Medication errors with the dosing of insulin: problems across the continuum.

Medication Errors with the Dosing of Insulin: Problems across the Continuum
Pa Patient Saf Advis 2010 Mar;7(1):9-17.
©2010 Pennsylvania Patient Safety Authority

PA-PSRS Patient Saf Advis. March 2010;7:9-17.

This article analyzed 2685 event reports involving insulin and found that the most common error types were drug omission, wrong-dose, and wrong-drug errors.



ABSTRACT

Controlling blood sugars with insulin is essential in the management of hyperglycemia in both diabetic and nondiabetic patients. However, studies have shown that the use of insulin has been associated with more medication errors than any other type or class of drug. From January 2008 to June 6, 2009, Pennsylvania healthcare facilities submitted 2,685 event reports to the Authority that mentioned medication errors involving the use of insulin products. The most common types of medication error associated with insulin were drug omission (24.7%) followed by wrong-drug errors (13.9%). More than 52% of the reported events led to situations in which a patient may have or actually received the wrong dose or no dose of insulin (e.g., dose omissions, wrong dose/overdosage, wrong dose/underdosage, extra dose, wrong rate errors), which could lead to difficulties in glycemic control. Strategies to address these problems include limiting the variety of insulin products on the organization’s formularies, developing standardized protocols and a standard format for prescribing insulin, avoiding the use of abbreviations or other shortcuts when communicating orders for insulin, and requiring an independent double check of all doses before dispensing and administering intravenous insulin.

Introduction

An estimated 23.6 million Americans (nearly 8% of the U.S. population) have diabetes mellitus. In 2007, approximately 17.9 million people have been diagnosed with the disease, and 5.7 million remain undiagnosed.1 Among adults diagnosed with type 1 or type 2 diabetes, 14% take insulin only, 13% take both insulin and oral medication, 57% take oral medication only, and 16% do not take either insulin or oral medication.2

Glycemic control is fundamental to the management of diabetes. Insulin is used to control blood sugars in both diabetic and nondiabetic patients. For example, it is used to manage hyperglycemia in intensive care unit (ICU) patients, a common finding caused by insulin resistance in the liver and muscle tissue. Some have considered insulin resistance to be an adaptive response, providing glucose for the brain, red blood cells, and wound healing.3

Due to a number of conflicting published studies, there has been an increased effort to determine the benefit of tightly controlled blood glucose levels, both in diabetic and nondiabetic patients. For example, in a large, single-center study of postoperative surgical patients, an initial investigation by van den Berghe et al. suggested that controlling blood glucose levels by intensive insulin therapy decreased mortality and morbidity in critically ill surgical patients.4 The study design employed a continuous infusion of insulin to maintain blood glucose between 80 mg/dL and 110 mg/dL. Patients receiving intensive insulin therapy were found to be less likely to require prolonged mechanical ventilation and intensive care. Also, rigorous insulin treatment reduced the number of deaths from multiple-organ failure with sepsis, regardless of whether there was a history of diabetes or hyperglycemia.

The NICE-SUGAR study evaluated whether there was a difference in mortality between subjects randomly assigned to either intensive glucose control (target blood glucose range of 81 mg/dL to 108 mg/dL) or conventional glucose control (target of 180 mg/dL or less).5 The study showed that the odds of dying with intensive control were 1.14 times greater than with conventional control. In addition, severe hypoglycemia (blood glucose level of 40 mg/dL) occurred in 6.8% of the intensive-control group and 0.5% in the conventional-control group. The NICE-SUGAR study also demonstrated that there was no significant difference between the two treatment groups in the median number of days in the ICU or hospital or in the median number of days of mechanical ventilation or renal-replacement therapy.

In a meta-analysis of randomized controlled trials of tight glucose control versus usual care in critically ill adults, the authors found no significant difference in hospital mortality or new need for dialysis. Although tight glucose control was associated with a significant reduction in septicemia overall, subgroup analysis suggested this benefit was limited to surgical ICU patients. Conversely, they found clear evidence that hypoglycemia increased roughly fivefold, regardless of the ICU setting, and was more common with patients receiving very, rather than moderately, tight glucose control.6

For many years, literature has shown that the use of insulin has been associated with more medication errors than any other type or class of drug. Cohen et al. reported in 1998 that 11% of harmful medication errors result from insulin misadministration.7 The U.S. Pharmacopeia MEDMARX 2008 data report showed that insulin was the leading product involved in harmful medication errors (i.e., National Coordinating Council for Medication Error Reporting and Prevention [NCC MERP] harm index8 E to I), representing 16.2% of all harmful medication error reports.9 In 2004, the Pennsylvania Patient Safety Authority established that 25% of all medications errors reported involve high-alert medications, and 16.3% involved insulin products.10 This article presents analysis of events involving insulin products reported to the Authority during an approximately 17-month period and describes the most common types of errors involving the use of insulin, as well as those events that could contribute to uncontrolled blood sugars.

A Look at the Numbers

Pennsylvania healthcare facilities submitted 2,685 event reports to the Authority from January 2008 to June 6, 2009, that mentioned medication errors involving the use of insulin products. Categorization by harm score, which is adapted from the NCC MERP harm index,8 shows that 78.7% (n = 2,113) of the events reached the patient (harm index = C to I) and 1.8% (n = 49) of the events resulted in patient harm (harm index = E to I). The care areas most often cited in these reports include medical/surgical units (22.3%, n = 599), pharmacy (8.7%, n = 234), and telemetry (7.1%, n = 191). Roughly 53% (n = 1,434) of the events involved elderly patients (ages 65 years and older), while 1.7% (n = 46) involved pediatric patients (ages younger than 17 years).

The predominant medication error event types associated with insulin (see Table) were drug omission (24.7%, n = 662) followed by wrong drug (13.9%, n = 374) and wrong dose/overdosage (13%, n = 348). More than 52% (n = 1,409) of the reported events led to situations in which a patient may have or actually did receive the wrong dose or no dose of insulin (e.g., dose omissions, wrong dose/overdosage, wrong dose/underdosage [5.1%, n = 137], extra dose [8.5%, n = 227], wrong rate [1.3%, n = 36]), which could lead to fluctuations in glycemic control.

Table. Predominant Medication Error Event Types Associated with the
Use of Insulin (N=2,057, 76.6%), January 2008 to June 6, 2009
please see the full-text


Wrong-Drug Errors Associated with Insulin Products

There are numerous case reports in the literature that discuss the issue of wrong-drug medication errors with insulin products due to similarities in the brand and generic names, as well as similarity in labeling and packaging.11-14 The Authority has noted mix-ups between names occurring in Pennsylvania facilities (e.g., Humalog® and Humalog 75/25, Humalog and Humulin® R, Humalog 75/25 and Humulin 70/30, Novolog® and Humalog, Novolog 70/30 and Novolin® 70/30).15

During review of the wrong-drug medication errors, analysts found that facilities did not enter the actual name of the insulin products consistently into the reports. In fact, 70% (n = 262) of the submitted reports did not list a specific insulin product (e.g.,“insulin,” “regular insulin,” “NPH insulin,” “insulin 70/30”) or listed names of products that do not exist (e.g., “Humalog 70/30,” “Humalog R,” “Humulin 75/25”). This imprecise data collection limits individual facilities and the Authority from accurately determining the most common pairs of insulin products involved in wrong-drug errors. In addition, many of these reports did not specifically state why the error occurred or what went wrong that led to the patient receiving the wrong insulin product. Therefore, it was not possible to determine the most common types of wrong-drug errors that occurred (e.g., wrong drugs that may have been written by prescribers, selected during order entry, mislabeled in the pharmacy, wrongly pulled from stock). Analysts were able to determine the following:

Seventy-five (20%) reports of wrong-drug insulin errors specifically mention that the breakdown occurred when retrieving the medication, for example, from stock or an automated dispensing cabinet (ADC). Specifically, 28 reports (37.3% of stock errors) mentioned the use of overrides to obtain the insulin product from an ADC.
Sixty-nine (18.4%) of wrong-drug insulin errors involved mix-ups between a rapid acting insulin (e.g., Novolog, Humalog) and regular insulin (e.g., Novolin R, Humulin R, regular insulin).
Sixty-five (17.4%) reports of the wrong-drug events specifically identified that the error occurred during the prescribing node. Most of these reports involved the clarification of nonspecific (e.g., a specific insulin product was not indicated) orders, such as the following:
The physician wrote an order for “Novolin 18 units bid.” The order was not clarified when taken off, and regular insulin was given for two doses. When the physician came in the following day, the order was clarified, and he ordered Novolin N insulin. The patient was given two doses of Novolin R.

Intravenous (IV) administration of insulin has some advantages over subcutaneous administration, namely (1) more rapid onset of effect in controlling hyperglycemia, (2) more overall ability to achieve glycemic control, and (3) improved nonglycemic patient outcomes.16 During IV insulin infusion to control hyperglycemic crises, hypoglycemia, if it occurs, is short-lived; however, repeated administration of subcutaneous insulin may result in “stacking” the insulin’s effect, causing protracted hypoglycemia.16

The stability of an IV insulin infusion is 24 hours and requires the production of insulin infusions by pharmacy when ordered. Unless this infusion is distinguished with highlighting or a prominent sticker, an insulin infusion will resemble other pharmacy-prepared infusions. Of the wrong-drug errors involving insulin reported to the Authority, infusion bags containing insulin were mentioned in 9.4% (n = 35) of the cases. Nearly 88.6% (n = 31) of these reports reached the patient, and 11.4% (n = 4) resulted in patient harm. Patients accidentally received insulin instead of a noninsulin-containing infusion (e.g., antibiotics) in 60% (n = 21) of these wrong-drug, infusion-related reports. Examples are as follows:

An IV insulin bag was hung when replacing the patient’s Versed® (midazolam) bag. Two bags of insulin were then hanging, one at rate of 8 (Versed rate) and one at 5 (insulin rate). A [mid-afternoon] accucheck showed that the [blood glucose] level decreased to 36. D50 was administered as per protocol, and the insulin drip was turned off. The wrong-bag error was found at [the next] change of shift.

A patient was ordered IV Lasix® (furosemide), as well as IV insulin. The nurse meant to hang the IV Lasix but [before midnight] hung a bag of IV insulin instead. The patient already had an insulin infusion running. Approximately [four hours later], the patient was noted to be hypoglycemic. Both IV drips were turned off at that time, and the patient was given 50 ml of 50% dextrose. The RN [registered nurse] still believed that one of the IV drips was Lasix at this time. Four hours later, the oncoming RN was checking and verifying the patient’s IV drips and discovered the error. The patient required several more doses of 50% dextrose throughout the morning to correct episodes of hypoglycemia.

Wrong-Dose Errors with Insulin

Analysis of events that resulted in patients receiving a type of wrong dose (e.g., wrong dose/overdosage, wrong dose/underdosage, extra dose) reveals a variety of breakdowns that occurred in the medication-use process, including problems with insulin coverage orders, ambiguous orders written by prescribers, transcription and order-entry errors, the obtainment and/or use of the incorrect blood glucose value of the patient, and the ways in which information about insulin products is displayed on pharmacy labels and medication administration records.

About 10.4% (n = 36) of the wrong dose/overdosage events reported to the Authority involved a tenfold overdose of insulin.

Insulin Coverage Orders

The Diabetes Control and Complications Trial, a prospective, randomized controlled trial of intensive versus standard glycemic control involving inpatients with relatively recently diagnosed type 1 diabetes, showed that improved glycemic control is associated with significantly decreased rates of microvascular (retinopathy and nephropathy) as well as neuropathic complications.17 This led to the recommendation that type 1 diabetes be treated by using multiple insulin injections (three to four injections per day of basal and prandial insulin) as well as by matching the dose of prandial insulin to carbohydrate intake, pre-meal blood glucose, and anticipated activity.

However, the use of multiple-dose injections of insulin throughout the day has added complexity to controlling a patient’s blood glucose. For example, correction doses, sometimes referred to as “coverage” or erroneously as “sliding scales,” are used to adjust glucose levels around mealtimes. Organizations often have multiple algorithms for corrections doses, such that a facility may have “low dose,” “medium dose,” and “high dose” algorithms that require the nursing staff to obtain and document each patient’s blood glucose reading, determine the patient’s ordered algorithm, and then select the proper dose based on the blood glucose reading.

The predominant theme mentioned in reports of wrong-dose events involves the dosing of insulin based on a range of blood glucose values with a corresponding coverage dose, determined by a patient’s blood glucose reading. Of the wrong-dose errors submitted to the Authority (n = 712), 26% (n = 185) mention coverage or sliding scales. (Many events reported to the Authority used the phrase “sliding scale” in the narratives to denote the method used to determine the dose of insulin to administer to patients. While this term may be used in place of “correction dose” or “coverage,” it should be noted that sliding-scale insulin regimens used alone are ineffective and potentially harmful. When using subcutaneous insulin injection therapy, scheduled or standing insulin regimens should be the standard of care.18-21) As mentioned previously, this recommended method of maintaining tight control of a patient’s blood sugar, regardless if the patient is diabetic or not, adds complexity to the medication-use process for all healthcare practitioners.

One problem often seen with coverage orders is the clarity of handwritten orders from physicians, a particular problem when an organization does not have a standardized protocol or order form to order insulin, including the type of coverage (e.g., low, high). Adding to the complexity of these orders are the multiple values often used for multiple ranges of blood sugars. Problems have also occurred when shortcuts are taken when writing these types of orders for insulin. For example, orders have been written stating doses of insulin as “6+1” or “6+2” instead of writing out “7” or “8” (see Figure 1).

Figure 1. Example of Ambiguous Insulin "Coverage Order"
please see the full-text

Once these complex orders have been written, problems have occurred when transcribing the orders to medication administration records (MARs) or entering them into computerized order-entry systems. Errors also have occurred when selecting the blood glucose range, dose, or algorithm from a pharmacy label, a handwritten MAR, or a computer-generated MAR (see Figure 2). Pennsylvania facilities are experiencing these types of errors as evidenced by these events reported to the Authority:

A patient was ordered insulin on sliding scale level 2, but the order was transcribed incorrectly on the MAR as sliding scale level 1. The patient received two doses at level 1 coverage instead of level 2. The error was found during the 24-hour MAR check.

A patient was changed from high-dose sliding scale coverage to moderate dose. Order was transcribed onto medication sheet as bedtime coverage, but original order was for no bedtime coverage. Patient received four units of insulin.

open here to see the full-text (large):
http://www.patientsafetyauthority.org/ADVISORIES/AdvisoryLibrary/2010/Mar7(1)/Pages/09.aspx