Vous êtes sur la page 1sur 13

Insulin Therapy: A Personal Approach

Mayer B. Davidson

f e at u r e a r t i c l e
In Brief Insulin therapy is challenging for providers as well as for patients.
This article describes a set of principles underlying appropriate insulin
treatment and a detailed discussion of how to use them.

T
he most recent information marize that information and provide
from 2014 indicates that 29.1 a more detailed (personal) approach
million Americans have diabe- to starting insulin, intensifying insu-
tes (1). Of these, 2530% are taking lin regimens when necessary, and
insulin (2,3). Primary care providers adjusting insulin doses.
are reluctant to start patients on insu-
Principles of Adjusting Insulin
lin (4). Numerous studies have shown
that, typically, 37 years elapse be- Doses
tween failure of oral antidiabetic med- Table 1 summarizes the most im-
ications and insulin initiation (58). portant relationships among various
A1C levels at the time of insulin ini- components of the insulin prescrip-
tiation in these studies ranged from tion, their injection times, and the
8.9 to 9.7% (5,9). Furthermore, once most appropriate self-monitoring of
insulin is started, many primary care blood glucose (SMBG) test to judge
providers are uncomfortable adjusting their effects. Preprandial testing is
doses and further intensifying insulin preferred because it is much easier
regimens. Mean A1C levels in patients for patients to remember to test when
receiving insulin have been found to they are stopping for a meal (and usu-
range from 7.9 to 9.3% (911). ally an insulin injection) rather than
In a recent article in Clinical when they are busy 12 hours after
Diabetes (12), Galdo et al. nicely eating. If preprandial glucose values
described the available insulin prepa- are at target but A1C levels remain
rations and their pharmacokinetic too high, postprandial SMBG is in-
and pharmacodynamic properties. dicated to determine whether high
They summarized the guidelines for glucose levels after meals may explain
using insulin published by several the above-target A1C.
diabetes organizations and discussed Each component of the insu-
various articles that described (to lin prescription is changed (or not)
Charles R. Drew University, Los Angeles, CA some extent) how to start insulin depending on the pattern of SMBG
Corresponding author: Mayer B. Davidson, and adjust doses. They concluded that results that reflects its maximal activ-
mayerdavidson@cdrewu.edu available instructions regarding how ity. How should this be done? First,
DOI: 10.2337/diaclin.33.3.123 to accomplish this were not very spe- one must decide on the appropriate
cific. Although there is no one right target ranges for each of these times.
2015 by the American Diabetes Association.
Readers may use this article as long as the work
way to use insulin, there are some The American Diabetes Association
is properly cited, the use is educational and not important principles and relation- (ADA) in 2015 changed its recom-
for profit, and the work is not altered. See http://
creativecommons.org/licenses/by-nc-nd/3.0
ships that underlie the appropriate mended preprandial target range of
for details. use of insulin. This article will sum- 70130 mg/dL to 80130 mg/dL

V OLU M E 3 3 , NU M BER 3 , s u mm e r 2 0 1 5 123


F e at u r e A r t i c l e

TABLE 1. Relationships Among Components of the Insulin Prescription, Injection Times, and Blood
Glucose Testing
Insulin Preparation Time Injected Period of Activity Preprandial Test Best Reflecting
Insulin Effect
Regular Before a meal Between that meal and Before next meal or bedtime
either the next one or (snack) if injected before supper
Lispro
bedtime (snack) if injected
Aspart before supper
Glulisine
NPH Before breakfast Between lunch and supper Before supper
U-500 Regular
NPH Before supper or bedtime Overnight Before breakfast
U-500 Regular
Glargine Any time Peakless: 24 hours for Before breakfast
glargine, ~18 hours for
Detemir
detemir

to better reflect new data comparing cohort of the current definition of based on a smaller number of glucose
actual average glucose levels with A1C overweight (25.129.9 kg/m2). results per time interval, perhaps as
targets (13). However, using this pre- Glucose concentrations at a spe- few as one-third.
prandial target range defines glucose cific time are considered too high if Conversely, glucose concentrations
values of 7079 mg/dL as hypoglyce- the number of values that exceed the at a specific time are considered too
mia, which contradicts the previous upper target level minus the number low if the number of values that are
ADA definition of hypoglycemia as of values that are less than the lower less than the lower end of the tar-
<70 mg/dL. Therefore, I continue to target level plus bona fide episodes get range, plus bona fide episodes of
use the older preprandial target range of unexplained hypoglycemia for unexplained hypoglycemia for which
of 70130 mg/dL. The ADA kept its which no measured low glucose level no measured low glucose value is
previous postprandial target of <180 is available constitute 50% or more available, minus the number of val-
mg/dL 12 hours after eating (13). I of the glucose concentrations at that ues that exceed the upper end of the
feel that their postprandial target is time of day during the period being target range constitute 50% of the
too high; instead, I use a postprandial considered. As examples, that would glucose concentrations at that time of
target of 100160 mg/dL. mean 2 such values over 3 days, day during a 1- to 6-week period. If
The next step is to decide under 4/7 days, 5/10 days, 7/14 days, the glucose concentrations at a specific
what circumstances changes in insu- 11/21 days, 14/28 days, 18/35 time of day are neither too high nor
lin doses should be made. If a test at days, 21/42 days, or 50% of the too low, no change is needed in the
a specific time of day (usually before a values during any number of days component of the insulin prescription
meal or bedtime snack) is consistently between 1 and 6 weeks. I prefer to that primarily affects the test at that
too high or too low, raise or lower the analyze SMBG results at least every time of day. Of course, no adjustment
appropriate insulin dose by 2 units 6 weeks once insulin doses are rea- is made in that component of the insu-
in lean patients or 4 units in over- sonably stable because, over longer lin regimen for which there are too few
weight or obese patients, or 10% of periods, patients (at least those in my available test results (i.e., the patient
the current dose, whichever is higher. practice) often become less adher- did not perform SMBG at that time
For insulin dose adjustments, I con- ent to the recommended insulin on enough days during the period
sider people to be overweight if they doses and SMBG testing frequency. under review). On the other hand,
are 120% of their desirable body However, if longer periods supervene, episodes of unexplained severe hypo-
weight (DBW). I chose this value the principle that 50% of the glu- glycemia warrant serious consideration
because mortality starts to increase cose concentrations at that time of day to reduce the dose of the appropriate
at weights exceeding this level. The during the time period being consid- component of the insulin regimen,
calculation of DBW (14) is described ered should be measured before an especially if the patient has hypogly-
in Table 2. Alternatively, one could adjustment of an insulin dose would cemia unawareness.
use the BMI for the designation of still hold. On the other hand, 50% The desired frequency of SMBG
overweight. I would suggest using the may be too stringent a criterion for testing for insulin dose adjustments
older BMI definition of obesity (27 some, and individual providers may is an important issue. Patients taking
kg/m2) because it captures the upper be comfortable with making decisions only bedtime NPH insulin or only

124 clinical.diabetesjournals.org
d av i d s o n

Table 2. Calculating Desirable Body Weight (DBW)


Weight for Females (lb) Weight for Males (lb)
Initial Calculation
First 5 feet 100 106
Each inch over 5 feet 5 6
Subsequent Adjustment for Frame Add 10% to DBW for large-framed individuals or subtract 10% for small-
Size framed individuals. Frame size can be estimated by having the patients
predominant hand grasp the other wrist and oppose the thumb and middle
finger. If the two fingers meet, the patient has a medium frame (and thus no
DBW adjustment is necessary). If they overlap appreciably, the patient is small
framed, and if they fail to meet, the patient is large framed.
Interpretation Patients who are 120% of their DBW are considered overweight/obese,
whereas those <120% of their DBW are considered lean.

f e at u r e a r t i c l e
a basal insulin need only test in the In my experience, it is unusual for ing some simple carbohydrate was
fasting state (i.e., before breakfast). patients taking a bedtime snack to ingested for treatment), this was prob-
Once two or more insulin injections experience overnight hypoglycemia. ably not hypoglycemia, but instead
are required, testing before each meal They are instructed to decrease their possibly anxiety, the symptoms of
and before the bedtime snack would supper calories to accommodate the which can mimic hypoglycemia. If
be ideal. However, it is usually unre- bedtime calories. Even if they should hypoglycemia is either documented
alistic to expect this frequency of gain a few pounds, the additional by SMBG or the episode seems bona
testing for most patients. I strongly cardiovascular risk in these obese fide by history, one must next deter-
recommend that patients perform individuals is minimal, especially mine whether it was explained or
SMBG at least twice daily, alter- compared to avoiding a potentially unexplained.
nating before breakfast and before serious episode of overnight hypo- The principle behind prescribing
supper with before lunch and before glycemia. For example, the increase an insulin regimen is to formulate it
the bedtime snack. in the risk of cardiovascular disease in around the patients usual practiced
I even more strongly recommend a 250-lb patient who may gain 10 lb lifestyle. Prescribing an insulin reg-
that all patients taking insulin eat a more is minimal. On the other hand, imen and expecting the patient to
small bedtime snack containing some patients experiencing hypoglycemia, change his or her lifestyle to accom-
protein to decrease the chances of especially overnight, may be reluc- modate it is seldom successful. This
overnight hypoglycemia. The key to tant to increase insulin doses when is not to say that one should not
necessary. A few discontinue insulin encourage patients to adopt a health-
effective insulin dosing is a fairly con-
altogether, and convincing them to ier lifestyle, but rather that one has
sistent eating pattern around which
re-start it may be difficult. to base the insulin regimen on what
to fit the insulin regimen. If a patient
Hypoglycemia is another important a patient is actually doing and not
always foregoes a bedtime snack, one factor affecting insulin dose adjust- what one hopes he or she will do.
might not require the patient to eat ments. Frankly, few patients perform From the patients perspective, this
one. However, for a patient in the SMBG when they are experiencing requires relatively consistent eating
habit of eating bedtime snacks inter- hypoglycemia; treating it typically and exercise patterns. (Carbohydrate
mittently, it would be impossible to overrides testing. Therefore, deciding counting is an exception and will
safely adjust the basal or evening whether the episode described was be discussed later.) If hypoglycemia
NPH insulin doses without know- bona fide hypoglycemia is important occurs because the patient has altered
ing whether there would be a snack for making appropriate insulin dose the usual eating or exercise pattern
on any given evening. Furthermore, adjustments. Readers of this journal (most commonly by missing a meal,
the longer the patient refrains from will certainly be familiar with the eating a meal later than usual, or
eating, the more likely (nocturnal) symptoms of hypoglycemia, but, eating less carbohydrate than usual),
hypoglycemia is to occur. Therefore, as is well known, a number of the this would be explained hypoglyce-
I recommend eating a bedtime snack symptoms are nonspecific. The key mia and therefore should be ignored
every night to maximize the chance question to ask is what the patient in the analysis. (Of course, one
of appropriate basal and evening did to treat the episode and how long would point this out and educate the
NPH insulin dosing and to min- it took for the symptoms to start to patient to avoid a similar episode in
imize the possibility of nocturnal improve. If >20 minutes or so passed the future.) On the other hand, if
hypoglycemia. before improvement started (assum- hypoglycemia occurs in the presence

V OLU M E 3 3 , NU M BER 3 , s u mm e r 2 0 1 5 125


F e at u r e a r t i c l e

of the patients usual lifestyle routine,


this would be unexplained hypogly-
cemia and, if it occurs frequently
enough, one must consider lowering
the appropriate insulin dose accord-
ingly. Often, patients perform SMBG
after starting to treat an episode of
hypoglycemia to determine if more
carbohydrate is needed. These results
(often high because of overtreatment)
should be ignored. Only bona fide
episodes of hypoglycemia need to be
considered as low glucose values and
entered into the analysis as such.
When and How to Initiate
Insulin Therapy
Patients with type 1 diabetes, of
course, need insulin immediately af- fIGuRe 1. One- to two-hour plasma glucose concentrations over a 24-hour
ter diagnosis. An exception is patients period in patients taking maximal doses of either glyburide (20 mg/day) or glipizide
with latent autoimmune diabetes of (40 mg/day) (upper curve) or bedtime NPH insulin (lower curve). Reprinted with
the adult (LADA), whose glucose permission from Ref. 40.
can be controlled on noninsulin
medications for a while, although for insulin may have been unnecessarily injected before breakfast or a short-
a much shorter period than for pa- initiated) will have been treated with or rapid-acting insulin is given before
tients with type 2 diabetes. Because a combination of two to four non- meals. Bedtime NPH or basal insulin
of the difficulties of using insulin insulin medications before insulin gives patients much more flexibility
(from both the patient and provider is started. Several older studies have in their eating and exercise patterns
perspectives), as long as patients with shown that, when patients blood during the day. Nocturnal hypoglyce-
LADA meet A1C targets, I do not use glucose fails to be controlled with mia occurs somewhat less frequently
insulin. As soon as noninsulin medi- maximal doses of metformin plus a with glargine and detemir compared
cations fail, I start insulin. Although sulfonylurea, multiple-dose insulin to NPH. However, I have had little
many guidelines recommend insu- regimens do not yield better control difficulty with overnight hypogly-
lin in type 2 diabetes patients with than bedtime NPH insulin plus oral cemia from NPH when employing
high A1C levels, this is not always antidiabetic drugs when compared for gradual dose increases, especially
necessary. Almost all patients with up to 1 year (1720). Therefore, if the when patients eat a small bedtime
newly diagnosed type 2 diabetes will patients fasting plasma glucose (FPG) snack.
respond quickly to maximal doses of concentration can be lowered appre- Although the ADA A1C target
a sulfonylurea with reduction of the ciably, overall diabetes control is often is <7.0% for most patients, I do not
initial dose often necessary to avoid significantly improved (Figure 1). start insulin until the A1C exceeds
hypoglycemia (15,16). This avoids Bedtime NPH or basal insulin 7.4% because the risk of microvas-
the hassle of starting insulin, making is an easy way to introduce insu- cular complications is only slightly
frequent down-titrations, andnot lin therapy to patients with type 2 increased between the A1C levels
uncommonlydiscontinuing insulin diabetes that is not well controlled of 7.0 and 7.5% (21), and starting
in patients with newly diagnosed type with oral antidiabetic medications. insulin is a big change in a persons
2 diabetes. Because of the progressive Such a regimen involves only one lifestyle (e.g., taking injections, per-
loss of insulin secretion in type 2 di- daily insulin injection and, initially, forming more frequent SMBG, and
abetes, however, patients with a lon- requires only one SMBG test per being mindful of the risk of hypo-
ger duration of disease and very high day. Because the peak effect of NPH glycemia). For patients whose A1C
A1C levels do require insulin. insulin occurs around breakfast time target should be less stringent (13),
Bedtime nPH and Basal Insulin and glargine and detemir insulins I use an A1C level of 8.0% before
Regimens are peakless, there is a much lower introducing insulin.
Almost all people with type 2 dia- chance of daytime hypoglycemia Overweight and obese patients are
betes (with the exception of those mediated by exogenous insulin com- started on 16 units, and lean patients
who are newly diagnosed, in whom pared to a regimen in which NPH is are started on 10 units of bedtime

126 clinical.diabetesjournals.org
d av i d s o n

NPH insulin or one of the peakless twice daily. In that case, overweight whether the regimen of bedtime
basal insulins. With the exception or obese patients are instructed to NPH or basal insulin is sufficient to
of thiazolidinediones (TZDs), I increase each dose by 2 units and control the patients diabetes. Because
maintain patients on their nonin- lean patients by 1 unit. Self-titration of the delay of A1C levels to plateau
sulin medications to maximize the ceases when there have been no dose (22), one must wait 3 months to
chances that daytime glycemia will changes for 1 week. make this decision.
be controlled after the FPG target Self-titration instructions for Soon after reaching the FPG tar-
is reached. TZDs are discontinued patients are provided in English get range, there is reason to perform
because they will enhance the weight and Spanish on p. 128 and p. 129, SMBG a few times before supper.
gain seen with insulin, increase the respectively, of this issue of Clinical With FPG levels this much lower,
degree of fluid retention, and thereby Diabetes. Providers can copy these the maximal dose of the patients
increase the risk of heart failure. The instructions and simply circle the sulfonylurea may cause daytime
initial insulin doses are almost always appropriate section for each patient. hypoglycemia. Before-supper values
less than patients eventually require When should a regimen including <80 mg/dL may presage daytime
but do prevent overnight hypoglyce- bedtime NPH or basal insulin plus

f e at u r e a r t i c l e
hypoglycemia, and the sulfony-
mia early on, which could discourage daytime oral antidiabetic drugs be lurea dose should be halved. If this
patients from remaining on insu- judged ineffective and replaced with persists, the sulfonylurea should be
lin therapy. Doses are increased as a multiple-dose insulin regimen? discontinued. Halving and possi-
described above (i.e., if 50% of the This should not occur until target bly subsequently discontinuing the
SMBG values before breakfast exceed FPG concentrations are achieved and sulfonylurea dose should also occur
the target, the dose is increased by 4 A1C levels 3 months later are still too with bona fide episodes of unex-
units for overweight or obese patients high. A common error is to give up plained (undocumented) daytime
or 2 units for lean patients, or by before reaching the target FPG level. hypoglycemia, as well. Conversely,
10% of the current dose, whichever This step usually occurs in obese once FPG levels reach the target
is higher). Ideally, test values should patients for whom not enough insu- range, before-supper glucose values
be acted on frequently (every 37 lin is prescribed, and the FPG levels that consistently exceed 180 mg/dL
days) when insulin is started; as val- hover around the mid- to high-100 strongly suggest that the target A1C
ues approach target levels, longer mg/dL range. Many of these patients level of <7.5% will not be attained 3
intervals between adjustments are are extremely obese with resulting months hence, and intensification of
reasonable. Ideally, I prefer adjusting severe insulin resistance. They often the insulin regimen could be consid-
the dose at least once per week for require more insulin in a dose than ered at this point.
several weeks after starting insulin, can be given in a single syringe (100
every 2 weeks as the dose is titrated units) or pen injection of glargine (80 Multiple-Injection Insulin
upward, every 3 weeks as the dose units) or detemir (60 units). I have Regimens
becomes stabilized, and every 6 weeks sometimes prescribed >100 units of As previously mentioned, I use an
on an ongoing basis. Of course, ones insulin at bedtime to achieve the A1C level 7.5% to signal the need
practice situation will largely deter- target FPG range. However, I usu- to switch to two or more injections
mine the frequency of adjustments. ally prescribe basal insulin twice of insulin because multiple-dose in-
Many patients with type 2 diabe- daily when the dose exceeds a sin- sulin regimens, when followed ap-
tes are very obese and consequently gle injection because large volumes propriately, are much more difficult
require large doses of insulin to reach of injectate impair absorption. (See for patients in that they require more
their target FPG levels. Because this discussion of U-500 insulin below.) frequent SMBG; increase the risk
can take many months to achieve, Overnight hypoglycemia is seldom a of hypoglycemia, especially during
I attempt to teach these patients to problem because, as previously men- the daytime; and, depending on the
self-titrate their insulin dose based tioned, the insulin doses are increased regimen implemented, may reduce
on their FPG levels. Overweight gradually; NPH peaks around break- flexibility with regard to eating and
or obese patients are instructed to fast time (instead of in the middle of exercise patterns. There are three
increase their bedtime dose by 2 units the night when it is injected before possible intensified insulin regimens
and lean patients by 1 unit each eve- supper); glargine and detemir are that could be used: basal/bolus, self-
ning if that mornings SMBG value peakless; and patients are strongly mixed/split, and premixed insulin. I
exceeded 130 mg/dL. Conversely, if encouraged to eat a small bedtime give patients the choice of the basal/
the value was <70 mg/dL, the dose snack. bolus or the self-mixed split regimen,
would be decreased by the same Once >50% of the fasting SMBG explaining the pros and cons of each.
amount. Patients on high doses of a values are within the FPG target I do not recommend premixed insu-
basal insulin sometimes inject these range, the A1C level determines lins for reasons explained below.

V OLU M E 3 3 , NU M BER 3 , s u mm e r 2 0 1 5 127


Insulin Self-Titration Instructions for Patients
Instructions for overweight or obese patients
Test blood glucose daily before breakfast.
For those using NPH insulin taken before supper or bedtime or glargine or detemir taken before
bedtime:
Every morning the result is >130 mg/dL, increase the insulin dose by 2 units that evening.
(If you are taking NPH insulin before breakfast, do not change that dose.)
Every morning the result is <70 mg/dL, decrease the insulin dose by 2 units that evening.
(If you are taking NPH insulin before breakfast, do not change that dose.)
If there have been no changes in the insulin dose for 1 week, do not change the dose anymore.
For those using glargine or detemir taken before breakfast:
Every morning the result is >130 mg/dL, increase the insulin dose that morning by 2 units.
Every morning the result is <70 mg/dL, decrease the insulin dose that morning by 2 units.
If there have been no changes in the insulin dose for 1 week, do not change the dose anymore.
For those using glargine or detemir taken both before breakfast and at bedtime:
Every morning the result is >130 mg/dL, increase both the morning and bedtime insulin doses
by 2 units each.
Every morning the result is <70 mg/dL, decrease both the morning and bedtime insulin doses
by 2 units each.
If there have been no changes in the insulin doses for 1 week, do not change them anymore.
Instructions for lean patients
Test blood glucose daily before breakfast.
For those using NPH insulin taken before supper or bedtime or glargine or detemir taken before
bedtime:
Every morning the result is >130 mg/dL, increase the insulin dose by 1 unit that evening.
(If you are taking NPH insulin before breakfast, do not change that dose.)
Every morning the result is <70 mg/dL, decrease the insulin dose by 1 unit that evening.
(If you are taking NPH insulin before breakfast, do not change that dose.)
If there have been no changes in the insulin dose for 1 week, do not change it anymore.
For those using glargine or detemir taken before breakfast:
Every morning the result is >130 mg/dL, increase the insulin dose that morning by 1 unit.
Every morning the result is <70 mg/dL, decrease the insulin dose that morning by 1 unit.
If there have been no changes in the insulin dose for 1 week, do not change it anymore.
For those using glargine or detemir taken both before breakfast and at bedtime:
Every morning the result is >130 mg/dL, increase both the morning and bedtime insulin doses
by 1 unit each.
Every morning the result is <70 mg/dL, decrease both the morning and bedtime insulin doses
by 1 unit each.
If there have been no changes in the insulin doses for 1 week, do not change them anymore.

Permission is granted to reproduce this material for nonprofit education purposes. Written permission
is required for all other purposes. Please send requests to permissions@diabetes.org, referencing
Clinical Diabetes, Vol. 33, issue 3.
Instrucciones para pacientes sobre la auto-titulacin de insulina
Instrucciones para los pacientes con sobrepeso u obesidad
Hacerse la prueba de glucosa en la sangre todos los das antes del desayuno.
Para aquellos que utilizan insulina NPH tomada antes de la cena o antes de acostarse o glargina o
detemir se toma antes de acostarse:
Cada maana que el resultado es >130 mg/dL, aumentar la dosis de insulina por 2 unidades en
esa tarde. (Si usted est tomando insulina NPH antes del desayuno, no cambie esa dosis.)
Cada maana que el resultado es <70 mg/dL, disminuir la dosis de insulina por 2 unidades en
esa tarde. (Si usted est tomando insulina NPH antes del desayuno, no cambie esa dosis.)
Si no ha habido cambios en la dosis de insulina durante 1 semana, no cambie la dosis ms.
Para aquellos que utilizan glargina o detemir tomado antes del desayuno:
Cada maana que el resultado es >130 mg/dL, aumentar la dosis de insulina en esa maana por
2 unidad.
Cada maana que el resultado es <70 mg/dL, disminuir la dosis de insulina en esa maana por 2
unidad.
Si no ha habido cambios en la dosis de insulina durante 1 semana, no cambie la dosis ms.
Para aquellos que utilizan glargina o detemir tomado antes del desayuno y antes de dormir:
Cada maana que el resultado es >130 mg/dL, aumentar tanto por la maana y antes de dormir
las dosis de insulina por 2 unidades cada uno.
Cada maana que el resultado es <70 mg/dL, disminuir tanto por la maana y antes de dormir
las dosis de insulina por 2 unidades cada uno.
Si no ha habido cambios en las dosis de insulina durante 1 semana, no cambie las dosis ms.
Instrucciones para pacientes delgados
Hacerse la prueba de glucosa en la sangre todos los das antes del desayuno.
Para aquellos que utilizan insulina NPH tomada antes de la cena o antes de dormir o glargina o
detemir tomada antes de dormir:
Cada maana que el resultado es >130 mg/dL, aumentar la dosis de insulina por 1 unidad en esa
tarde. (Si usted est tomando insulina NPH antes del desayuno, no cambie esa dosis.)
Cada maana que el resultado es <70 mg/dL, disminuir la dosis de insulina por 1 unidad en esa
tarde. (Si usted est tomando insulina NPH antes del desayuno, no cambie esa dosis.)
Si no ha habido cambios en la dosis de insulina durante 1 semana, no cambie la dosis ms.
Para aquellos que utilizan glargina o detemir tomado antes del desayuno:
Cada maana que el resultado es >130 mg/dL, aumentar la dosis de insulina en esa maana por
1 unidad.
Cada maana que el resultado es <70 mg/dL, disminuir la dosis de insulina en esa maana por 1
unidad.
Si no ha habido cambios en la dosis de insulina durante 1 semana, no cambie la dosis ms.
Para aquellos que utilizan glargina o detemir tomado tanto antes del desayuno y antes de
acostarse:
Cada maana que el resultado es >130 mg/dL, aumentar tanto por la maana y antes de dormir
las dosis de insulina por 1 unidad de cada uno.
Cada maana que el resultado es <70 mg/dL, disminuir tanto por la maana y antes de dormir
las dosis de insulina por 1 unidad de cada uno.
Si no ha habido cambios en las dosis de insulina durante 1 semana, no cambie la dosis ms.
Permission is granted to reproduce this material for nonprofit education purposes. Written permission
is required for all other purposes. Please send requests to permissions@diabetes.org, referencing
Clinical Diabetes, Vol. 33, issue 3.
F e at u r e A r t i c l e

Basal/Bolus Insulin Regimen reflect overall glycemia. This period Self-Mixed/Split Insulin
The basal/bolus insulin regimen usu- will be doubled if an injection before Regimen
ally consists of a basal insulin plus a a second meal is required, and tri- In the self-mixed/split insulin reg-
short- or rapid-acting insulin before pled if injections before all three imen, the insulin preparations are
meals. In patients with type 2 diabe- meals are deemed necessary. Because mixed in the same syringe and inject-
tes, NPH insulin at bedtime would only 2530% of patients who start ed twice daily. If used properly, this
also be appropriate because these pa- by adding a single preprandial dose approach can yield as tight a level
tients retain some insulin secretion. of rapid-acting insulin to optimized of control as a basal/bolus regimen
NPH insulin at bedtime would not be basal insulin achieve an A1C <7.0% (31,32). However, patients have less
appropriate for patients with type 1 (25,26), there will be long delays in flexibility with their eating and exer-
diabetes because, if there is a long peri- reaching A1C goals for most patients. cise patterns. To switch from a bed-
od of time between lunch and supper, Indeed, a recent study showed that time NPH or basal insulin regimen,
the effect of the pre-lunch dose might this can take 8 months (27). I take 80% of that dose as the total
wane, and patients would experience The second potential problem NPH dose of the self-mixed split regi-
high pre-supper glucose because they with regard to adding prandial men and give two-thirds before break-
have little or no endogenous insu- insulin one meal at a time is that fast and one-third before supper. This
lin secretion. This potential prob- the most important determinant of is most often less than the patient will
lem is much more likely if a rapid- postprandial glucose (and therefore eventually require but, again, avoids
acting insulin is the formulation used subsequent preprandial glucose) is hypoglycemia, which can be discour-
before lunch. the preprandial value. The increases aging to patients. Because almost all
The basal/bolus approach requires in postprandial over preprandial glu- patients will require prandial insulin
four daily injections of insulin, which cose values are similar regardless of to achieve their A1C target, I routine-
is a problem for some patients. On the starting preprandial glucose level ly add a small amount of short- or
the other hand, it should be strongly (2830). Therefore, postprandial rapid-acting insulin to each NPH
considered for patients with irregular hyperglycemia is best treated initially injection after several visits. I delay
eating and exercise patterns. by lowering preprandial glucose lev- adding the short- or rapid-acting in-
Some investigators have recom- els. When preprandial insulin before sulin to allow patients to more grad-
mended that, initially upon switching a single meal has controlled postpran- ually accommodate themselves to the
to a basal/bolus regimen, short- or dial glucose concentrations but A1C lifestyle changes that taking insulin
rapid-acting insulin is only needed is still not within the target range, a entails. Initially, patients are asked to
before the largest meal of the day second preprandial injection must be perform SMBG before breakfast and
(23,24). If the subsequent preprandial introduced. However, when the sec- supper to adjust their NPH doses.
(or before bedtime snack, if the larg- ond injection lowers the preprandial When these results are <200 mg/dL,
est meal is supper) target is reached values before the initial meal that patients are asked to alternate their
and the A1C remains above target, had just been successfully treated, twice-daily testing to before breakfast
short- or rapid-acting insulin is intro- the short- or rapid-acting insulin and supper one day and before lunch
duced before the next largest meal. dose given before that first meal may and bedtime snack the next day. The
This is repeated if the second prepran- be too high, leading to postprandial latter two tests allow for adjusting the
dial injection of short- or rapid-acting hypoglycemia. The same potential short- or rapid-acting insulin doses.
insulin achieves the subsequent pre- problem occurs when a third pre- The pre-supper and pre-breakfast
prandial target but the A1C target is prandial injection is introduced. glucose levels are brought down to
still not attained; at this point, short- For these reasons, I usually start <200 mg/dL with the appropriate
or rapid-acting insulin is taken before preprandial insulin before all three NPH doses before the short- or rapid-
all three meals. meals when instituting a basal/bolus acting insulin is increased. This is
However, there are two poten- regimen. Because the patients current because larger doses of the short- or
tial problems with this approach. bedtime NPH or basal insulin dose rapid-acting insulin will be needed to
Although logically appealing, it can has already achieved the FPG target, lower glucose levels to target before
lead to long delays in reaching target I continue it. The initial preprandial lunch and the bedtime snack if the
A1C levels. First, the target glucose insulin dose is 24 units in lean pre-breakfast and pre-supper glucose
level, either postprandially or before patients and 46 units in overweight levels, respectively, are higher than if
the subsequent meal (or bedtime or obese patients. These doses almost they are closer to target. For example,
snack when supper is the meal in always need to be increased, but start- reaching a pre-lunch glucose concen-
question), must be achieved. Then, at ing at this level avoids hypoglycemia, tration of <130 mg/dL will require
least another 3 months must elapse which may discourage patients from more short- or rapid-acting insulin
before the A1C level will accurately continuing their intensified regimen. when the FPG is 280 mg/dL than

130 clinical.diabetesjournals.org
d av i d s o n

when it is 120 mg/dL. Delaying the TABLE 3. Initial Correction Doses of Short-
increase of the short- or rapid-act- or Rapid-Acting Insulin
ing insulin until the doses of NPH
Blood Glucose Correction Dose for Lean Correction Dose for
have started to achieve better control (mg/dL) Patients* (units) Overweight or Obese
reduces the possibility of hypoglyce- Patients (units)
mic reactions. On the other hand, if
<70 1 2
one waits until target levels before
breakfast and supper are reached 70150 0 0
before adjusting the appropriate doses 151200 +1 +2
of prandial insulin, too much time 201250 +2 +4
often elapses before better overall
control is achieved. 251300 +3 +6
Given that the peak effect of NPH 301350 +4 +8
taken before supper occurs between >350 +5 +10

f e at u r e a r t i c l e
6 and 12 hours later, as one increases
*<120% of DBW, as calculated in Table 2.
this dose to control FPG, hypogly-
120% of DBW, as calculated in Table 2.
cemia may occur overnight before
the target is reached. Decreasing the
the pre-supper situation. Another mean a third daily injection, negating
pre-supper dose may prevent the over-
common pattern is high glucose at one of the advantages of the two-in-
night hypoglycemia, but will hinder
bedtime but acceptable or low lev- jection self-mixed/split regimen.
efforts to achieve the FPG target. If
els before breakfast. Raising the pre- The other dietary approach is a
the patient is already eating a bedtime
supper dose to lower the bedtime re- constant-carbohydrate eating pat-
snack, adding a snack is not an avail-
sults would increase the likelihood of tern, in which patients are instructed
able option to help prevent overnight overnight hypoglycemia. Therefore, about the carbohydrate content of
hypoglycemia. In that event, moving achieving near-euglycemia with pre- various food products and asked to
the NPH from before supper to bed- mixed insulins usually is not possi- ingest a relatively consistent amount
time will almost always take care of ble, and these preparations should be of carbohydrate at each meal (i.e.,
the problem because the peak effect used only by patients who cannot be a similar amount at every break-
of the evening NPH will now occur taught to mix insulins themselves and fast, lunch, and supper from day to
just before breakfast when the patient for whom no family or other caregiver daynot the same amount at every
is about to eat. This approach con- is available. In my experience, most meal of the day). There is no evi-
verts the two-injection self-mixed/ patients can be taught to mix insulin dence that carbohydrate counting
split regimen to a three-injection reg- with persistent instruction. leads to better glycemic control than
imen. However, it is often the only
Short- or Rapid-Acting Insulin the constant-carbohydrate approach
way to both avoid overnight hypo-
There are two ways to decide on the (34), which is similar to the older
glycemia and meet the FPG target
usual dose of short- or rapid-acting Exchange system, recently termed
without converting to a four-injection
insulin. More sophisticated patients experience-based estimation (35).
basal/bolus regimen.
can learn carbohydrate counting With experience-based esti-
Premixed Insulin Regimens and then determine the preprandial mation, the preprandial short- or
Although premixed insulin prepa- doses of these insulins based on the rapid-acting insulin dose (in either
rations obviate the need for pa- amount of carbohydrate to be ingest- a basal/bolus or a self-mixed/split
tients to mix two different insulin ed, usually 1 unit per 10 or 15 g of regimen) can be broken down into
preparations in the same syringe carbohydrate (33). In a self-mixed/ three components. The basic dose
before injection, they have a ma- split regimen, this would work for is the amount prescribed to be taken
jor drawback: one cannot adjust breakfast and supper but, in my view, before the meal and is the dose that
the doses of the intermediate- would not be suitable for lunch. This is adjusted based on the pattern of
acting and short- or rapid-acting is because, as mentioned previously, SMBG values, as described earlier.
insulin separately. For example, a both the morning NPH insulin and The correction dose or supplemen-
common SMBG pattern in patients the short- or rapid-acting insulin be- tal dose depends on the preprandial
taking a premixed insulin preparation fore lunch have their maximal effects glucose level. Some patients are also
is high glucose before lunch but ac- between lunch and supper. It would able to incorporate an anticipatory
ceptable or low glucose before supper. not be clear which of these insulin dose that depends on anticipated
Raising the morning dose to lower doses to adjust in response to the pre- activities to occur with the meal or
the pre-lunch levels would jeopardize supper SMBG values. It would also within several hours after the meal.

V OLU M E 3 3 , NU M BER 3 , s u mm e r 2 0 1 5 131


F e at u r e A r t i c l e

Start U-500 regular insulin (RI) preprandial glucose range of 201300


if: mg/dL. If subsequent experience
1) A1C >8.0% and
2) TDD >200 units/day shows that this correction dose is in-
adequate for the upper part of that
If TDD 200300 units/day If TDD is >300 units/day
U-500 RI 100 units in AM U-500 RI 100 units in AM range (251300 mg/dL), the correc-
+ 50 RI units in PM + 100 RI units in PM tion dose should be increased to +4
units for that range of preprandial
values. To simplify the evaluation, it
is helpful if patients record their usual
F/U every 2 weeks

prandial insulin dose and their correc-
tion dose separately, e.g., 3 + 2, before
the meal in their logbooks.
If >50% of BG 70130 mg/dL, If >50% of BG >130200 mg/dL, If >50% of BG >200 mg/dL, U-500 Regular Insulin
no change in dose increase by 25 units increase by 50 units It is not uncommon for very obese
(If any unexplained (If any unexplained hypoglycemia, (If any unexplained
hypoglycemia, decrease by 25 no change in dose) hypoglycemia, increase by only patients with type 2 diabetes to re-
units) 25 units) quire hundreds of units of insulin to
achieve satisfactory control. Some,
No change in insulin dose despite testing appropriately and
No unexplained
hypoglycemia increasing insulin doses as recom-
mended, are unable to achieve A1C

levels <8.0% because large volumes
F/U monthly of injectate impair insulin absorp-
tion (36). In my practice, patients re-
FIGURE 2. Initiation and dose adjustments of U-500 regular insulin. BG, blood quiring >200 units of U-100 insulin
glucose; F/U, follow-up; RI, regular insulin; TDD, total daily dose. Reprinted with per day are switched to injections of
permission from Ref. 41. U-500 regular insulin before breakfast
and before supper. The action profile
For example, if a patient is eating at start with to add to (or subtract from) of U-500 regular insulin is similar to
a Chinese restaurant, he or she may the basic dose to correct elevated pre- that of NPH (37).
1 Initial doses and dose adjust-
add a few units (in addition to any prandial glucose concentrations. To
correction dose) in anticipation of a evaluate the responses to these correc- ments for U-500 insulin are shown in
meal that is higher than usual in car- tions doses, I use a target range for the Figure 2. If A1C levels remain 7.5%
bohydrate. Alternatively, if a patient subsequent preprandial SMBG value after the pre-breakfast and pre-
is going to engage in more exercise of 100150 mg/dL. If the majority supper SMBG values reach target
than usual after supper, he or she may of responses to a specific correction levels, separate injections of short- or
dose are >150 mg/dL, the correction rapid-acting insulin are given along
reduce the short- or rapid-acting insu-
dose needs to be increased; if the ma- with the U-500 insulin. I usually
lin dose taken before the meal that
jority are <100 mg/dL (plus episodes start with 6 units, and, inexplicably,
would otherwise have been dictated
of undocumented, unexplained hypo- pre-lunch and bedtime SMBG values
by the calculations for basic and cor- show that these patients respond to
rection doses. There is no formula glycemia), the correction dose should
be decreased. A minimum of three re- lower doses of these U-100 insulins
for anticipatory doses. They must be (1030 units). A1C levels often can
arrived at empirically based on the sponses to a specific correction dose is
necessary before it can be evaluated. be lowered to <7.5% and routinely to
patients ongoing experiences. <8.0% with U-500 regular insulin
For example, if there are five instances
Correction (Supplemental) in which a patient added 2 units of (36).
Doses regular insulin for preprandial SMBG Single Pre-Breakfast Injection
Assuming no change in the carbo- values between 201 and 250 mg/dL of NPH Insulin
hydrate content of the anticipated and the subsequent preprandial results If a single morning injection of NPH
meal, high preprandial glucose con- were >150 mg/dL on four of those is used, near-euglycemia is seldom
centrations require additional short- occasions, the correction dose for achieved. This approach requires one
or rapid-acting insulin to lower the this preprandial range should be in- injection of NPH to control both
subsequent preprandial SMBG value creased to +3 units. Moving forward, the pre-supper glucose concentra-
into the target range. Table 3 shows +3 units of short- or rapid-acting tion and the next mornings fasting
the extra amounts of these insulins I insulin would be added for the wider glucose value. In the typical scenario,

132 clinical.diabetesjournals.org
d av i d s o n

TABLE 4. Doses for Initial Intensified Insulin Therapy (Twice-Daily or Multiple-Injection Regimens)
Pre-Breakfast Pre-Lunch Pre-Supper Bedtime
Lean patients
Twice-daily injection 10 units NPH/ 6 units NPH/
regimen 24 units short- or 24 short- or

rapid-acting insulin rapid-acting insulin

Multiple-injection 4 units short- or 4 units short- or 4 units short- or 10 units NPH,


regimen rapid-acting insulin rapid-acting insulin rapid-acting insulin glargine, or detemir
insulin
Overweight or obese patients
Twice-daily injection 20 units NPH/ 10 units NPH/
regimen 46 units short- or 46 units short- or

rapid-acting insulin rapid-acting insulin

f e at u r e a r t i c l e
Multiple-injection 68 units short- or 68 units short- or 68 units short- or 16 units NPH,
regimen rapid-acting insulin rapid-acting insulin rapid-acting insulin glargine, or detemir
insulin

the morning NPH dose is increased, than it influences the pre-supper glu- ily) be treated initially with insulin,
and pre-supper glucose levels become cose value (38). In this very unusual and the usual subsequent scenario
acceptable before fasting values do. situation, a single morning injec- then involves decreasing the insulin
As the NPH dose is increased fur- tion of NPH insulin is appropriate, doses and often safely discontinuing
ther to lower the fasting values, late- with short- or rapid-acting insulin insulin in favor of oral medications.
afternoon hypoglycemia occurs, and given preprandially as necessary. Patients with type 1 diabetes who
the NPH dose must be stabilized or Interestingly, the pharmacodynamic present with diabetic ketoacidosis
decreased before target fasting levels response to short- and rapid-acting (DKA) will be discharged from the
have been reached. At this point, an insulin remains normal. This sug- hospital on insulin. However, there
evening dose of NPH must be intro- gests that the reason for the delayed are some patients who may benefit
duced to further improve control. response involves a slowed release from an initial intensified insulin
This means that the patient is now of insulin from the protamine in regimen (e.g., patients with type 1
on a self-mixed/split regimen, which the NPH preparation rather than a diabetes diagnosed before DKA su-
should have been the initial approach. general delay in the egress of insulin pervenes, those with LADA who
There are two exceptions to this from the subcutaneous space into the are not controlled with noninsulin
typical scenario, however. A few circulation. medications, and those with type 2
patients with type 2 diabetes who Patients with a delayed response to diabetes of long duration who are
are not well controlled with non- NPH insulin are usually identified by poorly controlled on noninsulin
insulin medications have FPG values recognizing that the FPG concentra- medications). Initial doses for lean
at or very near the target range. Their tions remain low as the evening NPH and overweight or obese patients
elevated A1C levels are the result of insulin dose in a self-mixed/split reg- with regimens involving twice-daily
daytime hyperglycemia, most often imen continues to be reduced and or multiple injections are presented
manifested by high pre-supper SMBG remains normal or low even when in Table 4. As with other regimens,
results. In this case, pre-breakfast no evening NPH insulin is injected. these initial doses most often are set
NPH insulin is indicated, initially These patients can be challenging much lower than eventually will be
10 units in lean patients and 16 units to control. Fortunately, this delayed required to reduce the likelihood of
in those who are overweight or obese. response to NPH insulin does not hypoglycemia, which may discourage
Eventually, as insulin secretion con- occur very often. patients from continuing with insulin
tinues to decrease, evening NPH therapy.
insulin becomes necessary. Intensified Insulin Regimens as
A second rare exception to the Initial Therapy Conclusion
usual morning NPH scenario is a As discussed above, most patients Although there is no one right way
patient who has a delayed response to with type 2 diabetes will transition to to implement insulin therapy for pa-
NPH insulin (i.e., the peak effect of a insulin therapy via bedtime NPH or a tients with diabetes, this approach,
morning injection occurs overnight, basal insulin. A few newly diagnosed taught to more than 40 mid-level
such that the morning injection patients with markedly elevated glu- health care professionals (registered
affects the next days FPG more cose concentrations will (unnecessar- nurses, nurse practitioners, physi-

V OLU M E 3 3 , NU M BER 3 , s u mm e r 2 0 1 5 133


F e at u r e A r t i c l e

cian assistants, and clinical pharma- apy despite inadequate glycemic control. J over time in diabetic patients. Diabetes Care
Gen Intern Med 2007;22:453458 1993;16:13131314
cists), has been very effective. For
9. Harris SB, Kapor J, Lank CN, Willan A, 23. Raccah D, Bretzel RG, Owens D, Riddle
example, a registered nurse trained in Houston T. Clinical inertia in patients with M. When basal insulin therapy in type 2
this approach and placed in a family T2DM requiring insulin in family practice. diabetes mellitus is not enoughwhat next?
medicine clinic serving Latino and Can Fam Phys 2010;56:e418e424 Diabetes Metab Res Rev 2007;23:257264
African-American patients lowered 10. Ziemer DC, Miller CD, Rhee MK, et al. 24. Nathan DM, Buse JB, Davidson MB,
A1C levels in 111 insulin-requiring Clinical inertia contributes to poor diabetes et al.; American Diabetes Association;
control in a primary care setting. Diabetes
patients referred to her by the clin- Educ 2005;31:564571
European Association for the Study of
Diabetes. Medical management of hyper-
ic physicians, from an average 11.1 11. Chen Y, Abbott S, Nguyen M, Grabner glycemia in type 2 diabetes: a consensus
to 7.3% within 912 months (39). M, Dumbo R. Glycemic control of insulin algorithm for the initiation and adjustment
Currently, three mid-level profes- treated patients across the U.S.: epidemi- of therapy: a consensus statement of the
ologic analysis of a commercially insured
sionals supervised by the author treat population. Diabetes 2013;62(Suppl. 1):A704
American Diabetes Association and the
European Association for the Study of
~800 racial/ethnic minority patients Diabetes. Diabetes Care 2009;32:193203
12. Galdo JA, Thurston, MM, Bourg CA.
with diabetes, approximately half of Clinical considerations for insulin phar- 25. Owens DR, Luzio SD, Sert-Langeron C,
whom are taking insulin. The average macotherapy in ambulatory care, part Riddle MC. Effects of initiation and titra-
A1C level in these insulin-requiring one: introduction and review of current
tion of a single pre-prandial dose of insulin
products and guidelines. Clinical Diabetes
patients is in the mid-7% range. 2014;32:6675
glulisine while continuing titrated insulin
glargine in type 2 diabetes: a 6 month
13. American Diabetes Association. proof of concept study. Diabetes Obes
Glycemic targets. Sec. 6 in Standards of
Duality of Interest Medical Care in Diabetes2015. Diabetes
Metab 2011;13:10201027
M.B.D. serves on an advisory board for Care 2015;38(Suppl. 1):S33S40 26. Davidson MB, Raskin P, Tanenberg
Sanofi. No other potential conflicts of inter- RJ, Vlajnic A, Hollander P. A stepwise
14. Hamwi GJ. Therapy: changing dietary approach to insulin therapy in patients with
est relevant to this article were reported.
concepts. In Diabetes Mellitus: Diagnosis
type 2 diabetes and basal insulin treatment
and Treatment. Vol. 1. Danowski TS, Ed.
References New York, American Diabetes Association,
failure. Endocr Pract 2011;17:395403
1. Centers for Disease Control and 1964, p. 7378 27. Rodbard H, Visco VE, Andersen H,
Prevention. National diabetes statistics 15. Peters AL, Davidson MB. Maximal Hiort LC, Shu DHW. Treatment intensifi-
report2014. Available from http://www. dose glyburide in markedly symptomatic cation with stepwise addition of prandial
cdc.gov/diabetes/pubs/statsreport14.htm. patients with type 2 diabetes: a new use insulin aspart boluses compared with full
Accessed 20 March 2015 for an old friend. J Clin Endocrinol Metab basal-bolus therapy (FullSTEP Study): a
2. Centers for Disease Control and 1996;81:24232427 randomized, treat-to-target clinical trial.
Prevention. Age-adjusted percentage of Lancet Diabetes Endocrinol 2014;2:3037
16. Babu A, Mehta A, Guerrero P, et al. Safe
adults with diabetes using diabetes medica- and simple emergency department discharge 28. Cusi K, Cunningham GR, Comstock JP.
tion, by type of medication, United States, therapy for patients with type 2 diabetes Safety and efficacy of normalizing fasting
19972011. Available from http://www.cdc. mellitus and severe hyperglycemia. Endocr glucose with bedtime insulin alone in
gov/diabetes/statistics/meduse/fig2.htm. Pract 2009;15:696704 NIDDM. Diabetes Care 1995;18:843851
Accessed 20 March 2015
3. Turner LW, Nortey D, Stafford RS, Singh 17. Yki-Jarvinen H, Kauppila M, Kujansuu 29. Monnier L, Colette C, Dunseath GJ,
S, Alexander GC. Ambulatory treatment E, et al. Comparison of insulin regimens Owens DR. The loss of postprandial glyce-
of type 2 diabetes in the U.S., 19972012. in patients with non-insulin-depen- mic control precedes stepwise deterioration
Diabetes Care 2014;37:985992 dent diabetes mellitus. N Engl J Med of fasting with worsening diabetes. Diabetes
1992;327:14261433 Care 2007;30:263269
4. Jeavons D, Hungin APS, Cornford CS.
Patients with poorly controlled diabetes 18. Wolffenbuttal BH, Sets JP, 30. Bonomo K, DeSalve A, Fiova E, et al.
in primary care: healthcare clinicians Rondas-Colbers GJ, Menheere PP,
Evaluation of a simple policy for pre- and
beliefs and attitudes. Postgrad Med J Nieuwenhuijzen-Kruseman AC.
post-prandial blood glucose self-monitoring
2006;82:347350 Comparison of different regimens in elderly
patients with NIDDM. Diabetes Care in people with type 2 diabetes not on insu-
5. Khunti K, Wolden MI, Thorsted BL, lin. Diabetes Res Clin Pract 2010;87:246251
1996;19:13261332
Andersen M, Davies MJ. Clinical inertia in
people with type 2 diabetes: a retrospective 19. Yki-Jarvinen H, Ryysy L, Kauppila M, 31. Reeves ML, Seigler DE, Ryan EA,
cohort study of more than 80.000 people. et al. Effect of obesity on the response to Skyler JS. Glycemic control in insulin-de-
Diabetes Care 2013;36:34113417 insulin therapy in noninsulin-dependent pendent diabetes mellitus: comparison of
diabetes mellitus. J Clin Endocrinol Metab outpatient intensified conventional therapy
6. Calvert MJ, McManis RJ, Freemantle N. with continuous subcutaneous insulin infu-
1997;82:40374043
Management of type 2 diabetes with multi- sion. Am J Med 1982;72:673680
ple oral hyperglycaemic agents or insulin in 20. Yki-Jarvinen H, Ryysy L, Nikkila
primary care retrospective cohort study. Br K, Tulokas T, Vanamo R, Heikkila M. 32. Umpierrez GE, Hor T, Smiley D, et al.
J Gen Pract 2007;57:455460 Comparison of bedtime insulin in patients Comparison of inpatient insulin regimens
with type 2 diabetes mellitus: a random- with detemir plus aspart versus neutral
7. Rubino A, McQuay LJ, Gough SC, Kvasz
ized, controlled trial. Ann Intern Med protamine Hagedorn plus regular in med-
V, Tennis P. Delayed initiation of subcuta-
1999;130:389396 ical patients with type 2 diabetes. J Clin
neous insulin therapy after failure of oral
glucose-lowering agents in patients with 21. Skyler JS. Diabetic complications: the Endocrinol Metab 2009;94:564569
type 2 diabetes: a population-based analysis importance of control. Endocrinol Metab 33. Kulkarni K. Carbohydrate counting for
in the UK. Diabet Med 2007;24:14121418 Clin North Am 1996;25:243254 pump therapy. In A Core Curriculum for
8. Nichols GA, Koo YH, Shah SN. Delay of 22. Tahara Y, Shima K. The response of Diabetes Education: Diabetes Management
insulin addition to oral combination ther- GHb to stepwise plasma glucose change Therapies. 5th ed. Franz MJ, Ed. Chicago,

134 clinical.diabetesjournals.org
d av i d s o n

Ill., American Association of Diabetes 36. Binder C. Absorption of injected insulin: Duran P. Integrating nurse-directed diabe-
Educators, 2003, p. 265276 a clinical-pharmacological study. Acta tes management into a primary care setting.
Pharmacol Toxicol (Copenh) 1969;27 (Suppl. Am J Manag Care 2010;16:652656
34. Bergenstal RM, Johnson M, Powers
2):1187 40. Cusi K, Cunningham GR, Comstock JP.
MA, et al. Adjust to target in type 2 diabe-
tes: comparison of a simple algorithm with 37. Davidson MB, Navar MD, Echeverry Safety and efficacy of normalizing fasting
D, Duran P. U-500 regular insulin: clinical glucose with bedtime NPH insulin alone in
carbohydrate counting for adjustment of
experience and pharmacokinetics in obese, NIDDM. Diabetes Care 1995;18;843851
mealtime insulin glulisine. Diabetes Care
2008;31:13051310 severely insulin-resistant type 2 diabetic 41. Ballani, P, Tran, MT, Navar MD,
patients. Diabetes Care 2010;33:281283 Davidson MB. Clinical experience with
35. American Diabetes Association. U-500 regular insulin in obese, markedly
Standards of medical care in diabe- 38. Davidson MB. Letter to the editor.
insulin resistant type 2 diabetic patients.
tes2014. Diabetes Care 2014;37 (Suppl. Diabetes Res Clin Pract 2004;64:229
Diabetes Care 2006;29:25042505; erratum
1):S14S80 39. Davidson MB, Blanco-Castellanos M, 2007;30:455

f e at u r e a r t i c l e
Living With
Type 2 Diabetes
Free resources for
patients newly
diagnosed

This free, 12-month patient program is available in English and Spanish.

MONTH 1 MONTH 3 MONTH 6 MONTH 9 MONTH 12


-------------------------------------------------------------- -------------------------------------------------------------- -------------------------------------------------------------- -------------------------------------------------------------- --------------------------------------------------------------
Food, Emotional Types of Prevention Resources
nutrition, health and physical of diabetes on support,
and healthy diabetes activity complications discrimination,
choices management and financial
burden

Visit diabetes.org/atdx
to help your patients get started today.

V OLU M E 3 3 , NU M BER 3 , s u mm e r 2 0 1 5 135