Human factors in GMP (7 February 2014)

Human Factors in GMPAJAZ S. HUSSAIN, PH.D. | INSIGHT, ADVICE & SOLUTIONS, LLC

2/7/2014 [email protected] 1

Outline

Background

• Safety, Quality & Behavior

• Multiple-decision levels & proactive risk management in a dynamic society

• Human Factors Analysis and Classification System

Challenge of a ‘cluster’ of poor data integrity practices

• Noted 483 observations - Misguided or malicious?

• Need for improving CAPA?

Question

• What considerations will enhance preventive system and controls?

• Recommendations


Safety, Quality & BehaviorPaul O'Neill and the

story of Alcoa• The Power of Habit: Why We Do What We Do in Life and Business.

Charles Duhigg

Human reliability • Assessment tools & analysis of ‘accidents’

• Human Factor Analysis and Classification System

Concepts of ‘proactive risk management’

• Preventive controls

Applications and results in medical settings

• Reduction in health care process deviations and errors

[Occupational] Safety Climate and Quality

Culture• Similarities and relevance of concepts and tools


A ‘process of migration’ to explain how accidents can occur


Boundary to unacceptable

workload

Boundary to economic failure

Boundary to functionally acceptable

performance

Safety Zone

Gradient towards least effort

Gradient towards cost effectiveness

Rasmussen (1997)

“Proactive Risk Management in a Dynamic Society”*: Many nested levels of decision-making

Government Regulators Company Management Staff Work


Politicalclimate &

public opinion

Changing marketconditions and

financial pressure

ChangingCompetency & levels

of education

Changing product mix, technologies, and

regulatory requirements

Environmental pressure

Laws (e.g., FDASIA) Regulations

JudgmentCompany Policy Plans Action

Judgment Judgment JudgmentJudgment

Political Science;Law; Economics;

Sociology

Economics;Decision Theory;Organizational

Sociology

IndustrialEngineering;

Management &Organization

Psychology;Human factors;

Human-MachineInteraction

Chemistry;Pharmacy;

Biotechnology; Engineering,…

Research disciplines

Analysis Public opinion Incident or accidentReport Observations, DataLogs, ReportsOperations review

*Jens Rasmussen & Inge Svedung (2000)

Human Factors Analysis and Classification System

2/7/2014 6

Latent Conditions

Latent Conditions

Latent Conditions

Active Conditions

Failed orAbsent Controls

OrganizationalInfluences

Failure inSupervision

Preconditionsfor

Deviations

Deviations

Multiple contributors (the holes in cheese slices) must be aligned for any adverse events to occur. Barriers in a system (the slices themselves) are intended to prevent errors that result in these deviation.

Reason, J. (1990) The contribution of latent human failures to the breakdown of complex systems. Philosophical Transactions of the Royal Society (London), series B. 327: 475-484.Shappell & Wiegmann. (2000). The Human Factors Analysis and Classification System—HFACS. FAA. US Department of TransportationBogner, M. S. (2002). Stretching the search for the “why” of error: the systems approach. Journal of Clinical Engineering, 27, 110-115.

[email protected]

The Swiss Cheese analogy is useful

• the complexity of failure

• structure your search for distal contributors

• the effort it takes to make and keep a system safe

To think about

• where the holes are or what they consist of,

• why the holes are there in the first place,

• why the holes change over time, both in size and location,

• how the holes get to line up to produce an accident.

The layers of defenses should not be assumed to be static or constant or independent of each other. Investigations

need to consider


Dekker, S. (2002). The Field Guide to Human Error Investigations. Ashgate

Challenge of a ‘cluster’ of poor data integrity practices

Suspects are cultural attitudes, human hardships in an emerging economy, and/or growth ambition beyond what current systems

Effective correction and prevention necessitates a deeper understanding to identify the root cause(s) and to inform policy considerations at the organizational, regulatory and political levels

There is a serious and lingering concern that there may be a set of common factors and that such practices may be more widespread

More such practices may be noted as rigor and frequency of inspections is increased

Many pharma companies in India have state-of-art quality systems

However, a ‘cluster’ of poor data integrity practices were noted at several companies in India


An illustrative list of observations (FDA 483)

During inspection of your firm we observed

• “…trial injections…..”

• “…results failing specifications are retested until acceptable results are obtained….”

• “…over-writing electronic raw data…..”

• “…OOS not investigates per XYZ SOP”

• “…appropriate controls not established….”

• “….records are not completed contemporaneously”

• “…observed analyst back-date logbooks”


Malicious or misguided?

“….records are not completed

contemporaneously”

Misguided

Malicious


Behaviors and antecedentsPast behavior

Attitude towards behavior

Subjective NormsPerceived behavioral

controlIntention Future behavior

Shop-floor

Supervisor

Middle Management

Senior Management

Shareholders

Regulators (local)

Regulators (aboard)

Media

Health care providers

Patients


Attitude towards the behavior

Subjective norm

Perceived behavioral control

IntentionFuture

Behavior

usually found to predict behavioral intentions with a high degree of accuracy

intentions, in combination with perceived behavioral control, can account for a considerable proportion of variance in behavior.

Ajen, I. The theory of planned behavior. ORGANIZATIONAL BEHAVIOR AND HUMAN DECISION PROCESSES 50, 179-211 (1991)

PastBehavior

At the individual level, in a manufacturing/QC function– how frequently does this occur?

attitude toward performing the

behavior

Process validation is

done so quality is good;

test prone to error

“Batch failure means I made a

mistake”

the subjective norm

documentation not critical;

Compendial testing sufficient

Indian regulators collect & test samples – no issue there


Note: Organizational dynamics - manufacturing status below R&D (if so; this needs to change)

Why some pharma professionals may ‘rationalize’ deliberate non-compliance?

FDA 483 (05/02-07/01/2005): 12

Observations

March 9, 2007: Four pleaded

guilty - “duping the FDA for six

years.”

Federal Register Vol. 77, No. 66,

April 5, 2012 Notices :

Debarment

2/7/2014 [email protected]

Penalty is significant in the US!!Companies often don’t survive!!

Three case examples – from experience at FDA Preconditions for violation (I & II) and Successful control of preconditions (III)

Case IView point: Observer

“Conspiracy to commit an offense against the United States”

Bankruptcy and debarment of several individuals

Case IIView point: Expert witness for the prosecution

“Criminal prosecution” Bankruptcy and…

Case IIIView point: Arbitrator; to avoid drug shortage

Company complied with cGMPs for the product before a specification was changed (FDA/USP)

Had to re-develop their products to comply with cGMPs


Case I & II vs. Case IIILaboratory records do not include complete

data

Input to and output from the computer and records or data are not checked for accuracy

Employees are not given training in cGMP & written procedures

required

The quality control unit lacks authority to fully

investigate

Products failing to meet specifications not

rejected

Controls on processes not established

Root-cause analysis not done or is inconclusive

Annual report did not include reports of

investigations

2/7/2014 15

http://www.fda.gov/AboutFDA/CentersOffices/OfficeofGlobalRegulatoryOperationsandPolicy/ORA/ORAElectronicReadingRoom/ucm061813.htm

Quality control unit adequate authority

Adequate training in cGMP…..

Adequate records and checked

Laboratory records complete

Cas

e I &

IIC

ase

III

Case I:

[email protected]

http://www.fda.gov/AboutFDA/CentersOffices/OfficeofGlobalRegulatoryOperationsandPolicy/ORA/ORAElectronicReadingRoom/ucm061813.htm

What was the difference?

Behaviors at all levels

Empowerment through adequate training and visible

management support

Encouragement and means to escalate issues

without fear

Clear description of responsibility & accountability

Effective supervision


Preconditions to malice or disregard

Rationalization/Attitude

Pressure/Incentive

Opportunity/”Holes in the Cheese”


Organization & failure in supervisionO

rgan

izat

ion

Environment

Structure

Policies

Culture

Process

Oversight

Procedures

Operations

Resources

Human

Monetary

Facilities

Failu

re in

Su

per

visi

on

Unforeseen

Known

Inadequate

Planned inappropriate operation

Failure to correct

Violation


Co

mp

etency?

Ma

lice or

a d

isrega

rd?

Preconditions, deviations and violationsPr

eco

nd

itio

ns

for

Dev

iati

on

s

Environmental

Technological

Physical

Condition of operators

Cognitive

Psych-Behavior

Physical state

Personnel factors

Self-imposed stress

Planning, communication,..

Dev

iati

on

s

Error

Skills-based

Decision

Perceptual

Violation

Routine

Exceptional


System-w

ide?

Ind

ividu

al?

An example…Sociological contributions to the study of adverse events (hospital setting) and relevance to GMP

The greater the number of individuals (departments) involved in the care of a patient, the higher

the risk of adverse events.

[GMP – In a QC department multiple analysts share the same tasks without clear definition of accountability & responsibility]

The more complex (technologically sophisticated, demanding specialist expertise) the tasks involved in the care of a patient, the higher the risk

of adverse events.

[GMP – Alarms, data-interfaces & controls on newer fluid-bed ignored or not understood]

The more easily a named individual can be identified as responsible for coordinating the care of a patient,

the lower the risk of adverse events.

[GMP – sharing of password for data entry (or file deletion)]


Quality in Health Care 2000;9:120–126

An example…Sociological contributions to the study of adverse events (hospital setting)

The greater the emphasis placed on arrangements for formal communication, the

lower the risk of adverse events

[GMP-Day to day operations tend to have significant

informality at the shop-floor]

The more that status distinction is observed among professional groups and between men and women in an organization, the

higher the risk of adverse events

[GMP- Sharp status distinction in Indian manufacturing

organization settings]

The greater the environmental pressure on an organization to achieve targets that are not directly related to quality of care, the

higher the risk of adverse events

[GMP- Attitude, subjective norms towards GMP-

Quality….]


Quality in Health Care 2000;9:120–126

Rationalization/Attitude: Consider a facilitated and informed organizational discussion, debate and training

documentation not critical;

Compendial testing sufficient

Indian regulators collect & test samples – no issue there


Are these our organizationsBlind-spots?

Why this poses high risks?

What we do now and the risk it poses to us, our families, & our customers?

How to we ensure we reduce this risk at every level?

Intention

Perceived risk &

controls

Subjective norms

Attitude

Strategic response neededWhat is the current US FDA risk-profile of your company?

◦ Specifically in the context of quality culture and assurance of data integrity?

What are the key priority issues?◦ Those that require the full and immediate attention of the entire management team

What should be the objectives?◦ That clearly state what your organization must achieve to address the priority issues

Who will be accountable?◦ Senior Management Team members who will address the priority issues and achieve the defined

objectives

How will you review the implementation of the strategic response?◦ A formal process to hold regularly scheduled reviews of the plan and to refine it as necessary


A few recommendations: To enhance preventive system and controls?

Prevent

Detect

CorrectLearn

Improve


Prevent and Detect

Preventive controls (include the latent variables)

• An example: Tardiness in biometric attendance logbook

• Variability in human processing time and sequence (which may not be part of established SOP)*

• Training , tools (e.g., sms), encouragement and incentives for self-reporting

• Employee certification

• Structured review meeting (weekly/monthly) “Variances that are not necessarily Deviations” - why and how, improvement targets and results to be reviewed at the next meeting. Quarterly review by the supervisory chain of command. Link to performance incentives and development plan.

Rapid detection & response

Clearly defined self, peer, and supervisory detection criteria, reporting process and expected response and time to respond

Assess and tack product impact and consumer complaints

Plus Compliance monitoring

Compliance audits


*The inclusion of the human factor is focused on the manifestation of various behavior patterns. Behavior patterns can be derived from the analysis of user related data (including self-reporting)

Correct

Develop tools, skills and

processes for root cause

investigations

Metrics for effectiveness and time to complete

Create appropriate data and knowledge

bases

Development and validation data

considered in this analysis

Escalation process

Communicate resolutions and

lessons to be learned to ensure

system wide implementation


LearnContinuous

process verification and risk

assessment

Reports and actions

Policies, practices and procedures

against comprehension

Individual and Systems

improvement

Metrics and benchmarks


Improve Leader framing and ‘walking the talk’

• Incentives and rewards (tangible and intangible)

• Communication – ensure the right subjective norms, attitudes and behavioral controls

Quality policy and risks: Review and communicate (quarterly)

• Continuous process verification and risk assessment & product impact and consumer complaints

• Risk map - Employees, business process, manufacturing process and products and link to development plan

• Action plan to ensure risk-mitigation plans, policies, procedures are aligned

• Review plans for individual and systems improvement

• Metrics and benchmarks

• Communicate benchmarks, accomplishments, improvements needed and targets


Summary: Importance of Human Factors in GMP

Corporate quality culture will be a focus area for regulatory scrutiny

•Suspicion of data integrity can be a significant challenge even at companies that have rigorous corporate quality polices

People and facilities need to be ready every day a product is manufactured

•Rigorous, frequent, targeted and coordinated cGMP inspections by several regulatory authorities

Quality Metrics

•A useful tool for improving performance and also an effective tool to detect “too good to be true situations”

Question based review to ensure quality by design; a higher need for clear and effective communication

• Internal review and cross- functional alignment will need to address evolving requirements to reduce risk of failure
