The method for measuring the ESR recommended by the International Council for Standardization in Haematology (ICSH)⁴ and also by various national authorities⁵ is based on that of Westergren, who developed the test in 1921 for studying patients with pulmonary tuberculosis. ESR is the measurement of the sedimentation of red cells in diluted blood after standing for 1 h in an open-ended glass tube of 30 cm length mounted vertically on a stand.

Conventional Westergren Method

For the diluent, prepare a solution of 109 mmol/l trisodium citrate (32 g/l Na₃C₆H₅O₇.2H₂O). Filter through a micropore filter (0.22 mm) into a sterile bottle. It can be stored for several months at 4°C but must be discarded if it becomes turbid through the growth of moulds.

Method

The method described below, originally described by the International Council for Standardization in Haematology (ICSH)⁴ and now adopted by the Clinical and Laboratory Standards Institute (CLSI) as its approved method,⁵ is intended to provide a reference method for verifying the reliability of any modification of the test.

Either collect venous blood in ethylenediaminetetra-acetic acid (EDTA) and dilute a sample accurately in the proportion of 1 volume of citrate to 4 volumes of blood, or collect the blood directly into the citrate solution. The test should then be carried out on the diluted sample within 4 h of collecting the blood, although a delay of up to 6 h is permissible provided that the blood is kept at 4°C. EDTA blood can be used within 24 h if the specimen is kept at 4°C, provided that 1 volume of 109 mmol/l (32 g/l) trisodium citrate is added to 4 volumes of blood immediately before the test is performed.

Mix the blood sample thoroughly and then draw it up into the Westergren tube to the 200 mm mark by means of a teat or a mechanical device; mouth suction should never be used. Place the tube exactly vertical and leave undisturbed for exactly 60 min, free from vibrations and draughts and not exposed to direct sunlight. Then read to the nearest 1 mm the height of the clear plasma above the upper limit of the column of sedimenting cells. The result is expressed as ESR = X mm in 1 h. A poor delineation of the upper layer of red cells may sometimes occur, especially when there is a high reticulocyte count.

Range in health

The mean values and the upper limit for 95% of normal adults are given in Table 6.1. There is a progressive increase with age, but it is difficult to define a strictly healthy population for determining normal values in individuals older than 70 years.⁶ In the newborn, the ESR is usually low. In childhood and adolescence, it is the same as for normal men with no differences between boys and girls. It is increased in pregnancy, especially in the later stages, and independent of anaemia;⁷ this is due to the physiological effect of haemodilution (an increase in the plasma volume)

Table 6.1 Erythrocyte sedimentation rate ranges in health

Age range (years)	ESR mean MM IN 1 H
10–19	8
20–29	10.8
30–39	10.4
40–49	13.6
50–59	14.2
60–69	16
70–79	16.5
80–91	15.8
Pregnancy
Early gestation	48 (62 if anaemic)
Later gestation	70 (95 if anaemic)

In the newborn, the ESR has been reported to be 0–2 mm in 1 h, increasing to 4 mm in 1 h at 1 week, up to 17 mm in 1 h by day 14, and then 10–20 mm in 1 h for both girls and boys, until puberty.⁸ However, as the studies in infants were obtained by the capillary method, they are not strictly comparable to the Westergren method.

Modified methods

An elevated ESR occurs as an early feature in myocardial infarction.¹⁵ Although a normal ESR cannot be taken to exclude the presence of organic disease, the vast majority of acute or chronic infections and most neoplastic and degenerative diseases are associated with changes in the plasma proteins that lead to an acceleration of sedimentation. An increased ESR in subjects who are HIV seropositive seems to be an early predictive marker of progression toward acquired immune deficiency syndrome (AIDS).¹⁶ The ESR is less helpful in countries where chronic diseases are rife; however, one study has shown that very high ESRs (higher than 100 mm/h) have a specificity of 0.99 and a positive predictive value of 0.9 for an acute or chronic infection.¹⁷ The ESR is influenced by age, stage of the menstrual cycle and medications taken (corticosteroids, contraceptive pills). It is especially low (0–1 mm) in polycythaemia, hypofibrinogenaemia and congestive cardiac failure and when there are abnormalities of the red cells such as poikilocytosis, spherocytosis, or sickle cells. In cases of performance-enhancing drug intake by athletes (discussed below) the ESR values are generally lower than the usual value for the individual and as a result of the increase in haemoglobin (i.e. the effect of secondary polycythaemia).

Plasma Viscosity

The ESR and plasma viscosity generally increase in parallel with each other.¹ Plasma viscosity is, however, primarily dependent on the concentration of plasma proteins, especially fibrinogen, and it is not affected by anaemia. Changes in the ESR may lag behind changes in plasma viscosity by 24–48 h, and viscosity seems to reflect the clinical severity of disease more closely than does the ESR.¹⁸

There are several types of viscometers, including rotational and capillary types that are suitable for routine use ¹ and, as for ESR methods, automated closed-tube methods are available.¹⁹ The main use of plasma viscosity is in the investigation of individuals with suspected hyperviscosity, myeloma and macroglobulinaemia. In conjunction with the ESR and CRP, the plasma viscosity can be used as a marker for inflammation. The viscosity test should be carried out as described in the instruction manual for the particular instrument used.

Reference Values

Each laboratory should establish its own reference values for plasma viscosity. As a general guide, ICSH has recorded that with the Harkness capillary viscometer normal plasma has a viscosity of 1.16–1.33 mPa/s (if expressed in poise (P), 1 cP = 1 mPa/s) at 37°C and 1.50–1.72 mPa/s at 25°C.²⁰ Plasma viscosity is lower in the newborn (0.98–1.25 mPa/s at 37°C), increasing to adult values by the 3rd year; it is slightly higher in old age. There are no significant differences in plasma viscosity between men and women or in pregnancy. It is remarkably constant in health, with little or no diurnal variation, and it is not affected by exercise. A change of only 0.03–0.05 mPa/s is thus likely to be clinically significant.

Whole blood viscosity

The viscosity of whole blood reflects its rheological properties; it is influenced by PCV, plasma viscosity, red cell aggregation and red cell deformability. It is especially sensitive to PCV, with which it is closely correlated. The clinical interpretation of its measurement must also take into account the interaction of the red cells with blood vessels, which greatly influences blood flow in vivo.

Guidelines for measuring blood viscosity and red cell deformability by standardized methods have been published.²¹ Rotational and capillary viscometers are suitable for measuring blood viscosity; deformability can be measured by recording the rate at which red cells in suspension pass through a filter with pores 3–5 mm in diameter.

Heterophile antibodies in serum: diagnosis of infectious mononucleosis

Infectious mononucleosis (IM) is caused by Epstein–Barr virus.²² The immune response that develops in response to virus-infected cells includes not only antibodies to viral antigens but also characteristic heterophile antibodies. Before the nature of this reaction was understood, Paul and Bunnell²³ demonstrated the antibodies as agglutinins directed against sheep red cells. They are, in fact, not specific for sheep red cells but also react with horse and ox, but not human, red cells. They are IgM globulins, which are immunologically related to, but distinct from, antibodies that occur in response to the Forssman antigens. The latter are widely spread in animal tissue; they occur at low titre in healthy individuals and at high titre in serum sickness and in some leukaemias and lymphomas.^24,²⁵ In these non-IM conditions, the antibody can be absorbed out by guinea pig cells. Thus, for the diagnosis of IM, it is necessary to demonstrate that the antibody present has the characters of the Paul–Bunnell antibody (i.e. it is absorbed by ox red cells but not by guinea pig kidney). This is the basis of the absorption tests for IM (‘monospot’ test). Immunofluorescent antibody tests have been developed to distinguish the IgM antibody, which occurs at high titre in the early phase of IM and diminishes during convalescence, from the IgG antibody, which persists at high titre for years after infection^26,²⁷ and which also occurs in the non-IM infections.^22,²⁸