As we were preparing this column on problems with insulin, we encountered a 2007 safety alert by ISMP on insulin-heparin mixups and thought to ourselves “we haven’t seen much of this recently”. Literally 2 hours after we said that, our literature-scanning software started popping up just-released news articles about a serious event in Canada in which four infants were inadvertently given insulin instead of heparin. The 2007 ISMP alert was issued after a neonate suffered severe hypoglycemia after insulin rather than heparin had been added to a bag of total parenteral nutrition (TPN). Ironically, the recent Canadian episode was an almost identical occurrence. Heparin was intended to be added to TPN but insulin (Humulin R) was added instead and the resultant mixture was administered intravenously to 4 neonates at the Canadian hospital. Staff noted the four infants worsening in similar ways and correctly diagnosed and managed the problem, though it is unknown whether any future neurological problems might result from the events.
Details of the root cause analysis (RCA) done at the Canadian hospital are not available though press reports indicate that similarity of the labeling on the two drugs likely played a role, as did insufficient space in the inpatient pharmacy. The hospital has apparently changed the labeling on insulin and heparin, added another safety check (?an independent double check), and is looking for additional pharmacy space.
Interestingly, ISMP Canada had also just released this August a bulletin on inadvertent administration of insulin to a nondiabetic patient, in whom a mixup of heparin and insulin was considered a possibility.
ISMP’s 2007 alert highlighted multiple other instances of mixups between insulin and heparin. They noted that similar packaging, fact that both insulin come in 10 mL vials, fact that both are dispensed in “units”, and fact that vials of both are often kept near each other on countertops, in drug carts, and under pharmacy IV admixture hoods. They also noted that mental slips are involved in most instances. Perhaps contributing also is the fact that insulin infusions have been increasing in recent years.
ISMP made several recommendations to avoid such mixups during drug preparation:
However, probably the two most important issues are whether either heparin or insulin belong in the TPN solutions in the first place. The ISMP article discusses the pros and cons of using insulin separately or as part of TPN solutions. The issue of heparin (supposedly used to avoid IV catheter thrombosis) is also discussed in that ISMP alert. There is no evidence base to support the use of heparin for this purpose in peripheral IV lines. Though heparin flushes may improve umbilical catheter patency in neonates, there is no current good evidence to demonstrate its effectiveness and safety in peripherally placed central catheters in neonates. We’ve previously discussed the issue of heparin flushes on multiple occasions.
The ISMP alert also had recommendations to help avoid heparin/insulin mixups at the point of administration:
Our original intent for today’s column was to review multiple aspects of insulin use in the hospital setting. Insulin, of course, is one of the top 5 high-risk medications at virtually every hospital. It is of concern not only because of the frequency with which insulin is used but also the seriousness of the consequences of incorrect administration (or failure to administer).
Fortunately, in the past year there have been multiple great resources highlighting the pitfalls associated with insulin. There was an ASHP (American Society of Health-System Pharmacists) and a good insulin audit tool from or the Victorian Medicines Advisory Committee (Australia).on management of insulin, on insulin errors (many of the lessons learned that appear in the ISMP webinar came from the Pennsylvania Patient Safety Reporting System so there is a lot of overlap in these two resources), and on insulin issues. The NPSA site also has a good . And some older but equally important resources exist, including a monograph from
The most egregious and potentially fatal errors with insulin have historically been related to use of the abbreviation “U” for units. The problem arises when the “U” looks like a zero so the patient inadvertently is given 10 times too high a dose. The same kind of problems can arise when the abbreviation “IU” for international units is used. The solution is to never allow use of either abbreviation. It’s easy to put these abbreviations on your list of “Do Not Use” abbreviatons but in practice they still keep popping up. An area of particular vulnerability may be your IT systems. You need to purge these abbreviations from all your computer software. We have seen examples continue to pop up unexpectedly even years after we thought all such instances had been purged! They often show up in things like “customized” order sets or in materials coming from third party vendors. So you must have continuous surveillance for these occurrences. In addition, you need to make sure that your dictation systems automatically prevent use of these abbreviations. But even if you always use “units” rather than abbreviations, an order can still be misinterpreted depending on the presentation. If there is insufficient space between the number and “units”, particularly if the “U” in “units” is upper case, someone reading that order may still interpret the “U” as a zero. For example, “NPH insulin 10Units” might be interpreted as 100 units.
CPOE (computerized physician order entry) is the best way to reduce the risk of both handwriting errors and abbreviation errors. We also highly recommend using standardized order sets, whether paper or electronic, and glycemia management protocols whenever possible.
The large number and variety of insulin preparations is problematic. Add to that the look-alike/sound-alike (LASA) issue and it is no wonder that so many insulin errors occur. Think about all the insulin preparations that could easily be mixed up: Humulin, Humalog, Novolin, Novolog, Humulin 70/30, Humulin 50/50, Humulin R, Humulin N, Humulin L, Humulin U, Humalog Mix 75/25, Novolog 70/30, Novolog Mix 70/30, Novolin R, Novolin N, Novolin L. Use of tall-man lettering may help prevent some LASA mixups (eg. humuLIN, humaLOG, novoLIN, novoLOG, etc.)
One problem that may apply to a variety of insulin preparations occurs when the computer display screens, from which a provider is choosing an insulin dose and preparation, displays truncated information.
Many hospitals have attempted to limit the number of insulin preparations they keep in their formulary and some have sought to use multiple vendors to avoid confusion with look-alike vials. That, however, may be problematic because patients come into the hospital on certain preparations of insulin and expect to be continued on the same preparation while an inpatient. It can also become a nightmare for medication reconciliation when patients are discharged.
Another error being seen more frequently is related to the relatively new 500 unit insulin preparations. These preparations come in 500 unit per mL vials. The problem arises when healthcare workers use this dosage form with insulin syringes that a calibrated for 100 unit per mL insulin preparations (note there are currently no syringes calibrated for the higher concentrations). It is recommended that the 500 unit per mL preparations be stored completely separate from the other insulin preparations and only be used by specially designated healthcare workers who have received specific training in use of the higher concentration product. Certainly, they should never be stored as part of “floor stock” where they might be easily mixed up with the more common 100 unit per mL preparations. Better yet, don’t stock this 500 unit per mL insulin at all! Unfortunately, the need for that higher concentration has been increasing in recent years as more obese patients with insulin resistance have been requiring much higher insulin doses. The higher concentrations may also be needed for patients having implanted insulin pumps. If you need to stock the higher concentration, make sure you have some mechanism (like a “hard stop”) to ensure that a pharmacist is involved in validating and preparing the dose and always have an independent double check before such preparations are administered.
Not only are incorrect doses of insulin a problem, but omitted or delayed doses are also a concern. Delayed doses become a concern especially when patients are transported to other areas of the hospital for testing, procedures, etc. We’ve often seen patients spending several hours in the radiology suite, missing scheduled doses of insulin and becoming hyperglycemic or getting a full dose of insulin on return from radiology and then getting hypoglycemia. And in several of our columns on patient safety issues in the radiology suite we have mentioned that IV infusions sometimes get turned off during imaging procedures and these may not be restarted appropriately. If you use a structured tool for your inhospital transports, like “Ticket to Ride” (see our November 18, 2008 Patient Safety Tip of the Week “Ticket to Ride: Checklist, Form, or Decision Scorecard?”), be sure to include an item related to whether insulin should be given or held while the patient is off the floor.
Another problem in insulin omission we have encountered relates to the timing of admission orders. Some pharmacy computer systems are programmed to dispense the “every morning” doses of drugs at 8AM the next morning. We have seen examples where a diabetic patient has admission orders written in the emergency room at 7:30AM but does not have those orders transcribed and entered into the computer system until 8:30 AM. In such cases, there is a risk that a patient could go 24 hours without an insulin dose (and without his or her other daily medications). Handoffs in such situations are critical in ensuring that patients get their intended doses of medications on the correct day.
One error we have encountered on numerous occasions is failure to inquire during medication reconciliation about the time the last dose was taken. We have seen this lead to failure to give any insulin on the day of admission or to delays in insulin administration when patients undergo long delays between presentation to the emergency room and ultimate admission to the hospital. On the other hand, it can also lead to inadvertent administration of two doses in one day (i.e. the patient took a dose at home and gets a repeat dose in the hospital). Your medication reconciliation forms, whether paper or electronic, need a column for time last dose taken. While that may be of little consequence for many drugs, it is clearly very consequential for high-risk drugs like insulin.
An excellent patient safety intervention for high-risk drugs is setting dose range limits on your CPOE or pharmacy IT systems. This is very valuable in preventing, for example, overdoses of chemotherapy agents. For insulin, it is much more difficult than it sounds. That is because the dosages of insulin used are so variable across patients. But it is worth looking at your data and saying “we’ve seldom used a dose of insulin exceeding x units” and then adding an alert that helps physicians, pharmacists or nurses question orders for large doses of insulin.
A discussion about “sliding scale” insulin therapy is beyond the scope of today’s column. However, suffice it to say that such practice has become frowned upon (though every hospital we’ve ever visited has examples where someone is still using it!). The PPSA Advisory on insulin errors has a good discussion on this issue and provides numerous examples of problems arising from “sliding scale” or “insulin coverage” orders. When insulin is being given based on the results of bedside glucose testing (glucometers), there have been multiple instances where nursing notes scribbled on a pad for multiple patients have led to patient weights or even room numbers being mistaken for glucose levels!
The syringes used to administer insulin have been a source for errors in insulin management. Insulin is supposed to be prepared and administered in special syringes calibrated in units. These are usually one mL syringes standardized for 100 units per mL insulin preparations. A serious problem arises when someone tries to administer insulin from regular intravenous syringes. In the latter the calibrated markings indicate volume, not units. So a provider erroneously using such a syringe might fill it to 4 mL, thinking this means 4 units, and actually be administering 400 units of insulin!
Use of insulin pens has increased substantially in recent years, with many advantages. However, one must be very careful that insulin pens are never used on more than one patient since there have been numerous examples of cross contamination with blood-borne diseases from insulin pens (FDA 2009).
Another practice being used more and more is the insulin drip. A good recent article (Maguire 2010) discusses how in many hospitals insulin drips are not just being used in ICU settings but are also being managed on regular floors. Note, however, that the PPSA Advisory on insulin errors noted several cases where IV bags containing insulin were mistakenly hung instead of other IV medications.
A real problem nowdays is related to the use of hospitalists. While we are huge advocates of hospitalist programs, we are seeing problems arising when patients are co-managed by more than one physician. In particular, we often see patients on surgical services having their diabetes managed by hospitalists (or other consultants). In such cases it needs to be made very clear to nursing staff who will be responsible for the insulin and diabetes management orders.
Independent double checks are often recommended when we are dealing with administration of high-risk medications. Even though we have emphasized that double checks are a relatively weak intervention (we know from all industries that the error rate when a supervisor checks someone else’s work may be 10% or higher), the literature supports a medication error reduction of about 30% when using a double check system (see our July 15, 2008 Patient Safety Tip of the Week “Publicly Released RCA’s: Everyone Learns from Them” for a description of independent double checks). Another nice article on independent double checks in preventing medication errors (ISMP Canada 2005) describes the independent double check process and calculates that independent double checks would reduce the error rate of a process having an error rate of 5% all the way down to 1 in 400.”). Also, for any high-risk medications you need to do truly independent double checks (see our March 30, 2010 Patient Safety Tip of the Week “
One last problem in managing patients on insulin deals with those patients having insulin pumps. Both external and implantable insulin pumps may present risks during MRI procedures (Shellock 2010). The pump motors may be damaged by exposure to the high electromagnetic fields generated during MRI. So in many cases the pump (and in some cases the transmitters or sensors) may have to be removed prior to entering the MRI environment. The Shellock article provides details for each pump by manufacturer for what to do prior to MRI and also what to do if a patient with an insulin pump is inadvertently exposed to MRI.
So what should your organization be doing to reduce errors related to use of insulin? We have several recommendations:
We have developed over the years a healthy respect for the complexities associated with use of insulin and its potential dangers. Insulin is truly a high risk medication and deserves a unique position in your patient safety activities.
ISMP (Institute for Safe Medication Practices). Action needed to prevent dangerous heparin-insulin confusion. ISMP Medication Safety Alert Acute Care Edition 2007; May 3, 2007
Region's handling of error angers moms. Babies given insulin instead of blood thinner
By Lana Haight, The StarPhoenix October 23, 2010
Premature babies mistakenly given insulin in their IVs.
By Hannah Scissons, The StarPhoenix October 21, 2010
Changes made after 4 babies given wrong drug
The Canadian Press
21 October 2010
Baby dies 2 weeks after given wrong drug
The Canadian Press
22 October 2010
ISMP Canada. Inadvertent Administration of Insulin to a Nondiabetic Patient. ISMP Canada Safety Bulletin 2010; 10(6): 1-3 August 31, 2010
ISMP Teleconference October 15, 2009. High Alert Medications Series Part IV. Preventing Errors from Insulin: A Multidisciplinary Approach
(you can purchase the CD at this website)
Pennsylvnia Patient Safety Authority. Pennsylvania Patient Safety Advisory. Medication Errors with the Dosing of Insulin: Problems across the Continuum .Pa Patient Saf Advis 2010 Mar;7(1):9-17
NPSA (UK). Rapid Response Report. Safer administration of insulin. June 2010
the report: Safe administration of insulin.
e-learning module on safe use of insulin
American Society of Health-System Pharmacists; Hospital and Health-System Association of Pennsylvania. Professional practice recommendations for safe use of insulin in hospitals. Rockville (MD): American Society of Health-System Pharmacists
Victorian Medicines Advisory Committee. Subcutaneous insulin audit tool. Melbourne (Australia): State Government of Victoria Department of Health
FDA. Information for Healthcare Professionals: Risk of Transmission of Blood-borne Pathogens from Shared Use of Insulin Pens. FDA Alert March 19, 2009
ISMP Canada. Lowering the Risk of Medication Errors: Independent Double Checks. ISMP Canada Safety Bulletin. January 2005
Shellock FG. MRI safety: Patients with insulin pumps require special cautions. DiagnosticImaging.com October 5, 2010
Maguire P. IV insulin on the floor: not so scary after all. How a protocol first championed by hospitalists has been embraced by nurses. Today’s Hospitalist 2010 (October 2010)
Print “PDF version”