Reduced central line infections? Check.

By ThinkReliability Staff

Sinai-Grace Hospital in Detroit has achieved remarkable reductions in bloodstream infections associated with central lines. They’ve reduced the rate of infections significantly by implementing a simple procedure and checklist. We will perform a root cause analysis that shows how these gains were achieved.

First, the hospital needed to determine what was at stake. Over 18 months, it was estimated that more than 1500 patients would die from infections. This is an impact to the patient safety goal. There was non-compliance with procedures, which is an impact to the compliance and organization goals. Infections result in a longer intensive care unit (ICU) stay, which is an impact to the patient services goal. Lastly, the hospital estimated that, over the 18 months, it would spend $175 million in additional costs from these infections.

Next, the stakeholders in the central line insertion process (doctors and nurses) were asked to help determine why these infections were occurring. Bloodstream infections resulting from intravenous catheters result when a catheter is inserted (for vascular access) with bacteria on it. Generally, the bacteria is on the catheter from a missed step in the catheter process which prevents contamination. The steps that were not always being followed were: doctors washing their hands and donning protective wear, patients not being washed with antiseptic or fully draped, and insertion sites not being covered with sterile dressing after the catheter is inserted.

As a solution, a checklist was created that outlined the six steps of catheter insertion. (The outline, Cause Map, process map, solutions, and checklist are shown on the downloadable PDF. To view it, click on “Download PDF” above.) The six steps included the cleanliness steps discussed above. Additionally, the medical professionals noticed that sometimes the procedures weren’t being followed because the necessary equipment was not available in the ICUs. Senior executives from the hospital were assigned to each unit, and were able to properly stock the ICUs. Additionally, the executives got Arrow International to manufacture central line kits that contained the necessary antiseptic and patient drapes.

The progress at Sinai-Grace has been remarkable, by joining all the necessary parties to an effective root cause analysis. Click on “Download PDF” to see what they did. (Read more in The New Yorker Annals of Medicine.)

That’s not my baby!

By ThinkReliability Staff

New mothers and maternity centers alike live in fear that babies will be discharged to the wrong family. The Joint Commission considers discharge of an infant to the wrong person a never event, and it’s no surprise. Even if the mix-up is quickly rectified, huge problems can ensue.

Luckily, we don’t have to wait until an incident happens to us. We can instead perform a proactive root cause analysis, where we consider what COULD go wrong to result in an infant being discharged to the wrong person. A thorough root cause analysis built as a Cause Map can capture all of the causes in a simple, intuitive format that fits on one page.

For our very basic Cause Map, we can consider that two things have to happen in order for a infant to go home with the wrong family. First, the wrong baby has to be given to a family, and second, the matching system has to be ineffective. We’ll break each of these causes down into more detail.

The matching system may be ineffective because it isn’t being used. If a facility doesn’t have a matching system, it obviously won’t be effective. Additionally, if a computerized system is in use there is always the possibility that it won’t be working. Or the staff may not know how to use the system, possibly due to lack of training or insufficient staffing.

If the system is being used but not being checked, it won’t be effective. Again, this could be because the staff doesn’t know how to use the system, or it could just be that the discharge personnel forget.

The wrong baby could be given to a family if the matching system is on the wrong infant. (This is only likely to happen if the matching system is put on outside the delivery room.) Or, the wrong baby could be in a bassinet. This could happen if the baby is kept in an nursery and a nurse is transporting more than one baby..

It’s possible to add even more detail to this Cause Map as the analysis continues. As with any investigation the level of detail in the analysis is based on the impact of the incident on the organization’s overall goals. Once the Cause Map is complete, solutions can be brainstormed that match up with causes. The solutions are shown on the downloadable PDF.

Click on “Download PDF” above to download a PDF showing the Cause Map and Solutions.

How to prevent error? Cut your process in half

By ThinkReliability Staff

Congratulations to Lancaster General Hospital, who received the Best Practices Award In Health-System Pharmacy yesterday (December 7, 2009).  The award was to recognize the hospital’s success in auto-programming of infusion pumps.  As Lancaster’s Medication Safety Pharmacist Amanda Prusch has stated:

Auto-programming provides an additional layer of safety than the stand-alone technologies and by streamlining the IV medication administration process, the potential for error would be reduced.

Lancaster identified a problem – in this case the problem is medication errors, of which 56% are estimated to be from improperly inserted intravenously (I.V.) medications.  Because of the potential for serious adverse patient health outcomes due to medication errors, Lancaster knew something had to be done.  Lancaster looked at the infusion pump programming process, and discovered it had 17 steps, many of them manually performed.  At each manually-controlled step in a process is an opportunity for human error, no matter how careful people are.

By reducing the number of steps in a process, and by reducing the amount of human input required, the opportunity for human error is reduced.  In order to reduce this risk, Lancaster has implemented “fully auto-programmed smart infusion systems” .  The implementation of these systems has reduced the number of steps in the infusion pump programming process to 7 (from 17).  As a result, they have seen a greater than 90% reduction in reprogramming (or errors that were noticed and fixed).  This has surely led to a great reduction in I.V. medication administration errors.  In addition to the patient safety improvements, it has the added bonus of reducing the amount of time spent programming the pumps by 24.8%.  This also contributes to patient safety because it frees up nurses’ valuable time.

Read more about the new infusion pump system.   Lancaster’s solution is just one way to solve the problem of medication errors.  However, by removing steps from a process, not only does it require less time of your most valuable asset (employees), it can reduce the possibility for error.

The surgery went well . . . except it wasn’t supposed to be done on you.

By ThinkReliability Staff

A “never event” that should come as no surprise considering its profound implications on patient safety is performing a procedure on the wrong patient. Ordinarily there are many checks to ensure that a patient’s identity matches that on the procedure order.

However, sometimes a combination of oversights can result in these checks being ineffective. This is especially true when there are patients of similar names, as happened in this particular case.

A patient (patient 1) was scheduled for a procedure in the EP lab. The EP lab called up to Patient 1’s floor, and were directed (incorrectly) to another floor by a person on the telephone. Although Patient 1 was on that floor, another patient with a similar name (patient 2) had been moved to another floor. The EP lab then directed Patient 2’s nurse to bring Patient 2 to the lab. The nurse brought Patient 2 to the lab, over her objections, despite the lack of information in Patient 2’s chart, or a signed consent form.

The various staff members at the EP lab did not verify the identity of the patient, either. However, they did notice the lack of consent form, and convinced Patient 2 to sign a consent form, for a procedure she did not need, was not scheduled for, and had opposed throughout the process.

Although the procedure was stopped partway through, when the true identity was realized, and there was no lasting injury to Patient 2, any procedure performed on the wrong patient has huge risk for patient safety.

On the downloadable PDF page, a simplified version of the process for taking patients to procedures is shown. The individual causes of the incident are identified in the Cause Map, and, where applicable, the process map. This example shows how a process map can be used to identify the various causes that led to an event.

A thorough root cause analysis built as a Cause Map can capture all of the causes in a simple, intuitive format that fits on one page. Even more detail can be added to this Cause Map as the analysis continues. As with any investigation the level of detail in the analysis is based on the impact of the incident on the organization’s overall goals.

Click on “Download PDF” above to download a PDF showing the Cause Map and Process Map.