials 3% 5% 10% 15% CSM % 0f Overall Trial Costs $25.6 MM $42.6 MM $85.3 MM $127 MM Projected Headroom1,2,3 1 Based on 30% of Phase III cost estimate of lowest estimated drug (Eliquis) overall trial costs 3 Projected headroom from risk based monitoring are displayed as $MM April 2012 - http://www.manhattan- institute.org/html/fda_05.htm 2 CSM % presented in four savings tiers; erring on the side of being conservative in projections
- 6. “This combined strategy (Modified Site Management) resulted in a 21.1% reduction in total trial costs in our pharmaceutical industry simulations . . .” 1,2,3 Sensible approaches for reducing clinical trial costs. Clinical Trials 2008; 5; 75; DOI: 10.1177/1740774507087551 $33.5 MM $34.7 MM $39.8 MM Projected Headroom 1 Yusuf S. Randomized clinical trials: slow death by a thousand unnecessary policies? CMAJ 2004; 171: 889–92 3 Eisenstein EL, Lemons PW, Tardiff BE, et al. Reducing the costs of phase III cardiovascular clinical trials. Am Heart J 2005; 149: 482–8. April 2012 - http://www.manhattan- institute.org/html/fda_05.htm 2 Califf RM. Clinical trials bureaucracy: unintended consequences of well-intentioned policy. Clin Trials 2006; 3: 496–502. In connection with the management of clinical trials, we pay, on behalf of our clients, fees to investigators and test subjects as well as other out-of- pocket costs for items such as travel, printing, meetings and couriers . . . Amounts paid by us as a principal for out-of-pocket costs are included in direct costs and the reimbursements we receive as a principal are reported as reimbursed revenue. During the years ended December 31, 2008, 2009 and 2010, fees paid to investigators and other fees we paid as an agent and the associated reimbursements were approximately $319.0 million, $330.4 million and $380.7 million, respectively. 10-K (Annual Report) Filed 02/25/11 for the Period Ending 12/31/10 2008 2009 2010 10.5% 10.5% 10.5% Modified Site Management 4 4 10.5% selected as a conservative (versus use of 21.1% in study projected reduction) variable for calculated headroom
- 7. “In 2011, the average per-patient trial costs across all therapeutic areas in Phase I costs = $21,883, Phase II = $36,070, Phase IIIa = $47,523 . . . ” 1,2,3 July 2011 - http://www.pharmalot.com/2011/07/clinical-trial-costs-for-each-patient-rose-rapidly/ Model Site Activation 1,2 1 The average per-patient trial costs across all therapeutic areas, in Phase I, costs rose from $15,023 in 2008 to $21,883 in 2011. In Phase II, the cost rose from $21,009 to $36,070. In Phase IIIa, the cost increased from $25,280 to $47,523 and in Phase IIIb, cost jumped from $25,707 to $47,095. Finally, Phase IV expenses rose from $13,011 to $17.042. (July 2011 - http://www.pharmalot.com/2011/07/clinical-trial-costs-for-each-patient-rose-rapidly/). Phase II Patients Months Total Patient Costs End 2 months early End 3 months early 160 12 $5.8MM $.96MM $1.4MM Phase IIIa Patients Months Total Patient Costs End 2 month early End 3 months early 1000 24 $47.5MM $1.9MM $3.9MM Phase IIIb Patients Months Total Patient Costs End 2 month early End 3 months early 1600 36 $75.3MM $2.1MM $4.2MM 2 Patients and Months variables are representative; for illustration purposes only . 3 The amount of direct savings varies depending on trial specifics and capabilities. Modeling site activation translates into months saved in enrollment cycles which more importantly translates into direct savings and operating costs.
- 8. This document purposefully illustrates my view only on a premise versus any specific business, any specific product or any specific service recommendation. This document is provided for informational purposes only and not as any sort of recommendation (other than the author’s knowledge and skill sets). This document is provided for informational purposes only and its contents are subject to change without notice. The author makes no warranties, express or implied, in or relating to this document or any information in it, including, without limitation, that this document, or any information in it, is error-free or meets any conditions of merchantability or fitness for a particular purpose. Perceived, specific content themes, names, characters, businesses, places, events and incidents are either the products of the author’s imagination or used in a fictitious manner. Any resemblance to actual names, characters, businesses, places, events and incidents is purely coincidental. Any data referenced in slides is publically available data and should not be perceived as having any confidential or proprietary nature. Presentation themes and backgrounds have been purposefully removed for posting to a public forum. Original document may have included title pages, titles, borders, etc., lending a more professional, guided experience. This document has been marked as final for the purposes of posting to a public forum. Typing, editing commands, and proofing marks are disabled and the document is read-only. This document may not be reproduced or transmitted in any form or by any means without author’s prior written permission. PUBLIC DOMAIN PUBLISHED CONTENT DISCLAIMER