Clinical Chemistry 45: 1323-1330, 1999;
(Clinical Chemistry. 1999;45:1323-1330.)
© 1999 American Association for Clinical Chemistry, Inc.
Maximizing Efficacy of Endocrine Tests: Importance of Decision-focused Testing Strategies and Appropriate Patient Preparation
George G. Klee
Department of Laboratory Medicine and Pathology, Mayo Clinic and Mayo Foundation, 200 First Street S.W., Rochester, MN 55905. Fax 507-284-4542; e-mail klee.george{at}mayo.edu
 |
Abstract
|
|---|
The efficacy of endocrine tests depends on the choice of tests, the
preparation of the patients, the integrity of the specimens, the
quality of the measurements, and the validity of the reference data.
Close dialogue among the clinicians, the laboratory, and the patients
is a key factor for optimal patient care. The characteristics of urine
and plasma samples and the advantages and limitations of paired test
measurements are presented. The importance of test sequence strategies,
provocative or inhibitory procedures, and elimination of drug
interferences is illustrated with four cases involving Cushing
syndrome, pheochromocytoma, primary aldosteronism, and hypercalcemia.
For each of these scenarios, key clinical issues are highlighted, along
with discussions of the best test strategies, including which
medications are likely to interfere. The importance of targeting
laboratory tests to answer well-focused clinical decisions is
emphasized. The roles of some time-honored provocative procedures are
questioned in light of more sensitive and specific analytic methods.
The importance of decision-focused analytical tolerance limits is
emphasized by demonstrating the impact of analytic bias on downstream
medical resource utilization. User-friendly support systems to
facilitate the implementation of test strategies and postanalytic
tracking of patient outcomes are presented as essential requirements
for quality medical practice.© 1999 American Association for
Clinical Chemistry
|
Introduction
|
|---|
Consider the hypothetical situation in which laboratory tests are
limited (precious resources that are rationed) such that a healthcare
provider could use only a controlled number of resource units per
definable patient care outcome event. In this perhaps not so
hypothetical environment, a clinician would want to ensure that the
most appropriate tests are ordered to provide the best chance of
answering the key diagnostic and therapeutic questions and that the
patient is properly prepared for optimal testing. It would be best to
have these tests performed by reliable methods in a laboratory that has
well-defined reference information needed for interpreting the test
results. It also would be important to understand the specific patient
care outcome events that the healthcare system is tracking and to use
this information for continuous quality improvement.
This report describes two general concepts that are important in
choosing optimal laboratory test strategies: (a) the
advantages and limitations of urine and blood measurements and
(b) the utility of paired trophic and target hormone
measurements. Four endocrine disorders that involve specialized
laboratory tests are presented to illustrate potential test strategies.
For each of these, a series of questions is posed for investigation
before ordering laboratory tests:
- (a) What are the key clinical issues?
- (b) What sequence of tests would be optimal?
- (c) What drugs should be discontinued before testing?
- (d) What stabilizing, provocative, or inhibitory
procedures should the patient undergo before the specimen collection?
|
Choice of Specimen
|
|---|
Twenty-four-hour urine specimens are used for many endocrine
tests. Urine specimens represent a time average that integrates over
the multiple pulsatile spikes of hormone secretion occurring throughout
the day. The 24-h urine specimen also has the advantages of better
analytic sensitivity for some hormones (1)(2).
Urine often contains not only the original hormone, but also key
metabolites that may or may not have biologic activity. For some
hormones, mainly the free (unbound) component is secreted in the urine
(e.g., urinary free cortisol). The drawbacks of urine specimens are the
inconvenience and delays of collecting the 24-h specimen. (For some
analytes, the first-morning or second-morning urine may be a more
convenient alternative.) Another limitation of urine specimens is the
uncertainty of the collection completeness. Measurement of urinary
creatinine concentrations helps in monitoring collection completeness,
especially when it is compared with the muscle mass of the patient.
Many urinary hormones are conjugated to carrier proteins before
excretion. Therefore, both hepatic function and, to a lesser degree,
renal function may alter urinary hormone values.
Blood specimens have both the advantage and the limitation of time
dependency. Most hormones have substantial biologic variation,
including ultradian, diurnal, menstrual, and seasonal variations
(3)(4)(5). Many hormones have short half-lives and are rapidly
cleared from the blood. Half-lives are particularly important when the
response to a provocative drug, such as the effect of
gonadotropin-releasing hormone, is being measured (6).
Simultaneous measurement of trophic and target hormones may help to
determine the location of the abnormality. Hyperfunction and
hypofunction generally are defined in terms of the target hormone
concentrations. Primary hyperfunction means the target gland is
autonomously overproducing the stimulatory hormone, whereas secondary
hyperfunction means that the trophic gland is overproducing. Similarly,
primary hypofunction implies target gland failure, whereas secondary
hypofunction implies trophic gland failure. Fig. 1
illustrates how simultaneous target hormone and trophic hormone
measurements can be used to classify endocrine disorders.
View larger version (39K):
[in this window]
[in a new window]
|
Figure 1. Interrelationships of target and trophic hormone
concentrations for defining hyperfunction vs hypofunction and primary
vs secondary disease or hormone resistance.
|
|
An example of paired hormone measurements is serum thyrotropin from the
pituitary as the trophic hormone with thyroxine and triiodothyronine
from the thyroid gland as target hormones. An increased concentration
of serum thyrotropin with low thyroxine and triiodothyronine is
characteristic of primary hypothyroidism. Another example is
corticotropin
(ACTH)1
as the trophic hormone paired with cortisol as the target
hormone for Cushing disease (hyperfunction) and Addison disease
(hypofunction). A more complex example is growth hormone as the trophic
stimulus for the liver production of insulin-like growth factor I,
which can cause excess growth (acromegaly or giantism) or the deficient
growth of dwarfism.
Patients with hormone resistance may present as potentially confusing
inconsistencies between target hormone and trophic hormone
concentrations. For example, thyroid hormone resistance may present
with increased thyrotropin, with normal or increased thyroid hormone
concentrations. On the other hand, pseudohypoparathyroidism may present
with increased parathyroid hormone (PTH) concentrations and
hypocalcemia. These hormone resistance syndromes or "pseudo"
disease states require careful correlation of the clinical presentation
of the patient (and family members) with the laboratory test results.
The four endocrine disorders selected for these test strategy paradigms
are Cushing syndrome, pheochromocytoma, primary aldosteronism, and
hypercalcemia.
|
Cushing Syndrome
|
|---|
Cushing syndrome is relatively rare, with only ~10 cases per 1
million people (7). Therefore, it is important to have a
sensitive and specific screening test. Measurement of urinary free
cortisol is the recommended screening test (8). However,
measurement of urinary free cortisol will not work well as a screening
test in a general population with a low prevalence of disease;
therefore, it is important to test only patients who have specific
clinical signs and symptoms, such as central obesity, facial plethora,
proximal muscle weakness, wide purple striae, or unexplained
osteoporosis, rather than testing all patients or all patients with
nonspecific conditions such as obesity or hypertension. Before testing,
clinicians and laboratorians should ascertain that the patient is not
using exogenous glucocorticoids, including topical preparations such as
hemorrhoid medications, which could contaminate urine collections. Some
HPLC methodologies can separate exogenous steroids, but most
immunoassay methods cross-react with synthetic steroids. This
cross-reactivity can cause substantial interference. Chronic alcohol
abuse also causes hypercortisolism, which can mimic Cushing syndrome
(9). Clinicians should try to ensure that alcoholic patients
refrain from drinking at least 1 month before testing.
Some medical centers recommend the 1-mg overnight dexamethasone
suppression test followed by morning plasma cortisol measurement as the
preferred screening test for Cushing syndrome. However, this generally
does not perform as well as urinary free cortisol
(10)(11). False-positive results of the
dexamethasone suppression test can be caused by obesity, stress,
psychiatric illness, and increases in cortisol-binding globulin.
False-negative results of the dexamethasone suppression test can be
caused by chronic renal failure and liver failure. There can be
artifacts caused by poor absorption of the dexamethasone or altered
drug metabolism. The 1-mg overnight dexamethasone test is most useful
in patients with ambiguous urinary free cortisol values and/or patients
without the complicating factors previously listed. The historic 24-h
17-hydroxycorticosteroid test is not recommended as a screening test
for Cushing syndrome because of low diagnostic accuracy
(12).
Once a diagnosis of endogenous hypercortisolism has been made
(generally with two urinary free cortisol measurements), the next step
is to localize the abnormality (13). The paired measurement
of ACTH and cortisol determines whether the disease is ACTH-independent
or ACTH-dependent. ACTH-independent disease generally implies an
adrenal source, which typically can be documented with magnetic
resonance imaging. ACTH-dependent and borderline cases require further
testing, usually involving ACTH-releasing hormone stimulation and/or
dexamethasone suppression with either low- or high-dose protocols
(14)(15). The low-dose dexamethasone test
consists of baseline measurements followed by administration of nine
doses of 0.5 mg of dexamethasone every 6 h and measurements of
urinary free cortisol and plasma cortisol during the last 24 h.
Failure of cortisol suppression indicates Cushing syndrome. The
high-dose dexamethasone test consists of nine doses of 2 mg or 4 mg of
dexamethasone every 6 h with follow-up cortisol measurements. Most
patients with pituitary-dependent Cushing syndrome will have suppressed
cortisol concentrations, whereas patients with adrenal neoplasms or
ectopic ACTH syndrome usually have minimal suppression.
If the source of ACTH cannot be localized conclusively, bilateral
inferior petrosal sinus sampling may be necessary. This is a complex
procedure involving collection of blood from catheters placed into the
left and right parts of the inferior petrosal venous system draining
the pituitary gland (16)(17). Comparison of ACTH
concentrations in these specimens with the ACTH concentration in a
concurrently collected peripheral blood sample allows the calculation
of flow gradients and better determination of the source of the ACTH.
Measurements of ACTH before and after stimulation with ACTH-releasing
hormone enhance the diagnostic accuracy. A ratio >2.0 for the baseline
inferior petrosal sinus to the peripheral ACTH concentration or a ratio
>3.0 after administration of ACTH-releasing hormone is consistent with
pituitary-dependent Cushing syndrome.
|
Pheochromocytoma
|
|---|
Catecholamine-secreting tumors are another rare disorder, with
~2–8 cases per 1 million people per year (18). The term
"pheochromocytoma" technically refers to catecholamine-producing
tumors that arise from the chromaffin cells of the adrenal medulla, but
its use here also includes catecholamine-secreting paragangliomas.
The approach to suspected pheochromocytoma depends on the level of
clinical suspicion (19). Optimal test reliability is
achieved by eliminating interfering medications and timing the specimen
collection to be concurrent with the clinical spells. These
hypertension-related spells often are associated with headache,
palpitations, diaphoresis, pallor, nausea, anxiety, tremor, and
epigastric or chest pain (20). A spell usually lasts 10–60
min and may occur daily or only a few times a year.
Biochemical testing should precede imaging studies. The 24-h urinary
excretion rates of catecholamines and their metabolites are the tests
of choice (21). If the clinical suspicion is high, screening
with a combination of measurements of metanephrines, catecholamines,
and vanillylmandelic acid is recommended to expedite the workup. If the
clinical suspicion is lower or if time is not an issue, only
measurement of metanephrines is recommended for the initial screen, and
the catecholamine and vanillylmandelic acid tests are performed if the
metanephrines are increased. In either case, measurement of urinary
creatinine is recommended to ensure completeness of collection.
Patients with pheochromocytomas generally have a twofold increase in
epinephrine or norepinephrine excretion rates or an increased
metanephrine concentration (22). With current test
methodologies, this test strategy is as sensitive as the previously
used histamine and glucagon stimulation tests, rendering these
stimulation tests obsolete. In the past 20 years at the Mayo Clinic,
none of the patients with negative test results for 24-h urinary
catecholamine or catecholamine metabolites have had positive results on
these provocative tests (21).
Patients with suspected pheochromocytoma often are receiving numerous
medications, particularly antihypertensive agents. Labetalol, an
antihypertensive medication, interferes with many metanephrine and
catecholamine assays; therefore, it should be discontinued 4–7 days
before testing (see Table 1
). Tricyclic antidepressants also interfere with many of these
assays and should be tapered and discontinued ~2 weeks before
testing. Dopamine-related drugs also can interfere with these
measurements.
Measurement of urinary catecholamines may not be valid in patients with
advanced renal disease (23)(24). For such
patients plasma catecholamine concentrations may be helpful; however,
higher decision levels, such as a threefold increase for norepinephrine
and a twofold increase for dopamine, should be used to help identify
catecholamine-secreting tumors in hemodialyzed patients.
Special care should be taken when collecting plasma specimens for
catecholamine testing. Many patients autonomously activate their
fight or flight hormones when blood is collected. The protocol
outlined in Fig. 2
is designed to minimize interference in this collection
process.
View larger version (19K):
[in this window]
[in a new window]
|
Figure 2. Flow diagram for patient preparation and collection
process for plasma catecholamine specimens.
|
|
|
Primary Aldosteronism
|
|---|
The syndrome of hypertension, hypokalemia, suppressed plasma renin
activity, and increased aldosterone excretion is consistent with
primary aldosteronism. Although initially considered relatively rare,
more cases (0.05–2% of the population) are being diagnosed when the
symptoms are evaluated biochemically (25). The diagnosis of
this disorder is important because it is a curable form of
hypertension. Multiple subtypes of this disorder have been identified,
some of which are linked to specific genes (26). Some of
these subtypes may have normal potassium concentrations; therefore,
normokalemia does not exclude the diagnosis of primary aldosteronism
(27)(28).
The best screening tests are the paired measurements of plasma renin
activity (PRA) and plasma aldosterone concentration (PAC)
(19). When PAC is measured in units of nanograms per
deciliter and PRA is measured in units of nanograms per milliliter per
hour, a high ratio of PAC to PRA (>20) is a positive screening test
result. The PAC typically is >200 ng/L (>20 ng/dL), whereas the PRA
typically is low. If PAC and PRA concentrations are increased and the
ratio is 
10, the patient should be investigated for secondary causes
of hyperaldosteronism. On the other hand, if both PRA and PAC are
depressed, other adrenal and metabolic disorders should be considered
(Fig. 3
).
View larger version (30K):
[in this window]
[in a new window]
|
Figure 3. Use of the PAC-to-PRA ratio to subclassify the different
causes of hypertension and hypokalemia.
DOC, 11-ß-OHSD, 11ß-hydroxysteroid
dehydrogenase. Reprinted by permission of the publisher from: Young WF
Jr. Endocrinol Metab Clin N Am 1997;26:801–27.
|
|
Patients with a positive result of a screening test for primary
aldosteronism should undergo confirmation testing. The confirmatory
test involves measurement of PAC after 3 days of salt loading with
potassium supplementation. Spironolactone and angiotensin-converting
enzyme inhibitor medications should be replaced with other drugs before
testing. On the third day, a 24-h urine specimen is collected and used
to measure aldosterone, sodium, and potassium. Excretion of >200 mEq
of sodium/24 h assures adequate sodium load. Urinary aldosterone
excretions >12 mg/24 h are consistent with hyperaldosteronism.
|
Hypercalcemia
|
|---|
The differential diagnosis of hypercalcemia depends on the
clinical setting (29)(30). Overall, primary
hyperparathyroidism and malignancy account for 80–90% of
hypercalcemia cases; however, primary hyperparathyroidism is the cause
of ~60% of the ambulatory cases and of ~25% of the hospitalized
cases, whereas malignancy causes ~35% of the ambulatory cases and
65% of the hospitalized cases.
A 1990 NIH Consensus Development Conference Panel for the Diagnosis and
Management of Asymptomatic Primary Hyperthyroidism recommended two
calcium determinations followed by immunoassay of intact PTH as the
most efficient way to evaluate these patients (31). They
specifically emphasized the importance of proper specimen collection
and withdrawal of interfering medications.
The hypercalcemia should be confirmed under conditions of minimal
venous occlusion after withdrawal of potentially causal drugs such as
thiazide diuretics. A 24-h urinary calcium measurement and excretory
urograms often are helpful for characterizing patients with
PTH-mediated hypercalcemia. Measurement of the ionized calcium
concentration may be useful in patients with altered serum albumin
concentrations. Most patients with hyperparathyroidism have borderline
increased or high normal concentrations of intact PTH. Even high normal
PTH values should be considered abnormal because patients with
increased calcium values should normally have suppressed PTH
concentrations. However, high normal PTH and hypercalcemia may be
present in patients with familial hypocalciuric hypercalcemia
(32). A low 24-h urinary calcium concentration (<100 mg/24
h) suggests this disorder, but a normal concentration does not rule out
familial hypocalciuric hypercalcemia. Careful family history and
lifelong increased calcium values in the patient are valuable clues for
identifying familial hypocalciuric hypercalcemia.
Recently, interoperative measurements of PTH have been advocated for
assuring effectiveness of parathyroid surgery
(33)(34). Because PTH has a short half-life of
~3 min, there is a rapid decrease in PTH after resection of an
abnormal parathyroid gland. A drop of PTH concentrations by 50% after
resection is interpreted as removal of the affected gland. In practice,
this procedure generally is not needed by experienced surgeons for
first time operations when anatomical landmarks are intact, but the
procedure can be quite helpful for reoperations and unusual cases.
|
Support Systems for Laboratory Test Requests
|
|---|
Medical information systems should help promote quality
healthcare. A basic premise of medicine is that the profession wants to
provide quality service. Therefore, if there is agreement on a
reasonable testing approach and if there is an efficient system
available for implementing this approach, most medical professionals
are likely to use this system. On the other hand, if there is not
agreement on the approach or if the systems for implementing this
approach are awkward and inefficient (or both), then it will be
difficult to change medical practice.
Therefore, it is prudent for laboratories to engage in dynamic
interchange with clinicians and other healthcare professionals to help
develop best practice paradigms for the use of laboratory tests
(35). It also is important for laboratorians to be involved
in the design and implementation of support systems to help improve
medical practice. For example, electronic order entry systems could be
structured to provide structured test requests potentially following
the strategies outlined here (after incorporating modifications to
adjust to the preferences of the local practice). Test orders could be
packaged according to the signs and symptoms of patient presentations,
such as those that are characteristic of suspected Cushing syndrome or
pheochromocytoma. When available, electronic medical records could be
searched for appropriate clinical signs and symptoms or medications
that might interfere. When this information is not available, queries
could be made (in real time) to the healthcare provider for key data.
When appropriate signs and symptoms do not match the defined criteria
and when interfering medications are identified, real-time advisories
could be provided.
The structured laboratory request forms could include automatic
laboratory test cascades in which follow-up tests are performed when
the initial tests and clinical information meet certain prescribed
criteria. For example, follow-up free thyroxine, triiodothyronine, and
anti-thyroperoxidase antibody measurements could be triggered by
abnormal thyrotropin tests results (36). These structured
orders also should assure that key paired hormone measurements are
performed concurrently when that information is needed to interpret
test results efficiently.
|
Quality Performance Specifications
|
|---|
Effective healthcare strategies are critically dependent on
reliable laboratory measurements. The laboratory test strategies
outlined above could be viewed as being similar to public health
policies that are optimized for the care of a large population of cases
defined by certain common characteristics (such as recommendations for
pneumonia vaccine for patients 
65 years of age) (37).
Specific decision limits for these policies are chosen to optimize
sensitivity and specificity. Similarly, specific decision limits should
be defined for the laboratory test strategies that will optimize the
aggregate performance.
The current heterogeneity of laboratory test methodologies, combined
with the changes in test values over time as a result of calibration
and reagent lot differences, makes the optimization and generalization
of test strategies difficult. Consider the hypothetical laboratory test
depicted in Fig. 4
, with a gaussian distribution of test values. If this test is
used as a frontline test for identifying patients at risk for a
specific disease and the decision threshold for follow-up testing is
set at +2 SD, then 2.3% of the patients tested would be positive on
screening. On the other hand, if the laboratory test measurements were
"shifted" upward 1 SD, then 15.8% would be positive on screening.
In terms of healthcare policy, this represents an almost 700% increase
in the number of patients subjected to additional medical
investigation. This potential increase in expense should more than
offset the cost of providing laboratory testing with tighter analytical
performance specifications.
View larger version (20K):
[in this window]
[in a new window]
|
Figure 4. Illustration of the marked effect that small shifts in
analytic bias can cause in the number of patients having results
exceeding a decision threshold.
The solid curve shows 2.5% having values above 2 SD.
The dotted curve represents a 1-SD upward shift, which
has 15.8% of values above 2 SD. Reprinted by permission of the
publisher from: Klee GG. Tolerance limits for short-term analytical
bias and analytical imprecision derived from clinical assay
specificity. Clin Chem 1993;39:1514–8.
|
|
This hypothetical gaussian model is not very different from many
laboratory test systems. Consider serum calcium measurements. When used
in a general case-finding mode in an ambulatory patient population,
~19 cases per 1000 would have an increased calcium concentration at
102 mg/L (10.2 mg/dL). If the calcium measurement system shifted up
1 mg/L (0.1 mg/dL; perhaps because of recalibration or reagent lot
changes), then ~26 patients would have increased values. With an
upward shift of 2 mg/L (0.2 mg/dL), 36 patients of 1000 would be
positive on screening. With an upward shift of 3 mg/L (0.3 mg/dL), 49
of 1000 patients would be positive on screening. This last result
represents a 160% increase in the numbers of patients undergoing
further evaluation.
The performance limits currently maintained by most laboratories are
much wider than the limits required to optimally control the downstream
effects of laboratory-based test strategies. For example, the CLIA '88
performance limits are ± 10 mg/L (± 1 mg/dL) for calcium, ±
25% for cortisol measurements, and ± 20% for thyroxine
measurements. The limits set by most equipment and reagent systems also
are quite wide for endocrine tests. If tests are allowed to vary within
these performance limits, the number of patient values crossing key
decision thresholds could show extremely wide fluctuations. For
example, almost all calcium measurements would be abnormal with a 10
mg/L (1.0 mg/dL) upward shift. Tolerance limits need to be much tighter
to ensure more uniform practice over time
(38)(39).
A second laboratory performance issue for many endocrine tests is the
lack of uniformity across manufacturers. Currently, there is little
incentive for manufacturers to provide uniform test values. Actually,
there may be an incentive for manufacturers to provide unique values
because that makes it more difficult for laboratories to change test
methods. This variation across methods makes the implementation of
laboratory-based guidelines difficult because each guideline must have
method-dependent decision limits. This heterogeneity of test values
also makes it difficult for clinicians to work in an integrated health
system using multiple test methods. A major incentive for reagent
manufacturers to provide uniform test values would be a requirement for
absolute vs peer group grading criteria for proficiency tests. However,
a difficult issue with that concept is the establishment of the
"true" target values for the proficiency test specimens
(40).
|
Interpretive Reporting
|
|---|
An important part of the implementation of decision-focused test
strategies is the interpretation of the test results. Most endocrine
test strategies require integration of multiple laboratory and clinical
data elements. It would be logical to include these multiple data
elements in an integrated test report along with reference data for the
appropriate decision categories. Unfortunately, most laboratory
information systems are single-test oriented. In addition, most systems
do not provide graphics; therefore, it is difficult to aggregate test
information from multiple reference populations, especially test data
involving more than one variable.
A plot of reference data for concurrent measurements of calcium and PTH
concentrations is shown in Fig. 5
(41). Display of this form of reference data should
help the clinician interpret these test variables better than using
single variable reference limits. Fig. 5
helps to illustrate the
concept of inappropriately increased PTH values in patients with
hypercalcemia, although these PTH values would not be flagged as
abnormal by computer systems that only consider tests univariately.
View larger version (33K):
[in this window]
[in a new window]
|
Figure 5. Plot of serum calcium vs serum PTH that could be used to
help clinicians interpret test results.
Reprinted by permission of the Mayo Foundation for Medical Education
and Research from: Kao PC, van Heerden JA, Grant CS, Klee GG, Khosla S.
Mayo Clin Proc 1992;67:637–45.
|
|
Programs have been developed for automated computer diagnosis based on
laboratory data. Most of these programs are not used in patient care
because the reporting objective generally is not to provide a
diagnosis, which should involve the full expertise of the healthcare
provider, but rather to display information that can be integrated
efficiently with other patient information available to the clinician
to help them make optimal decisions. Most clinicians are skeptical of
diagnostic systems that do not openly display the diagnostic criteria
because they cannot be assured that their patients match the
assumptions used by these systems. These issues of open information
display may become more problematic as laboratory tests become more
complex and proprietary support systems are developed.
|
Outcomes Monitoring
|
|---|
Many endocrine test systems are well suited for outcome monitoring
(42). Test strategies can be tracked in terms of sensitivity
and specificity. Short-term tracking may overestimate sensitivity
because missed cases may not be identified. However, surgically
confirmed cases of endocrine tumors generally could be tracked,
especially in an integrated healthcare system. For example, pituitary,
adrenal, gonadal, and parathyroid tumors could be compared to their key
laboratory tests to track strategy sensitivity. The monitoring of test
specificity generally should be more complete, especially if patients
identified as positive on screening are appropriately followed up with
confirmatory tests. Other factors that could be tracked include the
incidence of test interference and the number of requests for extra
tests.
A major problem with outcome monitoring of laboratory tests is that
many patients do not present with only one-symptom complexes.
Therefore, a series of test strategies may be needed to resolve
patients' problems. The test strategies should be targeted at
presentation syndromes rather than at final diagnoses. It is much
easier to define appropriate test sequences retrospectively once the
diagnosis is known than it is to prospectively evaluate patients with
complex signs and symptoms. Therefore, clinicians must be careful in
deciding which tests are inappropriate or extra. On the other hand,
there are considerable opportunities for improving healthcare
strategies, and probably many of the tests currently performed are not
necessary.
|
Summary
|
|---|
In anticipation of closer scrutiny of laboratory practice, it is
prudent that laboratorians and clinicians work together closely to
ensure that tests are used for optimal care. The key issues for each
disease presentation must be well defined to determine the best
laboratory test strategies. Twenty-four-hour urine specimens (with
complete collection) offer advantages of integrating our pulsatile
secretion spikes and diurnal variations and, therefore, are better than
plasma tests for several endocrine disorders. On the other hand, plasma
samples are valuable for evaluating timed response and for evaluating
patients with impaired renal function. Concurrent, paired measurements
of stimulatory and target analytes are essential for classifying
abnormalities of endocrine systems controlled by negative feedback. The
specificity of modern immunoassays and chromatography systems minimizes
the need for some of the early nonspecific tests or certain provocative
endocrine tests.
Standardized healthcare delivery systems can help minimize undesirable
variations in practice. If consensus can be achieved for selected test
strategies in a given medical practice, laboratorians can help build
delivery systems that should improve the likelihood of best practice
occurring. Mechanisms for these delivery systems may include decision
support information to facilitate test orders and reports that
integrate test values with other clinical information. Decision-focused
test strategies also can help define the analytic performance standards
needed for optimal laboratory support of patient care.
|
Footnotes
|
|---|
1 Nonstandard abbreviations: ACTH, corticotropin; PTH, parathyroid hormone; PRA, plasma renin activity; and PAC, plasma aldosterone concentration. 
|
References
|
|---|
-
Demir A, Alfthan H, Stenman UH, Voutilainen R. A clinically useful method for detecting gonadotropins in children: assessment of luteinizing hormone and follicle-stimulating hormone from urine as an alternative to serum by ultrasensitive time-resolved immunofluorometric assays. Pediatr Res 1994;36:221-226.
[ISI][Medline]
[Order article via Infotrieve]
-
Hourd P, Edwards R. Current methods for the measurement of growth hormone in urine. Clin Endocrinol 1994;40:155-170.
[Medline]
[Order article via Infotrieve]
-
Maes M, Mommen K, Hendrickx D, Peeters D, D'Hondt P, Ranjan R, et al. Components of biological variation, including seasonality, in blood concentrations of TSH, TT3, FT4, PRL, cortisol, and testosterone in healthy volunteers. Clin Endocrinol 1997;46:587-598.
[Medline]
[Order article via Infotrieve]
-
Sebastian-Gambaro MA, Liron-Hernandez FJ, Fuentes-Arderiu X. Intra- and interindividual biological variability bank. J Clin Chem Clin Biochem 1997;35:845-852.
-
Leppaluoto J, Ruskoaho H. Atrial natriuretic peptide, renin activity, aldosterone, urine volume and electrolytes during a 24-hour sleep-wake cycle in man. Acta Physiol Scand 1990;139:47-53.
[ISI][Medline]
[Order article via Infotrieve]
-
Demers LM. Pituitary function. Burtis CA Ashwood ER eds. Tietz textbook of clinical chemistry 3rd ed. 1999:1470-1495 WB Saunders Philadelphia, PA. .
-
Nieman LK, Cutler GB. Cushing's syndrome. DeGroot LJ eds. Endocrinology 1995:1741-1769 WB Saunders Philadelphia, PA. .
-
Meier CA, Biller BMK. Clinical and biochemical evaluation of Cushing's syndrome. Endocrinol Metab Clin N Am 1997;26:741-762.
[ISI][Medline]
[Order article via Infotrieve]
-
Lamberts SW, Klijn JG, de Jong FH, Birkenhager JC. Hormone secretion in alcohol-induced pseudo-Cushing's syndrome: differential diagnosis with Cushing disease. JAMA 1979;242:1640-1643.
[Abstract]
-
Connolly CK, Gore MB, Stanley N, Wills MR. Single-dose dexamethasone suppression in normal subjects and hospital patients. Br Med J 1968;2:665-667.
-
Tsigos C, Chrousos GP. Differential diagnosis and management of Cushing's syndrome. Annu Rev Med 1996;47:443-461.
[ISI][Medline]
[Order article via Infotrieve]
-
Zadik Z, De Lacerda L, De Carmargo LA, Hamilton BB, Migeon CJ, Kowarski AA. A comparative study of urinary 17-hydroxycorticosteroids, urinary free cortisol, and the integrated concentration of plasma cortisol. J Clin Endocrinol Metab 1980;51:1099-1101.
[Abstract]
-
Findling JW, Doppman JL. Biochemical and radiologic diagnosis of Cushing's syndrome. Endocrinol Metab Clin N Am 1994;23:511-537.
[ISI][Medline]
[Order article via Infotrieve]
-
Flack MR, Oldfield EH, Cutler GB, Jr, Zweig MH, Malley JD, Chrousos GP, et al. Urine free cortisol in the high-dose dexamethasone suppression test for the differential diagnosis of the Cushing syndrome. Ann Intern Med 1992;116:211-217.
-
Orth DN. Cushing's syndrome. N Engl J Med 1995;332:791-794.
[Free Full Text]
-
Findling JW, Kehoe ME, Shaker JL, Raff H. Routine inferior petrosal sinus sampling in the differential diagnosis of adrenocorticotropin (ACTH)-dependent Cushing's syndrome: early recognition of the occult ectopic ACTH syndrome. J Clin Endocrinol Metab 1991;73:408-413.
[Abstract]
-
Oldfield EH, Doppman JL, Nieman LK, Chrousos GP, Miller DL, Katz DA, et al. Petrosal sinus sampling with and without corticotropin-releasing hormone for the differential diagnosis of Cushing's syndrome. N Engl J Med 1991;325:897-905.
[Abstract]
-
Stenstrom G, Svardsudd K. Phaechromocytoma in Sweden, 1958–81: an analysis of the National Cancer Registry Data. Acta Med Scand 1986;220(Suppl):225-232.
[ISI][Medline]
[Order article via Infotrieve]
-
Young WF, Jr. Pheochromocytoma and primary aldosteronism: diagnostic approaches. Endocrinol Metab Clin N Am 1997;26:801-827.
[ISI][Medline]
[Order article via Infotrieve]
-
Young WF, Jr. Pheochromocytoma: 1926 to 1993. Trends Endocrinol Metab 1993;4:122-127.
[Medline]
[Order article via Infotrieve]
-
Young WF, Jr. Phaeochromocytoma: how to catch a moonbeam in your hand. Eur J Endocrinol 1997;136:28-29.
[ISI][Medline]
[Order article via Infotrieve]
-
Sheps SG, Jiang N-S, Klee GG. Diagnostic evaluation of pheochromocytoma. Endocrinol Metab Clin North Am 1988;17:397-414.
[ISI][Medline]
[Order article via Infotrieve]
-
Chauveau D, Martinez F, Houhou S, Grunfeld JP. Malignant hypertension secondary to pheochromocytoma in a hemodialyzed patient. Am J Kidney Dis 1993;21:52-53.
[ISI][Medline]
[Order article via Infotrieve]
-
Stumvoll M, Radjaipour M, Seif F. Diagnostic considerations in pheochromocytoma and chronic hemodialysis: case report and review of the literature. Am J Nephrol 1995;15:147-151.
[ISI][Medline]
[Order article via Infotrieve]
-
Streeten DHP, Tomycz N, Anderson GH, Jr. Reliability of screening methods for the diagnosis of primary aldosteronism. Am J Med 1979;67:403-413.
[ISI][Medline]
[Order article via Infotrieve]
-
Young WF, Jr, Klee GG. Primary aldosteronism: diagnostic evaluation. Endocrinol Metab Clin N Am 1988;17:367-395.
[ISI][Medline]
[Order article via Infotrieve]
-
Gordon RD, Stowasser M, Tunny TJ, Klemm SA, Rutherford JC. High incidence of primary aldosteronism in 199 patients referred with hypertension. Clin Exp Pharmacol Physiol 1994;21:315-318.
[ISI][Medline]
[Order article via Infotrieve]
-
Ganguly A. Primary aldosteronism. N Engl J Med 1998;339:1826-1834.
-
Lafferty FW. Differential diagnosis of hypercalcemia. J Bone Miner Res 1991;6:S51-S59.
-
Johnson KR, Mascall GC, Howarth AT. Differential laboratory diagnosis of hypercalcemia. CRC Crit Rev Clin Lab Sci 1984;21:51-97.
-
Consensus Development Conference Panel. Diagnosis and
management of asymptomatic primary hyperparathyroidism: consensus
development conference statement. Ann Intern Med 1991;114:593–7..
-
Marx SJ, Spiegel AM, Levine MA. Familial hypocalciuric hypercalcemia: the relation to primary parathyroid hyperplasia. N Engl J Med 1982;307:416-426.
[ISI][Medline]
[Order article via Infotrieve]
-
Irvin GL, 3rd, Deriso GT, 3rd. A new, practical interoperative parathyroid hormone assay. Am J Surg 1994;168:466-468.
[ISI][Medline]
[Order article via Infotrieve]
-
Kao PC, Van Heerden JA, Taylor RL. Intraoperative monitoring of parathyroid procedures by a 15-minute parathyroid hormone immunochemiluminometric assay. Mayo Clin Proc 1994;69:532-537.
[ISI][Medline]
[Order article via Infotrieve]
-
Lundberg GD. Changing physician behavior in ordering diagnostic tests [Editorial]. JAMA 1998;280:2036.[Free Full Text]
-
Klee GG, Hay ID. Biochemical testing of thyroid function. Endocrinol Metab Clin N Am 1997;26:763-775.
[ISI][Medline]
[Order article via Infotrieve]
-
Anonymous. Pneumococcal polysaccharide vaccine. Morbid
Mortal Wkly Rep 1989;38:64–8,73–6..
-
Klee G. A conceptual model for establishing tolerance limits for analytic bias and imprecision based on variations in population test distributions. Clin Chim Acta 1997;260:175-188.
[ISI][Medline]
[Order article via Infotrieve]
-
Smith FA, Kroft SH. Optimal procedures for detecting analytic bias using patient samples. Am J Clin Pathol 1997;108:254-268.
[ISI][Medline]
[Order article via Infotrieve]
-
Steele BW, Witte DL, Whitley RJ, Klee GG, Chan DW. The effects of modifying proficiency testing materials on thyroid function test results: a College of American Pathologists ligand assay survey study. Arch Pathol Lab Med 1997;121:1241-1246.
[ISI][Medline]
[Order article via Infotrieve]
-
Kao PC, van Heerden JA, Grant CS, Klee GG, Khosla S. Clinical performance of parathyroid hormone immunometric assays. Mayo Clin Proc 1992;67:637-645.
[ISI][Medline]
[Order article via Infotrieve]
-
Wong ET. Improving laboratory testing: can we get physicians to focus on outcome?. Clin Chem 1995;41:1241-1247.
[Abstract/Free Full Text]
The following articles in journals at HighWire Press have cited this article:
|
|
|
|
 
G. Mengozzi, D. Rossato, C. Bertello, C. Garrone, A. Milan, R. Pagni, F. Veglio, and P. Mulatero
Rapid Cortisol Assay during Adrenal Vein Sampling in Patients with Primary Aldosteronism
Clin. Chem.,
November 1, 2007;
53(11):
1968 - 1971.
[Abstract]
[Full Text]
[PDF]
|
|
|
Top
Abstract
Introduction
