Diabetic Emergencies Case 1

Answer the questions at the bottom of the page when all data has been reviewed

History & Examination

An 92 year old woman is found collapsed at home. Very little history is available. She is accompanied to the Hospital by a member of the Salvation Army who had found her on the kitchen floor. The charity worker is able to give ER/ED the name of the patient’s GP (family physician). Her GP is contacted. He has been her doctor for 30 years and knows her well. In his opinion she is a remarkable lady. Throughout her life she had devoted much of her time to community work with the Salvation Army and had remained fully self-caring until very recently. She had, however, started worrying in recent months about difficulty in getting to the shops. This was due to episodes of chest tightness on exertion which he had ascribed to mild exertional angina. He had started her on aspirin. She developed type II diabetes some years ago which had been reasonably well controlled on oral hypoglycemics. Her GP felt that she had become somewhat frailer since the diagnosis of diabetes and perhaps had been developing some mild cognitive impairment. She has no immediate family and lives alone. The area in which she lives is ‘not the best’ and he is a little concerned about her ability to return home on discharge.

On examination she is confused. There is clinical evidence of marked dehydration. There are occasional bilateral inspiratory crepitations on auscultation of the lung fields. There is no ankle oedema. There is no evidence of head injury. There are no focal neurological signs.

Investigations

Course 2 Case 1 bloods

Master ECG CSi

Portable AP chest radiograph

Report: poor quality image. no gross abnormality

Observations
Scanned_20141013143831

Ethel Shelby fluids


Questions

Q1. What is the metabolic diagnosis at presentation?

Q2. What has triggered the condition in Q1?

Q3. Analyse her ABG properly. What, if any, abnormalities are present?

Q4. Can you explain the serum sodium level?

Q5. What additional blood tests will be useful?

Q6. Outline the key points in management of this case.


Answers

Q1. What is the metabolic diagnosis at presentation?

Probable Hyperosmolar Hyperglycemic State (HHS). In an elderly type II diabetic presenting with profound dehydration, hypernatremia, hyperglycemia (urinalysis: glucose ++++) and apparent absence of ketoacidosis (urinalysis: ketones trace) this is the most likely diagnosis. Diagnosis of HHS requires definitive demonstration of hyperglycemia with high effective plasma osmolality and the absence of significant ketoacidosis (Q5 and HHS article below)

Q2. What has triggered the metabolic condition in Q1?

Probable ST elevation MI. The admission ECG is highly abnormal. She has right bundle branch block (RBBB, ECG features circled in black below). We don’t know if this is old or new. However, in the presence of RBBB, we can still comment on the ST segments. There is ST elevation in V2 (arrow), V3 and probably V4. A troponin level will be of interest but should be interpreted in the context of her acute kidney injury. It is difficult to place a definite age on these changes. Acute coronary syndrome is a common precipitant of HHS (see below). The event is often ‘silent’ (no chest pain reported) possibly due to co-existent autonomic neuropathy in this elderly population of diabetics.

Master ECG exp

MI is common in the setting of HHS but it is often difficult to be sure if it is the precipitant of HHS or a consequence of the hypercoagulable state associated with HHS (chicken or egg?). Acute thrombotic events (stroke, MI PE etc) occur frequently in the setting of HHS. Therefore, thromboprophylaxis is mandatory in these patients in the acute phase.

Q3. Analyse her ABG properly. What, if any, abnormalities are present?

Ethel Shelby ABG revised

Clinically, this lady is dehydrated and running a low blood pressure for someone of her age. We might reasonably suspect that she has problems with tissue perfusion. Also, we will confirm that she is in a hyperosmolar hyperglycemic state (HHS), a condition often associated with poor tissue perfusion and increased lactate levels.  So, before looking at the ABG, we expect to see a metabolic acidosis and indeed that is what we see here. The pH is low (an acidosis is present). The level of bicarbonate (‘metabolic base’) is low. Therefore, this is a metabolic acidosis. Her anion gap, [Na+] – ([HCO3-] + [Cl-]), is grossly elevated at 31 mEq/l (normal <18). This metabolic acidosis is due to ‘fixed’ acid addition to the extracellular fluid.

The pCO2 (respiratory acid) is low consistent with respiratory compensation. Using US units (mmHg = 7.5kPa), we can apply Winter’s formula. The expected pCO2 is given by 1.5[HCO3-] + 8 (+/-2) = 30 – 34 mmHg. As the observed pCO2 lies within this range we conclude that respiratory compensation is partial (pH has not returned to normal) but is adequate (in the expected range).

We also note an elevated paO2. This is explained by two factors. Firstly, she is hyperventilating, blowing off CO2 (‘respiratory acid’) to compensate for the metabolic acidosis. Secondly, she was on 36% oxygen at the time the gas was taken (4L O2 via nasal prongs, see observation chart). Using US units and the alveolar gas equation (A-a gradient), we can show that this  level of oxygen therapy and alveolar ventilation results in a pAO2 of 217 mmHg. This means her observed A-a gradient is 40 mmHg, close to the predicted gradient for her age and level of oxygen therapy (A-a gradient). There is no evidence here of any problems with her lungs.

Q4. Can you explain the serum sodium level?

Glucose filtered at the glomerulus is acting as an osmotic diuretic pulling water into the tubules. This is combined with a failure to increase her oral fluid intake resulting in a rise in the concentration of Na+ in the extracellular fluid (ECF). In the elderly, thirst mechanisms are often impaired or access to water limited.  Hyperglycemia in a person with intact thirst mechanisms and access to water may result in dilution of Na+ in the ECF (hyponatremia).

See discussion of hypernatremia below

Q5. What additional blood tests will be useful?

In order to establish a diagnosis of HHS we need to show that hyperglycemia is associated with significantly elevated plasma osmolality in the absence of ketoacidosis.

So, we measure:

Plasma glucose

Plasma osmolality

Serum ketones

A troponin level may help us establish the presence of an underlying acute coronary syndrome

A creatinine kinase (CK) level is needed. As we’ll see, she has some features of rhabdomyolysis

Blood cultures should be sent as we are querying underlying infection

Q6. Outline the key points in management of this case. 

When managing HHS use your own local protocols. The steps outlined here are used to teach students. Do not employ them in clinical practice.

We can go through the management of this patient in principle. In reality, her management will be tricky and will require a little wisdom. We will touch on these issues in the discussion below.

She is unlikely to survive outside an ICU setting and should be transferred to ICU.

Principles of management of HHS

1. Maintain the airway.

2. Replace fluids. These patients are severely dehydrated. We can calculate the total body water deficit based on the serum sodium, total body water deficit = (Na+ – 140)/140 x weight x 0.4 ( x 0.5 in a male) (Kgs). In our patient, (155 – 140)/140 x 60 x 0.4 = 3.2 L. We gave her 1L NaCl over the first hour. We then aimed to correct half the deficit over the next 12 hours and the remainder over 24 hours. She is having an MI and must be at risk of pulmonary edema. Hence, during this period of resuscitation, we monitor her O2 saturation regularly, examine her chest and monitor her urine output closely.

Note that she has a relatively high K+ level on admission. We, therefore, do not initially add K+ to her fluids. However, we will need to keep a close eye on this. Her serum K+ may be deceptive. She has been found on the floor and is at risk of rhbdomyolysis. K+ release from damaged muscle may cause a raised serum K+ level in the presence of total body depletion. Hopefully, you noted the urinalysis abnormalities on her fluid balance chart. Myoglobinuria will give a positive result for ‘blood’ on urinalysis (microscopy, however, will be negative for red cells). Indeed, subsequent blood tests showed a markedly elevated serum CK. We will need to recheck her serum K+ early during the period of resuscitation. If she becomes hypokalemic, we may need to supplement her fluids with K+.

3. Administer insulin as appropriate. In this lady’s case, her serum glucose came back at 631 mg/dl (35 mmol/l). She was commenced on an intravenous sliding scale of insulin.

4. Treat the underlying cause. In this patient’s case the underlying trigger to HHS was an acute coronary syndrome. We involved cardiology and initially managed this conservatively. She subsequently underwent stenting of a lesion in the left anterior descending artery.

Could she have an underlying infection? This is a possibility. She as a low temperature on admission and a low blood pressure. Her urinalysis abnormalities we could account for on the basis of a degree of rhabdomyolysis but what about her elevated white cell count (16)? This could relate to infection, however, it is well worth knowing that hyperglycemia per se is associated with leukocytosis. In fact, her CRP was only mildly elevated. We did not start this lady on antibiotics but kept this decision under review while monitoring her response to fluid resuscitation and insulin.

5. DVT prophylaxis. These profoundly dehydrated immobile patients are at risk of DVT/PE, prophylaxis with heparin is essential. However, keep in mind that the combination of heparin and aspirin in the elderly carries a high risk of gastrointestinal hemorrhage.


Subsequent Course

WC Dover

Her renal indices improved rapidly in response to rehydration. Six hours post-admission, she underwent cardiac catheterisation with stenting of a stenosis in the left anterior descending artery. Two days later she is re-established on oral hypoglycemics.

During her recovery period, the nurses are concerned about her ability to cope at home. She is of her generation, although she is clearly worried about going home, she does not want to be ‘any trouble to anybody’. She is reviewed by the physiotherapists and occupational therapists who agree that she would be at risk living on her own. The Consultant in charge of her care is asked to speak with her. She has no  immediate family and is well used to being alone. She had not remarried following the death of her young husband, killed during the Blitz while on leave. With much persuasion, the doctor and social worker get her to agree to placement in a local residential home for the elderly.

Several weeks after discharge she writes to her consultant, Professor Jazzo (pronounced ‘Jozzel’),

‘…the home is a great success with aerobics, line dancing and daily walks around the grounds. There are many other social activities which I can’t join in because I simply don’t have the time…’


Articles

HHS: general features

HHS is a life-threatening metabolic derangement. It is usually seen in type II diabetics but may occur in patients with type I diabetes. In affected individuals, an intercurrent illness impairs glycemic control resulting in marked sustained hyperglycemia (plasma glucose >600 mg/dl, >30 mmol/l). Glucose is a major determinant of plasma osmolality and the condition is, therefore, characterized by a greatly increased plasma osmolality. This results in a shift of water from the intracellular to the extracellular fluid compartment. Cellular dehydration adversely affects cerebral function so neurological abnormalities, particularly altered conscious level, are a feature of the condition.  In addition, glucose is freely filtered at the renal glomerulus. The high level of glucose in the glomerular filtrate results in an osmotic diuresis with loss of large quantities of free water in the urine.

Therefore, clinically HHS is characterized by profound dehydration and a degree of neurological impairment which may be focal or global. On blood testing, there is markedly raised plasma glucose and a raised effective plasma osmolality*.

HHS is distinct from the second hyperglycemic emergency affecting diabetics, diabetic ketoacidosis (DKA). However, approximately one fifth of cases of HHS will have a coexistent DKA. In addition, many of these patients will have a degree of lactic acidosis resulting from multiple mechanisms.

HHS diagnosis

Based on the features of the condition outlined above, the American Diabetes Association has proposed that the following criteria may be used for the diagnosis of HHS

  1. A plasma glucose of 600 mg/dl or greater (>30 mmol/l).
  2. A raised effective plasma (serum) osmolality of 320 mOsm/kg or greater
  3. Profound dehydration
  4. Serum pH greater than 7.300
  5. Bicarbonate concentration greater than 15 mmol/l (mEq/l)
  6. Low or absent ketonuria and low or absent ketonemia
  7. Alteration of conscious level

HHS management

As for any patient with significant neurological impairment, attention to supportive measures is mandatory. The patient’s airway must be protected and oxygenation maintained.

The primary problems of dehydration and hyperglycemia are addressed with aggressive fluid resuscitation and administration of intravenous insulin. Most institutions will have protocols available to guide you.

The triggering pathologies are diverse in this group and may be of an infectious or non-infectious aetiology. The triggering cause must be sought and treated appropriately.

These patients are at high risk of deep venous thrombosis (DVT) due to profound dehydration and, usually, immobility. In the absence of contraindications, DVT prophylaxis is mandatory.

HHS and plasma sodium

The effects of HHS on plasma [Na+] are variable. The loss of free water in the urine, with no concomitant loss of Na+, may result in concentration of Na+ in the ECF with significant hypernatremia on blood testing. However, hypernatremia and raised plasma osmolality are poorly tolerated by human beings, particularly the young, and provided the affected person has access to water, increased thirst and water intake combined with an osmotic shift of water from the intracellular to the extracellular fluid compartment may on occasion result in hyponatremia in HHS. Perhaps because of the age profile of type II diabetics (impaired thirst mechanism with age) combined with the neurological consequences of raised plasma osmolality, many patients do not increase water intake in response to the raised plasma osmolality of HHS. Therefore, typically, in the elderly population we see with this condition, the loss of free water in the urine results in significant hypernatremia.

*The ‘plasma osmolality’ (aka ‘serum osmolality’) is useful in the diagnosis of several important conditions. The plasma osmolality is taken as a reflection of the osmolality of the extracellular fluid compartment.  The plasma osmolality may be measured directly by physical means or ‘calculated’ based on the level of the major osmotically active species present in the plasma. The major players determining plasma osmolality are glucose, urea and electrolytes. The concentration of glucose and urea in the plasma are measured directly. However, to calculate the plasma osmolality, the total concentration of electrolytes must be estimated as not all are measured on our standard blood tests. Na+ and K+ represent the bulk of positively charged electrolytes (cations) in the plasma. So, the sum of Na+ and K+concentrations represents an approximation of the total concentration of positively charged electrolytes present. As the plasma is electrically neutral the sum of positive and negative electrolytes are roughly equal. So, 2 x ([Na+] + [K+]) is taken as an approximation of the concentration of all electrolytes, cations and anions, present in plasma. By convention [K+] is often left out of the calculation.

Plasma osmolality is measured in the SI unit of Osmoles/kg (Osm/kg). Therefore, in parts of the world (most countries) were SI units are in regular use, the calculated plasma osmolality (in mOsm/kg) is given by 2[Na+] + [glucose] + [urea] (all concentrations in mmol/l). In the United States, SI units are not used to measure glucose and urea concentration (measured in mg/dl) so an adjustment must be made to convert their concentration to SI units.

Calculated plasma osmolality (calculated serum osmolality)

SI units: 2[Na+] + [glucose] + [urea] = calculated plasma osmolality (mOsm/kg)

US units: 2[Na+] + [glucose]/18 + [urea]/2.8 = calculated plasma osmolality (mOsm/kg)

You will appreciate that both the US and SI unit equations are an estimate and even in normal circumstances may differ significantly from the actual plasma osmolality assessed by direct physical measurement. The variation in either direction may be as much as 15 mOsm/kg (+/- 15 mOsm/kg). This is important, we will come across cases where a significant (>15 mOsm/kg) difference between measured and calculated osmolality (a significant ‘osmolar gap’) will be an important clue in the diagnosis of poisoning.

There is one final point of which you should be aware. Unlike electrolytes, urea diffuses freely across cell membranes. Changes in the concentration of urea in the plasma (extracellular fluid), therefore, result in no change in the difference between intra and extracellular osmolality and consequently induce no osmotic shift in water distribution between the intra and extracellular fluid compartments. Glucose is a special case. Glucose enters cells freely in the presence of an effective concentration of insulin. In situations characterised by loss of insulin action, glucose, therefore, become an effective osmolyte. In these situations (eg HHS), the plasma (ECF) osmolality generated by electrolytes and glucose is referred to as the ‘effective’ plasma osmolality.

SI units: 2[Na+] + [glucose] = effective plasma osmolality (mOsm/kg)

US units: 2[Na+] + [glucose]/18 = effective plasma osmolality (mOsm/kg)

Major changes in the concentration of these species in the plasma (ECF) will induce an osmotic gradient between the intra and extracellular compartments. This may result in the clinical consequences of cerebral dehydration or oedema depending on the direction of change in plasma osmolality.


A few words on ‘hypernatremia’ in clinical practice

Outside the hospital setting, hypernatremia is an unusual finding in adult patients. A rise in sodium concentration in the ECF is accompanied by a rise in plasma osmolality. This is a potent stimulus to thirst and even in the presence of significant pathology, provided the individual has access to water, increased water intake will usually reverse the rise in serum sodium. This response  is augmented by release of ADH from the pituitary gland with increased free water retention in the kidney further helping to reduce sodium to the normal range. Clinically, this situation may manifest as polyuria and polydipsia but the serum sodium will remain in the normal range or may even fall.

Therefore, our present case illustrates two key requirements which need to be present if a patient is to develop significant hypernatremia. 1) A pathological process causing loss of free water. That is fluid loss which is relatively low in sodium content 2) an altered level of access to water. In the case illustrated here, the patient is an elderly type II diabetic. In elderly type II diabetics, intercurrent illnesses (eg, UTI, silent MI) can result in a marked rise in plasma glucose levels. Glucose is freely filtered at the renal glomerulus and will act as an osmotic diuretic pulling water into the collecting tubules. The loss of water independent of sodium may result in dehydration with a consequent hypernatremia. The combination of cellular dehydration and intercurrent illness may also result in confusion which combined with an age dependent decline in thirst mechanisms means that the elderly person fails to increase water intake. The osmotic diuresis continues and in this way a vicious cycle of declining conscious level and worsening hypernatraemia develops.

In hospital medicine the combination of water loss and loss of control over fluid intake is relatively common. For example, in the unconscious patient with severe fluid losses we may supply fluids with a higher Na+ content than is required. In addition, many intravenous drug preparations contain a high Na+ content. For these reasons, it is not uncommon to see iatrogenic hypernatremia develop in the severely ill patient requiring IV fluid replacement (Table 1).

Diabetes insipidus (DI) represents a failure to produce ADH in the pituitary gland (central DI) or a failure of the kidney to respond to this hormone. In the absence of this input there is an increase in free water output from the kidney. However, the thirst mechanism remains intact. Therefore, increased water intake may keep the serum Na+ concentration in the upper limit of the normal range or just above the normal range. Again, the diagnosis may be suggested by a history of polyuria/polydipsia in the correct clinical context.

Hypernatremia: causes

Free water loss and suppression of/failure to act on thirst

Hyperosmolar hyperglycemic state (HHS)

Iatrogenic: inappropriately high Na+ administration

Burns

Postop

ICU

Free water loss (Na+ may be normal or slightly above normal range)

Diabetes Insipidus

Central

Head trauma

Intracranial inflammatory pathology

drugs

Nephrogenic

Hyperalcaemia

Lithium