We’ve had several recent columns on 1000-fold heparin overdoses related to heparin flushes. Our December 2007 “What’s New in the Patient Safety World” column discussed the continued occurrence of 1000-fold heparin overdoses. This included discussion about the Cedars-Sinai incident involving, among others, the children of actor Dennis Quaid and the previous Indianapolis incident that resulted in death of 3 infants. We discussed many of the ISMP recommendations and listed multiple lessons learned. There was also followup in the April 2008 “What’s New in the Patient Safety World” column.
Now the National Patient Safety Agency of the UK NHS has just issued an alert “Risks with Intravenous Heparin Flush Solutions”. Timely because of all the recent cases of heparin overdose in the US due to inadvertent administration of high dose heparin, the alert was actually issued after an extensive root cause analysis in the UK of inadvertent high dose heparin administered by an anesthetist during procedures of 4 successive pediatric patients. Though the investigator had difficulty identifying other cases from the UK incident reporting system (because of problems with data structure and taxonomy of the database), he did find at least one other comparable case. The cases are very similar to the ones in Indiana and California that have received substantial media attention in the US.
The alert and its additional information point out that heparin flushes have no advantage over normal saline for maintaining peripheral intravenous catheters (evidence is less clear for arterial catheters or central venous catheters). Though there is minimal drug cost savings in moving to use of normal saline for such flushes, keep in mind that unnecessary use of heparin also exposes patients to the risks of allergic reactions, bleeding complications, medication error, and heparin-induced thrombocytopenia. And some other medications may be incompatible with heparin when injected into the same catheter. The alert also stresses that heparin flushes should not be “routine” and should only be used when specifically prescribed. It also reaffirms previous NPSA recommendations on injectable medications such as only stocking low-dose heparin products in areas other than pharmacy, correctly labeling all syringes, and using double checking systems when preparing medication for injection and when administering such injections.
But the real value in this alert is the independent review/RCA, which is extremely thorough and highlights a whole host of circumstances, human factors, and errors that combined to produce the incident(s). This is a long (and often repetitious) investigation report but we highly recommend you read it in its entirety. It’s filled with many of the things we’ve talked about in numerous columns: time pressures, look-alike/sound-alike (LASA) drugs or packaging, information overload, workarounds, drugs showing up in unexpected places, confirmation bias, failure to communicate important patient safety policy changes, failure to convey important clinical information to those who needed to know, slips, latent errors, culture of safety issues and a host of others. Though it’s not presented in a structured RCA format, most of the key elements of the RCA are readily perceived. And the lessons learned and multiple recommendations made would be good ones for almost any healthcare facility.
In a nutshell, 4 pediatric oncology patients were to have procedures (bone marrow aspirates, lumbar punctures, etc.) performed under anesthesia in a treatment room in an outpatient area. Each patient was to receive a heparin “flush” of their access line. The anesthetist, unaware that vials of high-dose heparin (25,000 units in 5 ml) had inadvertently been placed in the controlled drug cupboard (CDC), administered the high-dose heparin to each child thinking he was using a low-dose heparin “flush” (50 units in 5 ml). He, nor other staff involved in the care of these patients, recognized the errors. By chance, because parents of the children had noted some excess bleeding from puncture sites and nurses performing routine inventory of the medication cabinets noticed missing high-dose heparin vials, the incident was discovered the following morning. Fortunately, no long-term harm came to any of the affected children.
Some enabling conditions were identified. The anesthetist had arrived to the facility somewhat late and, to avoid inconvenience to the patients and their families, proceeded without taking a rest break or lunch break (time pressure).
He examined the four patients who were to have procedures done that day, then went to the treatment room to begin preparation for the procedures. He laid out the anesthetics tray for each of the four cases and syringes for each of the three medications each patient was to receive (they were all to receive the same medications). He did not label any of the syringes. Rather he was using his own method of recognizing what was in the syringe by the size of the syringe. This is an example of a workaround. Not only are there dangers in the failure to label the syringes but the preparation of medications for multiple patients at the same time is also extremely dangerous.
The high-dose heparin was not supposed to be in the CDC. It had been placed there because another adult patient had needed it on several occasions. The box containers for the two heparin preparations were distinctive. However, the four vials of high-dose heparin were not in the box container. The actual vials were difficult to distinguish from the vials of low-dose heparin. The anesthetist filled a syringe from each tray with the high-dose heparin, not realizing it was the wrong heparin preparation (look-alike/sound-alike or LASA issue). In such cases, confirmation bias likely plays a significant role: the vials looked like the ones he expected to use and even the word heparin appeared on the vials. Failure to read the exact label obviously was an error. However, it is clear from multiple similar occurrences of these incidents, that such confirmation bias is common and may affect even experienced well-intentioned healthcare workers. In fact, experienced staff are probably more prone to confirmation bias than inexperienced staff.
At no point during either the preparation of the injectable medications or their administration was a double check performed (i.e. a second observer did not independently verify the correct medication, dosage, route, and patient). Though this was required by policy, this anesthetist and most others were unaware of the requirement (see below in section on policy issues). Also keep in mind that good medication safety also requires verification of the medication at the time of disposal. In this case, the anesthetist simply discarded the used heparin vials in a sharps container.
The investigation does note the controversial issues related to double checks (i.e. that the second checker is more prone to error and, in fact, both checkers may be more prone to error because they assume the other checker correctly verified all the information). However, he cites two research studies that showed a roughly 30% reduction in drug errors with a double checking system.
Another problem identified was that the controlled drug cupboard had over time become an area where many things other than controlled drugs were kept (including personal valuables) and had become vulnerable to access by extraneous staff. In fact, the keys to that CDC were used by so many people that it was often difficult to locate the keys. Hence, one wonders if this had made it more likely unexpected drugs would show up in the CDC. This is an issue of the nature of the local culture of safety. The CDC’s in similar rooms also differed. This was due to an old CDC being moved from the old facility. It exemplifies the importance of standardization.
The anesthetist involved and most of the other anesthetists were unaware of the policy that double checks were required for preparation and administration of intravenous drugs. This was largely a result of policies being too long (some as long as 60 pages), important points not being highlighted, and changes not being highlighted. As a result the policies often went unread by those who needed to see them. And the organization had no formal means of assessing who had read the policies or policy changes. This is an example of both failure of communication and information overload. Sound familiar in your facility?
The investigation also revealed some after-the-incident issues. Parents of two of the children called that evening noting excessive bleeding at puncture sites. That information was never conveyed to the day staff. The following morning the parents of one child brought their child in because of the bleeding. The bleeding had stopped and the child was discharged home. A short time later the vials of high-dose heparin were noted to be missing and it became apparent what had happened. That lack of communication between the evening on-call staff and the daytime staff almost allowed a serious incident to go undetected.
The staff at the outpatient area also had no guidance as to how to proceed with an investigation once the incident was recognized. Having a policy and checklist for what to do in such cases would have been usual (see our July 24, 2007 Patient Safety Tip of the Week “Serious Incident Response Checklist”).
The investigation report provides 51 recommendations regarding changes needed at multiple levels of the pertinent health care systems. The reader is referred to the full investigation report for all those details.
NHS NPSA Rapid Response Report.
Risks with Intravenous Heparin Flush Solutions. April 24, 2008
Rapid Response Report: Risks with Intravenous Heparin Flush Solutions
(Reference: NPSA/2008/RRR02) issued on 24 April 2008
Evidence on heparin flushes recently reviewed by UK Medicines Information (UKMi) indicates no advantage over normal saline for maintaining peripheral intravenous catheters. http://www.druginfozone.nhs.uk/Record%20Viewing/viewRecord.aspx?id=591809
Toft B. Independent review of the circumstances surrounding four serious adverse incidents that occurred in the Oncology Day Beds Unity, Bristol Royal Hospital for Children on Wednesday, 3 January 2007. (UHBT Final Report Heparin). August 2007
Other heparin-flush discussions on our website: