Info
🌱 來自: sxtodx
an acid-base abnormality🚧 施工中
an acid-base abnormality
Scott D. C. Stern, MD
CHIEF COMPLAINT
PATIENT
Mr. L is a 42-year-old man with type 1 diabetes mellitus (DM) who complains of weakness, anorexia, and vomiting. Laboratory studies demonstrate a HCO3− of 6 mEq/L.
His very low HCO3− suggests a significant acid-base abnormality. In addition to evaluating his abdominal pain, exploring his acid-base disorder is critical.
What is the differential diagnosis of acid-base disorders? How would you frame the differential?
CONSTRUCTING A DIFFERENTIAL DIAGNOSIS
The differential diagnosis of acid-base disorders is extensive (Table 4-1) but can easily be organized into 4 distinct subsets by first determining whether the primary disorder is a (1) metabolic acidosis, (2) metabolic alkalosis, (3) respiratory acidosis, or (4) respiratory alkalosis. The key pivotal feature that allows the clinician to narrow the differential to 1 of these subsets is to first evaluate the pH and then the HCO3− and PaCO2.
Table 4-1. Differential diagnosis of primary acid-base disorders.
Step 1: Determine Whether the Primary Disorder is an Acidosis or Alkalosis by Reviewing the pH1.
A. pH < 7.4 indicates the primary disorder is an acidosis.
B. pH > 7.4 indicates the primary disorder is an alkalosis.
Step 2: Determine Whether the Primary Acidosis or Alkalosis is Metabolic or Respiratory by Reviewing the HCO3− and PaCO2
A. Recall that CO2 + H2O ⇔ H2CO3 ⇔ HCO3− + H+; therefore
B. PaCO2 changes drive pH as follows:
1. Increased PaCO2 drives reaction to right: This increases H+ which lowers pH, resulting in a respiratory acidosis.
2. Decreased PaCO2 drives reaction to left: This decreases H+ which raises pH, resulting in a respiratory alkalosis.
C. HCO3− changes drive pH as follows:
1. Increasing HCO3− drives the reaction to left: This consumes H+ which raises the pH, resulting in a metabolic alkalosis.
2. Decreasing HCO3− drives the reaction to the right: This increases H+ which lowers the pH, resulting in a metabolic acidosis. This occurs in 2 situations:
a. Processes that produce H+ ion (and consume HCO3−) (ie, ketoacidosis, lactic acidosis)
b. Processes that lose HCO3− (ie, diarrhea)
D. For acidosis (pH < 7.4)
1. HCO3− < 24 mEq/L: The primary disorder is a metabolic acidosis
2. PaCO2 > 40 mm Hg: The primary disorder is a respiratory acidosis
E. For alkalosis (pH > 7.4)
1. HCO3− > 24 mEq/L: The primary disorder is a metabolic alkalosis
2. PaCO2 < 40 mm Hg: The primary disorder is a respiratory alkalosis
This is summarized in Table 4-2.
Table 4-2. Identifying the primary disorder in patients with acid-base abnormalities.
Step 3: Narrow the Differential Diagnoses of Metabolic Acidosis by Calculating the Anion Gap
The differential diagnosis for metabolic acidosis is extensive but can be narrowed based on whether the anion gap is normal or elevated. The anion gap is an estimate of the unmeasured anions. As noted above, metabolic acidoses may be caused by processes that either (1) produce acid (ie, ketoacids, lactic acid, sulfates, phosphates, or other organic acids), or by (2) processes that lose HCO3− in the urine or stool (ie, diarrhea). Processes that produce acid (eg, ketoacidosis) also produce their associated unmeasured anions, which accumulate, resulting in an increased anion gap. Therefore, an elevated anion gap suggests one of those processes is the cause of the metabolic acidosis. On the other hand, processes that lose HCO3− do not generate unmeasured anions and the anion gap remains normal. Therefore, this simple calculation can help focus the differential diagnosis (Table 4-1).
A. Anion gap = Na+ – (HCO3− + Cl−)
B. 12 ± 4 is often cited as an ideal cutoff, although in some institutions, a normal anion gap is only 7–9 mEq/L.
C. The normal anion gap is affected by the serum albumin level.
1. Albumin is negatively charged so that lower serum albumin levels are associated with a lower anion gap.
2. The normal anion gap is 2.5 mEq/L lower, for every 1 g/dL drop in the serum albumin (below 4.4 g/dL).
3. The reference range at the institution performing the tests should be used.
D. An increased anion gap suggests that an anion gap metabolic acidosis is present.
Step 4: Explore the Differential Diagnoses of the Primary Disorder
After identifying the primary disorder as a metabolic or respiratory acidosis or alkalosis the differential diagnosis should be explored, looking for risk factors, associated signs or symptoms of each of the possible diagnoses (Table 4-1). This information allows the clinician to rank the differential diagnosis and then determine the appropriate testing strategy.
Step 5: Diagnose Primary Disorder
Synthesize the clinical and laboratory information to arrive at a diagnosis of the primary acid-base disorder.
Step 6: Check for Additional Disorders
Unfortunately, many patients have more than one simultaneous acid-base disturbance. Two final steps help clinicians recognize these situations.
Step 6A: Calculate Anion Gap (Even in Patients Without Acidosis) to Uncover Unexpected Anion Gap Metabolic Acidosis
Patients may have a simultaneous metabolic alkalosis (raising the HCO3−) and an anion gap acidosis (lowering the HCO3−). Depending on which is more severe, the HCO3− can be low, normal, or high. If the HCO3− is normal or high, it is easy to overlook the anion gap metabolic acidosis (which may be quite important.) Since the anion gap would remain elevated if an anion gap metabolic acidosis is present, a quick check of the anion gap can be an important clue to an otherwise unnoticed metabolic acidosis.
Always check the anion gap. An elevated gap suggests an anion gap metabolic acidosis even when the HCO3− is above normal.
Step 6B: Calculate Whether Compensation Is Appropriate
A. The acid-base system attempts to maintain homeostasis. Alterations in the respiratory or metabolic system trigger compensatory changes in the other system to minimize the change of pH. For instance, in metabolic acidosis, the respiratory system hyperventilates to lower the PaCO2 and thereby create a respiratory alkalosis that returns the pH closer to normal (but never quite to normal). Table 4-3 illustrates the direction of the compensatory change in acid-base disorders.
Table 4-3. Expected compensation in acid-base disorders.
B. Formulas predict the expected change in PaCO2 to compensate for metabolic processes and the expected change in HCO3− to compensate for respiratory processes (Table 4-4).
Table 4-4. Compensation in acid-base disorders.1,2
C. Compensation that is greater or less than expected suggests that an additional acid-base abnormality is present not just compensation.
D. If an additional process is implicated, the differential diagnosis for that additional disorder should be explored.
Step 7: Reach Final Diagnosis
Figure 4-1 outlines the stepwise approach to acid-base disorders.
Figure 4-1. Stepwise approach to the diagnosis of acid-base disorders.
Mr. L reports that he has had diabetes since he was 10 years old. His diabetes has been complicated by peripheral vascular disease requiring a below the knee amputation and laser surgeries for retinopathy. Two days ago, he began experiencing nausea and some vomiting. He continued to take his insulin. Physical exam reveals supine BP of 90/50 mm Hg and pulse of 100 bpm. Upon standing, his vital signs are BP, 60/30 mm Hg; pulse, 150 bpm; RR, 24 breaths per minute; and temperature, 37.0°C. Retinal exam reveals dot-blot hemorrhages and multiple laser scars. Lungs are clear to percussion and auscultation. Cardiac exam reveals a regular rate and rhythm with a grade I/VI systolic murmur at the upper left sternal border. Abdominal exam is soft and nontender. Stool is guaiac-negative. Lab studies reveal Na+, 138 mEq/L; K+, 6.2 mEq/L; HCO3−, 6 mEq/L; Cl−, 100 mEq/L; BUN, 40 mg/dL; creatinine, 1.8 mg/dL; glucose, 389 mg/dL; WBC, 10,500/mcL; HCT, 42%; ALT, AST, and lipase are normal.
At this point what is the leading hypothesis, what are the active alternatives, and is there a must not miss diagnosis? Given this differential diagnosis, what tests should be ordered?
RANKING THE DIFFERENTIAL DIAGNOSIS
Step 1: Determine Whether the Primary Disorder Is an Acidosis or Alkalosis by Reviewing the pH
Although an arterial pH has not yet been obtained, the patient’s very low HCO3− strongly suggests a metabolic acidosis. Commonly, sick patients are discovered to have a metabolic acidosis when a basic metabolic panel reveals a very low serum HCO3−. Although theoretically, compensation for a respiratory alkalosis could also cause a slight reduction in HCO3−, a HCO3− in this range is almost never seen unless there is, in fact, a primary metabolic acidosis. Nonetheless, an arterial blood gas (ABG) measurement can document the acidosis and evaluate respiratory compensation.
ABG: pH of 7.15, PaO2 of 80 mm Hg, and PaCO2 of 20 mm Hg.
The low pH confirms that the primary disorder is an acidosis.
Step 2: Determine Whether the Primary Acidosis or Alkalosis Is Metabolic or Respiratory by Reviewing the HCO3− and PaCO2
HCO3− = 6 mEq/L and PaCO2 = 20 mm Hg.
Both the HCO3− and PaCO2 are low. Since only a low HCO3− would create an acidosis the primary disorder is a metabolic acidosis. (A low PaCO2 drives the pH up [see above].)
Step 3: For Patients With Metabolic Acidoses, Narrow the Differential Diagnoses by Calculating the Anion Gap
Anion gap = Na+ – (HCO3− + Cl−) = 138 − (6 + 100) = 32 (Normal = 12 ± 4)
Clearly, the primary disorder is an anion gap metabolic acidosis. By referring to Table 4-1, the differential diagnosis can be narrowed to the remaining possibilities of diabetic ketoacidosis (DKA), other ketoacidoses, lactic acidosis, uremia, or toxin.
Step 4: Explore the Differential Diagnoses of the Primary Disorder
The history of childhood-onset DM strongly suggests insulin-dependent DM. This form of DM is associated with total or near total insulin deficiency increasing the risk of DKA. This is the leading hypothesis. Active alternative hypotheses include other ketoacidoses (starvation, alcohol) and uremia from chronic kidney disease (potentially secondary to long-standing diabetes). Finally, lactic acidosis (from hypoxemia or shock) is a “must not miss diagnosis” that should always be considered in sick patients with metabolic acidosis. Table 4-5 ranks the differential diagnoses considering the available demographic information, risk factors, symptoms, and signs.
Table 4-5. Diagnostic hypotheses for Mr. L.
Is the clinical information sufficient to make a diagnosis? If not, what other information do you need?
Leading Hypothesis: DKA
Textbook Presentation
DKA often begins with an acute illness (ie, pneumonia, urinary tract infection, myocardial infarction [MI]) or nonadherence to insulin in a patient with type 1 DM. Patients often complain of symptoms related to hyperglycemia (polyuria, polydipsia, and polyphagia) and to the precipitating illness (eg, fever, cough, dysuria, chest pain). Nonspecific complaints are common (nausea, vomiting, abdominal pain, and weakness). Patients are profoundly dehydrated and exhibit orthostatic changes or frank hypotension. Confusion, lethargy, and coma may occur secondary to dehydration, hyperglycemia, acidosis, or the underlying precipitating event.
Disease Highlights
A. Occurs primarily in patients with complete or near complete insulin deficiency
1. Type 1 autoimmune insulin-dependent DM
2. Type 2 DM can cause DKA
a. Type 2 DM accounts for 12–47% DKA episodes
b. DKA complicating type 2 DM is more frequently found in diabetic blacks and Hispanics than non-Hispanic whites. (Up to 47% of Hispanic diabetics with DKA had type 2 DM.)
c. Many such patients can eventually be treated with oral hypoglycemics without insulin after a short period of insulin therapy.
Many cases of DKA occur in type 2 DM.
3. Diabetes secondary to severe chronic pancreatitis and near complete islet cell obliteration
B. Incidence is 4.6–8.0 cases/1000 person years in patients with DM.
C. Precipitated by low insulin levels or illnesses that increase hormones counterregulatory to insulin (cortisol, epinephrine, glucagon, and growth hormone), or both.
1. The precipitant is the most frequent cause of mortality in DKA.
2. Most common precipitants
a. Infection
(1) Urinary tract infections and pneumonia are most common.
(2) Patients may be afebrile.
b. Discontinuation of insulin or oral hypoglycemics
c. New-onset type 1 DM
3. Other common precipitants include:
a. Other infections
b. MI
c. Cerebrovascular accident
d. Acute pancreatitis
e. Pulmonary embolism
f. Gastrointestinal hemorrhage
g. Severe emotional stress
4. Medications and drugs implicated in DKA include:
a. Failure of insulin pumps, and use of corticosteroids, thiazides, cocaine, antipsychotics.
b. SGLT2 inhibitors used to treat diabetes have also caused DKA in some patients.
(1) The mechanism includes decreased release of insulin (due to the lower glucose levels), direct stimulation of glucagon release, and ketogenic effects).
(2) Risk may be accentuated in patients undergoing stressful events, such as surgery.
(3) May be appropriate to avoid with concurrent insulin use
(4) Risk may be higher in patients with lower insulin levels, such as LADA (latent autoimmune diabetes of the adult).
(5) Due to glucosuria, DKA in these patients is often associated with near normal glucose levels (mean 265 mg/dL). Therefore, a high index of suspicion of DKA must be maintained.
D. Pathogenesis of DKA: (Figure 4-2). The marked decrease in insulin levels together with an increase in counterregulatory hormones lead to the following events:
Figure 4-2. Pathogenesis of diabetic ketoacidosis (DKA).
1. Hyperglycemia: Caused by
a. Marked insulin deficiency reduces cellular uptake of glucose
b. Increased hepatic glycogenolysis and gluconeogenesis
c. Oliguria: Hyperglycemia leads to an osmotic diuresis (polyuria) and dehydration. Oliguria eventually develops, which limits urinary glucose loss and thereby aggravates hyperglycemia.
2. Ketoacidosis
a. Marked insulin deficiency increases glucagon which, in turn, increases acetyl CoA production within liver.
b. Massive production of acetyl CoA overwhelms Krebs cycle resulting in ketone production and ketonemia (primarily beta hydroxybutyric acid and to a lesser extent acetoacetic acid).
c. Ketonemia leads to anion gap metabolic acidosis.
3. Volume depletion: Ketonemia and hyperglycemia cause an osmotic diuresis, which results in profound dehydration and typical fluid losses of 3–6 L.
4. Potassium loss
a. The osmotic diuresis also causes significant potassium losses.
b. Dehydration-induced hyperaldosteronism aggravates potassium loss.
c. Typical potassium deficit is 3–5 mEq/kg body weight.
5. Hyperkalemia
a. Despite the total body potassium deficit, hyperkalemia is frequent.
b. The etiology is multifactorial.
(1) Insulin normally drives glucose and potassium into the cells. Insulin deficiency decreases cellular uptake and causes hyperkalemia.
(2) Plasma hypertonicity drives water and potassium out of the cells and into the intravascular compartment accentuating the hyperkalemia.
6. Hyponatremia
a. As noted above, the hyperglycemia and ketonemia create an osmotic diuresis and free water loss
b. Despite this free water loss, many patients with DKA have hyponatremia.
c. Hyponatremia occurs because the hyperglycemia forces an osmotic efflux of water from the intracellular space into the extracellular space, diluting the serum sodium.
d. The elevated serum osmolality also stimulates antidiuretic hormone (ADH) release, which further aggravates the hyponatremia.
e. With treatment, the glucose (and water) shifts intracellularly, increasing the serum sodium which may actually rise above normal due to the water deficit.
f. Correction factors help predict the serum sodium concentration after the hyperglycemia is treated.
g. Experiments suggest that the sodium concentration will increase by 2.4 mEq/L for every 100 mg/dL that the glucose falls with treatment. (See Pseudohyponatremia in Chapter 24.)
E. Mortality rate of DKA is 5–15%. Risk factors for death include:
1. Severe coexistent disease (adjusted OR 16.3)
2. pH < 7.0 at presentation (adjusted OR 8.7)
3. Depressed mental status after 24 hours (adjusted OR 8.6)
4. Glucose > 300 mg/dL after 12 hours (adjusted OR 8.3)
5. > 50 units of insulin required in first 12 hours (adjusted OR 7.9)
6. Fever (axillary temperature ≥ 38.0°C) after 24 hours (adjusted OR 5.8)
7. Increasing age
a. Mortality rate < 1.25% in persons younger than 55 years
b. Mortality rate 11.8% in persons older than 55 years
Evidence-Based Diagnosis
A. American Diabetes Association (ADA) Diagnostic criteria
1. Glucose > 250 mg/dL. (Rarely patients with DKA have near normal blood glucoses [euglycemic DKA]. This is more common in pregnant patients and patients taking SGLT2 inhibitors.)
2. pH ≤ 7.3
3. HCO3− ≤ 18 mEq/L
4. Positive serum ketones
5. Anion gap > 10 mEq/L
B. Signs and symptoms
1. Polyuria and increased thirst are common.
2. Lethargy and obtundation may be seen in patients with markedly increased effective osmolality (> 320 mOsm/L), especially in patients with significant acidemia.
a. Effective osmolality can be calculated:
(1) (2 × Na+) + Glucose/18
(2) For example, a patient with a Na+ of 140 mEq/L and a glucose of 720 mg/dL has an effective osmolality of 320 mOsm/L
b. Consider neurologic insult (eg, cerebrovascular accident, drug intoxication) if neurologic changes are present in patients with a serum osmolality < 320 mOsm/L or if the neurologic abnormalities fail to resolve with therapy.
3. Abdominal pain
a. Very common in DKA
b. May be caused by the DKA or another intra-abdominal process (ie, appendicitis, pancreatitis, cholecystitis, abscess)
c. More likely to be due to an intra-abdominal process rather than DKA, if the DKA is mild (the HCO3− is closer to normal)
d. The frequency of abdominal pain increases as the severity of DKA increases (Table 4-6).
Table 4-6. Frequency and etiology of abdominal pain in patients with DKA.
Abdominal pain and delirium may be complications or causes of DKA (ie, cholecystitis, pancreatitis, or cerebrovascular accident). Always search for the etiology of abdominal pain or delirium in patients with DKA, especially if it occurs in patients with mild acidosis (HCO3− > 10 mEq/L), or serum osmolality < 320 mOsm/L.
4. Nausea and vomiting are common and nonspecific.
C. Hyperglycemia
1. Glucose level is variable.
2. 15% of patients with DKA have glucose levels < 350 mg/dL, particularly in
a. Pregnant patients
b. Patients with poor oral intake
c. Patients taking SLGT2 inhibitors
d. Patients given insulin in route to the hospital
3. An isolated glucose > 250 mg/dL has poor specificity for DKA (11%).
D. Ketones
1. 3 ketones: beta-hydroxybutyrate, acetoacetate, acetone
2. Beta-hydroxybutyrate is the predominant ketone in severe DKA.
a. Test of choice to evaluate DKA.
b. Sensitivity, 98%; specificity, 79–85%; LR+ 6.5; LR–, 0.02 (cutoff > 1.5 mmol/L)
c. Compared to urine ketones, beta-hydroxybutyrate testing reduces frequency of hospitalization, shortens time to recovery, and lowers costs
3. Standard ketone test (with nitroprusside) is an older test but insensitive for beta-hydroxybutyrate
4. Urine ketones are sensitive for DKA (98%) but not specific (35–69%). Blood measurements are preferred.
E. Anion gap
1. Anion gap is elevated in most patients with DKA (even when nitroprusside reaction is negative).
2. In patients evaluated in the emergency department with glucose > 250 mg/dL, the anion gap is 84–90% sensitive and 85–99% specific; LR+, 6–84; LR–, 0.11–0.16.
3. If anion gap is elevated and ketones are negative, beta-hydroxybutyrate measurements should be measured. If beta-hydroxybutyrate measurements are not available (or negative), lactic acid should be measured to rule out lactic acidosis.
F. Nonspecific findings
1. Amylase: Nonspecific elevations in amylase are common.
2. Leukocytosis
a. Mild leukocytosis (10,000–15,000 cells/mcL) is common and may occur secondary to stress or infection.
b. One study documented higher WBCs in DKA patients with major infection than in patients without infection (17,900/mcL vs 13,700/mcL).
c. Band counts were also higher in patients with infection (23% vs 6%).
Treatment
A. Treatment of DKA includes the following principles outlined in detail below:
1. Initial evaluation and frequent monitoring
2. Detection and treatment of the underlying precipitant
The most common cause of death in patients with DKA is the underlying precipitant. It must be discovered and treated.
3. Fluid resuscitation
4. Insulin
5. Potassium replacement
B. Initial evaluation and monitoring
1. Vital signs and orthostatic vital signs should be measured.
2. Check electrolytes, glucose, serum ketones, serum beta-hydroxybutyric acid, serum lactate, ABG, anion gap, plasma osmolality, blood urea nitrogen (BUN), and creatinine.
3. Serum creatinine may be artificially elevated due to interference of assay by ketones.
4. The serum glucose should be checked hourly and the electrolytes should be measured frequently (every 2–4 hours) and the anion gap calculated.
C. Detection and treatment of the underlying precipitant
1. A careful physical exam, including examination of the feet should look for infection or other underlying precipitant.
2. Urinalysis and urine culture, chest radiograph, CBC with differential, ECG, and troponin levels are appropriate.
3. Human chorionic gonadotropin beta subunit should be measured in women of childbearing age.
4. Other tests as clinically indicated (blood cultures, lipase, etc.)
D. Fluid resuscitation
1. Evaluate dehydration: Check BP, orthostatic BP and pulse, monitor hourly urinary output
2. IV normal saline 1–2 L bolus initially.
a. Larger volumes (1–2 L) are useful for patients with significant hypotension.
b. Smaller volumes (500 mL) may allow for more rapid correction of acidosis in patients without marked volume depletion.
c. Reevaluate patients after each liter by rechecking BP, orthostatic BP and pulse, urinary output, cardiac and pulmonary exams. Repeat boluses until hypotension and oliguria resolve.
d. Once hypotension resolves, decrease normal saline to 500 mL/h for 4 hours and then 250 mL/h for 4 hours.
3. If patients become hypernatremic with therapy (see above), normal saline should be switched to 0.45% (after their volume is restored) to correct the free water deficit.
E. Insulin
1. The ADA recommends an IV bolus of regular insulin (0.1 units/kg) followed by IV regular insulin at 0.1 units/kg/h. Alternatively, the bolus may be omitted and the insulin initiated at 0.14 units/kg/h. If glucose fails to fall by ≥ 10% in first hour, adjust insulin therapy.
2. Marked hypokalemia (< 3.3 mEq/L) should be excluded before insulin therapy is administered (see below).
3. Administer in monitored setting
4. Monitor glucose levels hourly: Target reduction 75–90 mg/dL/h and adjust insulin dose accordingly.
5. The ADA recommends continued IV insulin until glucose < 200 mg/dL and 2 of the following criteria are met: anion gap ≤ 12, serum HCO3− is ≥ 15 mEq/L, and the venous pH > 7.3.
a. Premature discontinuation of IV insulin may result in rebound ketoacidosis.
b. If patient’s glucose normalizes (< 200 mmol/day) before the anion gap normalizes and before the HCO3− is ≥ 18 mEq/L, reduce (but do not stop) the insulin infusion and add glucose (D5W or D10W) to the IV to prevent hypoglycemia.
c. Patients should receive their first dose of SQ insulin 1–2 hours before IV insulin is discontinued in order to prevent an insulin-free window and recurrent ketoacidosis.
In DKA, it is important to continue IV insulin until the anion gap returns to normal. Administer glucose as necessary to prevent hypoglycemia.
F. Potassium replacement
1. Insulin shifts potassium back into the intracellular compartment. Fluid resuscitation and correction of the acidosis further lower the serum potassium concentration.
2. Despite hyperkalemia on presentation, profound and potentially life-threatening hypokalemia is a common complication of therapy and often develops within the first few hours.
3. Patients with normal or near normal serum potassium concentrations on admission are, therefore, at higher risk of life-threatening hypokalemia with treatment and should have cardiac monitoring due to the risk of arrhythmias.
4. Potassium levels should be monitored hourly, and replacement should be initiated when urinary output resumes and potassium is < 5.0–5.2 mEq/L.
5. Potassium therapy should be initiated immediately in patients with hypokalemia. In addition, insulin therapy should be delayed until the serum potassium is > 3.3 mEq/L.
G. HCO3− therapy
1. Use is controversial; if used, monitor patients for hypokalemia.
2. HCO3− has not been shown to improve outcomes in patients with a serum pH > 6.9. It may also paradoxically lower CNS pH.
3. The ADA recommends HCO3− therapy in patients with a pH < 6.9.
H. Phosphate therapy
1. Dramatic falls in serum phosphate are common during treatment.
2. Replacement should be considered in patients with marked hypophosphatemia (< 1.0 mg/dL) or with respiratory depression, cardiac dysfunction, or anemia.
Careful, frequent observation and evaluation of patients with DKA is critical.
MAKING A DIAGNOSIS
Have you crossed a diagnostic threshold for the leading hypothesis, DKA? Have you ruled out the active alternatives uremia, starvation ketosis, alcoholic ketoacidosis, or lactic acidosis? Do other tests need to be done to exclude the alternative diagnoses?
Alternative Diagnosis: Uremic Acidosis
Textbook Presentation
Typically, patients with chronic kidney disease have low HCO3− levels, high creatinine levels (often > 4–5 mg/dL), and elevated BUN and phosphate levels. Patients often complain of a variety of constitutional symptoms secondary to their kidney disease, including fatigue, nausea, vomiting, anorexia, and pruritus.
Disease Highlights
A. Pathophysiology
1. Each day, ingested nonvolatile acids neutralize HCO3−.
2. In health, the kidneys regenerate the HCO3− and maintain the acid-base equilibrium.
3. Kidney impairment results in failed HCO3− regeneration and a metabolic acidosis.
B. Acidosis in patients with kidney disease may be of the anion gap type or nonanion gap type.
1. In early kidney disease, ammonia-genesis is impaired, resulting in reduced acid secretion and a nonanion gap metabolic acidosis.
2. In more advanced chronic kidney disease, the kidney remains unable to excrete the daily acid load and also becomes unable to excrete anions such as sulfates, phosphates, and urate. Therefore, an anion gap acidosis develops. HCO3− levels stabilize between 12 mEq/L and 20 mEq/L.
C. The acidosis has several adverse effects.
1. Increased calcium loss from bone
2. Increased skeletal muscle breakdown
Treatment
A. NaHCO3− replacement
B. Hemodialysis
Alternative Diagnosis: Starvation Ketosis
Typically, starvation ketosis occurs in patients with diminished carbohydrate intake. Ketosis is usually mild (HCO3− ≥ 14 mEq/L) and serum glucose is usually normal. Serum pH is usually normal.
Alternative Diagnosis: Alcoholic Ketoacidosis
Alcoholic ketoacidosis usually occurs in advanced alcoholism when the majority of calories come from alcohol. Ketoacidosis develops due to the combined effects of inadequate carbohydrate intake, ethanol conversion to acetic acid and stimulated lipolysis. Ketoacidosis may be precipitated by decreased intake, pancreatitis, gastrointestinal bleeding, or infection and may be profound. The plasma glucose level is typically normal to low. (Significant elevations suggest concomitant DKA.) It is important to consider other causes of metabolic acidosis in alcoholic patients with acidosis. First, patients with alcoholic ketoacidosis often have concomitant lactic acidosis. Shock and hypoxia should be carefully considered. Lactic acidosis may also occur due to an increase in NADH levels and can be particularly severe in patients with thiamine deficiency. Second, toxic ingestions (methanol, ethylene glycol, or salicylate) should also be considered, especially in patients with a large osmolar gap. (The osmolar gap = measured serum osmolality – calculated serum osmolality. The calculated osmolality = (2 × Na+) + Glucose (mg/dL)/18 + BUN (mg/dL)/2.8) + ETOH (mg/dL)/3.7. A normal osmolar gap < 10 mOsm/kg.) The treatment for alcoholic ketoacidosis should include IV thiamine prior to IV glucose to avoid precipitating Wernicke encephalopathy or Korsakoff syndrome.
Mr. L’s serum ketones are large. He denies any history of heavy alcohol use or abuse. The serum lactate level is 1 mEq/L (nl 0.5–1.5 mEq/L).
Step 5: Diagnose Primary Disorder
The high serum ketones confirm ketoacidosis as the primary metabolic disturbance and the high glucose and diabetic history clearly suggest DKA as the cause of the primary acid-base abnormality. The high glucose and profound acidosis are not consistent with starvation ketoacidosis and the absence of a significant alcohol history argues against alcoholic ketoacidosis. The normal lactate effectively rules out lactic acidosis, and uremic acidosis is very unlikely with mild kidney disease (creatinine = 1.8).
Step 6: Check for Additional Disorders
Step 6A: Check Anion Gap
Already completed (see above)
Step 6B: Calculate Whether Compensation Is Appropriate
As shown in Table 4-4 the expected drop in PaCO2 to compensate for a metabolic acidosis is 1.2 mm Hg per 1 mEq/L fall in HCO3−. The patient’s HCO3− is 6 mEq/L (nl is 24 mEq/L), which is an 18 mEq/L decrement. The PaCO2 should fall by 1.2 × 18 = 21.6 mm Hg. Since the normal PaCO2 is approximately 40 mm Hg, the PaCO2 would be expected to be approximately 40 − 21.6 ≈ 18. The actual PaCO2 (20 mm Hg) is close to this predicted value suggesting that respiratory compensation is indeed appropriate.
Step 7: Reach Final Diagnosis
Therefore, Mr. L is suffering from an anion gap metabolic acidosis secondary to DKA with appropriate respiratory compensation.
CASE RESOLUTION
Evaluation and treatment identify the precipitant of DKA and treats the acidosis, hyperglycemia, and profound dehydration.
Mr. L confirms he has been taking his insulin. He reports no fever, rigors, dysuria, cough, shortness of breath, diarrhea, or abdominal pain. Urinalysis, chest radiograph, and lipase were sent to search for the precipitating event. All of the results were normal. An ECG revealed T wave inversion in leads V1–V4, suggesting anterior myocardial ischemia. Troponin T levels were elevated consistent with an acute MI (believed to be the precipitant of his DKA). He was transferred to the ICU for monitoring. He received fluid resuscitation, IV insulin until his ketoacidosis resolved, and supplemental potassium (when his potassium fell below 5.3 mEq/L). His MI was treated with beta-blockers and aspirin. Subsequent cardiac catheterization revealed triple vessel disease. After stabilization, he underwent coronary artery bypass grafting and did well.
CHIEF COMPLAINT
PATIENT
Ms. S is a 32-year-old woman who complains of nausea and vomiting. She reports that she felt well until 5 days ago when she noticed urinary frequency and burning on urination. She increased her intake of fluids and cranberry juice but noticed some increasing right back pain 2 days ago. Yesterday, she felt warm and noticed that she had a fever of 38.8°C and teeth-chattering chills. Subsequently, she has been unable to keep down any food or liquids and has persistent nausea and vomiting. She feels weak and dizzy. Physical exam: supine BP, 95/62 mm Hg; pulse, 120 bpm; temperature, 38.9°C; RR, 24 breaths per minute. On standing, her BP falls to 72/40 mm Hg with a pulse of 145 bpm. Cardiac and pulmonary exams are notable only for the tachycardia. She has 2+ right costovertebral angle tenderness. Abdominal exam is soft without rebound, guarding, or focal tenderness. Initial laboratory results include Na+, 138 mEq/L; K+, 3.8 mEq/L; HCO3−, 14 mEq/L; Cl−, 102 mEq/L; BUN, 30 mg/dL; creatinine, 1.2 mg/dL; glucose, 90 mg/dL.
The list of symptoms and signs can be grouped together to make evaluation more organized: (1) dysuria, urinary frequency, flank pain, fever, and chills, (2) nausea and vomiting, (3) hypotension and tachycardia, and (4) low serum HCO3−. In addition to investigating the probable urinary tract infection, it is critical to determine the nature of the acid-base abnormality.
At this point, what is the leading hypothesis, what are the active alternatives, and is there a must not miss diagnosis? Given this differential diagnosis, what tests should be ordered?
RANKING THE DIFFERENTIAL DIAGNOSIS
Step 1: Determine Whether the Primary Disorder Is an Acidosis or Alkalosis by Reviewing the pH
Similar to the first case, Ms. S has a low serum HCO3− suggesting a metabolic acidosis. On the other hand, it is conceivable (but unlikely), that the low serum HCO3− could occur in compensation for a profound respiratory alkalosis. An ABG can determine the primary disorder and assess compensation.
An ABG reveals a pH of 7.29, PaCO2 of 30 mm Hg, PaO2 of 90 mm Hg.
The low pH on the ABG confirms the primary process is an acidosis.
Step 2: Determine Whether the Primary Acidosis or Alkalosis is Metabolic or Respiratory by Reviewing the HCO3− and PaCO2
Ms. S’s serum HCO3− is 14 mEq/L, her PaCO2 is 30 mm Hg. Both are quite low but only the low HCO3− would create an acidosis. (A low PaCO2 would drive the pH up and cause an alkalosis.) Since her pH is low and the HCO3− is low the primary process is a metabolic acidosis.
Step 3: Narrow the Differential Diagnoses of Metabolic Acidosis by Calculating the Anion Gap
The next step in the differential diagnosis is to calculate the anion gap. Her anion gap = 138 − (102 + 14) = 22.
Clearly, Ms. S is suffering from an anion gap metabolic acidosis. This is alarming because metabolic acidosis in the face of infection suggests lactic acidosis due to severe sepsis.
Step 4: Explore the Differential Diagnoses of the Primary Disorder
The leading and must not miss hypothesis would clearly be lactic acidosis, especially given the patient’s hypotension. If confirmed, the cause of the lactic acidosis must be determined (which would most likely be sepsis for Ms. S) and then treated. Although unlikely, alternative causes of an anion gap metabolic acidosis (Table 4-1) that would be reasonable to consider include alcoholic ketoacidosis and toxin-related acidosis (including salicylates). The normal glucose and lack of history of diabetes rules out DKA, and the severity of acidosis is not consistent with starvation ketoacidosis. The normal creatinine rules out uremic acidosis. The differential diagnosis for Ms. S is listed in Table 4-7.
Table 4-7. Diagnostic hypotheses for Ms. S.
The patient denies any history of alcohol use, moonshine or antifreeze ingestion, or unusual salicylate use. Further lab studies include WBC, 18,500 cells/mcL with 62% granulocytes and 30% bands. Urinalysis reveals > 20 WBC/hpf.
Ms. S’s history does not suggest toxic ingestions and her history of fever, dysuria, and flank pain as well as leukocytosis and pyuria, clearly suggest urinary tract infection and pyelonephritis. Her teeth-chattering chills suggest bacteremia, which combined with her hypotension suggests sepsis. Sepsis can cause lactic acid production and thereby generate an anion gap metabolic acidosis.
Is the clinical information sufficient to make a diagnosis? If not, what other information do you need?
Leading Hypothesis: Lactic Acidosis
Textbook Presentation
The presentation of lactic acidosis depends on the underlying etiology. The most common causes are hypoxemia, septic shock, cardiogenic shock, or hypovolemic shock. Patients with shock usually have hypotension and tachycardia and often have impaired mentation and decreased urinary output. Patients with septic shock typically have fever and tachypnea. While patients with cardiogenic or hemorrhagic shock often have cold extremities, patients with septic shock often have warm extremities and bounding pulses after fluid resuscitation. (Pulses are bounding due to a widened pulse pressure.) See Chapter 25 for a review of sepsis.
Disease Highlights
A. Most common cause of metabolic acidosis in hospitalized adults
B. The most common causes of lactic acidosis are due to inadequate tissue oxygenation. This results in anaerobic metabolism and the production of lactic acid. Therefore, the differential diagnosis can be remembered by tracing the pathway of oxygen all the way from the environment through the blood to the cells and mitochondria. Any disease that interferes with oxygen delivery can cause lactic acidosis (Table 4-8).
Table 4-8. Differential diagnosis of lactic acidosis.
1. Low oxygen carrying capacity
a. Hypoxemia (from pulmonary or cardiac disease)
b. Severe anemia
c. Carbon monoxide poisoning (interferes with oxygen binding)
d. Methemoglobinemia
2. Inadequate tissue perfusion; causes include
a. Hypovolemic shock
b. Cardiogenic shock
c. Septic shock
d. Regional obstruction to blood flow (eg, ischemic bowel or gangrene)
3. Inadequate cellular utilization of oxygen (cyanide poisoning)
4. Occasionally, lactic acidosis develops secondary to unusually high demand exceeding oxygen supply (eg, intense exercise, seizures).
5. Less common causes include:
a. Severe liver failure
b. Malignancy
c. Thiamine deficiency
d. Certain medications (nucleoside reverse transcriptase inhibitors, linezolid, propofol, and beta-agonists.)
6. Metformin may cause lactic acidosis.
a. Risk factors include concomitant chronic kidney disease, liver disease, heart failure (HF), alcohol use, acute illness, and IV radiocontrast administration.
b. Current recommendations suggest the following:
(1) For patients with a glomerular filtration rate (GFR) of 30–45 mL/min:
(a) Do not start metformin
(b) Reduce dose by 50% in patients already tolerating metformin
(2) For patients with a GFR < 30 mL/min, metformin should not be used.
(3) Metformin should be held for 48 hours beginning at the time of administration of IV radiocontrast agents.
C. Lactate elevation is associated with a substantially increased mortality in multiple clinical scenarios including sepsis, cardiogenic shock, trauma, pulmonary embolism, and burns. The mortality rate of patients with shock and lactic acidosis is 70% compared with 25–35% in patients with shock without lactic acidosis.
D. Lactate levels can serve to risk stratify patients with suspected infection, even among those without apparent shock, presumably due to the identification of undetected hypoperfusion. The mortality in normotensive patients (systolic BP ≥ 90 mm Hg) with suspected infection and a lactate level ≥ 4.0 was 15% vs. 2.5% in those with lactate levels < 4.0 mmol/L.
Evidence-Based Diagnosis
A. Serum lactate levels are the gold standard and more sensitive and specific than an increase in the anion gap.
B. An elevated anion gap is 44–67% sensitive for lactic acidosis
C. An elevated anion gap may suggest lactic acidosis, but a normal anion gap does not exclude lactic acidosis.
The serum lactate level should be measured in critically ill patients in whom shock is suspected regardless of the anion gap.
Treatment
A. Treatment of lactic acidosis should target the underlying condition.
B. Buffering agents such as NaHCO3− do not improve hemodynamics or survival in patients with a pH of > 7.1. Although unproven, some experts recommend NaHCO3− in patients with a pH < 7.1.
MAKING A DIAGNOSIS
Have you crossed a diagnostic threshold for the leading hypothesis, lactic acidosis? Do other tests need to be done to exclude the alternative diagnoses?
Serum lactate level of 8 mEq/L (nl 0.5–1.5 mEq/L) confirms lactic acidosis. Blood cultures and urine cultures grew Escherichia coli.
Step 5: Diagnose Primary Disorder
The serum lactate confirms an anion gap metabolic acidosis due to lactic acidosis as the primary acid-base disorder. The clinical scenario and positive cultures strongly suggest that the diagnosis is lactic acidosis secondary to sepsis. Other tests are not necessary to confirm the diagnosis.
Step 6: Check for Additional Disorders
Step 6A: Check Anion Gap
Already completed (see above).
Step 6B: Calculate Whether Compensation is Appropriate
In a metabolic acidosis, the PaCO2 is expected to fall by 1.2 mm Hg per 1 mEq/L fall in HCO3− (see Table 4-4). The patient’s HCO3− is 14 mEq/L (10 mEq/L below normal). The PaCO2 should fall by 1.2 × 10 = 12. Since normal PaCO2 is approximately 40 mm Hg, we would expect the PaCO2 to be approximately 28 mm Hg (40 − 12 = 28 mm Hg). The actual PaCO2 is 30 mm Hg, quite close to the prediction. This suggests that respiratory compensation is appropriate.
Step 7: Reach Final Diagnosis
In summary, Ms. S is suffering from a lactic acidosis with appropriate respiratory compensation.
CASE RESOLUTION
Ms. S was treated with broad-spectrum antibiotics and IV fluid resuscitation. After initial stabilization, hypotension recurred and urinary output dropped. She was transferred to the ICU. Four hours later her oxygenation deteriorated and a chest film revealed a diffuse infiltrate consistent with acute respiratory distress syndrome. She was intubated and given IV fluids, norepinephrine, antibiotics, and mechanical ventilation. Over the next 24 hours, her BP stabilized and her anion gap lactic acidosis resolved. Seventy-two hours later she was extubated. She eventually made a full recovery.
CHIEF COMPLAINT
PATIENT
Mr. R is a 55-year-old man with chronic obstructive pulmonary disease (COPD) with a chief complaint of dyspnea. He reports that symptoms began 5 days ago with a cough productive of green sputum. The cough worsened, and 4 days ago he had a low-grade fever of 37.2°C. He noticed increasing shortness of breath 3 days ago. He reports that previously he was able to walk about 25 feet before becoming short of breath but now he is short of breath at rest. Last night his fever reached 38.8°C, and today his dyspnea intensified. He is unable to complete a sentence without pausing to take a breath. On physical exam, he appears older than his stated age. He is gaunt, sitting upright, breathing through pursed lips, and in obvious distress. Vital signs are temperature, 38.9°C; RR, 28 breaths per minute; BP, 110/70 mm Hg; pulse, 110 bpm. His pulsus paradox is 20 mm Hg. Lung exam reveals significant use of accessory muscles and markedly decreased breath sounds. Cardiac exam is notable only for diminished heart sounds.
Your resident is concerned about the adequacy of Mr. R.’s ventilation and suggests checking his pulse oximetry. You remind him that a pulse oximeter will not address the adequacy of the patient’s ventilation nor will it determine whether respiratory failure is present and suggest an ABG.
An ABG reveals a pH of 7.22, PaCO2 of 70 mm Hg, and PaO2 of 55 mm Hg.
Always check an ABG when the adequacy of a patient’s ventilation is a concern. Patients with adequate oxygenation may still be in respiratory failure.
Clearly, Mr. R has several problems that are easily identified, including (1) fever, cough, and history of COPD; (2) respiratory distress; and (3) acidosis. All of these problems are obviously potentially life-threatening. Furthermore, a thorough evaluation of the acidosis may shed light on the status of the other problems.
At this point, what is the leading hypothesis, what are the active alternatives, and is there a must not miss diagnosis? Given this differential diagnosis, what tests should be ordered?
RANKING THE DIFFERENTIAL DIAGNOSIS
Step 1: Determine Whether the Primary Disorder is an Acidosis or Alkalosis by Reviewing the pH
The low pH confirms the primary disorder is an acidosis.
Step 2: Determine Whether the Primary Acidosis or Alkalosis is Metabolic or Respiratory by Reviewing the HCO3− and PaCO2
PaCO2 of 70 mm Hg; Na+, 138 mEq/L; K+, 5.1 mEq/L; HCO3−, 27 mEq/L; Cl−, 102 mEq/L; BUN, 30 mg/dL; creatinine, 1.2 mg/dL.
The PaCO2 and HCO3− are both elevated. An elevated PaCO2 would lower pH and cause an acidemia (whereas an elevated HCO3− would cause alkalemia). Since the patient is acidemic, the primary process is a respiratory acidosis.
Step 3: Narrow the Differential Diagnoses of Metabolic Acidosis by Calculating the Anion Gap
This step is not relevant to this patient with a HCO3− of 27 mEq/L arguing against a metabolic acidosis. (This is supported by the normal anion gap.)
Step 4: Explore the Differential Diagnoses of the Primary Disorder
Respiratory acidosis may be caused by pulmonary diseases and a variety of neuromuscular diseases (see Table 4-1). His prior history of COPD and acute pulmonary complaints of cough and fever clearly suggest that his respiratory acidosis is due to a pulmonary process. Specifically, Mr. R’s history of very poor exercise tolerance at baseline suggests severe COPD. Such severe COPD could result in chronic carbon dioxide retention and chronic respiratory acidosis. Alternatively, a “must not miss” possibility is that his acute respiratory infection has precipitated acute respiratory failure (and acute respiratory acidosis). This is suggested by his worsening symptoms, respiratory distress, upright posture, pursed lip breathing, pulsus paradox, and decreased breath sounds. It is critical to distinguish acute respiratory acidosis from chronic respiratory acidosis because the former is more likely to progress rapidly to complete respiratory failure and respiratory arrest. Therefore, acute respiratory acidosis is both the leading hypothesis and the “must not miss” diagnosis. Table 4-9 ranks the differential diagnosis considering the available demographic information, risk factors, symptoms and signs.
Table 4-9. Diagnostic hypotheses for Mr. R.
Patients with a history of asthma or COPD should be asked about a prior history of intubation or ICU admission. Such patients are at greater risk for respiratory failure.
Is the clinical information sufficient to make a diagnosis? If not, what other information do you need?
Leading Hypothesis: Respiratory Acidosis
Textbook Presentation
The presentation of respiratory acidosis depends primarily on the underlying cause. The most common causes are severe underlying lung disease (eg, COPD, pneumonia, or pulmonary edema) and such patients are in respiratory distress. Respiratory acidosis may also present as altered mental status in patients with advanced respiratory failure and in those in whom the respiratory failure is due to CNS disorders (ie, intoxication.)
Disease Highlights
A. Insufficient ventilation results in increasing levels of PaCO2. This in turn lowers arterial pH. Renal compensation occurs over several days, with increased renal HCO3− regeneration.
B. Ventilation is assessed by measuring the arterial PaCO2 and pH. Significant hypoventilation and acidosis may occur without significant hypoxia.
C. Etiology: Although most commonly due to lung disease, respiratory acidosis may result from any disease affecting ventilation—from the brain to the alveoli (eg, narcotic overdose is an unfortunately common cause of respiratory failure and death. See differential diagnosis of acid-base disorders in Table 4-1.)
D. Manifestations are due to the primary disorder and the effects of hypercarbia on the CNS.
1. Patients are typically quite dyspneic, in distress, sitting upright, leaning forward, and anxious. The cardiac and pulmonary findings depend on the underlying etiology.
2. CNS manifestations
a. Severity depends on acuity. Patients with chronic hypercapnia have markedly fewer CNS effects than patients with acute hypercapnia.
b. Anxiety, irritability, confusion, and lethargy may be seen.
c. Headache may be prominent in the morning due to the worsening hypoventilation that occurs with sleep causing vasodilatation and increasing intracranial pressure.
d. Stupor and coma may occur when the PaCO2 is > 70–100 mm Hg.
e. Tremor, asterixis, slurred speech, and papilledema may be seen.
Evidence-Based Diagnosis
A. Since respiratory failure can be an indication for emergent mechanical ventilatory support, clinicians should have a low threshold for checking an ABG to obtain the PaCO2. This includes patients with respiratory distress, mental status changes, and hypersomnolence.
B. Respiratory failure is typically characterized by PaCO2 > 45 mm Hg, causing a respiratory acidosis.
C. However, occasionally, a normal PaCO2 also suggests respiratory failure.
1. For example, during asthma attacks, patients typically hyperventilate and present with a PaCO2 below normal. A normal PaCO2 in such a patient may reflect respiratory fatigue and herald the development of frank respiratory failure.
2. Patients with primary metabolic acidoses typically hyperventilate to compensate, lowering the PaCO2 below normal.
a. A PaCO2 of ≥ 40 mm Hg is inappropriate in such cases and suggests respiratory failure.
b. Inability to compensate (hyperventilate) during a metabolic acidosis is associated with an increased risk of requiring mechanical ventilation.
D. Pulsus paradox is an objective marker of severe respiratory distress.
1. Defined as > 10 mm Hg drop in systolic BP during inspiration
2. May be seen in patients using unusually strong inspiratory effort due to asthma, COPD, or other respiratory diseases
3. The fall in systolic BP during inspiration is caused by an exaggerated inspiratory effort, which increases the negative inspiratory intrathoracic pressure and augments venous return. This results in excessive RV filling causing the interventricular septum to bulge into the LV, limiting LV filling, LV cardiac output, and systolic BP.
4. When elevated in patients with asthma, it is highly specific for a severe attack but has poor sensitivity (Table 4-10).
Table 4-10. Pulsus paradox in severe asthma.
Treatment
A. Identify and treat underlying disease process (ie, bronchodilators for asthma, naloxone for narcotic overdose).
B. Supplemental oxygen should be given as necessary to prevent hypoxemia.
Supplemental oxygen occasionally worsens hypercapnia in patients with severe COPD, asthma, or sleep apnea but should never be withheld from hypoxic patients.
C. Mechanical ventilation with either intubation or biphasic positive airway pressure (BiPAP) is lifesaving in many patients.
1. Institution of mechanical ventilation is considered when pH < 7.1–7.25 or PaCO2 > 80–90 mm Hg or when indicated by patient symptoms.
2. In general, patients with chronic hypoventilation tolerate hypercapnia better than patients with acute hypercapnia.
D. Avoid hypokalemia and dehydration that may worsen metabolic alkalosis, raise the serum pH, and inadvertently further suppress ventilation.
Step 5: Diagnose Primary Disorder
The patient’s clinical picture and ABG clearly suggest the primary disorder is a respiratory acidosis. However, to diagnose the primary disorder, it is critical to distinguish whether this is an acute or chronic respiratory acidosis. Acute respiratory acidosis can be distinguished from chronic respiratory acidosis by evaluating the degree of metabolic compensation (provided there are no other acidoses also effecting HCO3−). Chronic respiratory acidoses are associated with more complete compensation (and higher HCO3− levels), than acute respiratory acidoses (because metabolic compensation takes time). Table 4-4 shows the formulas that can be used to calculate the HCO3− levels. In acute respiratory acidosis, the HCO3− increases by only 1 mEq/L for every 10 mm Hg increase in PaCO2 whereas in chronic respiratory acidosis, the HCO3− increases by 4 mEq/L for every 10 mm Hg increase in PaCO2. In Mr. R’s case, the PaCO2 is 70 mm Hg, up by 30 mm Hg (from a normal of 40 mm Hg), so if this were an acute respiratory acidosis, the HCO3− level would be expected to increase by only 3 mEq/L (from a normal of 24 mEq/L to 27 mEq/L). If, on the other hand, this is a chronic respiratory acidosis, an increase of 4 mEq/L of HCO3− per 10 mm Hg increase in PaCO2 would be expected. For a 30 mm Hg increase in PaCO2, the predicted increase in HCO3− would be 3 × 4 = 12 mEq/L increasing the serum HCO3− to 36 (24 + 12 mEq/L). Since Mr. R.’s HCO3− is 27 mEq/L, an increase of only 3 mEq/L from a normal baseline of 24 mEq/L, the primary disorder is an acute respiratory acidosis, an alarming diagnosis.
Step 6: Check for Additional Disorders
Step 6A: Calculate Anion Gap (Even in Patients Without Acidosis) to Uncover Unexpected Anion Gap Metabolic Acidosis
Another “must not miss” diagnosis for Mr. R would be sepsis. His symptoms of fever and cough suggest the possibility of pneumonia, which can be complicated by sepsis resulting in an anion gap metabolic lactic acidosis. Although his elevated HCO3− does not immediately suggest a metabolic acidosis from sepsis, the HCO3− may not be low if there is also a superimposed metabolic alkalosis generating HCO3−. These hidden acidoses can be discovered by evaluating the anion gap (which is usually elevated in patients with lactic acidosis) or by measuring the serum lactate level.
The anion gap = 138 − (102 + 27) = 9, and the serum lactate level is 0.8 mEq/L (nl 0.5–1.5 mEq/L).
Mr. R has a normal anion gap and normal lactate level, ruling out a coexistent hidden anion gap metabolic acidosis from sepsis.
Mr. R’s other laboratory results include a WBC, 16,500/mcL with 62% granulocytes and 10% bands. Chest radiograph reveals hyperinflated lung fields and a left lower lobe infiltrate.
Step 7: Reach Final Diagnosis
As noted above, the minimal metabolic compensation suggests that Mr. R is suffering from an acute respiratory acidosis with metabolic compensation. There is no evidence of a hidden anion gap acidosis. Therefore, Mr. R has an acute respiratory acidosis caused by pneumonia and COPD. He is at significant risk for complete respiratory failure and he is transferred to the ICU.
It is vital to distinguish acute from chronic respiratory acidoses.
CASE RESOLUTION
In the ICU, Mr. R is placed on ventilatory support with BiPAP and antibiotics. Over the next 5 days, his pneumonia improves. On day 8, BiPAP is discontinued and he is sent to the medical floors.
REVIEW OF OTHER IMPORTANT DISEASES
Renal Tubular Acidosis (RTA)
Textbook Presentation
Although there are a variety of RTAs, the most common type in adults is type IV RTA, caused most commonly by long-standing diabetes. Laboratory abnormalities include mild kidney disease, a mild nonanion gap acidosis (HCO3− ≈ 17 mEq/L) and hyperkalemia. Only the highlights of type IV RTA will be reviewed here.
Disease Highlights
A. Patients with type IV RTA have hypoaldosteronism.
B. Hypoaldosteronism decreases potassium and H+ excretion, resulting in hyperkalemia and acidosis.
C. The hyperkalemia also interferes with ammonia production (the major renal buffer) and further impairs acid secretion. The inability to excrete the daily acid load causes a nonanion gap acidosis.
D. In patients with diabetes mellitus, type IV RTA is also associated with low renin levels.
E. Etiologies of type IV RTA are numerous.
1. Diabetes with mild kidney disease is the most common.
2. Other causes include
a. Drugs
(1) Nonsteroidal anti-inflammatory drugs
(2) Angiotensin-converting enzyme inhibitors
(3) Angiotensin receptor blockers
(4) Potassium-sparing diuretics
(5) Trimethoprim
(6) Heparin
(7) Cyclosporine
b. Addison disease
c. Systemic lupus erythematosus
d. AIDS nephropathy
e. Chronic interstitial kidney disease
Treatment
Dietary potassium restriction, loop diuretics, and fludrocortisone are useful.
Metabolic Alkalosis
Textbook Presentation
The most common clinical situations that give rise to a metabolic alkalosis are recurrent vomiting or diuretic treatment. The metabolic alkalosis per se is usually asymptomatic. Muscle cramping due to coexistent hypokalemia may be seen.
Disease Highlights
A. Metabolic alkalosis develops only when there is both a source of additional HCO3− and a renal stimulus that limits its excretion.
1. Increased HCO3− production develops when H+ is lost from the gastrointestinal tract (eg, due to vomiting) or (2) lost from the genitourinary tract (eg, due to hyperaldosteronism) or (3) during administration of HCO3−. Volume contraction around a constant amount of HCO3− also serves to increase the HCO3−.
2. Decreased HCO3− excretion is most commonly caused by decreased renal perfusion. This occurs when the effective circulating volume is reduced.
a. Examples include dehydration or other pathologic states associated with decreased renal perfusion (ie, HF, nephrotic syndrome).
b. The mechanisms that interfere with HCO3− excretion are complex but include enhanced renal HCO3− reabsorption and decreased renal HCO3− secretion.
(1) Decreased effective circulating volume promotes avid Na+ absorption in the proximal tubule, which in turn facilitates HCO3− reclamation (Figure 4-3).
Figure 4-3. Reabsorption of HCO3− in hypovolemia. Hypovolemia increases reabsorption of sodium in exchange for hydrogen ion at the proximal convoluted tubule (PCT). The hydrogen ion reacts with HCO3− eventually forming CO2 which crosses the cell membrane. HCO3− is then regenerated and delivered to the bloodstream.
(2) Decreased effective circulating volume and Cl−depletion also decrease HCO3− secretion by the collecting cells, which compounds the metabolic alkalosis. This develops because HCO3− secretion occurs in exchange with Cl− reabsorption in the distal tubules (Figure 4-4). This requires Cl− delivery to the collecting tubules, which decreases both due to enhanced proximal Cl− reabsorption and gastrointestinal or diuretic-induced Cl− losses.
Figure 4-4. Chloride depletion interferes with HCO3− secretion. Distal HCO3− secretion is facilitated by Cl− delivery. Hypovolemia increases proximal NaCl reabsorption, limiting distal chloride delivery, in turn interfering with HCO3− secretion.
(3) Decreased effective circulating volume results in secondary hyperaldosteronism, which activates H+ secretion by the collecting tubule cells increasing HCO3− production which is reabsorbed into the blood.
(4) Low tubular Cl− also draws chloride into the tubular cells (from the plasma) and promotes HCO3− reabsorption.
(5) Hypokalemia is an important mechanism that promotes HCO3− reabsorption. In the collecting tubule, it stimulates potassium reabsorption in exchange for H+ secretion. HCO3− is produced and reabsorbed into the blood.
B. Pathologic states associated with metabolic alkalosis (Table 4-1)
1. Vomiting or nasogastric drainage. Pathophysiology:
a. Gastric acid production (and secretion) is matched by HCO3− production. The H+ ion enters the gastric lumen, whereas the HCO3− enters the bloodstream.
b. Dehydration decreases renal HCO3− excretion (see above).
2. Dehydration or other causes of reduced GFR (ie, HF, nephrotic syndrome)
3. Diuretics
4. Hypokalemia
5. Hyperaldosteronism
a. Adrenal adenoma
b. Licorice ingestion or chewing tobacco (Normally, a renal enzyme converts cortisol to cortisone in order to prevent cortisol from exerting a significant mineralocorticoid effect. Licorice contains the steroid glycyrrhetinic acid, which blocks this enzyme resulting in a heightened mineralocorticoid effect from endogenous cortisol.)
6. Bartter or Gitelman syndromes
7. Respiratory acidosis also promotes a compensatory metabolic alkalosis. Occasionally, rapid resolution of the respiratory failure will correct the hypercapnia, resulting in a transient inappropriate metabolic alkalosis (posthypercapnic metabolic alkalosis).
8. Milk-alkali syndrome
Treatment
A. Volume resuscitation with NaCl in patients with true volume depletion usually results in resolution.
B. Replete potassium deficiency.
C. Carbonic anhydrase inhibitors and low bicarbonate dialysis can be used in severe cases, particularly in patients with HF (and ineffective circulating volume) who cannot tolerate NaCl.
Respiratory Alkalosis
Textbook Presentation
The presentation of respiratory alkalosis depends on the underlying disorder. Most causes are associated with tachypnea, which can be dramatic or subtle.
Disease Highlights
A. Hyperventilation induces hypocapnia causing respiratory alkalosis.
B. The most common causes are pulmonary diseases, cirrhosis, fever, pain, or anxiety (Table 4-1).
C. Hypocapnia acutely reduces CNS blood flow.
D. Symptoms include paresthesias (particularly perioral), vertigo, dizziness, anxiety, hallucinations, myalgias, and symptoms reflective of the underlying disorder.
E. Adverse effects include hypokalemia, hypocalcemia, lung injury, seizures, angina, and arrhythmias.
Treatment
Therapy is directed at the underlying disorder.
Mixed Disorders and the “Delta-Delta Gap”
A. Occasionally, 2 distinct metabolic processes will be present in the same patient (eg, 2 distinct acidoses, 1 anion gap and 1 nonanion gap). Alternatively, a patient may have both a metabolic alkalosis and metabolic acidosis (eg, metabolic alkalosis develops in a patient with vomiting and dehydration; if these symptoms are prolonged sufficiently, severe dehydration, hypovolemic shock, and lactic acidosis also develop).
B. These multiple metabolic processes can be difficult to tease out.
C. One approach to this problem is to evaluate the delta-delta gap. Here the relative rise in the anion gap over and above the normal anion gap, (ΔAG, the first delta) is compared with the absolute fall in HCO3− (ΔHCO3−, the second delta2).
1. In simple anion gap acidoses, the deltas are similar (as patients create extra anions (ie, ketones) increasing the anion gap, the serum HCO3− is neutralized in equal proportions).
2. On the other hand, in a patient with both an anion gap and nonanion gap acidoses, the fall in HCO3− will be greater than the rise in the anion gap.
3. In patients with an anion gap acidosis and a metabolic alkalosis, the fall in HCO3− will be antagonized by the concomitant metabolic alkalosis whereas the anions will still accumulate. Therefore, the fall in HCO3− is less than the increase in the anion gap.
D. While occasionally useful, there are several limitations to applying the delta-delta gap.
1. The normal anion gap varies from institution to institution and with the patient’s serum albumin.
2. Even in simple anion gap acidosis, bone buffering of acid and renal excretion of anions complicate the delta-delta gap and make it difficult to interpret.
E. In simple anion gap acidosis (without concomitant metabolic alkalosis or nonanion gap acidosis) the typical ΔAG/ΔHCO3− is 1.6:1 in lactic acidosis and 1:1 in ketoacidosis.
REFERENCES
Anderson LW, Mackenhauer J, Roberts JC, Berg KM, Cocchi MN, Donnino MW. Etiology and therapeutic approach to elevated lactate levels. 2013. Mayo Clin Proc. 2013;88(10):1127–40.
Bates DW, Cook EF, Goldman L, Lee TH. Predicting bacteremia in hospitalized patients. A prospectively validated model. Ann Intern Med. 1990;113(7):495–500.
Bryne AL, Bennett M, Chatterji R, Symons R, Pace NL, Thomas PS. Peripheral venous and arterial blood gas analysis in adults: are they comparable? A systematic review and meta-analysis. Respirology. 2014;19:168–75.
Drage LA. Life-threatening rashes: dermatologic signs of four infectious diseases. Mayo Clin Proc. 1999;74(1):68–72.
Fall PJ, Szerlip HM. Lactic acidosis: from sour milk to septic shock. J Intensive Care Med. 2005;20(5):255–71.
Figge J, Jabor A, Kazda A, Fencl V. Anion gap and hypoalbuminemia. Crit Care Med. 1998;26(11):1807–10.
Howell MD, Donnino MW, Talmor D, Clardy P, Ngo L, Shapiro NI. Performance of severity of illness scoring systems in emergency department patients with infection. Acad Emerg Med. 2007;14(8):709–14.
Jaimes F, Arango C, Ruiz G et al. Predicting bacteremia at the bedside. Clin Infect Dis. 2004;38(3):357–62.
Kitabchi AE, Umpierrez GE, Murphy MB, Kreisberg RA. Hyperglycemic crises in adult patients with diabetes: a consensus statement from the American Diabetes Association. Diabetes Care. 2006;29(12):2739–48.
Leibovici L, Cohen O, Wysenbeek AJ. Occult bacterial infection in adults with unexplained fever. Validation of a diagnostic index. Arch Intern Med. 1990;150(6):1270–2.
Leibovici L, Greenshtain S, Cohen O, Mor F, Wysenbeek AJ. Bacteremia in febrile patients. A clinical model for diagnosis. Arch Intern Med. 1991;151(9):1801–6.
Levraut J, Bounatirou T, Ichai C et al. Reliability of anion gap as an indicator of blood lactate in critically ill patients. Intensive Care Med. 1997;23(4):417–22.
Mellors JW, Horwitz RI, Harvey MR, Horwitz SM. A simple index to identify occult bacterial infection in adults with acute unexplained fever. Arch Intern Med. 1987;147(4):666–71.
Naunheim R, Jang TJ, Banet G, Richmond A, McGill J. Point-of-care test identifies diabetic ketoacidosis at triage. Acad Emerg Med. 2006;13(6):683–5.
Rose BD PT. Clinical Physiology of Acid Base and Electrolyte Disorders, 5th edition. McGraw Hill; 2001.
Safdar N, Maki DG. Inflammation at the insertion site is not predictive of catheter-related bloodstream infection with short-term, noncuffed central venous catheters. Crit Care Med. 2002;30(12):2632–5.
Slovis CM, Mork VG, Slovis RJ, Bain RP. Diabetic ketoacidosis and infection: leukocyte count and differential as early predictors of serious infection. Am J Emerg Med. 1987;5(1):1–5.
Tokuda Y, Miyasato H, Stein GH, Kishaba T. The degree of chills for risk of bacteremia in acute febrile illness. Am J Med. 2005;118(12):1417.
Umpierrez G, Freire AX. Abdominal pain in patients with hyperglycemic crises. J Crit Care. 2002;17(1):63–7.
1 The pH and PaCO2 are typically measured utilizing an arterial blood gas (ABG.) Alternatively, a peripheral venous blood gas (VBG) can be used to estimate the pH. A meta-analysis suggested the VBG pH was similar but slightly lower (0.033) than the arterial pH (0.015–0.051). However, the venous PCO2 did not reliably predict the arterial PaCO2. Venous PaCO2 was higher with a wide and unpredictable range (from –10.7 to +2.4) and should not be used to evaluate acid-base disturbances.
2 ΔAG = Patients anion gap – normal anion gap; ΔHCO3− = 24 – patients’ HCO3−