Diabetic ketoacidosis in pregnancy

Mary Anne Carroll, MD; Edward R. Yeomans, MD

F ortunately for obstetriciansand intensivists, the vast ma-jority of women who becomepregnant are both young and

healthy. Of the relatively small percent-age of pregnant women with pregesta-tional medical complications, an evensmaller group will experience what wewould call a medical emergency, that is, acondition that poses an immediate threatto the life of mother, fetus, or both. Dia-betic ketoacidosis (DKA) is such an en-tity. Once diagnosed, the first order writ-ten by the admitting physician should be“Admit to ICU.”

In our review of this important sub-ject, we will attempt to present usefulinformation to two groups of readers: a)physiologic changes unique to pregnancyfor the medical intensivist, because thesechanges affect management; and b) a“cookbook” algorithm of medical man-agement for the obstetrician who mayhave to provide emergency care beforeconsulting an intensivist. Again, fortu-nately for us, the incidence of DKA inpregnancy has declined substantivelyover the years, and so has the perinatalmortality (Table 1) (1–5). However, the

importance of providing prompt, effectivemedical care is unaltered or perhaps evenincreased, due to the unrealistic expecta-tion for a perfect outcome for all preg-nancies.

Adaptation to Pregnancy

A detailed account of physiologicchanges induced by pregnancy is pre-sented elsewhere in this supplement.Here we review only a few pertinent facts.

The acid-base state of pregnancy is acompensated respiratory alkalosis. Nor-mal pH is 7.43, PCO2 is 30 mm Hg, andbicarbonate is 19–20 mEq/L. The drop inbicarbonate occurs to compensate for theprimary respiratory alkalosis. This reduc-tion in buffering capacity renders thepregnant woman more susceptible tometabolic acidosis, particularly DKA.

Normal pregnancy is a diabetogenicstate, marked by relative insulin resis-tance, enhanced lipolysis, elevated freefatty acids, and ketogenesis (6). Ketonebodies can be demonstrated in the serumand urine of normal pregnant womenthroughout the antepartum period (7).

Several hormones produced in preg-nancy, including human placental lacto-gen, progesterone, and cortisol, impairthe action of maternal insulin and con-tribute to this diabetogenic state. Insuli-nase, of placental origin, further depletesmaternal insulin (8). For these reasons,insulin production by the pancreas isaugmented during normal pregnancy. Ifthe woman is a known diabetic, the re-quirement for exogenous insulin also in-

creases with advancing gestation. Insulindose adjustments may not keep pace withincreasing requirements, a fact that mayexplain the higher incidence of DKA inthe second and third trimesters of preg-nancy (9).

Pathophysiology

An overview of the complex patho-physiologic process leading to DKA isprovided in Figure 1 (10).

DKA results from inadequate insulinaction and a failure of glucose utilizationat the cellular level. With insulin defi-ciency, the cell is unable to use glucose asan energy substrate. Insulin counter-regulatory hormones are released in re-sponse to this metabolic condition; glu-cagon, catecholamines, cortisol, andgrowth hormone act on insulin-sensitivetissues to facilitate the production of al-ternative substrates for cellular metabo-lism, using carbohydrates, proteins, andlipids.

Muscle, adipose tissue, and liver are theinsulin-sensitive tissues that respond toboth the decrease in insulin as well as theincrease in counterregulatory hormonesseen in DKA. Decreasing glucose utilizationin muscle results in hyperglycemia. Addi-tionally, protein catabolism provides freeamino acids for gluconeogenesis.

In adipose tissue, the combination ofdecreased insulin and increased counter-regulatory hormones results in the acti-vation of hormone-sensitive lipase. Thisenzyme acts on triglycerides stored in theadipocyte, releasing free fatty acids in

From the Department of Obstetrics, Gynecologyand Reproductive Science, University of Texas HealthScience Center-Houston, Houston, TX.

The authors have no financial or ethical conflict ofinterest regarding the contents of this submission.

Copyright © 2005 by the Society of Critical CareMedicine and Lippincott Williams & Wilkins

DOI: 10.1097/01.CCM.0000183164.69315.13

Objective: The development of diabetic ketoacidosis in preg-nancy is a medical emergency, requiring treatment in an intensivecare setting. Both the mother and the fetus are at risk forsignificant morbidity and mortality. Physiologic changes unique topregnancy provide a background for the development of diabeticketoacidosis. An understanding of these physiologic changesassists in the management of the two patients being treated.Treatment of the patient with diabetic ketoacidosis includes in-sulin therapy and careful fluid management; recommendationsfor management are presented.

Patients: Pregnant women, either with preexisting diabetes orwith diabetes diagnosed during pregnancy.

Conclusions: Prompt recognition of the clinical manifestationsof diabetic ketoacidosis, followed by appropriate, timely treat-ment will optimize outcome for the pregnant woman and herfetus. (Crit Care Med 2005; 33[Suppl.]:S347–S353)

KEY WORDS: diabetic ketoacidosis; pregnancy; anion gap; os-motic diuresis; ketone bodies; acidosis; dehydration; nitroprus-side test; insulin

S347Crit Care Med 2005 Vol. 33, No. 10 (Suppl.)

large quantities, which undergo oxida-tion in the liver to form ketone bodies.Hepatocytes respond to the insulin defi-ciency and high levels of glucagon byincreasing hepatic glucose production.Hyperglycemia results from the increasein gluconeogenesis and glycogenolysis inthe liver and a decrease in glucose utili-zation in peripheral tissues (10–12).

Ketone bodies (acetoacetate, 3-�-hydroxybutyrate and acetone) are pro-duced in the liver from the oxidation offatty acids, under the influence of an el-evated glucagon/insulin ratio (12). �-hy-droxybutyrate production is selectivelyaccelerated in DKA, to a ratio of up to10:1 compared with acetoacetate (7). Ac-etone is produced by spontaneous decar-boxylation of acetoacetate (12) and ispresent in DKA in much lower concen-tration than either �-hydroxybutyrate oracetoacetate. It is highly fat soluble and isexcreted slowly via the lungs. Acetoneproduction is clinically apparent as afruity odor on the breath.

In the laboratory, the conventional testfor the presence of ketone bodies is thenitroprusside test. This test only detectsacetoacetate in serum and urine. Given that�-hydroxybutyrate is present in three toten times greater concentration and that itis metabolized to acetoacetate, the mea-surement of serum ketones using the ni-troprusside reaction may greatly underes-timate the degree of ketonemia present inthe patient. Quantitative testing for �-hy-droxybutyrate is available and should beconsidered if there is concern for a false-negative nitroprusside result (7).

Ketone bodies are organic acids thatcompletely dissociate at physiologic pHand are neutralized by bicarbonate (7). Inhigh concentrations (as seen in DKA),they contribute a large hydrogen ion poolthat consumes the normal buffering ca-pacity of serum and tissue, leading tometabolic acidosis with a high anion gap.

The kidney plays an important role inthe pathogenesis of DKA. Glucose resorp-tion in the tubules of the kidney is effec-

tive up to a threshold of �240 mg/dL,designated Tm (tubular maximum), be-yond which glucosuria is evident (13).Glucose is an osmotically active sub-stance. With increasing hyperglycemia,an osmotic diuresis ensues, leading todehydration, hypovolemia, hyperosmola-lity, and electrolyte depletion. Furtherdepletion of electrolytes occurs becauseketoacid anions are first coupled to so-dium and potassium salts, which are thenexcreted (7). Left untreated, worseningacidosis, electrolyte abnormalities, andhypovolemia can have multiple-systemeffects, including cardiac dysfunction, al-tered vascular tone, poor tissue perfu-sion, and diminished renal function, lead-ing to shock, coma, and death.

Clinical Presentation

Armed with an understanding of al-tered pregnancy physiology and patho-physiology of DKA from the precedingtwo sections, the clinician is now pre-pared to encounter a pregnant womanwith DKA. Typical presenting features areshown in Table 2.

Patients with diabetic ketoacidosisclassically present with a triad of concur-rent abnormalities, which constitute anacronym: D � dehydration; K � ketosis;A � acidosis (metabolic). Althoughmarked hyperglycemia is often detectedas well, DKA is occasionally diagnosedwith only mild elevations in serum glu-cose (10). This is especially true in preg-nant women who develop DKA (5, 6).Careful history taking, physical examina-tion, and laboratory evaluation may re-veal many of the diagnostic features listedin Table 2 (6, 11, 12, 14).

Various triggers of DKA in pregnancyhave been reported in the literature (3,6). In a 15-yr series of consecutive casesof DKA in pregnancy, Montoro et al. (3)found that cessation of insulin therapy inpregnancy was the precipitating factor forDKA in 40% of cases, infection initiated20% of cases, and 30% of women had

previously undiagnosed diabetes untilthey presented with diabetic ketoacidosis(15, 16).

Later in this review, we will present acase of a pregnant woman whose firstmanifestation of diabetes was DKA. Thediagnosis of DKA is less likely to be con-sidered in nondiabetics (15). In additionto Montoro’s three categories, certain ob-stetrical interventions like �-mimetic to-colytic agents and corticosteroids to en-hance fetal lung maturity have beenreported to trigger DKA (17, 18). Finally,ketoacidosis has been reported as a com-plication of the use of a subcutaneousinsulin infusion pump; however, this riskappears to be small (19, 20).

The impact of DKA on pregnancy canbe artificially separated into maternalconcerns and fetal concerns. In reality,both patients are affected simultaneously.

Maternal Concerns

In the same way that DKA representsan acute emergency in the nonpregnantindividual, it poses an immediate threatto maternal well-being. Severe dehydra-tion can lead to hypotension, acidosis cancause organ dysfunction, and electrolyteimbalance can cause cardiac arrhyth-mias. Prompt initiation of treatment cancorrect all of the metabolic derange-ments.

Fetal Concerns

Maternal hyperglycemia results in fe-tal hyperglycemia and fetal osmotic di-uresis. The fetus can also become acidoticfrom ketoacids that cross the placenta.Acidemia decreases uterine blood flow,reduces tissue perfusion, and leads to de-creased oxygenation of the fetoplacentalunit. Furthermore, a leftward shift of thematernal oxy-hemoglobin dissociationcurve with decreased 2,3-diphosphoglyc-erate increases hemoglobin affinity foroxygen, decreasing fetal oxygen delivery(9, 21, 22).

Ketoacids dissociate into hydrogenions and organic anions, both of whichare transported across the placenta. Thus,with increasing maternal ketonemia, fe-tal metabolic acidosis may develop.

Because the fetus is not directly acces-sible, inferences regarding fetal status areoften made from the external recordingof the fetal heart rate. Often, decreased orabsent variability, absent accelerations,and late decelerations are observed onexternal fetal heart rate tracings with de-

Table 1. Incidence of diabetic ketoacidosis in pregnancy

Time Interval Incidence, % (No.)Perinatal Mortality

Rate, % (No.)

Lufkin et al. (1) 1950–1979 7.9 (18/228) 27.8 (5/18)Kilvert et al. (2) 1971–1990 1.7 (11/635) 22Montoro et al. (3) 1972–1987 3.9 (22/560) 35 (7/20)Chauhan et al. (4) 1976–1981 22 35

1986–1991 3 10Cullen et al. (5) 1985–1995 2 (11/520) 9 (1/11)

S348 Crit Care Med 2005 Vol. 33, No. 10 (Suppl.)

compensated maternal DKA. Doppler ul-trasound has also been used to look atblood flow in fetal vessels with DKA (23).

Transient fetal blood flow redistributionwas demonstrated in the umbilical andmiddle cerebral arteries, as measured by

pulsatility index. Reversal of the abnor-mal flow was demonstrated after treat-ment of the maternal DKA. Hagay et al.

Figure 1. Pathophysiologic process leading to diabetic ketoacidosis.

S349Crit Care Med 2005 Vol. 33, No. 10 (Suppl.)

(24) similarly describe improvements inthe fetal heart rate tracing followingtreatment of maternal DKA.

Immediate delivery in response to thepresence of nonreassuring fetal monitor-ing is therefore not indicated, until thematernal metabolic condition is cor-rected. Intervention for fetal indicationsshould be reserved for fetal compromisepersisting after maternal resuscitation.Emergency cesarean delivery in the set-ting of decompensated DKA could furtherworsen the maternal condition (21, 23,24).

A single episode of DKA poses consid-erable risk to the fetus. Several retrospec-tive studies have reported a perinatalmortality rate of 9–35% (1–5). Early rec-ognition and prompt treatment of DKAmight well avoid adverse fetal outcome. Arecent case of DKA in pregnancy that wasnot optimally managed resulted in a fetaldeath.

Case

A 23-yr-old woman gravida 3, para 2 at31 wks gestation presented to the Obstet-rical Service with complaint of “feelingweak” and “throwing up” for 3 days du-ration. The patient complained of ex-treme thirst, nausea, and abdominal pain.She was not known to be diabetic; shedenied the development of gestational di-abetes in either of her two previous preg-nancies. On exam, she was noted to havedry mucous membranes, heart rate of128 beats/min, and respiratory rate of 26breaths/min. Her weight was 104 kg, andblood pressure was mildly elevated (169/87), which was attributed to agitation.Her oral temperature was 100.3°C. Shewas noted to be tolerating oral fluid in-take. Ultrasound examination of the fetusconfirmed cardiac activity. A finger-stickblood glucose measurement was initiallyread as “too high to calculate,” followed

by a measured finger-stick value of �600mg%. Initial laboratory work confirmedhyperglycemia (serum glucose of 836 mg/dL), an anion gap of 23, bicarbonate of 15mmol/L, blood urea nitrogen of 8, andcreatinine of 1.8 mg/dL; urinalysis waspositive for glucose (3�) and leukocytes.Urine and serum ketones were negative.Several attempts were made to obtain anarterial blood gas, but all were unsuccess-ful. The patient was admitted with a di-agnosis of pyelonephritis and hyperglyce-mia.

Five hours after admission to laborand delivery triage, intravenous fluid andinsulin treatment were started. In thefirst hour, the patient received 500 mL ofnormal saline and 5 units of regular in-sulin intravenously. Blood glucose (byfinger-stick) values remained �600 mg%for the next 3 hrs, and the anion gapremained elevated at 20 during this time.Normal saline was maintained at 500mL/hr over this time interval, along with5 units of intravenous regular insulin ad-ministered hourly.

For each of the next 6 hrs, the patientreceived 10 units of regular insulin.Blood glucose values remained in therange of 375–440 mg/dL. The patient re-mained tachycardic (125 beats/min), andurine output was recorded at �100 mL/hr. The intravenous fluid was changed to0.45% saline, at a rate of 100 mL/hr overthis time period. Potassium chloride wasadded to intravenous fluids to correct aserum potassium of 3.1 mmol/L.

The fetus was monitored by an exter-nal fetal heart rate monitor, with a base-line fetal heart rate ranging from 125 to140 beats/min. Notably, this was the sameas the mother’s pulse. Ten hours afteradmission, after difficulty in obtaining acontinuous fetal heart rate tracing, ultra-sound confirmed an intrauterine fetal de-mise.

This case illustrates delayed recogni-tion of the clinical manifestations ofDKA. Had a timely diagnosis been made,appropriate fluid and insulin therapycould have been instituted, with correc-tion of the maternal metabolic distur-bances, and perhaps the fetal death wouldhave been prevented.

Treatment

A careful search for the presence ofinfection is required for every patient pre-senting in DKA. Infection has been dem-onstrated to reduce the rate of glucoseutilization by 50% (28). The presence ofbands on a differential is more telling ofinfection, as an elevated white blood cellcount may be the result of dehydration,rather than infection. Common sites ofinfection include the urinary tract, lungs,soft tissue, sinuses, skin, and teeth.Prompt treatment of infection with ap-propriate antibiotic coverage will assist inglucose control. Other precipitatingcauses of DKA are referenced earlier (17–20).

We now present our recommenda-tions for treatment of DKA. An algorith-mic approach is outlined in Figure 2,focusing on the treatment of the fetus,maternal acidosis and dehydration, andthe elucidation and treatment of under-lying causes of DKA. This is followed byrecommendations for therapy in a “cook-book” form, detailing fluid administra-tion, insulin therapy, and electrolyte re-placement (Table 3).

For the interested reader, we concludethis section with a rationale for our rec-ommendations for treatment of DKA, in-cluding pertinent calculations discussedthroughout the text (Table 4). Treatmentof the pregnant woman with DKA doesnot differ from the nonpregnant patient.The goals of DKA management are asfollows:

1. Improve circulating volume and tis-sue perfusion.

2. Decrease serum glucose.3. Clear the serum and urine of keto-

acids (correcting acidosis).4. Correct electrolyte imbalances.5. Treat initiating causes of DKA.6. Monitor therapeutic response in the

mother and fetus.These goals will be met with volume

replacement, insulin therapy, replace-ment of electrolytes, and treatment of theunderlying causes of DKA (10, 14).

Fluid Therapy. Volume replacementshould be vigorous and sustained to cor-

Table 2. Diagnosis of diabetic ketoacidosis (DKA)

Signs and Symptoms Labs

Malaise Hypotension Hyperglycemia (�250 mg/dL)a

Dehydration Tachycardia Acidosis (arterial pH � 7.3)Polyuria Tachypnea Anion gap (�12 mEq/L)Polydipsia Hyperventilation Low bicarbonate (�15 mEq/L)

(Kussmaul respiration) Elevated base deficit (base deficit �4 mEq/L)Nausea Drowsiness Ketonemia (�1:2 dilution)Vomiting Mental status changeAbdominal pain LethargyIleus ComaFruity breath odor Shock

aIn pregnancy, DKA can manifest at lower blood glucose values (may be as low as 180 mg/dL).

S350 Crit Care Med 2005 Vol. 33, No. 10 (Suppl.)

rect dehydration, replete circulating vol-ume, decrease serum glucose concentra-tion, restore renal function, and,importantly, prevent a recurrence of aci-dosis. The major cause of fluid loss inDKA is osmotic diuresis. Isotonic fluid(0.9% normal saline) should be used forinitial fluid replacement. Isotonic fluidhas been demonstrated to replete the ex-tracellular fluid compartment and cor-

rect circulating plasma volume most ef-fectively (11).

Insulin Therapy. Insulin is the main-stay of treatment, correcting the hyper-glycemia, acidosis, and the overproduc-tion of ketones that occur in DKA.Continuous insulin infusion of regular(short-acting) insulin is the preferredmethod of delivery. Intramuscular ad-ministration of insulin leads to a delay in

response to treatment due to the im-paired absorption of insulin in the settingof volume depletion and acidosis (11).Intravenous insulin therapy should bemaintained until after the first dose ofsubcutaneous insulin is administered, toprevent a recurrence of ketoacidosis. Sev-eral protocols describe the subcutaneousadministration of insulin after an episodeof DKA (11, 25).

• Admit to obstetric intensive care unit (or equivalent).• Begin monitoring vital signs, pulse oximetry, strict intake and output, patient weight.• Labs: arterial blood gas, serum glucose, serum ketones, serum electrolytes, anion gap, urine analysis, urine culture, urine ketones.• Fetal monitoring if viable gestational age.

Start a bedside flowsheet.

Figure 2. Algorithm for treatment of diabetic ketoacidosis. ABG, arterial blood gas; UA, urine analysis; CXR, chest radiograph; IV, intravenous; NS, normalsaline.

S351Crit Care Med 2005 Vol. 33, No. 10 (Suppl.)

Electrolyte Replacement

Potassium. Potassium replacement isindicated after fluid replacement and in-sulin have been initiated. The most rapidchanges in potassium occur in the firstfew hours of treatment (26). Insulin me-diates the shift in K� ion from the extra-cellular to intracellular space, along withglucose. With correction of acidosis, K�

ions are further shifted into the intracel-lular fluid. Losses of potassium also occurwith continued osmotic diuresis and asketone bodies are excreted by the kidneyas potassium salts. The development ofsignificant hypokalemia can precipitatelife-threatening arrhythmias (11); closeattention to serum electrolytes and potas-sium replacement is essential to the ef-fective treatment of DKA.

On the average, a potassium deficit of5–10 mEq/kg body weight occurs (11,12). Various guidelines are described forpotassium replacement (12, 25, 27).Maintenance of serum K� levels at4–5mEq/L will prevent hypokalemia andthe associated cardiac arrhythmias. Po-tassium chloride is most often used forreplacement, at a maximum rate of 40mEq/hr. Potassium phosphate may alsobe used for replacement, if concomitanthypophosphatemia is present or with thedevelopment of hyperchloremic acidosis.

Phosphate, Magnesium, and Calcium.Inorganic cations are also deficient inDKA; however, the need for replacementof these electrolytes is debated. Studiesdo not demonstrate the benefit of re-placement of these cations (26). As men-tioned, potassium phosphate can be usedfor potassium replacement, if phosphatelevels are �1.0 mg/dL (decrease dose ofKCl accordingly). Replacement has alsobeen recommended for patients with leftventricular dysfunction or with obtunda-tion of mental status that has not re-sponded after the initial treatment ofDKA (27).

Bicarbonate. The use of bicarbonate tocorrect acidosis has been the subject ofstudy and controversy (22, 26). The me-tabolism of ketone bodies via the citricacid cycle acts to rapidly regenerate bi-carbonate levels. Studies demonstratethat bicarbonate administration has notproven beneficial in the treatment of DKAand its use may be associated with risk,including a paradoxical central nervoussystem acidosis, hypokalemia, hyperto-nicity, and the development of cerebraledema (22, 26). Treatment of DKA withbicarbonate delays the correction of ke-

Table 3. Recommendations for therapy

Fluid therapyEstimate fluid deficit of �100 mL/kg body weight.Correct 75% of estimated fluid deficit over first 24 hrs.Record intake and output hourly.Initial 24 hrs:

Use isotonic saline (0.9% NS)First hour Give 1 L NSSecond hour Give 0.5–1 L NSThird hour Give 0.5 L NSThereafter (24 hrs) Give 0.25 L/hr 0.45% NS until 75% deficit corrected

Calculate corrected Na� during fluid administration.Monitor serum Na� and Cl�; adjust IV fluids if hypernatremia or hyperchloremia.Continue hydration for 24–48 hrs, until anion gap and acidosis have corrected and volume

deficit is replaced.Insulin therapy

Give 0.1 units/kg bolus regular insulin IV, then 0.1 units/kg/hr IV continuous infusion by pump.Mix 50 units of regular insulin in 500 mL of NS (10 mL � 1 unit).Flush IV tubing prior to infusing, as insulin binding to tubing occurs.Monitor serum glucose every 1–2 hrs.If serum glucose is not decreased by 20% within first 2 hrs, double insulin infusion rate.Aim for a decrease in serum glucose of 60–75 mg/dL/hr.With blood glucose at 250 mg/dL, add dextrose to IV fluids, to avoid hypoglycemia (maintain

insulin rate).Monitor serum ketones every 2 hrs.Continue insulin therapy until bicarbonate and anion gap normalize.After full resolution of ketosis, maintain insulin drip until after the first subcutaneous dose of

insulin is administered.Electrolyte replacement

Monitor serum electrolytes every 2–4 hrsPotassium

Potassium chloride is most often given for replacement; occasionally potassium phosphateor acetate is used.

Anticipate deficit of 5–10 mEq/kg.With replacement, maintain adequate urine output (0.5 mL/kg/hr).Maintain serum K� level at 4–5 mEq/L.Suggested protocol using serum K�:

�5 mEq/L No treatment4–5 mEq/L 20 mEq/L replacement3–4 mEq/L 30–40 mEq/L replacement�3 mEq/L 40–60 mEq/L replacement

PhosphateReplacement not normally required.If serum phosphate �1.0 mg/dL or if cardiac dysfunction or obtundation, replace using

potassium phosphate.Bicarbonate

Replacement is usually not necessary.

NS, normal saline; IV, intravenous.

Table 4. Calculations

● Anion gap � [Na � (Cl � HCO3)]: normal value 12 mEq/L � 2Anion gap measures the difference between unmeasured anions and unmeasured cations.An increased anion gap reflects the presence of a metabolic acidosis, if renal failure is not present.● Serum osmolality (mOsm/L) � [2(Na � K) � (glucose/18) � (BUN/2.8)]

Normal value � 290 � 5Serum osmolality is elevated less in DKA (300–330 mOsm/L) than in nonketotic hyperosmolar

coma (�350 mOsm/L). The difference in serum glucose accounts for almost the entire differencein osmolality.

● Corrected serum sodium (mEq/L) � measured Na � {([plasma glucose (mg/dL)] � 100)/100} 1.6Normal range � 135–145 mEq/L

Hyperglycemia dilutes plasma Na by 1.6 mEq/L for every 100 mg/dL increase in glucose.An elevated value indicates marked dehydration; below the range suggests too rapid administration

of fluids.● Total body water deficit � {[0.6 body weight (kg)] � [1 � (140/serum Na)]}

BUN, blood urea nitrogen; DKA, diabetic ketoacidosis.

S352 Crit Care Med 2005 Vol. 33, No. 10 (Suppl.)

tonemia (22). Further studies are neededto assess the benefits and risks of bicar-bonate administration when pH is �6.9or in cases complicated by cardiac dys-function, sepsis, or shock (10).

REFERENCES

1. Lufkin ES, Nelson RL, Hill LM, et al: Ananalysis of diabetic pregnancies at MayoClinic, 1950 –79. Diabet Care 1984;7:539–547

2. Kilvert JA, Nicholson HO, Wright AD. Keto-acidosis in diabetic pregnancy Diabet Med1993; 10:278–281

3. Montoro MN, Myers VP, Mestman JH, et al:Outcome of pregnancy in diabetic ketoacido-sis. Am J Perinatol 1993; 10:17–20

4. Chauhan SP, Perry KG, McLaughlin BN, etal: Diabetic ketoacidosis complicating preg-nancy. J Perinatol 1996; 16:173–175

5. Cullen MT, Reece EA, Homko CJ, et al: Thechanging presentations of diabetic ketoaci-dosis during pregnancy. Am J Perinatol1996; 13:449–451

6. ACOG Practice Bulletin No. 60. Pregesta-tional Diabetes Mellitus. Obstet Gynecol2005; 105:675–685

7. Laffel L: Ketone bodies: A review of physiol-ogy, pathophysiology, and application ofmonitoring to diabetes. Diab Metab Res Rev1999; 15:412–426

8. Ogata ES. Carbohydrate homeostasis. In:Neonatology Pathophysiology and Manage-ment of the Newborn. Avery GB, FletcherMA, MacDonald MG (Eds). Philadelphia, Lip-pincott Williams & Wilkins, 1999

9. Ramin K: Diabetic ketoacidosis in pregnancy.Obstet Gynec Clin NA 1999; 26:481–488

10. White NH: Management of diabetic ketoaci-dosis. Rev Endocr Metab Dis 2003;4:343–353

11. Kitabchi AE, Umpierrez GE, Murphy MB etal: Management of hyperglycemic crises inpatients with diabetes. Diabet Care 2001; 24:131–153

12. Foster DW, Mc Garry JD: The metabolic de-rangements and treatment of diabetic keto-acidosis. N Engl J Med 1983; 809:159–169

13. Rose BD. Proximal tubule. In: Clinical Phys-iology of Acid-Base and Electrolyte Disor-ders. Rose BD (Ed). New York, McGraw-Hill,1989, p 43–54

14. Moore TR: Diabetes in pregnancy. In: Mater-nal Fetal Medicine Principles and Practice.Fifth Edition. Creasy RK, Resnik R, Iams JD(Eds). Philadelphia, Saunders (Elsevier),2004, pp 1023–1063

15. Sills IN, Rappaport R: New onset IDDM pre-senting with diabetic ketoacidosis in a preg-nant adolescent. Diabet Care 1994; 17:904–905

16. Kamalakannan D, Baskar V, Barton DM, et al:Diabetic ketoacidosis in pregnancy. PostgradMed 2003; 79:699–714

17. Bernstein IM, Catalano PM: Ketoacidosis inpregnancy associated with the parenteral ad-ministration of terbutaline and betametha-sone. A case report. J Reprod Med 1990; 35:818–820

18. Bedalov A, Balasubramanyam B: Glucocorti-coid-induced ketoacidosis in gestational dia-betes: Sequela of the acute treatment of pre-term labor. Diabet Care 1997; 20:922–924

19. Gabbe SG, Holing E, Temple P, et al: Bene-fits, risks, costs, and patient satisfaction as-

sociated with insulin pump therapy for thepregnancy complicated by type I diabetesmellitus. Am J Obstet Gynecol 2000; 182:1283–1291

20. Lindenbaum C, Menzin A, Ludmir J: Diabeticketoacidosis in pregnancy resulting from in-sulin pump failure. A CASE report. J ReprodMed 1993; 38:306–308

21. Chauhan SP, Perry KG: Management ofdiabetic ketoacidosis in the obstetric pa-tient. Obstet Gynecol Clin N Am 1995; 22:143–155

22. Rodgers BD, Rodgers DE: Clinical variablesassociated with diabetic ketoacidosis duringpregnancy. J Repro Med 1991; 36:797–800

23. Takahashi Y, Kawabata I, Shinohara A, et al:Transient fetal blood flow redistribution in-duced by maternal diabetic ketoacidosis di-agnosed by Doppler ultrasonography. PrenatDiagn 2000; 20:524–525

24. Hagay ZJ, Weissman A, Lurie S, et al: Rever-sal of fetal distress following intensive treat-ment of maternal diabetic ketoacidosis. Am JPerinatol 1994; 11:430–432

25. Foley MR: Diabetic ketoacidosis in pregnancyIn: Obstetric Intensive Care Manual. SecondEdition. Foley MR, Strong TH, Garite TJ(Eds). New York, McGraw-Hill, 2004

26. Adrogue HJ, Madias NE: Management of life-threatening acid-base disorders. N EnglJ Med 1998; 1:26–33

27. Magee MF, Bhatt BA: Management of decom-pensated diabetes. Crit Care Clin 2001; 17:75–106

28. Hagay ZJ: Diabetic ketoacidosis in preg-nancy: Etiology, pathophysiology, and man-agement. Clin Obstet Gynecol 1994; 37:39–49

S353Crit Care Med 2005 Vol. 33, No. 10 (Suppl.)