Supplier Partnership Not Always an Option
Joel Ericson’s “Lean Inspection Through Supplier Partnership” (November, p. 36) seemed behind the times.
My company’s just-in-time customers do not perform incoming inspection, and supplier manual inspection is not an acceptable solution. Our customers expect error proofing either by prevention at a workstation or detection at subsequent stations throughout a given process.
We might use a step-down inspection process to verify a corrective action implemented over time. Otherwise, inspection is primarily initial containment.
ISO 9000, Baldrige Criteria Perfect for Healthcare
I enjoyed R. Dan Reid’s excellent November “Standards Outlook” column (“Developing the Voluntary Healthcare Standard,” p. 68) outlining the background and concept of a voluntary healthcare standard. Reid clearly explains the work ASQ’s Healthcare Division and the Automotive Industry Action Group (AIAG) have done in developing the IWA-1 ISO 9000 document tailored to healthcare.
Unfortunately, few healthcare institutions have adopted ISO 9000 concepts. When I last checked, fewer than 25 U.S. hospitals were registered to ISO 9001. Most hospitals and physician groups are unfamiliar with ISO 9000 and quality management systems (QMSs) concepts and do not consider pursuit of registration value added at this time because it is not tied to accreditation.
Our experience with ISO 9001 registration at Physicians’ Clinic of Iowa has been extremely positive. We registered in 2003 and have seen significant improvements in both the clinical and business aspects of our 50-physician group.
Because healthcare institutions generally are much more familiar with the Malcolm Baldrige National Quality Award and Baldrige healthcare criteria, Reid’s approach to developing a new standard that integrates the Baldrige criteria with ISO 9000 should be supported. Clearly, healthcare institutions are interested in Baldrige; 78 of 150 total applicants for the award in 2005 and 2006 were from healthcare, while only four applicants were from manufacturing and 10 from service organizations.
Baldrige and ISO 9000 are complementary and move an organization in the same direction. Although ISO 9000 represents a more basic system than Baldrige, it has important features that are not required by Baldrige, such as a controlled document system, internal auditing and management review.
Also of interest is TUV Healthcare Specialists’ recent application to the Centers for Medicare and Medicaid Services seeking authority to accredit hospitals using a standard tied to ISO 9001 requirements. If such an option is made available to hospitals, it likely will generate much more interest toward ISO 9000, the Baldrige criteria and the overall concept of QMSs in the healthcare sector.
JAMES M. LEVETT
Treasurer, ASQ Healthcare Division
Chief medical officer
Physicians’ Clinic of Iowa
Cedar Rapids, IA
‘QP’ Discussion Board
“QP Mailbag” occasionally publishes edited excerpts from Quality Progress’ online discussion board, open to all ASQ members. To see this entire discussion and post your thoughts, go to www.asq.org/pub/qualityprogress and click Discussion Board, under Resources.
I’d like to start a discussion on November’s “3.4 per Million” column (“Control Charting at the 30,000-Foot-Level,” Forrest Breyfogle, p. 59). I like the overall thrust and goals: control-charting key data to drive decisions; analyzing data to extract the most information; watching out for knee jerk reactions and suboptimization.
However, there are a few ideas and calculations I don’t really agree with, and I would like to get some feedback from other professionals (and perhaps the author).
In general, I don’t see just what makes 30,000-foot-level control charting different from or better than traditional control charting. A 30,000-foot-level control chart presents data “since at least the process’s last shift,” but that doesn’t seem particularly different from how most people use control charts. A 30,000-foot-level control chart “is not intended to be used to determine whether and when timely process input adjustments should be made,” but that seems to be removing one of the key reasons for creating control charts.
More specifically, I wonder about some of the data. First, Figure 1 does not show the data used in the rest of the column. Later analysis shows a three parameter lognormal distribution with location = 3.39, scale = 1, and threshold = -29. However, a plot of random data from this distribution shows a much sharper peak than seen in Figure 1. In addition, with a threshold of -29, the data cannot go below -29, but the plot clearly shows data down to about -60.
The distribution of the data makes a big impact on the interpretation of the X chart (which I believe many would call an I chart). The author states, “Because these data are randomly selected, we would expect the control chart to have a random scatter within the upper and lower boundaries.” This would only be true if the original distribution were symmetric, which it isn’t. Because the distribution is skewed, you actually would expect the data points below the centerline to hug the centerline.
The author also states, “Although you could experience an out of control condition for these random data through chance, you would not expect to experience this with only 35 tracked subgroupings.” Again, be-cause of the distribution involved, this will not be true. Because the smallest possible value is approximately 30 and the lower control limit is -94, it is impossible to get points below the lower limit. And there is a point above the upper control limit.
Barton County Community College
Great Bend, KS