HFAP requirment for alarm testing

[Dansket]

This is the HFAP requirement:

08.01.05 Blood and Blood Product Storage.

Blood and blood products must be stored under appropriate conditions that include an adequate temperature alarm system that is regularly inspected.

(1) An audible alarm system must monitor proper blood and blood product storage temperature over a 24-hour period.

(2) Inspections of the alarm system must be documented.

§493.1271(c)(1)(2)

 

This is the inspectors citation of deficiency:

This standard is not met as evidenced by the following:

"Review of records and interview with staff and Biomed revealed alarm testing is being performed quarterly.  However the blood bank refrigerator charts do not reflect any spikes and there was no documentation on the charts that indicated that there were no checks".

 

Her first sentence acknowledges that we are doing quarterly alarm testing.  In her second sentence, she states that the lack of "spikes" suggests that we are not doing quarterly checks, hence the citation of deficiency.

 

I'm chaffing at the bit because I believe the FDA requires that the audible temperature alarm system be tested regularly. I don't see the temperature chart recorder as part of the Audible Temperature Alarm system.  We use a digital thermometer (Fluke) to monitor the temperature of the liquid as it approaches and exceeds the 1.5C and 5.5C set points.  A digital thermometer will respond to temperature changes in milliseconds, whereas a temperature recording chart pen takes seconds to respond to temperature change.  Once the audible alarm sounds the alarm probe is immediately returned to 4C liquid and as a result, no spike in the tracing.

 

I understand that attempts to correlate the temperature chart recording with the audible alarm sounding as being one tactic for controlling the process, but not being prescribed by an inspector as the only acceptable method.  Our temperature charts are printed in one (1) degree intervals.  You can't read a temperature chart to a tenth of a degree! We record this reading each day as a whole number (no fractions).  We record the digital temperature display each day in the format n.n.

 

Again I believe the purpose of alarm testing is to verify that the audible alarm is sounded if the temperature sensed by the probe reaches the alarm set points.  It is not to check the calibration of the temperature chart recording pen!

 

Say you?

Edited April 19, 2016 by Dansket

[SMILLER]

I guess the thinking is, you do not use that digital thermometer to set off alarms on a 24/7 basis, so the inspector is not going to accept this as proof that the alarm system works.

And really, without a spike on a chart, what hard copy do you have to show for a valid alarm check? 

The whole process is really a chore, especially when you get to checking freezers.  Having said all that, it does seem to me that the inspector is being a bit too picky.  If the digital thermometer is certified, I am not sure what their problem was.

Scott

[Dansket]

Scott.. 

Do I understand correctly, that your process for determining the temperature of audible alarm activation is based solely on a reading taken from the temperature chart pen tracing?

Dan

 

Edited April 19, 2016 by Dansket

[SMILLER]

Not exactly.  We pretty much do what you do but with a regular glass NIST thermometer.  Put the thermometer in a cup of water with the probes (alarm and recorder) and slowly add ice to it/add warm water etc, to be able to document when the alarm goes off.  Naturally this will cause a spike on the chart recorder.  We have never had an issue with the FDA or JCAHO as far as I know. 

But, anyway, your inspector seems too picky to me.  If you have documentation that you checked the alarms with a NIST thermometer, it seems to me that you should be good, regardless of whether its on the chart or not.

Scott

Edited April 19, 2016 by SMILLER
forgot a t

[goodchild]

Dansket, as long as you have documentation that describes the specific range for each storage device, what the alarm set points are, the temperature at which the alarm activated, and that the temperature didn't violate any of these constraints then I completely agree with your system. Alarm activations are annoying enough, why do inspectors feel the need to make them more onerous?

[pinktoptube]

Can't you just document it in your procedure that the chart will not correlate with the thermometer reading due to the sensitivity/time to change?

[Dansket]

When performing tube testing, tubes are read for the presence/absence of agglutination.  Results of these observations are documented in computer/on paper.  Tubes are discarded.  At that point, what direct proof do you have that the tubes were graded correctly and results entered correctly in computer/on paper?

The only proof was in the contents of the test tube and it is gone!

When an audible alarm system is tested, the only proof is the brief sounding of the alarm and then it's gone too!

 

[AMcCord]

Does your fridge have a digital display that notes the alarm? My Helmer's do and will hold 50 records of alarm and recovery (25 events). You could do a screen shot of the display showing the temp alarm (time of activation, time of recovery and the associated temps).

OR, my personal favorite, contest the citation. I've been checking alarms for a very long time and we've never been cited for not having a spike on the chart. There often is one, but not always if I'm really quick that day. I've never had an inspector look for that, including the FDA.

[goodchild]

CAP TRM.42850:

Alarm Sensors To Trigger Action Needed

Alarms are adjusted to be triggered before the temperature falls outside the 1-6° C

acceptable temperature range for refrigerators, or outside the acceptable range for

freezers and platelet incubators.

NOTE: Refrigerators, freezers and platelet incubators must have alarm systems that provide

opportunity to take action before the temperature of blood or components is outside of

acceptable ranges. Red cell units stored at temperatures higher than 6° C may be subject to

accelerated bacterial growth. Temperatures below the freezing point may induce hemolysis.

Freezers need not be operated at their lowest possible temperature, since some plastic plasma

containers held at temperatures lower than -25° C may exhibit increased breakage rates upon

handling.

Evidence of Compliance:

✓ Records of trigger temperatures during alarm checks AND

✓ Records of corrective action, when appropriate

 

The CAP perspective when I last discussed this topic with them at great length:

Your alarm activation documentation must show the temperature at which the alarm activated, to prove that they are functioning within the constraints that this checklist item describes. We have an electronic continuous temperature monitoring setup here, so we can load up the alarm activation event and print a nice little report for every probe in every device to satisfy the inspectors.

Perhaps you could create a spreadsheet with all the devices/probes, their temperature ranges, their alarm setpoints, and the temp at which the alarm sounded. A site that I inspected earlier this year did exactly this.

Edited April 20, 2016 by goodchild
last sentence added

[Dansket]

Thanks for all your ideas!  I plan to contest this.  Will post outcome.

[David Saikin]

When I do alarm checks the probe that runs the chart is not affected as I take the alarm probe out of the solution and place it in my baths for temp checking.  The only charts that move are my freezer and plt storage as there is only one probe.  AND my water bath alarms do not have a chart.

If you have documentation that you are doing the procedure they cannot impune that you are not.  I would respond that you feel you are in compliance and then see how the regulatory agency feels.  Inspectors are not always correct in their interpretation of standards.

[Dansket]

HFAP headquarters determined that we do meet the intent of the standard.  So no deficiency cited or on record.  Thanks for you replies!!

[mollyredone]

I agree that if I'm quick, the chart does not spike, nor does it show up in Isensix.  Good to know it could be contested!  I have another question.  

COM.30725 Non-certified Thermometers

If digital or other displays of temperatures on equipment are used for daily monitoring, the laboratory must verify that the readout is accurate. So if we use the digital temperature, does that need to be checked separately against an NIST thermometer?  What if we also have Isensix and it is checked against an NIST thermometer annually?  Is that sufficient correlation if we show our digital readout matches Isensix?  Hope that makes sense!