Macedonian pharmaceutical bulletin, 65 (1) 39 - 60 (2019)

ISSN:
UDC: 614.35:519.2
DOI: 10.33320/maced.pharm.bull.2019.65.01.005
Professional paper

Trending and Out-of-Trend results in

pharmaceutical industry

Maja Velinovska Cadinoska1*, Nada Popstefanova1, Miroslava Ilievska2,

Elizabeta Karadzinska1, Marija Davceva Jovanoska1, Marija Glavas Dodov3
1
Alkaloid AD Skopje, blvd. Aleksandar Makedonski 12, 1000 Skopje, R. N. Macedonia
2
GMP Expert, Mite Bogoevski 64А, 1000 Skopje, R. N. Macedonia
3
Institute of Pharmaceutical technology, Faculty of Pharmacy,
Ss. Cyril and Methodius University, Majka Tereza 47, 1000 Skopje, R. N. Macedonia

Received: March 2019; Accepted: May 2019

Abstract

There are increasing demands from both, regulatory agencies and industry itself, for improving the quality monitoring procedures.
This focus on quality and risk management has prompted for re-evaluation of the systems and procedures to ensure compliance with the
proposed guidelines. One of the essential tools for quality monitoring is statistical tools with which we can support any conclusion with
regard to the variability and capability of a given process and thus we can ensure a state of control. There are numerous techniques which
can be used and one of them is Statistical Process Control (SPC). SPC is a highly structured approach for defining what is important, by
process screening; data collection and measuring the data; data analyzing and finding the variations; what needs to be done and how to
bring the process in control and maintain the state of control. Trending is part of SPC and enables to see the general direction towards
which the process is moving. Trending is used as a prevention of possible out-of-specification results and nonconforming products that
could have impact on the patient’s health. It is no longer acceptable to be only within specification limits and out of process control.

Keywords: trend, trending, control charts, SPC, variable data, discrete (attribute) data, standard deviation, upper control limit, lower
control limit

development and optimization through technology

Introduction
transfer and scale up to commercial manufacturing with
the implementation of the three quality quidelines
Ever since The International Council for
achieves the desired quality - “fit for its purpose” (Jean-
Harmonization of Technical Requirements for
Louis et all., 2010). Good scientific development (Q8) in
Pharmaceuticals for Human Use (ICH) guidelines Q8 -
combination with QRM (Q9) and PQS (Q10) could
Pharmaceutical Development, Q9 – Quality Risk
improve the drug quality and efficiency during
Management (QRM) & Q10 - Pharmaceutical Quality
pharmaceutical manufacturing. As a result, there were
System (PQS) have been officialy published major
numerous of changes in good manufacturing guidelines
changes were going through pharmaceutical industry.
from different agencies and different territories (for
From an early phase of development, defining Quality
example, Food and Drug Administration Department of
Target Product Profile (QTTP), pre-formulation studies, Health and Human Services (FDA) in United States of
formulation screening and development, process America (USA), EudraLex: The Rules Governing
__________________
* [email protected]
40 M. Velinovska Cadinoska, N. Popstefanova, M. Ilievska, E. Karadzinska, M. Davceva Jovanoska, M. Glavas Dodov

Medicinal Products in the European Union, Good one Standard Operating Procedure (SOP) for regulating
Manufacturing Practice for drug products (GUI-0001), OOS; but while the OOS could be defined as a situation
Health Canada etc.). where the result clearly is out of specification, OOT and
In 1960, Dr Genichi Taguchi introduced a new OOE are more related to predicting such situations.
definition of World Class Quality. It means “On-Target” Therefore, numerous procedures and guidelines were
with “Minimum Variance”. Operating “On-Target” written for detecting and handling OOT results (Burgess
requires a different way of thinking about the processes. et al., 2015; Oakland, 2003; Thomas et al., 1982).
Operating with “Minimum Variance” could be achieved
only when the process displays a reasonable degree of
statistical control (Education SPC course, 2016; Oakland, Regulatory requirements
2003).
The pharmaceutical industry is known that is both, The regulatory requirements for monitoring of
highly regulated and capital intensive, with significant product quality are laid down in the current guidelines of
funds for development and manufacturing of the good manufacturing process (cGMP) to ensure that a state
medicinal products. Yet, when compared to other of control is maintained throughout the product lifecycle
industries, the quality assurance in pharmaceutical with the relevant process trends evaluated. In the cGMP
industry is not highly proficient. For example, the quality guidelines of EU, Volume 4 Part I (EU cGMP), trending
defects and batch failures range from 3 – 15% and waste is addressed in several chapters. For example, Chapter 1:
has been reported as high as 50%. In comparison, the Pharmaceutical Quality System, point 1.10, which relates
semiconductor industry maintains waste below 1% to product quality review (PQR) states that one of the
(Winkle, 2007). purpose of the PQR is to highlight any trend and to
PQS should enable system to enhance quality and identify product or process improvement. As stated in
availability of medicines. As stated in ICH Q10 (Fig. 1), Chapter 6, Quality Control, data obtained during release
one of keys objectives of PQS is to encourage testing of the batch (e.g. assay, dissolution, impurities
manufacturers to develop effective monitoring and control etc.) or during environmental controls should be recorded
system for process performance and product quality, in a manner permitting trend evaluation. During data
thereby providing assurance of continued sustainability evaluation any OOT result should be addressed and
and capability of the processes. subject to investigation. Chapter 6 also addresses the
importance of trending the data which are obtained during
stability testing. It is pointed out that any significant
negative trend should be not only investigated but also
reported to the competent authorities. Trending during
stability testing can be used as a proactive approach for
evaluation of changes during time, which may help
predicting future trends.
EU cGMP Annex 15: Qualification and validation
and FDA Process Validation Guideline: General Principal
and Practices specify that manufacturers must establish
ongoing program to collect and analyze product and
process data that are related to product quality and the
data should be statistically trended and reviewed by
trained personnel.
Data integrity guidelines from Medicines &
Healthcare products Regulatory Agency (MHRA) ‘GXP’
Fig.1. ICH Q10 Diagram 1. (Ref. ICH Guidelines Data Integrity Guidance and Definitions and World
https://www.ich.org/fileadmin/Public_Web_Site/I Health Organization (WHO) Guidance on Good Data and
CH_Products/Guidelines/Quality/Q10/Step4/Q10 Record Management Practices recommends that data
_Guideline.pdf). should be reviewed and, where appropriate, statistically
evaluated after completion of the process in order to
determine whether outcomes are consistent and compliant
Good manufacturing practice guidelines require the with established standards. The evaluation should take
processes Out-of-specification (OOS), Out-of-expectation into consideration all data, including atypical or suspect
(OOE) and Out-of-trend (OOT) to be fully understood. It data or rejected data, together with the reported data. This
is no longer acceptable to be within the specification includes a review of the original paper and electronic
limits but out of statistical control because in such cases records.
the probability of producing a defective product is high. The insight of the regulatory inspection findings
Recently, these three terms were interlaced together, in shows that the regulators expect useful trend analysis to

Maced. pharm. bull., 65 (1) 39 – 60 (2019)

 Trending and Out-of-Trend results in pharmaceutical industry 41
be performed, in their publications regarding inspection of There should be written policies, procedures,
compliance with good manufacturing practice deficiency protocols and reports and associated records of measures,
data trend several types of deficiencies are related to or results from trending and OOT investigations. The
trending and OOT result. For example, not all data were same principle of good documentation practice, which is
monitored by trend (e.g. IPC), or the monitoring is limited given in EU cGMP Chapter 4: Documentation should be
to graphical presentation of data; statistical tools for followed, there must be written specification, test
tracking the trend have not been applied protocol and test report. The specification should describe
correctly/completely (e.g., only one WECO rule has been which parameters should be subject to trending analysis
used). In the stability studies, significant negative trends such as: data obtained from assay analysis, dissolution,
have not been reported in the Agencies, only OOS results; other quality metrics, key performance indicators (KPI).
which is considered as a major finding (OOT Forum, The parameters should have defined acceptance criteria.
2015). The purpose of the analyzing and monitoring should be
described, for example, whether it is drifting of the results
toward the specification limits, or increased degradability
Methodology of the active pharmaceutical ingredient, etc. (Burges et
al., 2015; OOT Forum, 2015).
In order to explore the trending techniques and to
detect OOT results we performed literature review within
database of governmental agencies, non-governmental Defining the parameters that should be trended
organizations and current legislation (guidelines, standard
operating procedures, books and handbooks, as well as Quality risk management is a tool for assessing which
materials from educational courses and training parameters should be monitored. There are several risk
workshops). management tools however; the selection of particular
risk management tools is completely dependent upon
specific facts and circumstances. ICH Q9 gives guidance
Definition of trend and trending on the approach and potential application. Mostly it
depends of the process complexity, for process
Trend is a sequence of time related events. It shows monitoring mostly used techniques are Failure Mode and
the general direction towards which a particular Effects Analysis (FMEA), Failure Mode, Effects and
situation/data is moving. In order to see if there is a Criticality Analysis (FMECA) and Hazard analysis and
certain direction in their movement, as well as the extent critical control points (HACCP). As a rule, all data, which
to which the data changes, procedures are established are defined as critical, should be included in trending. For
which enable collection of data during a certain time the products which are developed using Quality by
period and analyzed. Trend analysis refers to techniques Design (QbD) principle the defined CQAs are the ones
for detecting an underlying pattern of behavior in a time that should be trended. For the products which are
or batch sequence, which would otherwise be partly or developed using traditional approach and for legacy
nearly completely hidden by noise. These techniques product the CQAs are defined using risk management.
enable specific behaviors such as a shift, drift or excessive For example, assay of drug substance, dissolution, system
noise to be detected. OOT is test result or pattern of suitability, critical in-process controls such a tablet’s
results that are outside of pre-defined limits, historical or hardness, critical process parameters on batch by batch
expected results. Trending refers to the techniques of data basis, OOS results, deviations, batch rejection, reworks
collection; the analyzing the collected data and the etc. should be evaluated (Burges et al., 2015; Education
patterns that they follow and techniques of OOT results course SPC, 2016; ICH Q9, 2006; OOT Forum, 2015).
detection (Laney, 2002; Oakland, 2003; OOT Forum,
2015).
Good trending reflects manufacturer’s profound How to perform trend analysis?
knowledge of the process and thus build confidence in the
manufacturer’s quality risk management. Trending may There are various statistical tools of analyzing data;
mitigate the risk of failure to comply with marketing one of them is Statistical Process Control (SPC), a
authorization, which can lead to product recall, or failure scientific method of analyzing data and using analysis to
to comply with cGMP guidelines, which can lead to solve practical problems. SPC means with the help of
withdrawal of GMP certificate (Burges et al., 2015). numbers, or data, we study the characteristics of our
In practice, the expectation is that there is no trend of process in order to make it behave in a way that we want
the data (for example, for production or analytical process to behave. SPC gives insight when specific conditions for
data). However, sometimes the trend is expected, for variation occur in the process and build in the system if
example in stability testing (Burges et al., 2015; OOT there is a positive effect on the process or eliminate if
Forum, 2015).

Макед. фарм. билт., 65 (1) 39 – 60 (2019)

42 M. Velinovska Cadinoska, N. Popstefanova, M. Ilievska, E. Karadzinska, M. Davceva Jovanoska, M. Glavas Dodov

Traditional SPC
Approach New way of thinking
Customer Continuous
Objective
requirement Improvement

Strategy Short Term Long term

Common Cause,
Data Analysis Variances, OOS
Special Cause
Relationship Advisory Cooperative
Everybody who is
Responsibility Quality Control &
involved in the
for Quality Quality Assurance
process
Effect on
Quality ↑ Cost↑ Quality ↑ Cost↓
Productivity
Management’s Coach, Trainer,
Fire Fighting
Role Sponsor

Fig. 2. SPC quality approach 1 (Ref. Education course SPC, 2016).

there is a negative effect on the process (Burges et al., special cause and establishing the boundaries in which we
2015; Education course SPC, 2016; Oakland, 2003; OOT want the process to move. It means to get a stable process
Forum, 2015; Thomas et al. 1982). and consequently to establish control limits. The third
SPC is a process-orientated data driven method to phase is monitoring and controlling the process (Burges et
improve processes and to deliver day after day results, as al., 2015; Education course SPC, 2016; Oakland, 2003;
shown in Fig. 2. If in the past, the objective was the OOT Forum, 2015; Thomas et al., 1982).
customer requirements, where the analysis was after some
variances or OOS results occurred, in the SPC the
objective is the continual improvement and process Types of variability in the process
understanding, controlling and monitoring the stability of
the process. As a new way of thinking, SPC promotes There are two types of variability, common cause
continuously improvement of the processes by reducing and special cause variability. The common cause
their variability and prevents from occurring OOS results variability comes from the design of the process itself. It
by alerting before any batch is violating specifications can be affected from the variability of the materials,
(Burges et al., 2015; Education course SPC, 2016; equipment, and environmental conditions, physical and
Oakland, 2003; OOT Forum, 2015; Thomas et al., 1982). mental reactions of the people. Most of these differences
The SPC have several phases. First phase is process are small; they cause a pattern to fluctuate in what is
understanding, in order to monitor and control the process known as “natural” or “normal” manner. Every process is
it is necessary to understand its behavior and natural affected by common cause variability that can be reduced
pattern. It is crucial to understand the process variability, but not eliminated. Engineers often use the word “noise”
its source, whether it is common or special cause to describe it. The process is at statistical control and can
variability. In statistical terms, this phase is process of be measured by the process capability indexes: potential
understanding the capability of the process - how much capability (Cp); actual capability (Cpk), overall capability
the center is within the specification limits, how much is of the process (Pp) and measure of overall capability of
the target the center value. the process (Ppk). Process capability refers to the
Second phase is eliminating the variation of the performance of the process when it is operating in control.

Maced. pharm. bull., 65 (1) 39 – 60 (2019)

 Trending and Out-of-Trend results in pharmaceutical industry 43

Fig. 3. Selection of mathematical model (Ref. Burges et al., 2015).

Two capability indices are usually computed: Cp and sampling inspection and design of experiments (DoE)
Cpk. Cp measures the potential capability in the process (Thomas et al., 1982).
and it does not take into account where the process mean One of the key component in understanding the
is located relative to the specifications. Often the process process and the variability is the analyzing the data. They
can be off-centre it is toward the lower or upper should come from a measurement system, which has only
specification limits, than Cpk measures the actual consistent common cause variation and should be free
capability in a process. Ideally if a process is centered, from systematic errors (special cause variations). The
then Cp=Cpk. Pp evaluates overall capability based on the number of measurements should be large enough to
variation in the process, but not on its location. Ppk is a provide a means of estimating the common cause error
measure of the overall capability of the process and (Burges et al., 2015; Oakland, 2003; Thomas et al., 1982).
evaluates both the location and the overall variation of the There are two types of data continuous (variable) data
process (Burges et al., 2015; Education course SPC, 2016; and attribute (discrete) data. A continuous random
Oakland, 2003; OOT Forum, 2015; Thomas et al., 1982). variable is the one that can take any value over a range of
Sometimes unexpected, special cause happens which values. This type of variables is obtained by
has an effect on the process variability. For example, measurement, for example an assay, dissolution, pH,
material is taken from a different batch, or the operator tablet’s hardness, temperature or an impurity level. The
sets the machine with new settings, and in-experienced attribute is an integer where the set of possible values for
operator takes place of an experienced operator. These a discrete random variable is at most countable. These
causes make the pattern fluctuate in an “unnatural” or types of data are obtained mainly by counting, for
“abnormal” way. Finding the root cause opens the example a cosmetic defect on a tablet, number of
possibility to identify and study the behavior of the cause. defective products or the number of particles in a
The effort to find the root cause depends on the solution. Hence, the selection of the appropriate
complexity of the process. Consequently, ether we can mathematical distribution may be shown as a decision
eliminate the negative effect or keeping the positive, in tree (Fig. 3) according which the mathematical shape
any way it is a chance for improvement. Control Charts (model) of data that determines the applicability
are the only tools to differentiate between common and (suitability) of the statistical methodology (Burges et al.,
special cause variability (Burges et al., 2015; Education 2015).
course SPC, 2016; Oakland, 2003; OOT Forum, 2015; For our purposes, the most useful distribution for
Thomas et al., 1982). continuous variables is the Normal or Gaussian
distribution for a population whose properties are well
known.
SPC techniques For a true mean value (µ) of zero and a standard
deviation (Ϭ) of 1 the probability distribution is given by
There are several techniques that are essential in the equitation (1.1) and shown graphically in Fig. 4.
SPC: control charts, process capability studies, statistical

Макед. фарм. билт., 65 (1) 39 – 60 (2019)

44 M. Velinovska Cadinoska, N. Popstefanova, M. Ilievska, E. Karadzinska, M. Davceva Jovanoska, M. Glavas Dodov

a goodness of fit for the normal distribution, when

(1.1)
comparing the observed sample distribution with the
expected probability distribution. One of the most used
The areas under the curve indicate the probability of test and most efficient test of normality is Shapiro-Wilk’s
values lying ± Ϭ, ±2Ϭ and ±3Ϭ from the mean. This the most used. When the associated p-value is more than
distribution is the basis for control charting of continuous 0.05, then there is a strong statistical evidence of
random variables and stability trending. normality. For this example, the associated p-value is
0.2420, and then there is evidence of fit to normal
distribution (OOT Forum, 2015).
If the data are not normally distributed, then they
should be transformed either by logarithm: ln (X) or by
their reciprocal value: 1/X. The same should be applied
for specification values as well, simple flow diagram in
(Fig.6). With such transformed data, we should continue
to compute the indices and analyze the behavior of the
data (Fig. 7) (OOT Forum, 2015).

Fig. 4. Normal or Gaussian distribution 1 (Ref.

Burgess et al., 2015).

For attribute data, the Binomial or Poisson

distributions are preferred. If counted defects are to be
analyzed, the Binomial distribution is used. If the data are
defects expressed as a percentage for example then the use
of the Poisson distribution is indicated (Burges et al.,
2015).
Fig. 6. Flow diagram of normality testing 1 (Ref. OOT
Forum, 2015).

Fig.7. Logarithmic transformation 1 (Ref. OOT Forum,

2015).

Control Charts

Control chart is a plot of observations or

Fig. 5. Data normality 1 (Ref. OOT Forum, 2015). measurements (means, ranges etc.) against time or a
batch. The measurements can be ether variables or
attributes. Fig. 8 shows the control chart, for variable with
There are several ways to test the “normality” of the normally distributed data. The control chart is divided in
data, as shown in Fig. 5. The data are displayed in zones depending on the distance from the process average
histogram, Shapiro-Wilks Test and Normal Quantile- of the one, two or three standard deviations, σ. In general,
Quantile Plot. The first idea of the underlying distribution the distance of two standard deviations from the process
gives the histogram. Normal Quantile-Quantile Plot gives average is called warning limit, the yellow lines, and the

Maced. pharm. bull., 65 (1) 39 – 60 (2019)

 Trending and Out-of-Trend results in pharmaceutical industry 45
distance of three standard deviations from the process order of production. The central line is a process
average is called the control limit, the red lines on the average and is added as a visual reference for detecting
chart. Based on the location of the observations or shifts or trends – this is also referred to as the process
measurements in the control chart we can decide when it location. The process dispersion is between UCL and
may be necessary to take some action, as later explained LCL which are computed from available data and placed
in Control chart zones (Oakland, 2003; OOT Forum, equidistant from the central line (Education course SPC,
2015; Thomas et al., 1982). 2016; Burges et al., 2015; Oakland, 2003; OOT Forum,
2015; Thomas et al., 1982).

Fig. 8. Control Chart concept 1, where: +3Ϭ limit is

Upper Control Limit or Upper Action Limit,
-3Ϭ limit is Lower Control Limit or Lower
Fig. 9. Elements of control chart 1 (Ref. OOT
Action Limit, +2Ϭ limit is Upper Warning
Forum, 2015).
Limit or Upper Action Limit, -2Ϭ limit is
Lower Warning Limit or Lower Action
Limit (Ref. OOT Forum, 2015).
The control limits, UCL and LCL are calculated
according formulas:
This approach is based on the idea that no matter how
well a process is designed there is a certain amount of
(1.2)
nature variability in output measurements.
When the variation in process quality is due to
random causes alone, the process is said to be in-control. (1.3)
If the process variation includes both random and special
causes of variation, the process is said to be out-of-
control. The control chart is supposed to detect the where: – Process average of the sample, – Standard
presence of special causes of variation (Education course deviation of the sample and L – Constant, which is
SPC, 2016; Oakland, 2003; OOT Forum, 2015; Thomas dependent of the sample size.
et al., 1982).
The American Walter Shewhart was credited with the
invention of control charts for variable and attribute data The control limits
in the 1920s, at the Bell Telephone Laboratories, and the
term ‘Shewhart charts’ is in common use (Oakland, 2003; The control limits are used so it can be determined
Thomas et al.1982). whether the process is in the state of statistical control,
i.e. consistent. The UCL and LCL provide a range of
what is still acceptable for a result. Control charts are
Elements of the control chart therefore used to determine if the results that are coming
in are within the limits of what is acceptable or if the
There are three main elements of a control chart as process is out of control.
shown in Fig. 9. A control chart begins with plotting of a If a process is influenced by chance 99.73% of all
time series graph, where the process data, or results of the results do fall within 3 standard deviation (Ϭ), for
measurement, are plotted against time. The data practical reason, these limits are called “natural” limits of
comprising each individual point are random, but the the process (Fig. 8 and 9). The natural limits have no
points themselves are plotted in non-random arrangement, connection with the specification limits, they can be
they are plotted consecutively according time that is in broader or narrower, later is preferable. Within these limit

Макед. фарм. билт., 65 (1) 39 – 60 (2019)

46 M. Velinovska Cadinoska, N. Popstefanova, M. Ilievska, E. Karadzinska, M. Davceva Jovanoska, M. Glavas Dodov

Fig. 10. Diagram of control charts selection vs data type & sample size 1 (Ref. Burges et al., 2015).

the process is able to hold when operating normally under

influence of the non-assignable causes. UCL and LCL Control charts for variable type of data
must, wherever possible, be based on the values
determined for the Proven Acceptable Range (PAR) and The control charts for variable type of data are used
Normal Operating Range (NOR) for a process. While the to assess stability for variable data (e.g. HPLC potency
specification limits are used so that we can determine if assay, UV measurements), typically are used two types
the product is in regulatory compliance (Education course control charts (Burges et al., 2015; Education course SPC,
SPC, 2016; OOT Forum, 2015; Thomas et al. 1982). 2016; Oakland, 2003; OOT Forum, 2015; Thomas et al.,
The calculation of the Control Limits should be 1982). The first types are charts to monitor mean
based on 20 to 30 data points. Fewer data points will not consistency over time. Monitor between subgroup
give sufficient accuracy and the control limits will be too variation (subgroup mean): if the sample is represented by
wide. If there are more than 30 data points, there is a high one observation (or measurement) per batch, i.e. sample
possibility that special cause variability will affect the size is one, the control chart that is constructed is
process and the control limits will be to wide, too. Even if individual control hart (I chart) – chart for individual
the number of samples is very low the use of Control measurement. If data are collected in subgroups (several
Charts is highly recommended and Control Limits should observations or measurements on the same batch) i.e.
be re-calculated with every new data point. Control limits when the sample size is >1, than control chart – chart
are recalculated if several rules are applicable: if the for subgroup mean is constructed.
change is positive; the root cause of the change is known The second types of charts are charts to monitor
and there is assurance that the change is going to be common cause variation. These charts monitor within
permanent; and the forth rule is that there should be at subgroup variation (such as subgroup range or subgroup
least 20 samples that confirm the new process (Burges et standard deviation over time). If the sample is represented
al., 2015; Education course SPC, 2016; Oakland, 2003; by one observation (or measurement) per batch than the
OOT Forum, 2015; Thomas et al., 1982). chart that is used for monitoring the common cause
variation is Moving Range (MR) Control Charts which is
based on the difference between two consecutive
Types of control charts measurements or observation. With the increase the
sample size than Range (R) or Standard deviation (S)
Which type of the control charts shall be used Control Charts are used. The R-chart plots subgroup
depends on the type of data and the size of the sample as ranges (when subgroup sample size <9), and the S-chart
shown in diagram in Fig.10 (OOT Forum, 2015). The plots subgroup standard deviations (when subgroup
types of the control chart will be explained in subsequent sample size ≥9).
sections. Only the combination of both type of charts: chart

Maced. pharm. bull., 65 (1) 39 – 60 (2019)

 Trending and Out-of-Trend results in pharmaceutical industry 47

Fig. 11. Individual (I) chart 1(Ref. Burges et al., 2015).

for monitoring of mean consistency and chart for monitor size (Montgomery, 1982).
the common cause variation has all the information. In our example:
Individual chart is presented with Moving Range (MR) = 1.128 and MR = 3.29, therefore:
Control Charts and X is usually presented with Range (R
or S control chart (Education course SPC, 2016; Oakland,
2003; OOT Forum, 2015; Thomas et al., 1982).

Individual (I) / Moving Range (MR) Charts

(Subgroup Size n=1) Moving Range (MR) chart
In order to estimate the process variability, we can
The individual control charts are used for samples of use the Moving Range, which will indicate possible shifts
sizes n=1, each individual point is plotted on the graph; or changes in the process from one observation to the next
for example, release testing according specification of the (Burges et al., 2015; Education course SPC, 2016;
finished product. Also it can be used if data are already Oakland, 2003; OOT Forum, 2015; Thomas et al., 1982).
available but not collected in constant subgroup sizes The moving range (MR) is defined as the absolute
which are greater than 1 (OOT Forum, 2015). difference between two consecutive observations or
For example, if we have 84 individual measurements, measurements (Burges et al., 2015; Oakland, 2003; OOT
they are plotted in a chart where the control limits are Forum, 2015; SPC Education course, 2016; Thomas et al.,
established and the average value is calculated. 1982). The formula for calculation is given in equitation
The calculation of the mean value, or process center 1.6:
is according standard formula for calculation of average
value: (1.6)

In our example, MR chart for the same 84

measurements, the mean value of MR which represents
the center line (CL) would be calculated as follows:
The calculation of the UCL and LCL are directly
dependent of the calculation of the moving ranges MR:

(1.4)

(1.5) The calculation of UCL and LCL for MR chart and

sample size 1 is according formulas 1.7 and 1.8 (Burges
et al., 2015; Education course SPC, 2016; Oakland, 2003;
where is constant which is dependent of the sample OOT Forum, 2015; Thomas et al., 1982).
Макед. фарм. билт., 65 (1) 39 – 60 (2019)
48 M. Velinovska Cadinoska, N. Popstefanova, M. Ilievska, E. Karadzinska, M. Davceva Jovanoska, M. Glavas Dodov

Fig. 12. Moving Range (MR) chart 1 (Ref. OOT Forum, 2015).

(1.7) (1.9)

(1.8)
(1.10)
D3 and D4- constants which are dependent of the sample – constant for sample size (Montgomery, 1982).
size (Montgomery, 1982).
In our example, the UCL and LCL are as follows: R and S Charts are used to detect changes in the
variation within subgroups. R charts are used for subgroup
size between 2 and 8, and S charts are used for subgroup
size ≥ 9 (Eq. 1.11 and 1.12, respectively).

(1.11)
/ Range (R) control chart
(1.12)
When the data are collected in subgroups and the
sample size n is 2≤ ≤ 8, than / Range (R) control chart
For example, for assay of active ingredient in the
are recommended. The size of subgroups should be
formulation, 21 measurements with subgroup of 4, n=4:
rational, predefined and constant, for example items which
were produced under the same conditions. The subgroups
should be formed by consecutive measurements. Each
subgroup’s statistics are compared to the control limits,
and patterns of variation between subgroups are analyzed.
For example, in-process controls parameters of tablets
from same batch such are tablet’s mass, hardness,
thickness; or control sample in analytical analysis, for
example dissolution testing is performed on 6 tablets,
therefore the subgroup n=6 (Burges et al., 2015; Oakland,
2003; OOT Forum, 2015; SPC Education course, 2016;
Thomas et al.,1982).
Chart is used to detect changes in the mean value
between subgroups; the process average is averages of all
subgroups: Fig. 13. chart 1(Ref. OOT Forum, 2015).

Mean value of group would be:

The control charts are constructed according

equitation 1.9 and 1.10:

Maced. pharm. bull., 65 (1) 39 – 60 (2019)

 Trending and Out-of-Trend results in pharmaceutical industry 49
Average value of group means would be: According to our previously presented example of 21
measurements with subgroup of 4, the UCL and LCL
should be calculated as follows:

Since the UCL and LCL are dependent of , we must

first calculate (Fig. 14).

R chart of the same measurements would be

constructed according following equitation:

(1.13)
Fig. 15. Table with constants for control charts 1
(Ref. Montgomery, 1982).

The Shewart control chart detects large variability

with standard deviation between 0.5 – 2.5. For more
sensitive shifting in the process with standard deviation
less than 0.5 CUSUM (Cumulative Sums) and EWMA
(Exponentially Weighted Moving Average) charts are
used. These charts detect small but persistent changes in
the process, they are mainly used in trend analysis for
investigation. CUSUM and EWMA charts are useful for
analysis of historical data after the discovery of a problem
(post mortem). They are used in cases when there is large
time difference between two measurements or on daily
basis for quick detection. CUSUM and EWMA charts can
Fig.14. R chart for within subgroup variability 1 be used in variable and discrete type of data (Burges et
(Ref. OOT Forum, 2015). al., 2015; OOT Forum, 2015).

Mean value of Range, , UCL and LCL are CUSUM charts

calculated according subsequent formulas:
Cumulative Sum control chart, which displays
(1.14) cumulative sums of the deviations of measurements or
subgroup means from a target value. As measurements
are taken, the difference between each measurement and
(1.15)
the benchmark value which is process target ( 0) is
calculated, and this is cumulatively summed up - CuSum.
If the processes are in control, measurements do not
deviate significantly from the process target, so
measurements greater than the process target and those
(1.16) less than the process target averaged each other out, and
the CuSum value should vary narrowly around process
(1.17) target level. If the processes are out of control,

Макед. фарм. билт., 65 (1) 39 – 60 (2019)

50 M. Velinovska Cadinoska, N. Popstefanova, M. Ilievska, E. Karadzinska, M. Davceva Jovanoska, M. Glavas Dodov

measurements will more likely to be on one side of the If is close to 0, more weight is given to past
bench mark, so the CuSum value will progressively depart observations. If is close to 1, more weight is given to
from that of the bench mark (Fig. 16) (Burges et al., 2015; present information. When =1, the EWMA becomes the
OOT Forum, 2015). Individuals control chart. Typical values for are less
than 0.25 (Burges et al., 2015; OOT Forum, 2015).

Control charts for discrete type of data

Discrete distribution is where the set of possible

values for the random variable is at most countable
(Attribute). There are several types of mathematical
models of data distribution (Fig. 3), but most useful in the
Fig.16. Graphical presentation of CuSum charts 1 (Ref. SPC analysis are Binomial and Poisson’s distribution. An
Burges et al., 2015). advantage of attributes is that they are in general more
quickly assessed; sometimes variables are converted to
attributes for assessment. However, as we shall see,
attributes are not so sensitive a measure as variables and,
EWMA
therefore, detection of small changes is less reliable
(Burges et al., 2015; OOT Forum, 2015). Whenever the
Exponentially Weighted Moving Average (EWMA)
measured quantities for one item are not continuous but
chart, also referred to as a Geometric Moving Average
rather quality characteristics or count data, control chart
(GMA) chart are a good alternative to the Shewart control
for discrete data should be used. Usually, the classification
chart when we want to detect small shifts. It acts in the
of the inspected item is “conforming item” or
same way as a CuSum chart. Each point on a EWMA
“nonconforming item”. A nonconforming item is a unit of
chart is the weighted average of all the previous subgroup
product that does not satisfy one or more of the
means, including the mean of the present subgroup sample
specifications of the product (it contains at least one
(Fig. 17). The weights decrease exponentially going
nonconformity). If more than one defect can be observed
backward in time (Burges et al., 2015; OOT Forum,
on the same unit, one can be interested in the number of
2015).
nonconformities (defects) per unit, instead of the fraction
nonconforming for a single nonconformity (defect),
(1.18) Fig.18 (Burges et al., 2015; OOT Forum, 2015).
Hence, a defective is an item or ‘unit’ that contains
where 0 < ≤ 1 is a constant and the starting value is the one or more flaws, errors, faults or defects. A defect is an
process target: 0= 0. individual flaw, error or fault. The inspection of product
defects is based on acceptance quality level (AQL) with
defined acceptance quality limit.

Fig. 17. EWMA chart 1 Fig. 18. Defective vs Defects 1 (Ref. OOT Forum,
(Ref. https://support.minitab.com/en- 2015).
us/minitab/18/help-and-how-to/quality-and-
process-improvement/control-charts/how- Depending of the type of the nonconformity or
to/time-weighted-charts/ewma-chart/before- defect that are subject of the analysis there are several
you-start/example/). types of charts (Fig. 19) (OOT Forum, 2015):

Maced. pharm. bull., 65 (1) 39 – 60 (2019)

 Trending and Out-of-Trend results in pharmaceutical industry 51

Fig. 19. Shematic presentation of control charts for discrete data 1 (Ref. OOT Forum, 2015).

to be out of control when they are not. For a Laney

Control charts for single nonconformity: P-chart attributes chart, the definition of common cause variation
and NP-chart includes not only the within-subgroup variation, but also
the average variation between consecutive subgroups. If
The control charts for single nonconformity can be there is over dispersion, the control limits on a Laney P’
either fraction (percentage) of non-conforming (P-chart) chart are wider than those of a traditional attributes chart.
o, r total number of non-conforming units, if sample size The opposite, under dispersion occurs when there is less
is the same (NP-chart). These charts are used to than expected variation in the data, it occurs mostly when
monitor the proportion of defective items where each item the data are auto correlated with each other -
can be classified into one of two categories, such as pass autocorrelation. For example, as a tool wears out, the
or fail (Oakland 2003; OOT Forum, 2015; Thomas et al., number of defects may increase. The increase in defect
1982). It should be noted the importance of the sample counts across subgroups can make the subgroups more
size. Namely, if sample size is too large with respect to similar than they would be by chance (Laney, 2002;
the number of nonconforming units (e.g., 20 Oakland 2003; OOT Forum, 2015; Thomas et al., 1982).
nonconforming units out of 500000), then the p-chart will
not work properly because the control limits are inversely
proportional to the sample size. Therefore, they became
very small and process will look out of control, as data
plotted on the control chart will be out of control limits. If
the sample size is the same (or approximately the same)
the individuals control charts should be used where the
number of nonconforming units should be plotted. If the
sample size is significantly different from one sample
point to another, then one could use a Laney p-chart
(Laney, 2002; Oakland, 2003; OOT Forum, 2015;
Thomas et al.,1982).
Other important point that should be addressed is the
dispersion of the data. The over dispersion of the data
Fig. 20. Traditional P-chart 1
exists when is more variation than expected as a result
(Ref. https://support.minitab.com/en-
from the natural causes and not special causes of
us/minitab/18/help-and-how-to/quality-and-
variation. Since the P Chart and later explained U-chart
process-improvement/control-charts/how-
are based on the assumption that the variability is
to/attributes-charts/p-chart/before-you-
constant, therefore in situation of wide dispersion of the
start/overview/).
data the over dispersion of the data can cause some points
Макед. фарм. билт., 65 (1) 39 – 60 (2019)
52 M. Velinovska Cadinoska, N. Popstefanova, M. Ilievska, E. Karadzinska, M. Davceva Jovanoska, M. Glavas Dodov

For example, the subgroups are very large, with an Forum, 2015):
average of about 2500 observations in each (Fig. 20 and
21). Additionally, the P Chart diagnostic test reveals over (1.22)
dispersion in the data. The large subgroup sizes result in
very narrow control limits on the traditional P chart. With The distribution of the random variable can be
the narrow control limits, the over dispersion causes obtained from the binomial distribution.
several of the subgroups to appear out of control. The
Laney P' chart, however, corrects for over dispersion and
shows that the process is actually in control. No points fall
outside of the control limits (Laney, 2002; Oakland 2003;
OOT Forum, 2015; Thomas et al., 1982).

Fig. 22. P-chart with unequal sample size 1

(Ref. https://support.minitab.com/en-
us/minitab/18/help-and-how-to/quality-and-
process-improvement/control-charts/how-
to/attributes-charts/p-chart/before-you-
start/overview/).
Fig. 21. Laney P-chart 1
((Ref. https://support.minitab.com/en-
us/minitab/18/help-and-how-to/quality-and- The P-chart control limits are inversely proportional
process-improvement/control-charts/how- of the samples size, the formula for calculating the UCL
to/attributes-charts/p-chart/before-you- and LCL (Fig.22 and 23), (Laney, 2002; Oakland 2003;
start/overview/). OOT Forum, 2015):

(1.23)
P-charts – control chart for proportion of
nonconforming units
(1.24)
This chart shows the proportion of nonconforming or
defective product produced by a manufacturing process Depending on the values of and ni, sometimes the
(Laney, 2002; Oakland 2003; OOT Forum, 2015; Thomas LCL is less than 0. In these cases, we set LCL=0 and
et al., 1982). assume the control chart only has an upper limit.
Suppose m samples of sample size ni are available,
is the average sample size:

(1.20)

If the sample size is the same for each group, then

= .
The sample fraction nonconforming for sample i is
defined as the ratio of the number of non-conforming units
in the sample i, Di, to the sample size ni (Laney, 2002;
Oakland 2003; OOT Forum, 2015; Thomas et al., 1982).

(1.21) Fig. 23. P-chart with equal sample size 1

(Ref. https://support.minitab.com/en-
Suppose m samples are available, then the average us/minitab/18/help-and-how-to/quality-and-
fraction nonconforming, which is process average or process-improvement/control-charts/how-
centerline (CL) is (Laney, 2002; Oakland 2003; OOT to/attributes-charts/p-chart/before-you-
start/overview/).
Maced. pharm. bull., 65 (1) 39 – 60 (2019)
 Trending and Out-of-Trend results in pharmaceutical industry 53
NP-charts – control chart for number of non- sample j. The total nonconformities in a sample follow a
conforming units (defectives) Poisson distribution. The C-chart upper and lower control
limits and centerline are calculated according formulas
This chart shows the total number of nonconforming 1.29, 1.30 and 1.31 respectively. Where is the observed
units, with constant sample size, almost the same as the P- average number of non-conformities in a preliminary
chart. Each point on an NP-chart represents the number of sample of m inspection units, n is the constant sample size
defective items or units for one subgroup (Fig. 24). The and is the number of defects for inspection unit i
(Oakland, 2003; OOT Forum, 2015):
centerline on an NP-chart represents the average
number of defectives per subgroup, n . If the (1.29)
subgroups sizes are not equal, the centerline is
not straight on the NP-chart (Oakland, 2003; OOT (1.30)
Forum, 2015).
The NP-chart control limits are calculated according
(1.31)
formulas:

If LCL yields a negative value, then LCL is fixed to

(1.25)
0.

(1.26)

where the standard deviation (1.27)

Fig. 25. C chart 1

(Ref. https://support.minitab.com/en-
us/minitab/18/help-and-how-to/quality-and-
process-improvement/control-charts/how-
Fig. 24. NP -Chart 1
to/attributes-charts/c-chart/before-you-
(Ref. https://support.minitab.com/en-
start/example/).
us/minitab/18/help-and-how-to/quality-and-
process-improvement/control-charts/how-
to/attributes-charts/np-chart/before-you- U-Charts for number of defects/non-conformities
start/example/).
If the sample size is not constant and can vary from
one sample to another, the average number of
C-charts for number of defects/non-conformities nonconformities per unit of product should be taken
instead of total number (Fig.26). The average number of
This chart shows the number of defects or
nonconformities per unit is labeled with which also
nonconformities produced by a manufacturing process.
represent the centerline of the chart; we can calculate
Each point on a C chart represents the number of defects
according equation 1.32 (Oakland, 2003; OOT Forum,
for one subgroup. When we have a constant sample size,
2015):
n, of inspection units from one sample to another, we can
work with the total number of nonconformities per sample
and construct the c-chart. The total number of (1.32)
nonconformities in a unit is represented on the chart
(Oakland, 2003; OOT Forum, 2015):
is the total nonconformities in a sample of .
(1.28)
Calculation of the U-chart control limits is according
formula 1.33 and 1.34:
where xij is the number of defects for inspection unit i in

Макед. фарм. билт., 65 (1) 39 – 60 (2019)

54 M. Velinovska Cadinoska, N. Popstefanova, M. Ilievska, E. Karadzinska, M. Davceva Jovanoska, M. Glavas Dodov

(1.33) Control chart zones

Control charts are divided in three zones A, B and C,

(1.34) depending of the spread of the standard deviation, as
shown in Fig. 8 and Table 1. The A zone is the zone
which is most distant from the centerline, this zone is
where represents the observed average number of
between 2 and 3 standard deviations above and below
nonconformities per unit in a preliminary data set of m
centerline or mean value of the process, the limits of the
inspection units, n is the sample size of the current
zone A are the UCL and LCL. Zone B is between 1 and 2
inspected sample (Oakland, 2003; OOT Forum, 2015):
standard deviations from the centerline; its limits are the
warning limits UWL and LWL. The C zone is between
(1.35) centerline and one standard deviation. The zones A, B,
and C are sometimes called the three sigma zone, two
sigma zone, and one sigma zone, respectively (Education
course SPC, 2016; Oakland, 2003; OOT Forum, 2015;
Thomas et al.,1982).

Table 1. The three sigma zones

Zone Region
Zone A Between 2σ from the center line and
(three sigma zone) 3σ limits (which are control limits)
Zone B Between 1σ and 2σ from the center
(two sigma zone) line (which are the warning limits)
Zone C Between 1σ and center line
(one sigma zone)

Fig. 26. U chart 1 Based on the location of the sample point in ABC
(Ref. https://support.minitab.com/en- zones, there are several rules for detecting ou-of-trend
us/minitab/18/help-and-how-to/quality-and- results WECO and Nelson rules (OOT Forum, 2015;
process-improvement/control-charts/how- Education course SPC, 2016).
to/attributes-charts/u-chart/before-you-
start/example/).
WECO Western Electric rules
Interpretation of the Control Charts, trending rules WECO rules were codified by a specially -
appointed committee of the manufacturing division of
The control charts are designed for detection of the Western Electric Company and appeared in the first
change in the average value of the process, detection of edition of a 1956 handbook, that became a standard text of
change in the variability and both simultaneously: the field. Their purpose was to ensure that line workers
detection of average value and variability of the process. and engineers interpret control charts in a uniform way.
Through control charts, we can detect continuous changes, Today they are widely used in interpreting the control
drifting or trend, or frequent irregular changes. The charts. There are four (4) rules that are in use in analyzing
trending rules are taking in to account the differences and interpretation of data patterns. If anyone of the
between normal and unnatural patterns based on several violates the rules than it is considered that trend or some
criteria. For example, the absence of data around the kind of instability do exist and there is special cause
central line or too much data near the control limits variation (OOT Forum, 2015; Thomas et al., 1982).
indicates presence of more than one distribution, this is Rule 1 is that if any single data point falls outside the
called mixture. The absence of data on control limits 3σ limit from the centerline, i.e. outside the zone A or
indicates stratification i.e. abnormally small fluctuations. beyond the UCL and LCL, then the process is out of
The presence of data outside the control limits indicates control. Rule 2 is if 2 out of 3 consecutive points, which
instability, i.e. abnormally large fluctuations. Other are on the same side of the centerline, fall beyond 2σ
unnatural patterns are systematic movement, cyclical limit, in zone A or beyond, then the rule is violated. Rule
repetition, trend (Oakland, 2003; OOT Forum, 2015; SPC 3 is if four (4) out of five (5) consecutive points fall
Education course, 2016; Thomas et al., 1982). beyond 1σ limit, in zone B or beyond, on the same side of

Maced. pharm. bull., 65 (1) 39 – 60 (2019)

 Trending and Out-of-Trend results in pharmaceutical industry 55

Fig. 27 WECO Rules 1 (Ref. https://en.wikipedia.org/wiki/Western_Electric_rules).

the centerline. Rule 4 is if 9 nine (9) consecutive points Rule 3 is if six (6), or more, points in a row are
fall on the same side of the centerline, in zone C or continually increasing, or decreasing, it can be interpreted
beyond, then trend do exist. Fig. 27 shows examples of that a trend exists (OOT Forum 2015):
the WECO rules.

Nelson rules

The Nelson rules were first published by Lloyd S

Nelson in 1984 in the Journal of Quality Technology.
There are eight (8) Nelson rules for interpreting the
control chart and separate special cause from common
cause of variation (Nelson 1984; OOT Forum, 2015).
Rule 1 is one point is more than 3 standard deviations
from the mean it means that the sample is out of control
(Fig. 28). Fig.30. Nelson Rule 3 1.

Rule 4 is fourteen (14), or more, points in a row

alternate in direction, increasing then decreasing, which
usually means that the process is over-controlled (OOT
Forum, 2015) (Fig. 31).

Fig. 28. Nelson Rule 1 1.

Rule 2 is applied when nine (9), or more, points in a

row are on the same side of the mean, which means that
some prolonged bias exists (OOT Forum, 2015) (Fig. 29).
Fig. 31. Nelson Rule 4 1.

Rule 5 is two (2), or three (3) out of three (3) points

in a row are more than 2 standard deviations from the
mean in the same direction which can be interpreted as
that there is a medium tendency for samples to be out of
control (OOT Forum, 2015) (Fig.32).
Rule 6 is four (4) or five (5) out of five (5) points in a
row are more than 1 standard deviation from the mean in
the same direction. There is a strong tendency for samples
to be slightly out of control (OOT Forum, 2015) (Fig. 33).
Fig. 29. Nelson Rule 2 1.

Макед. фарм. билт., 65 (1) 39 – 60 (2019)

56 M. Velinovska Cadinoska, N. Popstefanova, M. Ilievska, E. Karadzinska, M. Davceva Jovanoska, M. Glavas Dodov

“Procedures should describe how trending and

calculations are to be performed and should guard
against overreaction to individual events as well as
against failure to detect unintended”.

Fig. 32. Nelson Rule 5 1.

Rule 7 is fifteen (15) points in a row are all within 1

standard deviation of the mean on either side of the mean Fig. 35. Nelson rules.
(OOT Forum, 2015) (Fig. 34).

Examples of interpreting control charts

Test 1 Nelson & WECO: Points out-side the control

limits (Fig. 36); the process is grossly out of control,
mainly as a large difference between two consecutive
results (OOT Forum, 2015; Thomas et al. 1982).

Fig. 33. Nelson Rule 6.

Rule 8 is that eight points in a row exist, but none

within 1 standard deviation of the mean, and the points are
in both directions from the mean. Rule 7 and 8 are
diagnostic tests for stratification, i.e. for abnormal
fluctuation in data (OOT Forum, 2015).
Fig. 36. Test 1 Nelson & WECO rules 1 (Ref. Burges
et.al., 2015).

Nelson Тест 2 = WECO Тест 4 (Fig. 29): At least 9

points in a row, below the average; these rules can detect
change in the average value of the process (Burges et al.,
2015; Nelson, 1984; Thomas et al., 1982).

Fig. 34. Nelson Rule 7.

From the eight Nelson’s rules only the first 3 of them

are used, the reason for this is that we should be cautious
not overreact and search out-of-trend results when they do
not exist (OOT Forum, 2015).
Even the FDA Guidance for Industry Process
Validation: General Principles and Practices in part D.
Stage 3: Continued Process Verification clearly states that Fig.37. Nelson test 2; WECO test 4 1 (Ref. Burges et
al., 2015).
Maced. pharm. bull., 65 (1) 39 – 60 (2019)
 Trending and Out-of-Trend results in pharmaceutical industry 57
Nelson Rule 3 (Fig.38): 6 or more points in a row are than the UCL and LCL can be accepted and for every
continually increasing or decreasing, this test indicates new measurement WECO or Nelson rules should be
trend, i.e. drifting of the average value. applied. If the rules are violated investigation should be
triggered and closed before batch disposition.

Basic Capability Indices

The process capability is another vast array in SPC

and in this review it is only mentioned as a connection
between process stability and capability. Process
capability refers to the performance of the process when it
Fig. 38. Nelson test 3. is operating in control. Process capability index is a
measure relating the actual performance of a process to its
specified performance. The process itself is a combination
of the materials used for production, the equipment, the
Steps for control charting and trending
method of production, the personnel involved in
manufacturing operation, the environment. The absolute
Control charting and trending can be explained in
minimum requirement is that three standard deviations
several short steps:
each side of the process mean are contained within the
 The samples should be collected at exact time
specification limits, which are the UCL and UCL. This
interval with defined sample size, n; The number
means that ca 99.7 per cent of output will be within the
of samples which are needed for establishing the
tolerances (Fig.8). A more stringent requirement is often
control chart limits are at least 30
stipulated for UCL and LCL to ensure that product of the
 After data collection, we should check the correct quality is consistently obtained over the long term.
normality of the data When a process is under statistical control i.e. only
 The data should not be auto correlated common causes of variation are present, a process
 Depending of the sample size we should construct capability index may be calculated (Fig. 39).
of I or charts for process average and MR or R Process capability indices are simply a means of
charts for process variability indicating the variability of a process relative to the
We should check the stability of the process on the product specification tolerance: Cp and Cpk. Where Cp is
MR, R charts, by applying WECO or Nelson rules and if process capability ratio for a centered process and
they are violated than the process is unstable. measures the potential capability of the process. Cpk is
Consequently, this means that UCL are LCL for I/X-bar process capability ratio for an off-center process and
charts are invalid. It should be noted that UCL and LCL measures the actual capability of a process (Burges et al.,
are valid only on stable process. If the process is stable 2015; Oakland, 2003).

UNSTABLE PROCESS STABLE PROCESS STABLE PROCESS , INCAPABLE AND CAPABLE

Fig. 39. Process stability and capability 1 (Ref. Burges et al., 2015).

Макед. фарм. билт., 65 (1) 39 – 60 (2019)

58 M. Velinovska Cadinoska, N. Popstefanova, M. Ilievska, E. Karadzinska, M. Davceva Jovanoska, M. Glavas Dodov

Calculation of the Cp index

In order the product to be manufactured with

parameters that are within specification it is important that
the difference between upper specification limit (USL)
and lower specification limit (LSL), or so called tolerance
limits (T) is smaller than the process standard variation. In
Fig.40 is shown how the widening of specification limits
is affecting the process capability. Usually in
pharmaceutical industry, pharmacopeia’s monographs Fig. 41. Non-centered process 1 (Ref. Oakland, 2003).
impose the product specification limits and changing the
limits would not be allowed. In that case, efforts should be
made to lower as much as possible the standard deviation. For upper and lower specification limits, there are
The calculation of Cp index takes in consideration of the two Cpk values, Cpku and Cpkl. These relate the
established specification limits and standard deviation σ. difference between the process mean, μ, and the upper and
The calculation is according equitation 1.36. When the lower specification limits respectively, to 3σ (half the
variables control charts are used in the capability studies total process variation). Cpk is the minimum value of the
the standard deviation, , is estimated by and constant Cpu and Cpl. It is calculated according formulas 1.37 and
2 as explained in Control charts for variable data (Burges 1.38 (Burges et al., 2015; Oakland, 2003).
et al., 2015; Oakland, 2003).

(1.36)
(1.37)

(1.38)

Many industries use a benchmark value of 1.33. If

Cpk is lower than 1.33 benchmark, it should be
considered ways to improve the process, such as reducing
its variation or shifting its location. In Fig. 42 examples
are shown the stability of the process and it location. For
stable and capable processes, Example 1, the control
limits are narrower than the specification limits and all
process data are within UCL and LCL and around the
target of the specification limit. For unstable but capable
process, Example 2, the UCL and LCL are too much
narrower and some data are out of control limits. All the
data are around the target value of the specification. In
Example 3 it is noticeable that all the data are toward the
upper specification limit al do all of them are within
Fig. 40. Effect on the process capability the widening control and specification limits. But since the data are
of specification limits 1 (Ref. Oakland, 2003). toward the upper specification limits the process is
incapable. In Example 4 it is shown that not only the data
are out of control but out of specification also. The
Calculation of the Cpk index process is unstable and incapable (OOT Forum, 2015).

The process capability ratio- does not take into

account where the process mean is located relative to the Conclusion
specifications. In some instances, we can see wide
specification limits, with low standard deviation, but part Trending is a trend in the GMP environment.
of the process lies outside the specification limit (Fig. 41). Particularly in process validation trending is a requirement
The index that is used to measure both the process form FDA since 2011. The control chart are essential tools
variation and the centering is Cpk, which is widely for data analyzing establishing control limits on CQAs,
accepted as process capability index (Burges et al., 2015; trending and detecting the unexpected or special cause
Oakland, 2003). variability in the process. They are used in context of gain

Maced. pharm. bull., 65 (1) 39 – 60 (2019)

 Trending and Out-of-Trend results in pharmaceutical industry 59

EXAMPLE 1. STABLE AND CAPABLE PROCESS EXAMPLE 2. UNSTABLE BUT CAPABLE PROCESS

EXAMPLE 3. STABLE BUT INCAPABLE PROCESS EXAMPLE 4. UNSTABLE AND INCAPABLE PROCESS

Fig. 42. Examples of process stability and capability (Ref. OOT Forum, 2015).

more knowledge of the process. The critical data are European Compliance Academy (ECA), 2015. OOT Forum,
monitored more frequently and thus adequate responses in education course, Barcelona, Spain.
case of violations based on risk shall be taken. We should European Compliance Academy (ECA), 2016. Education course
ensure that correct control chart is used, based on data SPC, Heidelberg, Germany.
type, their distribution and sample sizes. Along with this European Medicines Agency (EMA), 2016, Guideline on
process validation for finished products - information and
process it is imperative to ensure roust documentation and
data to be provided in regulatory submissions. Available at:
knowledge management. Continuous monitoring should https://www.ema.europa.eu/en/documents/scientific-
enable management to assess process capability more guideline/guideline-process-validation-finished-products-
frequently and drive improvements. information-data-be-provided-regulatory-
submissions_en.pdf.
International Council for Harmonisation of Technical
References Requirements for Pharmaceuticals for Human Use (ICH),
2009. Tripartite Guideline Q8 (R2) Pharmaceutical
Burges, C., 2015. Laboratory Data Management Guidance, Out development. Available at:
of Expectation (OOE) and Out of Trend (OOT) Results, https://www.ich.org/fileadmin/Public_Web_Site/ICH_Prod
ECA _AQCWG_ SOP 02_OOE OOT_v1.0. ucts/Guidelines/Quality/Q8_R1/Step4/Q8_R2_Guideline.p
EudraLex Volume 4 EU Guidelines for Good Manufacturing df.
Practice for Medicinal Products for Human and Veterinary International Council for Harmonisation of Technical
Use Annex 15: Qualification and Validation. Available at: Requirements for Pharmaceuticals for Human Use (ICH),
https://ec.europa.eu/health/sites/health/files/files/eudralex/v 2006. Tripartite Guideline Q9 Quality risk management.
ol-4/2015-10_annex15.pdf. Available at:
European Commission, 2010. EudraLex The Rules Governing https://www.ich.org/fileadmin/Public_Web_Site/ICH_Prod
Medicinal Products in the European Union EU Guidelines ucts/Guidelines/Quality/Q9/Step4/Q9_Guideline.pdf.
to Good Manufacturing Practice Medicinal Products for International Council for Harmonisation of Technical
Human and Veterinary Use, Part I Basic Requirements for Requirements for Pharmaceuticals for Human Use (ICH),
Medicinal Products. Available at: 2008. Tripartite Guideline Q10 Pharmaceutical quality
https://ec.europa.eu/health/documents/eudralex/vol-4_en. system. Available at:
European Compliance Academy (ECA), 2015. 6th GMP https://www.ich.org/fileadmin/Public_Web_Site/ICH_Prod
Conference, Heidelberg, Germany ucts/Guidelines/Quality/Q10/Step4/Q10_Guideline.pdf.
Макед. фарм. билт., 65 (1) 39 – 60 (2019)
60 M. Velinovska Cadinoska, N. Popstefanova, M. Ilievska, E. Karadzinska, M. Davceva Jovanoska, M. Glavas Dodov

International Council for Harmonisation of Technical https://doi.org/10.1080/002075499191832.

Requirements for Pharmaceuticals for Human Use (ICH), Nelson, L.S., 1984. The Shewhart Control Chart—Tests for
2012. Tripartite Guideline Q11 Development and Special Causes. Journal of Quality Technology 16 (4),
manufacture of drug substances. Available at: 238-239. Available at:
https://www.ich.org/fileadmin/Public_Web_Site/ICH_Pro https://doi.org/10.1080/00224065.1984.11978921.
ducts/Guidelines/Quality/Q11/Q11_Step_4.pdf. Oakland, J.S., 2003. Statistical Process Control, Fifth Edition.
International Medical Device Regulators Forum (IMDRF), Robert, J.L., 2010. ICH Integrated training workshop. Available
2004. Process Validation Guidance, Annex A: Statistical at:https://www.ich.org/products/guidelines/quality/training
methods and tools for process validation. Available at: -programme-for-q8q9q10/presentations.html.
http://www.imdrf.org/documents/doc-ghtf-sg3.asp. Thomas, D.W., 1982. Statistical Quality Control Handbook:
Laney, D.B., 2002. Improved Control Charts for Attributes. Second Edition, Western Electric Company Inc.
Quality Engineering 14(4), 531–537. US Department of Health and Human Services, Food and Drug
Medicines & Healthcare products Regulatory Agency (MHRA), Administration (FDA), 2011. Guidance for Industry,
2018, GXP- Data Integrity Guidance and Definitions. Process Validation: General Principles and Practices.
Available at: Winkle, H., 2007. Implementing QbD, Presented at PDA/FDA
https://assets.publishing.service.gov.uk/government/uploa Joint Regulatory Conference
ds/system/uploads/attachment_data/file/687246/MHRA_G World Health Organization (WHO), 2015. Guidance on Good
xP_data_integrity_guide_March_edited_Final.pdf. Data and Record Management Practices. Available at:
Montgomery, D.C., 1982. Economic Design of an x Control https://www.who.int/medicines/publications/pharmprep/W
Chart. International Journal of Quality & Reliability HO_TRS_996_annex05.pdf?ua=1.
Management 14(3), 234-259. Available at:

Резиме

Тренд и резултати вон тренд во фармацевтската

индустрија
Маја Велиновска-Чадиноска1*, Нада Попстефанова1, Мирослава Илиевска2,
Елизабета Караџинска1, Мaрија Давчева Јовановска1, Марија Главаш Додов3
1
Алкалоид АД Скопје, бул. Александар Македонски, бр.12,
1000 Скопје, Р. С. Македонија
2
Експерт од областа на Добрата Производна Пракса, Мите Богоевски 64А,
1000 Скопје, Р. С. Македонија
3
Институт за фармацевтска технологија, Фармацевтски Факултет,
Универзитет „Св.Кирил и Методиј”, Мајка Tереза 47, 1000 Скопје, Р. С. Македонија

Клучни зборови: тренд, контролни карти, СКП, варијабилни податоци, дискретни (атрибутивни) податоци,
стандардна девијација, горна контролна граница, долна контролна граница

Постои зголемување на барањата и од регулаторните агенции и од самата индустрија, за подобрување на

постапките за следење на квалитетот на лековите. Овој фокус на квалитетот и управувањето со ризиците
поттикна реевалуација на системите и процедурите за да се обезбеди усогласеност со предложените водичи и
регулаторни барања. Едни од основните алатки за следење на квалитетот се статистички алатки со кои можеме
да поддржиме било какви заклучоци во однос на варијабилноста и способноста на даден процес и на тој начин
може да се обезбеди состојба на контрола. Постојат бројни техники кои може да се користат, а една од нив е
Статистичката контрола на процеси (СКП). СКП е високо структуриран пристап за дефинирање на она што е
важно, преку скрининг на процесот, собирање на податоци, анализа на податоци и наоѓање на извор на
варијабилност на процесот; како и тоа да даде насоки што треба да се направи и како да се доведе процесот во
контрола или пак да се задржи состојбата на контрола. Трендот е дел од СКП и овозможува да се види општата
насока кон која се движи процесот. Трендот се користи како превенција на можни резултати вон спецификација
и неусогласени производи кои би можеле да имаат влијание врз здравјето на пациентот. Повеќе не е прифатливо
да се биде само во рамките на спецификациите, а надвор од контролата на процесот.

Maced. pharm. bull., 65 (1) 39 – 60 (2019)