Take blood to measure serum calcium and serum albumin concentrations

• This recommendation is based on a narrative review, Best practice in primary care pathology [Smellie et al, 2008].

Avoid prolonged application of a tourniquet

• Historically, it was thought that the apparent serum calcium concentration increased if a tourniquet was applied to distend the vein for venepuncture [McMullan et al, 1990].
• Although one regional Canadian guideline recommends that a tourniquet should be avoided when taking blood to measure serum calcium concentration [Alberta Medical Association, 2008], one small study found no important effect of tourniquet use on adjusted calcium concentrations [McMullan et al, 1990]. Two narrative reviews recommend avoiding prolonged application of a tourniquet [Smellie et al, 2008; Waters, 2009] presumably because of the uncertainty regarding this issue.

Ensure that the calcium concentration is adjusted for serum albumin

• The formula used to adjust the serum calcium concentration is based on narrative reviews [Rizzoli and Bonjour, 1992; Murphy et al, 2006; Waters, 2009].
• The reason for adjusting for serum albumin concentration is that a large proportion of circulating calcium is bound to albumin, and low albumin levels can affect the total serum calcium level [Carroll and Schade, 2003].
• The statement that the calculation may be inaccurate at extreme albumin concentrations, if the person has paraproteinaemia (for example in multiple myeloma) or if the person is acidotic, is derived from narrative reviews [Ralston, 1992; Bushinsky and Monk, 1998; Inzucchi, 2004; Smellie et al, 2008].

Repeat to confirm

• The recommendation to repeat the test (provided there are no indications for immediate management) is based on two narrative reviews [Klee et al, 1988; Smellie et al, 2008] and a study of 24,500 people who underwent testing of plasma calcium concentrations at a biochemistry department in Derby at the request of their physician [Harrop et al, 1982]. Three-quarters of samples were from hospital inpatients. Of 738 people whose initial plasma calcium concentration was greater than 2.60 mmol/L, 404 people had repeat tests. Of the people who had repeat tests, only 201 (50%) had a second plasma calcium concentration greater than 2.60 mmol/L. The higher the initial plasma calcium concentration, the more likely was the repeat test to confirm hypercalcaemia. The confirmation rate was 37% for the range 2.61–2.70 mmol/L, 72% for the range 2.71–2.80 mmol/L, 93% for 2.81–2.90 mmol/L, and 98% for greater than 2.90 mmol/L.
• The recommendation to admit people with severe or symptomatic hypercalcaemia is based on a narrative review [Smellie et al, 2008].

When to diagnose mild, moderate, and severe hypercalcaemia

• These recommendations are derived from a narrative review [Smellie et al, 2008]. The concentration of serum calcium affects management.