Statistical Process Control as a tool for research and

healthcare improvement

Tan Song Xian

Faculty

of Applied Science and Technology, University Tun Hussein Onn Malaysia,

Malaysia

Abstracts

.Nowadays, health issues care

are getting more concern by people. The awareness of health care among people

had increased. The Statistical process control (SPC) tools are being used in

health care industry and the usage is increasing. The SPC tools are applied

into health care improvement such as hospitals. This paper provided an overview of common uses of the

SPC tools in health care industry and example of applying SPC tools in health

care industry and how it is used for quality improvement.

Keywords.

Statistical control process; tools; health care; improvement

1

Introduction

Quality can be

achieved by evaluating and improving production processes or service delivery. 1

Changes are required in order to improve the process of health care and service

delivery, but not all the changes will result in improvement.1,2

Health care

industry is under increasing pressure to be more efficient and more effective.

Nowadays, hospitals are able to adopt the techniques and methods of Continuous

Quality Improvement (CQI) as a part of their trust requirements. One of the

main challenges of implementing CQI into health care industry is on how to

manage, control and improve the process by using Statistical process control

(SPC) techniques.3

SPC is a strategy, a philosophy and a set of

methods for on-going improvement of systems and processes. The SPC approach is

based on data and has its foundation in the theory of variation which are common

causes and special causes. The primary tools commonly used in SPC are Shewhart

charts which also known as control charts, run charts, histograms, Pareto

diagrams, scatter diagrams, flow charts4. By monitoring the systems or

process, it must also be able to minimize the false positive or false negative

that may arise and which could lead to inappropriate clinical decision making.

Recently, the

statistical process chart which also known as control charts are used in

monitoring for health care5. In year 1920, Walter A Shewhart developed a

theory of variation which then forms the basis of SPC.6,7 The SPC charts

are very useful tools for investigating and identifying the important process

variables and the quality improvement.8

The control charts was originally used as a tool for controlling and monitoring

manufacturing process. The control charts is a set of simple graphical tools.

Generally, control charts is consist of a central line which represent the mean

of the data, a lower and upper lines represent the lower and upper controls

limits respectively which are usually set at three-sigma from the mean. Any

points that fall outside the limits in the control charts, is considered as out

of control points.5

There are

several control charts that are applied into the monitoring processes which are

Shewhart chart, Cumulative Sum (CUSUM) and Exponentially Weighted Moving

Average (EWMA) which allows continuous real time assessment. The Shewhart chart

is used to discover large changes, while CUSUM and EWMA are more suitable to recognize

of small to moderate changes.9

2

Methods

The control charts are the most frequently used tools in the

statistical process. Control charts are used to monitoring processes in order

to achieve a better mean value of the process or to reduce the variability of

the processes to improve the quality. There are three horizontal lines in the

control charts, which are the center line which also known as mean, and lower

and upper limits. While the vertical axis consist of the values of the

appropriate sample characteristics.10

The sample size of the data must be specified before

designing a control charts. When the subgroup size of the data is 1, the

suitable control charts is Moving Range Chart. While for the subgroup size

between 2 to 10, the recommended charts is chart and R chart. If the subgroup size of the

data if greater than 10, the suitable control charts is chart and s chart.

For the chart, let be the average of each of the sample. Next the

process average is calculated by using. Then, the will be used as the center line of the control charts. To construct

the control limits, the estimated standard deviation ? is needed. The range of

the sample is the difference between the smallest and largest observations

which is R= . Then let R1, R2Rm be the ranges

of the sample and calculate the average of the range by . Then the upper control

limit (UCL) is calculated by using UCL, while lower control limit

(LCL) is computed by LCL= . The center line is equal

to For R chart, the control limits are calculate

by using UCL , LCL=and center line = .

For s chart, the control limits is calculated by using the

following formula which UCL= , while LCL= and center line=. 11

Figure 1: Example of

a control chart 8

The control

limits are used to determine whether the process is stable or not and identify

whether there is points out of control or not. If the result shows that the

process is a stable process which there is no out of control points and showed

a non-random pattern, the parameters of the statistical model is used and the

control limits are used for further monitoring process.8

3

Results and Discussion

The following examples illustrate the application of control

charts as a data analysis tool.

Example: Laboratory

turn around time (TAT)

Several clinician are complaining about the turn around time

(TAT) for complete blood counts has been out of control and the condition is

getting worse. Thus, the laboratory manager decided to investigate this

situation by collecting data. The data are stratified by shift firstly and the type

of request to ensure that the analysis is conducted by a reasonable processes.

Generally, the TAT data is always follow the normal distribution. The chart and s chart are the suitable control

charts for these data. 2

Figure 2: The control

charts of turn around time (TAT) for day shift routine orders for complete

blood counts.2

The mean , and standard deviation, s of TAT of each day

were calculated for three randomly selected order for complete blood counts.

From figure 2, the upper chart is chart, while the bottom chart is s chart. For chart, it shows the mean of TAT for the three

orders each day. While the s chart shows the standard deviation for the same

three orders. It can be seen that there is no points are out of control which

indicated that the turn around time for complete blood counts for each day are

in control.

However, from the statement above, it stated that the turn

around time for complete blood counts are out of control and are getting worse.

If the clinician’s complains are true, it will observed that there is points

out of control and an increasing trend will be observed from the control chart.

But, the result from the control chart shows that the process is having a good

performance and it is in a statistical control. Although this result may not

agree with the view of clinicians, but it is not necessarily meaning that the

result are acceptable. A process that are in control can be predictably as a bad

process. There may be exist common cause variation.

In this case, the process is stable and predictable but it

is not acceptable to the clinicians. It is appropriate to consider to lower the

mean of TAT and reduce the variation which to lower the center line and the aim

is to bring the control limits closer as an improvement strategies since the

process is exhibit common cause variation only. From the strategy, a new and

more acceptable control limits will be produced and hence the level of

performance will also increased. Then, the new process with new baseline

measurements is tested to decide whether the process is improved, remain the

same or getting worse.2

4

Conclusion

In general, the statistical

process control tools such as control charts could help the teams to make

decision on the correct improvement strategy whether to search for special

causes when the process is out of control or to work on more fundamental

process improvements when the process is in control. In the example above, the

control charts can be used as a simple monitoring tools to ensure the improvement

are remain over the time.2From figure 2, it can be seen that the process is

in control. However, the result is difference as what the clinician’s claim.

Thus, the department has to collect more data to get a more reasonable control

limits. But, it general, the example had helped to generate a simple overview

on how SPC had been applied into health care industry.

Process monitoring

by using SPC tools is an important process in evaluating and improving the

framework in health care industry. The control charts are using three sigma

control limits generally.7

In conclusion, the

result indicated how SPC had been applied to health care industry although

there is still having some barriers but in order to overcome the barriers some

changes are needed so that the application could improve patients’ health. The

SPC tools are very useful as a tools for evaluating the performance of health

care providers. The SPC tools like control charts are useful, user friendly and

easy to use and it is a statistically strict process analysis tools that could

be used by quality improvement teams. These SPC tools could help quality

improvement managers and also researchers to use the data and result to make

appropriate decisions for quality improvement.

ACKNOWLEDGMENT

I would

like to express a special thanks to my lecturer Dr. Shuhaida Binti Ismail who

had gave me a chance to do a topic on Statistical process control as a tools

for research and healthcare improvement. She had gave me guidance and

motivation when doing the paper. I would like to express a special to my parent

too as they had gave me motivation to do the paper.

References

1. Sigurd Fasting, Sven

E. Gisvold. Statistical process control methods allow the analysis and

improvement of anesthesia care, Canadian Journal of Anesthesia, pp. 767-774,

2003.

2. J.C Benneyan, R.C.

Lloyd, P.E. Plsek. Statistical process control as a tool for research and

healthcare improvement, Quality and Safety in Health Care, vol 12, pp.

458-464, 2003.

3. E.G.Tsacle, N.A.Aly. Statistical

process control in the Health Care Industry, Elsevier Ltd, vol. 31, pp.

447-450, 1996.

4. Carey RG. Improving

Healthcare with Control Charts: Basic and Avanced Statistical Process Control

Methods and Case Studies,Milwaukee:ASQ Quality Press 2003.

5. Ruth Tennant, Mohammed

A. Mohammed, Jamie J. Coleman, Una Martin. Monitoring patients using control

charts: a systematic review, International Journal for Quality in

Health Care, vol 19, No.4, pp. 187-194, 2007.

6. Shewhart WA. Economic

Control of Quality of Manufactured Product, New York: D. Van

Nostrand Company, 1931.

7. MA Mohammed. Using

Statistical Process Control to improve the quality of health care, Quality

and Safety of Health Care, pp. 243-245, 2004.

8. William H.Woodall,

Benjamin M. Adams, James C. Benneyan. The Use of Control Charts in

Healthcare, John Wiley & Sons. Ltd, 2012.

9. Anthony

P. Morton, MStatistical Process Control as a tool for research and

healthcare improvement

Tan Song Xian

Faculty

of Applied Science and Technology, University Tun Hussein Onn Malaysia,

Malaysia

Abstracts

.Nowadays, health issues care

are getting more concern by people. The awareness of health care among people

had increased. The Statistical process control (SPC) tools are being used in

health care industry and the usage is increasing. The SPC tools are applied

into health care improvement such as hospitals. This paper provided an overview of common uses of the

SPC tools in health care industry and example of applying SPC tools in health

care industry and how it is used for quality improvement.

Keywords.

Statistical control process; tools; health care; improvement

1

Introduction

Quality can be

achieved by evaluating and improving production processes or service delivery. 1

Changes are required in order to improve the process of health care and service

delivery, but not all the changes will result in improvement.1,2

Health care

industry is under increasing pressure to be more efficient and more effective.

Nowadays, hospitals are able to adopt the techniques and methods of Continuous

Quality Improvement (CQI) as a part of their trust requirements. One of the

main challenges of implementing CQI into health care industry is on how to

manage, control and improve the process by using Statistical process control

(SPC) techniques.3

SPC is a strategy, a philosophy and a set of

methods for on-going improvement of systems and processes. The SPC approach is

based on data and has its foundation in the theory of variation which are common

causes and special causes. The primary tools commonly used in SPC are Shewhart

charts which also known as control charts, run charts, histograms, Pareto

diagrams, scatter diagrams, flow charts4. By monitoring the systems or

process, it must also be able to minimize the false positive or false negative

that may arise and which could lead to inappropriate clinical decision making.

Recently, the

statistical process chart which also known as control charts are used in

monitoring for health care5. In year 1920, Walter A Shewhart developed a

theory of variation which then forms the basis of SPC.6,7 The SPC charts

are very useful tools for investigating and identifying the important process

variables and the quality improvement.8

The control charts was originally used as a tool for controlling and monitoring

manufacturing process. The control charts is a set of simple graphical tools.

Generally, control charts is consist of a central line which represent the mean

of the data, a lower and upper lines represent the lower and upper controls

limits respectively which are usually set at three-sigma from the mean. Any

points that fall outside the limits in the control charts, is considered as out

of control points.5

There are

several control charts that are applied into the monitoring processes which are

Shewhart chart, Cumulative Sum (CUSUM) and Exponentially Weighted Moving

Average (EWMA) which allows continuous real time assessment. The Shewhart chart

is used to discover large changes, while CUSUM and EWMA are more suitable to recognize

of small to moderate changes.9

2

Methods

The control charts are the most frequently used tools in the

statistical process. Control charts are used to monitoring processes in order

to achieve a better mean value of the process or to reduce the variability of

the processes to improve the quality. There are three horizontal lines in the

control charts, which are the center line which also known as mean, and lower

and upper limits. While the vertical axis consist of the values of the

appropriate sample characteristics.10

The sample size of the data must be specified before

designing a control charts. When the subgroup size of the data is 1, the

suitable control charts is Moving Range Chart. While for the subgroup size

between 2 to 10, the recommended charts is chart and R chart. If the subgroup size of the

data if greater than 10, the suitable control charts is chart and s chart.

For the chart, let be the average of each of the sample. Next the

process average is calculated by using. Then, the will be used as the center line of the control charts. To construct

the control limits, the estimated standard deviation ? is needed. The range of

the sample is the difference between the smallest and largest observations

which is R= . Then let R1, R2Rm be the ranges

of the sample and calculate the average of the range by . Then the upper control

limit (UCL) is calculated by using UCL, while lower control limit

(LCL) is computed by LCL= . The center line is equal

to For R chart, the control limits are calculate

by using UCL , LCL=and center line = .

For s chart, the control limits is calculated by using the

following formula which UCL= , while LCL= and center line=. 11

Figure 1: Example of

a control chart 8

The control

limits are used to determine whether the process is stable or not and identify

whether there is points out of control or not. If the result shows that the

process is a stable process which there is no out of control points and showed

a non-random pattern, the parameters of the statistical model is used and the

control limits are used for further monitoring process.8

3

Results and Discussion

The following examples illustrate the application of control

charts as a data analysis tool.

Example: Laboratory

turn around time (TAT)

Several clinician are complaining about the turn around time

(TAT) for complete blood counts has been out of control and the condition is

getting worse. Thus, the laboratory manager decided to investigate this

situation by collecting data. The data are stratified by shift firstly and the type

of request to ensure that the analysis is conducted by a reasonable processes.

Generally, the TAT data is always follow the normal distribution. The chart and s chart are the suitable control

charts for these data. 2

Figure 2: The control

charts of turn around time (TAT) for day shift routine orders for complete

blood counts.2

The mean , and standard deviation, s of TAT of each day

were calculated for three randomly selected order for complete blood counts.

From figure 2, the upper chart is chart, while the bottom chart is s chart. For chart, it shows the mean of TAT for the three

orders each day. While the s chart shows the standard deviation for the same

three orders. It can be seen that there is no points are out of control which

indicated that the turn around time for complete blood counts for each day are

in control.

However, from the statement above, it stated that the turn

around time for complete blood counts are out of control and are getting worse.

If the clinician’s complains are true, it will observed that there is points

out of control and an increasing trend will be observed from the control chart.

But, the result from the control chart shows that the process is having a good

performance and it is in a statistical control. Although this result may not

agree with the view of clinicians, but it is not necessarily meaning that the

result are acceptable. A process that are in control can be predictably as a bad

process. There may be exist common cause variation.

In this case, the process is stable and predictable but it

is not acceptable to the clinicians. It is appropriate to consider to lower the

mean of TAT and reduce the variation which to lower the center line and the aim

is to bring the control limits closer as an improvement strategies since the

process is exhibit common cause variation only. From the strategy, a new and

more acceptable control limits will be produced and hence the level of

performance will also increased. Then, the new process with new baseline

measurements is tested to decide whether the process is improved, remain the

same or getting worse.2

4

Conclusion

In general, the statistical

process control tools such as control charts could help the teams to make

decision on the correct improvement strategy whether to search for special

causes when the process is out of control or to work on more fundamental

process improvements when the process is in control. In the example above, the

control charts can be used as a simple monitoring tools to ensure the improvement

are remain over the time.2From figure 2, it can be seen that the process is

in control. However, the result is difference as what the clinician’s claim.

Thus, the department has to collect more data to get a more reasonable control

limits. But, it general, the example had helped to generate a simple overview

on how SPC had been applied into health care industry.

Process monitoring

by using SPC tools is an important process in evaluating and improving the

framework in health care industry. The control charts are using three sigma

control limits generally.7

In conclusion, the

result indicated how SPC had been applied to health care industry although

there is still having some barriers but in order to overcome the barriers some

changes are needed so that the application could improve patients’ health. The

SPC tools are very useful as a tools for evaluating the performance of health

care providers. The SPC tools like control charts are useful, user friendly and

easy to use and it is a statistically strict process analysis tools that could

be used by quality improvement teams. These SPC tools could help quality

improvement managers and also researchers to use the data and result to make

appropriate decisions for quality improvement.

ACKNOWLEDGMENT

I would

like to express a special thanks to my lecturer Dr. Shuhaida Binti Ismail who

had gave me a chance to do a topic on Statistical process control as a tools

for research and healthcare improvement. She had gave me guidance and

motivation when doing the paper. I would like to express a special to my parent

too as they had gave me motivation to do the paper.

References

1. Sigurd Fasting, Sven

E. Gisvold. Statistical process control methods allow the analysis and

improvement of anesthesia care, Canadian Journal of Anesthesia, pp. 767-774,

2003.

2. J.C Benneyan, R.C.

Lloyd, P.E. Plsek. Statistical process control as a tool for research and

healthcare improvement, Quality and Safety in Health Care, vol 12, pp.

458-464, 2003.

3. E.G.Tsacle, N.A.Aly. Statistical

process control in the Health Care Industry, Elsevier Ltd, vol. 31, pp.

447-450, 1996.

4. Carey RG. Improving

Healthcare with Control Charts: Basic and Avanced Statistical Process Control

Methods and Case Studies,Milwaukee:ASQ Quality Press 2003.

5. Ruth Tennant, Mohammed

A. Mohammed, Jamie J. Coleman, Una Martin. Monitoring patients using control

charts: a systematic review, International Journal for Quality in

Health Care, vol 19, No.4, pp. 187-194, 2007.

6. Shewhart WA. Economic

Control of Quality of Manufactured Product, New York: D. Van

Nostrand Company, 1931.

7. MA Mohammed. Using

Statistical Process Control to improve the quality of health care, Quality

and Safety of Health Care, pp. 243-245, 2004.

8. William H.Woodall,

Benjamin M. Adams, James C. Benneyan. The Use of Control Charts in

Healthcare, John Wiley & Sons. Ltd, 2012.

9. Anthony

P. Morton, Michael Whitby, Mary-Louse. The

Application of Statistical Process Control Charts to the detection and

monitoring of hospital-acquired infections, J Qual Clin Pract, pp. 112,

2001.

10. Pavol Gejdos. Continuous

Quality Improvement by Statistical Process Control, Procedia Economics

and Finance, vol. 34, pp. 565-572, 2015.

11. Montgomery. Douglas C.

Introduction to Statistical Quality Control, John Wiley &

Sons, pp. 228-262, 2009.

ichael Whitby, Mary-Louse. The

Application of Statistical Process Control Charts to the detection and

monitoring of hospital-acquired infections, J Qual Clin Pract, pp. 112,

2001.

10. Pavol Gejdos. Continuous

Quality Improvement by Statistical Process Control, Procedia Economics

and Finance, vol. 34, pp. 565-572, 2015.

11. Montgomery. Douglas C.

Introduction to Statistical Quality Control, John Wiley &

Sons, pp. 228-262, 2009.