Hyperosmolar Hyperglycemic State Workup

Updated: May 26, 2023
  • Author: Dipa Avichal, DO; Chief Editor: George T Griffing, MD  more...
  • Print

Approach Considerations

On the basis of the consensus statement published by the American Diabetes Association, diagnostic features of hyperosmolar hyperglycemic state (HHS) may include the following [6, 8] :

  • Plasma glucose level of 600 mg/dL or greater
  • Effective serum osmolality of 320 mOsm/kg or greater
  • Profound dehydration, up to an average of 9 L
  • Serum pH greater than 7.30
  • Bicarbonate concentration greater than 15 mEq/L
  • Small ketonuria and absent-to-low ketonemia
  • Some alteration in consciousness

HHS should be considered in children presenting with hyperglycemia and hyperosmolarity without significant ketoacidosis. It is particularly important to distinguish HHS from diabetic ketoacidosis (DKA) in children, because younger persons are at higher risk for the development of cerebral edema as a complication of aggressive fluid repletion.


Blood Studies

Serum glucose

A fingerstick blood sugar measurement is the simplest first step in the evaluation. The serum glucose level usually is elevated dramatically, most often to greater than 600 mg/dL. Many patients present with glucose concentrations greater than 1000 mg/dL. Blood sugar levels of 65-250 mg/dL exclude significant glycemic derangement and should prompt a search for other causes of present symptoms.

The concentration of glucose in the plasma is directly proportional to the degree of dehydration. Higher concentrations of glucose relate to higher degrees of dehydration, higher plasma osmolality, and a worse prognosis.

Monitor the plasma glucose concentration hourly during the first 24-48 hours of treatment.

Hemoglobin A1c

Although hemoglobin A1c (glycosylated hemoglobin) levels are not useful in the acute phase of therapy, they may be obtained as an indicator of the patient’s glucose control over the previous several weeks. An elevated A1c level may help in determining medication noncompliance or undiagnosed DM. A normal A1c is useful in determining whether the episode of HHS is secondary to an underlying acute process (ie, infection, myocardial infarction [MI]).  

Serum osmolarity or osmolality

Normal serum osmolality ranges from 280 to 290 mOsm/kg. A serum osmolality of 320 mOsm/kg or higher defines HHS. Rarely, serum osmolality may exceed 400 mOsm/kg. In HHS, higher serum osmolality relates to greater impairment of the level of consciousness. Serum osmolality may be calculated from sodium, blood urea nitrogen (BUN), and glucose values, as follows:

Osm = (2 × Na) + (glucose/18) + (BUN/2.8)

The osmole gap is the difference between the measured osmolality and the calculated osmolality (at low solute concentrations, they are nearly equivalent measures). Although the measured osmolality is very high in patients with HHS, the osmole gap should be unimpressive, because the calculated osmolality includes the elevated serum glucose concentration. If the osmole gap is very large, consider toxic alcohol ingestion.

Serum electrolytes

As HHS progresses and osmotic diuresis occurs, electrolytes are lost in the urine. All electrolytes are extremely deficient at the time of presentation, at which time the relative deficiencies of water and electrolytes determine their plasma concentrations. Additionally, the presence of hypertriglyceridemia affects the concentration of electrolytes. Triglycerides exert an osmotic drag and displace electrolytes in the plasma.


Hyponatremia or hypernatremia may be present. In the setting of hyperglycemia, pseudohyponatremia is common as a result of the osmotic effect of glucose drawing water into the vascular space. The measured serum sodium concentration can be corrected upward in proportion to increases in serum glucose to yield an estimate of what the serum sodium level would be in the absence of hyperglycemia and its associated osmotic effect. To correct sodium for hyperglycemia, the following calculation can be used [28]

Corrected Na = Na + ([Gluc - 5]/3.5)

Some patients may present with elevated serum sodium concentrations. Patients with hypernatremia usually have elevated plasma osmolality and are more often found with neurologic symptoms. [29]


Hypokalemia or hyperkalemia may be present. Commonly, at time of presentation of HHS, serum potassium may be elevated due to an extracellular shift caused by insulin deficiency. However, total body potassium is likely low regardless of its serum value. The average potassium deficit in normally about 300-600 mEq. A low measured serum potassium suggests profound total body losses, and patients should be placed on cardiac monitoring. During treatment, insulin drives potassium into cells, and intravenous (IV) hydration dilutes potassium in the circulation. Aggressively replace potassium to maintain plasma levels in the normal range during treatment.


Serum magnesium levels are also a poor indicator of true total body magnesium. In the presence of hypokalemia, concomitant hypomagnesemia should be presumed and treated.

Bicarbonate and anion gap

The bicarbonate concentration in a patient with HHS is usually normal or mildly reduced. This is because there is minimal ketone formation in the process of HHS, in contrast to DKA, in which bicarbonate levels can be markedly reduced (bicarbonate < 15mEq/L). 

The anion gap is calculated according to the following formula:

(Na+ + K+) - (Cl + HCO3)

The calculated anion gap in HHS is usually within normal limits (8-12 mmol/L). A wide anion gap can be observed in patients with HHS, reflecting mild metabolic acidosis. The mild acidosis in HHS is often multifactorial and results, in part, from the accumulation of minimal ketoacids in the absence of effective insulin activity. Some patients with profound dehydration may have high anion gaps, reflecting the additional contribution of lactic acid produced by hypoperfusion of tissues. Underlying renal disease with uremia also may contribute to a high anion gap.

Monitor plasma electrolyte levels at least every 4 hours during the first 24-48 hours of treatment.

Blood gases

Arterial blood gas (ABG) values are obtained to measure serum pH. In most cases of HHS, the blood pH is greater than 7.30. ABG values also indicate underlying diseases associated with HHS. Hypoxemia may be observed in association with cardiac or pulmonary diseases. Hypocarbia may be due to respiratory alkalosis as a compensatory mechanism to a primary metabolic acidosis. Hypocarbia also may be due to tachypnea in response to an elevated alveolar-arterial oxygen gradient from pulmonary disease. 

Venous blood gas (VBG) values may be substituted in patients with normal oxygen saturation on room air. VBGs provide comparable information, are easier to draw, and are less painful to the patient. The pH measured by a VBG assessment is 0.03 pH units less than the pH measured by ABG assessment [30] .

Serum ketones

A mild degree of ketosis is usually observed in any patient who is dehydrated. In those with HHS, despite the significant degree of dehydration, ketosis is mild and responds readily to treatment. Profound ketosis that does not respond readily to IV rehydration is the norm in persons with DKA. Mild to moderate ketosis can be present when the disease has features of both HHS and DKA (overlap cases).

Renal function studies

Patients with HHS present with acute elevations in BUN and creatinine secondary to prerenal azotemia from volume depletion. Initially, BUN and creatinine concentrations are likely to be elevated, and the BUN-to-creatinine ratio may exceed 30:1. When possible, these values should be compared with previous values; many patients with diabetes have baseline renal insufficiency. If a patient's renal function does not normalize after treatment, this may indicate irreversible or underlying renal damage.

Serum enzymes

Dehydration causes a rise in the plasma levels of albumin, amylase, bilirubin, calcium, creatinine kinase (CK), lactate dehydrogenase, lipase, total protein, and transaminases. Up to two thirds of patients with HHS have elevated serum enzyme levels. Accordingly, serum levels of CK and isoenzymes should be measured routinely because both MI and rhabdomyolysis can trigger HHS and both can be secondary complications of HHS. [31]  Also, elevated amylase and lipase does warrant an exclusion of underlying pancreatitis even though these enzymes may be elevated during HHS. Clinical correlation is needed.

Avoid the assumption that enzyme level elevation is due to dehydration. Exclude underlying disease associated with each of these abnormal blood levels in patients with HHS. This is especially true in the case of CK elevations.

Complete blood count (CBC)

Leukocytosis is frequently observed in HHS. It can be secondary to HHS itself or result from an underlying infection. Elevated levels of counterregulatory hormones, stress, dehydration, and demargination of leukocytes during HHS may give rise to leukocytosis. Even though HHS can cause leukocytosis, a leukocyte count of over 25,000 or bands greater than 10% may suggest an underlying infection, and workup is warranted. [32] A complete history and physical exam will help determine the source of infection. Obtain a chest radiograph, a urine culture, and possibly a blood culture, in patients with leukocytosis with suspicion of infection.


Urine Studies

Urinalysis can reveal elevated specific gravity (evidence of dehydration), glycosuria, small ketonuria, and evidence of urinary tract infection (UTI). However, urine for analysis may be difficult to obtain in a severely dehydrated patient with HHS. Catheterization of the urinary bladder may be necessary.

Urinalysis may provide further information about the patient’s metabolic state. Ketones are rarely present in persons with HHS. Glycosuria may be a sign of uncontrolled diabetes in a patient presenting with HHS. Gross proteinuria suggests underlying renal disease. Urinary osmolality and the urine specific gravity can be very high in patients with HHS secondary to dehydration.

Urine cultures may be obtained if clinical suspicion is high for UTI and if urinalysis shows signs of infection. Send cultures as clinically indicated.


Radiography of Chest and Abdomen

In the initial evaluation of patients with HHS, a chest radiograph is advisable to exclude pneumonia. Radiographic findings may be falsely negative at first because of the profound dehydration in some patients, and serial studies may document pneumonia once the patient has been volume resuscitated.

Abdominal radiographs are indicated if the patient has abdominal pain or is vomiting.


Computed Tomography of Head

Patients with HHS who present with altered mental status may have an underlying CNS disease. Computed tomography (CT) of the head is indicated in many patients with focal or global neurologic changes to help exclude hemorrhagic strokes, subdural hematoma, subarachnoid bleeding, intracranial abscesses, and intracranial masses. It may be useful for patients who show no clinical improvement after several hours of treatment, even in the absence of clinical signs of intracranial pathology.

Repeat CT scanning is indicated if cerebral edema is a concern during the treatment of HHS.



Electrocardiography (ECG) is indicated in all patients with HHS because myocardial infarction (MI) and pulmonary embolism (PE) can precipitate HHS. The height of the T waves in the ECG tracings may point to a potassium derangement. The duration of the QT interval may be abnormal as a consequence of calcium abnormalities.


Other Tests

Cerebrospinal fluid (CSF) cell count, glucose, protein, and culture are indicated in patients with an acute alteration of consciousness and clinical features suggestive of possible CNS infection. Patients who are immunocompromised may require additional studies of the CSF, such as polymerase chain reaction (PCR) assay for herpes simplex virus (HSV) and cryptococcal antigen.

When meningitis or subarachnoid hemorrhage is suspected, lumbar puncture (LP) is indicated. If meningitis is suspected clinically, do not withhold antibiotics while waiting for the LP to be completed.

A study by Hu and Lin indicated that in type 2 DM patients, the CHA2DS2-VASc score can be used to predict the incidence of new-onset atrial fibrillation but that this ability is diminished in individuals with comorbid HHS. [33]