## Algorithms or Sliding Scales

Type 2 Diabetes Defeated

Get Instant Access

A common element of intensive insulin programs is "insulin algorithms": short-term adjustments in doses for variable activity, diet, etc. Unlike pattern management, the dose changes are made only once, to cover the event. Each patient's algorithms are unique, and are based on the patient's and caregiver's identification of diet, exercise, and other habits that cause the blood glucose values to deviate from the usual range, then designing and validating diet and/or insulin fixes. Algorithms are operative when patients tell you, "I play basketball two

night a week after dinner with my friends. Dinner those nights is always pasta, and I reduce my lispro by 2 units. After the game I drink half a glass of Gatorade. Also, 1 always do a glucose test before the Gatorade and it's generally 130 to 150. Rarely it's less than that, and then I drink a full glass of Gatorade." The purpose of these short-term adjustments is to maintain stable glycemia. Thus, by necessity, the changes must be precise and based on testing—it needs to be proven that a 2-unit lowering of the pre-meal insulin dose following a 60-minute game of tennis works better than a 1- or 3-unit adjustment. In fact, as patients start frequent use of SBGM, identifying problem activities and "fixes" becomes second nature for many of them. Often the role of the physician is to make sure the changes are validated with blood glucose testing to prove they work.

Another common algorithm is to build in compensatory variations in the mealtime insulin dose for out-of-range blood glucose levels. An example is shown in Table 5. The dose changes are typically based on the rule of 1800, which states that the glycemia-lowering effect of a 1-unit change in lispro/aspart can be calculated as 1800 divided by the patient's 24-hour insulin dosage (1500

 Pre-meal blood glucose (mg/dl) Lispro/aspart adjustment (units) <60 -2 61-80 -1 81-140 0 141-200 + 1 201-260 +2 261-320 +3 >321 + 4

An algorithm is a supplement or subtraction from the usual mealtime insulin dosage that compensates for the premeal glycemia level. The dosing range is calculated by the rule of 1800 for lispro or aspart insulins and 1500 for Regular insulin: divide the total 24-hour insulin dose by 1500 or 1800 to get the glucose-lowering effect of 1 unit of that insulin. This algorithm was for an individual taking 30 units of insulin daily—thus, adjustments of 60-mg/dl increments in glycemia per 1 unit lispro/ aspart.

An algorithm is a supplement or subtraction from the usual mealtime insulin dosage that compensates for the premeal glycemia level. The dosing range is calculated by the rule of 1800 for lispro or aspart insulins and 1500 for Regular insulin: divide the total 24-hour insulin dose by 1500 or 1800 to get the glucose-lowering effect of 1 unit of that insulin. This algorithm was for an individual taking 30 units of insulin daily—thus, adjustments of 60-mg/dl increments in glycemia per 1 unit lispro/ aspart.

ft is used for Regular insulin). For example, if your patient takes 30 units of insulin each day, each unit of lispro/aspart will lower glycemia an average of 60 mg/ dl. The scale in Table 5 is constructed with 1 -unit adjustments of lispro/aspart for every 60-mg/dl increase or decrease from the ideal glycemia range. Key to this practice is appreciating that these are adjuncts to the patient's usual doses. Thus, my own practice introduces this insulin adjustment only after we have completed what has already been discussed: validation of the program and basic doses, pattern management for fine tuning, and introduction of algorithms for lifestyle variations.

In contrast to the above approach, "sliding scales" are not recommended although they are frequently used in primary-care settings. These are scales that patients use to determine their mealtime insulin dosage based only on their blood glucose level; their usual mealtime insulin need, what they are going to eat, recent activity, and other relevant factors are not considered. Further, sliding scales are rarely individualized or validated for a particular patient, so overdosing or underdosing is common. Thus, in most patients the net result is that sliding scales promote erratic blood glucose control and/or an inability to get the target HbA,c without unpredictable hypoglycemia (27,28). For all these reasons, sliding scales are viewed by diabetes specialists as being generally detrimental to the patient and not recommended.