Over the past decade or so, the topic of risk analysis has become an increasingly important subject. Recently and belatedly, the ISO 9001 committee decided to sprinkle throughout the standard requirements for risk analysis that will be part of the 2015 revision.
Risk analysis need not be a difficult subject to address and yet, a Google™ search of “risk analysis” will yield over 344 millions results! Even if only a tenth of a percent of these results is of interest, anyone interested in the subject may well need several lifetimes to read them all. Worse yet, should you be able to complete this Herculean task, you may still not know what to do or how much to do.
YOUR ALREADY DOING RISK ASSESSMENTS
In the majority of cases a risk analysis is nothing more than an extension of what many companies are currently doing in some other form. Indeed, whenever a process capability is conducted a risk of producing bad product or, out of tolerance product, is estimated.
The computation of capability indices such as Cpk, Cp, Cpl or any other similar indices computed by most statistical software is a form of risk quantification. For Six Sigma aficionados, if a process is for example, found to be operating at 4s the long term risk of producing bad product is approximately 0.6% or approximately 6,500 parts per million (ppm); one could also if needed, estimate associated costs.
Naturally, there are other types of risk assessments for which the computation of a process capability is either not available or not possible to determine. In such cases, other techniques such as FMEA could be used to assess, prioritize and quantify process risk.
RISK ASSESSMENT EXAMPLE
An example of risk associated with instrumentation reading
Monitoring someone’s blood pressure is a simple and yet crucial task especially for patients that are subject to strokes. A reading of the systolic and diastolic reading that is too high or too low invariably requires a corrective action. For stroke patients, the ideal systolic to diastolic ratio is around 120/60 or a little less. Ratios in the 170s/90s or even higher are considered potentially dangerous and require administering medication or perhaps, if the condition persists increasing the dosage.
However increasing the dosage is also risky because the blood pressure may fall below 100/50 or even lower thus creating the risk of fainting or worse if the ratio falls much lower. The medical staff of any hospital is thus faced with two risks when monitoring blood pressure for critical patients: insuring that the blood pressure does fall dangerously low because of an over dosage while at the same time insuring that the blood pressure is not dangerously high thus increasing the likelihood of a stroke; hence the importance of obtaining an accurate reading.
There are two traditional ways of obtaining blood pressure: using a stethoscope or using an electronic device of which there are several kinds including devices that are placed on the wrist. Although electronic devices are not recommended for patients suffering from an irregular heartbeat condition known as arrhythmia they are routinely used in hospital with patients with arrhythmia.[i]
In one particular case I observed a nurse used a stethoscope on a patient; the reading was 100/50. If such a low reading persists, the recommendation is to skip one cycle of the medication and keep monitoring. However in this particular case, the nurse waited about a minute and took another reading using an electronic monitoring device attached to the wrist; this time the reading was 175/92!
Which reading should the nurse believe? She took another reading with the stethoscope and obtained a low reading of around 105/54 and suggested to stop one cycle of the medication. Had the nurse only taken one reading with the electronic device the medication would have been prescribed at the risk of further lowering the blood pressure to perhaps a dangerous low level (which had happened once when the blood pressure came down to 70/45.) Fortunately, without the medication, the blood pressure remained stable at around 135/70 for another five days.
PROCESS RISK ANALYSES
This simple example illustrates the need to conduct risk analyses on processes including the simple process of taking blood pressure. The manufacturing sector has long been aware of the fact that it is not sufficient to calibrate equipment one also needs to assess repeatability and reliability of the measurement system by conducting gage R&R studies.[ii]
The above example from the health care industry, an industry that certainly knows about the need to calibrate equipment, illustrates the importance and benefits of conducting gage R&R to assess and quantify risk.
[i] http://en.wikipedia.org/wiki/Blood_pressure. “Oscillometric monitors may produce inaccurate readings in patients with heart and circulation problems, which include arterial sclerosis, arrhythmia, preeclampsia, pulsus alternans, and pulsus paradoxus.”
[ii] One study in the U.K revealed that as many as 25 to 30 percent of blood pressure monitoring equipment used in hospitals have unacceptable calibration error. Calibration accuracy of hospital-based non-invasive blood pressure measuring devices. de Greeff A, Lorde I, Wilton A, Seed P, Coleman AJ, Shennan AH. Source: Maternal and Fetal Research Unit, Division of Reproduction and Endocrinology, King’s College London, London, UK.
Bio:
James Lamprecht is a management consultant and Six Sigma Master Black Belt. In his career spanning over three decades, Dr. Lamprecht has worked as a consultant, teacher, and statistician. He has audited over one hundred companies here and abroad and has conducted hundreds of seminars and classes in applied industrial statistics, ISO 9001 and Six Sigma. He has authored 11 books including Interpreting ISO 9001:2000 with Statistical Methodology (ASQ Quality Press, 2001), Applied Data Analysis for Process Improvement: A Practical Guide to Six Sigma Black Belt Statistics (ASQ Quality Press, 2005) and Dare To Be Different: Reflections on Certain Business Practices with Renato Ricci (ASQ Quality Press, 2009). Dr. Lamprecht who has consulted in Europe, Canada and Latin America received his doctorate from UCLA.