It is important to understand normal or reference values/intervals and ranges.
Normal ranges can vary to some degree from laboratory to laboratory. Frequently, this is because of the particular type of equipment used. Theoretically, normal can refer to the ideal health state, to average reference values or intervals, or to types of statistic distribution. Normal values are those that fall within two standard deviations of the mean value for the normal population. The reference interval typically represents the upper and lower limits wherein 95% of healthy people would fall. Although establishing normal or reference values is complex, it is much more so in the pediatric population. There are many challenges related to gender- and age-specific reference intervals in the pediatric population.
The reported reference range for a test can vary according to the laboratory used, the method employed, the population tested, and methods of specimen collection and preservation.
Most normal blood test values are determined by measuring fasting specimens.
Specific factors can influence test results. For example, patient posture can significantly alter the values for plasma volume, hemoglobin, hematocrit, and a number of other laboratory values.
Laboratories must specify their own normal ranges. Many factors affect laboratory test values and influence ranges. Thus, values may be normal under one set of prevailing conditions but may exhibit different limits in other circumstances. Age, gender, race, environment, posture, diurnal and other cyclic variations, foods, beverages, fasting or postprandial state, drugs, and exercise can affect derived values. Interpretation of laboratory results must always be in the context of the patient’s state of being. Circumstances such as hydration, nutrition, fasting state, mental status, or compliance with test protocols are only a few of the situations that can influence test outcomes.
Scientific publications and many professional organizations are changing clinical laboratory data values from conventional U.S. units (also referred to as U.S. customary units) to Système International (SI) units. Currently, many data are reported in both units.
The SI system uses seven dimensionally independent units of measurement to provide logical and consistent measurements. For example, SI concentrations are written as amount per volume (moles or millimoles per liter) rather than as mass per volume (grams, milligrams, or milliequivalents per deciliter, 100 mL, or liter). Numeric values may differ between systems or may be the same. For example, chloride is the same in both systems: 95–105 mEq/L (conventional) and 95–105 mmol/L (SI).
Clinical laboratory data may be reported in conventional U.S. units, SI units, or both. Examples of conversion of data from the two systems are included in Table 1.4. To convert SI units to conventional U.S. units, divide by the factor; to convert conventional U.S. units to SI units, multiply by the factor.
Example:
To convert a digoxin (drug management) level of 0.6 nmol/L (SI units), divide by the factor 1.281 to obtain conventional units of 0.5 ng/dL.
To convert a Ca2+ (electrolyte) value of 8.6 mg/dL (conventional units), multiply by the factor 0.2495 to obtain the SI units of 2.15 mmol/L.
Recognize margins of error. For example, if a patient has a battery of chemistry tests, the possibility exists that some tests will be abnormal owing purely to chance. This occurs because a significant margin of error arises from the arbitrary setting of limits. Moreover, if a laboratory test is considered normal up to the 95th percentile, then 5 times out of 100, the test will show an abnormality even though a patient is not ill. A second test performed on the same sample will probably yield the following: 0.95 × 0.95, or 90.25%. This means that 9.75 times out of 100, a test will show an abnormality even though the person has no underlying health disorder. Each successive testing will produce a higher percentage of abnormal results. If the patient has a group of tests performed on one blood sample, the possibility that some of the tests will read abnormal due purely to chance is not uncommon.