Attribute Control Charts
- A hospital is analyzing nosocomial infections and wants to reduce their infection rate below the national average of 2.0%. They have sampled 40 cases every week for the last 20 weeks. The data are shown below. Create p and np control charts. Interpret the graphs. How is the hospital performing in relation to its stated goal?
- Walter Shewhart ( 1980) presented the following data in his classic Economic Control of Quality of Manufactured Product. Create a p chart for each machine. Does either of the two machines show evidence of special cause variation?
- From January 1846 through December 1848 Semmelweis ( 1983) recorded births and the number that resulted in the death of the mother at his hospital. The data is available in the Chapter07.xls file, in the Problem07–03 tab. Create a p chart to analyze performance. Interpret the chart. Was the system stable?
- Postsurgical infections have been reported to affect 2% to 5% of the 16 million patients who undergo surgery in U.S. hospitals. Infections increase the chance of complications and death. Antibiotics given one hour prior to surgery have been shown to reduce the probability of infection. The director of quality improvement has sampled 20 patients per week over the preceding 25 weeks. The data is available in the Chapter07.xls file in the Problem07–04 tab. The data collected records whether a patient contracted an infection after surgery. Create a p chart to analyze performance. Interpret the hospital’s performance based on your control chart and identify any issues that should be investigated. Assuming the aver- age rate of infection is 3.5%, is the hospital doing a good job?
- Readmission rates within one year for congestive heart failure have been documented at 35%. A local heart program wants to assess its performance against this standard. The pro- gram has randomly selected ten patients per month over a 24-month period for review. The data is available in the Chapter07.xls file, in the Problem07–05 tab. Some of the patients were deleted from the sample due to death, relocation, or other reasons that preclude follow-up. Create a p chart. Is the process stable? How is the program performing relative to the documented standard? Since not all months have 10 observations, either use 10 as the sample size or use the average sample size to calculate the control limits.