Issue link: https://www.e-digitaleditions.com/i/341135
factors affect an outcome. They do so using a bar chart with a line plot superimposed on the bars that corresponds to a second y-axis to show what percentage each element con- tributes to the total set of variables. For example, of 133 people surveyed about their medication compliance, say 47 reported non-compliance due to unclear dosing instruc- tions. To show that, the first bar would show a count of 47 respondents, and the line plot at that point would corre- spond to 35 percent on the second y-axis (133 ? 47 = 35.3 percent), as Figure 3 shows. Many Pareto charts show only the few factors that account for the bulk of the effects in order to separate the "vital few" from the "trivial many," as Joseph Juran said [5]. Some call it the "80-20 rule." 26 July 2014 Tablets & Capsules element "Dies improperly cleaned." Does that stem from a lapse in an automatic cleaning cycle (equipment-related error) or from a cleaning protocol that was not written correctly and clearly (method-related error)? Or perhaps there was no precise definition of what acceptably "clean" is (measurement-related error). But in making the chart, don't devote too much time to such questions. It's more important to understand your process and to recognize what is working correctly and incorrectly (defects). During the initial search for causal factors, just list as many potential causes of the problem as you can. Pareto charts. Named for Italian theoretician Vilfredo Pareto, Pareto charts show the degree to which different Figure 2 Cause-and-effect ("fishbone") diagrams Changed excipient providers Incorrect procedures Dies improperly cleaned Production room overly humid Changed excipient providers 0.15 Incorrect procedures 0.10 Dies improperly cleaned 0.05 Production room overly humid 0.70 a. The "head" is the problem, and the "ribs" are causal factors b. Assigning probabilities to causal factors indicates where to focus corrective action Man Material Method Tablet ejection malfunctions Machines Measurement Mother Nature Man Material Method Tablet ejection malfunctions Machines Measurement Mother Nature Figure 3 Pareto chart Dosing instructions unclear Tablet too large Unsure of benefit Ordering/refilling too complicated Side-effects Too many other meds Forgot Lost prescription Didn't fill prescription Didn't want to buy/too expensive Number of respondents Percent 100 80 60 40 20 0 140 120 100 80 60 40 20 0 Number of respondents 47 32 21 12 8 5 3 2 2 1 Percentage 35.3 24.1 15.8 9.0 6.0 3.8 2.3 1.5 1.5 0.8 Cumulative percentage 35.3 59.4 75.2 84.2 90.2 94.0 96.2 97.7 99.2 100.0 Reasons for refusal Man f-Locwinart_24-35_Masters 7/2/14 1:29 PM Page 26