Research REPORT

DATA:
The land of
lost opportunity
Survey reveals that despite a growing
awareness of data’s value, much of pharma
has not capitalized on the potential offered
by new data management technologies

Sponsored by
 Pharma Manufac tu rin g: Resea rc h REPORT

p Safely and efficiently manufac- and use data to its full potential. CURRENT LANDSCAPE
turing drugs is predicated on not Results revealed a need for a The pharma industry has always
only collecting, but leveraging the validation-ready, integrated been data rich. Drugmakers col-
right data. system for data aggregation, lect massive amounts of data at all
Process data, which can be contextualization and analytics, in stages of the process and rely on
defined as time series data gener- order to ensure data integrity and that data to do everything from
ated by process control systems, sound decision-making. make decisions to advance or halt
laboratory systems, and other
plant floor equipment, has EXHIBIT 1
increased dramatically in phar-
Why are you using your data management solution?
ma’s new digital age. But without
the right data management sys-
tems in place, data is nothing more   34%  
Because of regulatory guidelines
than spreadsheets of wasted oppor-
tunity for improvement.
  35%  
Are pharma manufacturers ade- For CPV / Proce ss and Product Monitoring
quately equipped to not only meet
industry data demands but to use   22%  
this data to improve operations? For APQR creation
The technology exists — including
predictive analytics, robotic process   44%  
For process understanding
automation, and AI-based tools, all
digitally connected via the Inter-
  62%  
net of Things (IoT) — but many For process monitoring
pharma companies are struggling
to fully leverage all this data.   54%  
For process optimization
To better understand the
pain points surrounding data
  46%  
management and software For process robustness/improvement
systems, Pharma Manufacturing,
in collaboration with BIOVIA,   34%  
recently conducted a survey of For process standardization

pharma manufacturers regarding

their process data solutions. The   12%   For “Golden Batch” identification
survey found that while pharma
understands the value of process
data, work still needs to be done in
  12%   For collaboration with partners

terms of adopting the right systems

— particularly systems that reduce   8%   No solution
compliance risks and inefficiencies,

www.Ph a r m a M a nufa c t ur i ng.c om

- 2-
 Pharma Manufac tu rin g: Resea rc h REPORT

the development of compounds, EXHIBIT 2

to understanding the efficacy Rank the level of importance of the

of marketed treatments on following features of a manufacturing
real patients. production/process analytics solution
But as the industry’s digital
transformation progresses, Most important Least important

data is piling up. Specifically, as

Regulatory compliance capability 54% 0.8%
equipment becomes “smarter” and
more connected, more process Validation readiness 41% 2%
data is being captured. It will
Integration with lab solutions 34% 2%
probably come as no surprise to
those in the industry how much Total cost 32% 1%

of pharma is managing all of this Integration with quality solutions 32% 3%

data. According to survey results,
Complete feature set in one package 29% 2%
close to 60 percent are still using
Excel to aggregate and analyze Lot/batch/material genealogy 25% 2%
process data. Excel is universal —
Automated document generation/population 25% 2%
it’s familiar, easy to use and easy
to understand — and prior to the Initial cost 24% 2%

explosion of big data, got the job Process-centric view 19% 7%

done for pharma.
Easy data exchange with partners 18% 10%
Excel, however, has limitations
— largely tied to the fact that Dashboarding 17% 4%
it’s a very manual method for
Flexible pricing/licensing models 15% 3%
data management. The survey
also revealed that 32 percent of Web-based client 12% 9%
respondents are using Minitab, a
Multi-tenant cloud architecture 11% 10%
statistical software. The ubiquity
of Minitab is likely tied to the Single-tenant or private cloud architecture 6% 13%

prevalence — and deficiencies —

of Excel. Since Excel cannot do academia, it is a familiar choice multiple commercial off-the-shelf
the advanced statistical analyses to many entering the pharma software packages, installed on
required in manufacturing, an industry from universities or internal systems. In order to meet
external tool is required. Minitab other companies. the complex, specific needs of the
is highly compatible with Excel, Just under 19 percent of those pharma industry, these multiple
allowing for auto import of data surveyed noted the use of in-house systems were cobbled together
from Excel spreadsheets into the or bespoke systems to manage internally by onsite IT staff. This of
Minitab system. Because Minitab data. This too is not surprising, as course becomes problematic when
is extraordinarily common in other traditionally, software models in it comes to updating or expanding
process industries as well as in pharma involved the purchase of the use of these systems.

www.Ph a r m a M a nufa c t ur i ng.c om

- 3-
 Pharma Manufac tu rin g: Resea rc h REPORT

For pharma, the noted between different teams, functions MOTIVATORS

methods of data management and partners reap substantial In order to more closely explore
largely represent “how it’s always strategic benefits. Disjointed pharma’s data management strug-
been done.” Culturally, change systems limit a company’s gles, it’s important to understand
in pharma can be difficult. ability to improve productivity the reasons behind pharma’s use
Financially, change is often and efficiency. of specific data management
perceived as costly. But this In terms of specific challenges solutions. (See Exhibit 1). Con-
unwillingness to change does not related to aggregation and analysis sidering how highly regulated
come without obstacles for the of process data, 61 percent of those the industry is, it’s interesting to
industry, many of which were surveyed say they still have data on note that survey results show that
highlighted in our survey results. paper — the top challenge noted. the desire to better control the
This was followed by 42 percent process (in the form of process
SURVIVAL CHALLENGES who struggle with validation. This management, process optimization,
Not surprisingly, pharma’s cur- is a situation where one challenge process improvement and process
rent state of data management has leads to another. Paper-based understanding) eclipse regulatory
resulted in myriad struggles. validation is a tedious, time-con- compliance as a prime motivator.
The situation is further compli- suming process that requires a What this likely reflects is that
cated by a lack of consistency when significant amount of resource pharma is no longer just managing
it comes to implementing data commitment from the validation data for the sake of compliance; the
management systems. Just 36 per- and document management teams. industry seeks to go beyond this
cent say they use the systems they It also creates delays in releasing and derive insight and improve-
identified for “some” products and critical manufacturing equipment ments from this data. Of course,
processes, while 23 percent say they for manufacturing operations. compliance remains top of mind:
use only the data solutions for a This type of validation also When it came to ranking features
“select few” products and processes. increases risk. If Excel spreadsheets of a manufacturing production/
Under 41 percent report using the are being utilized, as is often the process analytics solution (See
solutions identified to manage all case, they need to comply with a Exhibit 2), half of respondents
products and processes. number of requirements. Controls say that “regulatory compliance
Additionally, data management are quite often inadequate and this capability” (electronic records/
techniques vary by production site. may result in compliance risks. signatures, audit trails, user/device
Only around 42 percent are using Many respondents also noted controls, etc.) is the most important
the solutions identified across all disconnects, pointing out that data feature needed in a solution. “Val-
their company’s sites. Included aggregation and analysis are done idation readiness,” a feature closely
in this number are the rare six in separate systems that are not linked to compliance, is also being
percent that are using the same integrated (33 percent), disconnects prioritized by the industry. Other
system across their entire network, on departmental levels (30 percent) standouts in the results affecting
including partners. This highlights and data silos (28 percent) — all the decision-making process for
a huge missed opportunity in are perils of manual systems where data solution choice were costs
terms of collaboration. Companies collaboration is difficult, slow (initial and total), completeness of
that are able to share quality data and inconsistent. the feature set, and single-tenant

www.Ph a r m a M a nufa c t ur i ng.c om

- 4-
 Pharma Manufac tu rin g: Resea rc h REPORT

or private cloud architecture. companies report applying pre- The better news is that 28 per-
These latter responses could be dictive modeling in specific areas cent are in fact taking action,
interpreted as disconnect between only, such as adjusting process reporting that they are currently
what pharma companies expect to conditions based on raw material expanding use of their data man-
pay for a data solution and what characterization, forecasting cam- agement solutions.
vendors are able to provide at that paign outcomes for supply chain At the same time, around 41
price point. Alternatively, they purposes or correcting downstream percent feel confident that their
may simply highlight the common operations based on upstream companies are “on track with pro-
drivers in our industry and reflect process conditions (the most popu- cess understanding and meeting
the trend toward high ROI, plat- lar reason). its objectives.” This was further
form-based solutions. A shared predictive modeling corroborated by the 49 percent
environment — where teams can who report “no plans to change”
COMPLIANCE AND BEYOND collaborate effectively to bring their current use of data manage-
Predictive modeling represents the better treatments to market faster, ment solutions.
next-level in data management, at lower cost and in a way that is It’s encouraging that more than
offering an accurate and completely auditable and compliant half of respondents acknowledged
efficient method of applying and — requires giving up manual-based their company’s need to do more
interpreting data. Predictive systems and transitioning to a digi- with process understanding to
process modeling uses advanced tal technology platform. meet its objectives, and reflects
technology to make predictions a growing understanding of the
about process points, such THE NEXT STEPS importance of data as a key busi-
as performance indicators or In pharma, there is often a gap ness asset. But a large percentage
quality attributes, with the goal between aspirations and actions. of the industry still remains either
of improving and optimizing When survey respondents were unaware of or unconvinced of the
processes. Today’s predictive asked how they felt their compa- value of solutions that can reduce
modeling software allows users nies were doing meeting quality, inefficiencies, risks and costs.
to flag risks in real time, thereby productivity and profitability There is often a perception that
reducing or eliminating them. objectives for manufacturing, many the costs to change to a different
Unfortunately, pharma can’t (56 percent) acknowledged that system are higher than the poten-
reach this level of data management there were still deficiencies, noting tial payoff — in pharma, many are
by using manual-based systems. that their “company needs to do still disconnected from the value
Although modeling is relatively more with process understand- that such a change would bring to
established in API-production, ing to meet its objectives.” When their business and their customers.
most pharma plants still lack the broken down by manufacturing This value can come in many
data infrastructure required for sectors, those focusing on small different forms. In the quality
predictive process modeling. molecules were the least satisfied, assurance and control area, com-
Survey results indicate that much with 64 percent feeling that their panies can use new data-driven
of pharma has not reached the companies need to do more. This technologies to better understand
predictive modeling stage — 41 awareness is an important first step how input parameters such as
percent responded as such. Some towards improvement. machine settings, operator training

www.Ph a r m a M a nufa c t ur i ng.c om

- 5-
 Pharma Manufac tu rin g: Resea rc h REPORT

levels, or raw material options forecast demand, and automatically capitalizing on the benefits offered
will affect product quality and identify and mitigate supply risks. by new data solutions. p
outcomes. On the plant floor, dig- Big data is getting bigger by the
ital operations reduce the risk of minute, and there is tremendous — Karen Langhauser,
human-error. In the supply chain, potential for process improvement Chief Content Director,
Pharma Manufacturing
pharma operations executives can hidden in these piles of data.
leverage big data, external and Recognizing this is the first step
internal indicators, and machine — but truly evolved data manage-
learning algorithms to better ment requires going beyond this,

SURVEY DEMOGRAPHICS

EXHIBIT 3 EXHIBIT 4

What types of products does Which functional area most closely

your company manufacture? matches yours in your company?

  28.9%  
APIs

  25.6%   26.7% Quality / Regulatory

Small Molecule
17.5% Mfg / Process Sciences
and Technology
  30.6%  
Biologics 12.5% Manufacturing

10.0% Product /
  9.9%   Personalized Medicine Process Development

7.5% Corporate Management

  28.9%  
Combination Products 6.7% Supply Chain

5.0% (Technical) Operations

  38.8%  
Generics/Biosimilars
4.2% Information Technology

  17.4%  
Other 4.2% Project Management

4.2% Automation Engineering

1.7% Other

www.Ph a r m a M a nufa c t ur i ng.c om

- 6-