AbstractImprovement of health care are neededto become different in processes of service. The process performance iscomplicated by the existence of natural variation if it determined thesechanges are having the beneficial effects. Such as same measurements arerepeated will get different values and some of it will indicate a better orworst performance. Traditional statistical analysis methods will delay the timeof decision making due to a lots of measurements over time but it also considermethod that account for natural variation. Statistical process control (SPC) asa tool for research and healthcare improvement because it is a kind ofstatistics that combines presentation of data with time series analysis method.
The tool make the data more quickly and in a understand way to decision makers.There are 7 main tools that mentioned in the statistical process control (SPC).In this paper, we mentioned the primary tool which is the control chart.
This toolmake a better understanding and communicating data from the healthcareimprovement efforts to all the researchers and public. This paper also provides overview of statisticalprocess control and few examples of the healthcare applications which appliedprimary tool of SPC—the control chart. Introduction Nowadays,health care is quite important for everyone include younger, teenagers, andalso elder because they know body health are quite important and necessary tothem. Without health, they cannot do anything that they want like travel, enjoyinglife and so on. Maybe they just stay at the hospital or lying on the bed onlywithout do anything.
Last time, people only busy work for earn money withoutcare their health and pollution level of world is not so much serious comparedto now. In last few years, people noticed that health care is quite importantfor them to maintain a good and healthy lifestyle. They not only busy theirwork and also care about health care like eating a healthy meal, buy a healthcare insurance, always make appointment for body checkup and so on. Not onlythat, they also care about that due to pollution nowadays quite seriouscompared to last time. But, still a lot of people don’t know what is thehealthcare and why is it important. And nowadays health care should make someimprovement to benefit for everyone. Healthcare also called healthcare is a kind of action to maintain or make improvementof health via the prevention, diagnosis, treatment and other physical andmental impairments inside human body. Health professionals delivered healthcarein some occupation like physicians, optometry, audiology, pharmacy, psychology,physician associates, dentistry, midwifery, nursing, medicine, and other healthprofessions.
Economicand social condition mostly always affect the health care when the healthcaremay vary across individuals, groups and individuals. Every country have differentpolicies and plans to related personal or grouped health care within theirsocial. Due to meet the health needs for populations, health care system shouldbe prepared organization. Their exactconfiguration will make a difference between national and subnational entities.In some country, planning occurs more centralized among government and privateorganization while in some country, the plan is distributed among marketparticipants.
According to the World Health Organization (WHO), awell-functioning healthcare system requires a robust financing mechanism; awell-trained and adequately paid workforce; reliable information on which tobase decisions and policies; and well maintained health facilities andlogistics to deliver quality medicines and technologies.1Healthcare cancontribute to a large part of a country’s economy. It is important to promotephysical and mental health to people around the world.
Example, in 1980, thefirst disease in human history to be completely eliminated by health careinterventions was declared by WHO. 4Even thought,all improvement needed change, but not all change will carry out in theimprovement. The major components of measurement are:(1) Determine and define keyindicators(2) Collect data(3) Analyze and interpret data The third component which is analysisand interpretation of data by using statistical process control (SPC) wasfocused by this paper.
To determine whether changes in processes to make a realdifference in outcome, statistical process control (SPC) can do so. In thispaper, we discuss an overview of statistical process control (SPC) theory, explaincontrol charts which is a major tool of SPC, and provide examples of theirapplication to common issues in healthcare improvement. The basic theory of statisticalprocess control was developed in the late 1920s by Dr Walter Shewhart,3 astatistician at the AT&T Bell Laboratories in the USA, and was popularisedworldwide by Dr W Edwards Deming.4 Both observed that repeated measurementsfrom a process will exhibit variation—Shewhart originally worked withmanufacturing processes but he and Deming quickly realized that theirobservation could be applied to any sort of process. If a process is stable,its variation will be predictable and can be described by one of severalstatistical distributions. One such model of random variation isthe normal (or Gaussian) bell shaped distribution which is familiar to mosthealthcare professionals.
While repeated measurements from many processesfollow normal distributions, it is important to note that there are many othertypes of distributions that describe the variation in other healthcaremeasurements such as Poisson, binomial, or geometric distributions. Forexample, the random variation in the number of wound infections after surgerywill follow a binomial distribution since there are only two possibleoutcomes—each patient either did or did not have a postoperative infection withabout the same probability (assuming that the data are adjusted for patientacuity, surgical techniques, and other such variables). SPC theory uses the phrase “commoncause variation”to refer to the natural variation inherent in a process on aregular basis. This is the variation that is expected to occur according to theunderlying statistical distribution if its parameters remain constant overtime.
For example, the random variation between body temperatures within apopulation of healthy people is a result of basic human physiology, while therandom variation in week to week wound infection rates is a result of factorssuch as training, sources of supplies, surgical and nursing care practices, andcleanliness procedures. Processes that exhibit only common cause variation aresaid to be stable, predictable, and in “statistical control”, hence the majortool of SPC is called the “statistical control chart”. Conversely, the phrase “special causevariation”refers to unnatural variation due to events, changes, orcircumstances that have not previously been typical or inherent in the regularprocess. This is similar to the concept in traditional hypothesis tests of dataexhibiting statistically significant differences, a key distinction being thatwe now test for changes graphically and over time using small samples. Forexample, heavy demand for A&E services brought on by an influenza epidemicmay create special cause variation (statistically significant differences) inthe form of increases in A&E waiting times. As another example, supposethat the daily mean turn around time (TAT) for a particular laboratory test is64 minutes with a minimum of 45 minutes and a maximum of 83 minutes; this meanhas been observed for several months. One day the mean jumps to 97 minutesbecause a major power outage caused the computers to go down, the lights to goout, and the pneumatic tube system to become inoperative. On this particularday the process is said to be “out of control”and incapable of performing as ithad in the past due to the “special cause”of the power outage.
Note that special cause variation canbe the result of either a deliberate intervention or an external event overwhich we have little control. Special causes of variation can also be transient(being short staffed in A&E one day due to illness of a key person) or canbecome part of the permanent common cause system (eliminating a staff positionthrough a budget cut). Interventions in a research study orchange ideas in a quality improvement project are deliberate attempts tointroduce special causes of variation. Statistical tools are therefore neededto help distinguish whether patterns in a set of measurements exhibit common orspecial cause variation.
While statistical process control charts andhypothesis tests are both designed to achieve this goal, an importantdifference is that SPC provides a graphical, simpler, and often faster way toanswer this question. The basic principles of SPC are summarised in box 1. Box 1 Basic principles of SPCIndividual measurements from anyprocess will exhibit variation. If the data come from a stable commoncause process, their variability is predictable within a knowable range thatcan be computed from a statistical model such as the Gaussian, binomial, orPoisson distribution. If processes produce data withspecial causes, measured values will deviate in some observable way from theserandom distribution models. Assuming the data are in control, wecan establish statistical limits and test for data that deviate frompredictions, providing statistical evidence of a change.
These observations lead to twogeneral approaches for improving processes. Because processes that exhibitspecial cause variation are unstable and unpredictable, they should be improvedby first eliminating the special causes in order to bring the process “intocontrol”. In contrast, processes that exhibit only common cause variation willcontinue to produce the same results, within statistical limits, unless theprocess is fundamentally changed or redesigned. Moreover, if a process remains incontrol, future measurements will continue to follow the same probabilitydistribution as previously—that is, if a stable process produces data thatfollow a normal distribution and it is not further disturbed by special causes,we can expect about 95% of future measurements to fall within ±2 standarddeviations (SD) around the mean. We can make similar statements aboutprediction ranges associated with any other statistical distribution.
Ingeneral, regardless of the underlying distribution, almost all data will fallwithin ±3SD of the mean if the underlying distribution is stable—that is, ifthe process is in statistical control. The control chart therefore defineswhat the process is capable of producing given its current design andoperation. If a different level of performance is wanted in the future, we mustintervene and introduce a change in the process—that is, a special cause. If wesimply want to sustain the current level of performance, special causes ofvariation must be prevented or eliminated. Control charts can often help todetect special cause variation more easily and faster than traditionalstatistical methods, and therefore are valuable tools for evaluating theeffectiveness of a process and ensuring the sustainability of improvements overtime.
THE CONTROL CHART: THE KEY TOOL OFSPCShewhart developed a relativelysimple statistical tool—the control chart—to aid in distinguishing betweencommon and special cause variation. A control chart consists of two parts: (1)a series of measurements plotted in time order, and (2) the control chart”template”which consists of three horizontal lines called the centre line(typically, the mean), the upper control limit (UCL), and the lower controllimit (LCL). To interpret a control chart, datathat fall outside the control limits or display abnormal patterns (see later)are indications of special cause variation—that is, it is highly likely thatsomething inherently different in the process led to these data compared withthe other data. As long as all values on the graph fall randomly between theupper and lower control limits, however, we assume that we are simply observingcommon cause variation. Where to draw the UCL and LCL isimportant in control chart construction. Shewhart and other SPC expertsrecommend control limits set at ±3SD for detecting meaningful changes inprocess performance while achieving a rational balance between two types of risks.If the limits are set too narrow there is a high risk of a “type Ierror”—mistakenly inferring special cause variation exists when, in fact, apredictable extreme value is being observed which is expected periodically fromcommon cause variation. This situation is analogous to a false positiveindication on a laboratory test.
On the other hand, if the limits are set toowide there is a high risk of a “type II error”analogous to a false negativelaboratory test. For example, for the familiar normaldistribution, in the long run 99.73% of all plotted data are expected to fallwithin 3SD of the mean if the process is stable and does not change, with onlythe remaining 0.27% falling more than 3SD away from the mean. While points thatfall outside these limits will occur infrequently due to common causevariation, the type I error probability is so small (0.0027) that we insteadconclude that special variation caused these data.
Similar logic can be appliedto calculate the type I and type II errors for any other statisticaldistribution. Although traditional statisticaltechniques used in the medical literature typically use 2SD as the statisticalcriteria for making decisions, there are several important reasons why controlcharts use 3SD. For the normal distribution approximately 95% of the values liewithin 2SD of the mean so, even if the process was stable and in control, ifcontrol limits are set at 2SD the type I error (false positive) rate for eachplotted value would be about 5% compared with 0.
27% for a 3SD chart. Unlike onetime hypothesis tests, however, control charts consist of many points (20–25 iscommon) with each point contributing to the overall false positive probability.A control chart with 25 points using 3SD control limits has a reasonablyacceptable overall false positive probability of 1–(0.9973)25?=?6.5%, whereasusing 2SD limits would produce an unacceptably high overall false positiveprobability of 1–(0.95)25?=?27.
7%! The bottom line is that the UCL and LCL areset at 3SD above and below the mean on most common control charts.5 In addition to points outside thecontrol limits, we can also look more rigorously at whether data appearrandomly distributed between the limits. Statisticians have developedadditional tests for this purpose; for example, a common set of tests forspecial cause variation is: one point outside the upper or lowercontrol limits; two out of three successive pointsmore than 2SD from the mean on the same side of the centre line; four out of five successive pointsmore than 1SD from the mean on the same side of the centre line; eight successive points on the sameside of the centre line; six successive points increasing ordecreasing (a trend); or obvious cyclic behaviour. In return for a minor increase infalse positives, these additional tests greatly increase the power of controlcharts to detect process improvements and deteriorations. The statistical”trick”here is that we are accumulating information and looking for specialcause patterns to form while waiting for the total sample size to increase.
This process of accumulating information before declaring statisticalsignificance is powerful, both statistically and psychologically. A final important point about theconstruction of control charts concerns the mechanics of calculating the SD. Aswith traditional statistical methods, many different formulae can be used tocalculate the SD depending on the type of control chart used and the particularstatistical distribution associated with that chart. In particular, the formulafor the SD is not the one typically used to calculate the empirical SD as mightbe found in a computer spreadsheet or taught in a basic statistics class.
Forexample, if we are monitoring the proportion of surgery patients who acquire aninfection, the appropriate formulae would use the SD of a binomial distribution(much like that for a conventional hypothesis test of proportions); ifmonitoring a medication error rate the appropriate formulae would use the SD ofa Poisson distribution; and when using normally distributed data theappropriate formulae essentially block on the within sample SD in a mannersimilar to that used in hypothesis tests of means and variances. Details ofcalculations for each type of control chart, when to use each chart, andappropriate sample sizes for each type of chart are beyond the scope of thispaper but can be found in many standard SPC publications