Informational Control Chart Interpretation - General "Rules"

I would suggest that a straight forward SPC chart is probably "good enough". At the very least, it will start pointing you in the right direction. Sometimes the KISS principle is sufficient.

Steve Prevette said:

I would suggest that a straight forward SPC chart is probably "good enough". At the very least, it will start pointing you in the right direction. Sometimes the KISS principle is sufficient.

Click to expand...

Not sure how well it will work. Field reliability data are highly variable, usually contain mixtures from different failure modes, and are highly censored.

Thanks for the response. Below are my response:

Subgroup: do you use the calendar time period (month or week),

MONTHLY

Elapsed time since manufacture/shipment or the lot / time or serial number grouping of manufacture?

COMBINATION OF MANUFACTURE AND SERVICE TIMES

Base value: do you use install base, lot quantity or actual usage rate within the subgroup?

INSTALL BASE

Do you use actual failure rates or reported failure rates? Or are these complaint or return/service rates? Many people say they are calculating MTBF/MTTF when they are really plotting only those events that are reported by their Customers...

BOTH ACTUAL AND AS FOUND FALIURES

Anyone help, what type of control chart should i use for medical device in the field. (accumulative install base and service data)

Month Install base Repairs
Dec/14 5882 190
Jan/15 6119 203
Feb/15 6434 208
Mar/15 6877 259
Apr/15 7059 239
May/15 7183 228
Jun/15 7315 276
Jul/15 7331 301
Aug/15 7359 279
Sep/15 7403 256
Oct/15 7424 308
Nov/15 7450 290
Dec/15 7487 306
Jan/16 7517 302

I created two basic control charts. one a I, MR chart and the other is a p chart. they are attached. pretty straight forward and they look just fine.

It appears that there may be slight trend up or a shift in failure rate that occurred around January of 2015. It won't come out as an OOC point in the standard rules. Without knowing what might have happened (is a specific failure mode increased or a new one introduced) its' tough to know if it's a trend or not. Only logic reason and investigation can do that.
The difficulty with control limits for combined field failures (including MTBF) is knowing when the stable period is and then setting proper limits or using the limits from all of the data and making a few 'false calls' that can be resolved by investigation.

I do use these all of the time and quite successfully; it takes a little practice and some real understanding of the 'process'...just like anything else.

oops forgot the file.

Thanks for the advise. Your are correct, the Jan 2015 data was the intro of the device in the market. I agree that the there is a challenge trying to establish the control limits. I was thinking of getting the older version of device(since data are more stable-5yrs worth) and use the control limits across the board. That way any OOC can be more meaningful.

SPC_Newbie said:

Does anyone have access to the document that is pointed to here: http :// www. hanford .gov /safety/vpp/trend.htm - DEAD 404 LINK UNLINKED

It appears to be inactive

Click to expand...

Steve may know.

Hmm second request this week. They are actually hosted on the Cove.

Statistical Process Control Library

YES - I'm going through a lot of old threads checking for dead links in posts going back some years, so old dead links are popping up a lot.